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/4-574/v1	18 Aug 15	{ "type": "Software Tool Article", "title": "GOsummaries: an R Package for Visual Functional Annotation of Experimental Data", "authors": [ "Raivo Kolde", "Jaak Vilo", "Jaak Vilo" ], "abstract": "Functional characterisation of gene lists using Gene Ontology (GO) enrichment analysis is a common approach in computational biology, since many analysis methods end up with a list of genes as a result. Often there can be hundreds of functional terms that are significantly associated with a single list of genes and proper interpretation of such results can be a challenging endeavour. There are methods to visualise and aid the interpretation of these results, but most of them are limited to the results associated with one list of genes. However, in practice the number of gene lists can be considerably higher and common tools are not effective in such situations.We introduce a novel R package, 'GOsummaries' that visualises the GO enrichment results as concise word clouds that can be combined together if the number of gene lists is larger. By also adding the graphs of corresponding raw experimental data, GOsummaries can create informative summary plots for various analyses such as differential expression or clustering. The case studies show that the GOsummaries plots allow rapid functional characterisation of complex sets of gene lists. The GOsummaries approach is particularly effective for Principal Component Analysis (PCA).By adding functional annotation to the principal components, GOsummaries improves significantly the interpretability of PCA results. The GOsummaries layout for PCA can be effective even in situations where we cannot directly apply the GO analysis. For example, in case of metabolomics or metagenomics data it is possible to show the features with significant associations to the components instead of GO terms. The GOsummaries package is available under GPL-2 licence at Bioconductor (http://www.bioconductor.org/packages/release/bioc/html/GOsummaries.html).", "keywords": [ "Gene Ontology", "word cloud", "Gene Set Enrichment analysis", "visualisation", "Principal Component Analysis", "limma" ], "content": "Introduction\n\nAs technologies mature, the time and cost of performing microarray and next-generation sequencing experiments is greatly reduced. A wide range of biological questions can be addressed using these experimental approaches. However, several steps of the analysis are often conceptually similar across experiments. At some point of analysis, lists of genes are identified from the data that display interesting behaviour. These lists can represent differentially-expressed genes between two tissues, genes with similar methylation patterns, genes that are close to relevant mutations, etc. Next, these genes are being annotated functionally, by searching for functional terms that are associated with more of them than expected by chance. The latter procedure is called Gene Ontology (GO) enrichment analysis1, and there are many web based tools for this, for example, DAVID2, Babelomics3 and g:Profiler4. The result of GO enrichment analysis is a list of GO terms with associated significance scores. There can be hundreds of significant functional terms associated with one gene list.\n\nAnalysis methods often produce many lists of genes instead of only one. For example, clustering analysis can divide genes into tens of clusters, each one of them displaying a distinct biological pattern and potentially unique function. Proper interpretation of the functional analysis results requires that we would also take into account the complex relations between these gene lists. Thus, ideally the underlying experimental data and the functional annotations should be shown together. In practice, the experimental data is usually shown in a single plot while the functional annotations of associated gene lists are given in a series of (long) tables. With this type of representation it is complicated to scan through all the functional terms, while keeping in mind the biological relations between the gene lists and the degree of enrichment of various terms. Therefore, methods that can visually summarise the experimental data and combine it with relevant functional annotations can significantly improve interpretation of analysis results.\n\nFor visualising the numeric experimental data there are numerous options, such as heatmaps, barplots, boxplots, etc. However, visualising the GO enrichment analysis results is more complicated, as there are not many options to represent textual data graphically. Many GO visualisation tools aim to reveal the connections between the terms by overlaying them on GO graphs. For example, g:Profiler uses this structure to group the significant results, GOrilla5 overlays the GO graph with enrichment scores, several tools6–10 visualise the results as a network and REVIGO11 displays significant categories among other options as treemaps and 2D scatterplots. But as the term names would still have to be shown then the resulting plots are physically even larger than the original tables and would not help in comparison of multiple gene lists. To achieve a more compact presentation of results from multiple gene lists, it is possible to display them in a matrix format as a heatmap, where columns represent the lists of genes and rows significant categories.\n\nThis is implemented, for example, in the g:Cocoa tool in g:Profiler4 and PloGO12. Although this approach provides a high-level overview of relations between enrichment results, it still tends to create visualisations that are too large to fit a computer screen or a page in print. A promising idea is to represent the enrichment results as word clouds, where the strength of enrichment can be expressed using font size. This is implemented in several tools11,13–16, but in all of these cases the emphasis is on the single gene list analysis. One cannot easily combine the results of multiple gene lists or attach the word clouds to the plots of experimental data.\n\nHere we extend the idea of using word clouds to represent GO enrichment results. We implement custom methods to filter GO enrichment results and display them as word clouds. In addition, we define a specific layout to display multiple word clouds, together with the associated experimental data. This allows the creation of concise visual summaries for analyses such as differential expression, clustering or principal component analysis. All the methods are implemented as an R package GOsummaries.\n\n\nMethods\n\nExamples of the plots produced by GOsummaries can be seen in Figure 2–Figure 4. Although the plots correspond to different data types and analysis methods, the layout stays the same. The plot consists of blocks corresponding to either one or two closely related gene lists, such as a cluster from clustering analysis or up- and down-regulated genes from a differential expression analysis. Each block consists of one or two word clouds representing the GO enrichment results and optionally a panel showing the experimental data related to the lists. The blocks are stacked on top of each other to display multiple gene lists. Depending on configuration one can fit 5–6 blocks on one printed page, however, for exploratory analysis one can easily generate plots with tens of blocks (see Supplementary Figure S2 and Supplementary Figure S3). In this way it is easy to quickly go through and efficiently compare many functional annotations in parallel. As such it does not need to contain all detailed information, but rather aids higher-level understanding. For more detailed analysis users can always refer back to full results from tools like g:Profiler or others.\n\nThe content of the panel on top of word cloud(s) is customisable and can display any information that can fit to such space. For example, in case of differential expression and clustering, the panel displays expression of the genes in underlying gene list(s) as boxplots. The y-axis shows the expression level and each boxplot corresponds to one sample. If expression data is not available, then the panel just shows the number of genes.\n\nThe word clouds are designed to show the results of a GO enrichment analysis. By default, the GO enrichment analysis is performed by the GOsummaries package itself using the g:Profiler4 web service. However, it is possible to use other type of information, for example, results from other GO enrichment tools or names of the significant genes. The font of the term is sized according to the associated p-values. More specifically, the size of the terms in each cloud is proportional to the -log10 of the enrichment p-value. As the word placement algorithm tries to use the available space effectively, the term sizes are not comparable between the word clouds. The global strength of enrichment of the terms is color-coded in grayscale.\n\nTo make the identification of the lists and their characterisation easier, the content of the gene lists is reflected both in the block title as well as small text next to the panel. For example, in case of differential expression visualisation, the title identifies the groups that were compared and the number of genes that was found is given next to the panel.\n\nA typical GO-based characterisation of a gene list can contain hundreds of statistically significant GO terms, thus, it is not reasonable to display all of them in one word cloud. As GO defines a hierarchy of biological processes that range from very specific to more general, the GO enrichment analysis results usually contain a number of closely related GO terms. In addition, terms with many associated genes tend to be too general and terms with a small number of genes too specific to give useful information about a larger list of genes. Therefore, it is possible to filter out many terms without losing much information.\n\nThere are algorithms such as RedundancyMiner17 that allow to filter the GO enrichment results for redundant terms. However, as g:Profiler, which is used for the functional analysis, has rather good tools for filtering the results, we use those as default.\n\nGOsummaries filters the GO terms both based on their size and structural relations, a graphical example can be seen on Supplementary Figure S1. First, it applies the lower and upper limit on the number of genes in the GO terms. By default, it considers GO terms with more than 50 and less than 1000 genes. For removing redundancies GOsummaries uses the hierarchical filtering option of g:Profiler. This divides the results into groups where the terms share parents and takes the one with smallest p-value from every such group. Also by default GOsummaries considers only results from the Biological Process branch of GO and KEGG and Reactome pathway databases. If the number of significant terms is still too high after such filtering, then GOsummaries selects by default 30 terms with the most significant enrichment.\n\nApplying these steps effectively reduces the number of terms while retaining relevant information. The default parameters have proven to be practical for lists of few hundred genes, but all these parameters can be easily changed within the user interface. For example, if one has smaller gene lists, then more specific GO terms can give more appropriate information. Some relevant terms might be lost during the filtering process, but for more specific analysis users can always go back to original results.\n\nInstead of performing the GO enrichment analysis with g:Profiler as described above, a user can supply their own annotations for visualisation as a word cloud. For example, it is also possible for GOsummaries to display results of Gene Set Enrichment Analysis (GSEA)18 or DAVID, or use RedundancyMiner to apply an alternative redundancy reduction step.\n\nThe GOsummaries layout can be useful even in cases where we do not use the GO enrichment results. For example, it is natural to show the gene names instead of the GO terms in the word clouds. This option can be useful, for example, for visualising metabolomics or metagenomics data (see Figure 4). It is implemented in several GOsummaries subroutines. For convenience, it can automatically convert various gene identifiers into gene names, using g:Convert web service4.\n\nThere are several common analysis methods of high-throughput data that create sets of gene lists as a result. For several of such methods we have created specialised routines that extract the gene lists and relevant expression data from the input, run the GO enrichment analysis and display the results. For example, we have created functions that can parse the results from the k-means function for clustering and limma package19 for differential expression.\n\nIn both of these cases, the interpretation of the resulting plots is straightforward. The word clouds represent the clusters or significant genes and panels display the expression patterns of these genes.\n\nInterestingly, we can apply the GOsummaries approach to Principal Component Analysis (PCA). Usually the results of PCA are depicted as a scatterplot of samples along the first few principal components (PC). The distances between the samples on this low dimensional plot approximate the distance in the actual dataset. Therefore, these plots can reveal outliers and general similarity structure of the samples but very little else.\n\nActually, PCA reveals much more information than shown on a scatterplot. Each principal component is a weighted sum of original features, such as genes. Thus, the weights, also called loadings, directly show how much influence each feature has to a principal component. In other fields, like psychology, the loadings are routinely used to give an interpretation to the components. However, in bioinformatics this information is often neglected.\n\nIn GOsummaries we utilise the information in loadings as follows. First, we take 500 genes with largest positive and negative loadings and run GO enrichment analysis on them. Then we display the results within the GOsummaries layout, where each block represents one principal component. The distribution of samples along the principal component is shown as a stacked histogram, with colour indicating different classes of samples. An example of such visualisation can be seen in Figure 2.\n\nThis type of display can be considered complementary to the typical 2D scatterplot representation. If a scatterplot gives an overview of the similarity between the samples, then GOsummaries representation associates a functional interpretation with each of the components. Thus, instead of just observing that a principal component discriminates between two sets of samples, we can also identify the biological processes that underlie this separation. As another advantage, one can display even tens of components in one figure, making it easier to get a comprehensive overview about the PCA results.\n\nFor some data types PCA does not work, since the data does not follow its assumptions. Then it is possible to use some other multidimensional scaling (MDS) methods, like principal coordinate analysis. This approach is used, for example, with metagenomics data for visualising similarities in taxon abundances.\n\nIn general, the result of a MDS analysis is a matrix with lower dimensionality. As the transformation does not have to be a linear transformation of features, we do not always obtain loadings for the features that could be used for interpreting the new components. Still, we can find correlations between the features and the scaled down components, and perform a statistical test to measure the significance of the correlation. GOsummaries can be applied to the significantly correlated features, much like we use it in case of PCA. It is possible to display either the GO analysis results or the names of significantly correlated features as word clouds.\n\nAll the methods are implemented as an add-on package for R statistical computation environment. The GO enrichment analysis is performed through R with gProfileR package that interacts with g:Profiler web toolkit. The figures of experimental data are drawn using ggplot220 package.\n\nR was chosen as a platform, thanks to its popularity for genomic analyses. Many of the key statistical algorithms producing the gene lists are specifically implemented in R and, thus, it is natural to integrate the subsequent analyses with it. Unfortunately, this choice constrains the output to static plots, as R does not handle interactivity equally well.\n\nThe GOsummaries methodology itself is not restricted to R. For example, we are planning to implement the same approach as a web based tool as well that could take advantage of interactive capabilities of modern Javascript libraries.\n\nThe package can run on any platform with a relatively recent R installation. When starting from gene lists, k-means clustering or PCA results, the analysis is performed using two steps. First, the GO analysis and filtration is carried out. Then the plot is drawn based on the resulting object. Both steps are automated and usually the analysis can be performed using only two commands. At the same time, all the critical parameters can be customised.\n\nFor comparing the word clouds we used a list of 622 mouse genes. In REVIGO and Cytoscape we used the enrichment results given by g:Profiler. This was the same functional data that was used by GOsummaries. GeneCodis3 and Genes2WordCloud performed the enrichment analysis on their own.\n\nThe embryonic stem cell dataset used for clustering was downloaded from ArrayExpress (accession E-TABM-672). We used the processed data matrix and did not apply any additional preprocessing steps. The clustering was performed on 2012 probesets that had standard deviation larger than 1.0.\n\nThe gene expression compendium was downloaded from ArrayExpress (E-MTAB-62) as raw data. It was normalised with Robust Multiarray Analysis method21 using default settings.\n\nThe example microbiome dataset was provided as an example for the metagenomic biomarker discovery tool LefSe22 and was downloaded from http://huttenhower.sph.harvard.edu/webfm_send/129.\n\n\nResults\n\nThe idea to show GO enrichment results as word clouds is not new and several tools, like REVIGO, Genes2WordCloud, GeneCodis3 and Cytoscape WordCloud already implement it.\n\nHowever, the usefulness of such an approach depends heavily on the methods used for constructing the word cloud. Most of the published methods follow more or less the approach taken by original word cloud implementation in http://www.wordle.net/, where words are counted and their size reflects their count within the results. However, count of a word within the GO enrichment results is not a good measure of its association with a gene list. With GOsummaries we took a more direct approach, since strength of association is already defined by the enrichment p-value, we just show the full category name scaled according to the p-value.\n\nTo compare the word clouds produced by different tools, we tried Genes2WordCloud, REVIGO, Cytoscape WordCloud and GOSummaries on a cluster from our embryonic stem cell time series (Figure 1). The cluster represents genes that are turned on on days 3 and 4 during embryonic development. The GOsummaries word cloud nicely highlights terms that are related to the biological pattern, like “embryo development”, “organ morphogenesis”, “cardiovascular system development”, etc. The results of other word clouds, however, are much poorer. Most of the highlighted words and phrases have nothing to do with the specific expression pattern. GeneCodis3 word cloud emphasises the need for redundancy filtering as most of the largest terms correspond to the same biological process. The word cloud of GOsummaries is also more compact thanks to our custom word placement algorithm that is optimised for fitting longer terms.\n\nOn top of these word clouds we have defined a graphical layout that integrates functional annotations of multiple gene lists with experimental data. In summary, GOsummaries produces dense visualisations, summarising large quantities of information, that cannot be recreated easily with existing tools.\n\n\nUse cases\n\nAs a practical example we used data from an experiment, where gene expression was measured in developing embryoid bodies at nine time points starting from stem cells23. The goal of the experiment was to understand temporal patterns of gene regulation that guide the differentiation process.\n\nTo achieve these goals, it is natural to first cluster the genes from high-throughput analysis into groups with similar behaviour and then characterise the groups functionally using GO enrichment analysis. GOsummaries visualisation is helpful in the interpretation and presentation of the clustering results. Figure 2 shows the GOsummaries results of k-means clustering (k = 5) on the time series. The main trends in the data are immediately clear from the figure. The genes that are related to stem cell maintenance are gradually turned off in the first few days. At the same time developmental genes are turned on in waves: first the embryonic morphogenesis and mesoderm development genes, then in two waves the cardiovascular system-related genes and, finally, the cell adhesion and immune system related genes.\n\nFive clusters presented in Figure 2 may not give an adequate overview of the dataset, increasing the number of clusters can reveal more interesting patterns. Supplementary Figure S2 shows the results for clustering where k = 20. Using GOsummaries to display the results makes the comparison of the clusterings easy. It is possible to see what clusters were divided, how did the division influenced the annotations and if any new interesting patterns emerged. For example, Cluster 4 in Figure 2 has been divided into three clusters in Supplementary Figure S2 (clusters 14–16). Although the expression patterns look very similar, the annotations are somewhat different between the new clusters. Cluster 7, that has a very distinct functional profile is a nice example of a new pattern emerged in the second clustering. In some other cases the new clusters have weak annotations, suggesting that they can be either combined together or ignored.\n\nEach cluster is described by a boxplot showing expression of the genes and a word cloud showing the most significant GO results. In the boxplot, each box corresponds to one sample and the y-axis shows the expression values. In the word clouds the size of the words is proportional to -log10 of enrichment p-value within one word cloud. The absolute enrichment strength of terms (words) is color-coded in grayscale.\n\nTo illustrate the utility of the PCA visualisation of GOsummaries we used the gene expression compendium published by Lukk et al24. This dataset is a collection of publicly available gene expression data covering 5372 samples from 206 studies, with annotations that were thoroughly re-curated by the authors. The analysis in the original article was based on principal component analysis. They inspected the first four principal components and related them with the cell types and tissues by visual inspection of the distribution of samples.\n\nUsing GOsummaries on this data could improve the analysis in two aspects. First, the GO annotations would add another dimension to the interpretation of the principal axis. Second, a dataset that is as diverse as this one may enclose more interesting features beyond the first four principal components; and therefore its analysis could directly benefit from GOsummaries that can easily create plots with tens of principal components to be screened for interesting associations.\n\nWe applied the GOsummaries approach and plotted the first 20 principal components (Supplementary Figure S3). Then we selected 3 additional interesting components for Figure 3. The GO annotations of the first components match well with the names and descriptions of the components presented in the original article. First component was named “Hematopoietic axis”. Fittingly, the GO annotations were strongly related to immune function in the negative end of the axis where the blood cells were clustered. Second component was named “Malignancy axis” and the most dominant GO annotations related to cell line and cancer samples were cell cycle and DNA replication. But there are informative components beyond the first four that were studied in the original article. For example, the eighth component clearly distinguishes muscle cells from everything else and indeed the GO annotations are also muscle related. Several other cases where GO annotations match well with the distributions of samples along different principal axes can be found in the Supplementary Figure S3, where first 20 principal components are shown.\n\nEach component is described by a histogram showing the distribution of samples along the principal axis and word clouds showing the GO annotations for most influential genes. The left and right word clouds represent 500 genes with largest negative and positive loadings respectively. The percentages next to histograms show the percentage of variation explained by each component.\n\nIn these examples we already knew what to expect from the GO annotations. In practice, however, there are often situations, where we can identify clear subclasses from the PCA results, but cannot characterise them any further. In these cases, the GO annotations could give invaluable insights to explain the patterns in the data.\n\nPrincipal Coordinate Analysis (PCoA) is a common approach for visualising taxon abundance data in metagenomic studies. The method is closely related to PCA and its results are usually presented in a similar manner as two- or three-dimensional scatterplots, with the same shortcomings. Thus, using GOsummaries on PCoA results of metagenomics data could make the results more interpretable.\n\nAs an example, we use a small subset of Human Microbiome Project 16S dataset that contains samples from various body sites. We applied PCoA using Bray-Curtis dissimilarity on the data to identify three principal coordinate axes. Then we associated taxons from the original data to the principal coordinates using Spearman correlation test and displayed the results using GOsummaries (Figure 4).\n\nEach component is described by a histogram showing the distribution of samples along the principal coordinate axis and word clouds showing most correlated features. The sizes and colours of the taxons in word clouds are proportional to -log10 of the Spearman rank correlation test p-value. The right and left word clouds represent taxons that were significantly either correlated or anti-correlated respectively with the principal coordinate.\n\nThe traditional scatterplot view would have told us only that there is a clear difference in microbial composition in different body sites. However, the GOsummaries version also identifies the taxons that contribute to the difference. For example, according to the first principal coordinate, the skin, ear and nasal sites have more abundant Actinobacteria, previously shown to be a dominant component of skin microbiota25. According to second and third principal coordinates, vaginal sites tend to have more abundant Lactobacillus, previously shown to be an important part of healthy vaginal microflora26.\n\nGOsummaries visualisation added considerable analysis depth to the PCoA of microbiome data, by revealing underlying differences between experimental groups.\n\n\nSummary\n\nHere we describe an R package GOsummaries that can be used for visualising functional annotations. By showing the annotations as word clouds and combining them with plots of underlying experimental data it is possible to create concise summaries of common analyses. The approach can be applied to any gene list, but is especially useful for clustering, PCA and differential expression results. We show the utility and wide applicability of the tool through three case studies. In comparison with other tools, we demonstrate that GOsummaries word clouds are compact but still manage to convey most relevant biological information. As the analysis pipeline used by GOSummaries is highly automated, it is easy to use and can be useful in many practical situations.\n\n\nSoftware availability\n\nhttp://www.bioconductor.org/packages/release/bioc/html/GOsummaries.html\n\nhttps://github.com/raivokolde/GOsummaries\n\nhttps://github.com/F1000Research/GOsummaries\n\n(F1000Research TO GENERATE)\n\nGPL-2", "appendix": "Author contributions\n\n\n\nRK conceived and implemented the tool, and performed case studies, RK and JV drafted the article. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe research is funded by Estonian Research Council [IUT34-4], target funding [SF0180008s12], ESNATS [HEALTH-F5-2008-201619] and European Regional Development Fund through the EXCS and BioMedIT projects.\n\nI confirm that the 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 Dr Marcel Leist, Dr Jüri Reimand, Konstantin Tretyakov, Dr Kersti Jääger and members of BIIT working group for helpful discussions and suggestions.\n\n\nSupplementary material\n\n\n\n\nReferences\n\nGene Ontology Consortium. Creating the gene ontology resource: design and implementation. Genome Res. 2001; 11(8): 1425–1433. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDennis G Jr, Sherman BT, Hosack DA, et al.: DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 2003; 4(5): P3. PubMed Abstract \| Publisher Full Text\n\nAl-Shahrour F, Minguez P, Vaquerizas JM, et al.: BABELOMICS: a suite of web tools for functional annotation and analysis of groups of genes in high-throughput experiments. Nucleic Acids Res. 2005; 33(Web Server issue): W460–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReimand J, Arak T, Vilo J: g:Profiler--a web server for functional interpretation of gene lists (2011 update). Nucleic Acids Res. 2011; 39(Web Server issue): W307–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEden E, Navon R, Steinfeld I, et al.: GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics. 2009; 10(1): 48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMerico D, Isserlin R, Stueker O, et al.: Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One. 2010; 5(11): e13984. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchröder MS, Gusenleitner D, Quackenbush J, et al.: RamiGO: an R/Bioconductor package providing an AmiGO visualize interface. Bioinformatics. 2013; 29(5): 666–668. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBastos HP, Sousa L, Clarke LA, et al.: GRYFUN: a web application for GO term annotation visualization and analysis in protein sets. PLoS One. 2015; 10(3): e0119631. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang X, Li J, Lee Y, et al.: GO-Module: functional synthesis and improved interpretation of Gene Ontology patterns. Bioinformatics. 2011; 27(10): 1444–1446. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang J, Duncan D, Shi Z, et al.: WEB-based GEne SeT AnaLysis Toolkit (WebGestalt): update 2013. Nucleic Acids Res. 2013; 41(Web Server issue): W77–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSupek F, Bošnjak M, Škunca N, et al.: REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One. 2011; 6(7): e21800. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPascovici D, Keighley T, Mirzaei M, et al.: PloGO: plotting gene ontology annotation and abundance in multi-condition proteomics experiments. Proteomics. 2012; 12(3): 406–410. PubMed Abstract \| Publisher Full Text\n\nTabas-Madrid D, Nogales-Cadenas R, Pascual-Montano A: GeneCodis3: a non-redundant and modular enrichment analysis tool for functional genomics. Nucleic Acids Res. 2012; 40(Web Server issue): W478–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOesper L, Merico D, Isserlin R, et al.: WordCloud: a Cytoscape plugin to create a visual semantic summary of networks. Source Code Biol Med. 2011; 6(1): 7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaroukh C, Jenkins SL, Dannenfelser R, et al.: Genes2WordCloud: a quick way to identify biological themes from gene lists and free text. Source Code Biol Med. 2011; 6(1): 15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDesai J, Flatow JM, Song J, et al.: Visual presentation as a welcome alternative to textual presentation of gene annotation information. Adv Exp Med Biol. 2010; 680: 709–715. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZeeberg BR, Liu H, Kahn AB, et al.: RedundancyMiner: De-replication of redundant GO categories in microarray and proteomics analysis. BMC Bioinformatics. 2011; 12(1): 52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545–15550. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmyth GK: Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004; 3: Article3. PubMed Abstract \| Publisher Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer. 2009. Publisher Full Text\n\nIrizarry RA, Bolstad BM, Collin F, et al.: Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. 2003; 31(4): e15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSegata N, Izard J, Waldron L, et al.: Metagenomic biomarker discovery and explanation. Genome Biol. 2011; 12(6): R60. PubMed Abstract \| Free Full Text\n\nGaspar JA, Doss MX, Winkler J, et al.: Gene expression signatures defining fundamental biological processes in pluripotent, early, and late differentiated embryonic stem cells. Stem Cells Dev. 2012; 21(13): 2471–2484. PubMed Abstract \| Publisher Full Text\n\nLukk M, Kapushesky M, Nikkilä J, et al.: A global map of human gene expression. Nat Biotechnol. 2010; 28(4): 322–324. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGao Z, Tseng C, Pei Z, et al.: Molecular analysis of human forearm superficial skin bacterial biota. Proc Natl Acad Sci U S A. 2007; 104(8): 2927–2932. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVásquez A, Jakobsson T, Ahrné S, et al.: Vaginal Lactobacillus flora of healthy Swedish women. J Clin Microbiol. 2002; 40(8): 2746–2749. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "10188", "date": "08 Sep 2015", "name": "Shaillay Dogra", "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\nGOsummaries is another nice way to summarize visually biologic process related data. Novel tools to visualize information in innovative and meaningful ways are always a welcome addition. The application on top of data from PCA is interesting. So is the informative use on microbiome metagenomics data. Overall, GOSummaries sounds like a good package on top of other useful packages plus novel functions of its own merit. The manuscript is very well-written, is easy to read and follow.Specifically:Is it possible to add functionality to keep term sizes across different word clouds to a fixed-scale option that then enables a user to be able to compare across separately generated figures (I note that authors have already mentioned this is not possible currently and also the reasons thereof). In Figure 1, authors have generated word-clouds on same input data using different available software options including their own GOsummaries. For the benefit of the readers, could the authors discuss on why different tools emphasize different words in their vizualitaions? What underlying assumptions of the different tools make the same input list be represented in different word sizes? For ex. GOsummaries emphasized: cardio vascular development; Cytoscape emphasized: regulation; genes2wordcloud emphasized: frequency, structure --- why is this so? For a general reader more on experimental biology background this may seem confusing and as if the tools are unreliable. Some discussion from authors on this aspect will be beneficial to the readers.", "responses": [] }, { "id": "10691", "date": "14 Oct 2015", "name": "Hilary Ann Coller", "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 have developed a method for summarizing information Gene Ontology searches in a format that makes it easier to understand. They focus on the most relevant Gene Ontology terms and make a word picture with the size of the words indicating their strength in the Gene Ontology results. There is also a method to take principal components of a gene expression dataset and visualize the behavior of different components in each sample along with the word pictures describing the gene ontology categories. The formatting is easy to understand and I expect that this software will be valuable for scientists investigating gene expression analysis.", "responses": [] } ]	1	https://f1000research.com/articles/4-574
https://f1000research.com/articles/4-45/v1	13 Feb 15	{ "type": "Research Article", "title": "Structure and dynamics of the membrane attaching nitric oxide transporter nitrophorin 7", "authors": [ "Markus Knipp", "Hideaki Ogata", "Giancarlo Soavi", "Giulio Cerullo", "Alessandro Allegri", "Stefania Abbruzzetti", "Stefano Bruno", "Cristiano Viappiani", "Axel Bidon-Chanal", "F. Javier Luque", "Markus Knipp", "Hideaki Ogata", "Giancarlo Soavi", "Giulio Cerullo", "Alessandro Allegri", "Stefania Abbruzzetti", "Stefano Bruno", "Axel Bidon-Chanal", "F. Javier Luque" ], "abstract": "Nitrophorins represent a unique class of heme proteins that are able to perform the delicate transportation and release of the free-radical gaseous messenger nitric oxide (NO) in a pH-triggered manner. Besides its ability to bind to phospholipid membranes, the N-terminus contains an additional Leu-Pro-Gly stretch, which is a unique sequence trait, and the heme cavity is significantly altered with respect to other nitrophorins. These distinctive features encouraged us to solve the X-ray crystallographic structures of NP7 at low and high pH and bound with different heme ligands (nitric oxide, histamine, imidazole). The overall fold of the lipocalin motif is well preserved in the different X-ray structures and resembles the fold of other nitrophorins. However, a chain-like arrangement in the crystal lattice due to a number of head-to-tail electrostatic stabilizing interactions is found in NP7. Furthermore, the X-ray structures also reveal ligand-dependent changes in the orientation of the heme, as well as in specific interactions between the A-B and G-H loops, which are considered to be relevant for the biological function of nitrophorins. Fast and ultrafast laser triggered ligand rebinding experiments demonstrate the pH-dependent ligand migration within the cavities and the exit route. Finally, the topological distribution of pockets located around the heme as well as from inner cavities present at the rear of the protein provides a distinctive feature in NP7, so that while a loop gated exit mechanism to the solvent has been proposed for most nitrophorins, a more complex mechanism that involves several interconnected gas hosting cavities is proposed for NP7.", "keywords": [ "Nitrophorin 7", "X-ray crystallography", "ultrafast laser flash photolysis", "ligand rebinding", "molecular dynamics simulations", "inner cavities", "kinetics mechanism" ], "content": "Introduction\n\nNitrophorins (NPs) comprise a family of heme binding proteins that originate from the saliva of the Rhodnius species of Triatomid bugs1,2. These insects use the heme iron as an anchor to store nitric oxide (NO), which is released during feeding on a host while the saliva is continuously pumped into the host tissue3. NO assists the feeding process by inhibition of blood coagulation and increase of blood vessel diameter. The heme cofactor is located inside an antiparallel 8-stranded β-barrel, a typical motif of the lipocalin fold4. The lipocalins represent a widespread family of single domain proteins that typically bind hydrophobic small molecules inside their β-barrel. Prominent examples of lipocalins are the retinol binding protein or the cow milk allergen β-lactoglobulin4,5. For a long time NPs were regarded as the only lipocalin able to bind heme. Recently, we reported on the human lipocalin α1-microglobulin (α1m) for which heme binding was previously suggested6 and is now established7. However, heme coordination is accomplished in a completely different manner in α1m, which forms an unusual [(α1m)(heme)2]3 complex of yet unknown structure.\n\nAt present, five NPs from the species Rhodnius prolixus, designated NP1, NP2, NP3, NP4 and NP7, have been recombinantly expressed and characterized in detail. High-resolution X-ray crystal structures were solved from NP1, NP2 and NP4. While NP1 and NP2 share very high amino acid sequence identity with NP4 (88%) and NP3 (81%), respectively, NP7 exhibits notable differences, which are mostly reflected in its high pI of 9.2 compared to the range of 6.1–6.5 covered by NP1-48. This extreme divergence is a consequence of the presence of 27 Lys out of a total of 185 residues, which, as homology modeling has revealed, cluster mostly at the surface at the side opposite to the heme pocket8. Thus, in contrast to the other NPs, NP7 binds to negatively charged phospholipid membranes with high affinity (4.8 nM)9,10.\n\nSince the administration of NO as a pharmacologically active compound is of major interest for the treatment of cardiovascular diseases or cancer11, the understanding of the structural and dynamic aspects of NO transport and storage by NPs is a major motivation for their detailed investigation. Here we report on seven crystal structures of NP7 at low and high pH and with different heme ligands. Moreover, molecular dynamics (MD) simulations were carried out to examine specific features of NP7, including the structural impact of the three extra residues present at the N-terminus of NP7, which represents a unique distinctive trait among NPs, and the topology of inner cavities. Finally, the structural studies were accompanied by fast and ultra-fast laser-flash photolysis experiments, which allowed to characterize the ligand migration and binding in NP7 from the 10-12 to the 10-3 s time-scale.\n\n\nMaterials and methods\n\nThe protein was expressed, reconstituted, and purified as was described previously10,12. The extended purification by cation exchange chromatography using a Ca2+-charged chelating Sepharose (GE Healthcare)13 was essential for the successful crystallization. Protein preparations were judged by SDS-PAGE to be >95% pure. MALDI TOF MS (Voyager DE Pro, Applied Biosystems) confirms the correct molecular masses including an initial Met-0 residue and accounting for two Cys–Cys disulfides (calculated for [NP7 + H]+: 20,969 Da, observed: 20,966 ± 20 Da). Proteins were kept at –20ºC in 200 mM NaOAc/HOAc (Carl Roth), 10% (v/v) glycerol (Carl Roth) (pH 5.0) until use.\n\nProtein crystals were obtained from 10 mg mL-1 NP7 in 10 mM NaOAc/HOAc (pH 5.5) using the vapor-diffusion method12,14 upon mixing with an equal volume of crystallization solution (Hampton research) as indicated in Table 1. The crystals were soaked for 10 min on ice in the respective crystallization solution plus 15% glycerol as a cryo-protectant. Where needed, heme ligands were added into the cryo-protectant solution. Afterwards, the crystals were immediately frozen in liquid nitrogen and kept there until the measurement. Diffraction data sets were collected at 100 K using the beamlines, BL14.2 at BESSYII (Helmholtz-Zentrum Berlin, Germany) and PXII at SLS (Villigen, Switzerland). The data set was processed with XDS15 and CCP416. The molecular-replacement method was applied using MolRep16 and an initial model from NP4 (PDB code 3MVF)17. Model building and refinement were carried out using WinCoot18 and Phenix19, respectively. Data collection and refinement statistics are summarized in Table 1. PHENIX was used to check the stereochemical properties.\n\nSamples for laser flash photolysis experiments were prepared by equilibrating the solutions in a sealed 0.2×1 cm-quartz cuvette connected to a tonometer with 0.1 or 1 atm CO. Na2S2O4 was then titrated anaerobically into the solution while formation of the CO adduct was monitored by absorption spectroscopy. The solubility of CO in water was such that its concentration was 1.05 mM at 10°C when the solution was equilibrated with 1 atm CO. The temperature dependence of CO solubility20 was taken into account in the numerical analysis. The numerical analysis of the rebinding kinetics was described in detail previously21,22.\n\nThe experimental setup used for the pump-probe studies has been described23. The sample was excited with a 70-fs,-10 nm bandwidth pump pulse centered at 530 nm obtained by a home-made optical parametric amplifier, driven by a regeneratively amplified Ti:sapphire laser at 1 kHz repetition rate. The probe pulse was a broadband single-filament white-light continuum generated in CaF2, covering a spectral range from 350 nm to 700 nm approximately. The pump pulse passed through a delay line and was then overlapped with the probe beam on the sample. The transmitted probe light was dispersed on an optical multichannel analyzer equipped with fast electronics, allowing single-shot recording of the probe spectrum at the full 1 kHz repetition rate. By changing the pump-probe delay we recorded 2D maps of the differential absorption (ΔA), as a function of probe wavelength and delay. The temporal resolution was determined to be approximately 150 fs and the sensitivity, for each probe wavelength, was better than 10-4.\n\nThe experimental setup was described in detail elsewhere21. Photolysis of CO complexes was obtained using the second harmonic (532 nm) of a Q-switched Nd:YAG laser (Surelite-Continuum II-10) at 10 Hz repetition rate. We used a cw output of a 75 W Xe arc lamp as probe beam, a 5-stages photomultiplier (Applied Photophysics) for detection and a digital oscilloscope (LeCroy LT374, 500 MHz, 4 GS s−1) for digitizing the voltage signal. A spectrograph (MS257 Lot-Oriel) was used to select the monitoring wavelength (436 nm) and to remove the stray light from the pump laser. The overall temporal resolution was determined to be about 10 ns and the sensitivity was better than 10-4. The sample holder was accurately temperature-controlled with a Peltier element, allowing a temperature stability of better than 0.1°C. Experiments were performed at 20°C.\n\nThe analysis of the entire CO rebinding kinetics required the use of a detailed kinetic scheme (see section Ligand rebinding kinetics for details) in order to estimate the microscopic rate constants. The differential equations associated with the kinetic schemes were solved numerically, and the rate constants were optimized to describe simultaneously the experimental data at two different CO concentrations. Numerical solutions to set coupled differential equations associated with reaction schemes were determined by using the function ODE15s within Matlab 7.0 (The MathWorks, Inc., Natick, MA). Fitting of the numerical solution to experimental data was obtained with a Matlab version of the optimization package Minuit (CERN).\n\nSpectra acquired in the fs-ps time scale were analyzed by singular value decomposition (SVD), performed using the software Matlab. Our data matrix D, consists of differential absorption measured as a function of two variables: the wavelength of the probe beam and the time delay between the pump and the probe pulses. The singular value decomposition of D can be written as:\n\nD = USVT (1)\n\nThe meaning of the symbols is the following: columns of matrix U are a set of linearly independent, orthonormal basis spectra; columns of V represent the amplitudes of these basis spectra as a function of time; VT is the transpose of matrix V; matrix S is a diagonal matrix of non-negative singular values that give the extent of the contributions of each of the products of the i-th column vectors UiViT to the data matrix D. The selected components can be further screened by evaluating the autocorrelation of the corresponding columns of U and V, rejecting the component if the autocorrelation falls below 0.824.\n\nMD simulations were used to explore the conformational space of wild type NP7 and its NP7(Δ1–3) variant, specifically regarding the structural flexibility of the N-terminus and the loops that shape the mouth of the heme cavity. The molecular systems were modeled from the X-ray structures of NO-bound NP4 at pH 5.6 and 7.4 (PDB ID: 1X8O, resolution of 1.0 Å; 1X8N; resolution of 1.1 Å)25 as representative systems at low and high pH. Following previous studies26–28, standard ionization states were assigned to all the residues with the only exception of Asp32, which was protonated at low pH and ionized at high pH. The two native disulfide bridges were also imposed between residues Cys5 and Cys124 and between Cys42 and Cys17329. Finally, the heme was modeled in the A orientation, which was shown to be the thermodynamically favored orientation for both NP7 and Met-NP7(Δ1–3)8,30.\n\nSimulations were run using the charmm2231 force field and the NAMD program32. The protein was immersed in a pre-equilibrated cubic box (~70 Å per side) of TIP3P33 water molecules. Bonds involving hydrogen atoms were constrained at their equilibrium length using SHAKE and SETTLE algorithms, in conjunction with a 2 fs time step for the integration of the Newton's equations. Trajectories were collected in the NPT (1 atm, 300 K) ensemble using periodic boundary conditions and Ewald sums (grid spacing of 1 Å) for long-range electrostatic interactions. A multistep protocol was used to minimize and equilibrate the system. Thus, the energy minimization was first performed for the hydrogen atoms, then water molecules, and finally the whole system. The equilibration was performed by heating from 100 to 298 K in four 200-ps steps at constant volumen. Then, a 200 ps MD at constant pressure and temperature was run. The final structure was used for the production MD runs, which covered 100 ns. Frames were collected at 1 ps intervals, which were subsequently used to analyze the trajectories.\n\n\nResults and discussion\n\nConditions for the crystallization of NP7 at pH 7.8 were previously reported12. The addition of di- and polyanionic substances to compensate for the positively charged surface of NP7 that is responsible for the binding to negatively charged phospholipid membranes was crucial for the crystal formation9,10. The previously reported crystals diffracted to a resolution of 1.8 Å12. Further optimization of the crystallization conditions resulted in crystals that diffract to even higher resolution down to 1.29 Å (Table 1). It was also possible to crystallize the protein at low pH conditions, i.e., pH 5.8. Since charge compensation for the Lys surface patch was crucial, we also tried to add the amine coordinating bisphosphonate (“lysine tweezer”) 1 (Scheme 1), which is known to be a Lys-specific binder34, resulting in high resolution crystals. Another additive that turned out to be successful was the Gly–Gly–Gly tripeptide.\n\nThe highest resolution shell is shown in parenthesis.\n\nThe crystallization conditions (all reagents from Hampton Research) were; a) 25%(w/v) PEG 3350, 0.1M MES monohydrate pH 5.8, 0.25%(w/v) Gly-Gly, 0.25%(w/v) Gly-Gly-Gly, 0.25%(w/v) Gly-Gly-Gly-Gly, 0.25%(w/v) pentaglycine, 0.02 M HEPES sodium pH 6.8.\n\nb) 25%(w/v) PEG 3350, 0.1 M bis-Tris propane pH 7.8, 0.02 M HEPES pH 6.8, 0.25%(w/v) naphthalene-1,3,6-trisulfonic acid trisodium salt hydrate, 0.25%(w/v) 2,6-naphthalenedisulfonic acid disodium salt, 0.25%(w/v) 4-aminobenzoic acid, 0.25%(w/v) 5-sulfosalicylic acid dehydrate.\n\nc) 25%(w/v) PEG 3350, 0.1 M bis-Tris propane pH 7.8, 0.25%(w/v) Gly-Gly, 0.25%(w/v) Gly-Gly-Gly, 0.25%(w/v) Gly-Gly-Gly-Gly, 0.25%(w/v) pentaglycine, 0.02 M HEPES sodium pH 6.8.\n\nd) 25%(w/v) PEG 3350, 0.1 M bis-Tris propane pH 7.8. 0.25%(w/v) hexamminecobalt(III) chloride, 0.25%(w/v) salicylamide, 0.25%(w/v) sulfanilamide, 0.25%(w/v) vanilic acid, 0.02 M HEPES sodium pH 6.8.\n\nTo obtain different iron-liganded structures, the crystals were soaked with imidazole (ImH), histamine (Hm), or the NO donating compound DEA/NONOate for 10 min at room temperature. Upon incubation with the mother liquor containing 15% of glycerol as a cryo-protectant, crystals were frozen in liquid nitrogen. Diffraction experiments were carried out at two different synchrotrons at 100 K.\n\nAll crystals occupied the space group P21. Refinement of the crystal structures was obtained through the molecular replacement method using the structure of NP4 (PDB code 3MVF)17 as a template. The refinement statistics are summarized in Table 1. It should be mentioned that no electron density of the additives (see Table 1 footnotes) except for the Gly–Gly–Gly tripeptide was observed.\n\nThe overall fold of NP7 resembles that of other NPs. Eight anti-parallel β-strands (A to H) form a barrel that hosts the heme cofactor including the ligation by the proximal His60 residue. The structural identity reflected by the RMSD values obtained from the comparison of the backbone atoms of two of the isoforms correlates with the amino acid sequence identities (Table 2). Figure 1 displays the overall fold compared to those of NP2 and NP4. The core of the lipocalin fold is well superposed in all cases. The position of the two disulfide bridges, which is a common trait of the NPs, is very similar among all the NPs. The largest differences are found in the A-B, B-C and G-H loops. In detail, the bending of the β-strands (βB and βC) is significantly more similar in the case of NP7 and NP2 compared to NP4. This is also true for the AB-loop. On the other hand, significant differences are found in the spatial arrangement of the G-H loop, which is markedly bent in NP7 compared to NP2 and NP4 (Figure 1).\n\na The RMSD calculation is based on the following PDB files: NP1, 2NP1; NP2, 2A3F; NP4, 1YWD; NP7, 4XMC.\n\n(a) Overall fold of NP7. (b) Overall fold of NP7 (green) in comparison to NP2 (blue) (PDB code, 2A3F) and NP4 (orange) (PDB code, 1YWD).\n\nOne of the most interesting features of NP7 is its ability to bind to negatively charged membranes. The crystal structures demonstrate the extensive clustering of Lys side-chains at the protein surface opposite the heme pocket that accomplishes the strong protein-membrane interaction. On the other hand, the side of the heme mouth has a negative charge potential leading to a total bipolar charge distribution. It was previously noticed that NP7 tends to aggregate at elevated concentrations8,35, which can be explained by a charge stabilized aggregation process. X-ray crystallography indeed supports such a head-to-tail interaction, which involves a variety of salt bridges (Figure S1), leading to chain-like arrangement in the crystal lattice (Figure 2a).\n\n(a) Arrangement of molecules in the crystal lattice of NP7. (b) Crystal contact between two neighboring molecules in the NP7 crystal. (c) The electrostatic potential of NP7. The red and blue colors show the negative and positive potentials, respectively.\n\nThe presence of Glu27 residue in the heme pocket of NP7 is unique among NPs36,37. Upon homology modeling it was noticed that the Glu27 carboxylate must somewhat interfere with the hydrophobic site of the cofactor, i.e., its vinyl and methyl substituents. Only A orientation is observed in NP78,30, whereas B orientation is favored in NP238,39 (Figure 3). Interestingly, the mutant that converts Glu27 to Val, i.e. the residue found in all the other NPs (NP7(E27V)), demonstrated that the heme orientation was reversed from A to B, whereas the replacement of Glu27 by Gln, NP7(E27Q), did not change the heme orientation compared to the wild type, and replacement of Val24 by Glu in NP2, NP2(V24E), resulted in the B → A reorientation of the cofactor30.\n\nThe heme pocket of NP7 is indicated by the three green circles corresponding to the position of the three aliphatic side-chains pointing from the top of the distal pocket onto the heme plane. The two possible heme orientations are indicated as A and B.\n\nIn the crystal structure of NP7[ImH] the electron density map clearly reflects the presence of heme A orientation, as expected for NP7 (see above). However, to our surprise, the NP7 structure that crystallized with Gly–Gly–Gly has heme B orientation (Table 1 and Figure 4). Thus, it may be concluded that the insertion of this ligand leads to a widening of the heme pocket, which may allow the heme to turn. Since the structures of NP7[NO], NP7[Hm] and unliganded NP7 at pH 5.8 were derived from crystal soaking and displacement of the Gly–Gly–Gly ligand, the heme consequently was found in the heme B orientation as the crystal lattice does not allow the cofactor to turn. However for NP7[NO], A orientation was demonstrated by circular dichroism spectroscopy30. Caution is needed regarding the presence of B orientation in unliganded NP7 at pH 7.8, since it could not be properly assigned due to the loosely occupied heme (Table 1).\n\n2Fo-Fc electron density map (contoured at 1σ) of the heme pocket of (a) NP7 at pH 5.8, (b) NP7 at pH 7.8, (c) NP7[NO], (d) NP7[ImH], (e) NP7[Hm], and (f) NP7[Gly-Gly-Gly].\n\nIn the electron density map derived from the diffraction pattern of the unliganded NP7 at pH 7.8 the heme was very poorly defined, while the rest of the structure was very well defined (Figure 4b). In contrast, the electron density of the cofactor bound to ImH, Hm and NO was well defined (Figures 4c–e). Since the ferriheme iron undergoes rapid photo reduction during the recording of diffraction patterns40–43, the crystal structures display the FeII state. We had previously reported that the FeII–NHis60 bond in NP7 is surprisingly weak, so that the reduction of NP7 leads to the equilibrium noted in Equation 2 at neutral pH (pKa = 7.8)44.\n\nH2O + FeII(ppIX)–NHis60 ⇆ H2O–FeII(ppIX) + NHis60 (2)\n\nOverall, these findings suggest that the ill-defined heme density of the unliganded NP7 reflects the movement of the cofactor inside the heme pocket according to the two coordination states. The binding of the π donors ImH, Hm or CO has a positive trans effect, which stabilizes the FeII–NHis60 bond, so that the heme resides in the pocket. Interestingly, NO has a negative trans effect due to the overlap of its anti-bonding π* orbital with the iron dz² orbital, thus weakening the Fe–NHis σ bond45–47.\n\nThe crystal structures allow detailed examination of Glu27 (Figure 5a). The side-chain is folded away from the hydrophobic heme side toward the interior of the structure, where it is involved in a network of H-bonding contacts. This includes a single water that is further coordinated to Tyr175 near the protein surface. It was previously found that the mutation Glu27→Gln has a remarkable destabilizing effect on the NP7 fold30. It can now be understood that the negative charge of Glu27 attracts the Tyr175 hydroxyl group, which forms an important interaction. On the other hand, the dense packing of side-chains next to His60 was identified as another destabilizing factor of the FeII–NHis60 bond, where Phe43 plays a crucial role44. Figure 5b shows the arrangement in comparison to the crystal structures of NP2 and NP4. Phe43 is oriented parallel to the heme plane with a distance of 3.5 Ǻ leading to π-stacking between the two aromatic rings. Moreover, the phenyl ring is perpendicularly oriented toward the His60 plane with a distance of 3.6 Ǻ. The distance between Glu27:Cβ and Phe43:Cβ is 4.0 Ǻ, while Glu27 also H-bonds to Phe43:NH.\n\nSpatial location relative (a) to the heme cofactor and (b) with respect to His60 and Phe43. For comparison, the structures of NP7(green), NP2 (blue) and NP4 (orange) are displayed.\n\nIn the crystal structure, NO is bound with ∠(Fe–N–O) = 124°, which corresponds to a reduced iron, i.e. FeII(NO). In the case of NP1[NO] a similar angle of 123° to 135° was reported48. For NP4[NO] two orientations of 110° and 177° were assumed because the electron density could not be sufficiently fit with a single conformation49. The data were interpreted by the parallel co-existence of photoreduced FeII(NO) and some residual FeIII(NO) in the frozen crystal. However, according to our41 and other40,43 work, photoreduction occurs even in frozen crystals within the first seconds of data collection, thus the presence of significant amounts of NP4[FeIII(NO)] seems very unlikely. Also in the case of NP7[NO] some of the electron density could not be sufficiently fit with the presence of a single NO configuration (Figure 4c). However, this may reflect an incomplete occupancy of the axial Fe site with NO and displacement with water since the protein crystals were prepared and transported in the FeIII(NO) form to the synchrotron, whereby loss of NO gas is possible.\n\nBinding of imidazole and histamine. Similar to other NPs, the binding of Hm is accomplished not only through the coordination of heme iron, but also through the salt bridge of its ethylamine group with the Asp32 carboxylate (Figure 6a). When ImH is bound, the missing ethylamine is compensated by H-bonding to a water molecule, which is then coordinated by Asp32 (Figure 6b). However, in marked contrast to other NPs, the affinity constant for ImH and Hm was markedly decreased. While for NP2 the equilibrium constant (Keq) was found to be 2.5 × 107 M-150, for NP7 Keq was 1.0 × 106 M-18. This difference is even more pronounced for Hm, as the equilibrium constants are 1.0 × 108 M-150 and 1.0 × 105 M-1 for NP2 and NP7 [8], respectively. The remarkable difference observed between the two ligands in NP7 as well as between NP2[Hm] and NP7[Hm] is not explained by the structures. However, taking into account that the N-terminus is expected to move into the space between the A-B and G-H loops, the disruption of the Asp32-Hm interaction is feasible. This is further supported by the finding that deletion of the N-terminal residues increases the affinity for Hm (Keq(NP7(Δ1-3)) = 1.3 × 107 M-1). This trend is also found in the enhanced affinity for ImH (Keq(NP7(Δ1-3)) = 3.2 × 107 M-1)8, which compares with the value reported for NP2 (see above).\n\n(a) Hm, (b) ImH, (c) pH 5.8 and (d) pH 7.8. The numbers represent the bond distances (Å).\n\nIt was previously suggested that the pH can have significant influence on the A-B and G-H loop conformations in NP4 due to the change in the protonation state of Asp30 (part of the A-B loop)25,27,51–54. While in the protonated state Asp30 binds to the backbone C=O of Leu130 (part of the G-H loop), it is detached upon deprotonation leading to the opening of the heme pocket51. Based on this mechanism, a model for the pH-dependent NO release under biological conditions was developed. However, very limited pH dependences of the ligand release rates and equilibrium constants leave considerable doubts about this likely oversimplified model55.\n\nIn contrast to NP1-4, where the pH dependence of the equilibrium constant is less than an order of magnitude, the difference for NP7 spans three orders of magnitude8. However, no significant conformational differences were found between the X-ray structures collected at two different pHs (5.8 and 7.8; RMSD = 0.22 Å), even though this might simply arise from the packing of NP7 molecules in the crystal. On the other hand, in all NP7 structures Asp32 is bound to a water that is held by the backbone NH and C=O of Asp134, that is, the Asp32-Ile132 H-bond that corresponds to the Asp30-Leu130 H-bond in NP4 is not observed (Figures 6c and 6d). Thus, there are significant differences in the key structural features during the transition between closed and open forms of the protein. However, it is unclear whether these differences are due to the head-to-tail arrangement (see above and Figure 2), or alternatively might be influenced by the distinct N-terminus stretch of NP7, which fills the region between A-B and G-H loops (Figure S2).\n\nUpon maturation, a major difference of NP7 compared to all other NPs is its extended N-terminus (Figure S3), which is characterized by the extension of the tripeptide Leu1–Pro2–Gly38,56. It should be mentioned that the N-terminus of the recombinant NP7 contains an additional Met0 residue originating from the start codon of the expression system. In the X-ray structures, the residues Met0 and Leu1 were not seen. As noted above, the crystal lattice involves a crystal contact between the A-B and G-H loops of one molecule and the positive patch of the back side of another molecule (Figure 2b), where the sharp E-F loop Lys116 plays a major role, contributing to the head-to-tail arrangement via salt bridges with residues Asp34 and Asp32 (distances of 3.4 and 5.0 Å, respectively). As a consequence, the N-terminus is rather floppy sitting on top of the structure (Figure S3).\n\nA series of MD simulations were run to examine the conformational flexibility and potential interactions formed by the N-terminus stretch of NP7 in aqueous solution. To this end, MD simulations were carried out for models of the wild type protein representative of the closed and open states, and the results were compared with those obtained for the Δ(1-3) variant to ascertain the structural impact of the three extra residues. Furthermore, in order to avoid an artifactual bias due to packing effects, models of wild type and Δ(1-3) NP7 proteins were built up using the X-ray crystallographic structures of NO bound NP4 at pH 5.6 and 7.425, because the conformations of the AB and GH loops are characteristic of the closed and open states, respectively52. All the simulations led to stable trajectories, as noted in RMSD values comprised between 0.9 and 1.3 Å (data not shown).\n\nThe behavior observed for the open forms of wild type NP7 and its Δ(1-3) variant is highly similar, mainly reflecting significant fluctuations of the A-B loop irrespective of the length of the N-terminus (Figure 7a), which for NP7 fills the region between A-B and G-H loops, thus resembling the arrangement found in the X-ray structures of NP7 (Figure S2). This structural resemblance is reduced in the closed state, because the A-B and G-H loops exhibit larger fluctuations in the wild type protein compared to the Δ(1-3) species (Figure 7b), and this effect is accompanied by the enhanced conformational flexibility of the Leu1-Pro2-Gly3 stretch in NP7, which is in contrast with the more rigid structure adopted by the N-terminus in the Δ(1-3) variant. Interestingly, superposition of the snapshots sampled for wild type and Δ(1-3) proteins reveals a widening of the heme cavity mouth in the wild type protein (Figure 7b), which would suggest a higher probability for the formation of transient pathways that connect the heme cavity with the bulk solvent. This is noted, for instance, in the larger area estimated for the heme cavity mouth for the closed states of the wild type NP7 and its Δ(1-3) variant (Figure S4).\n\nSuperposition of the snapshots taken every 10 ns from the MD trajectories sampled for the (a) open and (b) closed states of wild type NP7 and its Δ(1-3) variant. The backbone of the extra Leu-Pro-Gly stretch found in NP7 is shown in magenta, and the backbone of the A-B and G-H loops for wild type and Δ(1-3) NP7 is shown in green and yellow, respectively.\n\nOverall, these findings could explain why deletion of the N-terminal residues increases the affinity for ImH and Hm, an effect that can presumably arise from the more compact nature of the heme cavity and lower fluctuations of the A-B and G-H loops in the Δ(1-3) protein. Finally, the enhanced flexibility of the closed form of wild type NP7 may also explain the larger sensitivity to pH dependence, since the access of water molecules to the heme cavity could facilitate breaking of the H-bond between Asp32 and Ile132, thus favoring deprotonation of Asp32 and hence the transition to the open state.\n\nThe availability of the NP7 structures now allows the combination with the studies of fast dynamics of the ligand-protein interaction. The FeII–CO complex was studied as a model for the isoelectronic FeII–NO because of its slower geminate rebinding rates. The change in absorption spectra of CO bound and unbound heme proteins (Figure S5) allows the determination of time resolved differential absorption spectra after photolysis of the Fe–CO bond. In this study, NP7–CO was pumped with 70-fs laser pulses at 532 nm. Previous studies were performed using nanosecond laser flash photolysis, but the geminate rebinding was not resolved57. We report here that geminate recombination starts in the picosecond range, and merging the subnanosecond kinetics with the nanosecond laser photolysis data allow us to obtain a full time course for ligand rebinding.\n\nFigure 8a shows time resolved differential absorption spectra measured after photolysis of the CO adduct of NP7 at selected time delays. Transient spectra correspond to the difference between the ground state absorption spectra of NP7[FeII–CO] and NP7[FeII], with clean isosbestic points at 402 nm and 426 nm. Accordingly, only one significant spectral component is retrieved from the SVD analysis of the spectra collected between 4 ps and 1 ns (Figure 8b). As a consequence, the time course of the corresponding amplitude V1 perfectly matches the kinetics measured at 436 nm, but with significant noise reduction, as shown in Figure 8c. Similar results are obtained when the experiment is conducted at pH 5.5, where the time resolved differential absorption spectra have the same shape as those measured at pH 7.5. Also at this pH, only one spectral component is obtained from SVD and the time course of the amplitude is shown as red solid circles in Figure 8c. It is easily observed that the time course of V1 at acidic pH occurs with a higher rate, leading to a larger fraction of rebinding at the subnanosecond time scale.\n\n(a) Time resolved differential absorption spectra for NP7–CO (pH 7.5, T = 20°C) following femtosecond photoexcitation at 532 nm at 4 ps (black line), 200 ps (red line), 400 ps (green line) and 800 ps (blue line) delay times. (b) First spectral component (U1) obtained from the SVD analysis of the time resolved differential absorption spectra multiplied by the corresponding singular value (S1 = 0.23). (c) Comparison between the time course of the amplitude V1 (open circles) of the main spectral component obtained from SVD, and the normalized transient absorbance at 436 nm (solid line) as a function of the delay time. The red filled circles report the amplitude V1 of the main spectral component obtained from the experiment conducted at pH 5.5.\n\nWe have previously reported the CO rebinding kinetics to NP7 after nanosecond laser photolysis, showing that the binding reaction can be followed by the absorbance change at 436 nm57. In this work we now aim at reconstructing the full time course of CO rebinding, from a few picoseconds to reaction completion, occurring on the millisecond time scale. We have accomplished this by merging the CO rebinding kinetics determined in the femtosecond pump-probe and in the nanosecond laser flash photolysis experiments.\n\nDue to inherent limitations of the methods employed to collect the kinetics in the two time regimes, the kinetics in the time interval between ~2 ns (the longest available delay in the pump-probe experiment) and 20 ns (the shortest time accessible without distortion due to the instrumental response function of the nanosecond laser flash photolysis setup) is not available. Thus, a connection of the two data sets is not straightforward. In order to obtain a single kinetic trace, we have first normalized the amplitude V1 retrieved from SVD analysis of the pump and probe experiment, taking advantage of the fact that the fraction of photoproduct with respect to that initially generated at time t0, i.e., N(t) = ΔA(t)/ΔA(t0), (with t>t0), decreases with time from unity. Unfortunately, it is not possible to extend this procedure to the nanosecond photolysis data since multiple photolysis events occur during the laser pulse under our experimental conditions, thus impairing a correct determination of the fraction of unliganded species surviving at 20 ns. We have therefore estimated the concentration of the unliganded molecules at the end of the temporal window of pump and probe experiments by fitting the kinetics in this time window with a double exponential decay function, and extrapolating the transient signal beyond 2 ns. The flash photolysis data were then scaled in order to match the extrapolated fraction of unliganded molecules. We should say that, unlike the method proposed by Champion and coworkers58, the current procedure for merging the two different time ranges has some degree of uncertainty since there is about one log10 time unit of missing data which may contain kinetic information. Nevertheless, given the shape of the signals, it is expected that this is not a major contribution for the kinetics reported in this work.\n\nFigure 9 shows the overall rebinding curve for NP7 obtained with this method. From a simple inspection of the progress curves, it is easy to notice that for NP7 the sub-ns kinetics becomes faster and larger when the pH is lowered to 5.5. It is comprised of two kinetic phases, which can be well described by a double exponential decay, whose amplitudes and apparent rates increase at pH 5.5 (not shown). The sub-ns rebinding phase is very similar to the previously described CO rebinding to the related NP458. On the longer time scales, the progress curve for NP7 shows the features previously reported, with a heterogeneous bimolecular rebinding which becomes faster as the pH is lowered to 5.5.\n\nComplete rebinding kinetics to NP7 at pH 7.5 and pH 5.5 (T = 20°C). The experimental progress curves recorded at 1 CO atm (black open circles) and 0.1 CO atm (green open circles) are reported.\n\nRecently we proposed that the reaction progress for NP7 can be described by a microscopic model which takes into account rebinding of photodissociated ligands from internal cavities that are accessible at room temperature57. The current data indicate that in addition to more remote internal cavities, capable of hosting the ligand for a relatively long time, at least one additional temporary docking site exists in the vicinity of the reaction site at the heme, which modulates sub-nanosecond geminate rebinding. From a structural point of view, this assumption is supported by the topological analysis of the inner cavities present in the snapshots sampled along the MD simulation of the closed form of NP7. Thus, Figure 10a shows the three major cavities that shape the tunnel leading from the heme pocket to the back of the protein59 (shown as magenta isocontour) determined by using the MDpocket tool60, and also confirmed by Implicit Ligand Sampling calculations61. Besides the inner tunnel, Figure 10a shows the presence of two additional pockets around the heme. Residues Ile121, Ile123, Leu135 and Ser137 shape the first pocket (orange isocontour), and the side chains of Glu27, Phe43, Phe45 and Leu139 contribute to the second pocket (blue isocontour). Hence, it can be expected that rebinding of photolyzed CO would occur from a variety of transient docking sites, reflecting the topological distribution of pockets located close to the heme (at around ~9 Å from the heme iron) as well as from inner cavities present at the rear of the protein (at around ~22 Å from the heme iron), which can be visited via the inner tunnel present as a distinctive feature in NP7 compared to other NPs.\n\n(a) Representation of the cavities that form the inner tunnel in NP7 (blue) as well as two additional pockets found around the heme (orange and magenta) identified by the MDpocket analysis. (b) Pathways connecting the heme cavity and the bulk solvent in the closed form of NP7.\n\nOn the basis of the previous findings, the reaction scheme previously proposed by us 59 is then expanded as shown in Figure 11 to accommodate for transient population T2, T3 and T4 of the ligand in docking sites located nearby the distal pocket (DP), and is found to describe well the rebinding kinetics under the investigated experimental conditions. Sample fits on CO rebinding kinetics to NP7 are reported in Figure 12.\n\nReaction intermediates T2, T3 and T4 denote docking sites inside the protein.\n\n(A: pH 7.5, B: pH 5.5), at T = 20°C and 1 (black) and 0.1 atm (gray). The fits (yellow lines) are superimposed to the experimental data. In the figure, the time course of the other relevant species shown in Figure 8 is reported: DP (black), T1 (blue), T3 (cyan), T4 (magenta), NP (red) and NP7* (green).\n\nThe microscopic rates reported in Table 3 show that exit to the solvent occurs with a remarkably high rate (9 × 108 s-1 and 8.2 × 108 s-1 at pH 7.5 and 5.5, respectively), which competes, although to a minor extent, with internal rebinding and migration processes. This suggests that the connection between the distal pocket and the solvent must be quite efficient, much more than, for instance, in the case of other proteins with well recognized rather open connections. These include, for instance, myoglobin, for which CO escapes through the His gate at a rate 2.7 × 106 s-162, and type 1 non symbiotic hemoglobin AHb1 for which exit through the His gate occurs at 5 × 107 s-121,63. This nicely parallels the finding of two large and accessible pathways connecting the distal pocket and the solvent (denoted cage-to-top and cage-to-fron in Figure 10b), which can be ascribed to the larger fluctuations of the A-B and G-H loop and widening of the heme cavity in the closed form of wild type NP7 (see Figure 7b).\n\nInspection of the microscopic rates in Table 3 also reveals that there are some substantial differences with respect to the values previously determined on the sole basis of nanosecond laser photolysis data57. Inclusion of the ps rebinding phase results in a major increase in rebinding rates, a fact that is not surprising and was anticipated in our previous analysis. The value of the rebinding rate k-1 is remarkably large compared to, for instance, rebinding to the highly reactive type II truncated Hb from Thermobifida fusca (TrHbO), for which rebinding occurs with a rate of 3 × 108 s-164, a value which is 20 fold smaller than the one for NP7. Mutation of distal cavity residues capable of stabilizing water molecules nearby the binding site lowers the barrier for rebinding and increases the rate to 2 × 109 s-165, a value which becomes comparable to the value for NP7 at neutral pH, but is still one order of magnitude smaller than observed for NP7 at pH 5.5.\n\nOur results compare well with the average rate constant for CO rebinding to NP4, 4.7 × 109 s-1 at pH 7.5 and increasing to 2.1 × 1010 s-1 at pH 5.5, retrieved by a single distribution model58. In their analysis, Champion and coworkers adopted a statistical model to describe the effect of the strong distortion of the heme in NP4, to explain the observed non-exponential rebinding. In the case of NP7 the ps phase is clearly described by a double exponential decay, and therefore the use of a kinetic model that introduces an additional discrete reaction step appears justified.\n\nFinally, it is important to stress that the kon values derived from the current determination of microscopic rate constants reproduces values that well agree with our previous estimates57. The combination of microscopic rates to obtain kon is such that the resulting parameter is mostly sensitive to k-2, whereas it is relatively insensitive to the actual values of k-1 and k2 as long as their ratio remains comparable.\n\n\nConcluding remarks and future directions\n\nThe analysis of the X-ray crystal structures of NP7 at low and high pH and with different heme ligands show the existence of a large resemblance in the overall fold, which is also similar to the spatial arrangement found in other NPs. These structures reveal the extensive clustering of Lys side-chains at the protein surface opposite the heme pocket, which is implicated in the charge-stabilized head-to-tail interaction observed in the crystal lattice. Furthermore, it provides a basis to realize the ability of NP7 to bind to negatively charged membranes. However, packing effects also explain the large structural similarity observed for the different ligand-bound and unbound structures, although such a structural similarity does not permit to rationalize the effect of deletion of the N-terminal residues on the binding affinities of certain ligands nor the pH sensitivity of NP7.\n\nThe peculiar structural and dynamical properties of NP7 suggest that the mechanism at the basis of the pH sensitivity of this protein exhibits differential features with regard to the one that is operative for the other characterized NP isoforms. The extremely high reactivity of the binding site has the consequence that much of the kinetics is compressed in the sub-nanosecond time scale. In the case of the wt protein, the presence of articulate cavities, endowed with temporary docking sites where photodissociated ligands can migrate, and tunnels connecting the distal pocket with the solvent are proposed to be key determinants for the shape of the observed kinetics. In particular, the additional Leu-Pro-Gly stretch at the N-terminus and the presence of a Glu27 residue in the heme pocket which interferes with the heme vinyl and methyl substituents, both unique among NPs, have profound consequences for the heme pocket structure and dynamics. In particular, the role of Glu27 is still poorly understood and deserves further investigations.\n\n\nData availability\n\nF1000Research: Dataset 1. Data of membrane attaching nitric oxide transporter nitrophorin 7, 10.5256/f1000research.6060.d4256766\n\n\nAbbreviations\n\nα1m, α1-microglobulin; DEA/NO, sodium (Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate; Hb, hemoglobin; Hm, histamine; ImH, imidazole; IPTG, isopropyl β-d-1-thiogalactopyranosid; L, distal ligand on heme iron; LFP, laser flash photolysis; MALDI, matrix assisted laser desorption ionization; Mb, myoglobin; MD, molecular dynamics; MOPS, 3-(N-morpholino)propanesulfonic acid; NOS, nitric oxide synthase; NP, nitrophorin; PDB, Protein Databank of the Research Collaboratory for Structural Bioinformatics (RSCB) at http://www.pdb.org/pdb/home/home.do; RMSD, residual mean square deviation; TOF, time-of-flight; wt, wild-type.", "appendix": "Author contributions\n\n\n\nHO and MK conducted the X-ray analysis of the NP7 structures. GS, GC, SA, SB and CV performed the spectroscopic studies and the flash photolysis assays. AA and AB-C contributed to the molecular dynamics simulations. MK, CV and FJL contributed to the experimental design and preparation of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was financially supported by the Deutsche Forschungsgemeinschaft (DFG) grant KN951/1-1, the Max Planck Society (both to M.K.), Spanish Ministerio de Economía y Competitividad (SAF2011-27642; FJL), Generalitat de Catalunya (2014SGR1189; FJL), Icrea Academia (FJL). The Consorci de Serveis Universitaris de Catalunya and the Barcelona Supercomputer Center are acknowledged for computational facilities.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupporting information\n\nResidues Pro and Gly of the extra N-terminus stretch (Leu-Pro-Gly) found in NP7 are shown as green-coloured spheres. The spatial position of Leu, and the additional Met0 residue originating from the start codon of the expression system, were not seen in the X-ray crystal.\n\nThe total sequence identity among all five protein sequences amount to 33% (indicated by ‘*’). The proximal His is indicated by ‘#’. An initial Met residue (in grey) results in case of the recombinantly expressed proteins of NP1, NP2, NP3, and NP7 is not present in the mature proteins in vivo (2, 3). Residues of relevance to this study are highlighted in color.\n\nWild type NP7 in open and closed states is shown as red and black lines; Δ(1-3) variant in open and closed states is shown as blue and green lines.\n\nBlack: NP7[FeIII]; red: NP7[FeII]; blue: NP7[FeII–CO].\n\n\nReferences\n\nLehane MJ: The Biology of Blood-Sucking in Insects. 2nd ed. Cambridge University Press: Cambridge (United Kingdom). 2005. Reference Source\n\nSoares RP, Sant'Anna MR, Gontijo NF, et al.: Identification of morphologically similar Rhodnius species (Hemiptera: Reduviidae: Triatominae) by electrophoresis of salivary heme proteins. Am J Trop Med Hyg. 2000; 62(1): 157–161. PubMed Abstract\n\nWalker FA: Nitric oxide interaction with insect nitrophorins and thoughts on the electron configuration of the {FeNO}6 complex. J Inorg Biochem. 2005; 99(1): 216–236. PubMed Abstract \| Publisher Full Text\n\nFlower DR, North AC, Sansom CE: The lipocalin protein family: structural and sequence overview. Biochim Biophys Acta. 2000; 1482(1–2): 9–24. PubMed Abstract \| Publisher Full Text\n\nSakurai K, Konuma T, Yagi M, et al.: Structural dynamics and folding of beta-lactoglobulin probed by heteronuclear NMR. Biochim Biophys Acta. 2009; 1790(6): 527–537. PubMed Abstract \| Publisher Full Text\n\nLarsson J, Allhorn M, Åkerström B: The lipocalin alpha(1)-microglobulin binds heme in different species. Arch Biochem Biophys. 2004; 432(2): 196–204. PubMed Abstract \| Publisher Full Text\n\nSiebel JF, Kosinsky RL, Åkerström B, et al.: Insertion of heme b into the structure of the Cys34–carbamidomethylated human lipocalin α(1)-microglobulin: formation of a [(heme)(2) (α(1)-Microglobulin)](3) complex. Chembiochem. 2012; 13(6): 879–887. PubMed Abstract \| Publisher Full Text\n\nKnipp M, Yang F, Berry RE, et al.: Spectroscopic and functional characterization of nitrophorin 7 from the blood-feeding insect Rhodnius prolixus reveals an important role of its isoform-specific N-terminus for proper protein function. Biochemistry. 2007; 46(46): 13254–13268. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndersen JF, Gudderra NP, Francischetti IM, et al.: Recognition of anionic phospholipid membranes by an antihemostatic protein from a blood-feeding insect. Biochemistry. 2004; 43(22): 6987–6994. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKnipp M, Zhang H, Berry RE, et al.: Overexpression in Escherichia coli and functional reconstitution of the liposome binding ferriheme protein nitrophorin 7 from the bloodsucking bug Rhodnius prolixus. Protein Expr Purif. 2007; 54(1): 183–191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeller A: Electrochemistry and nitric oxide mass transport in cancer: why ingestion of sodium nitrite could be effective in treating vascularized tumors. Phys Chem Chem Phys. 2010; 12(34): 9972–9975. PubMed Abstract \| Publisher Full Text\n\nOgata H, Knipp M: Crystallization and preliminary X-ray crystallographic analysis of the membrane-binding haemprotein nitrophorin 7 from Rhodnius prolixus. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2012; 68(Pt 1): 37–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRisse C, Taing JJ, Knipp M: Preparation of nitrophorin 7(Δ1-3) from Rhodnius prolixus without start-methionine using recombinant expression in Escherichia coli. Anal Biochem. 2014; 451: 28–30. PubMed Abstract \| Publisher Full Text\n\nBenvenuti M, Mangani S: Crystallization of soluble proteins in vapor diffusion for x-ray crystallography. Nat Protoc. 2007; 2(7): 1633–1651. PubMed Abstract \| Publisher Full Text\n\nKabsch W: Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J Appl Cryst. 1993; 26: 795–800. Publisher Full Text\n\nCollaborative Computational Project, Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 1994; 50(Pt 5): 760–763. PubMed Abstract \| Publisher Full Text\n\nHe C, Ogata H, Knipp M: Formation of the complex of nitrite with the ferriheme b beta-barrel proteins nitrophorin 4 and nitrophorin 7. Biochemistry. 2010; 49(28): 5841–5851. PubMed Abstract \| Publisher Full Text\n\nEmsley P, Cowtan K: Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004; 60(Pt 12 Pt 1): 2126–2132. PubMed Abstract \| Publisher Full Text\n\nAfonine PV, Grosse-Kunstleve RW, Adams PD: The Phenix refinement framework. CCP4 Newsletter. 2005; 42: Contribution 8. Reference Source\n\nSottini S, Viappiani C, Ronda L, et al.: CO Rebinding Kinetics to Myoglobin- and R-State-Hemoglobin-Doped Silica Gels in the Presence of Glycerol. J Phys Chem B. 2004; 108(24): 8475–8484. Publisher Full Text\n\nAbbruzzetti S, Bruno S, Faggiano S, et al.: Time-resolved methods in Biophysics. 2. Monitoring haem proteins at work with nanosecond laser flash photolysis. Photochem Photobiol Sci. 2006; 5(12): 1109–1120. PubMed Abstract \| Publisher Full Text\n\nBruno S, Faggiano S, Spyrakis F, et al.: The reactivity with CO of AHb1 and AHb2 from Arabidopsis thaliana is controlled by the distal HisE7 and internal hydrophobic cavities. J Am Chem Soc. 2007; 129(10): 2880–2889. PubMed Abstract \| Publisher Full Text\n\nPolli D, Laer L, Cerullo G: High-time-resolution pump-probe system with broadband detection for the study of time-domain vibrational dynamics. Rev Sci Instrum. 2007; 78(10): 103108. PubMed Abstract \| Publisher Full Text\n\nHenry ER, Hofrichter J: Singular value decomposition: Application to analysis of experimental data. In Numerical computer method Brand, L., Johnson, M. L., Eds Academic Press, Inc.: San Diego. 1992; 210. ; 129–192. Publisher Full Text\n\nKondrashov DA, Roberts SA, Weichsel A, et al.: Protein functional cycle viewed at atomic resolution: conformational change and mobility in nitrophorin 4 as a function of pH and NO binding. Biochemistry. 2004; 43(43): 13637–13647. PubMed Abstract \| Publisher Full Text\n\nMenyhárd DK, Keserü GM: Protonation state of Asp30 exerts crucial influence over surface loop rearrangements responsible for NO release in nitrophorin 4. FEBS Lett. 2005; 579(24): 5392–5398. PubMed Abstract \| Publisher Full Text\n\nMarti MA, González Lebrero MC, Roitberg AE, et al.: Bond or cage effect: how nitrophorins transport and release nitric oxide. J Am Chem Soc. 2008; 130(5): 1611–1618. PubMed Abstract \| Publisher Full Text\n\nMaes EM, Weichsel A, Andersen JF, et al.: Role of binding site loops in controlling nitric oxide release: structure and kinetics of mutant forms of nitrophorin 4. Biochemistry. 2004; 43(21): 6679–6690. PubMed Abstract \| Publisher Full Text\n\nKnipp M, Taing JJ, He C: Reduction of the lipocalin type heme containing protein nitrophorin -- sensitivity of the fold-stabilizing cysteine disulfides toward routine heme-iron reduction. J Inorg Biochem. 2011; 105(11): 1405–1412. PubMed Abstract \| Publisher Full Text\n\nYang F, Zhang H, Knipp M: A one-residue switch reverses the orientation of a heme b cofactor. Investigations of the ferriheme NO transporters nitrophorin 2 and 7 from the blood-feeding insect Rhodnius prolixus. Biochemistry. 2009; 48(2): 235–241. PubMed Abstract \| Publisher Full Text\n\nMacKerell AD, Bashford D, Bellott M, et al.: All-atom empirical potential for molecular modeling and dynamics studies of proteins. J Phys Chem B. 1998; 102(18): 3586–3616. PubMed Abstract \| Publisher Full Text\n\nPhillips JC, Braun R, Wang W, et al.: Scalable molecular dynamics with NAMD. J Comput Chem. 2005; 26(16): 1781–1802. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJorgensen WL, Chandrasekhar J, Madura JD, et al.: Comparison of simple potential functions for simulating liquid water. J Chem Phys. 1983; 79: 926–935. Publisher Full Text\n\nRenner C, Piehler J, Schrader T: Arginine- and lysine-specific polymers for protein recognition and immobilization. J Am Chem Soc. 2006; 128(2): 620–628. PubMed Abstract \| Publisher Full Text\n\nVarghese S, Yang F, Pacheco V, et al.: Expression, purification, and solid-state NMR characterization of the membrane binding heme protein nitrophorin 7 in two electronic spin states. Biochemistry. 2013; 52(40): 7031–7040. PubMed Abstract \| Publisher Full Text\n\nKnipp M, Yang F, Berry RE, et al.: A Single Mutation in Nitrophorins from Blood-Sucking Insects Flips the Heme Orientation by 180 degrees Around the Cmeso-alpha-Cmeso-gamma Axis. In 9th European Biological Inorganic Chemistry Conference Kozlowski, H., Ed Medimond International Proceedings: Wrocław (Poland). 2008; 15–20. Reference Source\n\nYang F, Knipp M, Shokhireva TK, et al.: 1H and 13C NMR spectroscopic studies of the ferriheme resonances of three low-spin complexes of wild-type nitrophorin 2 and nitrophorin 2 (V24E) as a function of pH. J Biol Inorg Chem. 2009; 14(7): 1077–1095. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndersen JF, Montfort WR: The crystal structure of nitrophorin 2. A trifunctional antihemostatic protein from the saliva of Rhodnius prolixus. J Biol Chem. 2000; 275(39): 30496–30503. PubMed Abstract \| Publisher Full Text\n\nShokhireva TKh, Weichsel A, Smith KM, et al.: Assignment of the ferriheme resonances of the low-spin complexes of nitrophorins 1 and 4 by 1H and 13C NMR spectroscopy: comparison to structural data obtained from X-ray crystallography. Inorg Chem. 2007; 46(6): 2041–2056. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarman EF: Radiation damage in macromolecular crystallography: what is it and why should we care? Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 4): 339–351. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHe C, Fuchs M, Ogata H, et al.: Guanidine-ferroheme coordination in the mutant protein nitrophorin 4(L130R). Angew Chem Int Ed Engl. 2012; 51(18): 4470–4473. PubMed Abstract \| Publisher Full Text\n\nMacedo S, Pechlaner M, Schmid W, et al.: Can soaked-in scavengers protect metalloprotein active sites from reduction during data collection? J Synchrotron Radiat. 2009; 16(Pt 2): 191–204. PubMed Abstract \| Publisher Full Text\n\nBeitlich T, Kühnel K, Schulze-Briese C, et al.: Cryoradiolytic reduction of crystalline heme proteins: analysis by UV-Vis spectroscopy and X-ray crystallography. J Synchrotron Radiat. 2007; 14(Pt 1): 11–23. PubMed Abstract \| Publisher Full Text\n\nHe C, Neya S, Knipp M: Breaking the proximal FeII-NHis bond in heme proteins through local structural tension: lessons from the heme b proteins nitrophorin 4, nitrophorin 7, and related site-directed mutant proteins. Biochemistry. 2011; 50(40): 8559–8575. PubMed Abstract \| Publisher Full Text\n\nMcCleverty JA: Chemistry of nitric oxide relevant to biology. Chemical Review. 2004; 104(2): 403–418. PubMed Abstract \| Publisher Full Text\n\nWyllie GR, Schulz CE, Scheidt WR: Five- to six-coordination in (nitrosyl)iron(II) porphyrinates: effects of binding the sixth ligand. Inorg Chem. 2003; 42(18): 5722–5734. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPraneeth VK, Näther C, Peters G, et al.: Spectroscopic properties and electronic structure of five- and six-coordinate iron(II) porphyrin NO complexes: Effect of the axial N-donor ligand. Inorg Chem. 2006; 45(7): 2795–2811. PubMed Abstract \| Publisher Full Text\n\nDing XD, Weichsel A, Andersen JF, et al.: Nitric Oxide Binding to the Ferri- and Ferroheme States of Nitrophorin 1, a Reversible NO-Binding Heme Protein from the Saliva of the Blood-Sucking Insect, Rhodnius prolixus. J Am Chem Soc. 1999; 121(1): 128–138. Publisher Full Text\n\nWeichsel A, Andersen JF, Roberts SA, et al.: Nitric oxide binding to nitrophorin 4 induces complete distal pocket burial. Nat Struct Biol. 2000; 7(7): 551–554. PubMed Abstract \| Publisher Full Text\n\nBerry RE, Ding XD, Shokhireva TKh, et al.: Axial ligand complexes of the Rhodnius nitrophorins: reduction potentials, binding constants, EPR spectra, and structures of the 4–iodopyrazole and imidazole complexes of NP4. J Biol Inorg Chem. 2004; 9(2): 135–144. PubMed Abstract \| Publisher Full Text\n\nDi Russo NV, Estrin DA, Martí MA, et al.: pH-Dependent conformational changes in proteins and their effect on experimental pKas: the case of Nitrophorin 4. PLoS Comput Biol. 2012; 8(11): e1002761. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartí MA, Estrin DA, Roitberg AE: Molecular basis for the pH dependent structural transition of Nitrophorin 4. J Phys Chem B. 2009; 113(7): 2135–2142. PubMed Abstract \| Publisher Full Text\n\nSwails JM, Meng Y, Walker FA, et al.: pH-dependent mechanism of nitric oxide release in nitrophorins 2 and 4. J Phys Chem B. 2009; 113(4): 1192–1201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKondrashov DA, Montfort WR: Nonequilibrium dynamics simulations of nitric oxide release: comparative study of nitrophorin and myoglobin. J Phys Chem B. 2007; 111(31): 9244–9252. PubMed Abstract \| Publisher Full Text\n\nKnipp M, He C: Nitrophorins: nitrite disproportionation reaction and other novel functionalities of insect heme-based nitric oxide transport proteins. IUBMB Life. 2011; 63(5): 304–312. PubMed Abstract \| Publisher Full Text\n\nKnipp M, Soares RP, Pereira MH: Identification of the native N-terminus of the membrane attaching ferriheme protein nitrophorin 7 from Rhodnius prolixus. Anal Biochem. 2012; 424(1): 79–81. PubMed Abstract \| Publisher Full Text\n\nAbbruzzetti S, He C, Ogata H, et al.: Heterogeneous kinetics of the carbon monoxide association and dissociation reaction to nitrophorin 4 and 7 coincide with structural heterogeneity of the gate-loop. J Am Chem Soc. 2012; 134(24): 9986–9998. PubMed Abstract \| Publisher Full Text\n\nBenabbas A, Ye X, Kubo M, et al.: Ultrafast dynamics of diatomic ligand binding to nitrophorin 4. J Am Chem Soc. 2010; 132(8): 2811–2820. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOliveira A, Allegri A, Bidon-Chanal A, et al.: Kinetics and computational studies of ligand migration in nitrophorin 7 and its Δ1-3 mutant. Biochim Biophys Acta. 2013; 1834(9): 1711–1721. PubMed Abstract \| Publisher Full Text\n\nSchmidtke P, Bidon-Chanal A, Luque FJ, et al.: MDpocket: open-source cavity detection and characterization on molecular dynamics trajectories. Bioinformatics. 2011; 27(23): 3276–3285. PubMed Abstract \| Publisher Full Text\n\nCohen J, Olsen KW, Schulten K: Finding gas migration pathways in proteins using implicit ligand sampling. Methods Enzymol. 2008; 437: 439–457. PubMed Abstract \| Publisher Full Text\n\nSottini S, Abbruzzetti S, Viappiani C, et al.: Determination of microscopic rate constants for CO binding and migration in myoglobin encapsulated in silica gels. J Phys Chem B. 2005; 109(41): 19523–19528. PubMed Abstract \| Publisher Full Text\n\nAbbruzzetti S, Spyrakis F, Bidon-Chanal A, et al.: Ligand migration through hemeprotein cavities: insights from laser flash photolysis and molecular dynamics simulations. Phys Chem Chem Phys. 2013; 15(26): 10686–10701. PubMed Abstract \| Publisher Full Text\n\nMarcelli A, Abbruzzetti S, Bustamante JP, et al.: Following ligand migration pathways from picoseconds to milliseconds in type II truncated hemoglobin from Thermobifida fusca. PLoS One. 2012; 7(7): e39884. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBustamante JP, Abbruzzetti S, Marcelli A, et al.: Ligand uptake modulation by internal water molecules and hydrophobic cavities in hemoglobins. J Phys Chem B. 2014; 118(5): 1234–1245. PubMed Abstract \| Publisher Full Text\n\nKnipp M, Ogata H, Soavi G, et al.: Data of membrane attaching nitric oxide transporter nitrophorin 7. F1000Research. 2015. Data Source" }	[ { "id": "7667", "date": "24 Feb 2015", "name": "Marten H. Vos", "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 work combines crystallography, time-resolved spectroscopy and MD simulations to investigate structure and CO dynamics in a member of the NO transporter NP family, NP7. Whereas the found properties are similar to other NPs, in particular the well-studied NP4, this work is useful as substantial differences had been suggested from previous studies.A few suggestions for improvements:Intro, 2nd par, 3rd phrase suggests that NP1 and NP2 share more identity with NP4 than NP7, which appears not the case for NP2 (Table 2). Please specify sample conditions (path length, concentration) for both types of spectroscopic measurements. MD simulations of NP7 are based on homology modelling from NP4 structures. A comparison of the homology and X-ray models, other than the flipped heme, would be interesting. Admittedly it will not change the order of magnitude of the results, but the problematic gluing of the gap between 1 ns and 20 ns by extrapolating a non-finished double exponential (fit parameters?) can be avoided by determining the effective CO dissociation yield at ~20 ns in the linear excitation regime using an established procedure (Brunori et al., 1973). If I understand well the kinetics in Fig. 9 can be fitted by four exponentials (two <10 ns and two μs-ms), and thus 7 independent free parameters. The model has as many as 11 rate parameters (Table 3, kon is strictly not a microscopic rate constant). This would suggest the model is underdetermined. Please discuss. Do the authors associate T1, T2 and T3 with specific modeled cavities? It is unclear why the rebinding rate is specifically or only compared with TrHbO. In a number of sensor proteins, including CooA, CO rebinding phases on the order of tens and hundreds of picoseconds have been observed.", "responses": [ { "c_id": "1502", "date": "18 Aug 2015", "name": "Cristiano Viappiani", "role": "Author Response", "response": "This work combines crystallography, time-resolved spectroscopy and MD simulations to investigate structure and CO dynamics in a member of the NO transporter NP family, NP7. Whereas the found properties are similar to other NPs, in particular the well-studied NP4, this work is useful as substantial differences had been suggested from previous studies.A few suggestions for improvements:Intro, 2nd par, 3rd phrase suggests that NP1 and NP2 share more identity with NP4 than NP7, which appears not the case for NP2 (Table 2).The sentence has been misinterpreted. The sequence identity is determined for NP1 versus NP4 (88%), and for NP2 versus NP3 (80%). The sentence has been rewritten in the revised version. Likewise, Table 2 has been modified to include the sequence identity with NP3, even though the lack of an X-ray structure impedes to include the data for the structural resemblance.Please specify sample conditions (path length, concentration) for both types of spectroscopic measurements.This relevant information was left out and has been included in the revised manuscript.MD simulations of NP7 are based on homology modelling from NP4 structures. A comparison of the homology and X-ray models, other than the flipped heme, would be interesting.The structural similarity of the protein backbones of the homology and X-ray model has been mentioned in the revised version (Materials and Methods. MD simulations).Admittedly it will not change the order of magnitude of the results, but the problematic gluing of the gap between 1 ns and 20 ns by extrapolating a non-finished double exponential (fit parameters?) can be avoided by determining the effective CO dissociation yield at ~20 ns in the linear excitation regime using an established procedure (Brunori et al., 1973).We agree that this would be the best way to handle the data. Indeed, this procedure was applied to merge the two time regimes in our previous works on trHb from Thermobifida fusca. (Marcelli et al., 2012; Bustamante et al., 2014).However, unlike that case, in the current nanosecond experiments we were unable to reach a low enough laser pulse energy range where multiple photolysis becomes completely negligible, yet with a reasonable signal to noise ratio. This impairs the possibility of using the above mentioned procedure. For this reason we resorted to the approach we adopted for the present experiments. We are aware that the present data normalization procedure may bias to some extent the kinetics, but this effect is expected to be minor, as already mentioned in the original submission. A further statement about this has been added in the “Ligand rebinding kinetics” section.If I understand well the kinetics in Fig. 9 can be fitted by four exponentials (two <10 ns and two μs-ms), and thus 7 independent free parameters. The model has as many as 11 rate parameters (Table 3, kon is strictly not a microscopic rate constant). This would suggest the model is underdetermined. Please discuss.Overall kinetics can be approximated by 4-5 exponential decays. Apparent rates and amplitudes are combinations of microscopic rates, some of which (in the bimolecular phase) are dependent on the reactant concentration. By performing a global analysis of the rebinding kinetics at two different CO concentrations it is possible to remove almost entirely the problem of under-determination of parameters due to their cross correlation.Do the authors associate T1, T2 and T3 with specific modeled cavities?Based on their location, it appears reasonable to associate transient population of the cavities T2 and T3 close to the distal pocket with fast reaction intermediates, whereas the longer lived reaction intermediate T4 is interpreted as rebinding from the cavity in the back tunnel. In the revised manuscript we have indicated putative locations of the cavities with reference to isocontours in Figure 10.Incidentally, we note that the caption to Figure 12 had a typo, where the cavities were erroneously labeled T1, T2 and T3 instead of T2, T3 and T4. This was noted also by Reviewer #3. The scheme in Figure 11 also contained a typo, which is now corrected in the revised version. We apologize for that.It is unclear why the rebinding rate is specifically or only compared with TrHbO. In a number of sensor proteins, including CooA, CO rebinding phases on the order of tens and hundreds of picoseconds have been observed.This remark is certainly relevant. The comparison of the rate constants proposed in the final paragraphs refers to trHbO only because this is the only available set of microscopic rates to compare with.Indeed, several excellent papers have appeared in the literature, reporting apparent lifetimes/rate constants for CO rebinding. We have expanded the discussion in the revised manuscript to include some recent works." } ] }, { "id": "9064", "date": "23 Jun 2015", "name": "Suman Kundu", "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\nMarkus Knipp et al applied a repertoire of biophysical and computational techniques to provide insight into the unique structural features of NP7, a transporter nitrophorin. The study also highlighted the differences with other nitrophorins and unraveled a likely mechanism for NO migration and exit in NP7, which represents advancement in the understanding of nitrophorins. This is a nice, elaborate piece of work that triggers the mind in many ways. The article will read better and assume greater significance if the following issues can be addressed:The second sentence (line 3) in the Abstract should be modified as “Besides its ability to bind to phospholipid membranes, the N-terminus of NP7, a member of the NO transporter nitrophorin family, contains…………..”; otherwise the “other” in that sentence does not make any sense. Moreover, placing NP7 in this sentence also introduces the target protein to the reader better, than mentioning it later. Without such an introduction of the subject of the study, the beginning of the abstract reads pretty abrupt and can be difficult for readers to grasp, unless they come back to it after reading the Introduction.How does one ensure that NO stays bound to heme iron in the crystal structures of NPs, given the photoreduction and the fact that Fe(II) might have altered affinity for NO than Fe(III)? The fact that the electron density could not be fit with a single configuration of NO indicates that NO is only partially bound, which means that the crystals contain a mixture of liganded and unliganded states? Does such mixed population influence conclusions about structural features? A sentence in this regard could be added in the relevant section in the manuscript to dispel doubts. Since the authors solved the crystal structures of NP7 at low (5.8) and high pH (7.8) and in NO-bound form, for MD simulations these structures could be used to model the molecular systems rather than using NP4 structures. Alternatively, the models obtained from NP4 could be compared with the crystal structures. Clarification would help readers.\n\nSince the liganded and unliganded structures are similar, is it safe to assume that the structure of NP7 is not adapted to regulate ligand binding, but rather acts as a reservoir for NO storage suitably endowed with tunnels and cavities for the NO release as and when required, with dependence on pH only? A line of discussion would help. Finally, as a future perspective, a crystal structure of the mutant of NP7 (with extra N-terminal residues deleted), both in unliganded and liganded forms, might help to validate the rational hypothesis laid down in the paper.I congratulate the authors on the wonderful manuscript.", "responses": [ { "c_id": "1503", "date": "18 Aug 2015", "name": "Cristiano Viappiani", "role": "Author Response", "response": "Markus Knipp et al applied a repertoire of biophysical and computational techniques to provide insight into the unique structural features of NP7, a transporter nitrophorin. The study also highlighted the differences with other nitrophorins and unraveled a likely mechanism for NO migration and exit in NP7, which represents advancement in the understanding of nitrophorins. This is a nice, elaborate piece of work that triggers the mind in many ways.We thank this reviewer for his appreciation of our work.The article will read better and assume greater significance if the following issues can be addressed:The second sentence (line 3) in the Abstract should be modified as “Besides its ability to bind to phospholipid membranes, the N-terminus of NP7, a member of the NO transporter nitrophorin family, contains…………..”; otherwise the “other” in that sentence does not make any sense. Moreover, placing NP7 in this sentence also introduces the target protein to the reader better, than mentioning it later. Without such an introduction of the subject of the study, the beginning of the abstract reads pretty abrupt and can be difficult for readers to grasp, unless they come back to it after reading the IntroductionThe underlined words can be added to improve the reability of the sentence.We thank reviewer # 2 for pointing this out, we completely agree that the sentence was missing a reference to the specific protein, which is the subject of the present study.How does one ensure that NO stays bound to heme iron in the crystal structures of NPs, given the photoreduction and the fact that Fe(II) might have altered affinity for NO than Fe(III)? The fact that the electron density could not be fit with a single configuration of NO indicates that NO is only partially bound, which means that the crystals contain a mixture of liganded and unliganded states? Does such mixed population influence conclusions about structural features? A sentence in this regard could be added in the relevant section in the manuscript to dispel doubts.The occupancy of NO ligand was refined to 0.44. The crystal structure in this case contains a mixture of the NO-bound and NO-unbound states, however, this does not influence conclusions. The occupancy of NO has been included in the revised manuscript.Since the authors solved the crystal structures of NP7 at low (5.8) and high pH (7.8) and in NO-bound form, for MD simulations these structures could be used to model the molecular systems rather than using NP4 structures. Alternatively, the models obtained from NP4 could be compared with the crystal structures. Clarification would help readers.We have already addressed this point which was also raised by reviewer #1. As stated above, the structural similarity of the protein backbones of the homology and X-ray model is now mentioned in the revised version (Materials and Methods. MD simulations).Since the liganded and unliganded structures are similar, is it safe to assume that the structure of NP7 is not adapted to regulate ligand binding, but rather acts as a reservoir for NO storage suitably endowed with tunnels and cavities for the NO release as and when required, with dependence on pH only? A line of discussion would help. We have modified the first paragraph of the “Concluding remarks and Future Directions” section to discuss this aspect.Finally, as a future perspective, a crystal structure of the mutant of NP7 (with extra N-terminal residues deleted), both in unliganded and liganded forms, might help to validate the rational hypothesis laid down in the paper.It has been mentioned in the last paragraph of the manuscript that availability of the crystal structure of the protein missing the three amino acids at the N-terminus may help understanding the role of this short sequence." } ] }, { "id": "9066", "date": "27 Jul 2015", "name": "Antonio Cupane", "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 article by Knipp et al. reports the results of an extensive investigation of the nitrophorin isoform NP7. The authors have resolved the X-ray crystallographic structure of NP7 at different pH values and ligation states, measured the ligand rebinding kinetics after laser photolysis of CO bound NP7 at different pH values and performed MD simulations on different NP7 conformations to assess the degree of flexibility of different protein domains and the topology of inner protein cavities that are shown to play a significant role in the binding of ligands to the protein.The presented results are new and significant to our understanding of the structure-function-dynamics relationship in the NP7 nitrophorin. However, the paper would benefit from a more clear discussion of an important aspect that emerges from the results presented by authors: the charge distribution across NP7, which peculiarly determines the ability of this particular nitrophorin isoform to bind negatively charged membranes, is likely to significantly affect both the way NP7 molecules are arranged in the crystalline state and their overall 3D structure in the crystal. As stated by the authors, the surprisingly small differences between the crystal structure of NP7-H2O at pH 5.8 and 7.8 could be due to the overwhelming contribution of electrostatic interactions between neighboring proteins within the crystal. The authors refrained from using their NP7 X-ray structures as starting points for their MD simulations since they admittedly do not account for the conformational change expected to occur upon increase of the pH from 5.8 to 7.8. Overall, these findings seem to indicate that, in the case of NP7 nitrophorin, the structural models retrieved from X-ray crystallography fail to accurately describe the 3D structure of the protein in solution.It would be worth to better clarify and discuss this point in the paper.A further important result of the paper comes from the analysis of ligand rebinding kinetics after photolysis of NP7-CO (presumably at a much lower protein concentration, at which interparticle interactions are negligible). A large fraction of CO ligands rebind geminately in the picosecond time-scale with a biphasic time course. Moreover, nanosecond flash-photolysis experiments clearly show that ligand migration through protein cavities modulates the rebinding process at longer time-scales. Although the main findings are clear already from inspection of the raw data, the procedure used to match the kinetics measured in the sub-nanosecond time-scale with those measured at longer time-scales seems problematic (as also noted by Marten H. Vos in his review of the paper). The detailed procedure used to match the two datasets should be added to the paper. In particular, a supplementary figure showing the two-exponentials fit of the picosecond data and its extrapolation beyond 2 ns, would be helpful in this respect. It would be also useful to have an idea of how much a different choice of the relative scaling between the two datasets would affect the estimation of microscopic rates.Minor points:as noted by Suman Kundu in his review of the paper there are sentences that need to be corrected in the abstract; apart from the second sentence (line 3), also the last sentence (line 17) “Finally, the topological distribution of pockets located around the heme as well as from inner cavities …” clearly needs to be corrected. page 3, last paragraph of the Introduction: “Here we report on seven crystal structures of NP7 …”. Only 6 different crystal structures are described in the paper and in Table 1. the caption of Figure 12 should be corrected in comparison to Figure 11: T1 should be T2.", "responses": [ { "c_id": "1504", "date": "18 Aug 2015", "name": "Cristiano Viappiani", "role": "Author Response", "response": "The article by Knipp et al. reports the results of an extensive investigation of the nitrophorin isoform NP7. The authors have resolved the X-ray crystallographic structure of NP7 at different pH values and ligation states, measured the ligand rebinding kinetics after laser photolysis of CO bound NP7 at different pH values and performed MD simulations on different NP7 conformations to assess the degree of flexibility of different protein domains and the topology of inner protein cavities that are shown to play a significant role in the binding of ligands to the protein.The presented results are new and significant to our understanding of the structure-function-dynamics relationship in the NP7 nitrophorin. However, the paper would benefit from a more clear discussion of an important aspect that emerges from the results presented by authors: the charge distribution across NP7, which peculiarly determines the ability of this particular nitrophorin isoform to bind negatively charged membranes, is likely to significantly affect both the way NP7 molecules are arranged in the crystalline state and their overall 3D structure in the crystal. As stated by the authors, the surprisingly small differences between the crystal structure of NP7-H2O at pH 5.8 and 7.8 could be due to the overwhelming contribution of electrostatic interactions between neighboring proteins within the crystal. The authors refrained from using their NP7 X-ray structures as starting points for their MD simulations since they admittedly do not account for the conformational change expected to occur upon increase of the pH from 5.8 to 7.8. Overall, these findings seem to indicate that, in the case of NP7 nitrophorin, the structural models retrieved from X-ray crystallography fail to accurately describe the 3D structure of the protein in solution.It would be worth to better clarify and discuss this point in the paper.In agreement with the reviewer, we believe that both crystal packing and electrostatic interactions impose a severe constraint on the AB loop, limiting the conformational flexibility of this region, and this likely explains why there seems to be no apparent major difference between the structures at pH 5.8 and 7.8. This has been explicitly stated in the revised version of the manuscript.A further important result of the paper comes from the analysis of ligand rebinding kinetics after photolysis of NP7-CO (presumably at a much lower protein concentration, at which interparticle interactions are negligible). A large fraction of CO ligands rebind geminately in the picosecond time-scale with a biphasic time course. Moreover, nanosecond flash-photolysis experiments clearly show that ligand migration through protein cavities modulates the rebinding process at longer time-scales. Although the main findings are clear already from inspection of the raw data, the procedure used to match the kinetics measured in the sub-nanosecond time-scale with those measured at longer time-scales seems problematic (as also noted by Marten H. Vos in his review of the paper). The detailed procedure used to match the two datasets should be added to the paper. In particular, a supplementary figure showing the two-exponentials fit of the picosecond data and its extrapolation beyond 2 ns, would be helpful in this respect. It would be also useful to have an idea of how much a different choice of the relative scaling between the two datasets would affect the estimation of microscopic rates.We have addressed this issue in response to the similar argument raised by Marten Vos (Reviewer #1). Within the “Ligand rebinding kinetics” section we have added a statement about the effect of a wrong estimate of the extrapolated value of the signal from the subnanosecond kinetics. We provide an estimate of the effect on the rates constants for over/under-estimating the residual absorption after the subnanosecond rebinding phase. As detailed in the text, even a relatively large 10% error results in modest effects on rate constants.Minor points:as noted by Suman Kundu in his review of the paper there are sentences that need to be corrected in the abstract; apart from the second sentence (line 3), also the last sentence (line 17) “Finally, the topological distribution of pockets located around the heme as well as from inner cavities …” clearly needs to be corrected.We have revised the sentence as suggested by Suman Kundupage 3, last paragraph of the Introduction: “Here we report on seven crystal structures of NP7 …”. Only 6 different crystal structures are described in the paper and in Table 1.This wrong statement was revised. We thank this reviewer for pointing that out.the caption of Figure 12 should be corrected in comparison to Figure 11: T1 should be T2.As noted above, we have revised the caption." } ] } ]	1	https://f1000research.com/articles/4-45
https://f1000research.com/articles/4-70/v1	16 Mar 15	{ "type": "Review", "title": "Resources, challenges and way forward in rare mitochondrial diseases research", "authors": [ "Neeraj Kumar Rajput", "Vipin Singh", "Anshu Bhardwaj", "Neeraj Kumar Rajput", "Vipin Singh" ], "abstract": "Over 300 million people are affected by about 7000 rare diseases globally. There are tremendous resource limitations and challenges in driving research and drug development for rare diseases. Hence, innovative approaches are needed to identify potential solutions. This review focuses on the resources developed over the past years for analysis of genome data towards understanding disease biology especially in the context of mitochondrial diseases, given that mitochondria are central to major cellular pathways and their dysfunction leads to a broad spectrum of diseases. Platforms for collaboration of research groups, clinicians and patients and the advantages of community collaborative efforts in addressing rare diseases are also discussed. The review also describes crowdsourcing and crowdfunding efforts in rare diseases research and how the upcoming initiatives for understanding disease biology including analyses of large number of genomes are also applicable to rare diseases.", "keywords": [ "rare disease", "mitochondria", "mitochondrial DNA", "genome variation", "crowdsourcing", "crowdfunding", "semantic-web", "next generation sequencing" ], "content": "Introduction\n\nMitochondria are organelles present in every cell of the body (except red blood cells) and generate almost all of the energy needed by the cells to grow and sustain life. In addition to adenosine triphosphate (ATP) generation, mitochondria are involved in a large number of specialized functions in major cellular pathways including apoptosis, urea cycle, pyrimidine biosynthesis, heme synthesis, etc1. The proteins that take part in these pathways are encoded both by mitochondrial DNA (mtDNA) and nuclear DNA (nDNA)2. mtDNA encodes only a limited number of genes (37) which code for 13 proteins, two rRNAs and 22 tRNAs3. These proteins code for four respiratory complexes of the oxidative phosphorylation (OXPHOS) system. The only non-coding segment of mtDNA is the displacement loop (D-loop, 1121 bp) that contains the origin of replication of the H-strand (OH) and the promoters for L and H-strand transcription. The diseases related to mitochondrial dysfunction are due to mutations in both the mtDNA and nDNA encoded components. Genetically, mitochondrial diseases are characterized as (i) those with sporadic or maternally inherited mtDNA mutations, (ii) those with abnormalities with Mendelian transmission of the trait, i.e., disorders believed to be due to mutations in nuclear genes that control mitochondrial biogenesis, and (iii) those that are caused by nuclear genes but are misinterpreted as mitochondrial based on the biochemical findings4,5. mtDNA point mutations which can either be maternally inherited or generated somatically have been associated with many diseases like A3243G for MELAS, A8344G for MERRF, T8993G for NARP, etc. Also, there appears to be a class of slightly deleterious mutations that modify the risks of developing certain complex diseases or trait6.\n\nHuman mtDNA and nDNA mutations causing mitochondrial dysfunction are implicated in a broad spectrum of diseases affecting various tissues like brain, heart, liver, skeletal muscles, etc7. The clinical symptoms of the disease depend on the cell type affected and range from loss of motor control, muscle weakness, cardiac disease to visual or hearing loss, etc.8. Given that mitochondria are involved in a large number of cellular pathways, it is always challenging to correlate the exact causative role of genome variation with the observed phenotype. Additionally, the varying spectrum of disease symptoms is a major deterrent in early disease diagnosis. This is clear from the fact that with over 5000 mtDNA variation reported across databases, pathogenicity assignments for most of the variation is only limited to the association with the phenotype without any conclusive evidence on its causative role. While there are a large number of resources available on various aspects of human mtDNA, a major bottleneck is the lack of documentation of genomic variation data across populations with clinical details to evaluate these variations for disease association. The need of the hour is to curate these resources using standards for data exchange over the web as also using standard ontologies for data analysis across platforms.\n\nMost of the mitochondrial diseases have rare occurrence in the population and hence are termed as ‘Rare diseases’. As of now there are nearly 7000 rare diseases reported worldwide arising from mutations in either nDNA or mtDNA. It is estimated that nearly 300 million people in the world are affected by rare diseases. In India alone there are nearly 70 million people diagnosed with rare diseases (http://www.rarediseasesindia.org/), despite the fact that there are no standard diagnostic tests available for most rare diseases. In addition, the definition of rare diseases vary, for e.g., the European Union considers diseases to be rare when they affect no more than 5 in 10,000 people, while the Unites States of America (USA) consider a disease to be rare when affecting fewer than 200,000 people. In Asia, the threshold is 1 in 10,000. However, the prevalence of most mitochondrial diseases is not known9,10.\n\nThis review elaborates on the available resources, the bottlenecks in rare disease research and proposes innovative ways to address these challenges. There are many reviews discussing the challenges involved in establishing genotype-phenotype correlations with mitochondrial dysfunction given its involvement in multiple pathways in a spatio-temporal manner3,5,8,11. Here we review our current understanding of the rare disease initiatives and efforts that are ongoing globally along with specific focus on mitochondrial disease resources.\n\nThe first human DNA to be completely sequenced was the human mitochondrial DNA (16569 base pairs) in 198112,13. Given the relatively small size and absence of repeats, sequencing and assembling the mitochondrial human genome was not as challenging and difficult as sequencing the human nuclear genome using the Sanger Sequencing technology. However, mitochondrial genome sequencing has its own unique problems given the high mutation rate and high levels of heteroplasmy. Precise determination of the levels of heteroplasmy is crucial since the level of heteroplasmy determines both the penetrance and severity of expression of some mitochondrial diseases. The next generation sequencing (NGS) technologies like Virtual terminator sequencing (Illumina), Pyrosequencing (Roche) and SOLID have allowed to overcome these limitations by providing massive parallelization, high coverage, high accuracy as compared to Sanger sequencing. Specific protocols, including long range PCR with mitochondria specific primers, and algorithms for reference based and de novo assembly have been developed to sequence mitochondrial DNA using NGS technologies14. NGS-based clinical targeted gene assay for the mitochondrial genome and 108 selected nuclear genes associated with mitochondrial disorders have also been designed to facilitate the analysis and understanding of nuclear and mitochondrial variations in mitochondrial diseases15. These emerging technologies offer an excellent opportunity to further dissect the molecular basis of disease manifestation4. With an increasing number of individuals that may be genetically screened across different populations, excellent datasets may be available to explore the genetic basis of disease. The genomics data along with clinical and biochemical profiles may also be used to identify disease biomarkers with high sensitivity and specificity, which is a major challenge in diagnosing mitochondrial dysfunction.\n\nOver the years, a large number of web-based resources have been developed on various aspects of mitochondrial diseases, most of them focusing on the data from mtDNA. Some of these include, MitoMap, a database on human mitochondrial variation6, MitoLSDB, the largest curated data on mtDNA variation with phenotype using LOVD16, MitoCarta, a resource on mitochondrial proteins based on localization17, MitoMiner, a mitochondrial protein identification system based on multiple evidences18,19, MitoBreak, a curated dataset on mtDNA rearrangements20, HmtDB, an online resource for data on mitochondrial genome sequences annotated with population and variation data21, Mitochondrial Database (mitoDB), the mitochondrial database on clinical features seen in mitochondrial diseases22, to name a few. Analysis pipelines and platforms have also been developed, including the MtSNPscore which assesses the role of variation in context of disease association using a combined evidence approach23, Mit-o-matic, an analysis pipeline for clinical evaluation of mitochondrial variations from the NGS datasets24. More recently, the United Mitochondria Disease Foundation (UMDF) (http://www.umdf.org/site/c.8qKOJ0MvF7LUG/b.7929671/k.BDF0/Home.htm) along with the National Institute of Child Health and Human Development (NICHD) (http://www.nichd.nih.gov/Pages/index.aspx) launched the Mitochondrial Disease Sequence Data Resource (MSeqDR) Consortium. The goals of this consortium is to facilitate deposition, curation, annotation and integrated analysis of genomic data for mitochondrial diseases for clinical and research communities25. The list of various mitochondrial resources may be seen in Table 1.\n\nThere are a large number of resources developed both for mitochondrial community and rare diseases community. The first section lists the resources available on mitochondrial diseases including Support and Advocacy groups, databases and analysis pipelines#, research& and patient networks$. The second section lists resources on rare diseases.\n\nGlobally, attempts have also been made to systematically address the problem of rare diseases by establishing focused programs and consortia-based approaches. Therapeutics of Rare and Neglected Diseases (TRND) (http://www.ncats.nih.gov/research/rare-diseases/trnd/trnd.html), a program led by the National Center for Advancing Translational Sciences (NCATS) (http://www.ncats.nih.gov/), supports the development of potential treatments for rare and neglected diseases to first-in-human trials. This approach provides a de-risking strategy making the downstream development efforts commercially viable. TRND also supports the pre-clinical studies including medicinal chemistry optimization, drug metabolism and pharmacokinetics, toxicology formulation, and others studies required to file Investigational New Drug (IND) application for regulatory approvals. The other initiatives by NCATS for rare diseases include Office of Rare Disease Research (ORDR) (http://www.orphadata.org/cgi-bin/inc/ordo_orphanet.inc.php) which coordinates a large number of collaborative research efforts towards rare diseases including support to institutes and centers, managing patient registry, human bio-specimen repository, to name a few major activities. Rare Diseases Clinical Research Network (RDCRN) (http://rarediseases.info.nih.gov/research/pages/41/rare-diseases-clinical-research-network), from ORDR, focuses on advancing medical research on rare diseases by facilitating collaboration, study enrollment and data sharing. It also connects scientists from multiple disciplines across various clinical sites globally to work with patient advocacy groups. The North American Mitochondrial Disease Consortium (NAMDC) (http://www.rarediseasesnetwork.org/namdc/), a part of RDCRN, specially works towards collecting information from mitochondrial disease patients in a clinical patient registry. In addition to periodically updating the patients on mitochondrial diseases, NAMDC also helps researchers to identify and recruit patients for future studies. Data generated as part of the various initiatives involving patient information are managed by the NCATS Global Research Patient Registry Data Repository (GRDR) (http://www.ncats.nih.gov/research/rare-diseases/grdr/grdr.html). This is a web-based resource that aggregates de-identified patient information across many registries and provides a Globally Unique Identifier (GUID) to each patient data. GUID allows for patient follow-up across different registries, diseases, studies and countries and also ensures that clinical information is also mapped to bio-specimen datasets. Bridging interventional Development Gaps (BrIDs) (http://www.ncats.nih.gov/research/rare-diseases/bridgs/bridgs.html) is a division of NCATS for pre-clinical innovation towards the development of new therapeutic agents both for common and rare diseases. A recent perspective on the NCATS TRND and BrIDGs programs highlights the role of team effort where academia, biotech and pharma industries, patient communities, advocacy groups, regulators, and government support, all are needed to navigate through the translational Valley of Death26. National Organization for Rare Disorders (NORD) (http://www.rarediseases.org/) is a non-profit organization with the aim to improve the lives of all people affected by rare diseases. The services offered by NORD include identification, treatment and cure of rare disorders through advocacy, research and programs of education. Similarly, Global genes is a non-profit organization for patient advocacy and work towards building awareness and providing connections and resources for rare disease patients (http://globalgenes.org/). Similar efforts in the European subcontinent have led to the establishment of Orphanet, a consortium of 40 countries coordinated by the French INSERM team and which hosts a reference portal for rare diseases and orphan drugs. Orphanet hosts a directory of information of expert clinics, medical labs, clinical trials, patient organizations, etc. (http://www.orpha.net/consor/cgi-bin/index.php). The joint effort of the European Commission and the NIH established IRDiRC, the International Rare Diseases Research Consortium, in 2011. This international consortium of researchers and organizations aims to deliver 200 therapies and means to diagnose most rare diseases by 2020, and as per their reports, the targets are being delivered (http://www.irdirc.org/). The list of various rare diseases resources may be seen in Table 1.\n\nAs mentioned earlier, there are no clear estimates of the number of mitochondrial rare diseases. In order to get an approximation, the list of 6537 rare diseases was taken from Global genes website. To find out which rare disease is also a mitochondrial disease, a list of mitochondrial diseases was taken from the Mitochondrial Database (mitoDB) (51)22. In addition, information on mitochondrial diseases is also referred from UMDF (41) and MITOMAP (51)6. On comparing rare disease list with the mitochondrial diseases lists, 18 rare mitochondrial diseases were identified. It is important to mention here that since an exact match was performed we could have missed diseases for which abbreviation or synonyms are used. This also highlights that standard ontologies and terms are not systematically followed making data intractable for automated data analysis. We further checked the prevalence of these 18 rare diseases from the Rare Diseases India (http://www.rarediseasesindia.org/) and Orphanet portals. As may be seen on Table 2, data are available for few diseases only.\n\nThe table provides a list of diseases that cause mitochondrial dysfunction and are also reported rare diseases.\n\n* indicates birth prevalence\n\nIt is evident from a simple search in the clinical trials registry (https://clinicaltrials.gov/) that only 0.2% of clinical trials ever done, ongoing, terminated or planned are for rare diseases. This clearly indicates that the major challenge for rare diseases research is the cost involved in research and development given the poor return on investment. It is also well known that clinical trials are prohibitively expensive even for common diseases where finding expert clinicians and acceptable number of patients is not as challenging as for rare diseases. These issues are addressed by de-risking research, developing platforms for sharing patient data, generating centralized patient registries and offering various incentives for making the development commercially viable. All these aspects have been discussed earlier. Here we discuss the various models that have been attempted to bridge the last mile of developing novel therapeutics.\n\nOne of the major challenges in the treatment of rare diseases is clinical trials patient recruitment. To overcome this challenge, patient advocacy groups and websites like PatientsLikeMe are turning out to be a major game changer. PatientsLikeMe is a patient powered research network that allows people to connect and share their experience with other people having the same disease or condition. PatientsLikeMe started its first online community for Amyotrophic Lateral Sclerosis (ALS)27–29 in 2006 where participants could ask specific questions about the treatment options and what to expect to fellow users (http://www.patientslikeme.com/). In addition, the patients also got involved in experimenting with drugs that have not received regulatory approval. Thus data generated by these self-reporting platforms can be used to establish the efficacy and safety of a compound for rare diseases. The self-reported data help provide evidence to support or refute treatment outcomes. Despite several limitations of self-reported data, like unmeasured covariates, data reliability and controlled settings, these approaches are promising as has been shown with the quality of data gathered with smartphone games which were comparable to data obtained from controlled laboratory environments30,31. In 2011, the company expanded its scope and allowed any patient with any condition to join the community. Currently there are more than 300,000 members registered on the PatientLikeMe community with more than 2000 health conditions (http://en.wikipedia.org/wiki/PatientsLikeMe). Besides, interactions of the patient advocates from the rare diseases group with major regulatory bodies has led to expedite review and approval process for rare disease treatments (http://www.forbes.com/sites/medidata/2014/09/25/rare-disease-patient-voices-bring-change-to-the-clinical-trials-process/).\n\nUsing another interesting strategy, Pfizer Inc. used a web-based interactive platform to evaluate the efficacy of a drug (tolterodine tartrate extended release capsules) to treat overacting bladder. This project called REMOTE is a phase IV trial under an Investigational New Drug (IND) application32. The participants for the trial were recruited through an interactive web-based platform from one clinical site overseen by physician33. Despite poor patient participation, the study reports that the trial outcomes are consistent with the results from the conventional trial. The observations from REMOTE is a learning experience for the trial community on the challenges involved at various levels from patient understanding to technical issues. These learnings are used to conduct another trial in Europe, REMOTE2.0, to overcome the bottlenecks faced with the first attempt. These initiatives are of immense significance in establishing the robustness of trial strategies where patient recruitment is a challenge. These mobile technologies also allow patients with complex disabilities to participate in trials. The mobile technologies in form of wearable sensors also make the trial monitoring and patient diagnosis at different time points more affordable34. Subsequent efforts to share patient data also have a positive impact on patient recruitment.\n\nIt is also proposed that the N-of-1 trial method may be used to evaluate new treatments. Under this model a single patient is the entire trial and the treatment outcome is measured at different time points for pre-treatment, treatment duration and post-treatment35. If the treatment outcome is measurable and the patient is cured, the treatment may be considered effective.\n\nThus, the crowdsourcing strategy is an attractive tool for patient recruitment as it is affordable and time effective. Likewise, crowdfunding is also proposed as a viable option for financing various aspects of rare diseases research. Crowdfunding initiatives not only generate awareness about rare diseases but also lead to funding for supporting research activities. One such example is the ALS Ice Bucket Challenge leading to 1.2 million videos on Facebook and 2.2 million tweets in a matter of 3 months, which raised over $100 million of donations (http://en.wikipedia.org/wiki/Ice_Bucket_Challenge). Through innovative ideas like ALS Ice Bucket Challenge it is possible to raise funds and also increase public awareness about the rare diseases. The Rare Genomics Institute (RGI) (http://raregenomics.org/) is an international non-profit providing expert network and an online crowdfunding mechanism to assist families pursue personalized research projects for diseases which would not be studied otherwise. RGI offers sequencing facilities for diagnosis, expert guidance on sequencing and systematic interpretation of the data generated. A recently published success story is that of a 3-year old girl showing symptoms of involuntary eye movement, small-sized head, involuntary muscle contraction, development delay and progressive decline. Whole exome sequencing of the family revealed a novel mutation that causes mental retardation and severe developmental delays. Through the RGI platform, $5000 was raised in 50 days to carry out the genome sequencing in the girl and her family, which led to identification of the novel mutation36. These examples underline the potential of crowdfunding in addressing scientifically challenging and socially important problems.\n\nWith the increasing disease burden, the need of the hour is to establish new models for ensuring that clinical trials are sustainable and transparent. Clinical trials are generally demanding on both human and other resources. It is not possible to apply the conventional methods to rare disease trials given the shear costs involved. To overcome these bottlenecks, efforts are being made lately to reverse the approach of rare disease patient recruitments for clinical trials by taking the trials to the patients and not vice versa. This approach demands establishing patient-centric sites which is again challenging given their extremely sparse distribution. Telemetry innovations may offer solutions to this problem, which allows health monitoring of the patients remotely. This model is not just suitable for rare diseases but is also applicable to common diseases trials given that many clinical trial sites never recruit patients (http://www.clinicalleader.com/doc/how-rare-disease-know-how-can-shape-big-pharma-clinical-trials-0001).\n\nOpen access platforms which allow for data integration and exchange of ideas to facilitate the process of bringing new therapeutic interventions and diagnostic methods to the market are needed to drive sustained research and development for rare diseases. It is also imperative to synchronize the efforts at all levels of research, diagnosis and regulatory guidelines for evidence-based clinical decision-making. As part of the platform, it is also important to involve regulatory agencies from very early on in the discovery pipeline so that the progression through development and approvals is quicker and more affordable, which is the bottleneck with rare diseases. One strategy which seems to work well is to apply approaches of drug repurposing where it is crucial to understand the mechanism of disease manifestation and subsequently applying various modeling approaches to find new indications of existing drugs in the market (http://www.fda.gov/ForIndustry/DevelopingProductsforRareDiseasesConditions/HowtoapplyforOrphanProductDesignation/ucm216147.htm). This strategy will be even more successful where one can systematically list the phenotypic parameters provided by patient groups in close collaboration with clinicians.\n\nThere is a need for innovation in technology that allows for interfacing the different stakeholders, namely, researchers, clinicians and patient groups. Such resources will go a long way in future to allow patients to perform diagnostic procedures with greater accuracy and help clinicians identify the best possible route for therapeutic interventions. This is only possible if the participating members use standard terms to share the data. Various ontologies have been developed over years for capturing function of genes like the Gene Ontology37, PAGE-OM38 and VariO39 for capturing features of variation, UMLS which includes MeSH40, RxNORM41 and SNOMED CT42 for capturing clinical details and Human Phenotype Ontology43 for phenotype data, to name a few. These ontologies are meant to function as a common set of vocabulary that needs to be shared on a community collaborative platform. The semantic network of these terms will allow deciphering and interpreting novel patterns in understanding disease biology. Such a semantic-based platform will be amenable for plugging in data and new methods with increasing understanding of disease biology. It will also ensure that the data generated as part of negative results are also shared systematically with scientific community. Orphanet Rare Disease Ontology (ORDO) (http://www.orphadata.org/cgi-bin/inc/ordo_orphanet.inc.php) is an effort to achieve these goals. ORDO offers definitions, classification of rare diseases, gene-disease relations, epidemiological data and connections with other terminologies like MeSH, UMLS (http://www.nlm.nih.gov/research/umls/), OMIM44, UniProtKB45, HGNC46, ensembl47, Reactome48, etc. It is imperative that existing resources utilize these ontologies for data sharing allowing for data interoperability across different platforms. Figure 1 illustrates the scope of the proposed integrative platform describing the scientific challenges involved in establishing genotype-phenotype correlations and how the existing resources and community collaborations may be converged towards a systems level understanding of the disease biology.\n\n(A) This section describes the challenges involved in dissecting the impact of genomic variation in disease association. As shown, mitochondria are involved in a large number of cellular processes including energy metabolism. Genes encoded by mitochondrial and nuclear genomes carry out these functions. The protein subunits of the complexes of the electron transport chain are encoded both by mtDNA (red) and nDNA (green). In order to understand the complex genotype-phenotype correlation it is imperative to identify the molecular interactions at systems level (protein-protein interaction network, metabolic network, signaling network, regulatory network). It is important to curate the information needed to generate these networks from literature and existing resources. (B) The resources currently available for mitochondrial dysfunction include databases, web servers and analysis pipelines as shown. For generating systems level models, these resources may be integrated systematically. (C) There are various ongoing efforts involving researchers, clinicians and or patient groups. It is proposed that a community collaborative open access platform is a must to interface these communities. In order to establish such a platform that allows geographically different communities to work together, globally accepted ontologies in a language independent representation are needed.\n\n‘Internet of DNA’ is listed as one of the top 15 breakthrough technologies of 2015 (http://www.technologyreview.com/featuredstory/535016/internet-of-dna/?utm_campaign=newsletters&utm_source=newsletter-weekly-biomedicine&utm_medium=email&utm_content=20150224). The technology needed to harness the power of genomics lies in comparing the genetic information from a large number of individuals with medical records. This currently is a huge challenge, partly because of the technical reasons of moving petabytes of data across different labs, but especially due to the privacy issues surrounding patient information. Both these issues should be addressed to ensure that the ever-increasing amount of genomic and clinical data piling up in laboratories and hospitals are utilized optimally. Upcoming initiatives like the MatchMaker Exchange are aiming to bring the genotype and phenotype data together on a common platform (http://matchmakerexchange.org/). Global Alliance for Genomics and Health also known as GA4GH is an organization which provides protocols, APIs and file formats for effective and responsible sharing of genomic and clinical data (http://genomicsandhealth.org/). The organization goal is to overcome challenges likes ethics and privacy involved with sharing of genomics data and to accelerate the potential of genomic medicine for advancement of human health. As discussed before, the need of the hour is to let the patients decide on who will access their data and how these may be used. This is only possible if the information generated using patient samples is made available in real-time. It is also important that other components of this major collaborative strategy are also part of a community platform that allows for gated access to patient data with patients deciding how and with whom their data should be shared.\n\nIn the changing paradigms of disease treatment, a very recent approval is made by Britain on Mitochondrial donation (http://www.theguardian.com/politics/2015/feb/24/uk-house-of-lords-approves-conception-of-three-person-babies). In this approach in vitro fertilization (IVF) technique is used with biological material coming from three parents, mother and father (contributing 98.8% genetic material) and a female donor (contributing 0.2% genetic material). This three-parent IVF approval in UK has received mixed reactions and only time will address the concern of the long-term implications of the same.\n\n\nConclusions\n\nRare diseases affect over 300 million people globally, however the true burden of these diseases on human health remains to be determined. Rare genetic variants are disease causing and lead to a personalized disease manifestation. Thus, it is time to review the disease definition considering both the molecular mechanisms involved and environmental factors leading to differential phenotypes. This will allow for a better understanding of both rare and common diseases. On the other end, a paradigm shift in drug discovery and development is also needed to translate the effort in understanding disease mechanisms to identify potential therapeutic routes. Newer models and platforms that allow involvement of patient communities in research and development is also expected to offer solutions to patients suffering from rare diseases who may then benefit from appropriate treatment options. Community collaborative approaches for research and funding offer an unprecedented opportunity for making new discoveries and translating to therapeutic interventions.", "appendix": "Author contributions\n\n\n\nAB conceived and designed the article outline. NKR and AB worked on the tables and figures. AB, NKR and VS wrote the manuscript. All authors have read and approved the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nNKR acknowledges the CSIR-GENESIS project for providing research fellowship.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nPalmieri F: Diseases caused by defects of mitochondrial carriers: a review. Biochim Biophys Acta. 2008; 1777(7–8): 564–578. PubMed Abstract \| Publisher Full Text\n\nBoengler K, Heusch G, Schulz R: Nuclear-encoded mitochondrial proteins and their role in cardioprotection. Biochim Biophys Acta. 2011; 1813(7): 1286–94. PubMed Abstract \| Publisher Full Text\n\nTaylor RW, Turnbull DM: Mitochondrial DNA mutations in human disease. Nat Rev Genet. 2005; 6(5): 389–402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGershoni M, Levin L, Ovadia O, et al.: Disrupting mitochondrial-nuclear coevolution affects OXPHOS complex I integrity and impacts human health. Genome Biol Evol. 2014; 6(10): 2665–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcFarland R, Taylor RW, Turnbull DM: Mitochondrial disease--its impact, etiology, and pathology. Curr Top Dev Biol. 2007; 77: 113–55. PubMed Abstract \| Publisher Full Text\n\nKogelnik AM, Lott MT, Brown MD, et al.: MITOMAP: a human mitochondrial genome database. Nucleic Acids Res. 1996; 24(1): 177–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWallace DC, Fan W: The pathophysiology of mitochondrial disease as modeled in the mouse. Genes Dev. 2009; 23(15): 1714–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRussell O, Turnbull D: Mitochondrial DNA disease-molecular insights and potential routes to a cure. Exp Cell Res. 2014; 325(1): 38–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGorman GS, Schaefer AM, Ng Y, et al.: Prevalence of nuclear and mtDNA mutations related to adult mitochondrial disease. Ann Neurol. 2015. PubMed Abstract \| Publisher Full Text\n\nSchaefer AM, McFarland R, Blakely EL, et al.: Prevalence of mitochondrial DNA disease in adults. Ann Neurol. 2008; 63(1): 35–9. PubMed Abstract \| Publisher Full Text\n\nGreaves LC, Reeve AK, Taylor RW, et al.: Mitochondrial DNA and disease. J Pathol. 2012; 226(2): 274–86. PubMed Abstract \| Publisher Full Text\n\nAnderson S, Bankier AT, Barrell BG, et al.: Sequence and organization of the human mitochondrial genome. Nature. 1981; 290(5806): 457–65. PubMed Abstract \| Publisher Full Text\n\nAndrews RM, Kubacka I, Chinnery PF, et al.: Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nat Genet. 1999; 23(2): 147. PubMed Abstract \| Publisher Full Text\n\nSosa MX, Sivakumar IK, Maragh S, et al.: Next-generation sequencing of human mitochondrial reference genomes uncovers high heteroplasmy frequency. PLoS Comput Biol. 2012; 8(10): e1002737. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDames S, Chou LS, Xiao Y, et al.: The development of next-generation sequencing assays for the mitochondrial genome and 108 nuclear genes associated with mitochondrial disorders. J Mol Diagn. 2013; 15(4): 526–34. PubMed Abstract \| Publisher Full Text\n\nShamnamole K, Jalali S, Scaria V, et al.: MitoLSDB: a comprehensive resource to study genotype to phenotype correlations in human mitochondrial DNA variations. PLoS One. 2013; 8(4): e60066. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPagliarini DJ, Calvo SE, Chang B, et al.: A mitochondrial protein compendium elucidates complex I disease biology. Cell. 2008; 134(1): 112–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith AC, Blackshaw JA, Robinson AJ: MitoMiner: a data warehouse for mitochondrial proteomics data. Nucleic Acids Res. 2012; 40(Database issue): D1160–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith AC, Robinson AJ: MitoMiner, an integrated database for the storage and analysis of mitochondrial proteomics data. Mol Cell Proteomics. 2009; 8(6): 1324–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDamas J, Carneiro J, Amorim A, et al.: MitoBreak: the mitochondrial DNA breakpoints database. Nucleic Acids Res. 2014; 42(Database issue): D1261–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRubino F, Piredda R, Calabrese FM, et al.: HmtDB, a genomic resource for mitochondrion-based human variability studies. Nucleic Acids Res. 2012; 40(Database issue): D1150–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScheibye-Knudsen M, Scheibye-Alsing K, Canugovi C, et al.: A novel diagnostic tool reveals mitochondrial pathology in human diseases and aging. Aging (Albany NY). 2013; 5(3): 192–208. PubMed Abstract \| Free Full Text\n\nBhardwaj A, Mukerji M, Sharma S, et al.: MtSNPscore: a combined evidence approach for assessing cumulative impact of mitochondrial variations in disease. BMC Bioinformatics. 2009; 10(Suppl 8): S7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVellarikkal SK, Dhiman H, Joshi K, et al.: mit-o-matic: A comprehensive computational pipeline for clinical evaluation of mitochondrial variations from next-generation sequencing datasets. Hum Mutat. 2015. PubMed Abstract \| Publisher Full Text\n\nFalk MJ, Shen L, Gonzalez M, et al.: Mitochondrial Disease Sequence Data Resource (MSeqDR): A global grass-roots consortium to facilitate deposition, curation, annotation, and integrated analysis of genomic data for the mitochondrial disease clinical and research communities. Mol Genet Metab. 2015; 114(3): 388–396. PubMed Abstract \| Publisher Full Text\n\nFagnan DE, Yang NN, McKew JC, et al.: Financing translation: Analysis of the NCATS rare-diseases portfolio. Sci Transl Med. 2015; 7(276): 276ps3. PubMed Abstract \| Publisher Full Text\n\nWicks P, Vaughan TE, Massagli MP, et al.: Accelerated clinical discovery using self-reported patient data collected online and a patient-matching algorithm. Nat Biotechnol. 2011; 29(5): 411–4. PubMed Abstract \| Publisher Full Text\n\nFrost JH, Massagli MP: Social uses of personal health information within PatientsLikeMe, an online patient community: what can happen when patients have access to one another's data. J Med Internet Res. 2008; 10(3): e15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrownstein CA, Brownstein JS, Williams DS 3rd, et al.: The power of social networking in medicine. Nat Biotechnol. 2009; 27(10): 888–90. PubMed Abstract \| Publisher Full Text\n\nBrown HR, Zeidman P, Smittenaar P, et al.: Crowdsourcing for cognitive science--the utility of smartphones. PLoS One. 2014; 9(7): e100662. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBhardwaj A: Harnessing the Crowd for Neurology Research. Sci Transl Med. 2014; 6(250): 250ec141. Publisher Full Text\n\nOrri M, Lipset CH, Jacobs BP, et al.: Web-based trial to evaluate the efficacy and safety of tolterodine ER 4 mg in participants with overactive bladder: REMOTE trial. Contemp Clin Trials. 2014; 38(2): 190–7. PubMed Abstract \| Publisher Full Text\n\nPfizer. Web-based Methodology Trial to Evaluate the Efficacy and Safety of Tolterodine ER in Subjects With Overactive Bladder (REMOTE). 2011. Reference Source\n\nPatel S, Park H, Bonato P, et al.: A review of wearable sensors and systems with application in rehabilitation. J Neuroeng Rehabil. 2012; 9: 21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLillie EO, Patay B, Diamant J, et al.: The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med. 2011; 8(2): 161–173. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIsakov O, Lev D, Blumkin L, et al.: Crowdfunding effort identifies the causative mutation in a patient with nystagmus, microcephaly, dystonia and hypomyelination. J Genet Genomics. 2015; 42(2): 79–81. PubMed Abstract \| Publisher Full Text\n\nGene Ontology Consortium. The Gene Ontology project in 2008. Nucleic Acids Res. 2008; 36(Database issue): D440–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrookes AJ, Lehvaslaiho H, Muilu J, et al.: The phenotype and genotype experiment object model (PaGE-OM): a robust data structure for information related to DNA variation. Hum Mutat. 2009; 30(6): 968–77. PubMed Abstract \| Publisher Full Text\n\nVihinen M: Variation Ontology for annotation of variation effects and mechanisms. Genome Res. 2014; 24(2): 356–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRogers FB: Medical subject headings. Bull Med Libr Assoc. 1963; 51: 114–6. PubMed Abstract \| Free Full Text\n\nLiu S, Wei Ma, Moore R, et al.: RxNorm: prescription for electronic drug information exchange. IT Professional. 2005; 7(5): 17–23. Publisher Full Text\n\nCote RA, Robboy S: Progress in medical information management. Systematized nomenclature of medicine (SNOMED). JAMA. 1980; 243(8): 756–62. PubMed Abstract \| Publisher Full Text\n\nKohler S, Doelken SC, Mungall CJ, et al.: The Human Phenotype Ontology project: linking molecular biology and disease through phenotype data. Nucleic Acids Res. 2014; 42(Database issue): D966–74. 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): D514–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoutet E, Lieberherr D, Tognolli M, et al.: UniProtKB/Swiss-Prot. Methods Mol Biol. 2007; 406: 89–112. PubMed Abstract \| Publisher Full Text\n\nGray KA, Yates B, Seal RL, et al.: Genenames.org: the HGNC resources in 2015. Nucleic Acids Res. 2015; 43(Database issue): D1079–85. PubMed Abstract \| Publisher Full Text\n\nCunningham F, Amode MR, Barrell D, et al.: Ensembl 2015. Nucleic Acids Res. 2015; 43(Database issue): D662–9. PubMed Abstract \| Publisher Full Text\n\nCroft D, Mundo AF, Haw R, et al.: The Reactome pathway knowledgebase. Nucleic Acids Res. 2014; 42(Database issue): D472–7. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8964", "date": "26 Jun 2015", "name": "Xin Qi", "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 very good and informative review paper, which summarized nicely on the resources on research of mitochondrial diseases ranging from clinical setting to basic science.", "responses": [] }, { "id": "9177", "date": "01 Jul 2015", "name": "Vanniarajan Ayyasamy", "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 article really serves as a good resource for researchers, clinicians and also patients.I would suggest the authors to compare the available databases of the mitochondrial research and may provide a better summary.", "responses": [ { "c_id": "1514", "date": "13 Aug 2015", "name": "Anshu Bhardwaj", "role": "Author Response", "response": "We thank the reviewer for encouraging comments. The scope of the current review was to provide an end-to-end view of the resources, challenges and way forward for addressing translational challenges in mitochondrial diseases. The comparison of databases of the mitochondrial research needs a separate review as the parameters for comparison need to be comprehensive and the information related to each resource should be discussed in details. The authors are working on the same and will share the details as soon as it is ready." } ] } ]	1	https://f1000research.com/articles/4-70
https://f1000research.com/articles/4-14/v1	16 Jan 15	{ "type": "Method Article", "title": "Development of an image-based screening system for inhibitors of the plastidial MEP pathway and of protein geranylgeranylation", "authors": [ "Michael Hartmann", "Elisabet Gas-Pascual", "Andrea Hemmerlin", "Michel Rohmer", "Thomas J. Bach", "Michael Hartmann", "Elisabet Gas-Pascual", "Andrea Hemmerlin", "Michel Rohmer" ], "abstract": "We have recently established an in vivo visualization system for the geranylgeranylation of proteins in a stably transformed tobacco BY-2 cell line, which involves expressing a dexamethasone-inducible GFP fused to the prenylable, carboxy-terminal basic domain of the rice calmodulin CaM61, which naturally bears a CaaL geranylgeranylation motif (GFP-BD-CVIL). By using pathway-specific inhibitors it was demonstrated that inhibition of the methylerythritol phosphate (MEP) pathway with oxoclomazone and fosmidomycin, as well as inhibition of protein geranylgeranyl transferase type 1 (PGGT-1), shifted the localization of the GFP-BD-CVIL protein from the membrane to the nucleus. In contrast, the inhibition of the mevalonate (MVA) pathway with mevinolin did not affect this localization. Furthermore, complementation assays with pathway-specific intermediates confirmed that the precursors for the cytosolic isoprenylation of this fusion protein are predominantly provided by the MEP pathway. In order to optimize this visualization system from a more qualitative assay to a statistically trustable medium or a high-throughput screening system, we established new conditions that permit culture and analysis in 96-well microtiter plates, followed by fluorescence microscopy. For further refinement, the existing GFP-BD-CVIL cell line was transformed with an estradiol-inducible vector driving the expression of a RFP protein, C-terminally fused to a nuclear localization signal (NLS-RFP). We are thus able to quantify the total number of viable cells versus the number of inhibited cells after various treatments. This approach also includes a semi-automatic counting system, based on the freely available image processing software. As a result, the time of image analysis as well as the risk of user-generated bias is reduced to a minimum. Moreover, there is no cross-induction of gene expression by dexamethasone and estradiol, which is an important prerequisite for this test system.", "keywords": [ "In eukaryotic cells", "certain proteins (i.e.", "members of the Ras superfamily of GTPases in vertebrates) are modified by a series of post-translational modifications", "leading to the creation of a lipidated", "hydrophobic domain at the carboxyl terminus of the protein. This post-translational processing", "also referred to as protein isoprenylation", "mediates protein-protein interactions", "increases the affinity of the prenylated proteins to cellular membranes and is therefore important for the targeting and function of such covalently modified proteins. Protein isoprenylation depends on the presence of a carboxy-terminal tetrapeptide in target proteins", "the CaaX motif (‘C’ refers to cysteine", "‘a’ denotes an aliphatic amino acid and ‘X’ represents a specific amino acid)1–4." ], "content": "Introduction\n\nIn eukaryotic cells, certain proteins (i.e., members of the Ras superfamily of GTPases in vertebrates) are modified by a series of post-translational modifications, leading to the creation of a lipidated, hydrophobic domain at the carboxyl terminus of the protein. This post-translational processing, also referred to as protein isoprenylation, mediates protein-protein interactions, increases the affinity of the prenylated proteins to cellular membranes and is therefore important for the targeting and function of such covalently modified proteins. Protein isoprenylation depends on the presence of a carboxy-terminal tetrapeptide in target proteins, the CaaX motif (‘C’ refers to cysteine, ‘a’ denotes an aliphatic amino acid and ‘X’ represents a specific amino acid)1–4.\n\nThis C-terminal CaaX motif is recognized by cytosolic CaaX protein isoprenyl transferases, which either attach a 15-carbon farnesyl unit (from farnesyl diphosphate, FPP), a reaction catalyzed by a protein farnesyl transferase (PFT), or a 20-carbon geranylgeranyl unit (from geranylgeranyl diphosphate, GGPP), a reaction catalyzed by protein geranylgeranyl transferase (PGGT1) to the cysteine of the CaaX motif via a thioether bond. The specificity of the reaction is mainly defined by the C-terminal “X”. As a general rule, proteins are geranylgeranylated when the “X” is a leucine residue, whereas any other amino acid - most probably a methionine, serine, alanine, glutamine or cysteine - will lead to the covalent attachment of a farnesyl residue5. A third mechanism is known for the members of the Rab family of small GTPases, which are isoprenylated at two different C-terminal cysteine residues by Rab geranylgeranyl transferase (also referred to as PGGT2)6–8. Both PFT and PGGT1 are heterodimeric enzymes that share a common α-subunit whereas their respective β-subunit is encoded by different genes3,4,6,9.\n\nFollowing prenylation in the cytosol, the newly lipidated protein is targeted to the endoplasmic reticulum (ER), where it usually undergoes two subsequent processing reactions; first, the C-terminal amino acid is removed by a specific endoprotease, known as RCE1 (RAS converting enzyme 1); second, the carboxyl group of the now exposed, lipidated cysteine residue is carboxyl-methylated by the enzyme isoprenylcysteine carboxyl methyltransferase (ICMT). In contrast to the prenylation reaction and the proteolytic removal of the -AAX tripeptide, this last step in the maturation of prenylated proteins can be reversed by certain methylesterase enzymes (ICME), which have been identified in animals and plants10–14. The proteolytic cleavage of the last three amino acids and the carboxyl-methylation are commonly referred to as “CaaX processing” or post-prenylation reactions15.\n\nIn addition to these post-translational modifications, certain proteins, such as NRAS, HRAS and KRAS4A in vertebrates or members of the Rop (Rho) family in plants can be palmitoylated as well or S-acylated, preceding their transfer to their cellular destination - most likely the plasma membrane16–19. Other proteins, such as KRAS4B do not require a second lipidic modification, but possess a polybasic, lysine-rich sequence in proximity to the C-terminal CaaX motif instead18.\n\nPrenylated proteins have been particularly well studied in vertebrates because of their implication in oncogenesis, as mutational constitutive activation of RAS GTPases is responsible for up to 20% of human cancers20–22. In comparison, less is known about prenylated proteins in plants, even though many studies suggest that these proteins play important roles in cellular processes similar to their yeast and mammalian counterparts, such as growth regulation, signal transduction, cell cycle regulation and membrane trafficking4,6,9,23–27.\n\nFollowing earlier observations with a specific calmodulin from Petunia (CaM6328,29), we have established a system for the visualization of protein geranylgeranylation, based on a dexamethasone-inducible GFP fusion protein, N-terminally fused to the C-terminal extension of rice calmodulin CaM61, which bears a basic domain (BD) plus a CVIL geranylgeranylation motif, using stably transformed tobacco BY-2 cells30–32. After the covalent modification of the GFP-BD-CVIL protein (or of its His6-tagged derivative H6-GFP-BD-CVIL) it becomes predominantly localized to the plasma membrane. The non-prenylatable control protein GFP-BD-C/S (GFP-BD-SVIL), in which the cysteinyl residue of the CVIL-motif had been replaced by a serinyl residue (thus removing the thiol group necessary for the covalent attachment of the prenyl moiety), nearly completely mislocated to the nucleus and in particular to the nucleoli of the cells. During the course of such studies it was revealed that the inhibition of the cytosolic MVA pathway by mevinolin had no effect on the geranylgeranylation-dependent insertion of the fusion protein into the plasma membrane30,31. This is in contrast to inhibition of the MEP pathway, for instance by oxoclomazone, which was previously recognized as an inhibitor of DXP synthase33,34 and with fosmidomycin, acting on MEP synthase35,36, the second enzyme in this pathway. We immediately thought of developing this system further into a versatile screening system capable of demonstrating the in vivo efficiency of MEP pathway (and protein geranylgeranylation) inhibitors30, though this required the miniaturization of the system to allow the use of fluorescence and confocal microscopy.\n\nOur contribution here is in focusing on the key steps that are necessary for the development of a screening pipeline37,38 and to demonstrate the feasibility of establishing a medium to high throughput compound screen on the basis of our in vivo protein isoprenylation assay30,32.\n\nOver the last decade, technological advances have dramatically changed the importance of image-based assays for modern cell biology. In the past, classical, non-microscopic approaches have been systematically used to investigate protein functions and interactions or to screen small-compound libraries in high-throughput. Thanks to the knowledge and tools developed during the evolution of various technologies in the last few years, including proteomics and genetics approaches39–42 and DNA microarrays43–45 or RNA interference experiments46–48, fluorescence microscopy has become a powerful method to study protein functions and interactions in the living cell45,49. This was in particular accomplished due to the availability of a great variety of fluorescent proteins and fluorophores50–54, which can be used to tag a protein of interest and to reveal information about its localization, its interaction with other cellular components and proteins or even to visualize biochemical reactions, e.g. the effect of a given treatment at a cellular level. This may permit conclusions on the physiological relevance of the protein, in contrast to classical, biochemical assays, where an isolated protein is tested in an artificial environment. In this way, data acquired by fluorescence microscopy can help to complement the above mentioned genetic and proteomic approaches.\n\nBecause of the availability of automated microscopes and more powerful image analysis software, multiple features of a cell can be measured, analyzed and compared, even over a certain period of time.\n\nPreviously, researchers were often forced to inspect their acquired microscopic images by eye, which is a time-consuming and subjective task, even for an experienced user55–57. Nowadays, modern, automated image acquisition platforms provide highly quantitative data and allow image-based screens of several thousand compounds a day, depending on the experimental set-up. In addition, fluorescence microscopy also allows to follow reactions at different scales. For instance, changes at the subcellular level can be observed using high resolution58,59, whereas a population of cells can be monitored using low resolution60,61, thereby providing a large set of data for every single image. This combination of high throughput-screening (HTS) methods and automated image acquisition and analysis has therefore been dubbed high-content-screening37,38,62–68.\n\nNevertheless, image-based chemical screening still remains a field in development, with the majority of users belonging to the pharmaceutical industry69. As the technology is being pushed by those companies, the variety of HCS systems available to the scientific community and the number of publications generated with them has considerably increased over the past few years69–73. However not every academic research unit having developed a biological assay with the potential for high-throughput/content screening will be able to use modern imaging platforms, as all the hardware and software components of a complete screening pipeline (not counting the time for development and validation) represent a significant financial investment68.\n\nAnother trend in high-content screening microscopy involves a multidimensional image read-out. This approach is also known as multiplexing74,75. Technically, it consists of using multi-color fluorescence microscopy and several distinct fluorescent markers in the same cell system. Of course, the spectral properties of the used fluorescent dyes and fluorophores as well as the detection gain and the computing ability of the microscopy platform are limiting factors, although this can be partially resolved by linear unmixing of fluorescent signals, “a method allowing the reliable separation of even strongly overlapping fluorescence signals...”76,77. Multiplexing is particularly interesting if more than one phenotype needs to be analyzed. As a side effect, changes in the morphology of the cells can also be monitored, which can reveal a potential toxicity event during the screening experiment. For example, multiple fluorophores, staining the nucleus, cytoplasm, microtubules, Golgi or endoplasmic reticulum could be detected in parallel, revealing additional information about cellular changes, as part of a small compound screen78,79. As a result, a pre-selection of drug-candidates can be performed at an early stage in the whole screening pipeline. This aspect can be quite important, considering that up to 30% of potential drug candidates are rejected because of toxicity issues80. Likewise, efficiency can be increased and the costs, significantly reduced64,80.\n\nHowever, one of the most critical steps remains the analysis of images, especially in purely cell-based assays. Although there is a variety of commercial software available for numerous purposes, many applications - especially those with complex cellular phenotypes - require custom-made solutions.\n\nBy successfully testing the robustness and selectivity of the test system based on transformed BY-2 cells, and also the efficiency of some novel drug candidates in vivo32, we had now established this bioassay as a qualitative and quantitative approach for the identification and evaluation of MEP pathway and eventually of protein geranylgeranylation inhibitors. However, all analyses performed so far with this bioassay were reliant on the performance of a single microscope user, observing biological processes and counting individual cells through the ocular of a microscope. Therefore further developments were required, in particular to reduce the involuntary bias of results by the user and to increase the speed and reproducibility of the application.\n\nSome important aspects can escape the human eye during a visual observation, and sometimes it is difficult to draw conclusions of biological significance by analysis of single cells. Therefore, it was important to check the reproducibility of the results with a random collection of cells. Furthermore, the quality of the acquired digital images had to be high enough for an automatic identification and measurement of biological features, such as the nuclear fluorescence observed in inhibitor-treated BY-2 cells. The use of BY-2 cells as a suitable model to investigate the effects of various treatments on the whole cell level has various advantages, including the efficient uptake of exogenous compounds and the short reproduction time. However, an important problem was the heterogeneity of fluorescence of liquid suspensions derived from the first-generation calli. We successfully solved this problem and this was an important prerequisite for the further development of a statistically significant screening system32.\n\nTherefore, a major emphasis of this study was to further improve the initial test system and to evaluate its potential for miniaturization, automatization and, if possible, high-throughput applications. However, the results obtained after treatment of the H6-GFP-BD-CVIL lines with squalestatin demonstrated that the effects on the morphology and localization of the fusion protein could be more complex than initially expected (due to partial delocalization, roundish cell shape and “all-or-nothing” – phenotype32).\n\n\nMethods & results\n\nIn order to demonstrate the feasibility and reproducibility of such an image-based approach, BY-2 cells expressing the H6-GFP-BD-CVIL fusion protein was treated with the MEP- or MVA pathway inhibitors, oxoclomazone (OC, 30 µM, synthesized by Dr. Eilers, Monsanto, St Louis, MO) or mevinolin (MEV, 5 µM, made available by MSD, Rahway NJ) respectively, as well as with a combination of both inhibitors according to standard protocols30. Images were acquired at low magnification using a Nikon E800 microscope. Low magnification fluorescence microscopy uses a resolution of 500 µm to 1 mm and is used to monitor phenotypic effects on entire cell populations. Medium-magnification fluorescence microscopy by definition is applied for subpopulation analysis at a resolution between 10–50 µm, whereas high-magnification fluorescence microscopy focuses on intracellular events and uses resolution of 1 µm or lower56. The phenotypes of the untreated control and the MEV-treated cells were more or less identical, with the majority of fluorescence at the cell periphery, especially at the boundary between cells in the files. In addition, GFP fluorescence was also clearly visible at the peri-nuclear membrane, whereas the GFP fluorescence in OC-treated cells was mostly localized in the nucleus. The combination of both inhibitors gave about the same or even more unequivocal results than cells treated with OC alone (Figure 1, cf.30).\n\nLow magnification fluorescence images of large numbers of cloned BY-2 cells expressing H6-GFP-DB-CVIL (cf. Gerber et al., 2009). Images were taken with a Nikon E800 microscope equipped with a DXM11200 CCD color camera (specifications: 20 × 0.45 objective; filters: Ex460-500, DM505, EM510-560). A: Untreated cells; B–D: cells treated with 30 µM oxoclomazone (OC, B), 5 µM mevinolin (MEV, C) and 30 µM OC plus 5 µM MEV (D). The white bar represents 50 µm.\n\nFor image analysis, we first used ImageJ, version 1.41 – 1.43, a publicly available Java-based image-processing program, which was inspired by NIH Image for the Macintosh81. The MBF ImageJ bundle used in this study provided an excellent selection of the most important plug-ins for microscopy users due to a well-illustrated online manual. At this point it should be noted that several other free image processing software programs are available nowadays that allow to customize the analysis of microscopic images and to generate tailor-made solutions for the identification and evaluation of complex phenotypes in experimental datasets82–84.\n\nOne of the major aspects when working in fluorescence microscopy is the level of saturation of the images. For instance, within an optimal scenario, all images taken in the green channel – displaying the localization of the geranylgeranylatable GFP fusion protein - should be acquired with a minimum of saturation to reduce signals corresponding to unspecific background noise or non-specific fluorescence associated with cellular structures. For the subsequent image analysis, it is quite important to differentiate any object from its surrounding background. This process is referred to as “segmentation”. Most of the software programs operating in the microscopes and used for image acquisition, such as the LSM image browser or the AxioVision software, offer the option to control the saturation levels of the images during or after the acquisition process.\n\nThe general principle of the particle analysis with ImageJ or similar software packages can be explained by using an image of 30 µM OC-treated cells (acquired with the E800 epifluorescence microscope from Nikon), where the delocalization of the GFP-BD-CVIL protein is clearly visible.\n\nThe overall process can be divided into the following parts:\n\ni “thresholding”\n\nii “watershed separation”\n\niii “particle analysis”.\n\nThresholding. The image requires conversion into a “binary” image, for instance into black and white (Figure 2). This can be achieved by using the menu command “Image/Type/8-bit”. Depending on the image parameters, such as luminosity and brightness, the contrast may be enhanced, with “Process>Enhance Contrast” (Figure 2). The thresholding is essential, as the software has to discriminate between the background and the borders of the object to be analyzed. The principle is quite simple: a threshold range is set, either manually or automatically, and all the pixels within this range are converted to black and those with values outside this range, converted to white (colors can be inverted). During this process, the pixels within the range are displayed in red, whereas the background remains black. By moving the scroll bars in either direction, pixels at the periphery of objects are added or deleted from the image, as demonstrated in Figure 3.\n\nA) Conversion of the RAW color image to a binary image file. B) Enhancement of the contrast.\n\nAutomatic thresholding (GFP-BD-CVIL BY-2 cell line treated with 30 µM OC) resulted in removing most of the pixels related to membranes. Only the strongest signals coming from the fluorescent nuclei remained visible. This step was particularly important as it removes the problems of user-generated bias. To guarantee maximum reproducibility, the MBF ImageJ bundle comes with a whole collection of plug-ins, using different algorithms for image thresholding (Figure 3).\n\nA) The threshold range can be adjusted automatically. Red pixels show objects that are within the threshold range. B) After pushing the “apply” button, the image is converted to black and white.\n\nSegmentation process - watershed separation. When using cell suspensions in microscopy, it is almost impossible to obtain images without overlapping cells or cells clumped and clustered into small groups, even if the dilution of the cells has been optimized beforehand. To minimize counting errors due to these problems, such as (in our case) nuclei that are slightly in contact with each other, watershed separation was used (“Process>Binary>Watershed”). The principle of watershed separation is quite simple85. It is based on the so-called “Euclidian Distance Map” (EDM). Any black pixel in the image is replaced with a grey pixel, whose intensity is proportional to its distance from the next white pixel. The intensity increases the closer it gets to the center of the object. In summary, this algorithm erodes objects from a binary image until they disappear. Then, it dilates them back, until they touch another black pixel. At the point of contact, a watershed line is drawn (Figure 4a).\n\nThe image from Figure 3 contains nuclei that are localized so closely together that they need to be separated by the watershed method describe in the main text.\n\nParticle analysis. There are several different options for counting particles with ImageJ. The menu command “Plug-ins>Particle analysis” shows a variety of plug-ins coming with the MBF ImageJ bundle. Among those are the “cell counter” and the “nucleus counter” plug-ins. In the context of our screening system, the default particle-counting menu “Analyze/Analyze particles” proved to be sufficient.\n\nThe desired parameters for particle-counting could easily be adapted for this purpose. By setting the minimum size of the object and the degree of circularity to a certain level, all objects not corresponding to the selected requirements would be excluded. Other menu points concern the visual output of the results in graphical or tabular form. The submenu routine “Show: Outlines” for example allowed to display the outlines of the detected objects with a reference number (Figure 4b).\n\nA) Particle analysis; B) Automatic particle-counting based on the data generated by the watershed method.\n\nThe “nucleus counter” plug-in mentioned before summarizes and automates many of the steps discussed before. Optionally, a summary of the results can be displayed and exported to other applications such as Microsoft Excel. In addition, ImageJ offers the possibility to create macros with the integrated macro editor (“Plugins>Macros>Record or Edit”). This feature allowed organization of the entire segmentation and particle analysis process by creating a customized step-by-step analysis pipeline.\n\nLimitations of the test system. As outlined above, detecting and counting cells that display a shift of GFP localization to the nucleus, for instance after a specific inhibitor treatment, is achievable with modern image analysis software like ImageJ. Nevertheless, these results also suggested that in order to get a statistically reliable picture of inhibitor treatment effects, it was necessary to develop a quick and inexpensive method to also detect those cells being unaffected by treatments. But, in contrast to other model systems like yeast or bacterial cells, tobacco BY-2 cells display a great variety of phenotypes in a given population of cells, such as significantly different sizes and shapes. In addition, they grow in files (optimally in tetrades, four cells in a row) or clusters. On a regular microscope slide these tend to overlap depending on the dilution of the cell suspension, which makes automatic detection/recognition of individual cells and features within them and their correct attribution extremely challenging with the software solutions available to us. Of course, there are several possible solutions, including the development of a custom-made algorithm, but this is usually very time-consuming and involves the work of software experts, as well as a whole cascade of follow-up validation experiments.\n\nAt this point we also thought about fixing the cells before examination by standard protocols, though this is inconvenient for monitoring a treatment over a certain period of time. A direction that has not been explored for eliminating at least the problem of heterogeneous cell shapes, is the generation of protoplasts and their analysis by flux cytometry. However, this approach was not considered here because the cell wall digestion might trigger a massive stress response in cells, which could have a direct effect on the localization of the fluorescent reporter protein H6-GFP-BD-CVIL. For instance, we observed that cells that were exposed to different stress factors (high temperatures, no agitation over-night) were less sensitive to inhibitor treatments. Preliminary tests showed that addition of methyl jasmonate (MeJA) to cells treated with 40 µM OC could partially complement the inhibitor effects, suggesting that different hormones might influence the pool of prenyl diphosphates in response to biotic or abiotic stress.\n\nTherefore, we decided to choose a less crude approach for counting untreated cells and to stain the nucleus with another fluorophore, as this technique had already proved valuable for the detection of inhibitor-treated cells. This approach should theoretically allow easy determination of the ratio of affected (GFP fluorescence in the nucleus) to non-affected cells (GFP fluorescence mainly localized at the PM). However, none of the commercially available and well-described nuclear stains (DAPI (4'-6-diamidino-2-phenylindole), Hoechst 33342 and 33268, etc.) worked satisfactorily with our model system.\n\nDue to the difficulties in efficiently staining nuclei in living tobacco BY-2 cells, a new strategy was chosen, consisting of transforming the existing GFP-BD-CVIL BY-2 cell line with an estradiol-inducible vector86, thereby driving the expression of a red fluorescent protein (RFP), C-terminally fused to a nuclear localization signal (NLS). The goal was then to stably transform the cell line (H6-GFP-BD-CVIL) with a second, inducible gene construct, which specifically stains the nuclear region of BY-2 cells. Therefore, the NLS sequence of the large simian virus (SV) 40 T-antigen (PPKKKRKV) known to be sufficient for targeting several proteins to the nucleus of mammalian and plant cells87–92 was fused N-terminally to the coding sequence of mRFP (monomeric RFP) and placed under the control of an estradiol-inducible promoter in the pER10 vector system86. The resulting vector was then used to stably transform the H6-GFP-BD-CVIL cell line via Agrobacterium-mediated transformation as described before30. Lee et al.93 showed that transformation of tobacco protoplasts with a SV40::RFP construct resulted in an efficient targeting of the fusion protein to the nucleus, making this protein a promising candidate. For the co-localization studies, we had chosen an enhanced RFP described by Campbell et al.94. The mRFP is a substantially modified version of the RFP from Discosoma coral (DsRed or drFP583)95, which was improved in regard to many major required characteristics, especially in the context of dual color imaging with GFP, one of the most important aspects relevant to a visual test system. For instance, the enhanced fluorophore mRFP is stable and bright, has a significantly shorter maturation time and its emission peak occurs at approximately 605 nm, thus facilitating the optical separation from the emission of the sGFP (synthetic GFP)87, which is a codon-optimized version of the green fluorescent protein (GFP) from Aequorea victoria97–100, with an emission peak at 511 nm101.\n\nThe N-terminal fusion of the SV40-NLS peptide efficiently targets mRFP to the nucleus of the double transformed tobacco BY-2 cell line. To investigate if the NLS fusion protein localizes to the nucleus and can be properly co-expressed with the GFP fusion protein of the pTA7001-H6-GFP-BD-CVIL line, 3-day-old BY-2 cells were transformed with the pER10-NLS-mRFP vector by agroinfection. (The original pER10-NLS-mRFP vector was kindly provided by Prof. Nam-Hai Chua, Rockefeller University, New York.) Calli were selected on BY-2 solid medium supplemented with 30 µg/ml hygromycin and 50 µg/ml kanamycin (Sigma). First calli appeared after 3 to 4 weeks of growth in the dark at 26°C and were subcultured twice on solid medium until liquid pre-cultures (10 ml with the same two antibiotics) were started. These pre-cultures were grown for 7 to 10 days under permanent shaking (154 rpm, routinely at 26°C, in constant darkness), until they reached a suitable optimal density. After parallel induction with estradiol and dexamethasone for 15 h, the pre-cultures were screened visually by fluorescence microscopy. Images were acquired using the microscope settings described in Table 1 (properties of the fluorochromes and fluorescent proteins used in this work). Twelve cultures out of 36 (33.3%) showed cells expressing both fluorescent proteins.\n\nThe most promising cell line (N-20) showed bright fluorescence in both channels set for the visualization of GFP and RFP, respectively, and had a significant ratio of fluorescent positive cells (> 50% of total cells), making it a good candidate for first tests and further re-selection efforts to obtain a performing cell line suitable for statistical approaches. In Figure 5a a typical cell tetrade is displayed, after induction by both dexamethasone (Dex, green fluorescence) and estradiol (Est, red fluorescence).\n\nAll channels: Four BY-2 cells (ideally) growing in a tetrade with red and green fluorescence induced at the same time (+ differential interference contrast, DIC). GFP: the GFP-BD-CVIL fusion protein is visualized after induction with 10 µM Dex and is mainly localized to the PM and cytosol. RFP: The monomeric RFP fused to the C-terminus of the SV40 NLS is visualized after induction with 5 µM estradiol and is predominantly found in the nuclear compartment. GFP/RFP: Overlay of the GFP and RFP channels. Induction time with both elicitors was 15 h. Images were acquired using a LSM510 confocal laser scanning microscope equipped with an inverted Carl Zeiss axiovert 100 M microscope. Dual color imaging was performed using dual excitation/emission scanning in the multitracking mode (Carl Zeiss Laser Scanning Microscope software). White bar represents 20 µm.\n\nThe newly created double fluorescent cell line NLS-mRFP/H6-GFP-BD-CVIL shows no visible cross-induction of fluorescence after treatment with estradiol or dexamethasone. In order to establish a reliable visual test system, it was necessary to verify that the two co-existing induction systems in the NLS-mRFP (Est) and H6-GFP-BD-CVIL (Dex) lines only respond to their specific inducers. As both chemical-inducible systems are based on a similar principle of induction - the action of chimeric trans-activators whose transcriptional activities are regulated by specific hormones and/or structurally related compounds86) - the newly generated cell line was separately treated with both inducers under standard conditions. In addition, fluorescent cell dyes were used in parallel as negative controls (Figure 6). Simultaneous treatment of the cell line with both inducers (24 h induction, 10 µM Dex, 6 µM Est) resulted in a phenotype with green fluorescence predominantly located at the peripheral membrane region and red fluorescence mainly located in the nuclear compartment.\n\nAfter treatment with Dex only (24 h, 10 µM), the observed cells did not show any detectable signal in the red channel (RFP). As a negative control, these cells were treated with the plasma membrane stain FM4-64 (5 µg/ml, 5 min treatment). FM4-64 fluorescence was clearly visible in the red channel, whereas no detectable signals appeared in the nuclear region (the cells had been treated with 50 µM OC to indicate the position of the nucleus).\n\nNo cross-induction of fusion gene expression was visible. Induction time for dexamethasone (Dex, 10 µM final) and estradiol (Est, 6 µM final) was 15 h in all experiments. + Dex/+ Est: control experiments with both inducers with the GFP fusion protein targeted to the PM (G) and the mFRP fusion protein located in the nucleus (R). + Dex (+ OC)/+ FM4-64: Dex alone is only inducing the expression of the GFP fusion protein (G). The negative control with FM4-64 (5 µg/ml; membrane stain) shows no signals from the nucleus (R). For a better understanding, cells were treated with 50 µM oxoclomazone (OC) 3 h before induction, to indicate the position of the nucleus. Fluorescein diacetate (FDA, 7.5 µM final) is used as negative control, also indicating that the cells were alive during the imaging process. G and R indicate the green and the red channel, respectively. Bars = 20 µm.\n\nCells treated with Est only (24 h, 6 µM) showed an accumulation of red fluorescence in the nuclear region and no detectable green fluorescence. As a negative control, the viability cell stain fluorescein diacetate (FDA102) was used (7.5 µM final, 2 min). Green fluorescence could be observed in the whole cell, which also indicated that the cells were living when the image was taken.\n\n24 h are sufficient to obtain homogenous red fluorescence. Another important aspect was the optimal duration of the induction times for both fusion proteins, in order to determine at which time point the intensities of the emitted signals were sufficiently strong and homogenous for detection and image acquisition. To address this issue, the cellular localization of the NLS:mRFP reporter protein was examined at various times after induction with Est by fluorescence microscopy. 7-day-old cells were diluted 6-fold in MS-BY-2 medium and 3 ml were dispatched into the wells of a 6-well plate. These cells were then cultured at 26°C in obscurity and under permanent shaking. Induction with 6 µM Est (solubilized in benzene/ethanol, 1:1) took place at 48, 36, 24, 18, 15, 12, 6 and 3 hours before observation. The settings for image acquisition in the red channel did not change during the whole experiment.\n\nTo determine the level of saturation, images were converted to a rainbow scale with the Zeiss LSM image browser. Red signals indicate saturation. As seen in Figure 7[A], nuclear localization of the NLS:mRFP protein could already be observed 3 h after induction. Nevertheless, it took at least 24 h until most of the signal arising from the nuclei was saturated (red dots, Figure 7[B]).\n\nIn vivo targeting of the NLS-mRFP fusion protein after induction with 6 µM Est (and in the presence of 10 µM Dex – not shown). The red fluorescent signals were examined at different time points after induction. A) Overlay of transmission microscopy images with the RFP channel. B) Same images as in A shown in LSM image browser rainbow mode (red indicates saturated areas). Saturation is already reached after 24 h of induction with both elicitors. All images were acquired with the same microscope settings (red channel, Carl Zeiss LSM510 microscope; EC-Plan-Neofluar, 10x/0,3 M27).\n\nCloning of transgenic BY-2 cells: generation of a cell line with strongly reduced heterogeneity. Generation of a performing double-transformed cell line was a process of constant engineering of the initial GFP-BD-CVIL BY-2 cell line in order to obtain a clonal selection of cells responding appropriately to different stimuli. In the course of this procedure, cloning of primary heterogeneous suspensions generated secondary homogenous lines. The resulting calli and suspensions derived thereof were both screened and evaluated by fluorescence. In this context it is important to consider that the terms “homogenous” and “heterogeneous” do not describe the intensity of the fluorescence, but rather if a given culture showed well-balanced and stable fluorescence within the cell population. To achieve this goal, the most promising double-fluorescent cell line (N-20) was chosen for a rigorous re-selection process. Two liquid cultures started from these secondary calli showed bright fluorescence in both channels as well as a high ratio of fluorescent cells (> 95%) and were maintained for further experiments (Figure 8). Nevertheless, it proved to be a time-consuming challenge, as within cell suspensions of supposedly clonal origin (primary suspensions derived from primary calli), important variations were regularly observed. These variations not only concerned the morphology of the cells, but also more importantly the homogeneity and intensity levels of the fluorescence. Therefore, a major focus was to re-select homogenous transgenic cell lines with high intensity levels of fluorescence. A relatively simple method to generate more homogenous cell lines, derived from secondary calli, was established and constantly improved over time, leading to the final protocol summarized in Table 2 together with a comparison of a procedure recently published by Nocarova and Fischer103.\n\nThe most promising cell line N-20 was re-selected in order to obtain a performing cell line. The resulting cell lines (N20-22 and N20-25) are characterized by a high ratio of bright fluorescent cells (> 95%). Images were taken as described in the legend to Figure 5. Cells were induced for 24 h with 10 µM Dex and 6 µM Est (final concentrations). White bars represent 50 µm.\n\nThis method proved to be quite efficient, as we succeeded in obtaining several secondary suspension cultures of the double-transformed mRFP-cell line that exhibited a high percentage of bright fluorescent cells (> 95%) under our experimental conditions.\n\nResponse of the doubly transformed cell line to various treatments. As the major emphasis of this project was the development of a tool to screen for inhibitors of the MEP pathway and the prenylation reaction itself, it was necessary to verify that the newly generated cell line showed the same phenotype in response to various treatments as the initial cell line H6-GFP-BD-CVIL despite the presence of a second, inducible fusion protein.\n\nThe doubly transformed BY-2 cell line was treated for 18 h with inhibitors like oxoclomazone (OC), fosmidomycin (FOS) and mevinolin (MEV) affecting key enzymes of the MEP and MVA pathways, respectively, as indicated in the legend to Figure 9. Induction with 10 µM Dex and 6 µM Est took place 24 h before observation (Figure 9).\n\nThe cell line was treated with different inhibitors affecting the MVA pathway (mevinolin) and the MEP pathway (fosmidomycin, oxoclomazone). Induction of the cell line with Est and Dex was carried out 24 h before observation. The cell line was treated with specific inhibitors 18 h before observation. Fosmidomycin (FOS) and oxoclomazone (OC) clearly shifted the localization of the GFP fusion protein to the nucleus, whereas mevinolin (MEV) treatment had no visible effect on the cells. In addition, we used two new cell lines with artificial micro interfering (amiRNA) silencing constructs as “biological controls”. For instance, DXR-silenced cells show the same phenotype as those inhibited in vivo by Fos. HMGR silencing however does not exert any visible effect on the localization of the GFP fusion protein. These results underline the specificity of the test system for the MEP pathway. White bars represent 20 µm.\n\nAs expected, inhibition of the cytosolic MVA pathway by mevinolin did not show any effect on the predominant localization of the green fluorescent fusion protein at the periphery of the treated cells, which is very similar to the fluorescence pattern of untreated cells. In contrast, inhibition of the first two enzymes in the MEP pathway (DXS and DXR) by OC and FOS at 40 µM final concentration each resulted in the previously described, nearly complete translocation of GFP-BD-CVIL to the nucleus, which is consistent with the results obtained in our previous study30. In all cases, the NLS-mRFP protein could be induced and co-expressed without affecting the expression or localization of the GFP-BD-CVIL protein.\n\nAs a further proof of concept, different cell lines were generated, targeting HMGR (MVA pathway) and DXR genes (MEP pathway) with an artificial micro-RNA (amiRNA) strategy104 in order to confirm the effects of the inhibitors on their target proteins on a biological level (Figure 9). DXR-silenced cells exhibited a phenotype similar to FOS and OC treatment, whereas the silencing of HMGR did not have any effect on the localization of GFP-BD-CVIL. However, this latter result, with only quite limited amiRNA silencing of all three HMGR genes in tobacco could be explained by recent observations on the dependency of miRNA formation on the biosynthesis of sterols: Apparently the formation of repressive complexes with ARGONAUTE (AGO) proteins needs a membrane association in which sterols are functional105, as was shown with miRNA action-deficient (mad) Arabidopsis mutants 3 and 4. MAD3 encodes HMGR1, and MAD4 encodes sterol C-8 isomerase, catalyzing an important step in cytoplasmic sterol biosynthesis105. Thus, silencing of a gene coding for a key enzyme like HMGR or downstream in the pathway might be impossible to achieve beyond a rather low degree105.\n\nThese results further confirmed the hypothesis that plastidial isoprenoid synthesis is the major contributor to the geranylgeranylation of GFP-BD-CVIL and once again proved the viability of using transformed BY-2 cells for the study of isoprenoid precursors.\n\nThe mRFP fluorescence can be used for the identification/quantification of cells. The proposed approach to determine the total number of fluorescent tobacco BY-2 cells in an image was to stain the nucleus, as this technique was successfully applied to identifying cells that showed a mislocalization of the GFP fluorescence to the nucleus in response to the treatment with the MEP pathway inhibitor oxoclomazone. Figure 10 shows and explains three different scenarios for the detection of cells, ranging from selected cells to whole-population images.\n\nThe flow of actions to be taken is shown left, and corresponding images and their conversion for particle counting is shown for A) a file of cells, B) a group of cells at higher magnification, and C) at lower magnification.\n\nTo obtain statistical data, we combined the analyzed images of the red and green channels. The detected nuclei in the red channel are used as a mask that serves as a landmark for the zone we intend to investigate in the green channel, the nucleus (Figure 10). This proved to be a promising approach for the analysis of rather simple scenarios and its applicability for more complicated scenarios has to be further investigated. Critical points are the homogeneity and intensity of the fluorescence. This protocol can however be saved and re-used for the quick analysis of multiple images or sets of images, given that the image acquisition settings are identical. Therefore, the more constant the treatment conditions, the fewer follow-up adjustments for the image analysis that are required.\n\nThe mRFP fluorescence emitted at the nucleus can be used to find the optimal focal plane for the acquisition of images in double fluorescence mode. Modern image-based screening approaches typically use multi-well plates for efficiently screening chemical libraries, at a medium to high throughput level38,56,64. This requires the acquisition of several images from every well, a task that is commonly accomplished by automated microscope platforms. One of the major challenges of these systems remains the focusing technology106. Depending on the application, individual routines often have to be developed in order to acquire images of adequate quality for later image analysis. Fluorescence-based focusing has several disadvantages, including photobleaching and possible phototoxicity107,108. However, observations made during manual focusing with the double-fluorescent cell lines indicated that the maximal intensity in the red channel (nuclear-localized mRFP) correlated with the focal plane, found by a human observer. These early results suggested that the fluorescence emitted by the mRFP could be used for later autofocusing purposes, keeping in mind that one of the main features we are interested in is the change of subcellular localization of the prenylable GFP fusion protein from the periphery of the cell to the nucleus after inhibition.\n\nTo confirm these early observations, a series of multichannel images of BY-2 cells (expressing both fluorescent proteins) spanning a total distance of 50 µm in the Z-plane was acquired at different focal planes. Afterwards, each optical slice of this Z-stack was analyzed using ImageJ software (Figure 11). The images of the green channel were analyzed by the edge-finding algorithm of ImageJ, whereas the integrated density was calculated for the red channel. The results clearly show that, for the green channel, the sharpest image (as perceived by a human observer) of the Z-stack (identified by the edge-finding algorithm of ImageJ) is also the image with the highest integrated density in the red channel, which is defined as “the sum of the values of the pixels in an image or selection” (ImageJ online manual). This correlates very well with general observations about fluorescence images that indicate a maximum image contrast at the Z-stage height corresponding to the focal position107,108. Therefore, a fluorescence-based autofocusing approach could use the nucleus-located maximum of red fluorescence to define the plane of focus and acquire additional pictures at an offset from this position (in both directions). This vertical series of images could then be summed up into a single projection or used to choose the best focal plane for each fluorophore. In an optimal scenario, two different images in different focal planes should be taken, when working with different fluorophores/wavelengths, due to the chromatic aberration of optical lenses (objectives), which means that different colors/wavelength of light are focused to different points109.\n\nImages are subsequently taken at different levels of the sample (Z-stack). The optimum sharpness of signals for RFP indicates the focus being set on nuclei. The determination of the sharpest image for the GFP signal provides an integrated optimum for the analysis of doubly transformed BY-2 cells.\n\nModern microscope-based screening approaches typically use multi-well plates, as these allow a significant increase in the number of tested compounds per day and save reagents and consumables at the same time due to the miniaturization of the experimental setup110.\n\nThe most common format for classical medium (up to 10,000 compounds/day) to high throughput (between 10,000–100,000 compounds/day) screening assays are 96- or 384-well plates with average working volumes of 100–200 µl and 50 µl per well, respectively. Despite the possible efficiency gains connected with these high-density formats, there are severe technical hurdles for their use in routine HTS assays, most importantly the adaptation of automated liquid handling and dispensing technology, which is better established for the bigger 96- and 384-well formats110,111.\n\nOur goal was to significantly decrease the working volume for the assay and to use a format fulfilling the requirements of modern cellular imaging platforms, capable of (automatically) acquiring images at a (reasonable) throughput rate. However, the image quality should still be sufficiently good to monitor whole cell populations, on the one hand and to measure intracellular events on the other.\n\nIdeally, multiple images at different positions and different magnifications should be acquired from each well. This kind of read-out however would be extremely time-consuming and take several hours to process an entire multi-well plate. Therefore, it was necessary to find the right balance between high content and a reasonable throughput or, in other words, between time, cost and quality, which Mayr and Fuerst110 called “the magic triangle of HTS”.\n\nCriteria and parameters to be considered. In order to find the format providing us with the greatest flexibility as far as the quality of the image acquisition, the growth conditions of the BY-2 cells, and the general liquid handling were concerned, it was first necessary to take a closer look at some of the characteristics of the model system used in the bio-assay.\n\nBY-2 cells usually grow in files or individual cells (in the exponential growth phase), easily reaching 50 to 100 µm in length and more than 30 µm in width, with the nucleus having a diameter between 10 and 20 µm (our observations, after measuring in average several hundred cells). Previous results already indicated that the images taken in a medium to low magnification- mode (10 × objective - resolution in the µm range) could be exploited by image analysis software and provide sufficient information for the analysis of the observed phenotypes. For images acquired with the 10 × objective (EC-Plan Neofluar 10 × /0.30 M27), the field of view ranges from about 900 µm × 900 µm to 1272 µm × 1272 µm (stack size: x-plane × y-plane) in the greatest possible number of configurations. A field of this size (approximately 1 mm × 1 mm) allows the study of up to 100 cells on a conventional microscope cover slide, depending on the dilution factor of the culture. In order to obtain data from a statistically significant number of cells for each treatment, images from multiple fields should be collected from each well. Typically, 96-well plates (i.e., Cellstar® Cat.-No.650 180, Greiner bio-one, Les Ulis, 91941 Courtaboeuf, France) have an internal diameter at the bottom of the well of approximately 5 mm. The diameter of the next largest format, the 384-well plate, is already significantly smaller at about 3 mm111. However, considering the size of the cells and the image field, as well as the need to acquire multiple images, all formats smaller than 96-wells did not make any sense for our experimental system.\n\nThe diameter of a conventional round-shaped well allows the acquisition of at least 9 independent fields of more than 1 mm × 1 mm, without interfering with the walls of the wells. However, one of the limitations remains the possible read-out pattern, which cannot exploit the whole surface of the well. In order to maximize the surface for the read-out, commercial square-shaped micro-array plates (with glass-bottoms) offered an interesting alternative to round 96-well plates (Figure 12).\n\nThe scheme explains the steps of how the LSM microscope can scan wells. Images can be taken at low resolution to indicate the behavior of many cells for statistical evaluation. But at the same time high resolution is achieved by zooming into the image for detailed analysis of fluorescent protein localization, either to the PM (control) or to nuclei (after addition of an appropriate inhibitor like oxoclomazone (OC) at 40 µM).\n\nFigure 12 shows a field of view of 1 mm2 that was acquired using an inverted fluorescence microscope (Carl Zeiss) and a 96-well plate with a glass bottom (No. 1.5, γ-irradiated, MatTek Corporation - Ashland, MA, USA) that resembles a conventional cover slide in thickness. To be able to use 96-well plates, a special stage adaptor for multi-well plates was purchased from Carl Zeiss (Invertoskop Microscope Specimen Holder). In contrast to the image acquisition for cells from a normal cover slide, the situation is a little more complex for cells in a wel, as hardware and software have to cope with cells in a more extensive three-dimensional space (a suspension of cells with a height of several mms). This means that not all of the visible cells are in the right focal plane and align perfectly parallel to the z-axes. Nevertheless, this lack of “substrate flatness”106 should partially be solved by the presence of a reference point for the focusing software. For that reason we generated the double-transformed cell line emitting red fluorescence through an inducible, nuclear-localization fusion protein.\n\nResults obtained by manual focusing clearly demonstrated that it is possible to resolve subcellular details at a satisfying resolution (Figure 12) (1024 × 1024 pixels × 8) and deliver enough information to distinguish the phenotypes of interest, here for instance the mislocalization of H6-GFP-BD-CVIL to the nuclei after treatment with OC (Figure 12).\n\nBesides the quality of the image acquisition process, another important factor for the miniaturization of the assay was the nature of the biological material, which sets distinct limits for the downscaling process and needs the adaptation of various parameters for the significantly smaller format (adjusting minimal and maximal fill volumes; agitation; minimizing the evaporation; liquid handling technology etc.)\n\nProblems in the optimization of culture conditions in microtiter plates. In this context, it must be kept in mind that this bioassay, in contrast to most cell-based imaging approaches, relies on the observation of living cells. The majority of image-based screens at a high-throughput rate are usually performed with fixed cells64. BY-2 cells are grown in liquid medium, which means that the treated and induced cultures require permanent shaking for more than 20 h to prevent cell sedimentation, as this may lead to sub-optimal nutrient and oxygen supply and could interfere with the expression of the fluorescent reporter proteins.\n\nWe examined the influence of suboptimal agitation in this small-scale system (volume 100 to 200 µl) on the expression of the reporter proteins in several independent experiments in which different shaking conditions for the BY-2 cultures were tested. Therefore, 7-day-old cells were diluted (1:10) into fresh BY-2 medium and then induced by the addition of 10 µM Dex and 5 µM Est. Then 200 µl of this dilution was transferred into the wells of a 96-well plate (conventional round-shaped wells, with conical bottom) and incubated for 20 h in the dark under permanent shaking (160 rpm or 320 rpm). Cells that were shaken at 160 rpm (which corresponds to the shaking frequency of 6-well plates and culture flasks) showed a normal induction of the GFP fusion protein, whereas the mRFP fusion protein was barely expressed. However, in cells that were cultivated at 320 rpm, the expression of the NLS-mRFP protein could clearly be detected by fluorescence microscopy (Figure 13). The gas-liquid mass transfer properties of shaken 96-well plates have been investigated in detail by Hermann et al.112 and revealed that the oxygen transfer rate (OTR) measured in the wells was strongly influenced by different parameters, such as the surface tension of the medium, the material of the well, the filling volume and the shape of the well. In round-shaped wells, for example, due to the high surface tension, no liquid movement occurred until a critical shaking intensity was reached: for 200 µl of water shaken at shaking diameter of 25 mm, the rate had to exceed 300 rpm. On the other hand, frequencies above 450 rpm could not be used without the risk of the liquid spilling out of the well112. As a general rule, one can say that the OTR increases proportionally with shaking amplitude and frequency due to an increase in the total surface that is available for oxygen (gas) transfer. The same effect was observed by replacing round-shape wells by rectangular or square wells, which can be explained by the increase of the turbulence of the system due to the effect of the corners. A higher fill volume on the other hand decreases the oxygen transfer rate if all other parameters are kept constant112–114. No agitation of the fluid (diluted cells in BY-2 medium) was observed for 250 µl until around 300 rpm (using a Heidolph unimax 1010 shaker, 10 mm). Therefore a frequency of 320 rpm and higher was used. However the limitation for further testing was the maximum speed of the available shaker (500 rpm). In addition, these results were obtained by using wells with a conical bottom. The use of an inverted microscope required 96-wells with a flat-bottom for the imaging process, and we found that the hydrodynamic behavior of a BY-2 culture in a flat-bottomed well differs significantly from a conventional deep-well. Preliminary results indicate that the speed has to exceed 500 rpm to assure an optimum agitation of the cells for a filling volume of 200 µl. This result prompted us to purchase 96-well glass-bottom plates with a square-shaped cross-section area/ground profile. Besides increasing the OTR at lower shaking frequencies compared to round wells, it should also confer an additional advantage to the read-out process by significantly increasing the total surface area of the well.\n\nShown is the expression of both fluorescent reporter proteins in transgenic tobacco BY-2 cells incubated over-night in the wells of a 96-well microtiter plate. A 7-day-old culture was diluted tenfold into a fresh culture medium before the inducers Dex (10 µM) and Est (5 µM) were added. Afterwards 200 µl of the cell suspension were added to the wells of a 96-well glass-bottom microtiter plate and shaken under different conditions (150 rpm and 320 rpm, respectively) before being examined by fluorescence microscopy as described in previous Figure legends. White bars = 10 µm.\n\nA prerequisite for an image-based screening system is a certain degree of automatization as far as repetitive tasks are concerned. The use of the AutofocusScreen for LSM macro provided by Zeiss allows the automation of different steps of the image acquisition process. The first tests performed with the 96-well glass bottom plates indicated that all features could be used, including the autofocus routine and the automatic well-readout (Figure 14). However, in order to find the right balance between speed and image quality, the protocol still requires refinement and further validation before reproducible and exploitable data sets may be obtained.\n\nImages from multiple locations can be taken using the “AutofocusScreen for LSM” macro, developed in collaboration between Carl Zeiss MicroImaging GMBH (Jena, Germany) and the group of Dr. Jan Ellenberg at the European Molecular Biology Laboratory (EMBL, Heidelberg, Germany). It is freely downloadable at http://www.zeiss.de/LSM-Macros. Well positions selected to be scanned can be defined by simply clicking check boxes. Steps X and Y define the distance between two acquisition locations. The exact position (here marked by a red cross) will be defined by the user, at the beginning of the image acquisition procedure. The user also has the option to define multiple tiles (with no or partial overlap to each other) around this position. The size of the scanned image will also be defined by the user and the scanning settings he chooses. Autofocus can be hardware- or cell-based, acquiring the emitted laser light or the emitted light of the sample respectively (the above shown image displays a cell-based scan across the Z-axis, using the green channel). Initial experiments with tobacco- BY-2 cells however showed that at multiple positions in the Z-axis above the glass bottom, a significant amount of cells could be visualized in the respective focal plane. This factor however will have to be adjusted for the scanning of multiple wells, as BY-2 cells will sediment quite fast (within minutes) to the bottom of the well, resulting in a significant change in the conditions and number of cells in the Z-axis.\n\n\nDiscussion\n\nAccording to Carpenter37,38, in chemical screening based on fluorescence-imaging, intensity might fluctuate from cell to cell for various reasons, including differences in cell cycle position, stochastic variations in gene expression, pre-existing amounts of proteins and metabolites in each cell and micro-environmental differences (due to cell medium or cell-to-cell-contacts). However, these are only a few examples, and in order to better understand the impact of multiple factors on the growth and expression capacity of transgenic tobacco BY-2 cells, possible causes of variations need a discussion in the context of observations made during our studies.\n\nExpression noise and cell-to-cell variations. Variations in the expression of proteins in a population of genetically identical cells may occur for various reasons and there are many aspects that may contribute to this behavior115. For instance, several studies in bacterial and yeast model systems have shown that a certain amount of this cell-to-cell variation resulted from so-called “expression noise”, that may be defined as stochastic fluctuations in the expression of a gene115–118 focused on the expression noise in gene networks, and showed that these stochastic variations were caused by intrinsic noise at the level of the gene (e.g. number of mRNA copies), transmitted noise from upstream genes and global noise affecting all the genes.\n\nOther studies however suggested that expression noise was rather a minor source of total cell-to-cell variations119 and showed that the differences may be caused by other factors, such as the capacity of individual cells to express proteins from genes (expression capacity). For instance, differences may occur in the levels of cellular components needed for protein expression (e.g. variations in the global pool of housekeeping genes, cell cycle position or fluctuations of environmental conditions discussed later). As an example, Gordon et al.108 monitored the expression levels and maturation rates of YFP (yellow fluorescent protein) in exponentially growing yeast cells with a pheromone-inducible gene expression system. Interestingly, they observed that the total amount of the reporter protein YFP could vary by up to a 4-fold in inducer-treated yeast cells (Saccharomyces cerevisiae), whereas the maturation rates of the protein only showed little variations (39 min +/- 7 min). Even though these studies used less complex model systems than plant cells, this could explain the variations in fluorescence levels observed during this work.\n\nNatural heterogeneity of transgene expression in tobacco BY-2 cells. Tobacco BY-2 cells are often referred to as the HeLa cells of plant molecular biology and were used in hundreds of studies focusing on various aspects of plant physiology120,121. Under standard growth conditions, the cell duplication time is around 14 h120 and the cell divisions can be synchronized, which allows cell cycle-related studies121–124. Although they are not able to form chloroplasts and have to be grown under heterotrophic conditions, they nevertheless contain active proplastids and leucoplasts120 and have been shown to be an excellent system to study the synthesis of sterols and isoprenoids125,126.\n\nGiven the higher complexity of plants, such as tobacco, compared to bacteria or yeast, variations in the expression levels of endogenous and reporter proteins from one cell to another can easily be imagined. This aspect is particularly interesting in connection with our observations, where transgene expression in individual cells of the newly generated H6-GFP-BD-CVIL BY-2 cell line proved to be unstable and heterogeneous in many cases (Figure 6).\n\nThese often dramatic changes in the brightness of fluorescent cells, as well as the unstable ratio of fluorescent to non-fluorescent cells in a supposedly clonal cell line soon led us to hypothesize that the cultured cell suspensions, derived from primary calli (the first calli obtained after A. tumefaciens-mediated transformation), could contain (epi-)genetically different cells. The heterogeneity of callus cultures is a well-known phenomenon for many plant species and consequently, the selection of highly productive cell lines that show the desired attributes is a common step, especially for commercial efforts to produce secondary metabolites in vitro127–129.\n\nSuch variations in expression levels among independent transgenic lines (from the same initial transformation event) might also arise from the inserted sequence itself. Changes can indeed be induced by the methylation degree of chromosomal insertion regions130, the locus of the insertion131, and the number of insertion copies or transgene silencing132–134. For instance, only recently have experiments demonstrated that the integration site of a transgene significantly influences its susceptibility to RNA silencing, rather than affecting its initial expression level135.\n\nIn the past, several studies136,137 helped to get a clearer picture of the variations of transgene expression in genetically identical clones. However, the most interesting contribution to this poorly understood topic came very recently from Nocarova and Fischer103. They reported a method to clone transgenic tobacco BY-2 cells with the goal of reducing the high natural heterogeneity of transgene expression. The cell lines generated in their laboratory “repeatedly produced only a low frequency of cell lines with well-balanced and stable fluorescence in all cells”. These observations corroborated the results obtained during the generation of both cell lines (H6-GFP-BD-CVIL and SV40-mRFP/H6-GFP-BD-CVIL) during the study presented here.\n\nIn order to identify the sources of such heterogeneity, Nocarova and Fischer103 transformed tobacco BY-2 cells with a gene encoding a free GFP under the control of the CaMV 35S promoter driving the constitutive expression of the transgene. Then they monitored the expression levels of GFP in primary calli and the derived suspension cultures (primary suspensions) as well as in secondary calli and suspensions they obtained by a simple cloning/selection procedure (Table 2). Interestingly, only about 40% of the (primary) calli obtained after Agrobacterium-mediated transformation showed homogenous GFP fluorescence. The remaining calli displayed heterogeneous GFP expression, either in a mosaic (m) or sectorial (s) distribution pattern within the calli. In addition, up to 90% of the (primary) suspension culture lines derived from all primary calli consisted of cells with heterogeneous levels of GFP fluorescence. On the other hand, secondary calli, obtained with their cloning method, showed homogenous fluorescence in approximately 90% of the cases, whereas only a little more than 40% of secondary suspensions had homogenous GFP fluorescence intensities. Molecular analysis of the primary clones by Nocarova and Fischer103 by Southern hybridization identified two causes for the observed heterogeneity:\n\nThe first was genetic heterogeneity due to the presence of cells with different T-DNA insertions, and the second was epigenetic heterogeneity, caused by transgene silencing at the transcriptional level in connection with DNA-methylation, as treatment with the DNA-demethylation drug 5-azacytidin138 reactivated GFP expression in some lines. In many cases this heterogeneity could be resolved by subsequent cloning, but nevertheless a certain fraction showed what the authors called a “permanent expression heterogeneity” which could, for example, be due to temporal changes in the accessibility of promoter sequences to transcription factors132.\n\nTo resume this part, these results are interesting in different ways: first, they help to explain the heterogeneity in the fluorescence levels of the transgenic cell lines generated in this work. This heterogeneity was observed in our lab independently from other sources, and various approaches were discussed on how to eliminate or reduce it. Finally, we developed a simple and inexpensive method to generate secondary calli, and the resulting transgenic lines proved to be more homogenous. For the GFP-BD-CVIL line, 23 out of 48 suspension cultures (~48%) derived from primary calli showed fluorescent cells. However, the majority of these lines displayed weak ratios of fluorescent cells to non-fluorescent cells. The four lines with the best ratios and highest fluorescence intensities were chosen to reselect calli with our protocol. Two out of 15 secondary lines showed strong, homogenous fluorescence and a very high ratio of fluorescent to non-fluorescent cells (Figure 16).\n\nFor the double transformation of the GFP-BD-CVIL line with the SV40-mRFP gene construct, 20 appearing calli were screened, and seven out of them showed strong fluorescence in the first generation of cell suspension cultures (35%). In order to obtain a more homogenous line, calli from suspension cultures were reselected and four promising cell lines were obtained, but only the best-performing line was subcultured in liquid medium.\n\nOptimization of culture conditions in microtiter plates. Another important factor, which plays a key role in the successful set-up and miniaturization of the assay, are changes in growth conditions, such as in the inoculum ratio, agitation, aeration and temperature. Growth conditions for tobacco BY-2 cells, subcultured on a weekly basis in 250 ml Erlenmeyer flasks, were more or less constant for several years: 0.75 to 2 ml of a 7-day-old, stationary liquid suspension culture was used to inoculate 40 ml of BY-2 medium supplemented with the required selection markers and the cells were grown in the dark on a rotary shaker at 154 rpm and routinely at 26°C30,126. Nevertheless, given various problems with the culture room facilities, conditions were not always as stable as desired. In addition, downscaling the whole test system from commercial 6-well plates to high-tech glass bottom 96-well plates made it necessary to re-examine all growth conditions to determine optimal conditions for this smaller format.\n\nFor instance, agitation of plant culture cells plays an essential role as it provides homogeneity of the culture in respect to nutrients, enhances mass and heat transfer, and reduces cell clumping and formation of aggregates, a phenomenon often observed in plant cell cultures, due to secretion of extracellular polysaccharides (EPS) (see139–141). For example, insufficient mixing triggers the formation of aggregates and some heterogeneity in oxygen and nutrient supply inside the cell population. Therefore, sub-optimal agitation conditions may also explain the differences in transgene expression levels observed during this study.\n\nAeration of plant suspension cultures is also an important parameter as it leads to desorption of volatile products and removes metabolic heat. Oxygen supply is hence a very crucial factor, as excess or lack can both have negative effects. Without going into detail, the mass transfer coefficient, KLa is a function of agitation and aeration at the same time and is part of an equation commonly used to optimize growth conditions in modern bioreactor systems129. Plant cells grow relatively slowly and are known to be particularly sensitive as to their optimal oxygen supply. They adapt their metabolism even to minor changes in gas composition. This may result in an alteration of growth characteristics and in production of secondary metabolites127. Under sub-optimal growth conditions, one can easily imagine that an insufficient supply of the prenylation precursor, geranylgeranyl diphosphate (GGPP) and its hydrolyzed product geranylgeraniol (GGol) in some cells could result in the shift to the nucleus of a part of the fluorescent GFP-BD-CVIL protein. This view is supported by the fact that exogenous GGol is able to completely reverse inhibition of the MEP pathway, whereas the control experiments always had a fraction (< 5%) of cells with signals from the nuclei142.\n\nFinally, the temperature is a major factor for the cultivation of plant cell cultures. Conditions are dependent on the plant species, and even for the same species, optimal temperatures may vary as far as the synthesis of a distinct metabolite is concerned129,143,144.\n\nOf course, all these parameters have to be adapted to the scale of the system, the cell culture line, the culture conditions and the growth phase.\n\nAcquisition of images at low magnification as a prerequisite. An essential aspect and one of the novel contributions of this study was the acquisition of images at low magnification, showing groups and sub-populations of cells. The advantage of such an approach is the fact that an image can be stored and used to give the relevant biological information that might just be overseen or misinterpreted by a microscope user at a given moment. With the software ImageJ, it was possible to detect and quantify cells displaying a mislocalization of the fusion protein by a rather simple image analysis approach, using the nucleus as a reference point. But a major challenge for our approach was the detection of untreated cells, due to the fact that BY-2 cells are very diverse in shape and size and grow in files of different size. In addition, because they are growing in a liquid medium, it is nearly impossible to avoid superimposed cells, whatever the dilution. This is a hard challenge for fully automated image analysis software, even for specific plug-ins for the detection of cellular features that are adapted to simpler model systems and scenarios. Most likely a custom-made algorithm might cope with this problem and resolve it. A possible solution to the problem of cell shape heterogeneity might be the generation of protoplasts and their analysis by flux cytometry, but from our experience the necessary digestion of the cell wall would trigger stress responses critically biasing the observations.\n\nIn a previous study30,31 it had been demonstrated that inhibition of MEP pathway enzymes and of protein geranylgeranylation led to essentially identical phenomena, viz., the mislocalization of GFP-BD-CVIL (and of its derivative H6-GFP-BD-CVIL) from the plasmalemma to the nucleus. By simple chemical complementation it was possible to distinguish between inhibitors of at least the first two enzymes in the MEP pathway and isoprenylation itself. Thus, given that there is a hit in screening unknown compounds, this method could be applied again. Within a short time, the range of putative molecular targets could be minimized by such follow-up experiments. It also became obvious that the mislocalization was perfect in the presence of a low concentration of mevinolin, which suggests that the MVA pathway seems to contribute to, or promote, the formation of isoprenyl residues (< 5%). However, in an extension of the test method it is also possible to treat all cells with MEP pathway inhibitors and then screen for low molecular weight compounds that alleviate the inhibition, visualized by targeting the GFP fusion protein back into the plasmalemma, as we have seen for instance with exogenous GGol, which efficiently overcame the MEP pathway inhibition at < 5 µM. Preliminary observations argue also for a hormonal regulation of protein isoprenyl transferase activity and for some adaptation of enzymes to the pool size of substrates, indicative of some kinetic flexibility in distinguishing between C15 and C20 substrates.\n\nThe availability of this bioassay should be useful to address open questions and for many applications in agriculture and medicine. For instance, one can imagine applying this test to the search of compounds that interfere with the bio-activation of the herbicide clomazone to oxoclomazone, which is P450-mediated, or more generally the search of new herbicides. To elucidate mechanisms underlying the intracellular transport of prenylated protein in BY-2 cells is also an interesting topic. According to Gerber142, the classical secretory pathway appeared not to be involved as a transit route of the GFP-BD-CVIL to the plasmalemma of BY-2 cells. Interestingly, human KRAS4B shares several features with the GFP-BD-CVIL fusion protein (e.g. the polybasic domain and the CaaX-motif)145,146. Several prenyl-binding proteins are known, such as Rho-GDI or the δ-subunit of phosphodiesterase (PDEδ) in animals147–149 that are involved in the binding of isoprenylated Ras protein and the transport to its final membrane destination. Similar proteins might be candidates for the transport of H6-GFP-BD-CVIL from the ER to its final cellular destination. The use of our bioassay in combination with inhibitors of CaaX-processing and compounds interfering with the membrane anchorage of H6-GFP-BD-CVIL might be a promising strategy to investigate this aspect in the development of anti-cancer strategies.\n\nThe bioassay may also be very useful to screen unknown chemical compounds with cytotoxic properties. By testing potential inhibitors derived from plant extracts of the Clusiaceae family that are traditionally used for the treatment of parasitic diseases in Cameroon, we made an interesting observation (Figure 15). Whereas none of the compounds was efficiently inducing a mislocalization of the fusion protein to the nucleus (compare OC treatment), several treatments triggered toxic effects. One compound in particular displayed an unusual phenotype, with barely detectable signals from the plasma membrane (PM) and nucleus (N/Nu); instead, the GFP fluorescence seemed to be trapped in the cytoplasm (Cyt). Interestingly, this compound was identified as a prenylated anthracenoid, isolated from Psorospermum glaberrium (Clusiaceae) found in Cameroon, and shown to have the highest activity in tests against leishmaniasis149,151, further validating the experimental system as a tool to detect cytotoxic effects. This could be an indication that the transport of H6-GFP-BD-CVIL might be mediated by protein-protein interactions, which are impaired by the prenylated compound.\n\nOn the left side as controls cells without and with oxoclomazone (OC) treatment showing plasma membrane (PM) localization or its mislocalization to the nucleus (N), especially in nucleoli (Nu). On the right side cells are depicted that have been treated with an inhibitor that has shown a high efficiency in killing protozoan Leishmania cells, responsible of leishmaniasis. The H6-GFP-BD-CVIL is mainly accumulated in the cytoplasm (Cyt) that is for instance surrounding the nucleus. At the higher magnification some punctuate structures become visible. That might result from coagulated cytoplasm typical of cell death. White bars in the combined images represent 20 µm.\n\nAnthroquinolol derivatives being structurally similar to the above-mentioned anthracenoid, isolated from Antrodia camphorata, an endemic fungus found in Taiwan, have very recently been shown to block Ras and Ras-related GTP-binding protein activation in human lung cancer (A549 and H838), liver cancer (HepG2 and Hep3B), and leukemia (K562 and THP-1) cell lines by direct binding to FTase and to GGTase I152.\n\nComparison with other screening methods for inhibitors of the MEP pathway and potential for further applications. Screening methods for MEP pathway inhibitors have already been described in the past, but were based on in vitro assays, using enzymes from E. coli alone and fluorescent compounds153 or in combination with downstream enzymes of the MEP pathway together with auxiliary enzymes necessary for optical measurements in medium-throughput methods154. Extracts from a series of Mediterranean plants were for instance examined for their inhibitory potency of E. coli MEP synthase (ispC = deoxyxylulose phosphate reductoisomerase, DXR), with the most efficient one coming from Cercis siliquastrum154. However, the truly active individual components could not yet be identified. A similar system based on cloned and heterologously expressed and combined DXS and DXR from Mycobacterium tuberculosis was recently described155. While such approaches are certainly suited to identify inhibitors that react with an enzyme, their activity in vivo could well be disappointing if they are not absorbed to penetrate internal cell membranes to arrive at their molecular target(s). This is the principal advantage of using a whole-cell method, as described in this article. We have recently become aware of a faintly related approach, based on the measurable growth inhibition of specifically engineered strains of Salmonella typhimurium having a separately inducible (native) MEP pathway and an inducible (non-native) MVA-utilizing pathway156. However, compounds that are efficient in a plant cell-based system like oxoclomazone30 are only mildly or not at all inhibitory to bacterial counterparts of MEP pathway enzymes. Furthermore, our screening system is also suitable to check for inhibitors of protein geranylgeranylation, an important aspect in for instance cancer research157 and even in studies on the recruitment of the eukaryotic cell protein isoprenylation machinery by pathogenic bacteria injecting isoprenylatable proteins (cf.7,8,21). More recently, a test system was described, based on inhibitors affecting the accumulation of phytoene in barley leaf cuttings, after treatment of plants with a known herbicide (norflurazon at 200 µM) that blocks the conversion of this intermediate into end-of-chain carotenoids158. By this approach compounds that interfere with the accumulation of phytoene at steps preceding phytoene synthase could be detected, which was verified by applying representative inhibitors like derivatives of clomazone. However, the procedure required the rather tedious extraction and separation from lipids by saponification, another solvent extraction from the MeOH/KOH solution and finally the photometric quantification158.\n\nAfter some problems with the maintenance of culture conditions due to a failure of the temperature control system of the growth chamber, we observed a sudden loss of all the mRFP fluorescence in one of the double-transformed cell lines, whereas the level of GFP fluorescence remained completely untouched (Figure 16). After elimination of all evident sources of error (replacing the inducer, subculturing the two-week old line, replacing the medium), the culture was incubated over a whole 7-day-growth cycle in presence of 10 µM 5-azacytidine, a nucleotide analog that cannot be methylated, and, remarkably, the mRFP fluorescence could be partially restored (Figure 17). As this result clearly suggested a DNA methylation event, we screened the literature for common sources of such sudden drops in gene expression levels of transgenes in plant cultures. As we had already observed the same phenomenon in the original H6-GFP-BD-CVIL cell lines during the inhibitor tests with fosmidomycin-derived prodrugs, this point was quite important, given all the effort put into the generation of the cell lines. Schmitt et al.159 reported that the antibiotics kanamycin, hygromycin and cefotaxim caused a DNA hypermethylation at CpG sites in the genome of tobacco plants grown in vitro, as shown by the SssI methylase accepting assays and genomic sequencing with sodium bisulfite. Interestingly, these methylations occurred in a time and dose-dependent manner and were not reversed when the progeny was not grown anymore in the presence of the antibiotics160.\n\nA- and C- Heterogeneity of GFP and RFP fluorescence in a suspension of transgenic BY-2 cells derived from primary calli. Arrows indicate cells with missing fluorescence or significant variations in fluorescence intensities (red and green fluorescence). B- and D- Cell suspensions derived from re-selected calli (secondary calli). The fluorescence is strong and homogenous in both channels. Nevertheless some cells (less than 5%) show heterogeneity in fluorescence. Possible reasons are discussed in the main text. White bars indicate 20 μm.\n\nA: Double fluorescent cell line that suddenly lost mRFP expression. B and C: Cells were treated with 10 µM azacytidine for one week. RFP fluorescence could be recovered, but many dead cells were due to overall toxic effects. Even though the concentration was scaled down 10–20x compared to values that are indicated in the literature for treatments, the concentration still seemed to be too high for the use in tobacco BY-2 cells (which have a good uptake rate and fast metabolism). All images are shown as merged images taken in green and red fluorescence, as well as white light mode. White bars = 50 µm.\n\nThe methylation of plant genomes is a common process, which can affect up to 30% of the cytosine residues161. It also occurs as part of “natural” gene regulation in plants162,163. However, increased DNA methylation was observed in several cases associated with PTGS (post-transcriptional gene silencing) and TGS (transcriptional gene silencing) or different forms of stress164–167. For instance, transgene silencing was induced in Petunia, after a period of high light intensity and temperature168, whereas high temperatures alone were shown sufficient to silence different transgenes in tobacco169–171. Thus, major breakdowns of the air-conditioning system of our growth chambers (several weeks during summer, at > 27°C), might well have contributed to the observed silencing. However, although recovery of the mRFP expression by azacytidine treatment had confirmed our assumption that the sudden loss of fluorescence, due to a DNA methylation event, could explain the heterogeneity of the cells, we did not maintain and subculture the recovered lines, as azacytidine is a powerful mutagenic agent172, which might exert pleiotropic effects on the cells.\n\n\nMaterials and methods\n\nThe sources of chemicals and biochemicals were already described in the preceding article32, as well as the basic methods like cultivation of tobacco BY-2 cells, their stable transformation by a gene encoding H6-BD-GFP-CVIL placed under the control by a dexamethazone-inducible promoter. This also holds true for microscopic techniques. Only what methods had been newly introduced in this study will be described here.\n\nDXR and HMGR artificial microRNA silencing strategy. Artificial microRNAs (amiRNA) designed to silence DXR and HMGR genes were predicted using the Web MicroRNA designer WMD2. For DXR, we used the Nicotiana tabacum DXR cDNA sequence (Genbank accession number DQ839130) as a template and four potential miRNA were selected. For the design of amiRNA capable of silencing all of the HMG-CoA reductase (HMGR) genes HMGR1- NTU60452, HMGR2- AF004232, HMGRL- AF004233, looked for the regions with highest homology between all of the isoforms and submitted it to WMD2. In this case, two potential miRNA were selected. The selected amiRNAs sequences were amplified by recursive PCR as described by Schwab et al.104 using the primers depicted in Table 3 and Table 4, and cloned into pBSK.\n\nThe amiRNA sequences were subcloned into the pER10 inducible vector86, kindly provided by Prof. Nam-Hai Chua, Rockefeller University) by XhoI-SpeI and transformed into an Agrobacterium tumefaciens LBA4404 hypervirulent strain. Agrobacterium transformants were checked by PCR (using oligos pER for 5’- GCTCGACTCTAGGATCTTCG - 3’ and pER rev 5’- GTAGGATTCTGGTGTGTGG-3’) and used for stable transformation of BY-2 cell lines.\n\nStable BY-2 cell lines expressing the construct H6-GFP-BD-CVIL were previously reselected for good fluorescent intensity. 3-day-old selected cell suspensions were transformed through co-culture with pER10-DXR or pER10-HMGR transformed agrobacteria and plated into MS-hygromycin/kanamycin (30 µg/ml each). Plates were incubated in the dark at 28°C until calli appeared. At least 10 transformed calli were re-selected three times through fresh MS plates (supplemented with hygromycin and kanamycin, both at 30 µg/ml) and derived minicultures were sub-cultured weekly (10 ml, MS supplemented with hygromycin and kanamycin, both at 30 µg/ml). Genomic DNA was extracted from each BY-2 clone and t-DNA insertion was checked by PCR (using oligos pER for and pER rev). For fluorescence screening we first treated the cells with 5 µM estradiol (Est) (to induce amiRNA expression) and/or 50 µM fosmidomycin (FOS), as control of the expected phenotype. After 2 h, GFP fluorescence was induced by addition of Dex at 10 µM final concentration and cell cultures were screened by confocal laser scanning microscopy for lines showing high levels of GFP fluorescence and also exhibiting a DXR silencing phenotype (e.g. mislocalization of the H6-GFP-BD-CVIL from the plasma membrane to the nucleus). HMGR silenced lines were selected among those lines showing high levels of GFP fluorescence at the plasma membrane. Selected lines were subsequently used for inhibition tests, including treatments with EST, FOS or mevinolin (MEV).\n\nAs a control of such silencing experiments, stably transformed lines were generated with the pER10 empty vector. Induction by EST did not result in mislocalization of H6-GFP-BD-CVIL from the plasma membrane to nuclei, as expected.", "appendix": "Author contributions\n\n\n\nMichael Hartmann was responsible for the experimental design and carried out the majority of analyses with confocal microscopy, especially including the development of a semi-automated screening method based on culture of cells in 96-well microtiter plates. In addition he transformed, selected and maintained BY-2 cell lines. He generated the Figures, wrote major parts of the initial manuscript as part of his PhD thesis and contributed to the final version of the manuscript. Andréa Hemmerlin helped to transform, select and maintain original BY-2 cell lines. Elisabet Gas-Pascual was responsible of all amiRNA experiments and the selection of suitable BY-2 cell lines. Michel Rohmer was instrumental in the interpretation of results and had designed inhibitors of MEP pathway enzymes. Thomas J. Bach initiated and supervised all those studies with the GFP fusion protein expressing BY-2 cells and their use as to the cross-talk between the cytoplasmic MVA and the plastidial MEP pathways. He brought the manuscript into its updated and final form. The project, especially inhibitor testing, represented an essential part of the close collaboration between the research groups of Michel Rohmer and Thomas J. Bach. All authors critically revised the manuscript and agreed the final version for publication.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe are grateful for a Ph.D. fellowship attributed to Michael Hartmann provided by the Région Alsace. The research conducted in the laboratories of T.J. Bach and M. Rohmer was supported by grants from the Agence Nationale de la Recherche (ANR-05-BLAN-0217-01 and -02 TERPENE; ANR-06-BLAN-0291-02 BIOSIS). General support was also provided by the Centre National de la Recherché Scientifique (CNRS) and the Université de Strasbourg.\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 wish to thank Prof. Nam-Hai Chua (Rockefeller University, New York) for providing us with dexamethasone- and estrogen-inducible vectors. Thanks go to Dr. Jérôme Mutterer (Microscopy Platform at the IBMP) for his advices and the maintenance of confocal microscopes and corresponding software. Prof. T. Nagata, Tokyo University, provided the BY-2 cell line to the IBMP many years ago, which is greatly appreciated.\n\n\nReferences\n\nClarke S: Protein isoprenylation and methylation at carboxy-terminal cysteine residues. Annu Rev Biochem. 1992; 61: 355–386. PubMed Abstract \| Publisher Full Text\n\nSchafer WR, Rine J: Protein prenylation: genes, enzymes, targets and functions. Annu Rev Genet. 1992; 26: 209–237. PubMed Abstract \| Publisher Full Text\n\nZhang FL, Casey PJ: Protein prenylation: molecular mechanisms and functional consequences. Annu Rev Biochem. 1996; 65: 241–269. PubMed Abstract \| Publisher Full Text\n\nCrowell DN: Functional implications of protein isoprenylation in plants. Prog Lipid Res. 2000; 39(5): 393–408. PubMed Abstract \| Publisher Full Text\n\nRoskoski R Jr: Protein prenylation: a pivotal posttranslational process. Biochem Biophys Res Commun. 2003; 303(1): 1–7. PubMed Abstract \| Publisher Full Text\n\nYalovsky S, Rodr Guez-Concepción M, Gruissem W: Lipid modification of proteins-slipping in and out of membranes. Trends Plant Sci. 1999; 4(11): 439–445. PubMed Abstract \| Publisher Full Text\n\nMaurer-Stroh S, Washietl S, Eisenhaber F: Protein prenyltransferases. Genome Biol. 2003; 4(4): 212. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaurer-Stroh S, Washietl S, Eisenhaber F: Protein prenyltransferases: anchor size, pseudogenes and parasites. Biol Chem. 2003; 384(7): 977–989. PubMed Abstract \| Publisher Full Text\n\nGalichet A, Gruissem W: Protein farnesylation in plants--conserved mechanisms but different targets. Curr Opin Plant Biol. 2003; 6(6): 530–535. PubMed Abstract \| Publisher Full Text\n\nTan EW, Rando RR: Identification of an isoprenylated cysteine methyl ester hydrolase activity in bovine rod outer segment membranes. Biochemistry. 1992; 31(24): 5572–5578. PubMed Abstract \| Publisher Full Text\n\nVan Dessel GA, De Busser HM, Lagrou AR: On the occurrence of multiple isoprenylated cysteine methyl ester hydrolase activities in bovine adrenal medulla. Biochem Biophys Res Commun. 2001; 284(1): 50–56. PubMed Abstract \| Publisher Full Text\n\nLamango NS: Liver prenylated methylated protein methyl esterase is an organophosphate-sensitive enzyme. J Biochem Mol Toxicol. 2005; 19(5): 347–357. PubMed Abstract \| Publisher Full Text\n\nDeem AK, Bultema RL, Crowell DN: Prenylcysteine methylesterase in Arabidopsis thaliana. Gene. 2006; 380(2): 159–166. PubMed Abstract \| Publisher Full Text\n\nHuizinga DH, Omosegbon O, Omery B, et al.: Isoprenylcysteine methylation and demethylation regulate abscisic acid signaling in Arabidopsis. Plant Cell. 2008; 20(10): 2714–2728. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYoung SG, Ambroziak P, Kim E, et al.: Postprenylation protein processing: CXXX (CaaX) endoproteases and isoprenylcysteine carboxyl methyltransferase. In: Tamanoi F, Sigman DS, editors. Protein Lipidation. San Diego. 2000; pp. 153–213. Reference Source\n\nHancock JF, Magee AI, Childs JE, et al.: All ras proteins are polyisoprenylated but only some are palmitoylated. Cell. 1989; 57(7): 1167–1177. PubMed Abstract \| Publisher Full Text\n\nHancock JF, Paterson H, Marshall CJ: A polybasic domain or palmitoylation is required in addition to the CAAX motif to localize p21ras to the plasma membrane. Cell. 1990; 63(1): 133–139. PubMed Abstract \| Publisher Full Text\n\nHancock JF, Cadwallader KA, Paterson H, et al.: A CAAX or a CAAL motif and a second signal are sufficient for plasma membrane targeting of RAS proteins. EMBO J. 1991; 10(13): 4033–4039. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSorek N, Poraty L, Sternberg H, et al.: Activation status-coupled transient S acylation determines membrane partitioning of a plant Rho-related GTPase. Mol Cell Biol. 2007; 27(6): 2144–2154. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBos JL: Ras oncogenes in human cancer: a review. Cancer Res. 1989; 49(17): 4682–4689. PubMed Abstract\n\nDownward J: Targeting RAS signalling pathways in cancer therapy. Nat Rev Cancer. 2003; 3(1): 11–22. PubMed Abstract \| Publisher Full Text\n\nMalumbres M, Barbacid M: RAS oncogenes: the first 30 years. Nat Rev Cancer. 2003; 3(6): 459–465. PubMed Abstract \| Publisher Full Text\n\nYalovsky S, Kulukian A, Rodríguez-Concepción M, et al.: Functional requirement of plant farnesyltransferase during development in Arabidopsis. Plant Cell. 2000; 12(8): 1267–1278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGalichet A, Gruissem W: Developmentally controlled farnesylation modulates AtNAP1; 1 function in cell proliferation and cell expansion during Arabidopsis leaf development. Plant Physiol. 2006; 142(4): 1412–1426. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGalichet A, Hoyerová K, Kamínek M, et al.: Farnesylation directs AtIPT3 subcellular localization and modulates cytokinin biosynthesis in Arabidopsis. Plant Physiol. 2008; 146(3): 1155–1164. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrowell DN, Huizinga DH: Protein isoprenylation: the fat of the matter. Trends Plant Sci. 2009; 14(3): 163–170. PubMed Abstract \| Publisher Full Text\n\nAndrews M, Huizinga DH, Crowell DN: The CaaX specificities of Arabidopsis protein prenyltransferases explain era1 and ggb phenotypes. BMC Plant Biol. 2010; 10: 118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodríguez-Concepción M, Yalovsky S, Zik M, et al.: The prenylation status of a novel plant calmodulin directs plasma membrane or nuclear localization of the protein. EMBO J. 1999; 18(7): 1996–2007. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodríguez-Concepción M, Toledo-Ortiz G, Yalovsky S, et al.: Carboxyl-methylation of prenylated calmodulin CaM53 is required for efficient plasma membrane targeting of the protein. Plant J. 2000; 24(6): 775–784. PubMed Abstract \| Publisher Full Text\n\nGerber E, Hemmerlin A, Hartmann M, et al.: The plastidial 2-C-methyl-D-erythritol 4-phosphate pathway provides the isoprenyl moiety for protein geranylgeranylation in tobacco BY-2 cells. Plant Cell. 2009; 21(1): 285–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGerber E, Hemmerlin A, Bach TJ, et al.: The role of plastids in protein geranylgeranylation in tobacco BY-2 cells. In: The Chloroplast: Basics and Applications, CA Rebeiz et al. (eds), Advances in Photosynthesis and Respiration, Springer Science + Business Media B.V. 2010; 31: 127–137. Publisher Full Text\n\nHartmann M, Hemmerlin A, Gas-Pascual E, et al.: The effect of MEP pathway and other inhibitors on the intracellular localization of a plasma membrane-targeted, isoprenylable GFP reporter protein in tobacco BY-2 cells [v2; ref status: indexed, http://f1000r.es/2af]. F1000Res. 2013; 2: 170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMüller C, Schwender J, Zeidler J, et al.: Properties and inhibition of the first two enzymes of the non-mevalonate pathway of isoprenoid biosynthesis. Biochem Soc Trans. 2000; 28(6): 792–793. PubMed Abstract \| Publisher Full Text\n\nZeidler J, Schwender J, Müller C, et al.: The non-mevalonate isoprenoid biosynthesis of plants as a test system for drugs against malaria and pathogenic bacteria. Biochem Soc Trans. 2000; 28(6): 796–798. PubMed Abstract \| Publisher Full Text\n\nKuzuyama T, Shimizu T, Takahashi S, et al.: Fosmidomycin, a specific inhibitor of 1-deoxy-D-xylulose 5-phosphate reductoisomerase in the nonmevalonate pathway for terpenoid biosynthesis. Tetrahedron Lett. 1998; 39(43): 7913–7916. Publisher Full Text\n\nZeidler J, Schwender J, Müller C, et al.: Inhibition of the non-mevalonate 1-deoxy-D-xylulose-5-phosphate pathway of plant isoprenoid biosynthesis by fosmidomycin. Z Naturforsch C. 1998; 53: 980–986. Reference Source\n\nCarpenter AE: Software opens the door to quantitative imaging. Nat Methods. 2007; 4(2): 120–121. PubMed Abstract \| Publisher Full Text\n\nCarpenter AE: Image-based chemical screening. Nat Chem Biol. 2007; 3(8): 461–465. PubMed Abstract \| Publisher Full Text\n\nUetz P, Hughes RE: Systematic and large-scale two-hybrid screens. Curr Opin Microbiol. 2000; 3(3): 303–308. PubMed Abstract \| Publisher Full Text\n\nRual JF, Venkatesan K, Hao T, et al.: Towards a proteome-scale map of the human protein-protein interaction network. Nature. 2005; 437(7062): 1173–1178. PubMed Abstract \| Publisher Full Text\n\nFoster LJ, de Hoog CL, Zhang Y, et al.: A mammalian organelle map by protein correlation profiling. Cell. 2006; 125(1): 187–199. PubMed Abstract \| Publisher Full Text\n\nGavin AC, Aloy P, Grandi P, et al.: Proteome survey reveals modularity of the yeast cell machinery. Nature. 2006; 440(7084): 631–636. PubMed Abstract \| Publisher Full Text\n\nHughes TR, Marton MJ, Jones AR, et al.: Functional discovery via a compendium of expression profiles. Cell. 2000; 102(1): 109–126. PubMed Abstract \| Publisher Full Text\n\nSmith MG, Jona G, Ptacek J, et al.: Global analysis of protein function using protein microarrays. Mech Ageing Dev. 2005; 126(1): 171–175. PubMed Abstract \| Publisher Full Text\n\nStarkuviene V, Pepperkok R, Erfle H: Transfected cell microarrays: an efficient tool for high-throughput functional analysis. Expert Rev Proteomics. 2007; 4(4): 479–489. PubMed Abstract \| Publisher Full Text\n\nCarpenter AE, Sabatini DM: Systematic genome-wide screens of gene function. Nat Rev Genet. 2004; 5(1): 11–22. PubMed Abstract \| Publisher Full Text\n\nBaum B, Craig G: RNAi in a postmodern, postgenomic era. Oncogene. 2004; 23(51): 8336–8339. PubMed Abstract \| Publisher Full Text\n\nMoffat J, Sabatini DM: Building mammalian signalling pathways with RNAi screens. Nat Rev Mol Cell Biol. 2006; 7(3): 177–187. PubMed Abstract \| Publisher Full Text\n\nTerjung S, Walter T, Seitz A, et al.: High-throughput microscopy using live mammalian cells. Cold Spring Harb Protoc. 2010; 2010(7): pdb–p84. PubMed Abstract \| Publisher Full Text\n\nZhang J, Campbell RE, Ting AY, et al.: Creating new fluorescent probes for cell biology. Nat Rev Mol Cell Biol. 2002; 3(12): 906–918. PubMed Abstract \| Publisher Full Text\n\nVerkhusha VV, Lukyanov KA: The molecular properties and applications of Anthozoa fluorescent proteins and chromoproteins. Nat Biotechnol. 2004; 22(3): 289–296. PubMed Abstract \| Publisher Full Text\n\nMiyawaki A, Nagai T, Mizuno H: Engineering fluorescent proteins. Adv Biochem Eng Biotechnol. 2005; 95: 1–15. PubMed Abstract\n\nDay RN, Davidson MW: The fluorescent protein palette: tools for cellular imaging. Chem Soc Rev. 2009; 38(10): 2887–2921. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoss U, Larrieu A, Wells DM: From jellyfish to biosensors: the use of fluorescent proteins in plants. Int J Dev Biol. 2013; 57(6–8): 525–533. PubMed Abstract \| Publisher Full Text\n\nEggert US, Kiger AA, Richter C, et al.: Parallel chemical genetic and genome-wide RNAi screens identify cytokinesis inhibitors and targets. PLoS Biol. 2004; 2(12): e379. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLang P, Yeow K, Nichols A, et al.: Cellular imaging in drug discovery. Nat Rev Drug Discov. 2006; 5(4): 343–356. PubMed Abstract \| Publisher Full Text\n\nLamprecht MR, Sabatini DM, Carpenter AE: CellProfiler: free, versatile software for automated biological image analysis. Biotechniques. 2007; 42(1): 71–75. PubMed Abstract \| Publisher Full Text\n\nWouters FS, Verveer PJ, Bastiaens PI: Imaging biochemistry inside cells. Tends Cell Biol. 2001; 11(5): 203–211. PubMed Abstract \| Publisher Full Text\n\nWatson P, Jones AT, Stephens DJ: Intracellular trafficking pathways and drug delivery: fluorescence imaging of living and fixed cells. Adv Drug Deliv Rev. 2005; 57(1): 43–61. PubMed Abstract \| Publisher Full Text\n\nBurnett P, Robertson JK, Palmer JM, et al.: Fluorescence imaging of electrically stimulated cells. J Biomol Screen. 2003; 8(6): 660–667. PubMed Abstract \| Publisher Full Text\n\nWu CC, Reilly JF, Young WG, et al.: High-throughput morphometric analysis of individual neurons. Cereb Cortex. 2004; 14(5): 543–554. PubMed Abstract \| Publisher Full Text\n\nGiuliano KA, DeBiasio RL, Dunlay RT, et al.: High-content screening: a new approach to easing key bottlenecks in the drug discovery process. J Biomol Screen. 1997; 2: 249–259. Publisher Full Text\n\nGiuliano KA, Haskins JR, Taylor DL: Advances in high content screening for drug discovery. Assay Drug Dev Technol. 2003; 1(4): 565–577. PubMed Abstract \| Publisher Full Text\n\nStarkuviene V, Pepperkok R: The potential of high-content high-throughput microscopy in drug discovery. Br J Pharmacol. 2007; 152(1): 62–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh S, Carpenter AE, Genovesio A: Increasing the Content of High-Content Screening: An Overview. J Biomol Screen. 2014; 19(5): 640–650. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLjosa V, Caie PD, Ter Horst R, et al.: Comparison of methods for image-based profiling of cellular morphological responses to small-molecule treatment. J Biomol Screen. 2013; 18(10): 1321–1329. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZanella F, Lorens JB, Link W: High content screening: seeing is believing. Trends Biotechnol. 2010; 28(5): 237–245. PubMed Abstract \| Publisher Full Text\n\nGough AH, Johnston PA: Requirements, Features, and Performance of High Content Screening Platforms. In High Content Screening, DL Taylor, JR Haskins and KA Giuliano, eds. (Humana Press). 2006; pp. 41–61. Reference Source\n\nHoffman AF, Garippa RJ: A pharmaceutical company user’s perspective on the potential of high content screening in drug discovery. Methods Mol Biol. 2007; 356: 19–31. PubMed Abstract \| Publisher Full Text\n\nGururaja TL, Goff D, Kinoshita T, et al.: R-253 disrupts microtubule networks in multiple tumor cell lines. Clin Cancer Res. 2006; 12(12): 3831–3842. PubMed Abstract \| Publisher Full Text\n\nRichards GR, Smith AJ, Parry F, et al.: A morphology- and kinetics-based cascade for human neural cell high content screening. Assay Drug Dev Technol. 2006; 4(2): 143–152. PubMed Abstract \| Publisher Full Text\n\nWilson CJ, Si Y, Thompsons CM, et al.: Identification of a small molecule that induces mitotic arrest using a simplified high-content screening assay and data analysis method. J Biomol Screen. 2006; 11(1): 21–28. PubMed Abstract \| Publisher Full Text\n\nSmith C, Eisenstein M: Automated imaging: data as far as the eye can see (Technology Feature). Nat Methods. 2005; 2: 547–555. Publisher Full Text\n\nLevenson RM, Mansfield JR, Tsien RY, et al.: Multiplexing fluorescent agents with multispectral imaging: microscopy and in-vivo examples. AACR Meeting Abstracts. 2005; 2005: 906–907. Reference Source\n\nMansfield JR, Gossage KW, Hoyt CC, et al.: Autofluorescence removal, multiplexing, and automated analysis methods for in-vivo fluorescence imaging. J Biomed Opt. 2005; 10(4): 41207. PubMed Abstract \| Publisher Full Text\n\nZimmermann T: Spectral imaging and linear unmixing in light microscopy. Rietdorf J editor. Springer-Verlag, Heidelberg. 2005; 95. : 245–265. Publisher Full Text\n\nZimmermann T, Rietdorf J, Pepperkok R: Spectral imaging and its applications in live cell microscopy. FEBS Lett. 2003; 546(1): 87–92. PubMed Abstract \| Publisher Full Text\n\nPerlman ZE, Slack MD, Feng Y, et al.: Multidimensional drug profiling by automated microscopy. Science. 2004; 306(5699): 1194–1198. PubMed Abstract \| Publisher Full Text\n\nTanaka M, Bateman R, Rauh D, et al.: An unbiased cell morphology-based screen for new biologically active small molecules. PLoS Biol. 2005; 3(5): e128. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKola I, Landis J: Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov. 2004; 3(8): 711–715. PubMed Abstract \| Publisher Full Text\n\nAbramoff MD, Magelhaes PJ, Ram SJ: Image processing with ImageJ. Biophotonics Int. 2004; 11(7): 36–42. Reference Source\n\nCarpenter AE, Jones TR, Lamprecht MR, et al.: CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol. 2006; 7(10): R100. PubMed Abstract \| Free Full Text\n\nJones TR, Kang IH, Wheeler DB, et al.: CellProfiler Analyst: data exploration and analysis software for complex image-based screens. BMC Bioinformatics. 2008; 9: 482. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMisselwitz B, Strittmatter G, Periaswamy B, et al.: Enhanced CellClassifier: a multi-class classification tool for microscopy images. BMC Bioinformatics. 2010; 11: 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoerdink JB, Meijster A: The watershed transform: definitions, algorithms and parallelization strategies. Fundam Inform. 2001; 41: 187–228. Reference Source\n\nZuo J, Niu QW, Chua NH: Technical advance: An estrogen receptor-based transactivator XVE mediates highly inducible gene expression in transgenic plants. Plant J. 2000; 24(2): 265–273. PubMed Abstract \| Publisher Full Text\n\nKalderon D, Roberts BL, Richardson WD, et al.: A short amino acid sequence able to specify nuclear location. Cell. 1984; 39(3 Pt 2): 499–509. PubMed Abstract \| Publisher Full Text\n\nKalderon D, Richardson WD, Markham AF, et al.: Sequence requirements for nuclear location of simian virus 40 large-T antigen. Nature. 1984; 311(5981): 33–38. PubMed Abstract \| Publisher Full Text\n\nRaikhel N: Nuclear targeting in plants. Plant Physiol. 1992; 100(4): 1627–1632. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHicks GR, Smith HM, Shieh M, et al.: Three classes of nuclear import signals bind to plant nuclei. Plant Physiol. 1995; 107(4): 1055–1058. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXiao CY, Jans DA: An engineered site for protein kinase C flanking the SV40 large T-antigen NLS confers phorbol ester-inducible nuclear import. FEBS Lett. 1998; 436(3): 313–317. PubMed Abstract \| Publisher Full Text\n\nLaunholt D, Merkle T, Houben A, et al.: Arabidopsis chromatin-associated HMGA and HMGB use different nuclear targeting signals and display highly dynamic localization within the nucleus. Plant Cell. 2006; 18(11): 2904–2918. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee YJ, Kim DH, Kim YW, et al.: Identification of a signal that distinguishes between the chloroplast outer envelope membrane and the endomembrane system in vivo. Plant Cell. 2001; 13(10): 2175–2190. PubMed Abstract \| Free Full Text\n\nCampbell RE, Tour O, Palmer AE, et al.: A monomeric red fluorescent protein. Proc Natl Acad Sci U S A. 2002; 99(12): 7877–7882. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMatz MV, Fradkov AF, Labas YA, et al.: Fluorescent proteins from nonbioluminescent Anthozoa species. Nat Biotechnol. 1999; 17(10): 969–973. PubMed Abstract \| Publisher Full Text\n\nChiu W, Niwa Y, Zeng W, et al.: Engineered GFP as a vital reporter in plants. Curr Biol. 1996; 6(3): 325–330. PubMed Abstract \| Publisher Full Text\n\nShimomura O, Johnson FH, Saiga Y: Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aequorea. J Cell Comp Physiol. 1962; 59: 223–239. PubMed Abstract \| Publisher Full Text\n\nPrendergast FG, Mann KG: Chemical and physical properties of aequorin and the green fluorescent protein isolated from Aequorea forskalea. Biochemistry. 1978; 17(17): 3448–3453. PubMed Abstract \| Publisher Full Text\n\nTsien RY: The green fluorescent protein. Annu Rev Biochem. 1998; 67: 509–544. PubMed Abstract \| Publisher Full Text\n\nPrasher DC, Eckenrode VK, Ward WW, et al.: Primary structure of the Aequorea victoria green-fluorescent protein. Gene. 1992; 111(2): 229–233. PubMed Abstract \| Publisher Full Text\n\nChiu W, Niwa Y, Zeng W, et al.: Engineered GFP as a vital reporter in plants. Curr Biol. 1996; 6(3): 325–330. PubMed Abstract \| Publisher Full Text\n\nKovarik A, Fojtova M: Estimation of viable cell count after fluorescein diacetate staining using phosphorimager analysis. Biotechniques. 1999; 27(4): 685–688. PubMed Abstract\n\nNocarova E, Fischer L: Cloning of transgenic tobacco BY-2 cells: an efficient method to analyse and reduce high natural heterogeneity of transgene expression. BMC Plant Biol. 2009; 9: 44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchwab R, Ossowski S, Riester M, et al.: Highly specific gene silencing by artificial microRNAs in Arabidopsis. Plant Cell. 2006; 18(5): 1121–1133. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrodersen P, Sakvarelidze-Achard L, Schaller H, et al.: Isoprenoid biosynthesis is required for miRNA function and affects membrane association of ARGONAUTE 1 in Arabidopsis. Proc Natl Acad Sci U S A. 2012; 109(5): 1778–1783. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiron Y, Paran Y, Zatorsky NG, et al.: Laser autofocusing system for high-resolution cell biological imaging. J Microsc. 2006; 221(Pt 2): 145–151. PubMed Abstract \| Publisher Full Text\n\nSantos A, Ortiz de Solórzano C, Vaquero JJ, et al.: Evaluation of autofocus functions in molecular cytogenetic analysis. J Microsc. 1997; 188(Pt 3): 264–272. PubMed Abstract \| Publisher Full Text\n\nGordon A, Colman-Lerner A, Chin TE, et al.: Single-cell quantification of molecules and rates using open-source microscope-based cytometry. Nat Methods. 2007; 4(2): 175–181. PubMed Abstract \| Publisher Full Text\n\nDunn KW, Wang E: Optical aberrations and objective choice in multicolor confocal microscopy. Biotechniques. 2000; 28(3): 542–4, 546, 548–50. PubMed Abstract\n\nMayr LM, Fuerst P: The future of high-throughput screening. J Biomol Screen. 2008; 13(6): 443–448. PubMed Abstract \| Publisher Full Text\n\nHouston JG, Banks MN, Binnie A, et al.: Case study: impact of technology investment on lead discovery at Bristol-Myers Squibb, 1998–2006. Drug Discov Today. 2008; 13(1–2): 44–51. PubMed Abstract \| Publisher Full Text\n\nHermann R, Lehmann M, Buchs J: Characterization of gas-liquid mass transfer phenomena in microtiter plates. Biotechnol Bioeng. 2003; 81(2): 178–186. PubMed Abstract \| Publisher Full Text\n\nDuetz WA, Ruedi L, Hermann R, et al.: Methods for intense aeration, growth, storage, and replication of bacterial strains in microtiter plates. Appl Environ Microbiol. 2000; 66(6): 2641–2646. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBetts JI, Baganz F: Miniature bioreactors: current practices and future opportunities. Microb Cell Fact. 2006; 5: 21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLevsky JM, Singer RH: Gene expression and the myth of the average cell. Trends Cell Biol. 2003; 13(1): 4–6. PubMed Abstract \| Publisher Full Text\n\nElowitz MB, Levine AJ, Siggia ED, et al.: Stochastic gene expression in a single cell. Science. 2002; 297(5584): 1183–1186. PubMed Abstract \| Publisher Full Text\n\nRaser JM, O’Shea EK: Control of stochasticity in eukaryotic gene expression. Science. 2004; 304(5678): 1811–1814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPedraza JM, Van Oudenaarden A: Noise propagation in gene networks. Science. 2005; 307(5717): 1965–1969. PubMed Abstract \| Publisher Full Text\n\nColman-Lerner A, Gordon A, Serra E, et al.: Regulated cell-to-cell variation in a cell-fate decision system. Nature. 2005; 437(7059): 699–706. PubMed Abstract \| Publisher Full Text\n\nNagata T, Nemoto Y, Hasezawa S: Tobacco BY-2 cell line as the “Hela” cell in the cell biology of higher plants. Int Rev Cytol. 1992; 132: 1–30. Publisher Full Text\n\nNagata T, Sakamoto K, Shimizu T: Tobacco BY-2 cells: The present and beyond. In Vitro Cell Dev Biol - Plant. 2004; 40(2): 163–166. Publisher Full Text\n\nHemmerlin A, Bach TJ: Effects of mevinolin on cell cycle progression and viability of tobacco BY-2 cells. Plant J. 1998; 14: 65–74. PubMed Abstract \| Publisher Full Text\n\nKumagai-Sano F, Hayashi T, Sano T, et al.: Cell cycle synchronization of tobacco BY-2 cells. Nat Protoc. 2006; 1(6): 2621–2627. PubMed Abstract \| Publisher Full Text\n\nKuthanova A, Fischer L, Nick P, et al.: Cell cycle phase-specific death response of tobacco BY-2 cell line to cadmium treatment. Plant Cell Environ. 2008; 31(11): 1634–1643. PubMed Abstract \| Publisher Full Text\n\nWentzinger LF, Bach TJ, Hartmann MA: Inhibition of squalene synthase and squalene epoxidase in tobacco cells triggers an up-regulation of 3-hydroxy-3-methylglutaryl coenzyme a reductase. Plant Physiol. 2002; 130(1): 334–346. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHemmerlin A, Gerber E, Feldtrauer JF, et al.: A review of tobacco BY-2 cells as an excellent system to study the synthesis and function of sterols and other isoprenoids. Lipids. 2004; 39(8): 723–735. PubMed Abstract\n\nVerpoorte R, Memelink J: Engineering secondary metabolite production in plants. Curr Opin Biotechnol. 2002; 13(2): 181–187. PubMed Abstract \| Publisher Full Text\n\nGeorgiev M, Pavlov A, Ilieva M: Selection of high rosmarinic acid producing Lavandula vera MM cell lines. Process Biochem. 2006; 41(9): 2068–2071. Publisher Full Text\n\nGeorgiev MI, Weber J, Maciuk A: Bioprocessing of plant cell cultures for mass production of targeted compounds. Appl Microbiol Biotechnol. 2009; 83(5): 809–823. PubMed Abstract \| Publisher Full Text\n\nProls F, Meyer P: The methylation patterns of chromosomal integration regions influence gene activity of transferred DNA in Petunia hybrida. Plant J. 1992; 2(4): 465–475. PubMed Abstract \| Publisher Full Text\n\nIglesias VA, Moscone EA, Papp I, et al.: Molecular and cytogenetic analyses of stably and unstably expressed transgene loci in tobacco. Plant Cell. 1997; 9(8): 1251–1264. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Leeuwen W, Ruttink T, Borst-Vrenssen AW, et al.: Characterization of position-induced spatial and temporal regulation of transgene promoter activity in plants. J Exp Bot. 2001; 52(358): 949–959. PubMed Abstract \| Publisher Full Text\n\nSchubert D, Lechtenberg B, Forsbach A, et al.: Silencing in Arabidopsis T-DNA transformants: the predominant role of a gene-specific RNA sensing mechanism versus position effects. Plant Cell. 2004; 16(10): 2561–2572. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrancis KE, Spiker S: Identification of Arabidopsis thaliana transformants without selection reveals a high occurrence of silenced T-DNA integrations. Plant J. 2005; 41(3): 464–477. PubMed Abstract \| Publisher Full Text\n\nFischer U, Kuhlmann M, Pecinka A, et al.: Local DNA features affect RNA-directed transcriptional gene silencing and DNA methylation. Plant J. 2008; 53(1): 1–10. PubMed Abstract \| Publisher Full Text\n\nMüller E, Lörz H, Lütticke S: Variability of transgene expression in clonal cell lines of wheat. Plant Science. 1996; 114(1): 71–82. Publisher Full Text\n\nDown RE, Ford L, Bedford SJ, et al.: Influence of plant development and environment on transgene expression in potato and consequences for insect resistance. Transgenic Res. 2001; 10(3): 223–236. PubMed Abstract \| Publisher Full Text\n\nChristman JK: 5-Azacytidine and 5-aza-2'-deoxycytidine as inhibitors of DNA methylation: mechanistic studies and their implications for cancer therapy. Oncogene. 2002; 21(35): 5483–5495. PubMed Abstract \| Publisher Full Text\n\nScragg AH: The problems associated with high biomass levels in plant cell suspensions. Plant Cell Tissue Organ Cult. 1995; 43(2): 163–170. Publisher Full Text\n\nKieran PM, MacLoughlin PF, Malone DM: Plant cell suspension cultures: some engineering considerations. J Biotechnol. 1997; 59(1–2): 39–52. PubMed Abstract \| Publisher Full Text\n\nAbdullah MA, Ali AM, Lajis NH, et al.: Issues in plant cell culture engineering for enhancement of productivity. Dev Chem Eng Mineral Process. 2005; 13(5–6): 573–587. Publisher Full Text\n\nGerber E: Localisation cellulaire de protéines fluorescentes isoprénylables dans les cellules de tabac BY-2. Thèse doctorale, Université Louis Pasteur (ULP) Strasbourg, France. 2005. Reference Source\n\nTen Hoopen HJG, Vinke JL, Moreno PRH, et al.: Influence of temperature on growth and ajmalicine production by Catharanthus roseus suspension cultures. Enzyme Microb Technol. 2002; 30(1): 56–65. Publisher Full Text\n\nGeorgiev M, Pavlov A, Ilieva M: Rosmarinic acid production by Lavandula vera MM cell suspension: the effect of temperature. Biotechnol Lett. 2004; 26(10): 855–856. PubMed Abstract \| Publisher Full Text\n\nJames GL, Goldstein JL, Brown MS: Polylysine and CVIM sequences of K-RasB dictate specificity of prenylation and confer resistance to benodiazepine peptidomimetic in vitro. J Biol Chem. 1995; 270(11): 6221–6226. PubMed Abstract \| Publisher Full Text\n\nLeventis R, Silvius JR: Lipid-binding characteristics of the polybasic carboxy-terminal sequence of K-ras4B. Biochemistry. 1998; 37(20): 7640–7648. PubMed Abstract \| Publisher Full Text\n\nZimmermann G, Papke B, Ismail S, et al.: Small molecule inhibition of the KRAS-PDEδ interaction impairs oncogenic KRAS signalling. Nature. 2013; 497(7451): 638–642. PubMed Abstract \| Publisher Full Text\n\nZimmermann G, Schultz-Fademrecht C, Küchler P, et al.: Structure guided design and kinetic analysis of highly potent benzimidazole inhibitors targeting the PDEδ prenyl binding site. J Med Chem. 2014; 57(12): 5435–5448. PubMed Abstract \| Publisher Full Text\n\nSpiegel J, Cromm PM, Zimmermann G, et al.: Small-molecule modulation of Ras signaling. Nat Chem Biol. 2014; 10(8): 613–622. PubMed Abstract \| Publisher Full Text\n\nLenta B, Weniger B, Kaiser M, et al.: Antileishmanial natural prenylated anthranoids. Planta Med. 2012; 78(11): PI290. Publisher Full Text\n\nLenta B, Weniger B, Kaiser M, et al.: Antileishmanial natural prenylated anthranoids. Planta Med. 2012; 78(11): PI290. Publisher Full Text\n\nHo CL, Wang JL, Lee CC, et al.: Antroquinonol blocks Ras and Rho signaling via the inhibition of protein isoprenyltransferase activity in cancer cells. Biomed Pharmacotherapy. 2014. PubMed Abstract \| Publisher Full Text\n\nKaiser J, Yassin M, Prakash S, et al.: Anti-malarial drug targets: screening for inhibitors of 2C-methyl-D-erythritol 4-phosphate synthase (IspC protein) in Mediterranean plants. Phytomedicine. 2007; 14(4): 242–249. PubMed Abstract \| Publisher Full Text\n\nllarionova V, Kaiser J, Ostrozhenkova E, et al.: Nonmevalonate terpene biosynthesis enzymes as antiinfective drug targets: substrate synthesis and high-throughput screening methods. J Org Chem. 2006; 71(23): 8824–8834. PubMed Abstract \| Publisher Full Text\n\nHumnabadkar V, Jha RK, Ghatnekar N, et al.: A high-throughput screening assay for simultaneous selection of inhibitors of Mycobacterium tuberculosis 1-deoxy-D-xylulose-5-phosphate synthase (Dxs) or 1-deoxy-D-xylulose 5-phosphate reductoisomerase (Dxr). J Biomed Screening. 2011; 16(3): 303. PubMed Abstract \| Publisher Full Text\n\nTesta CA, Johnson LJ: A whole-cell phenotypic screening platform for identifying methylerythritol phosphate pathway-selective inhibitors as novel antibacterial agents. Antimicrob Agents Chemother. 2012; 56(9): 4906–4913. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDi Paolo A, Danesi R, Caputo S, et al.: Inhibition of protein farnesylation enhances the chemotherapeutic efficacy of the novel geranylgeranyltransferase inhibitor BAL9611 in human colon cancer cells. Br J Cancer. 2001; 84(11): 1535–1543. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCorniani N, Velini ED, Silva FM, et al.: Novel bioassay for the discovery of inhibitors of the 2-C-Methyl-D-erythritol 4-phosphate (MEP) and terpenoid pathways leading to carotenoid biosynthesis. PLoS One. 2014; 9(7): e103704. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmitt F, Oakeley EJ, Jost JP: Antibiotics induce genome-wide hypermethylation in cultured Nicotiana tabacum plants. J Biol Chem. 1997; 272(3): 1534–1540. PubMed Abstract \| Publisher Full Text\n\nAdams RLP, Burdon RH: Molecular Biology of DNA Methylation. New York: Springer Verlag. 1985. Reference Source\n\nFinnegan EJ, Brettell RI, Dennis ES: The role of DNA methylation in the regulation of plant gene expression. EXS. In: DNA Methylation: Molecular Biology and Biological Significance. Jost J-P and Saluz, H-P, eds. (Birkhäuser, Basel, Switzerland), 1993;64: 218–261. PubMed Abstract\n\nFinnegan EJ, Genger RK, Peacock WJ, et al.: DNA methylation in plants. Annu Rev Plant Physiol Plant Mol Biol. 1998; 49: 223–247. PubMed Abstract \| Publisher Full Text\n\nDe Wilde C, Van Houdt H, De Buck S, et al.: Plants as bioreactors for protein production: avoiding the problem of transgene silencing. Plant Mol Biol. 2000; 43(2–3): 347–359. PubMed Abstract \| Publisher Full Text\n\nMatzke M, Aufsatz W, Kanno T, et al.: Genetic analysis of RNA-mediated transcriptional gene silencing. Biochim Biophys Acta. 2004; 1677(1–3): 129–141. PubMed Abstract \| Publisher Full Text\n\nHalpin C: Gene stacking in transgenic plants--the challenge for 21st century plant biotechnology. Plant Biotechnol J. 2005; 3(2): 141–155. PubMed Abstract \| Publisher Full Text\n\nMeng L, Ziv M, Lemaux PG: Nature of stress and transgene locus influences transgene expression stability in barley. Plant Mol Biol. 2006; 62(1–2): 15–28. PubMed Abstract \| Publisher Full Text\n\nMeyer P, Linn F, Heidmann I, et al.: Endogenous and environmental factors influence 35S promoter methylation of a maize A1 gene construct in transgenic petunia and its colour phenotype. Mol Gen Genet. 1992; 231(3): 345–352. PubMed Abstract \| Publisher Full Text\n\nVaucheret H, Palauqui JC, Elmayan T, et al.: Molecular and genetic analysis of nitrite reductase co-suppression in transgenic tobacco plants. Mol Gen Genet. 1995; 248(3): 311–317. PubMed Abstract \| Publisher Full Text\n\nNeumann K, Droge-Laser W, Köhne S, et al.: Heat treatment results in a loss of transgene-encoded activities in several tobacco lines. Plant Physiol. 1997; 115(3): 939–947. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKöhne S, Neumann K, Pühler A, et al.: The heat-treatment induced reduction of the pat gene encoded herbicide resistance in Nicotiana tabacum is influenced by the transgene sequence. J Plant Physiol. 1998; 153(5–6): 631–642. Publisher Full Text\n\nvan der Fits L, Deakin EA, Hoge JH, et al.: The ternary transformation system: constitutive virG on a compatible plasmid dramatically increases Agrobacterium-mediated plant transformation. Plant Mol Biol. 2000; 43(4): 495–502. PubMed Abstract \| Publisher Full Text\n\nLandolph JR, Jones PA: Mutagenicity of 5-azacytidine and related nucleosides in C3H/10T 1/2 clone 8 and V79 cells. Cancer Res. 1982; 42(3): 817–823. PubMed Abstract" }	[ { "id": "7343", "date": "19 Mar 2015", "name": "Patricia León", "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 manuscript by Hartmann et al. describes the establishment of an in vivo assay using plant cell culture cell system to estimate the geranylgeranylation of proteins. The geranylgeranilation is a posttranslational modification required for particular targeting and interactions of some proteins but in plants its function is not fully characterized. Interestingly, previous work by the same group has demonstrated that this modification relays in the activity of the chloroplast isoprenoid MEP pathway and not in the Mevalonate pathway, that operates in the cytoplasm. This manuscript describes the implementation of an in vivo method that allows screening for novel drugs or external signals that affect the isoprenylation process. This method takes advantage of the change in the subcellular localization of a genarygeranylated protein from the cytoplasm to the nucleus. The authors analyze different parameters in order that this bioassay could be use in a qualitative and quantitative manner. In particular this is interesting because there are not too many examples that uses plant culture cells for massive screenings based on fluorescence. My impression is that this manuscript reports a valuable methodology that will be of interest for many groups interested in understanding basic aspects of the prenylation and isoprenoid synthesis in general, and also for selection of novel inhibitors of the pathway. The data included is of good quality data and also this manuscript provides a very detailed discussion of the different parameter of the implemented system. However, a general comment of this manuscript is that is very long for the reader. I consider that some aspects could be shortened without loosing critical points. For example, the general considerations from the introduction (page 3) could be shortened. This is also true for some points in the Results and Methodology such as the description of the ImageJ method or the modified RFP reporter protein that perhaps does not require to be described in such detail and just providing the corresponding references or refer to the related table might be sufficient.- Another aspect I found unusual from the manuscript is that in the Discussion the authors again included additional results. I think that the authors should present the Results in the corresponding section. For example the influence of the putative methylation over the cells is an aspect that I think should be described in the Result section. Additionally of these general aspects related to the organization of the manuscript I have some major points that would be important to address:- The authors select a secondary calli line that is homogenous in the fluorescence and nicely demonstrate that this line exhibits high fluorescence ratio. An important aspect to described is how stable this secondary suspension culture is and how long have the authors followed it. I wonder whether this could be a problem in the transgene expression over time. It is well know that liquid cultures tend to have polyploidization, chromosomal aberrations and maybe changes in chromatin condition. I think this is an important information to be included, particularly if somebody is interested in the selected line. The authors mention some of this aspect in the Discussion but its stability over time is an important information.- To reinforce that the changes in localization observed with the inhibitors is due to a specific blockage of the MEP pathway the authors also used microRNA interference of the HMGR and DXR genes. The authors report that the localization of the GFP-BD_CVIL is very similar to that observed with fosmidomycin and oxoclomasone, supporting the role of the MEP pathway in the geranylgenarlynation. However in the case of the ami HMGR the authors state to not detect any change in the localization. If I understand correct no difference in localization was found with the MVA inhibitor (Figure 9G) and this is exactly what they were expecting. From this result the authors discuss about the inhibition levels for the HMGR that appear to be limited and the possible role of sterols for silencing. I think that this aspect in the current form and probably requires rewriting. Also since the levels of RNA the different ami lines tested are not shown this part remains not fully clear. Thus I think that if this aspect remains the authors should at least mention the inhibition level observed in each line. Minor commentIn the abstract the authors state “Furthermore, complementation assays with pathway-specific intermediates confirmed that the precursors for the cytosolic isoprenylation of this fusion protein are predominantly provided by the MEP pathway” Maybe it escapes me but I did not find any mention in the text of complementation with the intermediates and maybe this paragraph maybe deleted from the Abstract. Page 10. As a negative control the cells were treated with the plasma membrane stain FM4-64. I presume that this data is not shown and this should be indicated in the text. Page 10. In the induction with 6µM Est the time laps used maybe better to be reported in increasing order after induction. In the legend of Figure 7 it is stated “Saturation is already reach after 24 h of induction with both elicitors”. Since both elicitors are not shown in this Figure as stated before I suggest that this statement is moved to the text and not in the Legend.", "responses": [ { "c_id": "1480", "date": "12 Aug 2015", "name": "Thomas J. Bach", "role": "Author Response", "response": "First of all we would like to thank Prof. León for her thoughtful comments and suggestions to further improve the manuscript. We are aware that the manuscript is very detailed in some sections. In fact we faced some dilemma, namely to make specialists in the field of plant isoprenoid biosynthesis and function acquainted with rather advanced imaging techniques (and in this special case even with aspects of protein isoprenylation in general), which might not be in the focus of phytochemists and plant physiologists. At the same, specialists in imaging techniques might also be interested in the particular situation of subcellular metabolic organization in plant cells and the challenges arising from the use of living plant cell cultures that had to be adapted to such analyses, for instance through diminishing the size of cultures to fit into 96-well plates. We would not expect the gentle readers to analyze the publication systematically, but rather to focus on those aspects that seem particularly interesting or where detailed information is essential for evaluation and understanding.As to the DNA methylation problem: We will reorganize the manuscript accordingly. This data was originally integrated in a supplementary part and only later found its way into the main manuscript, which led now to this justified criticism by Prof. León. This part has now been integrated into the Results section where we think it fits.As to selection of call and so forth: Admittedly, this aspect cannot be neglected and additional information concerning the process of generation and maintenance of the cell cultures will be included in the updated version of the manuscript, in particular a scheme illustrating the process of obtaining the doubly transformed cell lines used in this study (Supplemental Figure 1). As far as the maintenance of the cell lines is concerned, we needed to restart liquid cultures from calli several times to recover good levels of fluorescence. This correlated well with documented failures in the maintenance of temperature in the growth chambers. As mentioned in the manuscript, the main culprits for the observed loss of fluorescence were likely changes on the molecular level, in particular DNA methylation events, which became evident after the prolonged treatment with 5-azacytidine recovered some red fluorescence. However, at the same time it led to cell death in a considerable number of cells.We cannot exclude that other, more severe events such as chromosomal alterations might have contributed as well to the loss in fluorescence. We were aware of this possibility but decided to focus on a quick way to save and re-select our “precious” cell lines from cells that might not have been affected by those “somaclonal” events instead of using the whole repertoire of available techniques to figure out the individual causes for those changes.In addition, it should not be forgotten that maintaining the cell culture as a suspension in the long term is time-consuming and costly and does not assure high and homogenous levels of transgene expression. Thus, cell suspension establishment from (re-)selected calli - as suggested in the manuscript - was implemented as a standard procedure once the protocol for it was fully optimized and is highly recommendable. As for the amiRNA constructs now expression data has been added. Up to 10 lines per construct were selected and tested and none achieved better levels of silencing than those presented. It should be noted however that this was not the main point goal of the publication! If we take for instance these amiRNA experiments, then the initial idea was based on a request by a reviewer commenting on the publication that appeared in 2009 (Gerber et al. – reference 30 in the article). To us it was more important to show an independent activation of promoters by dexamethasone (GFP-CVIL/M) and the amiRNA construct to silence either HMGR or DXR (cf. Figure 11 new). The next step was then to express a nucleus-targeted RFP under the control of estrogen, without any cross induction…We discussed this part in view of more recent observations that suggested some dependence of the RNAi machinery on the presence of sterols in the membrane to which essential proteins in this process. One Arabidopsis mutant was deficient in HMGR, known to coarse-regulate sterol biosynthesis. Thus the inhibition by mevinolin might have mimicked such an effect, and thereby a “negative“ result (means no visible inhibition by amiRNA) would be no proof per se. We think that this aspect is properly discussed, even though the visible outcome is what was more or less “expected“ (see Gerber et al. 2009). We admit that the percentage of inhibition (quantitation of corresponding mRNA by PCR) was limited, which in the case of HMGR silencing could also be interpreted to mean that inhibition would be partial at best. By the way, when such an approach had been used with an enzyme downstream of HMGR, viz., cycloartenol synthase (CAS) it resulted in about 35% diminished mRNA, that of HMGR1 and 2 by about 60% after induction by estradiol (Gas-Pascual et al., 2015). This paper is now quoted accordingly.The criticized part in the Abstract relates to the preceding publication, in which such inhibitors had been specifically tested, also chemical complementation by addition of intermediates whose formation had been blocked. At the same this approach could be extended in adding effector molecules, i.e. hormones and growth regulators and watch of whether the mislocalization of the indicator protein could be overcome, for instance through activation of pathways and transporters… Addition of “preceding” where needed should now have clarified why this part was added to the Abstract. A citation number was not allowed to be added by the publisher.The staining by FM4-64 is shown in figure 6, and this information has been added once more where appropriate.The time laps problem: A misunderstanding? No, We respectfully disagree - because the values are related to the time point when the cells were observed after treatment, thus this sequence is more logical as it describes the real time course of this experiment.As to Figure 7: We feel the statement in the legend is sufficiently clear: Both elicitors were used, however the images were only examined for the red fluorescence induced by estradiol." } ] }, { "id": "8261", "date": "16 Apr 2015", "name": "Wolfgang Eisenreich", "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 article provides detailed information about an in vivo based assay system of protein prenylation using stably transformed tobacco BY-2 cells. Notably, the authors show that the method is also useful in finding potential inhibitors against enzymes catalyzing upstream reactions in the MEP pathway of IPP/DMAPP biosynthesis. On this basis, the assay is a welcome addition to existing high-throughput screening assays using photometry. The paper is well written and nicely illustrated. I congratulate the authors on their findings.", "responses": [] } ]	1	https://f1000research.com/articles/4-14
https://f1000research.com/articles/4-550/v1	12 Aug 15	{ "type": "Review", "title": "Cadherin signaling: keeping cells in touch", "authors": [ "Olga Klezovitch", "Valeri Vasioukhin", "Olga Klezovitch" ], "abstract": "Cadherin-catenin complexes are critical for the assembly of cell-cell adhesion structures known as adherens junctions. In addition to the mechanical linkage of neighboring cells to each other, these cell-cell adhesion protein complexes have recently emerged as important sensors and transmitters of the extracellular cues inside the cell body and into the nucleus. In the past few years, multiple studies have identified a connection between the cadherin-catenin protein complexes and major intracellular signaling pathways. Those studies are the main focus of this review.", "keywords": [ "cell-cell interactions", "cadherin", "adherin junctions", "cell-cell adhesion", "catenin", "intracellular signaling pathways" ], "content": "Introduction\n\nThe ability of cells to communicate and adhere to each other represents an ultimate prerequisite for the formation and maintenance of a multicellular organism. By sensing their microenvironment, cells can decide whether to continue or stop proliferating, change shape, accept a new identity, move out of the neighborhood, or simply cease to exist. How do the external signals get transmitted inside and prompt the cells to respond accordingly? In the past several years, cadherin-catenin protein complexes emerged as important regulators of morphogenesis and adult tissue homeostasis, linking cell-cell adhesion to multiple major signaling networks. In this short review, we will focus on the most recent studies that address the mechanisms and the functional relevance of the cadherin-mediated intracellular signaling.\n\n\nAdherens junctions: structural organization and association with the actin cytoskeleton\n\nCadherin-catenin complexes comprise the core of a specialized type of adhesion junction named an adherens junction (AJ) (Figure 1). Among the family of classic cadherins, which includes E (epithelial)-, N (neural)-, P (placental)-, VE (vascular-endothelial)-, R (retinal)-, and K (kidney)-cadherins, E-cadherin is the most frequently employed in the formation of AJs in epithelial cells. To initiate the adhesion process, extracellular domains of cadherins engage in the Ca2+-dependent homophilic trans-interaction with identical cadherin molecules on an adjacent cell, while their cytoplasmic tails bind to p120- and β- (or its homolog γ-) catenin proteins. In turn, β-catenin interacts with α-catenin, which contains an actin-binding domain and physically links AJ complexes to the actin cytoskeleton1,2. Interaction between the actomyosin cytoskeleton and the AJs is prominently regulated by the mechanical forces and Rho-family of small GTPases (covered in detail in 3–6). This regulation is necessary for proper tissue morphogenesis and is highly dynamic, facilitating not only the coupling but also the detachment of cadherin-catenin complexes from actomyosin cytoskeleton, allowing cell-cell separation, cell sorting, and cell migration.\n\nThe diagram depicts protein members of the adherens junctions clustered at the plasma membranes of two juxtaposed cells and summarizes their individual roles in the intricate network of intracellular signaling pathways. Note that, despite their unique structural features and separate functions, both cadherins and catenins often work in concert and may also participate in the regulation of the same signaling pathway though via a distinct mechanism. Abbreviations: MAPK, mitogen-activated protein kinase; NFκB, nuclear factor-kappa-B; RTK, receptor tyrosine kinase; YAP1, yes-associated protein 1.\n\n\nCadherin-mediated intracellular signaling has a pivotal role in contact inhibition of cell proliferation\n\nThe ability of cadherins to transmit signals from the extracellular microenvironment inside the cell body is likely a direct consequence of their adhesive function, which stimulates clustering of cadherin molecules involved in AJ formation. In cell culture experiments, formation of a confluent cell monolayer results in prominent clustering of cadherin-catenin molecules at the AJs. This clustering not only strengthens cell-cell adhesion but also provides important cues for apical-basal cell polarization and significantly influences the downstream signaling events (for review, see 3,5,7). It was noticed a long time ago that formation of a confluent cell monolayer results in cell cycle withdrawal8. This phenomenon is known as “contact inhibition of cell proliferation”7. Re-expression of E-cadherin in human epithelial cancer cell lines that lack E-cadherin expression or disruption of E-cadherin with neutralizing antibodies in cell lines that maintained endogenous E-cadherin demonstrated that cadherin-mediated cell-cell adhesion plays a pivotal role in execution of contact inhibition of cell proliferation9. Similarly, activation of cadherin-catenin-mediated cell-cell adhesion by re-expression of α-catenin in a carcinoma line that was missing endogenous α-catenin resulted in retardation of cell proliferation10. A negative impact of E-cadherin expression on tumor progression was also revealed in genetic mouse experiments in vivo11. Since restoration of cadherin-catenin-mediated cell-cell adhesion results in prominent changes in cell morphology and re-establishment of apical-basal cell polarity, these early experiments were unable to determine whether cadherin clustering plays a direct or indirect role in negative regulation of cell proliferation. This question was later addressed by elegant experiments in Dr. Gumbiner’s laboratory, which demonstrated that clustering of cellular cadherins by E-cadherin-coated extracellular beads is sufficient to induce proliferation inhibitory signaling, thus directly implicating cadherin clustering in cell signaling events12.\n\n\nCadherin-catenin adhesion and growth factor receptor signaling pathways\n\nHow do cadherins exert their signaling functions? Multiple signaling molecules are located at the cell-cell contact sites in direct proximity to the AJ complexes. Many growth- and proliferation-promoting signaling pathways are initiated at the cell surface by receptor-type tyrosine kinases (RTKs). Cadherins can physically interact with several RTKs and they prominently impact their signaling abilities. For example, E-cadherin associates with epidermal growth factor receptor (EGFR) and negatively regulates its kinase activity12–14. Tumor-suppressor protein neurofibromatosis type 2 (NF2 or Merlin) promotes association between E-cadherin and EGFR, links EGFR to the cortical actin cytoskeleton, and blocks its internalization, which is necessary for EGFR activation and signaling15,16. Loss of Merlin in mouse liver results in prominent activation of EGFR signaling, expansion of progenitors, and development of liver cancer17. In addition to EGFR, E-cadherin can also negatively impact signaling of other RTKs, including ErbB2, insulin-like growth factor receptor (IGFR), and c-Met14. Similar to E-cadherin in epithelial cells, VE-cadherin in endothelial cells interacts with vascular-endothelial growth factor receptor 2 (VEGFR2) and negatively regulates its mitogen-activated protein kinase (MAPK) signaling by preventing the clathrin-dependent internalization of VEGFR2 and promoting the association between VEGFR2 and tyrosine phosphatase PTPRJ, which dephosphorylates and inactivates VEGFR218,19.\n\nIt is important to note that in some cases cadherins can promote growth factor receptor signaling. For example, N-cadherin stimulates fibroblast growth factor receptor signaling by preventing ligand-induced receptor internalization20. Both E-cadherin and VE-cadherin can promote PI3-kinase (PI3K) signaling and protect cells from apoptotic cell death21,22. VE-cadherin associates with the transforming growth factor-beta (TGF-β) receptor complex and potentiates cell proliferation inhibitory TGF-β signaling events23.\n\n\nβ- and p120-catenins and the direct line of communication between cell-cell junctions and transcriptional regulation of gene expression\n\nBy acting at the plasma membrane, cadherins are ideally positioned to attract and retain their cytoplasmic partners, thus modulating their activation, stability, or nuclear accumulation or a combination of these.\n\nThis is important because some of these intracellular proteins are pivotal signaling molecules in their own right. For example, β-catenin is a very potent transcriptional co-activator and a key member of the canonical Wnt signaling pathway (for review, see 24–26). The levels of cytoplasmic β-catenin available for signaling are tightly controlled by the activity of the β-catenin-destruction protein complex, which is inhibited by activation of Wnt signaling24,25. Sequestration of β-catenin at the cell junctions can attenuate its ability to enter the cell nucleus and participate in transcriptional regulation. Indeed, multiple studies demonstrated that the loss of cadherin-mediated cell adhesion can promote β-catenin release and signaling26. The exact relationship between cadherin-mediated adhesion and β-catenin signaling is highly complex and context-dependent. In some cases, not only do cadherins not inhibit but they actually potentiate the β-catenin signaling pathway (for review, see 27).\n\nSimilarly to β-catenin, cadherins can sequester at the plasma membrane and prevent cytoplasmic accumulation of another member of AJs, p120-catenin (for review, see 28). p120-catenin binds to the transcriptional repressor KAISO and inhibits its function29–31. In addition, p120-catenin is a potent regulator of Rho-family GTPases and the nuclear factor-kappa-B (NFκB) signaling pathway28,32. p120-catenin is critical for stabilization of cadherin-catenin complexes and formation of AJs, and this function is likely to be responsible for its tumor-suppressor function in squamous cell carcinoma (SCC), which was revealed by genetic loss-of-function experiments in mice33.\n\n\nα-catenin and regulation of cellular signal transduction pathways\n\nα-catenin is crucial for AJ formation because it is necessary for the direct linkage of cadherin-catenin complexes at the membrane with the actin cytoskeleton34. Although there are three α-catenin genes in mammalian genomes (alpha E-catenin CTNNA1, alpha N-catenin CTNNA2, and alpha T (testis)-catenin CTNNA3), most epithelial cells express only one α-catenin (CTNNA1), and the knockout of this gene is usually sufficient for the complete loss of AJ function and loss of cell polarity35,36. This is different from inactivation of E-cadherin or β-catenin, which may often have redundant functions in the AJs because of the expression of other cadherins and γ-catenin. Notably, this is not the case in the adult heart, where inactivation of all expressed alpha-catenins (Ctnna1 and Ctnna3) does not cause a severe cell adhesion defect comparatively to N-cadherin knockout mice37,38.\n\nSimilar to p120 catenin (Ctnnd1), genetic loss-of-function experiments in mice revealed prominent tumor-suppressor activity of epithelial α-catenin (Ctnna1), as epidermal stem cell-specific deletion of α-catenin in mice results in the development of SCC tumors35,39,40. Like p120-catenin, α-catenin has been linked to NFκB signaling pathway in skin39 and in E-cadherin-negative basal-like breast cancer cells41, where it interacts with and stabilizes IκBα by preventing its ubiquitylation and association with proteasomes41. In addition to its critical role in cell-cell adhesion, via direct interaction with the dynactin protein complex, α-catenin can regulate dynactin-dynein-mediated traffic and integrate the microtubule and actin cytoskeletons during intracellular trafficking events42.\n\nLoss-of-function experiments in vivo and in vitro revealed an important role of α-catenin in regulation of several major signaling networks, including Ras-MAPK35, canonical Wnt27,43, and Hedgehog44 pathways. Since α-catenin acts as a tumor suppressor in skin epidermis, our laboratory performed a small interfering RNA (siRNA) screen for genes necessary for this function in keratinocytes, which revealed a connection between α-catenin and yes-associated protein 1 (YAP1), a pivotal target of the Hippo signal transduction pathway40. The connection between cadherin-catenin proteins and the Hippo pathway components has been demonstrated by multiple studies and we will discuss these findings in more detail45–47 (see below).\n\n\nMeet the Hippo: the new darling of the cadherin signaling\n\nFirst identified in Drosophila, the Hippo signaling pathway is evolutionarily conserved and functions as a key regulator of organ size and tumorigenesis by inhibiting cell proliferation and promoting (and, in some cases, inhibiting) apoptotic cell death (for review, see 48,49). In vertebrates, the core of the canonical Hippo pathway consists of two sequentially acting sets of kinases, MST1/2 and LATS1/2 (Hippo and Warts in Drosophila), and several associated co-activators and scaffold proteins. The MST1/2 kinases phosphorylate and activate LATS1/2, which in turn phosphorylates the growth-promoting transcriptional co-activator YAP1 (Yorkie in Drosophila) and its homolog TAZ (also known as WWTR1), leading to their cytoplasmic retention. When the Hippo pathway is inhibited, YAP1 translocates to the nucleus, where it binds multiple transcriptional factors and promotes their transcriptional activity48,49. It is important to note that, in addition to the canonical Hippo signaling pathway, YAP1/TAZ nuclear localization and activity can be regulated independently from MST1/2 and LATS1/245,50,51. In both Drosophila and mammalian model systems, the Hippo signaling is exquisitely sensitive to changes in the actin cytoskeleton or cellular tension which functions as a pivotal regulator that integrates and transmits upstream signals to the Hippo signal transduction pathway (for review, see 49,52). Increase in F-actin and actomyosin contractility blocks Hippo signaling and prominently activates Yorkie/YAP1/TAZ51,53.\n\nFor a long time, it remained largely unknown whether extracellular cues play any role in activating the Hippo pathway in mammals. The identity of the upstream transmembrane receptors responsible for transmitting the external signals inside the cell was undetermined. Elegant experiments in Dr. Guan’s laboratory identified G-protein-coupled receptors as important upstream regulators of Hippo signaling in mammalian cells54. The evidence that the nuclear localization and activity of YAP1 are inversely correlated with cell density55 pointed in the direction of the cell-cell junctions as potential upstream regulators of the Hippo signaling pathway. Indeed, it was recently demonstrated that E-cadherin homophilic binding at the cell surface in mammalian MDA-MB-231 cells is sufficient to control the subcellular localization of YAP1 independently of other cell interactions46. In addition, two recent studies using primary mouse keratinocytes revealed that α-catenin can bind to YAP1 and sequester it in the cytoplasm, thus modulating the level of YAP1 phosphorylation and its activity40,45 (for review, see 56,57). Importantly, there was an inverse correlation between α-catenin levels and nuclear YAP1 localization in both cultured keratinocytes and human SCC tumors, indicating that α-catenin may act as an inhibitor of YAP1 both in vitro and in vivo40. Of interest, although Ca2+ depletion, which abolishes cadherin homophilic interactions, triggered translocation of YAP1 into the nucleus, the depletion of E/P-cadherin or β-catenin in cultured keratinocytes did not affect the cellular localization of YAP145, pointing at the possibility that the expression of other cadherins and catenins might be sufficient to maintain AJs in E/P-cadherin or β-catenin knockdown keratinocytes.\n\nIn addition to α-catenin, β-catenin interacts with YAP1 and these proteins prominently impact each other’s nuclear localization and activity47,58,59. Constitutive activation of β-catenin in human cancer cells results in the formation of a β-catenin-YAP1-TBX5 transcriptional complex, which is essential for cancer cell survival60.\n\nIn Drosophila, the Hippo pathway can be regulated by multiple upstream transmembrane modules, which include atypical cadherins Dachsous and Fat (for review, see 61). Recently, another AJ protein, Echinoid, was shown to activate Hippo signaling via its physical interaction with and stabilization of the Hippo-binding partner Salvador62. This interaction is triggered by cell-cell contacts and requires the dimerization of Echinoid cytoplasmic domain. It is of interest to mention that, although there is no known Echinoid homolog in mammals, this protein is able to interact with Drosophila E-cadherin, thus contributing to the formation and maintenance of AJs63. Overall, although there are a lot of similarities between Drosophila and mammalian Hippo signaling pathways, at least some of the upstream regulators may be quite different64. Drosophila Yorkie is missing the C-terminal PDZ-binding motif, which is necessary for the connection between YAP1/TAZ and tight junction (TJ) proteins in mammalian cells. Although α-catenin is a potent negative regulator of YAP1 in mammalian cells38,40,45,46,65, it is a positive regulator of Yorkie in Drosophila66,67. While E-cadherin is a cell autonomous-positive regulator of Hippo pathway in mammalian cells46, it is a cell autonomous-negative regulator of Hippo in Drosophila67. Fat4, the mammalian ortholog of Drosophila fat gene, does not regulate the Hippo pathway in mouse liver, the organ highly sensitive to changes in the canonical Hippo signaling pathway64. However, mammalian FAT4 and Dachsous cadherins appear to negatively regulate YAP1 in neural progenitor cells68,69, indicating that at least some of the important connections in Hippo signaling may be tissue- and species-specific.\n\nAs discussed above, one of the ways for cadherins to regulate contact inhibition of cell proliferation is by antagonizing the activity of a variety of RTKs, including the EGFR. Interestingly, changes in RTK activity may indirectly impact Hippo signaling. For example, it was recently demonstrated that, in immortalized mammary cells, EGF treatment triggers the nuclear accumulation of YAP1 through activation of PI3K and phosphoinositide-dependent kinase (PDPK1) and this is largely independent of AKT signaling70. Interestingly, in Drosophila, EGF signaling also inhibits the Hippo pathway but through a different mechanism, which uses MAPK and the inhibitor of Warts, Jub71. Taken together, those findings point at the important connection between AJs, mitogenic factor pathways, and growth-inhibitory Hippo signaling. Of note, the Drosophila Jub was also shown to associate with α-catenin in a cytoskeleton tension-dependent manner, thus linking the actomyosin cytoskeleton, regulation of Hippo pathway activity, and AJs66.\n\nIn addition to the AJs, cadherin-mediated adhesion plays an important role in the formation of TJs and the apical-basal cell polarity domains. In turn, the polarity complex proteins can interact with structural components of both AJs and TJs, thus potentially centralizing the regulation of several signaling pathways (for review, see 72), although it is possible that the AJs and cell polarity regulate the Hippo signaling via multiple, genetically separable mechanisms67. The TJ-associated proteins angiomotin and angiomotin-like 1 and 2 directly interact with YAP1/TAZ, localize them to the cytoplasm and TJs, and negatively regulate their transcriptional activity73–76. Remarkably, at least in some cases, angiomotin proteins promote YAP1 activity by antagonizing YAP1-LATS2 interaction and increasing YAP1 dephosphorylation and translocation to the nucleus77. Interestingly, via its interaction with Merlin, angiomotin can localize to the AJs and facilitate AJ-specific recruitment and activation of LATS78. In both Drosophila and mammals, Merlin promotes Hippo signaling by targeting LATS to the cell membrane79. However, since angiomotin proteins are missing in the Drosophila genome, the angiomotin-mediated localization and activation of LATS at the AJs are likely to be species-specific, and this may potentially explain the differences in AJ-mediated regulation of YAP1 between Drosophila and mammalian model systems.\n\n\nFuture directions\n\nThe unique aspect of cadherin-mediated signaling is that the clustering of cadherin molecules is mediated by the direct cell-cell contacts. This enables cells to identify and map the positions of their immediate neighbors, helping to integrate individual cells into the tissues not only at physical but also at biochemical levels. Although we are continually learning about novel aspects of cadherin-mediated signaling, it is clear that the unifying picture is still not within reach. Knowledge remains highly fragmented with distinct and frequently seemingly opposite findings generated in different model organisms, tissues, or cell culture conditions. Future studies are clearly necessary to accumulate more data in the hope that the sheer quantity of information will inevitably result in a qualitative change in our understanding of how individual cells use their cell-cell adhesion structures to coordinate their behavior in building and homeostatic maintenance of multicellular organisms.\n\n\nAbbreviations\n\nAJ, adherens junction; E, epithelial; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; MAPK, mitogen-activated protein kinase; N, neural; NFκB, nuclear factor-kappa-B; P, placental; PI3K, PI3-kinase; RTK, receptor tyrosine kinase; SCC, squamous cell carcinoma; TGF-β, transforming growth factor-beta; TJ, tight junction; VE, vascular-endothelial; VEGFR2, vascular endothelial growth factor receptor 2; YAP1, yes-associated protein 1.", "appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nWork on this article was supported in part by National Institutes of Health grants CA179914 and CA188452.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nStepniak E, Radice GL, Vasioukhin V: Adhesive and signaling functions of cadherins and catenins in vertebrate development. Cold Spring Harb Perspect Biol. 2009; 1(5): a002949. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCapaldo CT, Farkas AE, Nusrat A: Epithelial adhesive junctions. F1000Prime Rep. 2014; 6: 1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaître J, Heisenberg CP: Three functions of cadherins in cell adhesion. Curr Biol. 2013; 23(14): R626–33. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nTakeichi M: Dynamic contacts: rearranging adherens junctions to drive epithelial remodelling. Nat Rev Mol Cell Biol. 2014; 15(6): 397–410. PubMed Abstract \| Publisher Full Text\n\nPriya R, Yap AS: Active tension: the role of cadherin adhesion and signaling in generating junctional contractility. Curr Top Dev Biol. 2015; 112: 65–102. PubMed Abstract \| Publisher Full Text\n\nRöper K: Integration of cell-cell adhesion and contractile actomyosin activity during morphogenesis. Curr Top Dev Biol. 2015; 112: 103–27. PubMed Abstract \| Publisher Full Text\n\nMcClatchey AI, Yap AS: Contact inhibition (of proliferation) redux. Curr Opin Cell Biol. 2012; 24(5): 685–94. PubMed Abstract \| Publisher Full Text\n\nEagle H, Levine EM: Growth regulatory effects of cellular interaction. Nature. 1967; 213(5081): 1102–6. PubMed Abstract \| Publisher Full Text\n\nSt Croix B, Sheehan C, Rak JW, et al.: E-Cadherin-dependent growth suppression is mediated by the cyclin-dependent kinase inhibitor p27KIP1. J Cell Biol. 1998; 142(2): 557–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWatabe M, Nagafuchi A, Tsukita S, et al.: Induction of polarized cell-cell association and retardation of growth by activation of the E-cadherin-catenin adhesion system in a dispersed carcinoma line. J Cell Biol. 1994; 127(1): 247–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerl AK, Wilgenbus P, Dahl U, et al.: A causal role for E-cadherin in the transition from adenoma to carcinoma. Nature. 1998; 392(6672): 190–3. PubMed Abstract \| Publisher Full Text\n\nPerrais M, Chen X, Perez-Moreno M, et al.: E-cadherin homophilic ligation inhibits cell growth and epidermal growth factor receptor signaling independently of other cell interactions. Mol Biol Cell. 2007; 18(6): 2013–25. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nHoschuetzky H, Aberle H, Kemler R: Beta-catenin mediates the interaction of the cadherin-catenin complex with epidermal growth factor receptor. J Cell Biol. 1994; 127(5): 1375–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQian X, Karpova T, Sheppard AM, et al.: E-cadherin-mediated adhesion inhibits ligand-dependent activation of diverse receptor tyrosine kinases. EMBO J. 2004; 23(8): 1739–48. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCole BK, Curto M, Chan AW, et al.: Localization to the cortical cytoskeleton is necessary for Nf2/merlin-dependent epidermal growth factor receptor silencing. Mol Cell Biol. 2008; 28(4): 1274–84. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCurto M, Cole BK, Lallemand D, et al.: Contact-dependent inhibition of EGFR signaling by Nf2/Merlin. J Cell Biol. 2007; 177(5): 893–903. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBenhamouche S, Curto M, Saotome I, et al.: Nf2/Merlin controls progenitor homeostasis and tumorigenesis in the liver. Genes Dev. 2010; 24(16): 1718–30. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGrazia Lampugnani M, Zanetti A, Corada M, et al.: Contact inhibition of VEGF-induced proliferation requires vascular endothelial cadherin, beta-catenin, and the phosphatase DEP-1/CD148. J Cell Biol. 2003; 161(4): 793–804. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLampugnani MG, Orsenigo F, Gagliani MC, et al.: Vascular endothelial cadherin controls VEGFR-2 internalization and signaling from intracellular compartments. J Cell Biol. 2006; 174(4): 593–604. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nSuyama K, Shapiro I, Guttman M, et al.: A signaling pathway leading to metastasis is controlled by N-cadherin and the FGF receptor. Cancer cell. 2002; 2(4): 301–14. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nPece S, Chiariello M, Murga C, et al.: Activation of the protein kinase Akt/PKB by the formation of E-cadherin-mediated cell-cell junctions. Evidence for the association of phosphatidylinositol 3-kinase with the E-cadherin adhesion complex. J Biol Chem. 1999; 274(27): 19347–51. PubMed Abstract \| Publisher Full Text\n\nCarmeliet P, Lampugnani MG, Moons L, et al.: Targeted deficiency or cytosolic truncation of the VE-cadherin gene in mice impairs VEGF-mediated endothelial survival and angiogenesis. Cell. 1999; 98(2): 147–57. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nRudini N, Felici A, Giampietro C, et al.: VE-cadherin is a critical endothelial regulator of TGF-beta signalling. EMBO J. 2008; 27(7): 993–1004. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nClevers H, Nusse R: Wnt/β-catenin signaling and disease. Cell. 2012; 149(6): 1192–205. PubMed Abstract \| Publisher Full Text\n\nClevers H: Wnt/beta-catenin signaling in development and disease. Cell. 2006; 127(3): 469–80. PubMed Abstract \| Publisher Full Text\n\nHeuberger J, Birchmeier W: Interplay of cadherin-mediated cell adhesion and canonical Wnt signaling. Cold Spring Harb Perspect Biol. 2010; 2(2): a002915. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCrea PD, Maher MT, Gottardi CJ: Nuclear signaling from cadherin adhesion complexes. Curr Top Dev Biol. 2015; 112: 129–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchackmann RC, Tenhagen M, van de Ven RA, et al.: p120-catenin in cancer - mechanisms, models and opportunities for intervention. J Cell Sci. 2013; 126(Pt 16): 3515–25. PubMed Abstract \| Publisher Full Text\n\nDaniel JM, Reynolds AB: The catenin p120(ctn) interacts with Kaiso, a novel BTB/POZ domain zinc finger transcription factor. Mol Cell Biol. 1999; 19(5): 3614–23. PubMed Abstract \| Free Full Text\n\nPark JI, Kim SW, Lyons JP, et al.: Kaiso/p120-catenin and TCF/beta-catenin complexes coordinately regulate canonical Wnt gene targets. Dev Cell. 2005; 8(6): 843–54. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nSpring CM, Kelly KF, O'Kelly I, et al.: The catenin p120ctn inhibits Kaiso-mediated transcriptional repression of the beta-catenin/TCF target gene matrilysin. Exp Cell Res. 2005; 305(2): 253–65. PubMed Abstract \| Publisher Full Text\n\nPerez-Moreno M, Davis MA, Wong E, et al.: p120-catenin mediates inflammatory responses in the skin. Cell. 2006; 124(3): 631–44. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nStairs DB, Bayne LJ, Rhoades B, et al.: Deletion of p120-catenin results in a tumor microenvironment with inflammation and cancer that establishes it as a tumor suppressor gene. Cancer cell. 2011; 19(4): 470–83. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBuckley CD, Tan J, Anderson KL, et al.: Cell adhesion. The minimal cadherin-catenin complex binds to actin filaments under force. Science. 2014; 346(6209): 1254211. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nVasioukhin V, Bauer C, Degenstein L, et al.: Hyperproliferation and defects in epithelial polarity upon conditional ablation of alpha-catenin in skin. Cell. 2001; 104(4): 605–17. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nNemade RV, Bierie B, Nozawa M, et al.: Biogenesis and function of mouse mammary epithelium depends on the presence of functional alpha-catenin. Mech Dev. 2004; 121(1): 91–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKostetskii I, Li J, Xiong Y, et al.: Induced deletion of the N-cadherin gene in the heart leads to dissolution of the intercalated disc structure. Circ Res. 2005; 96(3): 346–54. PubMed Abstract \| Publisher Full Text\n\nLi J, Gao E, Vite A, et al.: Alpha-catenins control cardiomyocyte proliferation by regulating Yap activity. Circ Res. 2015; 116(1): 70–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKobielak A, Fuchs E: Links between alpha-catenin, NF-kappaB, and squamous cell carcinoma in skin. Proc Natl Acad Sci U S A. 2006; 103(7): 2322–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSilvis MR, Kreger BT, Lien WH, et al.: α-catenin is a tumor suppressor that controls cell accumulation by regulating the localization and activity of the transcriptional coactivator Yap1. Sci Signal. 2011; 4(174): ra33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPiao H, Yuan Y, Wang M, et al.: α-catenin acts as a tumour suppressor in E-cadherin-negative basal-like breast cancer by inhibiting NF-κB signalling. Nat Cell Biol. 2014; 16(3): 245–54. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLien W, Gelfand VI, Vasioukhin V: Alpha-E-catenin binds to dynamitin and regulates dynactin-mediated intracellular traffic. J Cell Biol. 2008; 183(6): 989–97. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nChoi SH, Estarás C, Moresco JJ, et al.: α-Catenin interacts with APC to regulate β-catenin proteolysis and transcriptional repression of Wnt target genes. Genes Dev. 2013; 27(22): 2473–88. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLien WH, Klezovitch O, Fernandez TE, et al.: alphaE-catenin controls cerebral cortical size by regulating the hedgehog signaling pathway. Science. 2006; 311(5767): 1609–12. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nSchlegelmilch K, Mohseni M, Kirak O, et al.: Yap1 acts downstream of α-catenin to control epidermal proliferation. Cell. 2011; 144(5): 782–95. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nKim NG, Koh E, Chen X, et al.: E-cadherin mediates contact inhibition of proliferation through Hippo signaling-pathway components. Proc Natl Acad Sci U S A. 2011; 108(29): 11930–5. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nAzzolin L, Panciera T, Soligo S, et al.: YAP/TAZ incorporation in the β-catenin destruction complex orchestrates the Wnt response. Cell. 2014; 158(1): 157–70. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBarry ER, Camargo FD: The Hippo superhighway: signaling crossroads converging on the Hippo/Yap pathway in stem cells and development. Curr Opin Cell Biol. 2013; 25(2): 247–53. PubMed Abstract \| Publisher Full Text\n\nYu FX, Guan KL: The Hippo pathway: regulators and regulations. Genes Dev. 2013; 27(4): 355–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeung CY, Zernicka-Goetz M: Angiomotin prevents pluripotent lineage differentiation in mouse embryos via Hippo pathway-dependent and -independent mechanisms. Nat Commun. 2013; 4: 2251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDupont S, Morsut L, Aragona M, et al.: Role of YAP/TAZ in mechanotransduction. Nature. 2011; 474(7350): 179–83. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nGaspar P, Tapon N: Sensing the local environment: actin architecture and Hippo signalling. Curr Opin Cell Biol. 2014; 31: 74–83. PubMed Abstract \| Publisher Full Text\n\nAragona M, Panciera T, Manfrin A, et al.: A mechanical checkpoint controls multicellular growth through YAP/TAZ regulation by actin-processing factors. Cell. 2013; 154(5): 1047–59. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nYu F, Zhao B, Panupinthu N, et al.: Regulation of the Hippo-YAP pathway by G-protein-coupled receptor signaling. Cell. 2012; 150(4): 780–91. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nZhao B, Wei X, Li W, et al.: Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control. Genes Dev. 2007; 21(21): 2747–61. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRobinson BS, Moberg KH: Cell-cell junctions: α-catenin and E-cadherin help fence in Yap1. Curr Biol. 2011; 21(21): R890–2. PubMed Abstract \| Publisher Full Text\n\nEckert F, Wolff H, Ring J, et al.: Das atypische Fibroxanthom. Hautarzt. 1990; 41(1): 39–42. PubMed Abstract\n\nHeallen T, Zhang M, Wang J, et al.: Hippo pathway inhibits Wnt signaling to restrain cardiomyocyte proliferation and heart size. Science. 2011; 332(6028): 458–61. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nImajo M, Miyatake K, Iimura A, et al.: A molecular mechanism that links Hippo signalling to the inhibition of Wnt/β-catenin signalling. EMBO J. 2012; 31(5): 1109–22. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRosenbluh J, Nijhawan D, Cox AG, et al.: β-Catenin-driven cancers require a YAP1 transcriptional complex for survival and tumorigenesis. Cell. 2012; 151(7): 1457–73. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGumbiner BM, Kim N: The Hippo-YAP signaling pathway and contact inhibition of growth. J Cell Sci. 2014; 127(Pt 4): 709–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYue T, Tian A, Jiang J: The cell adhesion molecule echinoid functions as a tumor suppressor and upstream regulator of the Hippo signaling pathway. Dev Cell. 2012; 22(2): 255–67. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nWei SY, Escudero LM, Yu F, et al.: Echinoid is a component of adherens junctions that cooperates with DE-Cadherin to mediate cell adhesion. Dev Cell. 2005; 8(4): 493–504. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBossuyt W, Chen CL, Chen Q, et al.: An evolutionary shift in the regulation of the Hippo pathway between mice and flies. Oncogene. 2014; 33(10): 1218–28. PubMed Abstract \| Publisher Full Text\n\nHerr KJ, Tsang YN, Ong JW, et al.: Loss of α-catenin elicits a cholestatic response and impairs liver regeneration. Sci Rep. 2014; 4: 6835. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRauskolb C, Sun S, Sun G, et al.: Cytoskeletal tension inhibits Hippo signaling through an Ajuba-Warts complex. Cell. 2014; 158(1): 143–56. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nYang CC, Graves HK, Moya IM, et al.: Differential regulation of the Hippo pathway by adherens junctions and apical-basal cell polarity modules. Proc Natl Acad Sci U S A. 2015; 112(6): 1785–90. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCappello S, Gray MJ, Badouel C, et al.: Mutations in genes encoding the cadherin receptor-ligand pair DCHS1 and FAT4 disrupt cerebral cortical development. Nat Genet. 2013; 45(11): 1300–8. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nVan Hateren NJ, Das RM, Hautbergue GM, et al.: FatJ acts via the Hippo mediator Yap1 to restrict the size of neural progenitor cell pools. Development. 2011; 138(10): 1893–902. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nFan R, Kim NG, Gumbiner BM: Regulation of Hippo pathway by mitogenic growth factors via phosphoinositide 3-kinase and phosphoinositide-dependent kinase-1. Proc Natl Acad Sci U S A. 2013; 110(7): 2569–74. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nReddy BV, Irvine KD: Regulation of Hippo signaling by EGFR-MAPK signaling through Ajuba family proteins. Dev Cell. 2013; 24(5): 459–71. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nWhite MD, Plachta N: How adhesion forms the early mammalian embryo. Curr Top Dev Biol. 2015; 112: 1–17. PubMed Abstract \| Publisher Full Text\n\nChan SW, Lim CJ, Chong YF, et al.: Hippo pathway-independent restriction of TAZ and YAP by angiomotin. J Biol Chem. 2011; 286(9): 7018–26. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nWang W, Huang J, Chen J: Angiomotin-like proteins associate with and negatively regulate YAP1. J Biol Chem. 2011; 286(6): 4364–70. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nZhao B, Li L, Lu Q, et al.: Angiomotin is a novel Hippo pathway component that inhibits YAP oncoprotein. Genes Dev. 2011; 25(1): 51–63. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nVarelas X, Samavarchi-Tehrani P, Narimatsu M, et al.: The Crumbs complex couples cell density sensing to Hippo-dependent control of the TGF-β-SMAD pathway. Dev Cell. 2010; 19(6): 831–44. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nYi C, Shen Z, Stemmer-Rachamimov A, et al.: The p130 isoform of angiomotin is required for Yap-mediated hepatic epithelial cell proliferation and tumorigenesis. Sci Signal. 2013; 6(291): ra77. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nHirate Y, Hirahara S, Inoue K, et al.: Polarity-dependent distribution of angiomotin localizes Hippo signaling in preimplantation embryos. Curr Biol. 2013; 23(13): 1181–94. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nYin F, Yu J, Zheng Y, et al.: Spatial organization of Hippo signaling at the plasma membrane mediated by the tumor suppressor Merlin/NF2. Cell. 2013; 154(6): 1342–55. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "9951", "date": "12 Aug 2015", "name": "Andrea I. McClatchey", "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", "responses": [] }, { "id": "9952", "date": "12 Aug 2015", "name": "Stefano Piccolo", "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", "responses": [] }, { "id": "9953", "date": "12 Aug 2015", "name": "Glenn L. Radice", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-550
https://f1000research.com/articles/4-546/v1	11 Aug 15	{ "type": "Research Article", "title": "Pain and sickness behavior associated with corneal lesions in dairy calves", "authors": [ "Brandon J. Woods", "Suzanne T. Millman", "Natalia A. da Silva", "Reneé D. Dewell", "Rebecca L. Parsons", "Chong Wang", "Annette M. O'Connor", "Brandon J. Woods", "Suzanne T. Millman", "Natalia A. da Silva", "Reneé D. Dewell", "Rebecca L. Parsons", "Chong Wang" ], "abstract": "Infectious bovine keratoconjunctivitis (IBK) is a common corneal disease of calves that adversely affects animal welfare by causing pain and weight loss. Identifying behavioral indicators of pain and sickness in calves with IBK is necessary for designing studies that aim to identify effective means of pain mitigation. Consistent with principles of the 3Rs for animal use in research, data from a randomized blinded challenge study was used to identify and describe variation of behaviors that could serve as reliable indicators of pain and sickness in calves with corneal injuries. Behavioral observations were collected from 29 Holstein calves 8 to 12 weeks of age randomly allocated to one of three treatments: (1) corneal scarification only, (2) corneal scarification with inoculation with Moraxella bovoculi and (3) corneal scarification with inoculation with Moraxella bovis. Behavior was continuously observed between time 1230 - 1730 h on day -1 (baseline time period) and day 0 (scarification time period). Corneal scarification and inoculation occurred between 0800 - 1000 h on day 0. Frequency of head-directed behaviors (head shaking, head rubbing, head scratching) and durations of head rubbing, feeding, standing with head lifted, lying with head lifted and sleeping were compared between study days and groups. Following scarification, the frequency of head-directed behavior significantly increased (p = 0.0001), as did duration of head rubbing (p=0.02). There was no significant effect of trial, trial day, treatment or treatment-day interaction on other behaviors studied. Our study demonstrated that head-directed behavior, such as head shaking, rubbing and scratching, was associated with scarification of eyes using an IBK challenge model, but sickness behavior was not observed.", "keywords": [ "calves", "infectious bovine keratoconjunctivitis", "ocular pain", "sickness behavior" ], "content": "Preamble\n\nThe authors affirm that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the review as planned have been documented and explained. The authors have indicated where results from these study animals are reported in other publications, and have including citations for these publications where relevant.\n\n\nIntroduction\n\nInfectious bovine keratoconjunctivitis (IBK) is a disease of cattle, causing corneal edema and ulceration, photophobia, blepharospasm and ephiphora (Gelatt, 2008; George, 1984). IBK can occur in 20–30% of calves in a beef calf crop, with an estimated 30% of beef herds affected annually (Brown et al., 1998). Incidence of IBK has been associated with 6.8 to 13.6 kg decreased weaning weight (Funk et al., 2009; Snowder et al., 2005). As there is evidence that vaccination is an ineffective strategy in isolation (Funk et al., 2009; Kizilkaya et al., 2013; O'Connor et al., 2012); producers must identify non-responding animals and minimize the impact of IBK once diagnosed with antibiotic treatment (O'Connor et al., 2006).\n\nBlepharospasm and photophobia suggest IBK is painful (Williams, 2010) and pain mitigation therapies may be useful adjuncts to antibiotic therapy by improving animal welfare and reducing weight loss. Since blepharospasm, photophobia and ocular discharge are the earliest indications of IBK (Postma et al., 2008), suggesting that detection occurs only once the condition is quite advanced. In livestock and horses, ocular injury can occur as a result of irritation of the corneal surface by dust, tall grasses, weeds or contact with other elements in the environment, such as fencing. Mechanical injury to the eye increases susceptibility of cattle to IBK infection (Postma et al., 2008), and identification of behavior responses to injury may provide opportunity for early detection of corneal injury and preventive treatment. Although subjective scoring of behavior associated with acute IBK infection has been described in the literature as an aspect of signalment and clinical assessment, scientific investigation of behavioral responses to ocular injury and infection is needed.\n\nWe have postulated that pain and sickness behavior associated with IBK might reduce nursing and forage consumption and explain the weight loss commonly associated with this disease. Cattle display behavioral changes in response to pain that may be specific to the nature of the injury (Millman, 2013). For example, calves display significant increases in ear flicking and head rubbing behaviors (Duffield et al., 2010; Faulkner & Weary, 2000; Heinrich et al., 2010), that occur concurrently with increases in physiological and biochemical responses (Heinrich et al., 2009) following cautery disbudding surgery, and which are mitigated by postsurgical analgesia with nonsteroidal anti-inflammatory drugs. Further, the cytokine cascade associated with the inflammatory response evokes characteristic “sickness behaviors”, such as anorexia and increased rest (Millman, 2007; Watkins & Maier, 2000). Associations between morbidity and changes in feeding and social behavior have been identified in cattle (Weary et al., 2009). However, a paucity of information about pain and sickness behavior specifically associated with IBK or ocular insults presents barriers to testing our hypothesis.\n\nPrevious research by our team suggests that mechanical ocular injury is painful to calves (Dewell et al., 2014). A randomized and blinded disease challenge study was conducted to assess putative causal organisms for IBK incidence in calves (Gould et al., 2013). A mechanical ocular injury was administered in one eye (“scarification”), followed by inoculation with Moraxella bovis (M. bovis), Moraxella bovoculi (M. bovoculi) or no inoculation. Only calves in the M. bovis treatment developed IBK-associated corneal abnormalities. Concurrent with the microbiological study, we evaluated clinical approaches for qualifying ocular pain in calves, using pressure algometry, a Cochet Bonet aesthesiometer, blepharospasm and photophobia (Dewell et al., 2014). Significant changes in mechanical nociception threshold scores were observed following scarification relative to baseline values prior to treatment suggesting increased pain sensitivity, but neither IBK inoculation nor corneal ulceration were associated with differences in nociception responses. Retrospective video analysis of the calves enrolled in this study presented a unique opportunity for detailed investigation of behavioral changes in the home pen associated with ocular injury and IBK infection.\n\nThe objective of this study was to maximize the value of data obtained from a prior study by describing the magnitude and variation of pain and sickness behaviors in calves with experimental induced ocular injury and infection. Such information will facilitate early detection of affected animals by animal caregivers and veterinarians. Furthermore, this information will enable researchers to appropriately design studies to assess the effectiveness of pain mitigation strategies and design studies to assess the extent to which pain and sickness behaviors contribute to weight loss associated with IBK.\n\n\nMaterials and methods\n\nThis study is a hypothesis generating study and represents a secondary use of animals enrolled in an experimental study conducted to assess putative causal organisms for IBK in calves (Gould et al., 2013). This experimental study population provided a unique controlled setting for pain and sickness behaviors in calves with corneal scarification and IBK. Three trials (replicates) were conducted in January 2012 (Trial 1), May 2012 (Trial 2) and August 2012 (Trial 3). Dairy breed calves, predominantly Holstein genetics with some Jersey influence, and 8 to 12 weeks of age, were sourced from the Iowa State University (ISU) Dairy Farm (Trial 1 and Trial 2) and a private Iowa-based owner (Trial 3). Calves were housed in a biosecurity Level 3 facility at ISU in Ames, Iowa. For each trial, all enrolled calves were housed in a single room maintained at 20–21°C (68–70°F). Each calf was housed separately in raised 0.9 × 1.8 meter (3 × 6 foot) pens that provided no opportunity for calf-to-calf contact. Auditory contact among calves was not restricted, and visual contact among calves was limited to the unique position and location of each pen. Calves were provided free choice water and were fed mixed grass hay and a pre-mixed calf starter (Heartland Co-op, Des Moines, IA). To avoid cross contamination, caretakers and research personnel wore protective gloves and clothing when working with the calves. If personnel had physical contact with calves during animal husbandry and study related activities, protective items were changed before a new calf or equipment or facilities associated with another calf was contacted. Other biosecurity measures included providing separate feeders and individual automatic watering troughs. Approval for this study was obtained from the Iowa State University (ISU) Institutional Biosafety Committee (IBC#11-D-0017-A) and the Institutional Animal Care and Use Committee (IACUC 8-11-7187-B).\n\nPrior to enrollment in each replicate on day - 4, calves received an extensive ophthalmic examination by a board certified veterinary ophthalmologist and a supervised veterinary ophthalmology intern as described previously (Gould et al., 2013). Only calves without identified ocular abnormalities on day - 4 were enrolled.\n\nThe sample size for the original study was based on estimated IBK risk of infection between groups. The study enrolled 36 calves to obtain 80% power to detect an estimated 60% difference in risk between groups based on an expected 10% IBK risk in controls and at least 70% IBK risk in inoculated animals with significance level 0.05. It was not possible to calculate the power of the study to detect meaningful differences in pain or sickness behavior outcomes as estimates of normal levels or normal variation associated with IBK were not available prior to the study.\n\nUpon enrollment on day - 4, calves were allocated to one of three treatments using a random number generator (Microsoft Excel, 2007) and then the left or right eyes were randomly allocated for corneal scarification. The three treatments were:\n\na. 1) corneal scarification only (control)\n\nb. 2) corneal scarification with inoculation with Moraxella bovoculi (ATCC strain: BAA- 1259; Origin: California, depositor: Dr. J Angelos) (M. bovoculi)\n\nc. 3) corneal scarification with inoculation with Moraxella bovis (strain Epp63-300; Dr. Rosenbusch lab; Origin: National Animal Disease Center) (M. bovis)\n\nA concurrent negative control treatment was not relevant to the question posed by the original experiment therefore for the question assessed by this study in lieu of a concurrent negative control treatment it was necessary we used the response prior to scarification to represent behaviors expected under the no pain condition.\n\nCorneal scarification took place between 0800–1000 h on day 0. Only one eye from each calf was scarified. Scarification was accomplished by a researcher trained in the procedure and according to a published protocol (Rosenbusch & Ostle, 1986). Calves were individually restrained using a portable modified head restraint placed on the front of each pen. Prior to scarification, the cornea of one eye of each calf was anesthetized 3–5 minutes prior to the scarification procedure with topical 0.5% proparacaine hydrochloride (Bausch & Lomb Inc., Rochester NY). A sterilized wire brush approximately 5 mm in length was used to create 3–4 horizontal and vertical superficial epithelial scratches. To inoculate eyes with M. bovoculi or M. bovis, a sterile swab was rolled across a blood agar plate containing the organism. For each scarified eye, the swab was rolled or wiped across the cornea as well as introduced into the medial conjunctiva sac. To blind the allocation status of each eye, the researcher preparing the swabs concealed the allocation status from the researcher conducting the scarification procedure. Calves were restrained and observed for development of centrally located corneal ulcerations consistent with IBK on days +1, 3, 6, 8 and 10 relative to scarification. The results of the original microbiological causation study (Gould et al., 2013) and the mechanical nociception thresholds (Dewell et al., 2014) are published elsewhere. If a corneal ulcer was identified and reached 15mm diameter or wider, the calf was euthanized on the same day using an appropriate dose of sodium pentobarbital administered intravenously by or under the supervision of a licensed veterinarian. At the conclusion of the study, all remaining calves were euthanized using an appropriate dose of sodium pentobarbital administered intravenously by or under the supervision of a licensed veterinarian.\n\nCalf behavior was recorded using digital video recording. Video images were captured, utilizing three Noldus portable labs (Noldus Information Technology, Wageningen, NL), one for each cohort of four calves. One color Panasonic camera (WV-CP484, Kadoma, Japan) was mounted above each stall, and positioned to ensure maximum stall and calf visibility. The 12 cameras were divided into three zones, one for each Noldus video portable lab, based on the location of the stalls. Every zone contained four cameras each directed at a specific stall and fed into a multiplexer, which allowed the image to be recorded onto a PC using HandiAvi (v4.3, Anderson’s AZcendant Software, Tempe, AZ) at 30 frames per second. Color video with no audio was continuously recorded between 0500–2000 h from day -1 to day 10 relative to scarification. However, due to the rapid development of IBK ulcers in one scarification treatment group, behavior outcomes were collected from video recordings on day -1 and on day 0 after the scarification procedure only.\n\nBehavioral observations were collected by a single trained technician (BW) from 22 May 2013 until 3 July 2013, using Observer® (v10.1.548, Noldus Information Technology, Wageningen, NL). This person was unaware of the actual treatments the calves received and the allocation groups until after all data were collected and the preliminary data analysis conducted. The individual was aware that the study related to ocular lesions. Prior to data collection, an ethogram was developed by two co-authors (SM and RP). The behaviors were selected based on prior behavior studies involving pain and sickness behaviors (Duffield et al., 2010; Millman, 2007; Todd et al., 2007; Millman, 2013). Table 1 describes the ethogram. Blink rates and eye movements (open, closed, fixed and rolled) would be relevant behaviors associated with ocular pain; however, these behaviors were not included because the eyes of the calves could not be reliably seen throughout the duration of the video. Ear-flicking was omitted based on similar logic. Frequencies were recorded for behavioral events whereas durations were recorded for behavioral states. An event is a behavioral pattern of relatively short duration, and in our study independent events were separated by five or more seconds. In contrast, a state is a behavioral pattern of a longer duration such as posture or prolonged activity (Martin & Bateson, 2007).\n\nResearch technician training on the Observer® program and the ethogram comprised two weeks, and data collected was compared relative to data collected by a trained research associate with several years of experience (RP). Behavioral data collection formally began when the inter-observer reliability values were: proportion of agreements = 0.72, Kappa = 0.68, Rho = 0.98. Intra-observer reliability was checked periodically throughout the video observation period and were: proportion of agreements = 0.79, Kappa = 0.78, and Rho = 0.98.\n\nBehavioral data was collected using continuous sampling of hour segments of video recorded from 1230–1730 on day -1 (hereafter referred to as baseline time period) and on day 0 (hereafter referred to as scarification time period). See Figure 1 for timeline of study. These days of interest were selected to enable identification of changes in behavior associated with the scarification treatments when the full cohort of calves was present in each treatment. The time of day, 1230–1730 h, was selected because calves were least likely to be disturbed by caretakers or research personnel during this time period, there was sufficient light to enable detailed behavior observations and to accommodate the timing of the scarification procedure (0800–1000 h).\n\nPrior to data collection, video was cut using the Virtual Dub® software (Avery Lee, compiled with Microsoft Visual Studio 2005 for X86, version 1.9.9) into approximately 90 minute blocks to facilitate blinding of the observer to treatment, day and time. The order of video blocks observed was randomized using a random number generator in MS Excel®. Calves were observed in groups of four in real time and frequencies and durations according to the ethogram were recorded. The observer took periodic 5–10 minute breaks after observing 90 minutes of video.\n\nThe frequencies of head shaking, head scratching, and drinking over the five-hour period for baseline and scarification time periods were calculated as categorical variables. The durations for all other behaviors were calculated as continuous variables in minutes for the total observed time for each day over the five-hour period. For each behavior, the distribution was evaluated using visual examination of a box plot and presented based on current recommendations for reporting (Lang & Altman, 2013). After this preliminary analysis and data check, the researchers were unblinded to the treatments.\n\nThe results of an analysis of the pressure algometry data suggested that the calves experienced increased pain sensitivity after scarification as indicated by reduced mechanical nociception thresholds (Dewell et al., 2014). However, since only calves in the M. bovis treatment developed IBK-ACA (Gould et al., 2013), we hypothesized these calves would show sickness behavior associated with inflammation and proinflammatory cytokines. These observations lead to the a priori hypotheses (i.e. before the data were unblinded) that pain related behaviors should differ between baseline time period and scarification time periods for all calves. We also hypothesized that sickness behavior would occur in calves enrolled that received M. bovis, since these calves subsequently developed IBK, but would not occur in calves enrolled in the M. bovoculi or control treatments since these calves did not develop IBK. The statistical analysis were designed a priori to test these hypotheses.\n\nBehaviors observed that were expected to be important indicators of ocular pain (or irritation) were head-directed behaviors: head shaking, head rubbing and head scratching. The frequencies of these three head-directed behaviors (head scratching, head rubbing, and head shaking) were summed to create a variable that described the total frequency of any head-directed behavior. Similarly, although measured separately in the ethogram, the behavioral states standing and head rubbing and lying and head rubbing were combined into a single state termed head rubbing for analyses purposes. Head rubbing was a unique behavior of interest because we considered it both an event and a state so its frequency and duration were calculated.\n\nBehaviors in the ethogram that were expected to be important indicators of sickness behaviors were feeding, standing with head lifted, lying with the head lifted, and sleeping. Lying with head in sternal recumbency and lying and licking were recorded, but were rare and removed from further analysis. The behaviors lying with the head tucked to the left and lying with the head tucked to the right were observed separately, but were later combined for the analyses and are described in this study as sleeping based on previous research and simplicity (Ternman et al., 2012).\n\nFor all analyses the unit of analysis was the calf. All models were executed using PROC GLIMMIX (SAS 9.3, Inst. Inc., Cary, NC). Explanatory variables for all models were treatment (three scarification treatments), time period (two time periods prior to and following scarification i.e. baseline and scarification) and trial, which refers to the three trials conducted. All models included the fixed effects treatment, time period and the interaction between treatment and time period. Random effects were trial and animal. When main effects were significant based on a p value < 0.05, least square means were calculated and the differences between means were tested using Turkey-Kramer adjustment for multiple comparison when appropriate.\n\nFor head-directed behavior, two potential generalized linear models were assessed i.e., Poisson and negative binomial. A scaled Pearson’s statistic (χn2df where df is the degrees of freedom) was used to select the preferred modeling approach. We preferred the model with the statistics closer to 1. The negative binomial was chosen.\n\nCount data were analyzed with mixed effect negative binomial regression models, with treatment, time and interaction between treatment and time as fixed effects and animal nested in treatment and trial as random effects. Logarithm of total time of observation was used as offset in the negative binomial models.\n\nFor behaviors in which duration was calculated (head rubbing, feeding, standing with head lifted, lying with head lifted and sleeping), duration data were log transformed and analyzed with linear mixed models, with treatment, time and interaction between treatment and time as fixed effects and animal nested in treatment and trial as random effects. Logarithm of total time of observation was used as offset.\n\nFixed effects were not excluded based on non-significance. Diagnostic tests were conducted to check the model assumptions for all models. In addition, the normality assumptions and alternative models were checked. The akaike information criterion (AIC) criteria or likelihood ratio tests (LRT) were used to select the final model when the models were nested.\n\n\nResults\n\nOf the 36 animals purchased for the study, five calves were ineligible for enrollment due to pre-existing corneal abnormalities. Thirty-one enrolled calves were randomly allocated to the three treatments over the three trials, resulting in uneven numbers of calves per treatment: scarification only (n = 11), scarification and inoculation with M. bovoculi (n = 10), scarification and inoculation with M. bovis (n = 10). Behavior data from two calves enrolled in trial 3 were missing due to camera malfunctions. These missing data were from one calf in the M. bovis treatment and one calf in the M. bovoculi treatment. Therefore data for this study included scarification only (n = 11), scarification and inoculation with M. bovoculi (n = 9) and scarification and inoculation with M. bovis (n = 9). Details from results of other analysis associated with the study are available elsewhere (Gould et al., 2013).\n\nBased on visual assessment of the data, neither event nor state behaviors were normally distributed, therefore the data were reported as median, minimum, 25th quartile, 75th quartile and maximum values, compliant with current statistical reporting guidelines (Lang & Altman, 2013). Table 2 reports this summary information for event behaviors. Because drinking events were observed, but were too rare to be relevant to the hypothesis they are reported here in text. For drinking events, the median (minimum, 25th percentile, 75th percentile, maximum) during baseline time period for control, M. bovoculi and M. bovis treatments were 4.0 (0.0, 1.5, 6.5, 17.0), 3.0 (0.0, 2.0, 5.0, 9.0), and 4.0 (1.0, 2.0, 5.0, 6.0) respectively. The median (minimum, 25th percentile, 75th percentile, maximum) during scarification time period for control, M. bovoculi and M. bovis treatments were 4.0 (1.0, 3.0, 7.0, 11.0), 6.0 (2.0, 4.0, 7.0, 13.0), and 4.0 (2.0, 3.0, 6.0, 9.0) respectively.\n\nThe descriptive statistics for behavioral states during baseline and scarification time periods are reported in Table 3 and Table 4 respectively. All behaviors listed in the ethogram were observed in calves in each treatment during both time periods. Lying with the head lifted and feeding were the two most commonly recorded behaviors. Lying in sternal recumbency, lying and head rubbing, and lying and licking were very rare behaviors. The maximum disturbance time for all trials was 6.5 minutes; however the median disturbance time for all trials during the baseline time period and scarification time period was 0. Standing out of view was an extremely rare observation (median for all treatments = 0), and lying out of view was never observed.\n\nThe frequencies of the three separate head-directed behaviors were too rare to enable modeling. Consequently, we combined the three separate head-directed behaviors into one measure, and the results of the analysis of the combined head-directed outcome are reported. No significant interaction between treatment and time period was observed (p = 0.18). However, the effect of time period was significant (p = 0.0001), whereas the effect of treatment was not significant (p = 0.42). The least squares means comparison of the baseline time period and scarification time period was not significant for the control treatment (p = 0.75). However, it was significantly different for the M. bovoculi group (p value = 0.02) and the M. bovis group (p = 0.04). Table 5 shows the regression-based estimates and 95% confidence interval of the head-directed behavior frequencies for the three treatments during the baseline time period and scarification time period. These estimates can be used as a basis for future sample size determination. The random effect for animal was (0) and the estimate and standard error for trial term were 0.21 and 0.25 respectively.\n\nHead-directed behaviors are events in units of log of link. Head-directed behaviors back transformed are in units of frequency of event per hour. Head rubbing, feeding, standing with head lifted, lying with head lifted and sleeping are behavioral states in units of log of proportion.\n\na = units (log of link)\n\nb = units (frequency of event per hour)\n\nc = units (log of proportion)\n\nThe original proposed model for duration of time spent head rubbing, with trial included as a random effect, did not converge. Therefore, the model was modified to include trial as a fixed effect rather than a random effect. The p values for trial, time period, treatment and time period by treatment interaction were 0.65, 0.02, 0.23, and 0.08 respectively. The duration of time spent head rubbing did not differ significantly between baseline time period and scarification time period for the control treatment (p=1) and for the M. bovis treatment (p=0.92). However, duration of time spent head rubbing was significantly increased after scarification for the M. bovoculi treatment (p = 0.041). The variance component for the random effect and its standard error for animal were 0.36 and 0.58 respectively.\n\nIn no model was there a significant difference among treatments or between time periods for the total duration spent feeding, standing with head lifted, lying with head lifted or sleeping. The model-derived estimates of the transformed data are reported in Table 5. These estimates may inform future study design.\n\nFor the feeding outcome, the p values for the fixed effects were the following for time period, treatment and time periods by treatment interaction: 0.62, 0.56 and 0.09. The random effect variance estimate and its standard error for animals were 0.21 and 0.09 respectively. The random effects estimate and standard error for trial were 0.01 and 0.04 respectively.\n\nFor the standing with head lifted outcome, the p values of the fixed effects for time period, treatment and time period by treatment interaction were 0.01, 0.55 and 0.81 respectively. The random effects estimate and standard error for animal were 0.06 and 0.09 respectively. The random effects for estimate and standard error for trial were 0.07 and 0.09 respectively.\n\nFor the lying with head lifted outcome, trial was calculated as a fixed effect. The model estimate of the p value for the fixed effects for trial, time period, treatment and time periods by treatment interaction were 0.92, 0.37 and 0.76 and 0.69 respectively. The random effect variance estimate and its standard error for animal were 0.02 and 0.01 respectively.\n\nFor the sleeping outcome, the p values for the fixed effects time period, treatment and time period by treatment interaction were 0.09, 0.99, and 0.63 respectively. The estimate and standard error of the variance component for animal were 0.19 and 0.69 respectively. The random effect variance estimate and its standard error for time period were 0.11 and 0.30 respectively.\n\n\nDiscussion\n\nThe objectives of this research were to report the magnitude and variation of behavioral changes in calves with ocular injury (corneal scarification), infection and IBK-ACA. The motivation for such information is the need to improve detection of IBK and ocular injury, design studies that assess pain mitigation strategies to reduce animal suffering and decrease production losses associated with disease. In order to do this, it is necessary to first identify behavioral changes associated with ocular pain and sickness behavior, and then design studies with sufficient power to detect meaningful differences in these behaviors. An essential element of sample size calculations is specification of the alternative hypothesis, which describes what is considered a meaningful difference in pain and sickness associated changes in behavior and an estimate of expected variation. This study provides data that can be used for such purposes.\n\nFirst, our results suggest that a combined index of head-directed behavior has the potential to be used as a measure of ocular pain in calves. This inference comes from the finding that there was a significant time period effect for head-directed behaviors. Specifically, the frequency of head-directed behaviors increased during the scarification time period. If the same data form was used to compare interventions, then these values should be used to determine expected samples sizes for intervention studies. This finding is consistent with those obtained during the mechanical nociception threshold component of this study, as determined by pressure algometry applied to landmarks surrounding the scarified and non-scarified eyes on day -4 and day +1 relative to scarification (Dewell et al., 2014). Interestingly, mechanical nociception thresholds were affected at all landmarks (surrounding treated and healthy eyes, as well as on the center of the face) suggesting a calf-level change in response, perhaps due to general hyperalgesia associated with the scarification procedure or due to habituation to stressors of handling and restraint. The findings of this behavioral component of the study suggest the former interpretation, since the head-directed behaviors occurred in the home pen when handlers were not present and prior to the post-scarification nociception tests.\n\nOur results are consistent with other studies quantifying behavioral changes associated with head wounds resulting from disbudding and dehorning surgeries. Frequencies of head shaking, head rubbing, and ear flicking by calves increased after hot-iron cautery disbudding surgery relative to behaviors observed after calves received a sham procedure with an unheated dehorner (Duffield et al., 2010). Behavioral responses were lower in calves that received the nonsteroidal anti-inflammatory drug (NSAID) meloxicam at the time of disbudding relative to calves that received a saline treatment (Duffield et al., 2010). Similarly, frequencies of head shaking, ear flicking, and head rubbing following cautery disbudding were found to be lower in calves that received ketoprofen versus those that did not receive an NSAID (Duffield et al., 2010; Faulkner & Weary, 2000). Head rubbing has also been reported when disbudding is performed using caustic paste, and the response was not mitigated by the NSAID flunixin meglumine (Stillwell et al., 2008). These other studies suggested that the changes in head-directed behavior are not specific to ocular pain. We interpret these changes in behavior as indicators of ocular pain, but it is also possibly due to irritation or itching from the head-restraint or procedure. Calves in our study were examined daily for any physical abnormalities and no lesions associated with trauma from the head-restraint or dermatological conditions of the head were noted. The consistency between behavioral and nociception responses in this study during, before and after scarification, together with the similar consistency in behavioral and nociception responses before and after disbudding which are mitigated when calves are provided with NSAID analgesia (Duffield et al., 2010), supports the pain interpretation.\n\nWe hypothesized that sickness behavior would be expressed by calves that developed IBK lesions, due to cytokines resulting from the inflammatory response. We expected to see decreases in the time spent feeding and standing, together with increases in time spent resting because febrile animals with infections commonly display depression, anorexia, altered grooming patterns and increased time sleeping (Hart, 1988). These sickness behaviors are presumed to help the body conserve energy and recover from the infection or disease, and are mediated by pro-inflammatory cytokines that act in a paracrine and endocrine manner at the site of inflammation, but also as neurotransmitters that can be produced by glial cells within the central nervous system (Dantzer & Kelley, 2007). M. bovis is a Gram-negative coccobacillus bacterium, and would be expected to produce the classical sickness behavior response (Postma et al., 2008). There are two primary reasons that could explain why our data did not document an association between sickness behavior and treatment.\n\nFirst, it is possible the calves were experiencing sickness motivation, but did not demonstrate changes in the variables we measured to indicate sickness behavior. The behaviors selected for our ethogram are consistent with expression of sickness behavior in cattle (Borderas et al., 2008; Hart, 1987; Proudfoot et al., 2012; Weary et al., 2009). Calves challenged with lipopolysacchride displayed reductions in hay eating, self-grooming and increased duration of lying (Borderas et al., 2008). However, the small pens in this study limited the behavioral repertoire of the calves, making changes from baseline behavior difficult to observe. Changes from baseline (healthy) behavior to sickness behavior may have been more apparent if calves were housed in larger and more complex pens in which they could interact with more stimuli. For example, play behavior by calves, such as bucking and running, is suppressed after cautery disbudding (Mintline et al., 2013). Calves were individually housed so they did not compete for access to feed and they were weaned calves so their dam’s influence on behavior was absent. Individual calf variation may potentially explain the lack of increased sickness behavior during scarification time period due to individual variation in regard to temperament, physiology and tolerance. Numerous environmental factors were controlled for in the biosecure research facility, including climactic factors, which may influence the perception of pain, development of IBK and expression of behavior. Further, our study controlled for insect interference because other studies have shown confounding between increase counts of ear flicks following painful procedures when high insect burden is present (Theurer et al., 2013a). Since calves were housed individually, social facilitation of behavior was less likely, but may have occurred due to auditory and visual contact with neighboring calves. However, since all treatments were housed in a single room and randomly assigned to pens, social facilitation of behavior would not explain differences in behavioral outcomes reported.\n\nAnother possible reason why sickness behaviors were not significantly different between time periods was the length of time between onset of inoculation and development of systemic infection. As expected, M. bovis calves developed IBK ulcers, but it is possible that the time from scarification to observation was insufficient to detect sickness behavior. Previous experience with this challenge model suggested that M. bovis infected animals would develop IBK (Rosenbusch & Ostle, 1986), and we expected to observe a longer duration of infection (days versus hours) before reaching the IBK ulcer diameter (15mm), which was identified as the humane endpoint for objective of the primary focus of the challenge study (Gould et al., 2013). In our study some calves developed IBK lesions within four hours of scarification, and others did not develop lesions until 24 hours after scarification. Behavioral changes in calves challenged with Mannheimia haemolytica, such as duration of lying, occur from d0 through d+ 8 (Theurer et al., 2013b), suggesting that a longer interval may have provide more opportunity to observe sickness behavior. However, increases in frequency of occurrence and decreases in duration of lying bouts of cattle infected with the gastrointestinal parasite Osteragia osteragi occurred only for animals inoculated at high versus moderate or low doses (Szyszka & Kyriazakis, 2013). Furthermore, nursery age pigs were found display significant behavioral changes within hours of inoculation with swine influenza virus (Millman, 2012), suggesting modulation of sickness behavior according to host, agent and environmental factors deserves greater scrutiny.\n\nContinuous video observation was chosen for this study over other options such as scan, time, and focal animal sampling for many reasons. The rationale for this choice, was that continuous video monitoring provided advantages of avoiding the potential for suppression of behavior due to the presence of a human observer in the room, facilitating collection of subtle behavior patterns and rare behaviors that are difficult to capture with instantaneous scan sampling, opportunity for breaks and greater concentration when data are collected over long periods of time, blinding of the observer to the treatments and time period segments to avoid bias, and validation of the ethogram and data collected using inter- and intra-observer reliability tests prior to and during data collection phases. Given that so little information is known about the topic of ocular pain in cattle, the labor and resources were justified. It may be advisable to sample smaller segments of video in subsequent studies to reduce the time-consuming and labor-intensive aspects of continuous scanning. However, we were unable to conduct statistical analyses for some behaviors due to their rarity. Biotelemetry may provide opportunities to collect some behavior automatically, such as with accelerometers attached to a limb to measure bouts of general activity and rest, or more specifically to quantify head only activity if possible affix the device to the calf’s head using a halter or adhesive. If feasible, biotelemetry could reduce the labor associated with behavior data collection substantially.\n\nAlthough the single observer was trained, misclassification bias is always a possibility in observational studies due to the difficulty in consistently observing the same behavior through multiple videos. If bias did occur due to a single reader, then we hypothesized that the direction of bias was to decrease the frequency of events, i.e. underestimation. That is, while one event was recorded, another may have been missed. This occurrence is an artifact of taking observations off of a quad unit where four unique animals were observed simultaneously, rather than observing one animal from a single camera view. However, we reduced the potential for observation bias by training the observer to a standard kappa compared to a very experienced reader (RP) and to periodically check the intra-reader reliability.\n\n\nConclusion\n\nOur study showed that frequency of head-directed behaviors (head shaking, head rubbing and head scratching) was associated with ocular scarification, and together with changes in nociception thresholds, suggestive of ocular pain. As beef calves are often observed at a distance on pasture and corneas are difficult to evaluate remotely, head rubbing, shaking and scratching, may be early behavioral indicators that producers could use to identify calves to be more closely evaluated for the presence of corneal lesions or other ocular abnormalities. Sickness behaviors were not found to be significantly associated with scarification or IBK ulcers when measured up to 8 h post-scarification. These results describe the magnitude and variability associated with behavioral responses to ocular scarification and IBK ulcers in a challenge model, and can be used for determining sample size calculations for future studies addressing pain mitigation for cattle suffering from ocular injury or disease, such as IBK or for exploring associations between behavior and performance during naturally occurring IBK infections. In conclusion, our research expands the breadth of knowledge for pain and sickness behaviors in cattle, specifically behaviors associated with IBK.\n\n\nData and software availability\n\nZenodo: Dataset and Code: Pain and sickness behavior associated with corneal lesions in dairy calves. 10.5281/zenodo.18854 (Woods et al., 2015).", "appendix": "Author contributions\n\n\n\nPrimary: B.J. Woods\n\nContributing authors in alphabetical order: N.A. da Silva, R.D. Dewell, R.L. Parsons, C. Wang\n\nSenior/supervisory author: S.T. Millman and A.M. O’Connor\n\nAll authors acknowledge that they met the following ICMJE standards for authorship (http://www.icmje.org/roles_a.html):\n\na. 1) Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND\n\nb. 2) Drafting the work or revising it critically for important intellectual content; AND\n\nc. 3) Final approval of the version to be published; AND\n\nd. 4) Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was partially supported by a grant to Dr. O’Connor from the Iowa Healthy Livestock Advisory Council 109-05-49.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors thank Stacie Gould for valuable technical assistance during the trials.\n\n\nReferences\n\nBorderas TF, de Passillé AM, Rushen J: Behavior of dairy calves after a low dose of bacterial endotoxin. J Anim Sci. 2008; 86(11): 2920–7. PubMed Abstract \| Publisher Full Text\n\nBrown MH, Brightman AH, Fenwick BW, et al.: Infectious bovine keratoconjunctivitis: a review. J Vet Intern Med. 1998; 12(4): 259–66. PubMed Abstract \| Publisher Full Text\n\nDantzer R, Kelley KW: Twenty years of research on cytokine-induced sickness behavior. Brain Behav Immun. 2007; 21(2): 153–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDewell RD, Millman ST, Gould SA, et al.: Evaluating approaches to measuring ocular pain in bovine calves with corneal scarification and infectious bovine keratoconjunctivitis-associated corneal ulcerations. J Anim Sci. 2014; 92(3): 1161–72. PubMed Abstract \| Publisher Full Text\n\nDuffield TF, Heinrich A, Millman ST, et al.: Reduction in pain response by combined use of local lidocaine anesthesia and systemic ketoprofen in dairy calves dehorned by heat cauterization. Can Vet J. 2010; 51(3): 283–88. PubMed Abstract \| Free Full Text\n\nFaulkner PM, Weary DM: Reducing pain after dehorning in dairy calves. J Dairy Sci. 2000; 83(9): 2037–41. PubMed Abstract \| Publisher Full Text\n\nFunk L, O'Connor M, Maroney M, et al.: A randomized and blinded field trial to assess the efficacy of an autogenous vaccine to prevent naturally occurring infectious bovine keratoconjunctivis (IBK) in beef calves. Vaccine. 2009; 27(34): 4585–90. PubMed Abstract \| Publisher Full Text\n\nGelatt KN: Essentials of Veterinary Ophthalmology. Wiley-Blackwell: Ames, IA, 2008. Reference Source\n\nGeorge LW: Clinical Infectious Bovine Keratoconjunctivitis. Comp Cont Educ Pract. 1984; 6: 712–20.\n\nGould S, Dewell R, Tofflemire K, et al.: Randomized blinded challenge study to assess association between Moraxella bovoculi and infectious bovine keratoconjunctivitis in dairy calves. Vet Microbiol. 2013; 164(1–2): 108–15. PubMed Abstract \| Publisher Full Text\n\nHart BL: Behavior of sick animals. Vet Clin North Am Food Anim Pract. 1987; 3(2): 383–91. PubMed Abstract\n\nHart BL: Biological basis of the behavior of sick animals. Neurosci Biobehav Rev. 1988; 12(2): 123–37. PubMed Abstract \| Publisher Full Text\n\nHeinrich A, Duffield TF, Lissemore KD, et al.: The impact of meloxicam on postsurgical stress associated with cautery dehorning. J Dairy Sci. 2009; 92(2): 540–7. PubMed Abstract \| Publisher Full Text\n\nHeinrich A, Duffield TF, Lissemore KD, et al.: The effect of meloxicam on behavior and pain sensitivity of dairy calves following cautery dehorning with a local anesthetic. J Dairy Sci. 2010; 93(6): 2450–7. PubMed Abstract \| Publisher Full Text\n\nKizilkaya K, Tait RG, Garrick DJ, et al.: Genome-wide association study of infectious bovine keratoconjunctivitis in Angus cattle. BMC Genet. 2013; 14: 23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLang TA, Altman DG: Basic statistical reporting for articles published in biomedical journals: The “Statistical Analyses and Methods in the Published Literature” or The SAMPL guidelines. In Science Editors' Handbook, European Association of Science Editors, 2013. Reference Source\n\nMartin P, Bateson P: Measuring Behavior: An Introductory Guide. Cambridge University Press, 2007. Reference Source\n\nMillman ST: Sickness behaviour and its relevance to animal welfare assessment at the group level. Animal Welfare. 2007; 16(2): 123–25. Reference Source\n\nMillman ST: How do we aid convalescence and improve pig welfare? Proceedings of the American Association of Swine Veterinarians 2012 Annual Meeting, General Session – Integrating Science, Welfare, and Economics in Practice. Denver, CO. 2012; 423–426.\n\nMillman ST: Behavioral responses of cattle to pain and implications for diagnosis, management, and animal welfare. Vet Clin North Am Food Anim Pract. 2013; 29(1): 47–58. PubMed Abstract \| Publisher Full Text\n\nMintline EM, Stewart M, Rogers AR, et al.: Play behavior as an indicator of animal welfare: Disbudding in dairy calves. Appl Anim Behav Sci. 2013; 144(1–2): 22–30. Publisher Full Text\n\nO'Connor AM, Shen HG, Wang C, et al.: Descriptive epidemiology of Moraxella bovis, Moraxella bovoculi and Moraxella ovis in beef calves with naturally occurring infectious bovine keratoconjunctivitis (Pinkeye). Vet Microbiol. 2012; 155(2–4): 374–80. PubMed Abstract \| Publisher Full Text\n\nO'Connor AM, Wellman NG, Evans RB, et al.: A review of randomized clinical trials reporting antibiotic treatment of infectious bovine keratoconjunctivitis in cattle. Anim Health Res Rev. 2006; 7(1–2): 119–27. PubMed Abstract \| Publisher Full Text\n\nPostma GC, Carfagnini JC, Minatel L: Moraxella bovis pathogenicity: an update. Comp Immunol Microbiol Infect Dis. 2008; 31(6): 449–58. PubMed Abstract \| Publisher Full Text\n\nProudfoot KL, Weary DM, von Keyserlingk MAG: Linking the social environment to illness in farm animals. Appl Anim Behav Sci. 2012; 138(3–4): 203–15. Publisher Full Text\n\nRosenbusch RF, Ostle AG: Mycoplasma bovoculi infection increases ocular colonization by Moraxella ovis in calves. Am J Vet Res. 1986; 47(6): 1214–16. PubMed Abstract\n\nSnowder GD, Van Vleck LD, Cundiff LV, et al.: Genetic and environmental factors associated with incidence of infectious bovine keratoconjunctivitis in preweaned beef calves. J Anim Sci. 2005; 83(3): 507–18. PubMed Abstract\n\nStillwell G, Lima MS, Broom DM: Comparing plasma cortisol and behaviour of calves dehorned with caustic paste after non-steroidal-anti-inflammatory analgesia. Livest Sci. 2008; 119(1–3): 63–69. Publisher Full Text\n\nSzyszka O, Kyriazakis I: What is the relationship between level of infection and ‘sickness behaviour’ in cattle? Appl Anim Behav Sci. 2013; 147(1–2): 1–10. Publisher Full Text\n\nTernman E, Hanninen L, Pastell M, et al.: Sleep in dairy cows recorded with a non-invasive EEG technique. Appl Anim Behav Sci. 2012; 140(1–2): 25–32. Publisher Full Text\n\nTheurer ME, Amrine DE, White BJ: Remote noninvasive assessment of pain and health status in cattle. Vet Clin North Am Food Anim Pract. 2013a; 29(1): 59–74. PubMed Abstract \| Publisher Full Text\n\nTheurer ME, Anderson DE, White BJ, et al.: Effect of Mannheimia haemolytica pneumonia on behavior and physiologic responses of calves during high ambient environmental temperatures. J Anim Sci. 2013b; 91(8): 3917–29. PubMed Abstract \| Publisher Full Text\n\nTodd CG, McKnight DR, Millman ST, et al.: An evaluation of meloxicam (Metacam (R)) as an adjunctive therapy for calves with neonatal calf diarrhea complex. J Anim Sci. 2007; 85: 369–69.\n\nWatkins LR, Maier SF: The pain of being sick: implications of immune-to-brain communication for understanding pain. Annu Rev Psychol. 2000; 51: 29–57. PubMed Abstract \| Publisher Full Text\n\nWeary DM, Huzzey JM, von Keyserlingk MA: Board-invited review: Using behavior to predict and identify ill health in animals. J Anim Sci. 2009; 87(2): 770–77. PubMed Abstract \| Publisher Full Text\n\nWilliams DL: Welfare issues in farm animal ophthalmology. Vet Clin North Am Food Anim Pract. 2010; 26(3): 427–35. PubMed Abstract \| Publisher Full Text\n\nWoods B, Millman S, da Silva N, et al.: Dataset and Code: Pain and sickness behavior associated with corneal lesions in dairy calves. Zenodo. 2015. Data Source" }	[ { "id": "9945", "date": "13 Oct 2015", "name": "Richard Laven", "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\nIBK is a common painful condition calves, yet there is limited evidence of its impact on calf behaviour. Previous research by this group has looked at the changes in pain sensitivity associated with ulceration. This paper sets out to look at behavioural changes associated with IBK.The study is clear and well-planned with three distinct treatment groups (scarification, scarification + apparently non-pathogenic bacteria; scarification + pathogenic bacteria). Unfortunately, as with the previous pain sensitivity study, the impact of the lesions of IBK resulting from infection of a scarified cornea is obscured by the impact of the scarification required as part of the disease-producing model. This limits the value of the study as a study of IBK rather than ocular pain. This is further compounded by the short time interval over which the data were collected.However, the data collection process is well-planned; in particular, the choice of measures to include in the ethogram is well explained and based on previously published data. The statistics are, for the most part, clearly explained, even though they are complex. One are that was not clear was the choice of model. The paper states “A scaled Pearson’s statistic (χ2n/df where df is the degrees of freedom) was used to select the preferred modeling approach. We preferred the model with the statistics closer to 1” to which the immediate questions for the non-statistician are how and why?There is one statistical misstep, which is repeated on multiple occasions. After demonstrating no effect of treatment or its interaction with time on behaviour, they then compare the baseline and scarification results for each treatment and report that there were treatment effects for the differences. For example, for the combined head behaviours there was no effect of treatment or interaction with treatment and time (p=0.42 and 0.18, respectively); yet it is reported that for control group baseline and scarification were not different (p=0.75) but they were for the two other groups (p=0.02 and 0.04). This is effectively saying there is an interaction between treatment and time when it’s just been shown that there isn’t. As it’s not really referred to later losing this analysis does not affect the conclusions of the data.The conclusion that head based behaviours could be used to detect early signs of ocular disease is valid, but its utility is reduced by the problems with the model of producing the disease as all of the changes found in this study are really a response to the scarification required to produce corneal damage that can then become infected by M. bovis. A better model with a less painful initial stage would be useful – perhaps introducing naïve calves to a group of previously infected cattle (either naturally or artificially induced); this should be discussed in the paper.", "responses": [] }, { "id": "12455", "date": "17 Feb 2016", "name": "Richard B Evans", "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\nOne reviewer (Dr. Laven) of this article noted an apparent conflict in the statistical reporting. He states, There is one statistical misstep, which is repeated on multiple occasions. After demonstrating no effect of treatment or its interaction with time on behaviour, they then compare the baseline and scarification results for each treatment and report that there were treatment effects for the differences. For example, for the combined head behaviours there was no effect of treatment or interaction with treatment and time (p=0.42 and 0.18, respectively); yet it is reported that for control group baseline and scarification were not different (p=0.75) but they were for the two other groups (p=0.02 and 0.04). This is effectively saying there is an interaction between treatment and time when it’s just been shown that there isn’t. As it’s not really referred to later losing this analysis does not affect the conclusions of the data. This is not a misstep. A lack of statistical significance for the interactions means that there wasn’t enough evidence to detect an interaction, not that an interaction doesn’t exist. The interaction test will have lower power than other tests. The authors correctly double checked the group significance cross-sectionally. It’s a good thing they did, because there is a slight interaction effect. That careful analysis is typical of this paper. The authors methodically verify each step of the study, leaving the reader with little doubt as to the strength of the conclusions. Enough details are provided so that another research team could replicate the study.", "responses": [] } ]	1	https://f1000research.com/articles/4-546
https://f1000research.com/articles/3-292/v1	01 Dec 14	{ "type": "Research Article", "title": "Countries’ Biomedical Publications and Attraction Scores", "authors": [ "Qinyi Xu", "Andrea Boggio", "Andrea Ballabeni", "Qinyi Xu", "Andrea Boggio" ], "abstract": "Studying publication volumes at the country level is key to understanding and improving a country’s research system. PubMed is a public search engine of publications in all life sciences areas. Here, we show how this search engine can be used to assess the outputs of life science-related research by country. We have measured the numbers of publications during different time periods based on the country of affiliation of the first authors. Moreover, we have designed scores, which we have named Attraction Scores, to assess the relative focus either toward particular types of studies, such as clinical trials or reviews, or toward specific research areas, such as public health and pharmacogenomics, or toward specific topics, for instance embryonic stem cells; we have also investigated a possible use of these Attraction Scores through a correlation analysis with regulatory policies. We have weighed the statistics against general indicators such as country populations and gross domestic products (GDP). During the 5-year period 2008-2012, the United States was the country with the highest number of publications and Denmark the one with the highest number of publications per capita. Among the 40 countries with the highest GDPs, Israel had the highest publications-to-GDP ratio. Among the 20 countries with the most publications, Japan had the highest Attraction Score for induced pluripotent stem (iPS) cells and Italy the highest proportion of review publications. More than 50% of publications in English were from countries in which English is not the primary language. We show an assorted and extensive collection of rankings and charts that will inform scholars and policymakers in studying and improving the research systems both at the national and international level.", "keywords": [ "Publication output at the country level has been assessed by means of different tools and methodologies. Both academic groups and private companies", "by using different and complementary approaches", "have offered valuable information to scholars and policymakers1–6. Some of these efforts have focused specifically on the life science sector by taking advantage of PubMed (http://www.ncbi.nlm.nih.gov/pubmed/)6–9", "a free and public search engine that provides access to over 24 million citations in all fields of life sciences", "mostly located in the MEDLINE (Medical Literature Analysis and retrieval System Online) bibliographic database. PubMed became accessible to the public at no charge in June 1997 and has been maintained by the United States National Library of Medicine (NLM) at the National Institutes of Health (NIH). Over the years", "it has become a very popular database among scientists and nonscientists interested in retrieving research publications in various biological and biomedical fields. One of the most valuable features is users’ ability to restrict searches through different filters and settings. Users can target their searches by restricting the query to publications that meet various requirements—including time of publication", "type of publication", "type of research", "language", "authors’ affiliation", "journal", "and other criteria—and by searching for specific keywords in the full text of the article or in sub-parts", "such as title and abstract. The system uses the Boolean operators AND", "OR and NOT to facilitate filtering and refining of the searches." ], "content": "Introduction\n\nPublication output at the country level has been assessed by means of different tools and methodologies. Both academic groups and private companies, by using different and complementary approaches, have offered valuable information to scholars and policymakers1–6. Some of these efforts have focused specifically on the life science sector by taking advantage of PubMed (http://www.ncbi.nlm.nih.gov/pubmed/)6–9, a free and public search engine that provides access to over 24 million citations in all fields of life sciences, mostly located in the MEDLINE (Medical Literature Analysis and retrieval System Online) bibliographic database. PubMed became accessible to the public at no charge in June 1997 and has been maintained by the United States National Library of Medicine (NLM) at the National Institutes of Health (NIH). Over the years, it has become a very popular database among scientists and nonscientists interested in retrieving research publications in various biological and biomedical fields. One of the most valuable features is users’ ability to restrict searches through different filters and settings. Users can target their searches by restricting the query to publications that meet various requirements—including time of publication, type of publication, type of research, language, authors’ affiliation, journal, and other criteria—and by searching for specific keywords in the full text of the article or in sub-parts, such as title and abstract. The system uses the Boolean operators AND, OR and NOT to facilitate filtering and refining of the searches.\n\nWith the goal of assisting scholars and policymakers in studying and improving research systems at the national and international level, we present a methodology that deploys PubMed to assess “bio” publication output and the sharing of publications for certain types of research and topics of interest. We also present data that can be retrieved using this methodology. Though PubMed search engine and MEDLINE database had been previously used to assess the quantity of research publications of countries7–11, our study constitutes the most recent, assorted and refined assessment of publication output in the life sciences. Thanks to the ability to attribute a paper to various countries, we present data of publication output by country and supranational regions. Further, we have used various stringency criteria in order to check the method’s robustness. Thus, we have analyzed publication output in different time ranges and calculated publication output in relation to various country-specific statistics (populations, GDPs, research and development (R&D) expenditures and presence/absence of English as a primary language). Finally, we have created a new score (Attraction Score) that measures the relative weight of publications for certain kinds of research or certain topics of interest. By analyzing the correlation between the Attraction Scores for human embryonic stem cells and the regulatory policies, we propose an example of how these Attraction Scores could be used to assess the research impact of regulatory policies. We believe that the methodology and the graphic representations of data will provide valuable and easy-to-grasp information that can assist professionals and the general public in understanding and improving biomedical research governance.\n\n\nMethods\n\nThe number of publications by country was determined by inserting the name/s of the countries in the “affiliation” field in the “advanced” section of the PubMed search engine. In the event of publications with authors based in different countries, we chose to attribute the paper only to the country of the leading author rather than collaborators. This was facilitated by the fact that PubMed only reports affiliation information of the first author of articles published before 2014 (http://www.ncbi.nlm.nih.gov/books/NBK3827/#pubmedhelp.Affiliation_AD) (see also Discussion section). Since this study focuses mainly on publications from 1993 to 2012, we have attributed a paper to a certain country based on the first author’s affiliation. We do not believe that this is a problem: in the vast majority of cases, first author and corresponding author are either the same person or they work in the same research institution/geographical area and therefore have the same country affiliation. To address the problem of countries recorded under multiple names, we have used the different designations separated by the Boolean operator OR. For example, publications of the United Kingdom were searched by inserting “united kingdom OR uk” in the “affiliation” field. Publications of the United States were searched by inserting “us OR usa OR united states” (we observed only small differences when inserting “usa OR united states”, see Discussion section). To identify the year of publication, we used the “Custom range” function (that is equivalent to using the “year:year[dp]” syntax in the search field). Most of the searches were made for the 5-years period 2008–2012, but we also investigated the 5-years periods 1993–1997, 1998–2002, and 2003–2007 to determine changes in publications volumes over time. Searches were also performed for individual years by using the “Custom range” function or by using the CSV downloads of the automatically retrieved yearly counts. We constrained publication output primarily on papers reporting original research. Therefore, unless otherwise indicated, we excluded reviews from our queries. We did so by using the operator NOT before the word “review” typed in the “Publication Type” field.\n\nIn generating publication data by continents, we added the number of publications of all countries in a certain continent. America was divided into North-Central America and South America. The publications of Russia and Turkey were allocated half to Asia and half to Europe. With regard to the European Union, we added the number of publications of the 28 countries that have joined the EU to present day. It should be noted, however, that some of the countries joined the European Union between 1993 and 2012. (One country, Croatia, joined the EU in 2013, and thus, it was not part of the EU at any time during the time periods under investigation). The total numbers of publications for the entire world were calculated by leaving the affiliation fields blank.\n\nIn addition to the standard criteria discussed so far (criteria A), we also conducted searches using various levels of stringency to criteria A. We thus added the “Journal Article” filter (in “Article types”) and “English” (in “Languages”) to criteria A to generate criteria B; the “Journal Article” filter (in “Article types”) to generate criteria C; the “English” filter (in “Languages”) to generate criteria D. The percentages of publications written in English were estimated by dividing the number of publications obtained using criteria D by the number of publications obtained using criteria A and then by multiplying the result by 100. The countries considered to have English as primary languages are shown in the data file. We observed only small differences between the counts obtained by using these four criteria (See Results section). We also decided to identify papers that reported the results of clinical trials. To this end, we created criteria E by adding the “Clinical Trial” filter to criteria A. This way, we were able to retrieve only publications based on clinical trial studies.\n\nWe chose to retrieve publications in certain specific areas or that discuss specific topics. To this end, we inserted in quotes the chosen research areas or topics of interest (for example, “public health” or “personalized medicine”) in the “Title/Abstract” field in the “Advanced” section of PubMed. This step allowed as to generate Attraction Scores for areas or topics (named “Area or Topic Attraction Scores”), which were calculated by dividing the number of publications in certain research areas or discussing certain topics, in a certain country, and within a certain period, by the total number of publications in the same country/period and by multiplying it by 10,000 to obtain easy-to-read numbers. The Area/Topic Attraction Scores were calculated by using criteria B because, as the searched areas or topics were English words, we wanted to exclude the (few) publications written in languages other than English. Nonetheless, we run the same query also using criteria A, and we obtained very similar results. Since the results are irrelevant, we decided not to show data of queries with criteria A in this paper. In the case of the “human embryonic stem cells” topic, we also used a “hESC/ESC Score”, defined as the ratio between number of publications obtained by using the search term “human embryonic stem cells” in the “Title/Abstract” field and the number of publications obtained by using the search term “embryonic stem cells” in the “Title/Abstract” field.\n\nSimilarly, we also generated “Clinical Trial Attraction Scores” by dividing the numbers of publications based on clinical trials studies retrieved using Criteria E by the total number of publications, retrieved using criteria A, in the same country and in the same period, and by multiplying the result by 10,000.\n\nFinally, we also generated Review Attraction Scores by comparing the numbers of publications obtained by criteria A with the numbers of publications obtained by criteria A+Reviews (i.e. without the exclusion of reviews) and by multiplying the result by 100 (in this case the score is the percentage).\n\nWe studied publication output during the 5-year period 2008–2012 relative to population, gross domestic products (GDP), and R&D expenditure. Population data (http://data.worldbank.org/indicator/SP.POP.TOTL) and GDP data (http://data.worldbank.org/indicator/NY.GDP.MKTP.CD) were retrieved from World Bank databases. We chose 2011 as a reference year. R&D expenditure data were retrieved from Battelle (a private nonprofit science and technology development company) 2014 Global R&D Funding Forecast (http://www.battelle.org/docs/tpp/2014_global_rd_funding_forecast.pdf) and expressed as GERD (Gross Expenditure on Research and Development) of year 2012 with Purchasing Power Parity in US$ billion. Data on population by continent were obtained from estimates published by the US Census Bureau (http://www.census.gov/compendia/statab/2012/tables/12s1330.pdf) for the year 2010. The numbers of total citable documents by country were retrieved from the SCImago Journal & Country Rank (http://www.scimagojr.com/index.php).\n\nWe repeated several searches one year apart to evaluate the consistency of the results across time. Several searches were therefore run a first time in July/August of 2013 and then repeated in September/October of 2014. We obtained very similar results, including for the year 2012 (See Results section). This shows that PubMed is rapidly updated and stable across time.\n\nThe classification of policies that regulate the use of human embryonic stem cells is based on the one previously developed by the Hinxton Group (http://www.hinxtongroup.org), a consortium fostering international cooperation in stem cell research. Other sources for country-specific policies are referenced in the Results section.\n\n\nResults\n\n\n\nWe determined the number of publications of countries for the 5-year periods 1993–1997, 1998–2002, 2003–2007, and 2008–2012. We divided the sums of publications for all countries by the total numbers of world publications (obtained without searching any country name in the “Affiliation” field, see Methods section), in the same time periods. We used both criteria A and criteria B for this analysis. We observed that the percentage of publications containing a country name in the affiliation increases with time. In the time period 2008–2012, the proportion of papers with a country name in the affiliation is 87.8% and 97.7% by using criteria A and criteria B, respectively (Figure S1) (sheet 1, including also the number of publications for all the countries for the time periods 2003–2007 and 2008–2012) (sheets of the database show results either from 2013 or 2014 searches. When not specified, the searches were made in 2013. When not specified, searches were for the time period 2008–2012). These data indicate that the methodology can plausibly effectively estimate the volumes of life science and biomedical publications of countries.\n\nWe determined the number of total world publications (obtained without searching any country name in the “Affiliation” field, see Methods section) for the time periods 1993–1997, 1998–2002, 2003–2007, and 2008–2012. Publications increased with a nearly constant rate during the four time periods, with the volume of the time period 2008–2012 corresponding to over 3.8 million publications (reviews excluded according to our above described searching criteria), roughly the double of the volume of the time period 1993–1997 (Figure 1) (sheet 2).\n\nCriteria A.\n\nWe determined the number of world publications based on clinical trial studies (obtained without searching any country name in the “Affiliation” field, see Methods section), for the time periods 1993–1997, 1998–2002, 2003–2007, and 2008–2012. Also, this type of publication increased at a nearly constant rate throughout this time span. Interestingly, the share of publications based on clinical trial studies has remained nearly constant (close to 5% of total publications) during this time span (Figure 2) (sheet 2).\n\nThe proportions of publications that are Clinical Trial publications are shown.\n\nCriteria E.\n\nWe estimated the share of world publications written in English by calculating the ratio between the numbers of publications determined by criteria D and the numbers of publications determined by criteria A for the time periods 1993–1997, 1998–2002, 2003–2007, and 2008–2012. The number of publications in English slightly increased during the four time periods, being 88.6% in the time period 1993–1997 and 93.3% in the time period 2008–2012 (Figure S2) (sheet 2).\n\nWe ranked all the countries according to their numbers of publications. In Figure 3 and Figure S3 (sheets 3 and 4), we show the charts with the 20 and 40 countries, respectively, with the most publications in the time period 2008–2012. With over one million publications, the United States represents by far the country with more publications than any other country, representing almost one-third of all world publications during the time period 2008–2012. The second-ranked country is China with a share of publication that is 28.5% of those attributed to the United States. In Figure 4 (sheet 4), we show a pie chart with the 25 countries with the most publications, including a “slice” representing the rest of the world (representing 10.0% of the total publications). We also tested the four different criteria (see Methods section) to determine this ranking. We noticed only minor differences in the numbers. The relative standard deviation was, on average, 1.9% for the 20 countries with the most publications (sheet 3). The only substantial differences between the four different criteria were for China (with a relative standard deviation equal to 13.7%) and France (with a relative standard deviation equal to 7.1%). These differences can be attributed primarily to the activation of the English language filter (See results below).\n\nCriteria A.\n\nThe part of publications other than the top 25 countries is shown as “rest of the world” (in pale red). In this case, differently from Figure 3, the data were obtained from 2013 searches.\n\nCriteria A.\n\nIn order to control for the consistency of these data over time, we ran several searches at two separate times at least one year apart (July/August 2013 and September/October 2014). In sheet 3 we show the differences in the total world publications (i.e. no affiliation specification) for the four different criteria. The differences were very small: below 1.5% for all four criteria. We also compared 2013 vs 2014 searches for the publications of countries with the most publications. Even in this case the results were very similar, and the average difference for the 20 countries with the most publications in the time period 2008–2012 was only 0.6%, with a maximum difference for Iran (6.1%) (sheet 4). These data confirm that PubMed is a reliable search engine that accurately retrieves information from databases that are promptly updated.\n\nUsing the same method described above, we estimated the percentage of publications in English for the 20 countries with the most publications in the time period 2008–2012. With the exceptions of France (88.3%) and China (78.8%), more than 95% of publications of all other countries were written in English (Figure S4) (sheet 3). The PubMed search engine seems to be fairly accurate in the classification of languages of articles; indeed, we saw that the proportions of publications in English of Anglophone countries like the United States, the United Kingdom and Australia were over 99%.\n\nWe also determined the proportion of review publications for the 20 countries with the most publications in the time period 2008–2012. They were calculated by comparing searches with or without the exclusion of reviews in the “Publication Types” field (See Methods section). On average, review publications constituted 9.9% of the publications, ranging from a maximum of 14.7% (Italy) to a minimum of 2.9% for South Korea (Figure S5) (sheet 4).\n\nWe measured the publications per capita of countries (sheet 5). We divided the country publications in the time period 2008–2012 by the country population and multiplied by 1,000 in order to obtain the numbers of publication per 1,000 people. Figure 5 shows the publications per 1,000 people for the 20 countries with the most publications. Switzerland (4.8), Sweden (4.4), the Netherlands (4.2), Israel (3.9), the United Kingdom (3.8), and Australia (3.6) were, in descending order, the countries with the highest publications per capita. Iran (0.4), Brazil (0.3), China (0.2), and India (less than 0.1) were, in descending order, the countries with lowest publications per capita. We also ranked all other countries based on their publications per capita. Figure 6 and Figure S6 show the number of publications per 1,000 people for the 20 and the 40 countries with the highest publications per capita, respectively. The top ranked country was Denmark (which is not part of the group of 20 countries with the most publications) with 4.8 publications per 1,000 people in the time period 2008–2012. Switzerland (4.8), Sweden (4.4), and the Netherlands (4.2) closely followed Denmark in this ranking.\n\nThe numbers represent publications per 1,000 people and were obtained by dividing the number of publications (reviews excluded) by the country populations and multiplying by 1,000. In this case, differently from Figure 3, the data were obtained from 2013 searches.\n\nCriteria A.\n\nThe numbers represent publications per 1,000 people and were obtained by dividing the number of publications (reviews excluded) by the country populations and multiplying by 1,000.\n\nCriteria A.\n\nWe analyzed the publication output of the 40 countries with the highest GDPs for the year 2011 and reported the data for the 20 and 40 countries with the highest GDPs, respectively in Figure 7 and Figure S7 (sheet 6). Almost every country in this group is also in the group of countries with the most publications in the time period 2008–2012; the exceptions are Russia, Mexico, Indonesia, and Saudi Arabia, which are not part of the group with the most publications, and Sweden, Israel, Iran, and Belgium, which are not in the group of 20 countries with the highest GDPs. We also calculated the publications per 1,000 people for the 20 countries with the highest GDPs: Switzerland (4.8), the Netherlands (4.2), and the United Kingdom (3.8), in descending order, had the highest ratios, whereas China (0.2), Russia (less than 0.1), India (less than 0.1), and Indonesia (less than 0.1), in descending order, had the lowest ratios in the time period 2008–2012 (Figure 8) (sheet 6).\n\nMoreover, we calculated the number of publications per GDP for the 20 and 40 countries with the highest GDPs (Figure 9 and Figure S8) (sheet 6). In the group of 40 countries with the highest GDPs, the country with the highest publications-to-GDP ratio in the time period 2008–2012 was Israel, followed by the United Kingdom, the Netherlands, and South Korea whereas Russia, the United Arab Emirates, Venezuela, and Indonesia were, in descending order, the countries with the lowest ratios. The ratio of the lowest ranking country (Indonesia) was less than 1% of the ratio of the highest ranking country (Israel).\n\nCriteria A.\n\nThe numbers represent publications per 1,000 people and were obtained by dividing the number of publications (reviews excluded) by the country population and multiplying by 1,000.\n\nCriteria A.\n\nThe numbers of publications (reviews excluded) were divided by the GDPs of 2011 (US$) from World Bank database and multiplied by 109.\n\nCriteria A.\n\nWe also calculated the publications per R&D expenditure for the 20 countries with the highest GDPs. The numbers on the chart represent the ratios between the numbers of publications and these R&D expenditures. The United Kingdom, Italy, Turkey and the Netherlands were, in descending order, the countries with the highest ratios, whereas Indonesia and Russia were, in descending order, the countries with the lowest ratios in the time period 2008–2012 (Figure 10) (sheet 9). The ratio of the lowest ranking country (Russia) is 5.1% of the ratio of the highest ranking country (the United Kingdom).\n\nThe numbers of publications (reviews excluded) were divided by the R&D expenditure of 2012 expressed as GERD (Gross Expenditures on Research and Development) in billion US$ at Purchasing Power Parity (PPP). The data of R&D expenditures of countries were taken from the Battelle nonprofit private company (http://www.battelle.org/docs/tpp/2014_global_rd_funding_forecast.pdf).\n\nCriteria A.\n\nWe measured the publication output based on clinical trial studies. First, similarly to what we did with the general publications (Figure S1), we calculated the proportions of clinical trials publications with a country name in the affiliation. The proportions were 96.5% and 97.8% for the time periods 2003–2007 and 2008–2012, respectively (Figure S9) (sheet 10). In the time period 2008–2012, there were over 180,000 publications based on clinical trial studies (sheet 10). We then determined the numbers of clinical trial publications for all the countries of the world (sheet 10) and ranked the 20 countries with the most clinical trial publications (Figure 11) (sheet 10) in the time period 2008–2012. The United States was the country with the most publications of clinical trial studies, with over 58,000 publications in the time period 2008–2012, over four times the volume of the United Kingdom (with over 14,000 clinical trial publications), the second in the ranking. In Figure 12 (sheet 11), we show the clinical trial publications of the 20 countries with the highest GDPs.\n\nCriteria E.\n\nCriteria E.\n\nWe compared the volumes of publications of time periods 2003–2007 and 2008–2012 for the 20 countries with the most publications. Figure 13 (sheet 12) shows the relative change (as the percentage of the volume of time period 2003–2007) of general publications (as usual, reviews were excluded). The four countries with the highest increases were, in descending order, Iran (220.4%), China (119.5%), India (115.2%), and South Korea (108.6%); these countries more than doubled the volume of publications from time period 2003–2007 to time period 2008–2012. The volumes of publications did not decrease in any of the 20 countries with the most publications. In this group the country with the lowest increase was Japan, with a 9.7% increase. We also determined the relative changes with regard to the clinical trial publications. Figure 14 (sheet 13) shows these relative changes. The country with the highest relative change was Iran, with a 179.6% increase from time period 2003–2007 to time period 2008–2012. South Korea (104.3%) and China (99.3%) were second and third in this ranking. The volume of clinical trial publications decreased only for Israel (-14.6%).\n\nThe relative changes are expressed as percentage change relative to the 5-year period 2003–2007.\n\nCriteria A.\n\nThe relative changes are expressed as percentage change relative to the 5-year period 2003–2007.\n\nCriteria E.\n\nIn order to determine the proportion of publications that are clinical trial studies, a proxy for the level of “attractiveness” towards clinical trial investigations, we created the Clinical Trial Attraction Score, defined as the ratio of clinical trial publications to the general publications multiplied by 10,000 (to make these scores comparable to the Topic Attraction Scores, see below). We calculated these scores for the 20 countries with the most publications (Figure 15) (sheet 14). The Netherlands, Italy, and Sweden were, in descending order, the countries with the highest Clinical Trial Attraction Scores. China and India, in descending order, were the ones with the lowest. The Attraction Score of the highest ranking country (the Netherlands) was 3.9 times the Attraction Score of the lowest ranking country (India).\n\nThe “Clinical Trial Attraction Scores” were calculated by dividing the numbers of Clinical Trials publications (reviews excluded) by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000.\n\nCriteria A and E.\n\nAs a proxy for the level of “attractiveness” towards specific research areas or topics, we determined Attraction Scores for research areas or topics. We used the same method used to determine the Clinical Trial Attraction Score. Basically, the number of publications related to a specific area or topic was divided by the total number of publications of the same country and then multiplied by 10,000 (to get easily readable scores).\n\nFigure 16 (sheet 15) reports the Topic Attraction Scores for “pharmacogenomics” in the 20 countries with the most publications in the time period 2008–2012. The Netherlands, Spain, Sweden, and the United States were, in descending order, the countries with the highest scores while Turkey and Iran were, in descending order, the countries with the lowest scores. The Attraction Score of the highest ranking country (the Netherlands) was 14.1 times the Attraction Score of the lowest ranking country (Iran).\n\nThe “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “pharmacogenomic” OR “pharmacogenomics” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000.\n\nCriteria B.\n\nFigure 17 (sheet 16) reports the Topic Attraction Scores for “personalized medicine” of the 20 countries with the most publications in the time period 2008–2012. United States, Israel, and Switzerland were, in descending order, the countries with the highest scores while Turkey, Iran and Brazil were, in descending order, the ones with the lowest scores. The Attraction Score of the highest ranking country (the United States) was 17.3 times the Attraction Score of the lowest ranking country (Brazil).\n\nThe “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “personalized medicine” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000.\n\nCriteria B.\n\nFigure 18 (sheet 17) reports the Topic Attractions Scores for “health” and “public health” for the 12 countries with the most publications in the time period 2008–2012. Australia, the United Kingdom, Canada, and the United States (remarkably, all Anglo-Saxon countries) had, in descending order, the highest Public Health Attraction Scores (with very similar results for the Health Attraction Score). The Attraction Score of the highest ranking country (Australia) was 5.6 times the Attraction Score of the lowest ranking country (Japan).\n\n(A) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “health” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000. (B) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “public health” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying the obtained quotient by 10,000.\n\nCriteria B.\n\nFigure 19 (sheet 18) reports the Topic Attraction Scores for “induced pluripotent stem cells” and “human induced pluripotent stem cells” of the 12 countries with the most publications in the time period 2008–2012. In both cases Japan was the country with by far the highest score. The Attraction Score for “iPS cells” of the highest ranking country (Japan) was 34.4 times the Attraction Score of the lowest ranking country (India). The Attraction Score for “hiPS cells” of the lowest ranking country (India) was 0.\n\n(A) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “induced pluripotent stem cells” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000. (B) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “human induced pluripotent stem cells” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-years period 2008–2012 and multiplying by 10,000.\n\nCriteria B.\n\nFigure 20 (sheet 19) reports the Topic Attraction Scores for “embryonic stem cells” and “human embryonic stem cells” in the 12 countries with the most publications in the time period 2008–2012. In both cases, South Korea was the country with the highest score. The Attraction Score for “ES cells” of the highest ranking country (South Korea) was 4.6 times the Attraction Score of the lowest ranking country (India). The Attraction Score for “hES cells” of the highest ranking country (South Korea) was 8.1 times the Attraction Score of the lowest ranking country (Italy).\n\n(A) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “embryonic stem cells” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000. (B) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “human embryonic stem cells” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000.\n\nCriteria B.\n\nWe compared publication output in the life sciences with the overall output of all disciplines. Data on the latter were obtained from the SCImago Journal & Country Rank, which includes journals and country scientific indicators from the information contained in the Scopus® database (http://www.scimagojr.com/). We ran the comparison for a single year (2012). Figure S10 (sheet 21) reports data for the 20 countries with the most publications in year 2012. The pattern, as expected, was very similar to the pattern for the time period 2008–2012 (Figure 3), and the group of 20 countries was the same with the exception of Israel, which was only in the 2008–2012 group and Denmark, which was only in the 2012 group. In addition, this time we also included reviews as these are included in the data from SCImago Journal & Country Rank. Figure S11 (sheet 21) reports the percentage of reviews (estimated as previously indicated) of the 20 countries with the most publications in 2012. With an average of 9.7%, a minimum of 3.0% (South Korea) and a maximum of 15.1% (Italy), these results are similar to those from 2008–2012 (Figure S5). We then included reviews and re-ranked the same 20 countries shown in Figure S10 (Figure S12) (sheet 21) and found that, given that the percentage of reviews was low, rankings and relative differences are not substantially affected by the inclusion of reviews. We therefore estimated the ratio of publications in the life sciences with publications in all fields and reported the results in Figure S13 (sheet 21). Data show an average of 36.7% publications in the life sciences with China (22.7%) and Iran (26.4%) scoring at the bottom and the United States (54.0%) and the United Kingdom (48.2%) scoring at the top.\n\nWe calculated the numbers of publications of countries where English is or is not the primary language. Figure 21 (sheet 22) shows the proportions of publications from countries where English is the primary language for the two time periods 2003–2007 and 2008–2012. The percentage of publications of these countries slightly decreased from 50.0% in time period 2003–2007 to 46.1% in time period 2008–2012. As a consequence, the percentages of publications of countries where English is not the primary language increased from 50.0% to 53.9%. This increase was expected given the large increases in publications volumes in non-English native speaking, high-volume publishing countries like China, India, South Korea, and Iran (Figure 13). For a list of countries considered to have English as primary language, see the data file.\n\nThe proportions are expressed as percentages in the 5-year periods 2003–2007 and 2008–2012. The proportions were determined by dividing the sum of the numbers of publications (reviews excluded) in all the countries where English is the primary language by the sums of the numbers of publications (reviews excluded) in all countries of the world. For countries that we considered to have English as primary language see the data file (sheet 21).\n\nCriteria B.\n\nData on publication output in each continent in the time period 2008–2012 appear in Figure 22 (Sheet 23). To calculate publication output for each continent, we divided the American continent into North-Central America and South America. North-Central America and Europe were the continents with the greatest number of publications, representing 35.5% and 33.7% of the total numbers of publications (sum of all the countries), respectively. Figure 23 shows the relative changes in publications from time period 2003–2007 to time period 2008–2012. Africa was the continent with the biggest relative change in publications, with a 78.5% increase in publication volume from time period 2003–2007 to time period 2008–2012, whereas Europe (30.4%) and North-Central America (24.2%) were the continents with the smallest increases. We also calculated the publications per capita in the time period 2008–2012. Figure 24 (sheet 23) reports the publications per 1,000 people in the continents. Oceania was the continent with the highest publications per capita with 2.7 publications per 1,000 people, followed by North-Central America with 2.1 and Europe with 1.5. Africa was the continent with the smallest publications per capita, with less than 0.1 publications per 1,000 people.\n\nThe proportions are expressed as percentages. America was divided into North-Central America and South America. The numbers of publications (reviews excluded) for Asia and Europe were approximated by equally dividing the publications (reviews excluded) of Russia and Turkey between the two continents.\n\nCriteria B.\n\nThe relative changes are expressed as percentages relative to the number of publications (reviews excluded) of the 5-year period 2003–2007.\n\nCriteria B.\n\nThe numbers represent the numbers of publications (reviews excluded) per 1,000 people by continent in the 5-year period 2008–2012.\n\nCriteria B.\n\nWe calculated the number of publications of the 28 countries that are currently part of the European Union (EU Countries) in time periods 2003–2007 and 2008–2012. The publications during the time period 2008–2012 were more than one million (Figure S14A) (sheet 24) and represented 32.1% and 30.6% of the world’s total publications in the time periods 2003–2007 and 2008–2012, respectively (Figure S14B). The relative increase from time period 2003–2007 to time period 2008–2012 of the EU Countries was 29.6%, therefore smaller than the relative increase for the whole world (36.2%) (Figure S14C). We also calculated the total clinical trials study publications of the EU Countries in time periods 2003–2007 and 2008–2012 (Figure S15A). In time period 2008–2012, there were over 65,000 clinical trial publications. These publications represent 39.2% and 36.3% of the world’s total clinical trials publications in time periods 2003–2007 and 2008–2012, respectively (Figure S15B). The relative change in clinical trial publications from time period 2003–2007 to time period 2008–2012 was a 5.7% increase for the 28 countries, compared to a 14.2% increase for the whole world (Figure S15C).\n\nPubMed provides a function to automatically download CSV files with the numbers of publications year by year. We compared the numbers of publications obtained by using this function with the numbers of publications obtained by using the custom range function to define time periods that we described above. An analysis of the 2012 reveals that the CSV download yielded fewer publications than the custom range approach. Figure S16 (sheet 25) shows the differences for the 20 countries with the most publications in 2012. On average the difference was -18.4%, with a maximum difference of -21.7% for Turkey and a minimum difference of -13.8% for Iran.\n\n\nDiscussion\n\nAssessing the quantity and quality of a country’s scientific output is key to understanding and improving its research system. In this paper, we present a methodology that focuses on publication in biology and related disciplines that contribute to advancing that field. Based on the search engine PubMed, the method was used to count publications at the country level and at the supranational level (continents and European Union), in different time periods, to trace changes in publication outcome over time and to measure publication ratios relative to country populations, country GDP and R&D expenditure, the share of publications based on clinical trials, whether the publication was a review article, and whether the publication was written in English. We demonstrated how various stringency criteria could be deployed to check data robustness. Box 1 presents an excerpt of the main findings.\n\n\n\nThe total number of publications during years 2008–2012 (almost four million, excluding review publications) is nearly the double the number of publications during years 1993–1997\n\nThe proportion of publications based on clinical trials has remained constant around 5% since 1993\n\nThe proportion of review publications for the 20 countries with the most publications is 9.9%, with a maximum for Italy (14.7%) and a minimum for South Korea (2.9%) (years 2008–2012)\n\nThe United States, with over one million publications in years 2008–2012 (reviews excluded) is by far the country with the most publications, having almost four times the publications of the second ranking country, China\n\nThe country with the most publications per capita is Denmark (4.81 per 1,000 people), followed by Switzerland (4.76 per 1,000 people) and Sweden (4.40 per 1,000 people) (years 2008–2012)\n\nThe country with the highest publications-to-GDP ratio, among the 40 countries with the most publications, is Israel, followed by the United Kingdom, the Netherlands, and South Korea (years 2008–2012)\n\nThe country with the highest publications-to-R&D expenditure (GERD PPP), among the 20 countries with the highest GDPs is the United Kingdom, followed by Italy, Turkey, and the Netherlands\n\nAmong the 20 countries with the most publications, Iran is the country with the highest relative increase (+220%) in the number of publications from years 2003–2007 to years 2008–2012. Japan is the country with the lowest relative increase (+9.7%)\n\nAmong the 20 countries with the most publications, Iran is the country with the highest relative increase (+179.6%) in the number of clinical trial publications from years 2003–2007 to years 2008–2012. Israel is the only country with a decrease (-14.6%)\n\nThe United States is the country with the most publications based on clinical trials (almost 60,000), nearly four times the number of publications based on clinical trials in the second ranking country, the United Kingdom (years 2008–2012)\n\nAmong the 20 countries with the most publications, the Netherlands is the country with the highest “Clinical Trial Attraction Score”, followed by Italy (years 2008–2012)\n\nAmong the 20 countries with the most publications, the Netherlands is the country with the highest “Pharmacogenomics Attraction Score” (years 2008–2012)\n\nAmong the 20 countries with the most publications, the United States is the country with the highest “Personalized Medicine Attraction Score” (years 2008–2012)\n\nAmong the 20 countries with the most publications, Japan is the country with the highest “iPS cells Attraction Score” (years 2008–2012)\n\nAmong the 20 countries with the most publications, South Korea is the country with the highest “hES cells Attraction Score”, (years 2008–2012)\n\nThere is some correlation between policies regulating the use of human embryonic stem cells and the “hES cells Attraction Scores”\n\nAmong the 12 countries with the most publications, Australia, the United Kingdom, and Canada, in descending order, have the highest “Public Health Attraction Scores” (years 2008–2012)\n\nThe estimated proportion of publications written in English in countries where English is the primary language has decreased from 50.0% (years 2003–2007) to 46.1% (years 2008–2012)\n\nOceania is the continent with the most publications per capita (2.7 per 1,000 people) followed by North-Central America (2.1 per 1,000 people) (years 2008–2012)\n\nNorth-Central America has the highest proportion of world publications (35.5%); Africa has the lowest (1.4%) (years 2008–2012)\n\nThe proportion of world publications of the present day 28 countries of the European Union is 30.6% (years 2008–2012)\n\nWe have also created the Attraction Scores. These assess the focus of publication output on certain types of studies (e.g., clinical trials) or areas or topics of interest (e.g., hESC). Attraction Scores express a proportion of the total publication output rather than the number of papers. For this reason, we believe that they better represent the real attraction of certain types of studies or certain topics of research to scientists.\n\nAttraction Scores can be correlated to many different factors. It is no surprise that this attraction for a cutting-edge topic like “pharmacogenomics” is higher in more technologically developed countries (Figures 16). Higher Attraction Scores may also be determined by past path-breaking discoveries that attract investments and the attention of researchers. This seems to be the case of iPS cells and Japan, a country in which these cells were first obtained and then highly researched thanks to massive research and technological investment12–14. The predictable result is that Japan has indeed the highest Attraction Score for iPS cells of all surveyed countries. Variations in Attraction Scores can also be correlated with other factors, including the prevalence of certain diseases, the structure of research workforces (e.g. small groups vs. big collaborative efforts), regulatory policies, and cultural and political factors. Attraction Scores can be used not only to assess the effect of putative determining factors but also to predict future trends. For example, the focus toward certain areas (or topics) or types of research can have different consequences such as the development of new avenues of research or new types of interaction between academia and industry.\n\nThough a detailed and comprehensive analysis of correlations and predictive uses of the Attraction Scores is beyond the purposes of this study, we want to provide an example of how the Attraction Scores can be used to assess the effect of specific policies. To this end, the policies regulating the use of hESCs in research present an interesting cases study. Over the past 15 years, countries have adopted a wide range of policies that can be divided into four categories—permissive, permissive compromise, restrictive compromise, and prohibitive15. In restrictive policy countries, human embryo research, derivation of hESCs from supernumerary embryos, and somatic cell nuclear transfer (SCNT) are usually prohibited. When permitted, research is possible with imported hESC lines or cell lines that were created before the enactment of any embryo research ban. Predictably, the two countries with the most restrictive policies (Italy and Germany) report very low Attraction Scores (Figure S17) (sheet 20). Attraction Scores are progressively higher as countries’ policies become more permissive. The highest Attraction Scores belong to South Korea and Australia. This conclusion is reinforced by the analysis of a second score (hESC/ESC score), which assesses the relative focus on human embryonic stem cells. This score is the ratio between the number of publications on “human embryonic stem cells” and the number of publications on “embryonic stem cells” (searches for these key words were, as usual, in the “Title/Abstract” field). The highest hESC/ESC scores belong to Spain and Australia, in which SNCT and derivation of hESC from supernumerary embryos are permissible, with the score of the highest ranking country (Australia) being 4.7 times higher than the score of the lowest ranking country (Italy). Germany reported the third lowest hESC/ESC score.\n\nThe ranking represents three cases of apparent outliers: Canada, Japan, and the United States. Canada and the United States report Attraction Scores that are apparently higher than their policies would intuitively suggest. However, a closer look at the policies of these countries explains their position in the ranking. Canada explicitly permitted research with human pluripotent stem cell research since 2002 and since 2006 legalized research on supernumerary embryos. Federal funding for hESC research has been available since the early 2000s and has been comparatively generous since then as evidenced by the fact that, in 2009, the MaRS Regenerative Medicine 2009 Industry Briefing report ranked Canada 4th (after the United States, the United Kingdom, and South Korea) for government funding for stem cell research (www.marsdd.com/mars-library/regenerative-medicine-industry-briefing/)16. In 2013, “at least 68 centers” with more than 350 researchers were “investigating stem cells and regenerative medicine” (http://www.ic.gc.ca/eic/site/lsg-pdsv.nsf/eng/hn01746.html Table 1, 4 (last updated June 18, 2013))17. The United States, another apparent outlier, is classified as a case of “permissive compromise.” Yet, the reality has been different. Policies were restrictive only at the federal level with the Dickey-Wicker Amendment and President Bush’s a ban on federal funding for research on certain stem cells. However, many US states funded and promoted a significant amount of stem cell research18, which resulted in a high Attraction Score for both 2008–2012 (Figure S17) and 2003–2007 (sheet 19). On the other hand of the spectrum, Japan adopted more liberal policies but reports a comparatively low Attraction Score. This is due to the way in which the regulatory requirement for prior approval before using hESC translated into an “excessively burdensome approvals process”, which is often blamed for Japan lagging behind other countries19. In 2010, Kawakami et al.19 worried that “[a]lthough direct, quantitative effects are difficult to demonstrate, it seems reasonable that these regulatory delays have presented serious challenges to Japanese researchers working, or seeking to work, in these fields, and ultimately impeded progress and competitiveness”. In addition, Japan has focused its research efforts in regenerative medicine on iPSs rather than hESCs as discussed above. Based on these results, we submit that Attraction Scores reveal a correlation between hESC policies and research output in this area.\n\nAs the other metric methodologies1–6,9–11,20, the one described in this report presents “pros” and “cons” as well as strengths and weaknesses. Indeed, there are some caveats to take in consideration when using this methodology or interpreting the data.\n\nFirst, as also shown in Figure S1, the country name is not always present in the affiliation. This would not be a problem if all the countries had the same proportions of publications without the country name in the affiliation. However, it is possible that different countries have different proportions of affiliations without the country name. In any case, we think that this is likely not a significant problem given that the vast majority of the publications have the country name in the affiliation and that it is unlikely that research groups around the world have very different habits when indicating the affiliation details on papers. In any case, small differences (if any) should be fully compensated when calculating the ratios between counts of the same country, as in the case of measuring ratios relative to changes in publication output over time or Attraction Scores.\n\nSecond, this method may not be perfectly accurate with regard to countries that are recorded under multiple names including acronyms. If a country name is missing then the count for that country would be underestimated. However, in our study we tried to include all possible country names. For example, the United States was searched including not only searches for “united states” but also for “us” OR “usa” (notably when we searched only for “united states” OR “usa” the volumes of publications were decreased by about 10%). Moreover, in some cases the name/acronym of a country could be the same as other affiliation specifications. This particular case may lead to overestimating the number of publications for that country. For example, the acronym US could be present in affiliations not related to the United States, and therefore, even if we believe that (based also on pilot tests we have performed; data not shown) this effect is conceivably negligible, there might be the possibility that the searches for the United States are slightly overestimated. Also, there are cases in which a name can be both a country and a region of another country. This is the case of Georgia, whose number of publications are likely overestimated because “georgia” may be listed as a sovereign country or a state that is part of the United States.\n\nThird, this method relies on first authors’ affiliation information to attribute publications to a certain country. Given that the first author ordinarily either matches the corresponding author or has the same affiliation country of the corresponding author and that the first author (even when she/he is not the corresponding author) frequently plays a prominent role in the project design or execution, this approach has the advantage of classifying papers based on the effective location of the main ideation and execution of research projects. Thus, this method is useful to determine publication outputs based on leading contributions; however, it is less accurate when all types of contributions need to be taken in account. It should also be noted that at the beginning of 2014, PubMed started inserting affiliation information for every author of published articles, however without the possibility of limiting the search to first authors (http://www.ncbi.nlm.nih.gov/books/NBK3827/#pubmedhelp.Affiliation_AD). Therefore, the method used in this paper cannot be used for papers published in 2014 or after. While PubMed’s decision to include information for all authors of a paper is welcomed, it would be desirable to be able again to select affiliation information only for the first (and possibly also specifically for the last) author.\n\nFourth, PubMed’s records may be incomplete and thus not perfectly accurate. This is the case of information such as the language in which the paper is written and whether the paper is a review.\n\nFifth, even if we excluded reviews from our counts (except where otherwise indicated), we did not exclude all publications that are not based on original research. This is the case of comments and editorials. Our choice was motivated by reasons of simplicity and because we believed that PubMed’s tagging of publications as “letters” or “editorials” could be not fully accurate and possibly unequal among countries. However, we believe this choice does not have an impact on data because it is plausible that these types of publications do not significantly affect the relative differences between countries as suggested also by the fact that, when we chose to count only the publications published as “article journal” (criteria C), we did not observe substantial differences with the standard criteria (criteria A) (sheet 3).\n\nSixth, the quality (however one defines “quality” in this context) of publications is not taken in account. This methodology quantifies the research output by determining the numbers of publications or ratios between numbers of publications and other variables. Proxies for the quality of the papers, such as numbers of citations, numbers of downloads, and impact factors of the journals, are not taken in consideration. Even if indexes based on the quality of research have been already proposed1–6,10,11,20, we argue that any means of measuring the quality of science will always be partial and controversial and for this reason it will always be useful to take in consideration also (or, in specific circumstances, only) the total volumes of publications.\n\nIf the limitations are taken into account, we believe that the methodology and information presented in this paper can be used, in conjunction with other metrics, to assess research systems in terms of publication output. In particular, we think that the volumes of publications, the relative changes in time, and the Attraction Scores described here provide valuable and unique information about the biomedical and biological research systems of countries. This will assist scholars in studying research systems and policymakers in designing policies to improve scientific production and its benefits to society.\n\n\nData availability\n\nfigshare: Biomedical publication and Attraction Score data based on PubMed searches, http://dx.doi.org/10.6084/m9.figshare.124689821", "appendix": "Author contributions\n\n\n\nAndrea Ballabeni designed the research. Qinyi Xu and Andrea Ballabeni collected and analyzed the PubMed and the other country-based bibliometric or non-bibliometric data. Andrea Boggio collected and analyzed the data on the regulatory policies. Andrea Ballabeni and Andrea Boggio wrote the manuscript. All authors critically reviewed the entire manuscript. All authors approved the final manuscript for publication.\n\n\nCompeting interests\n\n\n\nAndrea Boggio and Andrea Ballabeni are members of the board and of the council, respectively, of the Luca Coscioni Association; they do not receive any income from this organization.\n\n\nGrant information\n\nThis work was initiated at Bentley University and was funded, in part, by a grant from the Luca Coscioni Association for Freedom of Scientific Research to the Jeanne and Dan Valente Center for Arts and Sciences at Bentley University. The grant from the Luca Coscioni Association was for Qinyi Xu.\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 David Hemenway (Harvard School of Public Health) and Michele Rhee (National Brain Tumor Society) for helpful comments during the manuscript preparation.\n\n\nSupplementary figures\n\nCriteria A and B were used to calculate the proportions in the 5-year periods 1993–1997, 1998–2002, 2003–2007, and 2008–2012. The total world publications (reviews excluded) were obtained by searches without any specification in the affiliation field. The total world publications (reviews excluded) with the country name in the affiliation field were obtained by summing the numbers of publications (reviews excluded) of all countries of the world.\n\nThe percentages were calculated by determining the proportions between the numbers of publications obtained with criteria D and the numbers of publications obtained with criteria A.\n\nIn this case, differently from Figure 3, the data were obtained from 2013 searches.\n\nCriteria A.\n\nThe percentages were calculated by determining the proportions between the numbers of publications obtained with criteria D and the numbers of publications obtained with criteria A.\n\nThe percentages were calculated by determining the proportion between the numbers of publications obtained with Criteria A and the numbers of publications obtained with Criteria A but without the exclusion of the reviews.\n\nThe numbers represent publications per 1,000 people and were obtained by dividing the number of publications (reviews excluded) by the country populations and by multiplying by 1,000.\n\nCriteria A.\n\nCriteria A.\n\nThe numbers of publications (reviews excluded) were divided by the GDP of 2011 (US$) from World Bank database and multiplied by 109.\n\nCriteria A.\n\nCriteria E were used to calculate the proportions in the 5-year periods 2003–2007 and 2008–2012. The total world Clinical Trials publications (reviews excluded) were obtained by searches without any specification in the affiliation field. The total world Clinical Trials publications (reviews excluded) with the country name in the affiliation field were obtained by summing the numbers of Clinical Trials publications (reviews excluded) in all countries of the world.\n\nCriteria A.\n\nThe percentages were calculated by determining the proportions between the numbers of publications obtained with Criteria A and the numbers of publications obtained with Criteria A but without the exclusion of the reviews.\n\nCriteria A (without reviews exclusion).\n\nThe proportions were calculated by dividing the numbers of publications (reviews included) and all the citable documents of the SJR database (SCImago Journal & Country Rank) of the 20 countries with the most publications in the year 2012.\n\n(A) Numbers of publications (reviews excluded) of the present day 28 countries of the European Union in the 5-year periods 2003–2007 and 2008–2012. (B) Share of total world publications (reviews excluded) that came from the present day 28 countries of the European Union in the 5-year periods 2003–2007 and 2008–2012. (C) Relative changes in numbers of publications (reviews excluded) of the present day 28 countries of the European Union and of the World from the 5-year period 2003-2007 to the 5-year period 2008–2012. The changes are expressed as percentage changes relative to the 5-year period 2003–2007.\n\nCriteria A.\n\n(A) Numbers of Clinical Trials publications (reviews excluded) of the present day 28 countries of the European Union in the 5-year periods 2003–2007 and 2008–2012. (B) Share of total world Clinical Trials publications (reviews excluded) that came from the present day 28 countries of the European Union in the 5-year periods 2003–2007 and 2008–2012. (C) Relative changes in numbers of Clinical Trial publications (reviews excluded) of the present day 28 countries of the European Union and of the World from the 5-year period 2003–2007 to the 5-year period 2008–2012. The changes are expressed as percentage changes relative to the 5-year period 2003–2007.\n\nCriteria A and E.\n\nNumbers of publications (reviews excluded) of the 20 countries with the most publications (reviews excluded) in the year 2012 are shown, either obtained by automatic CSV downloads (pale green) or by custom range searches (pale green + pale blue). The pale blues portions of the bars represent the differences between the two methods.\n\nCriteria A.\n\n(A) The “Topic Attraction Scores” were calculated by dividing the numbers of publications (reviews excluded) with “human embryonic stem cells” in the title/abstract field by the total numbers of publications (reviews excluded) in the 5-year period 2008–2012 and multiplying by 10,000. The countries are ranked according to the scores. (B) The “hESC/ESC Scores” were calculated by dividing the numbers of publications (reviews excluded) with “human embryonic stem cells” in the title/abstract field by the numbers of publications (reviews excluded) with “embryonic stem cells” in the title/abstract field in the 5-year period 2008–2012 and multiplying by 10,000. The countries are ranked according to the scores. Criteria B.\n\nCriteria A.\n\nThe percentages were calculated by determining the proportions between the numbers of publications obtained with Criteria A and the numbers of publications obtained with Criteria A but without the exclusion of the reviews.\n\nCriteria A (without reviews exclusion).\n\nThe percentages were calculated by determining the proportions between the numbers of publications obtained with Criteria A and the numbers of publications obtained with Criteria A but without the exclusion of the reviews.\n\nCriteria A. Note that starting from 2014 PubMed reports affiliation information not only of the first author (see Methods and Discussion sections for a correct interpretation of 2014 counts).\n\n\nReferences\n\nLópez-Illescas C, de Moya Anegón F, Moed HF: Comparing bibliometric country-by-country rankings derived from the Web of Science and Scopus: the effect of poorly cited journals in oncology. J Inf Sci. 2009; 35: 244–256. Publisher Full Text\n\nNejati A, Jenab SMH: A two-dimensional approach to evaluate the scientific production of countries (case study: the basic sciences). Scientometrics. 2010; 84: 357–364. Reference Source\n\nAnonymous (SCImago Journal & Country Rank). Reference Source\n\nAnonymous (THE RESEARCH & INNOVATION PERFORMANCE OF THE G20). Reference Source\n\nAnonymous (Knowledge, networks and nations Global scientific collaboration in the 21st century). Reference Source\n\nJazayeri SB, Alavi A, Rahimi-Movaghar V: Situation of medical sciences in 50 top countries from 1996 to 2010--based on quality and quantity of publications. Acta Med Iran. 2012; 50(4): 273–278. PubMed Abstract\n\nRahman M, Fukui T: Biomedical research productivity: factors across the countries. Int J Technol Assess Health Care. 2003; 19(1): 249–252. PubMed Abstract \| Publisher Full Text\n\nRahman M, Fukui T: Biomedical publication--global profile and trend. Public Health. 2003; 117(4): 274–280. PubMed Abstract \| Publisher Full Text\n\nLoria A, Arroyo P: Language and country preponderance trends in MEDLINE and its causes. J Med Libr Assoc. 2005; 93(3): 381–385. PubMed Abstract \| Free Full Text\n\nFalagas ME, Pitsouni EI, Malietzis GA, et al.: Comparison of PubMed, Scopus, Web of Science, and Google Scholar: strengths and weaknesses. FASEB J. 2008; 22(2): 338–342. PubMed Abstract \| Publisher Full Text\n\nHightower C, Caldwell C: Shifting Sands: Science Researchers on Google Scholar, Web of Science, and PubMed, with Implications for Library Collections Budgets. Issu Sci Technolog Librarianship. 2010. Publisher Full Text\n\nTakahashi K, Yamanaka S: Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell. 2006; 126(4): 663–676. PubMed Abstract \| Publisher Full Text\n\nCyranoski D: Japan ramps up patent effort to keep iPS lead. Nature. 2008; 453(7198): 962–963. PubMed Abstract \| Publisher Full Text\n\nSong P, Inagaki Y, Sugawara Y, et al.: Perspectives on human clinical trials of therapies using iPS cells in Japan: reaching the forefront of stem-cell therapies. Biosci Trends. 2013; 7(3): 157–158. PubMed Abstract \| Publisher Full Text\n\nMoon S, Cho SB: Differential impact of science policy on subfields of human embryonic stem cell research. PLoS One. 2014; 9(4): e86395. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnonymous (Ontario’s MaRS Regenerative Medicine 2009 Industry Briefing report).\n\nAnonymous Government of Canada, Canadian Asset Map for Stem Cell and Regenerative Medicine: Overview of Stem Cell and Regenerative Medicine Research in Canada. Reference Source\n\nJohnson JA, Williams ED: Stem cell research: State initiatives. 2006. Reference Source\n\nKawakami M, Sipp D, Kato K: Regulatory impacts on stem cell research in Japan. Cell Stem Cell. 2010; 6(5): 415–418. PubMed Abstract \| Publisher Full Text\n\nAnonymous. Introducing the index. Nature. 2014; 515: S52–S53. Publisher Full Text\n\nXu Q, Boggio A, Ballabeni A: Biomedical publication and Attraction Score data based on PubMed searches. figshare. 2014. Data Source" }	[ { "id": "7052", "date": "16 Dec 2014", "name": "Vafa Rahimi Movaghar", "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 read the paper of Qinyi Xu et al. entitled \"Countries’ Biomedical Publications and Attraction Scores\" with great interest. Authors evaluated biomedical publications of all countries based on the PubMed and performed great job on different criteria to assess changes during different time periods based on the country of affiliation of the first authors especially during the five-year time period of 2008-2012 versus 2003-2007. The study is an addition to the literature and fruitful. I would recommend reading the paper to all medical scientists in the world. However, there are some concerns. PubMed is not the only search engine to be evaluated. There are other valuable sources such as ISI, Embase and Scopus which may change the results of the study. We would know that there are many journals indexed in the above-mentioned search engines especially written in the developing countries. On the other hand, quantity evaluation does not show the real scientific situation of the countries. I myself prefer the citation without self-citation for the best index of quality assessment. Other indices such as H-index reveal the mixture of quantity and quality assessment. The affiliation of the first author is acceptable, but not ideal to show the comprehensive situation of the countries. All authors’ affiliation or at least first and corresponding authors affiliations are better replacement in the future studies. Authors mentioned that they reported original articles. Therefore, they deleted reviews. However, I would recommend not deleting the review articles unless simultaneously deleting letters and correspondences.", "responses": [ { "c_id": "1134", "date": "22 Dec 2014", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would like to thank Professor Vafa Rahimi Movaghar for reviewing our article. We are glad that the reviewer appreciated our study. Here are some preliminary answers to his few concerns.The reviewer is right about the fact that PubMed is not the only source of biomedical publications and that some journals are not indexed in PubMed. We will certainly add remarks on this point in the revised version of the manuscript, which we will submit after receiving feedback from all reviewers.The reviewer mentions about the fact that our method measures only the quantity of publications. He is right but we have already addressed this issue in the Discussion section of the paper. In particular, we do not claim that our data can fully measure the efficiency of countries in biomedical research. Our data and methods need to be complemented with other data and methods. Our revised draft will certainly emphasize this point further.The reviewer mentions the limitations of using first authors’ affiliations. His comment has merit and in fact we have already acknowledged the pros and cons of this approach in the Discussion section.The reviewer mentions that it would be better to include reviews publications in the publication count. While we decided not to include reviews in the majority of calculations, we included them in some other calculations. We have observed that reviews amount to a rather small proportion of publications and consequently their inclusion does not substantially affect the rankings (this was done for years 2012 and 2013). We discuss this issue in the Results and Discussion sections. We would like to thank again Professor Vafa Rahimi Movaghar for taking the time to read the article and providing his feedback SincerelyThe authors" }, { "c_id": "1487", "date": "11 Aug 2015", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would like to confirm our first response to Professor Vafa Rahimi Movaghar. We have waited for a second review before revising the paper based on the reviewers’ suggestion. We now provide a revised version based on the feedback received from both referees.We would like to thank again professor Vafa Rahimi Movaghar for taking the time to read and comment our paper and for this helpful feedback." }, { "c_id": "1520", "date": "12 Aug 2015", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would also like to bring to the attention of our reviewers (Dr Vafa Rahimi Movaghar and Dr Youngim Jung) that we have updated figure S22 and added new figures S23-S25. In these figures we provide an analysis of year 2014. In particular, we observed that the patterns of publications were similar in the years 2008–2012, 2013 (the last year with the old settings regarding the affiliation) and 2014 (the first year with the new settings regarding the affiliation). We have accordingly slightly updated the paragraph added to the Discussion section: “In any case, this recent PubMed change offers the possibility to assess the countries’ biomedical publications by taking in consideration all contributing authors, including authors with minor contributions (that are usually placed in the middle of the list of authors). Even if assessing the countries’ biomedical publications based on the first author provides slightly different information from an assessment based on all authors, we expect that the patterns presented in this paper will not substantially change even if the new settings of PubMed were to be used. This is also suggested by the very similar patterns of publications in the years 2008–2012, 2013 (the last year with the old settings) and 2014 (the first year with the new settings) (Figure S18–Figure S25). At any rate, repeating this study by taking in consideration all authors could be informative to determine which countries have a propensity for leading (i.e. most of ideation and execution) versus assisting (i.e. least of ideation and execution) roles in biomedical research. However, this can only be done in the future, once that enough literature will be archived under the new settings (as of now, still not every currently published paper can be retrieved with the affiliation of all authors).”" } ] }, { "id": "9097", "date": "02 Jul 2015", "name": "Youngim Jung", "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 analyzed the volume of publications in the field of life sciences by country and continent in order to better understand regional differences in life sciences research infrastructure. The authors additionally allocated scores to quantify focal areas of interest and examined the potential use of these scores by exploring the relationship between Attraction Score and research policy on hESCs. This report is interesting, providing a detailed analysis with a graphical representation of data that facilitates a smooth understanding of the study's conclusions. However, as the first reviewer has mentioned, a limitation seems to be present in terms of the data source used for the study. PubMed is a comprehensive and authoritative database but is generally unable to provide an exhaustive data package for scientometric studies. Although the authors have claimed that the study covers the most recent, diverse and refined research output (via PubMed), the number of publications is likely to be only illustrative of the quantitative aspect of the research output. Similarly, the affiliation of the first author provides partial information on the contribution of each country to the publication. Instead, WoS or SCOPUS may be more useful data sources, allowing analysis of the citations and journal impact factors for all authors, as shown in previous studies. In addition, whether the Attraction Scores designated by the authors distinguish between the proportion of studies focused on the topic of interest and the entire number of publications remains open to speculation. Finally, the 'correlation analysis' referred to in this study is not an actual correlation analysis (which is a statistical term). It is recommendable to change the term in order to avoid confusion for the reader.", "responses": [ { "c_id": "1488", "date": "11 Aug 2015", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would like to thank Professor Youngim Jung for taking the time to read the paper and provide their helpful feedback. Here we provide our responses to their comments: We agree with the referee (as well as with the other referee) that, although PubMed is a search engine based on comprehensive and authoritative databases, some published papers might not be retrieved by searching in these databases. Having said this, we think that given the comprehensiveness and breadth of the PubMed system, the results presented in this paper are a bona fide representation of the literature. In any case, we agree that it will be important to crosscheck these data with data obtained by using other databases such as Scopus (Elsevier) and Web of Science (Thomson Reuter). To address this point, we have added the following paragraph to the Discussion section:\"Seventh, though PubMed is a search engine that is based on authoritative and comprehensive databases, it might not retrieve some publications. For this reason, it will be important to confirm these results by analyzing also other authoritative and comprehensive databases like Scopus (Elsevier) and Web of Science (Thomson Reuters) that might contain a few publications not retrieved by PubMed. In any case, given the comprehensiveness and breadth of PubMed databases, we believe that the results presented in this paper are a bona fide representation of the whole literature. Moreover, the fact that the paper is based on a database that is free and easily accessible by everyone in the world, certainly adds value to the results as it provides tools that can be freely used by anyone and it facilitates readers access to the underlying data, reproducibility, and comparison with our approaches.\" Moreover, in order to stress the point that our method and results are based on PubMed, we have slightly changed the title of the paper. We agree with the referee that the analysis based on the first author has, together with pros, some cons; for this reason we think that it will be useful to complement this analysis with analyses that include all authors. We have therefore added the following paragraph to the Discussion section:\"In any case, this recent PubMed change offers the possibility to assess the countries’ biomedical publications by taking in consideration all contributing authors, including authors with minor contributions (that are usually placed in the middle of the list of authors). Even if assessing the countries’ biomedical publications based on the first author provides slightly different information from an assessment based on all authors, we expect that the patterns presented in this paper will not substantially change even if the new settings of PubMed were to be used. At any rate, repeating this study by taking in consideration all authors could be informative to determine which countries have a propensity for leading (i.e. most of ideation and execution) versus assisting (i.e. least of ideation and execution) roles in biomedical research. However, this can only be done in the future, once that enough literature will be archived under the new settings (as of now, still not every currently published paper can be retrieved with the affiliation of all authors).\" We agree with the referee that analyses of citations and impact factors as well as of other putative quality indicators are not present in this paper. However, this was beyond the purposes of our research and we have already acknowledged this in the first version of the paper with the following paragraph of the Discussion section:\"Sixth, the quality (however one defines “quality” in this context) of publications is not taken in account. This methodology quantifies the research output by determining the numbers of publications or ratios between numbers of publications and other variables. Proxies for the quality of the papers, such as numbers of citations, numbers of downloads, and impact factors of the journals, are not taken in consideration. Even if indexes based on the quality of research have been already proposed 1– 6, 10, 11, 20 , we argue that any means of measuring the quality of science will always be partial and controversial and for this reason it will always be useful to take in consideration also (or, in specific circumstances, only) the total volumes of publications.\"Moreover, we have now added the following paragraph:\"Regardless, the methods and results presented in this paper are not to be intended as neither exclusive nor the perfect means of assessing countries’ research output. In fact, they are better seen as complementary to other methods and results.\"Regarding the comment of the referee on the “Attraction Scores”, we do not claim they are a perfect means of measuring the relative focus towards certain topics or areas of research. To avoid any ambiguity on this point we have made a few edits to the text. In particular, the term “assess” has been replaced with “appraise” or “study”. We agree with the referee that the term “correlation analysis’ can create confusion. For this reason, as suggested by the referee, we have changed the term throughout the manuscript. In same cases we have substituted it with the term “connection” while in other cases we have replaced it with the term “relationship”. We would like to thank again Professor Youngim Jung for taking the time to read our paper and provide her helpful feedback. We hope we have addressed all her reservations." }, { "c_id": "1519", "date": "12 Aug 2015", "name": "Andrea Ballabeni", "role": "Author Response", "response": "We would also like to bring to the attention of our reviewers (Dr Vafa Rahimi Movaghar and Dr Youngim Jung) that we have updated figure S22 and added new figures S23-S25. In these figures we provide an analysis of year 2014. In particular, we observed that the patterns of publications were similar in the years 2008–2012, 2013 (the last year with the old settings regarding the affiliation) and 2014 (the first year with the new settings regarding the affiliation). We have accordingly slightly updated the paragraph added to the Discussion section: “In any case, this recent PubMed change offers the possibility to assess the countries’ biomedical publications by taking in consideration all contributing authors, including authors with minor contributions (that are usually placed in the middle of the list of authors). Even if assessing the countries’ biomedical publications based on the first author provides slightly different information from an assessment based on all authors, we expect that the patterns presented in this paper will not substantially change even if the new settings of PubMed were to be used. This is also suggested by the very similar patterns of publications in the years 2008–2012, 2013 (the last year with the old settings) and 2014 (the first year with the new settings) (Figure S18–Figure S25). At any rate, repeating this study by taking in consideration all authors could be informative to determine which countries have a propensity for leading (i.e. most of ideation and execution) versus assisting (i.e. least of ideation and execution) roles in biomedical research. However, this can only be done in the future, once that enough literature will be archived under the new settings (as of now, still not every currently published paper can be retrieved with the affiliation of all authors).”" } ] } ]	1	https://f1000research.com/articles/3-292
https://f1000research.com/articles/4-148/v1	10 Jun 15	{ "type": "Antibody Validation Article", "title": "Immunoblotting validation of research antibodies generated against HS1-associated protein X-1 in the human neutrophil model cell line PLB-985.", "authors": [ "Kristina Inman", "Peter Cavnar", "Kristina Inman" ], "abstract": "HS1-associated protein X-1 (Hax1) is a 32 kDa protein that is ubiquitously expressed. Hax1 is an anti-apoptotic protein with additional roles in cell motility, and autosomal recessive loss of Hax1 results in Kostmann syndrome, a form of severe congenital neutropenia. Because of the important role of Hax1 in neutrophils we demonstrate here validation of two research antibodies directed against human Hax1 in the human neutrophil model cell line PLB-985 cells. We show that both the mouse anti-Hax1 monoclonal IgG directed against amino acids 10-148 of Hax1 and a rabbit anti-Hax1 polyclonal IgG antibody directed against full-length Hax1 reliably and consistently detect Hax1 during immunoblotting of three different PLB-985 cell densities. Using shRNA mediated Hax1 knockdown, we demonstrate the specificity of both Hax1 antibodies. In addition, our results suggest that the rabbit anti-Hax1 polyclonal antibody is provides a stronger intensity in detecting Hax1 protein, with detection in as few as 0.1 x 106 cells in 6 total replicates we have performed.", "keywords": [ "Hax1", "neutrophil", "PLB-985", "tubulin" ], "content": "Introduction\n\nHS1-associated protein X-1 (Hax1) is a 32 kDa protein consisting of 279 amino acids that is ubiquitously expressed1. Hax1 has been demonstrated to be a negative regulator of apoptosis in many immune cell types2–4. Furthermore, Hax1 has been shown to have additional roles in regulating cell motility and adhesion5,6, and is overexpressed in many types of cancer7. Patients with autosomal recessive mutations in the HAX1 gene have a form of severe congenital neutropenia called Kostmann syndrome. Severe congenital neutropenia is characterized by early recurrent bacterial infections and decreased neutrophil counts in the blood stream8.\n\nBecause of the recent increase in Hax1 investigations, it is important to identify reliable antibodies directed against Hax1. Using the human neutrophil model cell line PLB-985 cells, which can be terminally differentiated into neutrophil-like cells after treatment with DMSO, we demonstrate the applicability and selectivity of two antibodies against Hax1. A mouse Hax1 monoclonal antibody (BD Biosciences) that is routinely used in publications investigating Hax15,6,9–11 directed against Hax1 amino acids 10-148, and a rabbit polyclonal antibody (Proteintech Group, Inc.) directed against the full length Hax1 protein6.\n\n\nMaterials and methods\n\nDetails of all reagents used in the Western blotting procedures can be found in Table 1.\n\nAnti-tubulin (beta-) is a mouse monoclonal IgG1 [E7 was deposited to the DSHB by Klymkowsky, Michael (DSHB Hybridoma Product E7)] and was used as a loading control for all Western blots at a dilution of 1:1000 resulting in a final concentration of 45 ng/mL. Rabbit anti-Hax1 (Proteintech Group, Inc, Table 2) is a polyclonal antibody generated to full length Homo sapiens Hax1. The lot number used was 1, and a dilution of 1:1000 was used for all Western blots resulting in a final concentration of rabbit anti-Hax1 of 230 ng/mL. Mouse anti-Hax1 (BD Biosciences) is a mouse monoclonal IgG1 raised against Homo sapiens Hax1 amino acids 10–148. The lot number used was 3266979, and a dilution of 1:1000 was used for all Western blots resulting in a final concentration of 250 ng/mL. Goat anti-rabbit IgG IRDye 680LT and Goat anti-mouse IgG IRDye 800CW (Li-Cor Biosciences, Table 2) were used at a dilution of 1:40,000 (25 ng/mL).\n\nPLB-985 cells were maintained in RPMI 1640 (Mediatech, Inc.) supplemented with 10% fetal bovine serum, 60 μg/mL penicillin, and 100 μg/mL streptomycin (Mediatech, Inc.) at a concentration of 0.1–1 × 106 cells/mL. To differentiate PLB-985 cells into “neutrophil-like” cells 1.25% DMSO (Fisher Scientific) was added to 2 × 105 cells/mL for 6 days. Lentiviral Hax1 shRNA targets were purchased from Open Biosystems. Targets used; Hax1 shRNA (5'-ACAGACACTTCGGGACTCAAT-3') and control shRNA (5'-TGTCTCCGAACGTGTCACGTT-3'). HEK293-FT cells were grown to 70% confluency in a 10cm tissue culture dish for each lentiviral target and transfected using 6μg Hax1, 0.6μg vesicular stomatitis virus (VSV)-G, and 5.4μg cytomegalovirus (CMV) 8.9.1. 72 hour viral supernatant was collected and concentrated using Lenti-X concentrator (Clontech, Inc.) following the manufacturer’s instructions. 1 × 106 PLB-985 cells were infected with viral supernatant for 3 days in the presence of polybrene (4 μg/mL, Santa Cruz Biotechnology). Stable cell lines were generated with puromycin (1 μg/mL, Sigma Aldrich) selection.\n\nDifferentiated PLB-985 cells were counted and 0.1 × 106, 0.5 × 106, and 1 × 106 cells were pelleted by centrifugation.\n\nCells were lysed in Triton X-100 lysis buffer with protease inhibitors (25 mM HEPES, pH 7.5, 150 mM NaCl2, 1% TX-100, 10 mM MgCl2, 1 mM EDTA, 10% glycerol, 1 μg/mL pepstatin A, 2 μg/mL aprotinin, 1 μg/mL leupeptin) on ice for 10 minutes and clarified by centrifugation.\n\nCellular lysate was then removed and added to Laemmli sample buffer, boiled at 90°C for 5 minutes, and run on 10% SDS-PAGE gels.\n\nProteins were then transferred to 0.45μm nitrocellulose membranes (Santa Cruz Biotechnology) at 400mA for 1 hour at 4°C.\n\nFollowing transfer, the membrane was blocked in 5% BSA in 1× T-TS for 1 hour at room temperature with gentle rocking.\n\nMembranes were then probed with mouse anti-tubulin [(beta-) (45 ng/mL)], and either mouse anti-Hax1 (BD Biosciences, 250 ng/mL) or rabbit anti-Hax1 (Proteintech Group, Inc., 230 ng/mL) at room temperature for 1 hour.\n\nAfter primary antibody incubation the membranes were washed 3 × 5 minutes with 1× Tris-HCL/NaCl saline buffer (1× T-TS), see Table 1.\n\nThe membranes were incubated with goat anti-rabbit IgG IRDye 680LT and goat anti-mouse IgG IRDye 800CW (Li-Cor Biosciences, 25 ng/mL).\n\nAfter secondary antibody incubation the membranes were washed 3 × 5 minutes with 1× T-TS.\n\nBlots were imaged with an infrared imaging system (Odysssey Fc; Li-Cor Biosciences) using a 2-minute exposure time.\n\n\nResults\n\nTo determine the reproducibility and sensitivity of the mouse and rabbit anti-Hax1 antibodies on the PLB-985 cells, we performed Western blot analysis using three separate cell densities, 0.1 × 106, 0.5 × 106, and 1 × 106 cells. In our research using the PLB-985 cell system, we routinely use 1 × 106 – 10 × 106 cells in a Western blot. Using beta-tubulin as a loading control our Western blots illustrate an increasing protein concentration in the three samples as would be expected with increasing cell densities. We found that the mouse anti-Hax1 antibody (BD Biosciences) is visible as low as 0.5 × 106 cells, binding to a protein band at the expected Hax1 size of 32 kDa (Figure 1). In six different experiments (Figure 1 and Figure 4) we found inconsistency in protein detection with the Ms anti-Hax1 antibody. In all blots Hax1 was visible, however with varying degrees of intensity. Conversely, when the rabbit anti-Hax1 antibody (Proteintech Group, Inc.) was used, the antibody gave consistent and robust detection (Figure 2 and Figure 4). In some cases, Hax1 can be detected in as low as 0.1 × 106 cells using the Rb anti-Hax1 antibody (Figure 2C). We do not believe the difference between the two antibodies is due to variations in the cell extract or imaging software because when the same cell extract is immunoblotted on two different blots and scanned simultaneously the difference in sensitivity can be observed (Figure 3A). Using the Odyssey imaging system (Li-Cor Biosciences) to measure the intensity of each band, we calculated the intensity ratio of Hax1 relative to the tubulin loading control (Figure 3B). In both blots the levels of tubulin are similar, however it is evident that the rabbit anti-Hax1 antibody exhibits a stronger signal compared to the mouse monoclonal antibody. However, it should be noted that both antibodies reliably detect Hax1 in differentiated PLB-985 cells.\n\nWestern blot analysis of differentiated PLB-985 cell lysates from 0.1 × 106, 0.5 × 106, and 1 × 106 cells from three independent replicates (A–C). Mouse anti-tubulin (beta-) is used as a loading control and can be seen present at 55 kDa. Mouse anti-Hax1 detects a band at approximately 32 kDa as predicted. Hax1 can be detected in densities of 0.5 × 106 and 1 × 106 cells.\n\nWestern blot analysis of differentiated PLB-985 cell lysates from 0.1 × 106, 0.5 × 106, and 1 × 106 cells from three independent replicates (A–C). Mouse anti-tubulin (beta-) is used as a loading control and can be seen present at 55 kDa. Rabbit anti-Hax1 detects a band at approximately 32 kDa as predicted. Hax1 can be detected in densities as low as 0.1 × 106 cells.\n\n(A) Western blot analysis of differentiated PLB-985 cell lysates from 0.1 × 106, 0.5 × 106, and 1 × 106 cells comparing mouse and rabbit anti-Hax1 antibodies. Lysates from the same cell extractions were run on a single SDS-PAGE gel and blotted onto a single nitrocellulose membrane. After transfer, the membrane was cut and probed with either mouse anti-Hax1 or rabbit anti-Hax1. The membranes were imaged simultaneously. (B) Quantification of the band intensities was measured and the ratios of Hax1 to tubulin were graphed.\n\nWestern blot analysis of differentiated PLB-985 cell lysates from 0.1 × 106, 0.5 × 106, and 1 × 106 cells expressing either control shRNA or Hax1 shRNA from three independent replicates (A–C). Mouse anti-tubulin (beta-) is used as a loading control and can be seen present at 55 kDa. Both mouse and rabbit anti-Hax1 detects a band at approximately 32 kDa as predicted. The specificity of each antibody can be observed by the reduction in intensities in cells expressing Hax1 shRNA.\n\nTo demonstrate the specificity of both Hax1 antibodies we generated stably-expressing control shRNA and Hax1 shRNA PLB-985 cells (Figure 4). As described previously using the mouse anti-Hax1 antibody the control shRNA cells show inconsistent staining intensity, however it is clear that in these samples it is more robust than in the wild-type PLB-985 cells. Both the mouse anti-Hax1 and rabbit anti-Hax1 antibodies show reduced detection in the Hax1-deficient PLB-985 cells, which demonstrates that the antibodies are highly specific for Hax1. In many of the experiments we observed additional background bands that could be attributed to the goat anti-rabbit IgG secondary antibody (Figure 5).\n\n(A) Western blot analysis using goat anti-rabbit IgG 680LT only on cell lysates from 0.1 × 106, 0.5 × 106, and 1 × 106 differentiated PLB-985 cells (dPLB-985), and (B) control shRNA and Hax1 shRNA expressing PLB-985 cells. Two predominant background bands can be observed at approximately 60 and 70 kDa, and one band around 30 kDa. These background bands can also be seen in Figure 1, 2, and Figure 4.\n\n\nConclusion\n\nHere we show validation and comparison results of two antibodies generated against HS1-associated protein X-1, an anti-apoptotic protein that has a multi-factorial role in regulating cell proliferation and differentiation, cell motility, and cancer. Homozygous loss-of-function of Hax1 results in severe congenital neutropenia, a life threatening loss of circulating neutrophils in the blood stream. Thus studying the function of Hax1 in primary neutrophils and the neutrophil model cell line PLB-985 will help elucidate the disease pathogenesis of neutropenia syndromes. We demonstrate that mouse anti-Hax1 (BD Biosciences) and rabbit anti-Hax1 (Proteintech Group, Inc.) are both specific for Hax1. Furthermore we show that as little as 0.5 × 106 differentiated PLB-985 cells can be used to reliably detect Hax1 expression with both of the antibodies. We have evidence that the rabbit anti-Hax1 (Proteintech Group Inc.) results in a more robust and consistent detection of Hax1, likely due to the polyclonal nature of the antibody. Finally, lentiviral knockdown of endogenous Hax1 expression results in loss of Hax1 detection by both mouse anti-Hax1 and rabbit anti-Hax1 demonstrating the specificity of each antibody.\n\nIn conclusion we recommend the use of either mouse or rabbit anti-Hax1 antibodies shown here for studies using the PLB-985 cells as a neutrophil model cell line. Furthermore, it is our conclusion that a minimum cell density of 0.5 × 106 neutrophils should be used as a starting point for immunoblotting of Hax1, with greater than or equal to 1 × 106 cells being optimal.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for Figure 3, 10.5256/f1000research.6516.d4758212", "appendix": "Author contributions\n\n\n\nPC and KI co-wrote and conceived of the article. PC developed the figures. KI performed Western blotting and cell culture. All authors agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nKI and PC are funded by the University of West Florida.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to acknowledge the University of West Florida, for resources and lab support to do this study. The authors would like also like to thank John Steele and Roberta Palau for tissues culture assistance and aid in preparing the materials for this publication.\n\n\nReferences\n\nSuzuki Y, Demoliere C, Kitamura D, et al.: HAX-1, a novel intracellular protein, localized on mitochondria, directly associates with HS1, a substrate of Src family tyrosine kinases. J Immunol. 1997; 158(6): 2736–44. PubMed Abstract\n\nChao JR, Parganas E, Boyd K, et al.: Hax1-mediated processing of HtrA2 by Parl allows survival of lymphocytes and neurons. Nature. 2008; 452(7183): 98–102. PubMed Abstract \| Publisher Full Text\n\nBaumann U, Fernández-Sáiz V, Rudelius M, et al.: Disruption of the PRKCD-FBXO25-HAX-1 axis attenuates the apoptotic response and drives lymphomagenesis. Nat Med. 2014; 20(12): 1401–9. PubMed Abstract \| Publisher Full Text\n\nKlein C: Genetic defects in severe congenital neutropenia: emerging insights into life and death of human neutrophil granulocytes. Annu Rev Immunol. 2011; 29: 399–413. PubMed Abstract \| Publisher Full Text\n\nRadhika V, Onesime D, Ha JH, et al.: Galpha13 stimulates cell migration through cortactin-interacting protein Hax-1. J Biol Chem. 2004; 279(47): 49406–13. PubMed Abstract \| Publisher Full Text\n\nCavnar PJ, Berthier E, Beebe DJ, et al.: Hax1 regulates neutrophil adhesion and motility through RhoA. J Cell Biol. 2011; 193(3): 465–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMirmohammadsadegh A, Tartler U, Michel G, et al.: HAX-1, identified by differential display reverse transcription polymerase chain reaction, is overexpressed in lesional psoriasis. J Invest Dermatol. 2003; 120(6): 1045–51. PubMed Abstract \| Publisher Full Text\n\nCarlsson G, Elinder G, Malmgren H, et al.: Compound heterozygous HAX1 mutations in a Swedish patient with severe congenital neutropenia and no neurodevelopmental abnormalities. Pediatr Blood Cancer. 2009; 53(6): 1143–6. PubMed Abstract \| Publisher Full Text\n\nCilenti L, Soundarapandian MM, Kyriazis GA, et al.: Regulation of HAX-1 anti-apoptotic protein by Omi/HtrA2 protease during cell death. J Biol Chem. 2004; 279(48): 50295–301. PubMed Abstract \| Publisher Full Text\n\nFadeel B, Grzybowska E: HAX-1: a multifunctional protein with emerging roles in human disease. Biochim Biophys Acta. 2009; 1790(10): 1139–48. PubMed Abstract \| Publisher Full Text\n\nHan Y, Chen YS, Liu Z, et al.: Overexpression of HAX-1 protects cardiac myocytes from apoptosis through caspase-9 inhibition. Circ Res. 2006; 99(4): 415–23. PubMed Abstract \| Publisher Full Text\n\nKristina I, Cavnar P: Dataset 1 in: Immunoblotting validation of research antibodies generated against HS1-associated protein X-1 in the human neutrophil model cell line PLB-985. F1000Research. 2015. Data Source" }	[ { "id": "8986", "date": "18 Jun 2015", "name": "Andrew D. Chalmers", "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 work presented by Kristina and Cavnar presents data validating two commercial antibodies raised against Hax1. Their experimental approach is well explained and there is a good description of the methods and reagents used. The data is also convincing and well presented, I particularly liked the fact that complete blots are shown and knockdown used to show specificity. Overall, it is a very strong example of an antibody validation paper.I believe the manuscript is suitable for indexing subject to a few fairly minor changes, which are detailed below:I would make it clear that the antibodies are commercially available in the abstract. It would be worth double checking that details are provided for all of the western blotting reagents and methods. There is plenty of evidence that small changes can alter antibody behaviour and it really helps those trying to reproduce the work. One thing I noticed is that I don’t think the contents of Laemmli loading buffer are spelt out, does it have reducing agent in it? Another example is how long were secondaries incubated for? Figure 3- Is the quantification based on one or more experiments- this should be clear in the legend. Dataset 1- should be “plotted” and not “graphed”. I was interested in the background which appears to be caused by the secondary antibodies. The authors state it is due to the anti-rabbit antibody, but it seems there is at least some with the anti-mouse staining? I would be interested if this is common with this cell line or is it this set of secondaries- I think this would be worth briefly mentioning in the discussion. The authors should present the quantification of the knockdown (figure 4). If both antibodies are specific then you would predict that they will show roughly similar levels of reduction in signal. This would suggest that the remaining signal is caused by a lack of knockdown rather than low levels of non-specific staining.", "responses": [ { "c_id": "1493", "date": "10 Aug 2015", "name": "Peter Cavnar", "role": "Author Response", "response": "Thank you for the kind review of our manuscript. We have addressed your comments below:Commercial availability of the antibodies has now been included in the abstract, introduction, and discussion. We have now included the composition of the Laemmli loading buffer and also our transfer buffer composition to Table 1. Figure 3 includes quantification from three independent experiments. All references to datasets and graphs have been changed to \"plotted\". This was a mistake on our part in our first submission the two channels were overlays. This has been corrected and all blots have been separated into their subsequent mouse and rabbit channels. We hope it is clear now that the background bands we see are due to the goat anti-rabbit 680 secondary antibody with no detectable background in the mouse channel. This was a great suggestion and has been completed. The quantification is highly variable at low cell densities, however at 1 x 10^6 cells the level of the Hax1 KD is relatively constant and significant using the mouse or rabbit Hax1 antibodies. This demonstrates the specificity of the Hax1 antibodies used in this study." } ] }, { "id": "9249", "date": "06 Jul 2015", "name": "Mautusi Mitra", "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 paper titled “\"Immunoblotting validation of research antibodies generated against HS1-associated protein X-1 in the human neutrophil model cell line PLB-985.\" by Kristina and Cavnar focuses on the validation of two commercial antibodies generated against the human Hax1 protein. The experiments are in most part well designed. The research approaches, methods and the reagents used for the experiments are well described. The researchers have shown the entire Western blot and have also used Hax1-deficient PLB-985 cell lines to demonstrate the specificity of the two Hax1 antibodies. The data presented in the article strongly indicate that both the antibodies, especially the rabbit anti-Hax1 polyclonal antibody, can be used for specific detection of the human Hax1 protein and also shows the minimum cell density of neutrophils that can be used for immunoblotting experiments.The manuscript is suitable for indexing, provided the authors address few issues, which are described below:It is not clear if the antibodies were purchased commercially or were generated by the research laboratory with the help of a commercial company. This should be clarified. Hax1 interacts with the polycystic kidney disease protein PKD2, located in the ER membrane/plasma membrane protein. Have the authors checked the efficiency of protein solubilization in their experiments? Have the authors done western analyses on a blot which has proteins transferred from a SDS-PAGE gel that has the intact stacking gel? If the solubilization is incomplete, insoluble proteins will be detected by the respective antibody in the stacking wells in the blot. This experiment can help to confirm the minimum cell density of neutrophils to be used for immunoblotting and might help to improve the sensitivity of the Hax1 detection by these antibodies. The authors should specify the detailed composition of the Laemmli sample buffer as the composition of this buffer varies slightly from lab to lab. Some research labs add 2M-4M urea in the Laemmli buffer for complete solubilization of membrane bound- hydrophobic proteins that have a tendency to precipitate in SDS when heated at 95°C. If urea plus SDS are used for sample solubilization, samples are incubated at room temperature for 30 minutes [as heating causes urea to break down]. Have the authors tried immunoblotting with the pre-immune serum to make sure that these antibodies do not have any non-specific interactions? The authors should make this clear in the manuscript. The authors should clearly state the duration of the secondary antibody incubation which is missing in the manuscript. The authors state that the background protein bands are due to the goat anti- rabbit IgG secondary antibody (Figure 5). But there are background bands with the goat anti-mouse IgG secondary antibody (Figure 5). Have the authors performed the same experiment with the goat-anti-mouse secondary antibody? Is the background due to the cell line used in the experiment or due to the specific batch of the antibody? It would be nice to discuss this issue in the discussion section of the manuscript. No explanation was given for the observation why the control shRNA cells show more intense staining intensity than the wild type cells. This should be addressed in the discussion. Figure 3B does not clearly state how many replicates were used for the quantification of the band intensities. The authors should quantify the Hax1 knockdown in the Hax1 shRNA cell line (Figure 4). Without quantification, it is hard to accurately distinguish the specific Hax1 reduction from the background non-specific weak binding of antibodies.", "responses": [ { "c_id": "1492", "date": "10 Aug 2015", "name": "Peter Cavnar", "role": "Author Response", "response": "Thank you for your kind review of our manuscript. Below I have addressed your comments:We have clarified in the abstract, introduction, and discussion that these antibodies are commercially available. We have performed these experiments using gels that have not had the stacking gel removed and we do not observe any noticeable signal in the stacking gel. Furthermore, Gallagher et al. use a 1% TX-100 lysis buffer in their experiments similar to our lysis conditions. It would be interesting to purify the nuclear extract and compare to the cytosolic fractions to determine this further. The composition of the Laemmli sample buffer has now been added to Table 1. Thank you for this suggestion. Pre-immune sera experiments have been added to figure 5. In short we find that there is increased background using the pre-immune sera that we don't typically see. This demonstrates that the commercially available antibodies are relatively clean after the purification process. The 1 hour incubation of the secondary antibody has been added to the manuscript. The apparent background bands noticed in the mouse channel was a mistake on our part. For many of the figures the rabbit and mouse channels were overlays. This has been fully corrected and every figure now contains separate blots for the mouse and rabbit channels. As can be seen the background is only observed in the rabbit channel. We have repeated these experiments and on average we do not see any significant increase in band intensities between the wild-type PLB-985 cells and the control shRNA cells (see Figure 5C). This could have been variation due to loading in the past figures. We hope the new figures are more clear in this regard. Figure 3B is quantification from three independent replicates. This has been clarified in the figure legend, and the included dataset 1 contains the raw numerical information. This has been completed. Quantifying the knockdown from the 0.1 x 10^6 and 0.5 x 10^6 cell samples was highly variable due to the low level of protein available, however at the 1 x 10^6 cell density quantification was reliable and both antibodies showed a relatively equal level of knockdown." } ] }, { "id": "9247", "date": "08 Jul 2015", "name": "Lawrence L. LeClaire III", "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\nKristina and Cavnar present data to validate two commercial antibodies for HSA-associated protein X-1 in a human neutrophil cell line. The authors describe the conditions for the experiments as well as provide sufficient information/protocols for reader about the conditions used in that assays. In addition, the authors present the entire blots and describe the secondary bands that were detected. Only minor changes in the text should be made (below).The commercial source of the antibody should be listed in the abstract. It currently gives the impression that the antibodies used in the study were generated by the authors. The electrophoretic transfer conditions should be listed. Transfer buffer contents (%MeOH) as well as the model of transfer equipment (dry, semi-dry) will cause variability in protein transfer/antibody detection. Did the authors attempt to vary blocking conditions (milk, commercial blocking agents)? Is purified HAX1 available to use as a positive control or to build a standard curve for quantifying total HAX1 in the samples? Not necessary for this publication but would be extremely useful. Figure 3B suggests that detection is close to linear (especially with LiCor technology). Authors should denote the size of bands on an SDS-PAGE as \"relative mobility.\" For example, a band with a relative mobility of 32 kDa was detected... In Figure 3, change \"After transfer, the membrane was cut..\" To \"the membrane was divided\" Are the authors certain that the band detected by the secondary antibody is the lower of the two bands in the experiments? A single lane could be divided and one side probed with primary and the other with only secondary to confirm.", "responses": [ { "c_id": "1491", "date": "10 Aug 2015", "name": "Peter Cavnar", "role": "Author Response", "response": "Thank you for your kind review. I have revised and submitted a new version of this manuscript. Per your questions I will address each one:I have included the commercial availability of the antibodies in the abstract, introduction, and discussion. The transfer buffer conditions have now been included in Table 1. We have tried blocking with non-fat dry milk and do not find any differences in the selectivity of the Hax1 antibodies, however we did get considerable background from the milk in the goat anti-rabbit 680 channel. We have tested this with a blank membrane. Therefore we will traditionally use 5% BSA to block. We do have bacterial constructs that could be used to purify Hax1. Naturally that was not the purpose of this study, but I concur that these antibodies do behave quite linearly. Thank you for this suggestion, we have included all sizes as a relative mobility in our results and figure legends. We have edited figure 3 legend to reflect this. We have revised figure 5 which demonstrates that the background band that has a relative mobility of 32 kDa is from the secondary antibody." } ] } ]	1	https://f1000research.com/articles/4-148
https://f1000research.com/articles/4-517/v1	07 Aug 15	{ "type": "Review", "title": "Novel Insights and Therapeutics in Multiple Sclerosis", "authors": [ "Catriona A. Wagner", "Joan M. Goverman", "Catriona A. Wagner" ], "abstract": "The last twelve years have witnessed the development of new therapies for relapsing-remitting multiple sclerosis that demonstrate increased efficacy relative to previous therapies. Many of these new drugs target the inflammatory phase of disease by manipulating different aspects of the immune system. While these new treatments are promising, the development of therapies for patients with progressive multiple sclerosis remains a significant challenge. We discuss the distinct mechanisms that may contribute to these two types of multiple sclerosis and the implications of these differences in the development of new therapeutic targets for this debilitating disease.", "keywords": [ "multiple sclerosis", "immune", "therapeutics" ], "content": "Introduction\n\nMultiple sclerosis (MS) is an inflammatory, demyelinating disease of the central nervous system (CNS) and is the most common cause of non-traumatic neurologic disability in young adults. Although the etiology of MS has been debated in the past, recent data from GWAS studies provide exceptionally strong evidence that MS is an autoimmune disease1–3. This view was originally based on the observation that experimental autoimmune encephalomyelitis (EAE), an animal model that recapitulates many important features of MS, is induced by the activation or adoptive transfer of self-reactive CD4+ T cells specific for myelin proteins4. Furthermore, susceptibility to MS is most strongly associated with MHC class II alleles1. Inflammatory lesions and plaques of demyelination in the CNS are considered hallmark features of MS; however, substantial heterogeneity (with respect to clinical course, pathology, and response to therapies) is seen among patients. This review will focus on current ideas regarding the role of different pathological mechanisms in shaping these different manifestations of MS. Achieving an understanding of the specific pathogenic pathways that are relevant to individual patients with MS is critically important in order to predict which patients will respond well to current therapies, and to identify new therapeutic targets that can be tailored for patients with different types of MS.\n\n\nHeterogeneity in MS\n\nFor the majority of patients with MS, disease course begins with a relapsing-remitting phase with intermittent, discrete periods of neurological symptoms that coincide with the appearance of inflammatory lesions. Over time, most patients with relapsing-remitting MS convert to a progressive stage called secondary progressive MS, characterized by a decrease in the frequency, or complete cessation, of relapses and gadolinium-enhancing MRI lesions, and the gradual accumulation of disability associated with brain and spinal cord atrophy. A small subset of patients with MS initially present with a progressive disease course that is not preceded by clinical exacerbations. This form of MS is called primary progressive MS because it is not preceded by a relapsing-remitting phase5,6.\n\nSignificant heterogeneity is also observed in the structure of lesions and types of tissue injury seen in patients with MS. Lesions have been grouped into four different patterns, based on the presence or absence of antibody deposition and complement activation, differential loss of certain myelin proteins, whether lesions occur at perivenous sites, and whether oligodendrocytes are spared or die by apoptosis or necrotic cell death7,8. The observation that individual patients with MS exhibit only one pattern of lesions led to the proposal that different patterns may arise from distinct pathogenic pathways. However, these patterns have not been associated with particular clinical disease courses. Differences are also seen among patients in the distribution of lesions between the brain and spinal cord. The vast majority of patients with relapsing-remitting MS exhibit numerous lesions in the cerebral white matter. However, a small subset (10–15%) of patients with MS exhibit lesions predominantly in the spinal cord with relatively sparse brain involvement.\n\n\nRelapsing-remitting MS pathogenesis and therapies\n\nRelapsing-remitting MS is the best studied form of MS, as the majority of patients initially exhibit this form of disease, and the EAE model recapitulates many aspects of the inflammatory lesions seen in relapsing-remitting MS. Both MRI and immunohistochemical analyses of tissue sections indicate that inflammation is the key component leading to tissue injury and clinical relapses in relapsing-remitting MS patients. Perivascular lesions are comprised of inflammatory infiltrates dominated by lymphocytes and myeloid cells. Our conceptual framework for understanding how these lesions arise is based on studies in EAE. In EAE, activation of myelin-specific T cells induces expression of adhesion molecules and integrins that facilitate their extravasation across the blood brain barrier (BBB). Upon entry into the CNS, myelin-specific T cells are reactivated by the small number of antigen-presenting cells (APCs) in the healthy CNS that constitutively present myelin antigens in the perivascular and subarachnoid spaces. This reactivation triggers the T cells to produce soluble, inflammatory mediators that cause BBB permeability and recruitment of a range of inflammatory leukocytes4. Formation of a localized, inflammatory environment within lesions results in plaques of demyelination and axonal damage.\n\nIn EAE, CD4+ T cells are the predominant lymphocyte in the infiltrate, as the protocol for inducing EAE specifically primes CD4+ T cells. However, there is an abundance of CD8+ T cells in lesions in MS patients, and clonal expansion indicative of antigen-driven activation is more evident in the CD8+ T cell subset in the blood and cerebrospinal fluid of MS patients. Furthermore, depletion of CD4+ T cells resulted in limited therapeutic efficacy in MS patients9, although the results of this trial may not be conclusive. In contrast, treatments that deplete all leukocytes demonstrated greater efficacy in MS patients10. These observations raise key questions (Box 1) about the role for different lymphocyte subsets in CNS autoimmunity.\n\n\n\nDo both CD4 and CD8 T cells contribute to the pathogenesis of MS? What are their relevant effector mechanisms?\n\nDo CD8+ T cells exhibit both regulatory and pathogenic activity as seen for different subsets of CD4+ T cells?\n\nHow do B cells contribute to the pathogenesis of MS?\n\nAn important role for CD4+ T cells in MS is clear from the strong association of MHC class II molecules with disease susceptibility1. Studies of effector CD4+ T cells in EAE implicated both IFN-γ-producing Th1 and IL-17-producing Th17 cells as pathogenic mediators11. Originally, Th1 cells were considered the main effector cells, as adoptive transfer of Th1 cells could induce EAE12,13. This finding was consistent with an earlier observation that MS was exacerbated by the administration of IFN-γ14. However, mice deficient in cytokines associated with the differentiation and function of Th1 cells developed severe EAE15. In contrast, deficiency in IL-23, which is important for Th17 cell stabilization, conferred resistance to EAE, suggesting that Th17 cells may be the true effector cells16. An increase in IL-17 transcripts was also reported in chronic MS lesions17. However, IL-17A-/- and IL-17F-/- mice, treated with anti-IL-17A blocking antibody, are still susceptible to EAE18. Subsequently, GM-CSF was reported to be essential for EAE pathogenesis19–21, and recent studies have suggested that T cells producing GM-CSF may represent a distinct T cell subset22,23.\n\nThe studies described above were all carried out in C57BL/6 mice. Our studies of EAE in C3Heb/Fej mice have provided further insight into the complexity of EAE pathogenesis. While the incidence of EAE in C3Heb/Fej mice is strongly reduced in mice lacking both IL-17RA and IFN-γ receptors, disease incidence was only modestly reduced in mice that lacked either only IL-17 or only IFN-γ signaling. Importantly, we found that IFN-γ and IL-17 signaling had a much greater impact on the pattern of lesion localization within the CNS. IL-17 promoted inflammation in the brain via induction of chemokines that recruit neutrophils, and neutrophils contributed significantly to parenchymal tissue damage in the brain24. In contrast, IFN-γ inhibited inflammation in the brain by inhibiting neutrophil recruitment. Surprisingly, IFN-γ exerted the opposite influence in the spinal cord by promoting both neutrophil recruitment and inflammation in this microenvironment. Despite the enhanced, IFN-γ-mediated neutrophil recruitment to the spinal cord, neutrophils contributed less to spinal cord tissue damage compared to their role in the brain24. Recent work by Segal and colleagues in C57BL/6 mice also suggested that neutrophils may be more important effector cells in the brain compared to the spinal cord25. Interestingly, in contrast to EAE in C57BL/6 mice, we have observed only a modest reduction in disease incidence in C3Heb/Fej mice when EAE is induced by adoptive transfer of GM-CSF-/- T cells in C3Heb/Fej mice, suggesting that the stringent requirement for GM-CSF seen in C57BL/6 mice may be a strain-specific finding (Pierson E.R., Johnson M.C., and Goverman J.M., unpublished observations). Collectively, these studies demonstrate that it is critical to study different mouse strains in order to understand the complexity of disease manifestation in MS patients.\n\nThe finding that the brain and spinal cord microenvironments in mice respond very differently to cytokines produced by infiltrating T cells suggests that patients with distinct neuroinflammatory patterns may respond quite differently to therapies that target specific cytokines. There are also other challenges in designing cytokine-based therapies for MS patients. Despite the substantial data supporting key roles for Th1 and Th17 cells in EAE, a clinical trial administering ustekinumab (an antibody that neutralizes cytokines that promote differentiation of both Th1 and Th17 cells), had no beneficial effect26. It is difficult to draw conclusions from this one trial, however, especially in light of the fact that it is not known whether effector T cell differentiation occurs in the periphery or the CNS, or how important ongoing T cell differentiation is in patients with established MS. The dramatic benefit seen in patients with psoriasis following administering of an IL-17-neutralizing antibody has also not yet been reported for similar clinical trials in patients with MS. It is possible that better stratification of patients with MS, with respect to their neuroinflammatory pattern and other key disease characteristics, is needed to properly evaluate the effectiveness of therapeutic targeting of specific cytokines.\n\nCD8+ T cells often predominate in tissue sections and in CSF of MS patients, and clonal expansion is more commonly observed in the CD8+ compared to the CD4+ T cell subset4,27,28. However, the role of CD8+ T cells is still unclear, as EAE models have pointed to both pathogenic and regulatory functions. Global elimination of CD8+ T cells using either CD8-/- mice or antibody-mediated depletion of CD8+ T cells in vivo suggested a regulatory role for CD8+ T cells29,30. The observations that Qa-1-deficient mice exhibit increased susceptibility to EAE, and that adoptive transfer of Qa-1-restricted CD8+ T cells ameliorates disease, suggested that there may be distinct regulatory subsets of CD8+ T cells31,32. Other studies have reported a pathogenic role for myelin-specific CD8+ T cells in CNS autoimmunity33–35, and animal models using neo-antigens expressed in the CNS and CD8+ T cells that recognize the neo-antigen support a pathogenic role for CD8+ T cells36–40. We identified CD8+ T cells that recognize a MHC class I-restricted myelin basic protein (MBP) epitope and showed that these CD8+ T cells were pathogenic and produced lesions distinct from those seen in conventional EAE but similar to some lesions seen in patients with MS33,41,42. We also showed that both dendritic cells and oligodendrocytes presented the MHC class I-restricted epitope of MBP within the CNS of mice with CD4+ T cell-initiated EAE43. We speculate that CD8+ T cells could be pathogenic if they are triggered to produce inflammatory cytokines upon encountering dendritic cells and/or lyse oligodendrocytes, but they might ameliorate disease if they subsequently lyse dendritic cells that present antigen to both CD4+ and CD8+ T cells within the CNS. Our preliminary data suggest that recruitment of MBP-specific CD8+ T cells during disease induction can exacerbate CD4+ T cell-initiated EAE and may enhance brain inflammation (Wagner C.A. and Goverman J.M., unpublished data). However, CD8+ T cells may play different roles at different phases of disease, and it is important to identify their specific effects during each disease stage in order to therapeutically target (or harness) their activity.\n\nA pathogenic role for B cells in MS is suggested by the therapeutic benefit observed in patients treated with anti-CD20 monoclonal antibodies (rituximab or ocrelizumab) that deplete B cells44,45. As anti-CD20 does not deplete antibody-producing plasma cells, the ability of B cells to present antigen to T cells may be critical in CNS autoimmunity. In animal models, B cells have been shown to promote EAE induction by acting as antigen-presenting cells for T cells in both the periphery46–49 and CNS50. We found that in the healthy CNS, B cells comprise the majority of MHC class II+ cells, and that they play a role in the initial reactivation of infiltrating myelin-specific T cells, specifically the Th1 subset50. Despite the therapeutic efficacy of B cell depletion in MS, B cells that produce IL-10 and ameliorate EAE have been described51,52. If these regulatory B cells retain a stable phenotype in patients, transfer of this subset may be beneficial, in addition to depleting B cells that function only to present antigen to T cells53.\n\n\nSecondary progressive MS pathogenesis and therapies\n\nFollowing the success of anti-inflammatory therapies in relapsing-remitting MS, clinical trials were initiated in secondary progressive MS cohorts. However, clinical trials analyzing the effects of beta-interferon found no significant treatment effect, particularly in patients that had not exhibited MRI lesions for several years54–57. The lack of efficacy of anti-inflammatory therapies in secondary progressive MS suggested that neurodegeneration proceeds independently of inflammation in this stage58. However, further analyses of normal-appearing white matter and meninges tissue sections revealed that inflammation was still present, albeit in a different form59. Staining, using a marker that selectively stains for leaky endothelial cells, indicated that inflammation is compartmentalized behind a less permeable BBB60. Differences in the pathology, clinical signs and response to current therapies suggest that different mechanisms predominate in progressive MS, requiring new therapeutic approaches.\n\nSecondary progressive MS is characterized by increasing brain atrophy and accumulation of irreversible axonal and neuronal degeneration. Gadolinium-enhancing MRI lesions subside61 and a diffuse pattern of inflammation predominates in normal-appearing white and grey matter62. The mechanisms underlying these changes in pathology are poorly understood. Axonal transection begins early in relapsing-remitting MS63; however, the mechanisms may differ in these two stages of disease. During relapsing-remitting MS, inflammatory cells are thought to mediate demyelination via secretion of degradative enzymes, production of oxidative products, and increased levels of glutamate that can damage oligodendrocytes via excessive NMDA receptor signaling64. Thus, transected axons are more abundant in active lesions where inflammatory cells are localized versus chronic lesions in relapsing-remitting MS63. In contrast, demyelination and axonal injury is primarily associated with microglia activation in secondary progressive MS65. The pattern of diffuse parenchymal inflammation, as well as the presence of T and B cells in the meninges, may contribute to microglial activation; however the role of adaptive immune cells in facilitating, versus responding to, microglial activation remains to be established. A major consequence of microglial activation is the production of reactive oxygen species. Mitochondria and mitochondrial DNA are very susceptible to oxidative injury66,67, and axons are extremely susceptible to mitochondrial dysfunction due to their high demand for ATP production to propagate action potentials68. Failure of remyelination in progressive disease is another key difference between relapsing-remitting MS and secondary progressive MS. The state of chronic demyelination leads to diffusion of Na+ channels away from nodes of Ranvier and a subsequent influx of sodium and increased ATP consumption. The resulting energy imbalance ultimately leads to axonal degeneration and tissue damage64. Thus, remyelination may be a critical therapeutic target in secondary progressive MS, together with strategies to dampen microglial activation. Potential contributions of adaptive immune cells may also need to be addressed in secondary progressive MS.\n\nAn additional challenge in designing effective therapies for secondary progressive MS is our lack of understanding of mechanisms that lead to subpial demyelination. Therapies that prevent or resolve subpial lesions may be very important in treating secondary progressive MS patients as increased progression in disability is associated with cortical atrophy69, and subpial lesions are represented to a greater extent than leukocortical lesions in the total cortical lesion load62,70. However, subpial lesions typically lack peripheral inflammatory cells and the mechanism of demyelination that produces these lesions, while unknown, may be distinct from mechanisms of demyelination in white matter. Additionally, meningeal inflammatory aggregates are present in some patients with secondary progressive MS and some studies have identified lymph follicle-like structures. Post-mortem analyses of brain and spinal cord tissue revealed proliferating B cells and follicular dendritic cells in the follicles, suggesting germinal center formation71. The presence of these follicles correlates with disease severity and the extent of demyelination72. However, the contributions of these follicles to disease pathogenesis are still unclear.\n\n\nConcluding remarks\n\nIn this review, we have highlighted potential differences in pathogenic mechanisms between relapsing-remitting MS and secondary progressive MS, and how these differences may require distinct therapeutic approaches. Lesions in relapsing-remitting MS patients arise from a complex interplay of both CD4+ and CD8+ T cells, with B cells and innate immune cells playing critical roles in orchestrating T cell responses. The efficacy of therapies such as Natalizumab73 and fingolimod74,75 that reduce T cell entry into the CNS highlight the key role played by T cells in relapsing-remitting MS. While current therapies are relatively effective at reducing new MRI lesion formation and relapse rates, they are broadly anti-inflammatory and often associated with side effects. New therapies capable of targeting inflammation relevant only to the CNS are needed. In addition, patients with relapsing-remitting MS exhibit heterogeneity in CNS lesions, including the distribution of lesions within the CNS, reinforcing the need for treatments tailored to the individual patient. Finally, the long-term impact of reducing inflammatory lesions and clinical relapses during relapsing-remitting MS on the progression of neurological disability remains to be firmly established.\n\nThe distinct pathology seen in secondary progressive MS patients suggests that targeting different disease mechanisms may be important. Specifically, therapies that promote remyelination and prevent microglia activation, mitochondria dysfunction, and oxidative damage, while beneficial in relapsing-remitting MS, appear particularly crucial in treating patients with secondary progressive MS. It is also important to determine the exact role of inflammation during this phase to prevent relapse after tissue damage has been resolved. Because the BBB is more intact in secondary progressive MS, developing therapeutic agents capable of crossing the BBB is an additional challenge. A key to developing therapies for secondary progressive MS is the generation of new animal models that better reproduce the key features seen in this disease. While we have focused on secondary progressive MS, similarities in the pathology seen between secondary progressive MS and primary progressive MS support the hope that the same therapies will be beneficial in both types of disease.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by grants to J.M. Goverman from NIAID (R37 AI107494-01) and the National Multiple Sclerosis Society (RG 4792A6/1), and by a grant from NIAID supporting C.A. Wagner (T32 AI106677).\n\nThe content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.\n\n\nReferences\n\nInternational Multiple Sclerosis Genetics Consortium, Hafler DA, Compston A, et al.: Risk alleles for multiple sclerosis identified by a genomewide study. N Engl J Med. 2007; 357(9): 851–862. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nInternational Multiple Sclerosis Genetics Consortium, Wellcome Trust Case Control Consortium, Sawcer S, et al.: Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature. 2011; 476(7359): 214–219. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nBrynedal B, Duvefelt K, Jonasdottir G, et al.: HLA-A confers an HLA-DRB1 independent influence on the risk of multiple sclerosis. PLoS One. 2007; 2(7): e664. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nGoverman J: Autoimmune T cell responses in the central nervous system. Nat Rev Immunol. 2009; 9(6): 393–407. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAktas O, Kieseier B, Hartung HP: Neuroprotection, regeneration and immunomodulation: broadening the therapeutic repertoire in multiple sclerosis. Trends Neurosci. 2010; 33(3): 140–152. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nLublin FD, Reingold SC: Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis. Neurology. 1996; 46(4): 907–911. PubMed Abstract \| Publisher Full Text\n\nLucchinetti C, Brück W, Parisi J, et al.: Heterogeneity of multiple sclerosis lesions: implications for the pathogenesis of demyelination. Ann Neurol. 2000; 47(6): 707–717. PubMed Abstract \| Publisher Full Text\n\nMetz I, Weigand SD, Popescu BF, et al.: Pathologic heterogeneity persists in early active multiple sclerosis lesions. Ann Neurol. 2014; 75(5): 728–738. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nvan Oosten BW, Lai M, Hodgkinson S, et al.: Treatment of multiple sclerosis with the monoclonal anti-CD4 antibody cM-T412: results of a randomized, double-blind, placebo-controlled, MR-monitored phase II trial. Neurology. 1997; 49(2): 351–357. PubMed Abstract \| Publisher Full Text\n\nCAMMS223 Trial Investigators, Coles AJ, Compston DA, et al.: Alemtuzumab vs. interferon beta-1a in early multiple sclerosis. N Engl J Med. 2008; 359(17): 1786–1801. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nKroenke MA, Carlson TJ, Andjelkovic AV, et al.: IL-12- and IL-23-modulated T cells induce distinct types of EAE based on histology, CNS chemokine profile, and response to cytokine inhibition. J Exp Med. 2008; 205(7): 1535–1541. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nBaron JL, Madri JA, Ruddle NH, et al.: Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma. J Exp Med. 1993; 177(1): 57–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSegal BM, Shevach EM: IL-12 unmasks latent autoimmune disease in resistant mice. J Exp Med. 1996; 184(2): 771–775. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPanitch HS, Hirsch RL, Schindler J, et al.: Treatment of multiple sclerosis with gamma interferon: exacerbations associated with activation of the immune system. Neurology. 1987; 37(7): 1097–1102. PubMed Abstract \| Publisher Full Text\n\nSteinman L: A brief history of TH17, the first major revision in the TH1/TH2 hypothesis of T cell-mediated tissue damage. Nat Med. 2007; 13(2): 139–145. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nCua DJ, Sherlock J, Chen Y, et al.: Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature. 2003; 421(6924): 744–748. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nLock C, Hermans G, Pedotti R, et al.: Gene-microarray analysis of multiple sclerosis lesions yields new targets validated in autoimmune encephalomyelitis. Nat Med. 2002; 8(5); 500–508. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nHaak S, Croxford AL, Kreymborg K, et al.: IL-17A and IL-17F do not contribute vitally to autoimmune neuro-inflammation in mice. J Clin Invest. 2009; 119(1): 61–69. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nCodarri L, Gyülvészi G, Tosevski V, et al.: RORγt drives production of the cytokine GM-CSF in helper T cells, which is essential for the effector phase of autoimmune neuroinflammation. Nat Immunol. 2011; 12(6): 560–567. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nEl-Behi M, Ciric B, Dai H, et al.: The encephalitogenicity of TH17 cells is dependent on IL-1- and IL-23-induced production of the cytokine GM-CSF. Nat Immunol. 2011; 12(6): 568–575. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nPonomarev ED, Shriver LP, Maresz K, et al.: GM-CSF production by autoreactive T cells is required for the activation of microglial cells and the onset of experimental autoimmune encephalomyelitis. J Immunol. 2007; 178(1): 39–48. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSheng W, Yang F, Zhou Y, et al.: STAT5 programs a distinct subset of GM-CSF-producing T helper cells that is essential for autoimmune neuroinflammation. Cell Res. 2014; 24(12): 1387–1402. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nNoster R, Riedel R, Mashreghi MF, et al.: IL-17 and GM-CSF expression are antagonistically regulated by human T helper cells. Sci Transl Med. 2014; 6(241): 241ra280. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSimmons SB, Liggitt D, Goverman JM: Cytokine-regulated neutrophil recruitment is required for brain but not spinal cord inflammation during experimental autoimmune encephalomyelitis. J Immunol. 2014; 193(2): 555–563. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStoolman JS, Duncker PC, Huber AK, et al.: Site-specific chemokine expression regulates central nervous system inflammation and determines clinical phenotype in autoimmune encephalomyelitis. J Immunol. 2014; 193(2): 564–570. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nSegal BM, Constantinescu CS, Raychaudhuri A, et al.: Repeated subcutaneous injections of IL12/23 p40 neutralising antibody, ustekinumab, in patients with relapsing-remitting multiple sclerosis: a phase II, double-blind, placebo-controlled, randomised, dose-ranging study. Lancet Neurol. 2008; 7(9): 796–804. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nFriese MA, Fugger L: Pathogenic CD8+ T cells in multiple sclerosis. Ann Neurol. 2009; 66(2): 132–141. PubMed Abstract \| Publisher Full Text\n\nMars LT, Saikali P, Liblau RS, et al.: Contribution of CD8 T lymphocytes to the immuno-pathogenesis of multiple sclerosis and its animal models. Biochim Biophys Acta. 2011; 1812(2): 151–161. PubMed Abstract \| Publisher Full Text\n\nJiang H, Zhang Sl, Pernis B: Role of CD8+ T cells in murine experimental allergic encephalomyelitis. Science. 1992; 256(5060): 1213–1215. PubMed Abstract \| Publisher Full Text\n\nKoh DR, Fung-Leung WP, Ho A, et al.: Less mortality but more relapses in experimental allergic encephalomyelitis in CD8-/- mice. Science. 1992; 256(5060): 1210–1213. PubMed Abstract \| Publisher Full Text\n\nHu D, Ikizawa K, Lu L, et al.: Analysis of regulatory CD8 T cells in Qa-1-deficient mice. Nat Immunol. 2004; 5(5): 516–523. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nTang X, Maricic I, Purohit N, et al.: Regulation of immunity by a novel population of Qa-1-restricted CD8alphaalpha+TCRalphabeta+ T cells. J Immunol. 2006; 177(11): 7645–7655. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nHuseby ES, Liggitt D, Brabb T, et al.: A pathogenic role for myelin-specific CD8+ T cells in a model for multiple sclerosis. J Exp Med. 2001; 194(5): 669–676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSun D, Whitaker JN, Huang Z, et al.: Myelin antigen-specific CD8+ T cells are encephalitogenic and produce severe disease in C57BL/6 mice. J Immunol. 2001; 166(12): 7579–7587. PubMed Abstract \| Publisher Full Text\n\nFriese MA, Jakobsen KB, Friis L, et al.: Opposing effects of HLA class I molecules in tuning autoreactive CD8+ T cells in multiple sclerosis. Nat Med. 2008; 14(11): 1227–1235. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nMorgan DJ, Liblau R, Scott B, et al.: CD8(+) T cell-mediated spontaneous diabetes in neonatal mice. J Immunol. 1996; 157(3): 978–983. PubMed Abstract\n\nSaxena A, Bauer J, Scheikl T, et al.: Cutting edge: Multiple sclerosis-like lesions induced by effector CD8 T cells recognizing a sequestered antigen on oligodendrocytes. J Immunol. 2008; 181(3): 1617–1621. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nNa SY, Cao Y, Toben C, et al.: Naive CD8 T-cells initiate spontaneous autoimmunity to a sequestered model antigen of the central nervous system. Brain. 2008; 131(Pt 9): 2353–2365. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSobottka B, Harrer MD, Ziegler U, et al.: Collateral bystander damage by myelin-directed CD8+ T cells causes axonal loss. Am J Pathol. 2009; 175(3): 1160–1166. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nNa SY, Eujen H, Göbel K, et al.: Antigen-specific blockade of lethal CD8 T-cell mediated autoimmunity in a mouse model of multiple sclerosis. J Immunol. 2009; 182(10): 6569–6575. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nPerchellet A, Stromnes I, Pang JM, et al.: CD8+ T cells maintain tolerance to myelin basic protein by 'epitope theft'. Nat Immunol. 2004; 5(6): 606–614. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nJi Q, Perchellet A, Goverman JM: Viral infection triggers central nervous system autoimmunity via activation of CD8+ T cells expressing dual TCRs. Nat Immunol. 2010; 11(7): 628–634. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nJi Q, Castelli L, Goverman JM: MHC class I-restricted myelin epitopes are cross-presented by Tip-DCs that promote determinant spreading to CD8+ T cells. Nat Immunol. 2013; 14(3): 254–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHauser SL, Waubant E, Arnold DL, et al.: B-cell depletion with rituximab in relapsing-remitting multiple sclerosis. N Engl J Med. 2008; 358(7): 676–688. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nBar-Or A, Calabresi PA, Arnold D, et al.: Rituximab in relapsing-remitting multiple sclerosis: a 72-week, open-label, phase I trial. Ann Neurol. 2008; 63(3): 395–400. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nWeber MS, Prod'homme T, Patarroyo JC, et al.: B-cell activation influences T-cell polarization and outcome of anti-CD20 B-cell depletion in central nervous system autoimmunity. Ann Neurol. 2010; 68(3): 369–383. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nBettelli E, Baeten D, Jäger A, et al.: Myelin oligodendrocyte glycoprotein-specific T and B cells cooperate to induce a Devic-like disease in mice. J Clin Invest. 2006; 116(9): 2393–2402. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nKrishnamoorthy G, Lassmann H, Wekerle H, et al.: Spontaneous opticospinal encephalomyelitis in a double-transgenic mouse model of autoimmune T cell/B cell cooperation. J Clin Invest. 2006; 116(9): 2385–2392. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nPöllinger B, Krishnamoorthy G, Berer K, et al.: Spontaneous relapsing-remitting EAE in the SJL/J mouse: MOG-reactive transgenic T cells recruit endogenous MOG-specific B cells. J Exp Med. 2009; 206(6): 1303–1316. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nPierson ER, Stromnes IM, Goverman JM: B cells promote induction of experimental autoimmune encephalomyelitis by facilitating reactivation of T cells in the central nervous system. J Immunol. 2014; 192(3): 929–939. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nMatsushita T, Yanaba K, Bouaziz JD, et al.: Regulatory B cells inhibit EAE initiation in mice while other B cells promote disease progression. J Clin Invest. 2008; 118(10): 3420–3430. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nFillatreau S, Sweenie CH, McGeachy MJ, et al.: B cells regulate autoimmunity by provision of IL-10. Nat Immunol. 2002; 3(10): 944–950. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nTedder TF: B10 cells: a functionally defined regulatory B cell subset. J Immunol. 2015; 194(4): 1395–1401. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nPlacebo-controlled multicentre randomised trial of interferon beta-1b in treatment of secondary progressive multiple sclerosis. European Study Group on interferon beta-1b in secondary progressive MS. Lancet. 1998; 352(9139): 1491–1497. PubMed Abstract \| Publisher Full Text\n\nKappos L, Weinshenker B, Pozzilli C, et al.: Interferon beta-1b in secondary progressive MS: a combined analysis of the two trials. Neurology. 2004; 63(10): 1779–1787. PubMed Abstract \| Publisher Full Text\n\nPanitch H, Miller A, Paty D, et al.: Interferon beta-1b in secondary progressive MS: results from a 3-year controlled study. Neurology. 2004; 63(10): 1788–1795. PubMed Abstract \| Publisher Full Text\n\nSecondary Progressive Efficacy Clinical Trial of Recombinant Interferon-Beta-1a in MS (SPECTRIMS) Study Group: Randomized controlled trial of interferon- beta-1a in secondary progressive MS: Clinical results. Neurology. 2001; 56(11): 1496–1504. PubMed Abstract \| Publisher Full Text\n\nTrapp BD, Nave KA: Multiple sclerosis: an immune or neurodegenerative disorder? Annu Rev Neurosci. 2008; 31: 247–269. PubMed Abstract \| Publisher Full Text\n\nFrischer JM, Bramow S, Dal-Bianco A, et al.: The relation between inflammation and neurodegeneration in multiple sclerosis brains. Brain. 2009; 132(Pt 5): 1175–1189. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nHochmeister S, Grundtner R, Bauer J, et al.: Dysferlin is a new marker for leaky brain blood vessels in multiple sclerosis. J Neuropathol Exp Neurol. 2006; 65(9): 855–865. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nFilippi M, Rossi P, Campi A, et al.: Serial contrast-enhanced MR in patients with multiple sclerosis and varying levels of disability. AJNR Am J Neuroradiol. 1997; 18(8): 1549–1556. PubMed Abstract \| Publisher Full Text\n\nKutzelnigg A, Lucchinetti CF, Stadelmann C, et al.: Cortical demyelination and diffuse white matter injury in multiple sclerosis. Brain. 2005; 128(Pt 11): 2705–2712. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nTrapp BD, Peterson J, Ransohoff RM, et al.: Axonal transection in the lesions of multiple sclerosis. N Engl J Med. 1998; 338(5): 278–285. PubMed Abstract \| Publisher Full Text\n\nDutta R, Trapp BD: Relapsing and progressive forms of multiple sclerosis: insights from pathology. Curr Opin Neurol. 2014; 27(3): 271–278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFischer MT, Wimmer I, Höftberger R, et al.: Disease-specific molecular events in cortical multiple sclerosis lesions. Brain. 2013; 136(Pt 6): 1799–1815. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nLin MT, Beal MF: Mitochondrial dysfunction and oxidative stress in neurodegenerative diseases. Nature. 2006; 443(7113): 787–795. PubMed Abstract \| Publisher Full Text\n\nFischer MT, Sharma R, Lim JL, et al.: NADPH oxidase expression in active multiple sclerosis lesions in relation to oxidative tissue damage and mitochondrial injury. Brain. 2012; 135(Pt 3): 886–899. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nMahad DH, Trapp BD, Lassmann H: Pathological mechanisms in progressive multiple sclerosis. Lancet Neurol. 2015; 14(2): 183–193. PubMed Abstract \| Publisher Full Text\n\nGeurts JJ, Calabrese M, Fisher E, et al.: Measurement and clinical effect of grey matter pathology in multiple sclerosis. Lancet Neurol. 2012; 11(12): 1082–1092. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nBø L, Vedeler CA, Nyland HI, et al.: Subpial demyelination in the cerebral cortex of multiple sclerosis patients. J Neuropathol Exp Neurol. 2003; 62(7): 723–732. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSerafini B, Rosicarelli B, Magliozzi R, et al.: Detection of ectopic B-cell follicles with germinal centers in the meninges of patients with secondary progressive multiple sclerosis. Brain Pathol. 2004; 14(2): 164–174. PubMed Abstract \| Publisher Full Text\n\nMagliozzi R, Howell O, Vora A, et al.: Meningeal B-cell follicles in secondary progressive multiple sclerosis associate with early onset of disease and severe cortical pathology. Brain. 2007; 130(Pt 4): 1089–1104. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nRansohoff RM: Natalizumab for multiple sclerosis. N Engl J Med. 2007; 356(25): 2622–2629. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nCohen JA, Barkhof F, Comi G, et al.: Oral fingolimod or intramuscular interferon for relapsing multiple sclerosis. N Engl J Med. 2010; 362(5): 402–415. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nKappos L, Radue EW, O'Connor P, et al.: A placebo-controlled trial of oral fingolimod in relapsing multiple sclerosis. N Engl J Med. 2010; 362(5): 387–401. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation" }	[ { "id": "9920", "date": "07 Aug 2015", "name": "Lars Fugger", "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", "responses": [] }, { "id": "9926", "date": "07 Aug 2015", "name": "Melissa A. Brown", "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", "responses": [] }, { "id": "9922", "date": "07 Aug 2015", "name": "Benjamin M. Segal", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-517
https://f1000research.com/articles/4-516/v1	07 Aug 15	{ "type": "Review", "title": "Membrane manipulations by the ESCRT machinery", "authors": [ "Greg Odorizzi" ], "abstract": "The endosomal sorting complexes required for transport (ESCRTs) collectively comprise a machinery that was first known for its function in the degradation of transmembrane proteins in the endocytic pathway of eukaryotic cells. Since their discovery, however, ESCRTs have been recognized as playing important roles at the plasma membrane, which appears to be the original site of function for the ESCRT machinery. This article reviews some of the major research findings that have shaped our current understanding of how the ESCRT machinery controls membrane dynamics and considers new roles for the ESCRT machinery that might be driven by these mechanisms.", "keywords": [ "ESCRT", "ESCRT machinery", "endosomal sorting complexes required for transport", "membrane dynamics", "intralumenal vesicle budding pathway", "retrovirus budding", "cytokinesis" ], "content": "\n\nGenes encoding the proteins that came to be known collectively as the endosomal sorting complexes required for transport (ESCRT) machinery were discovered in the yeast Saccharomyces cerevisiae1. Genetic screens revealed these genes to be required for the sorting of transmembrane proteins into intralumenal vesicles (ILVs) that bud into the endosome lumen2,3. Biochemical studies assigned the different yeast gene products into distinct protein complexes that were dubbed ESCRTs4–7. Although they were defined in yeast, almost every protein subunit that belongs to an ESCRT complex can also be identified in every other eukaryotic genome, although ESCRT-0 appears restricted to metazoans and fungi8, and some orthologs are also found in archaeal species9,10. Such comprehensive distribution implies that at least some of the functions executed by the ESCRT machinery are highly conserved.\n\n\nESCRTs in the intralumenal vesicle budding pathway\n\nA key insight at the time of its discovery was that the ESCRT-I complex binds ubiquitin4. ESCRT-0 was subsequently found to have the same ubiquitin-binding property7, as was ESCRT-II11. Shortly before the ESCRTs were described as ubiquitin-binding protein complexes, studies revealed that many transmembrane proteins at the plasma membrane are ubiquitinated on their cytosolic domains and that this modification was essential for their degradation at lysosomes12. Together, these observations established that one of the functions of the ESCRT machinery is to target ubiquitinated transmembrane proteins for lysosomal degradation by sorting them at endosomes into ILVs, which then are exposed to degradative hydrolytic enzymes when late endosomes fuse with lysosomes (Figure 1A).\n\nESCRTs are generally represented by a spiral, which reflects the conformation adopted by ESCRT-III. (A) In the intralumenal vesicle (ILV) budding pathway, ESCRTs sort transmembrane proteins at endosomes into ILVs that are degraded when endosomes fuse with lysosomes. (B) In the viral budding pathway, ESCRTs are required at the plasma membrane (PM) for the release of infectious viral particles. (C) In the abscission step of cytokinesis, ESCRTs are required at the PM for membrane scission that separates dividing cells. (D) In the exosome budding pathway, ILVs created by ESCRTs at endosomes are released into the extracellular environment when endosomes fuse with the PM. (E) In the ectosome budding pathway, extracellular vesicles are released directly from the PM. Recent studies implicate the ESCRT machinery functioning in the repair of PM damage (F) and in the elimination of dysfunctional nuclear pore complex (NPC) intermediates (denoted in red) from the inner nuclear membrane (G).\n\nLagging behind the advances in understanding how ESCRTs recognize ubiquitinated ILV cargoes was insight into how the ILVs are actually created. This impasse was breached 10 years later with the discovery that purified subunits of the yeast ESCRT-III complex, when added to synthetic liposomes, can catalyze the membrane scission reaction required for the detachment of ILVs from the limiting membrane13. This observation explained why dominant-negative alleles that affect ESCRT-III function in vivo had been seen in earlier studies to disrupt the budding of retroviruses, a process that is topologically similar to ILV budding and also dependent upon the ESCRT machinery (see below). The same experimental system showed that purified yeast ESCRT-I and ESCRT-II complexes can cooperate with one another to induce the membrane invagination step that initiates ILV budding14. Thus, a broad outline of the division of functions executed by the ESCRT machinery at endosomes was proposed: ESCRT-0, ESCRT-I, and ESCRT-II concentrate ubiquitinated transmembrane proteins at membrane microdomains, where ESCRT-I and ESCRT-II produce incipient buds that are pinched off by the membrane scission activity of ESCRT-III.\n\nDespite the assignment of activities performed by certain ESCRTs that was suggested in vitro13,14, many of the details about their operations remain fuzzy. Perhaps the wooliest thinking concerns ESCRT-III structure-function. What makes it troublesome to study is that ESCRT-III is not a stable protein complex. Instead, its subunits exist in two states that are in equilibrium with one another: subunits either are monomeric and soluble in the cytosol or associate with membranes, where they polymerize into the ESCRT-III complex6,15. All of the ESCRT-III subunits are homologous to one another and are predicted to have the same three-dimensional (3D) structure, yet structures have been determined for only some of the subunits (either in whole or in part), and these structures were solved either for an individual subunit or for two different subunits bound to one another16–19 but not for all of the subunits co-assembled together. Macromolecular structures that correspond to ESCRT-III on membranes in vivo have been visualized by electron microscopy (EM)20,21, but no studies have unambiguously solved a structure of the complete ESCRT-III complex in relative isolation. Even the stoichiometry of its subunits has been defined only in relatively loose terms22.\n\nMurkier still is the mechanism by which ESCRT-III drives the membrane scission reaction. This topic has been reviewed often (e.g., 23) and can be boiled down to two working models. One proposes that the polymerization of ESCRT-III subunits mediates scission. This model draws its support from in vitro studies, including the original demonstration that purified ESCRT-III subunits assembled on synthetic membranes catalyze membrane scission; in this assay, the disassembly of ESCRT-III was necessary only for the replenishment of subunits so that they can participate in further rounds of complex assembly13. An assembly-driven model for scission is also supported by EM of purified ESCRT-III subunits that, when combined with one another, polymerize into tubules constricted at one end to form a dome shape17. This particular conformation led to the idea that the assembly of membrane-associated subunits at the neck of an ILV invagination would narrow the membrane to terminate in constriction and culminate in scission24, but whether a dome-shaped structure is formed by ESCRT-III in vivo is unknown.\n\nThe other model contends that membrane scission by ESCRT-III is coupled to disassembly of the complex by VPS4, a member of the broader family of AAA+ ATPases that are known for their roles in the disassembly of macromolecular complexes. This model is supported by studies conducted in intact cells showing that scission is stalled when VPS4 activity is inhibited20,25,26. Like other AAA+ ATPases, ATP-bound VPS4 assembles into a ring-shaped oligomer with a central pore. The amino terminus of VPS4 is oriented toward the pore of the oligomer, and this region has a domain that binds directly to a motif located in each ESCRT-III subunit9,27,28. Via this interaction, it is thought that VPS4 rips the ESCRT-III complex apart by extruding each individual subunit in succession through its pore when it hydrolyzes ATP. Conceivably, this action could shorten the ESCRT-III complex to gradually draw together the attached membrane in the neck of an ILV constriction, ultimately resulting in scission29. An alternative scenario consistent with studies conducted in vivo is that the engagement of VPS4 with ESCRT-III substrates alters the conformation of the ESCRT-III polymer, potentially serving to catalyze the membrane scission reaction20,30,31.\n\nDespite uncertainties surrounding the native structure of ESCRT-III in vivo and the mechanism by which it drives membrane scission, its function in this process is well supported by studies revealing that the role of ESCRT-III is not restricted to the ILV budding pathway. Indeed, the membrane scission activity of ESCRT-III appears to have essential roles in cellular processes that are unrelated to transmembrane protein degradation. These activities are considered below.\n\n\nESCRT activity in retrovirus budding\n\nVery shortly after they were reported to function in the ILV budding pathway, ESCRT-I and VPS4 were discovered to have a role in the budding of human immunodeficiency virus-1 (HIV-1) from the plasma membrane of infected cells32. In addition to exposing ESCRTs as having non-endosomal functions, this landmark report opened the door for an explosion of studies revealing that a subset of ESCRTs are generally required for the budding of all retroviruses and for many non-retroviral classes of viruses (reviewed in 33). Exploitation of the ESCRT machinery is driven by virally encoded proteins that recruit one or more ESCRT subunits to the membrane microdomain where new viruses are being packaged, the goal being to nucleate a protein interaction network that mediates recruitment of ESCRT-III34. Viral budding critically depends upon ESCRT-III (and VPS4) to catalyze the membrane scission reaction necessary for the release of a virion from the host cell membrane (Figure 1B). Thus, like the ILV budding pathway, the retroviral budding pathway depends on ESCRT-III/VPS4 at the final step to sever the membrane.\n\nThe identities of ESCRTs, aside from ESCRT-III/VPS4, recruited by each retrovirus are idiosyncratic. For instance, HIV-1 recruits both ESCRT-I32 and ALIX, the latter of which interacts with ESCRT-III35, whereas the equine infectious anemia virus recruits only ALIX35. The reason for selective utilization of components outside of the core ESCRT-III/VPS4 machinery might be that viruses do not need all of the activities that are performed by ESCRT complexes 0, I, and II in the ILV budding pathway33. For example, the structural proteins encoded by viruses can target themselves to the site of viral assembly and autonomously generate the membrane curvature needed to produce a virion. That different viruses recruit different ESCRTs to facilitate their exploitation of the membrane scission activity performed by ESCRT-III/VPS4 is a reflection of how evolution allows different solutions for the same problem.\n\nNotably, studies of how viruses take advantage of ESCRTs have yielded fundamental advances in understanding mechanisms that are likely common to all pathways that use ESCRT-III/VPS4. For example, the HIV-1 budding pathway was found to be affected by truncation of ESCRT-III subunits, leading to structure-function analyses demonstrating that the carboxyl termini of several ESCRT-III subunits make intramolecular contacts with their amino-terminal core regions to maintain the proteins as inactive subunits that are incapable of assembling into the ESCRT-III complex16,36,37. More recently, 3D super-resolution microscopy and correlative EM outlined the nanoscale organization within the head of budding HIV-1 virions, suggesting that VPS4 functions at least in part to remodel subunits assembled into the ESCRT-III complex during the membrane constriction process that leads to scission38. A direct role for VPS4 in membrane scission during retrovirus budding was also supported by light microscopy studies that tracked HIV-1 recruitment of ESCRTs over time to the plasma membrane of live cells39,40.\n\nIn addition to characterizing the mechanisms of viral exploitation of the ESCRT machinery, investigations of this process helped lead to the unexpected discovery that ESCRT-III/VPS4 has a role during cytokinesis in animal cells, as discussed in the next section.\n\n\nESCRT activities during cytokinesis\n\nIn metazoans, dividing daughter cells are connected by a relatively thin (200 nm) membrane tubule known as the midbody. A process termed abscission severs the midbody connection, and this step appears to depend on the membrane scission activity of ESCRT-III/VPS4 (Figure 1C; reviewed in 41). This discovery originated from serendipitous observations in a number of disparate studies, not all of which were aimed at understanding ESCRT functions per se, and was bolstered by proteomic studies that identified several ESCRT-binding proteins implicated to function at the midbody. Focused analyses of ESCRTs in cytokinesis revealed that the TSG101 subunit of human ESCRT-I binds directly to CEP55, a microtubule bundling protein, and this interaction mediates ESCRT-I recruitment to the midbody42–44. TSG101 also binds ALIX, the ESCRT-III-associated protein mentioned above45,46. Colocalization of ALIX with TSG101 to the midbody was found to require CEP55, and, consistent with their function during cytokinesis, knocking down TSG101 or ALIX expression results in multi-nucleated cells, and this is a hallmark cytokinetic defect42–44.\n\nThe point of CEP55 mediating recruitment of TSG101/ALIX appears to be so that ALIX can mediate recruitment of CHMP4B, which is the predominant subunit comprising the ESCRT-III complex. As in the ILV and viral budding pathways, ESCRT-III assembly within the midbody neck is expected to constrict the membrane. Three-dimensional electron tomography of the midbody revealed 17-nm-thick spiral filaments adjacent to membrane constrictions21, analogous to spirals of CHMP4B at the plasma membrane that were observed by deep-etch EM20. However, immunolabeling confirmed that the spirals imaged in the latter study were comprised of CHMP4B, and these spirals measured only 5 nm in thickness20, which is close to the 9-nm-thick spirals of the purified yeast ortholog of CHMP4B that were seen in vitro by negative staining EM47 and the 4-nm-thick spirals of the Caenorhabditis elegans ortholog of CHMP4B observed in vitro by cryo EM31. The severe limitations of immunolabeling in electron tomography made it impossible to confirm that CHMP4B (or other ESCRT-III subunits) comprise the 17-nm spiral filaments in the midbody, although this identity seems likely, given that formation of these filaments was dependent upon expression of a different subunit of the ESCRT-III complex21. Given that several cytoskeletal elements also oligomerize into filaments within the midbody, something else might comprise the 17-nm spiral structures presumed to be ESCRT-III; alternatively, ESCRT-III might co-assemble with something else (or with itself in parallel polymers) to form the 17-nm filaments seen by tomography.\n\nNotwithstanding the uncertainties described above, a role for ESCRT-III/VPS4 in membrane scission during cytokinesis is strongly supported by other lines of evidence. First and foremost is the discovery that orthologs of VPS4 and ESCRT-III subunits mediate cytokinesis during cell division in archaebacterial species10,48. Thus, the function of ESCRT-III/VPS4 most likely originated for this purpose. Additionally, time-lapse imaging by high-resolution light microscopy showed sequential recruitment of TSG101 and CHMP4B to the intercellular membrane bridge of the midbody connecting daughter cells, followed by VPS4, whereupon cell separation occurs49. Curiously, the kinetics involved during abscission are considerably slower than what has been observed in retroviral budding, potentially to allow for checkpoints that ensure the fidelity of cytokinesis. For instance, the Aurora B kinase that regulates chromosome contacts with microtubules also phosphorylates CHMP4C (a CHMP4B paralog), which causes CHMP4C to concentrate at the midbody50. As a result, abscission is inhibited, possibly signifying that phosphorylated CHMP4C interferes with the function of CHMP4B. Another intriguing possibility is that the timing of scission is regulated by tension: pulling forces between daughter cells were found to prolong their connection, whereas relaxation of this tension coincided with ESCRT-III assembly and subsequent abscission51. A role for tension in functions of the ESCRT machinery that drive the ILV budding pathway has also been modeled by using data derived from a variety of studies52.\n\n\nESCRTs and the biogenesis of extracellular vesicles\n\nComponents of the ESCRT machinery have also been linked to the formation of extracellular vesicles (EVs) that are secreted by many (if not all) cell types. EVs serve as shuttles that mediate intercellular exchange of proteins, RNAs, and lipids, and their functions in humans have been shown to have critical physiological roles in the immune, cardiovascular, and nervous systems53,54. Two types of EVs have been defined, and they can be readily distinguished by the way in which they are secreted. EVs known as exosomes originate as ILVs within endosomes and are released from the cell when late endosomes fuse with the plasma membrane rather than a lysosome (Figure 1D); EVs known as ectosomes (or shedding vesicles or microvesicles) bud directly from the plasma membrane (Figure 1E).\n\nA connection between exosome biogenesis and the ILV budding pathway is easy to imagine because they have a common origin, and several proteomic studies had identified ESCRT proteins in purified exosomes (reviewed in 55). However, an ESCRT-independent lipid-driven model for exosome biogenesis had originally been invoked by the observation that ceramide, which resides in the inner leaflet of the exosomal membrane, has a small head group and forms extended hydrogen bond networks that cluster this lipid species into microdomains favorable toward budding into the endosome lumen56. The ESCRT machinery was subsequently realized to have a role in exosome biogenesis when it was discovered that syndecan heparan sulfate proteoglycans (one type of cargo packaged into exosomes) interact via a cytoplasmic adaptor protein with ALIX, and exosome budding is blocked upon knocking down expression of ALIX, VPS4, or CHMP4 subunits of the ESCRT-III complex57. This finding indicates that exosomal cargoes can nucleate exosome budding through recruitment of at least a subset of the ESCRT machinery, analogous to the way in which retroviral proteins recruit ESCRTs to bud from infected cells.\n\nEvidence was also obtained that the ESCRT machinery functions in ectosome biogenesis. One report showed that knocking down VPS4 expression in C. elegans inhibits ectosome production58, whereas another showed that a human arrestin domain-containing protein that binds the TSG101 subunit of ESCRT-I is packaged into ectosomes, and the budding of these ectosomes was blocked by knocking down expression of TSG101 or VPS459. TSG101 and VPS4 were similarly reported to be required for the release of ectosomes loaded with human T-cell receptors into the immunological synapse60.\n\nOur understanding of the roles played by ESCRTs in the biogenesis of EVs is still in its infancy, and many details need to be worked out. For instance, do all cell types employ the same set of ESCRTs for exosome or ectosome budding (or both)? In ectosome biogenesis, do ESCRTs function in membrane deformation, as they appear to do in the ILV budding pathway13, or do they function predominately in membrane scission? And beyond ESCRTs, what determines whether an endosome fuses with the plasma membrane to release exosomes instead of fusing with a lysosome to deliver ILVs to their destruction?\n\n\nA new frontier?\n\nThe past year witnessed two discoveries revealing new potential roles for the ESCRT machinery. The repair of small (<100 nm) lesions in the plasma membrane was found to require ESCRT-III (Figure 1F; 61). In this case, however, it remains unclear whether ESCRT-III has a direct role, as the kinetics of its recruitment to wounded plasma membranes are a few seconds slower than the observed amount of time required for repair to occur62,63. One possibility is that remodeling of the plasma membrane occurs in response to wounding and that this process depends directly on ESCRT-III function64. More radically, ESCRT-III and VPS4 were found to be required at the inner nuclear membrane for the removal of defective nuclear pore complex (NPC) assembly intermediates (Figure 1G; 65), but how they might function in this process is unknown. One possibility is that nucleoporins extracted from defective NPCs are packaged by ESCRT-III/VPS4 into vesicles, but nucleoporin degradation mediated by this quality-control pathway appears to involve proteasomal degradation, the substrates of which cannot be membrane-bound. Thus, although it is still too early to tell whether the additional duties ascribed to ESCRTs in plasma membrane wound repair and NPC quality control are mediated by any of the functions already known for the ESCRT machinery, these unexpected findings signal a new frontier of understanding how ESCRTs have evolved from their apparently primordial beginning in cell division.", "appendix": "Competing interests\n\n\n\nThe author declares that he has no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nRaymond CK, Howald-Stevenson I, Vater CA, et al.: Morphological classification of the yeast vacuolar protein sorting mutants: evidence for a prevacuolar compartment in class E vps mutants. Mol Biol Cell. 1992; 3(12): 1389–402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOdorizzi G, Babst M, Emr SD: Fab1p PtdIns(3)P 5-kinase function essential for protein sorting in the multivesicular body. Cell. 1998; 95(6): 847–58. PubMed Abstract \| Publisher Full Text\n\nOdorizzi G, Katzmann DJ, Babst M, et al.: Bro1 is an endosome-associated protein that functions in the MVB pathway in Saccharomyces cerevisiae. J Cell Sci. 2003; 116(Pt 10): 1893–903. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKatzmann DJ, Babst M, Emr SD: Ubiquitin-dependent sorting into the multivesicular body pathway requires the function of a conserved endosomal protein sorting complex, ESCRT-I. Cell. 2001; 106(2): 145–55. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBabst M, Katzmann DJ, Snyder WB, et al.: Endosome-associated complex, ESCRT-II, recruits transport machinery for protein sorting at the multivesicular body. Dev Cell. 2002; 3(2): 283–9. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBabst M, Katzmann DJ, Estepa-Sabal EJ, et al.: Escrt-III: an endosome-associated heterooligomeric protein complex required for mvb sorting. Dev Cell. 2002; 3(2): 271–82. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBilodeau PS, Urbanowski JL, Winistorfer SC, et al.: The Vps27p Hse1p complex binds ubiquitin and mediates endosomal protein sorting. Nat Cell Biol. 2002; 4(7): 534–9. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nLeung KF, Dacks JB, Field MC: Evolution of the multivesicular body ESCRT machinery; retention across the eukaryotic lineage. Traffic. 2008; 9(10): 1698–716. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nObita T, Saksena S, Ghazi-Tabatabai S, et al.: Structural basis for selective recognition of ESCRT-III by the AAA ATPase Vps4. Nature. 2007; 449(7163): 735–9. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nSamson RY, Obita T, Freund SM, et al.: A role for the ESCRT system in cell division in archaea. Science. 2008; 322(5908): 1710–3. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nAlam SL, Sun J, Payne M, et al.: Ubiquitin interactions of NZF zinc fingers. EMBO J. 2004; 23(7): 1411–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHicke L: Ubiquitin-dependent internalization and down-regulation of plasma membrane proteins. FASEB J. 1997; 11(14): 1215–26. PubMed Abstract\n\nWollert T, Wunder C, Lippincott-Schwartz J, et al.: Membrane scission by the ESCRT-III complex. Nature. 2009; 458(7235): 172–7. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nWollert T, Hurley JH: Molecular mechanism of multivesicular body biogenesis by ESCRT complexes. Nature. 2010; 464(7290): 864–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLin Y, Kimpler LA, Naismith TV, et al.: Interaction of the mammalian endosomal sorting complex required for transport (ESCRT) III protein hSnf7-1 with itself, membranes, and the AAA+ ATPase SKD1. J Biol Chem. 2005; 280(13): 12799–809. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMuziol T, Pineda-Molina E, Ravelli RB, et al.: Structural basis for budding by the ESCRT-III factor CHMP3. Dev Cell. 2006; 10(6): 821–30. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nLata S, Schoehn G, Jain A, et al.: Helical structures of ESCRT-III are disassembled by VPS4. Science. 2008; 321(5894): 1354–7. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBajorek M, Schubert HL, McCullough J, et al.: Structural basis for ESCRT-III protein autoinhibition. Nat Struct Mol Biol. 2009; 16(7): 754–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartinelli N, Hartlieb B, Usami Y, et al.: CC2D1A is a regulator of ESCRT-III CHMP4B. J Mol Biol. 2012; 419(1-2): 75–88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHanson PI, Roth R, Lin Y, et al.: Plasma membrane deformation by circular arrays of ESCRT-III protein filaments. J Cell Biol. 2008; 180(2): 389–402. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGuizetti J, Schermelleh L, Mäntler J, et al.: Cortical constriction during abscission involves helices of ESCRT-III-dependent filaments. Science. 2011; 331(6024): 1616–20. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nTeis D, Saksena S, Emr SD: Ordered assembly of the ESCRT-III complex on endosomes is required to sequester cargo during MVB formation. Dev Cell. 2008; 15(4): 578–89. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nHanson PI, Cashikar A: Multivesicular body morphogenesis. Annu Rev Cell Dev Biol. 2012; 28: 337–62. PubMed Abstract \| Publisher Full Text\n\nFabrikant G, Lata S, Riches JD, et al.: Computational model of membrane fission catalyzed by ESCRT-III. PLoS Comput Biol. 2009; 5(11): e1000575. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWemmer M, Azmi I, West M, et al.: Bro1 binding to Snf7 regulates ESCRT-III membrane scission activity in yeast. J Cell Biol. 2011; 192(2): 295–306. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdell MA, Vogel GF, Pakdel M, et al.: Coordinated binding of Vps4 to ESCRT-III drives membrane neck constriction during MVB vesicle formation. J Cell Biol. 2014; 205(1): 33–49. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nStuchell-Brereton MD, Skalicky JJ, Kieffer C, et al.: ESCRT-III recognition by VPS4 ATPases. Nature. 2007; 449(7163): 740–4. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKieffer C, Skalicky JJ, Morita E, et al.: Two distinct modes of ESCRT-III recognition are required for VPS4 functions in lysosomal protein targeting and HIV-1 budding. Dev Cell. 2008; 15(1): 62–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaksena S, Wahlman J, Teis D, et al.: Functional reconstitution of ESCRT-III assembly and disassembly. Cell. 2009; 136(1): 97–109. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCashikar AG, Shim S, Roth R, et al.: Structure of cellular ESCRT-III spirals and their relationship to HIV budding. Elife. 2014; 3: e02184. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nShen QT, Schuh AL, Zheng Y, et al.: Structural analysis and modeling reveals new mechanisms governing ESCRT-III spiral filament assembly. J Cell Biol. 2014; 206(6): 763–77. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGarrus JE, von Schwedler UK, Pornillos OW, et al.: Tsg101 and the vacuolar protein sorting pathway are essential for HIV-1 budding. Cell. 2001; 107(1): 55–65. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMcCullough J, Colf LA, Sundquist WI: Membrane fission reactions of the mammalian ESCRT pathway. Annu Rev Biochem. 2013; 82: 663–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarlson LA, Hurley JH: In vitro reconstitution of the ordered assembly of the endosomal sorting complex required for transport at membrane-bound HIV-1 Gag clusters. Proc Natl Acad Sci U S A. 2012; 109(42): 16928–33. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nStrack B, Calistri A, Craig S, et al.: AIP1/ALIX is a binding partner for HIV-1 p6 and EIAV p9 functioning in virus budding. Cell. 2003; 114(6): 689–99. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nZamborlini A, Usami Y, Radoshitzky SR, et al.: Release of autoinhibition converts ESCRT-III components into potent inhibitors of HIV-1 budding. Proc Natl Acad Sci U S A. 2006; 103(50): 19140–5. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nShim S, Kimpler LA, Hanson PI: Structure/function analysis of four core ESCRT-III proteins reveals common regulatory role for extreme C-terminal domain. Traffic. 2007; 8(8): 1068–79. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nVan Engelenburg SB, Shtengel G, Sengupta P, et al.: Distribution of ESCRT machinery at HIV assembly sites reveals virus scaffolding of ESCRT subunits. Science. 2014; 343(6171): 653–6. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBaumgärtel V, Ivanchenko S, Dupont A, et al.: Live-cell visualization of dynamics of HIV budding site interactions with an ESCRT component. Nat Cell Biol. 2011; 13(4): 469–74. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nJouvenet N, Zhadina M, Bieniasz PD, et al.: Dynamics of ESCRT protein recruitment during retroviral assembly. Nat Cell Biol. 2011; 13(4): 394–401. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMierzwa B, Gerlich DW: Cytokinetic abscission: molecular mechanisms and temporal control. Dev Cell. 2014; 31(5): 525–38. PubMed Abstract \| Publisher Full Text\n\nCarlton JG, Martin-Serrano J: Parallels between cytokinesis and retroviral budding: a role for the ESCRT machinery. Science. 2007; 316(5833): 1908–12. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMorita E, Sandrin V, Chung HY, et al.: Human ESCRT and ALIX proteins interact with proteins of the midbody and function in cytokinesis. EMBO J. 2007; 26(19): 4215–27. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLee HH, Elia N, Ghirlando R, et al.: Midbody targeting of the ESCRT machinery by a noncanonical coiled coil in CEP55. Science. 2008; 322(5901): 576–80. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMartin-Serrano J, Yarovoy A, Perez-Caballero D, et al.: Divergent retroviral late-budding domains recruit vacuolar protein sorting factors by using alternative adaptor proteins. Proc Natl Acad Sci U S A. 2003; 100(21): 12414–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvon Schwedler UK, Stuchell M, Müller B, et al.: The protein network of HIV budding. Cell. 2003; 114(6): 701–13. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nHenne WM, Buchkovich NJ, Zhao Y, et al.: The endosomal sorting complex ESCRT-II mediates the assembly and architecture of ESCRT-III helices. Cell. 2012; 151(2): 356–71. PubMed Abstract \| Publisher Full Text\n\nLindås AC, Karlsson EA, Lindgren MT, et al.: A unique cell division machinery in the Archaea. Proc Natl Acad Sci U S A. 2008; 105(48): 18942–6. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nElia N, Sougrat R, Spurlin TA, et al.: Dynamics of endosomal sorting complex required for transport (ESCRT) machinery during cytokinesis and its role in abscission. Proc Natl Acad Sci U S A. 2011; 108(12): 4846–51. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCarlton JG, Caballe A, Agromayor M, et al.: ESCRT-III governs the Aurora B-mediated abscission checkpoint through CHMP4C. Science. 2012; 336(6078): 220–5. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLafaurie-Janvore J, Maiuri P, Wang I, et al.: ESCRT-III assembly and cytokinetic abscission are induced by tension release in the intercellular bridge. Science. 2013; 339(6127): 1625–9. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nRóżycki B, Boura E, Hurley JH, et al.: Membrane-elasticity model of Coatless vesicle budding induced by ESCRT complexes. PLoS Comput Biol. 2012; 8(10): e1002736. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRaposo G, Stoorvogel W: Extracellular vesicles: exosomes, microvesicles, and friends. J Cell Biol. 2013; 200(4): 373–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCocucci E, Meldolesi J: Ectosomes and exosomes: shedding the confusion between extracellular vesicles. Trends Cell Biol. 2015; 25(6): 364–72. PubMed Abstract \| Publisher Full Text\n\nChoi DS, Kim DK, Kim YK, et al.: Proteomics of extracellular vesicles: Exosomes and ectosomes. Mass Spectrom Rev. 2015; 34(4): 474–90. PubMed Abstract \| Publisher Full Text\n\nTrajkovic K, Hsu C, Chiantia S, et al.: Ceramide triggers budding of exosome vesicles into multivesicular endosomes. Science. 2008; 319(5867): 1244–7. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBaietti MF, Zhang Z, Mortier E, et al.: Syndecan-syntenin-ALIX regulates the biogenesis of exosomes. Nat Cell Biol. 2012; 14(7): 677–85. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nWehman AM, Poggioli C, Schweinsberg P, et al.: The P4-ATPase TAT-5 inhibits the budding of extracellular vesicles in C. elegans embryos. Curr Biol. 2011; 21(23): 1951–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nNabhan JF, Hu R, Oh RS, et al.: Formation and release of arrestin domain-containing protein 1-mediated microvesicles (ARMMs) at plasma membrane by recruitment of TSG101 protein. Proc Natl Acad Sci U S A. 2012; 109(11): 4146–51. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nChoudhuri K, Llodrá J, Roth EW, et al.: Polarized release of T-cell-receptor-enriched microvesicles at the immunological synapse. Nature. 2014; 507(7490): 118–23. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nJimenez AJ, Maiuri P, Lafaurie-Janvore J, et al.: ESCRT machinery is required for plasma membrane repair. Science. 2014; 343(6174): 1247136. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nLennon NJ, Kho A, Bacskai BJ, et al.: Dysferlin interacts with annexins A1 and A2 and mediates sarcolemmal wound-healing. J Biol Chem. 2003; 278(50): 50466–73. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nCorrotte M, Almeida PE, Tam C, et al.: Caveolae internalization repairs wounded cells and muscle fibers. Elife. 2013; 2: e00926. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nAndrews NW, Almeida PE, Corrotte M: Damage control: cellular mechanisms of plasma membrane repair. Trends Cell Biol. 2014; 24(12): 734–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWebster BM, Colombi P, Jäger J, et al.: Surveillance of nuclear pore complex assembly by ESCRT-III/Vps4. Cell. 2014; 159(2): 388–401. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "9923", "date": "07 Aug 2015", "name": "David Teis", "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", "responses": [] }, { "id": "9925", "date": "07 Aug 2015", "name": "Winfried Weissenhorn", "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", "responses": [] }, { "id": "9921", "date": "07 Aug 2015", "name": "Jon Audhya", "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", "responses": [] }, { "id": "9924", "date": "07 Aug 2015", "name": "Phyllis Hanson", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-516
https://f1000research.com/articles/4-512/v1	07 Aug 15	{ "type": "Opinion Article", "title": "An evolutionary perspective on signaling peptides: toxic peptides are selected to provide information regarding the processing of the propeptide, which represents the phenotypic state of the signaling cell", "authors": [ "Keith Daniel Harris", "Ari Barzilai", "Amotz Zahavi", "Ari Barzilai", "Amotz Zahavi" ], "abstract": "Structurally similar short peptides often serve as signals in diverse signaling systems. Similar peptides affect diverse physiological pathways in different species or even within the same organism. Assuming that signals provide information, and that this information is tested by the structure of the signal, it is curious that highly similar signaling peptides appear to provide information relevant to very different metabolic processes. Here we suggest a solution to this problem: the synthesis of the propeptide, and its post-translational modifications that are required for its cleavage and the production of the mature peptide, provide information on the phenotypic state of the signaling cell. The mature peptide, due to its chemical properties which render it harmful, serves as a stimulant that forces cells to respond to this information. To support this suggestion, we present cases of signaling peptides in which the sequence and structure of the mature peptide is similar yet provides diverse information. The sequence of the propeptide and its post-translational modifications, which represent the phenotypic state of the signaling cell, determine the quantity and specificity of the information. We also speculate on the evolution of signaling peptides. We hope that this perspective will encourage researchers to reevaluate pathological conditions in which the synthesis of the mature peptide is abnormal.", "keywords": [ "Signaling peptides", "handicap principle", "signal selection", "evolution" ], "content": "Introduction\n\nSignaling peptides are amino acid chains with diverse structures that serve as signaling molecules. The lengths of signaling peptides vary greatly from less than ten amino acids (such as oxytocin and vasopressin) to over 100 amino acids (such as the neurotrophic factors). The mature signaling peptide which is secreted is processed from a longer propeptide which contains other domains (prodomains) which are not part of the mature form (Figure 1).\n\nMature peptides of similar structure may function as a neurotransmitter, an endocrine or a paracrine signal within a multicellular organism, and also as a signal between unicellular organisms. For instance, the gonadotropin-releasing hormone (GnRH), which has a significant structural similarity to the yeast mating factor-alpha peptide1, serves as both a hormone in mammals and as a mating pheromone in yeast1. It also serves as a paracrine signal in the periphery of the multicellular organism2.\n\nThus, though the structure of the mature signal of signaling peptides such as GnRH is conserved, its specific signaling role is not, and their prodomains differ markedly. Moreover, in the same organism, structurally similar signaling peptides may regulate a diverse range of signaling pathways, such as the structurally similar oxytocin and vasopressin3, which also function as signals in unicellular organisms4.\n\nAssuming that signals elicit a response because they provide specific information that benefits the organism5,6, how may similar peptides provide information regarding such diverse metabolic processes?\n\nWhile the mature peptides of GnRH, oxytocin and vasopressin are short (9-10 amino acids), their propeptides are large proteins (100-160 amino acids). The cleavage of the propeptide to form the comparatively short mature peptide is often dependent on the completion of post-translational modifications, such as sequential enzymatic modification7, glycosylation, glycosulfation or the pairing of S-S bonds8.\n\nAs the cleavage of the mature peptide depends on the propeptide completing its various post-translational modifications, and as there is a fixed stochastic relationship between the mature peptide and the propeptide that is determined by the number of repeats of the mature domain within the propeptide, it is reasonable to assume that the ability of the signaling cell to complete the synthesis of the propeptide is the information provided by the mature peptide.\n\nWe suggest that while the synthesis and modifications of the propeptide are related to the phenotypic state of the signaling cell, the role of the mature peptide is to stimulate cells to be attentive to this information. In this opinion paper we briefly review a number of signaling peptides to support our suggestion and, in addition, speculate why and how a mature peptide is selected to serve as a stimulating molecule.\n\n\nDifferent propeptides produce similar mature peptides\n\nSimilar signaling peptides are used in different species to affect diverse metabolic processes; however, in many cases these similar peptides are processed from entirely different propeptides. Such is the case of the 10-amino acid GnRH, a hormone produced by the hypothalamus and also by cells in the periphery in vertebrates, which is structurally similar to the mature peptide of the yeast mating-alpha factor1. Also within the yeast genome, a similar mature peptide (mating-alpha factor) is produced by two different propeptides encoded by the genes MFAL-1 and MFAL-2. MFAL-1 has four tandem repeats of the mature domain, while MFAL-2 contains two repeats of the mature domain with a slight variation in sequence (Figure 2).\n\nSmall boxes represent the mature domain.\n\nEven in the same species, the structurally similar forms of the GnRH peptide are connected to different prodomains. There are two genes in humans that encode the sequence of GnRH: GNRH-I and GNRH-II. Each gene has a highly similar mature peptide, but a different prodomain (Figure 3). The synthesis of the mature form (GnRH) requires an intricate series of enzymatic modifications7.\n\nBlue-filled boxes indicate high conservation. Red box contains the mature peptide sequences.\n\nOxytocin and vasopressin are structurally similar peptides of nine amino acids. The mature peptides of oxytocin and vasopressin are highly similar, and they share an accessory protein, neurophysin, yet the other domains have little similarity in sequence (Figure 4).\n\nBlue-filled boxes indicate high conservation. Red box contains the mature peptide sequences.\n\nThe mature peptides of oxytocin and vasopressin are also active in unicellular organisms4. When comparing oxytocin and vasopressin propeptides across the phylogenetic tree, it is evident that the mature domain is more conserved than the prodomain12.\n\nIn addition, within the same species, within a conserved family of signal peptides, such as the neurotrophic factors BDNF, NT-3 and NGF, the variation of the sequence of the mature peptide between the different peptides is significantly lower than the variation among the prodomains of the respective propeptides (Figure 5). The variation in glycosylation sites is depicted schematically in Figure 6. This variation is also evolutionarily conserved: the sequence of the prodomain is unique while the sequence of the mature peptide is common to the family.\n\nBlue-filled boxes indicate high conservation. Red boxes contain the respective mature peptide sequences.\n\nFilled arrows denote cleavage site between prodomain and mature domain, empty arrows denote glycosylation sites.\n\n\nThe phenotypic state of the signaling cell affects the synthesis of the mature peptide\n\nThis variation is not only genetic, but may also be phenotypic. In the case of BDNF, there are several alternative transcripts that produce isoforms of BDNF. These isoforms differ in their prodomain but not in their mature domain, and this variation depends on metabolism and physiological parameters of the signaling cell. The production of specific isoforms of BDNF has been correlated to various pathological conditions in which the synthesis of BDNF is altered13.\n\nBDNF, NGF and NT-3 each have a conserved N-glycosylation site in the prodomain that is proximal to the processing site at which the propeptide is cleaved to form the mature peptide8. When N-glycosylation is blocked, cleavage of the propeptide is affected, and less of the mature form is synthesized. What accumulate in the Golgi are truncated forms of proBDNF8. The secreted mature form of BDNF is therefore a representation of the properly processed and cleaved proBDNF.\n\n\nPropeptide non-signaling functions\n\nWe know of two cases in which a whole functional protein that has a non-signaling role serves as the propeptide of a mature peptide signal. One of them is the sex pheromone system of Enterococcus faecalis14, the other the extracellular death factor (EDF) of Escherichia coli15.\n\nAs far as we are aware, there is no known non-signaling function for the propeptides of the neurotrophic factors. Likewise is the case for oxytocin and vasopressin, and also GnRH. However, cases in which a non-signaling function is known may illustrate how peptide signaling systems evolved.\n\nE. faecalis is a bacterium that has a sophisticated mechanism of plasmid transfer governed by signaling peptides of 7-8 amino acids in length. These peptides are produced from specific membranal proteins that perform non-signaling functions in the cell that are unrelated to the plasmid (Table 1). The mature peptide of the E. faecalis pheromones is part of the sequence of the propeptide which anchors it to the membrane. The cleavage of the propeptide from the membrane releases the mature peptide (the pheromone), which provides via a complex transduction mechanism14 reliable information that the signaling cell does not possess the plasmid.\n\n* Based on similarity\n\nAnother example of a signaling peptide whose propeptide serves a non-signaling function in the cell is the extracellular death factor (EDF) pentapeptide that activates the mazEF pathway in E. coli. The sequence of the pentapeptide is NNWNN, and it is synthesized by the proteolytic cleavage of the enzyme glucose-6-phosphate dehydrogenase17, which includes the sequence NNWDN. The mature peptide requires the modification of the aspartic acid in the propeptide to asparagine.\n\nAs the propeptide of the pentapeptide is a functional protein, the relationship between its synthesis and degradation links the production of the pentapeptide directly to cell metabolism, and specifically the metabolism of glucose. Since the pentapeptide appears in the E. coli genome only within this enzyme, and is an essential component of the enzyme, its secretion from E. coli is reliable information that the signaling bacterium does not need the enzyme for its current metabolism.\n\n\nDiscussion\n\nThe main purpose of this article is to propose a solution to the problem we faced when trying to understand how very similar short peptides may provide information that is relevant to receiver cells designed to serve very different roles. Mature peptides are often conserved across the phylogenetic tree, from unicellular organisms to mammals. Hence, it is tempting to attempt to identify what properties of these mature peptides cause them to be adapted for their role. Our perspective is derived from the assumption that signals provide reliable information regarding the behavior of the signaling cell6,18.\n\nWe suggest that mature peptides were selected as optimal carriers for transferring information due to their ability to stimulate the receiving cell to attend to the information they represent. We speculate that their advantage as stimulating agents is due to their harmful effects which force the receiver to attend to the information (Figure 7).\n\nA A functional protein is degraded to produce short peptides, one of which is toxic and is secreted. The toxic peptide harms neighboring cells by interfering with the cell membrane or diffusing into the cell and interfering with intracellular processes. B Mechanisms evolve that counter the toxicity of the peptide, such as binding proteins on the cell membrane or in the cytoplasm. C The interaction between the secreted toxic peptide and the binding proteins may further evolve into a signaling system, as the secretion of the toxic peptide reflects the phenotypic state of the secreting cell.\n\n\nPredictions\n\nOur suggestion that the mature peptide does not directly represent information regarding the signaling cell's metabolism, but reflects information regarding the synthesis of the prodomain, can be tested. Here we suggest experiments that may test this suggestion:\n\n1. Replacing the mature peptide of one propeptide with a similar mature peptide from the same family of peptides (such as oxytocin with vasopressin), may reveal that the mature peptide has the same effect due to its specificity for its receptor; however, we expect that the timing and quantity of the effect will be determined by the protein to which it is attached.\n\n2. The harmful effects of the mature peptides may be assessed by increasing their quantities beyond the ability of the receiving cells to counter their noxious effects to which they are usually exposed or by removing any other mechanisms that under normal circumstances counter these harmful effects, such as enzymes that degrade the mature peptide or bind it.\n\n\nSome speculations on the stages of the evolution of signal peptides\n\nIt is tempting to speculate how the sequence of mature peptides evolved, even though at present we have succeeded in collecting limited data to support this speculation.\n\nIt is reasonable to assume that the first generation of mature peptides were part of large proteins that had non-signaling functions in the cell. During the proteolysis of these large functional proteins, short peptides were secreted. Among the short peptides cleaved and secreted from cells, the peptides that harmed neighboring cells selected, in neighboring cells, for mechanisms that counter the harmful effects of the peptides. The level of the response to the harmful effects is correlated to the level of the secreted peptide and, hence, could be used by the neighboring cells as a source of information regarding the metabolism of the proteins from which they were cleaved in the signaling cells. These large proteins still serve their initial non-signaling function, yet they also serve as propeptides. These are the cases mentioned previously of the EDF pentapeptide of E. coli which is a functional element of glucose-6-phosphate dehydrogenase16 and the enterococcus pheromones, which anchor the propeptide to the membrane17.\n\nWe suggest that a signal which harms the receiver will force the receiver to respond to a smaller change in its concentration than a signal that provides a positive or neutral effect. Zahavi19 and Zahavi & Zahavi5 found this to be the case for many signals used by birds and humans. In addition, Harris et al.20 pointed out that several non-peptide signals such as glutamate and dopamine may cause harm to cells that do not counter their harmful effects in relation to their level of release. They suggested that the toxicity ensures that the response of the receiver cell is correlated to the level of the release from the signaling cell.\n\nWe also suggest that in the case of signaling peptides, the toxicity of the mature peptide ensures that the response of the receiving cell is correlated to the concentration of the secreted peptide. At present we are aware of only one case in which a peptide has a known direct toxicity that is crucial to its function, the EDF pentapeptide. The pentapeptide kills E. coli by binding to the mazF toxin and interfering with the ability of the mazE antitoxin protein to inhibit the activity of the mazF toxin15.\n\nOnce the sequences of mature peptides evolved as carriers of information due to their ability to stimulate cells to attend to information, natural selection could transpose a mature peptide to be a part of other proteins whose synthesis reflected information regarding the phenotypic state of the signaling cell, granted that this information benefitted the organism through its effect on the receiving cells. Small modifications in the mature peptide of the first generation were required to prevent the binding of the new mature peptide to the receptors of the first generation peptides, if the first generation and the second generation mature peptide were to function in the same organism (Figure 8).\n\nA A function protein (Protein A) contains a toxic signaling peptide (M) which has a complementary receptor (Receptor A) B A mutation causes the toxic signaling peptide M to be transferred into a different functional protein (Protein B), so that both proteins produce the same toxic peptide that interacts with Receptor A. The toxic peptide M no longer provides accurate information regarding the processing of Protein A. C A mutation in the toxic peptide M in Protein B (yielding the toxic peptide M) prevents its binding to the same receptor as M, and a receptor specific to M (Receptor A*) evolves to counter its toxicity, which may provide information on the processing of Protein B.\n\nThe third generation of mature peptides is secreted from central glands and from the brain to activate peripheral cells. Wessler et al.21 suggested that acetylcholine, which serves as a paracrine signal to coordinate activities between epithelial cells in the airpipe, evolved first as a paracrine signal and was only later adopted by the brain to activate peripheral cells that already respond to acetylcholine. Following this suggestion, we recently proposed a general model for the evolution of non-peptide signals20. It is reasonable to assume that signals which function between cells in the periphery, and are also used by the brain and central glands for their similar effect on peripheral cells, evolved first as paracrine signals in the periphery18,21 (i.e., neurons and endocrine cells adopted the mature peptides that served as signals in the periphery to activate cells that were already adapted to respond to them).\n\nIn several cases it is known that the mature peptide by which the brain stimulates peripheral cells is also synthesized in small quantities by the cells that respond to it. Oxytocin and GnRH for example are synthesized in peripheral cells that respond to the same peptides that are secreted in the brain. The present function of the synthesis of the small amounts of mature peptides in peripheral cells is unclear.\n\nIt is reasonable to assume that the function of the propeptides in the brain is not to provide information relating to the metabolism of the secreting neuron, but to allow the synthesis of the mature peptide in large quantities to regulate and synchronize the activities of various organs to respond to decisions made in the brain. Hence, we expect that the structure of the propeptides may differ between the brain and peripheral cells, as each serves a different need: neurons need to synthesize large quantities of the mature peptide, while in the periphery the propeptide reflects the phenotypic state of the signaling cell.\n\n\nHow our perspective changes the focus of treatments for pathological conditions in which the mature peptide is lacking\n\nMany neuropathologies, including amyotrophic lateral sclerosis, Alzheimer's diseases and Parkinson's disease, have been associated with a reduction in the synthesis of neurotrophic factors22–24. Current treatments of these diseases often involve administering synthesized mature peptides (the neurotrophic factors). However, if the inability to process the propeptide correctly (such as the inability to perform the glycosylations) reflects reliably the phenotypic state of the signaling cell, then these treatments provide false information to the neuron. The false information may mask the underlying problem, which may be the deterioration of the signaling cell’s ability to serve its function. If the mature peptide provides retrograde information to the neuron on the ability of the peripheral cell to be activated by the neuron, then supplementing it may mislead the organism that the system is functioning correctly even though it is not. In addition, if, as we suggest, the mature peptide functions due to its harmful effects, then providing an abnormal concentration of it to the receiver cells may cause unnecessary damage to the signaling system. Hence, the attempts to counter the disease should focus on attempting to improve the phenotypic state of the signaling cells, for instance, by providing anti-oxidants25.\n\n\nMethods\n\nAll sequences were taken from the UniProt database9. Alignments were calculated using Clustal Omega10 implemented in UniProt9. Jalview v2.8.211 was used for the graphical representation of the alignments.", "appendix": "Author contributions\n\n\n\nKeith D. Harris collected information from the literature, conducted the bioinformatic work and co-wrote the manuscript. Ari Barzilai advised and provided guidance regarding the content and composition of the manuscript. Amotz Zahavi supervised the research and co-wrote the manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting 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\nAcknowledgements\n\nWe would like to thank Adva Yeheskel from the Bioinformatics Unit, George S. Wise Faculty of Life Sciences, Tel Aviv University, for her assistance with the bioinformatics.\n\n\nReferences\n\nLoumaye E, Thorner J, Catt KJ: Yeast mating pheromone activates mammalian gonadotrophs: evolutionary conservation of a reproductive hormone? Science. 1982; 218(4579): 1323–1325. PubMed Abstract \| Publisher Full Text\n\nCheng CK, Leung PC: Molecular biology of gonadotropin-releasing hormone (GnRH)-I, GnRH-II, and their receptors in humans. Endocr Rev. 2005; 26(2): 283–306. PubMed Abstract \| Publisher Full Text\n\nZimmerman EA, Nilaver G, Hou-Yu A, et al.: Vasopressinergic and oxytocinergic pathways in the central nervous system. Fed Proc. 1984; 43(1): 91–96. PubMed Abstract\n\nCsaba G: The hormonal system of the unicellular Tetrahymena: a review with evolutionary aspects. Acta Microbiol Immunol Hung. 2012; 59(2): 131–156. PubMed Abstract \| Publisher Full Text\n\nZahavi A, Zahavi A: The Handicap Principle: A missing piece of Darwin’s puzzle. Oxford University Press, 1997. Reference Source\n\nZahavi A, Zahavi A: The Logic of Analog Signaling and the Theory of Signal Selection. Isr J Ecol Evol. 2012; 58: 269–278. Reference Source\n\nWetsel WC, Srinivasan S: Pro-GnRH processing. Prog Brain Res. 2002; 141: 221–241. PubMed Abstract \| Publisher Full Text\n\nMowla SJ, Farhadi HF, Pareek S, et al.: Biosynthesis and post-translational processing of the precursor to brain-derived neurotrophic factor. J Biol Chem. 2001; 276(16): 12660–12666. PubMed Abstract \| Publisher Full Text\n\nUniProt Consortium. UniProt: a hub for protein information. Nucleic Acids Res. 2014; 43(Database issue): D204–D212. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcWilliam H, Uludag M, Squizzato S, et al.: Analysis Tool Web Services from the EMBL-EBI. Nucleic Acids Res. 2013; 41(Web Server issue): W597–W600. PubMed Abstract \| Publisher Full Text \| Free 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\nGruber CW, Muttenthaler M: Discovery of defense- and neuropeptides in social ants by genome-mining. PLoS One. 2012; 7(3): e32559. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarlino D, Leone E, Di Cola F, et al.: Low serum truncated-BDNF isoform correlates with higher cognitive impairment in schizophrenia. J Psychiatr Res. 2011; 45(2): 273–279. PubMed Abstract \| Publisher Full Text\n\nChandler JR, Dunny GM: Enterococcal peptide sex pheromones: synthesis and control of biological activity. Peptides. 2004; 25(9): 1377–1388. PubMed Abstract \| Publisher Full Text\n\nBelitsky M, Avshalom H, Erental A, et al.: The Escherichia coli extracellular death factor EDF induces the endoribonucleolytic activities of the toxins MazF and ChpBK. Mol Cell. 2011; 41(6): 625–635. PubMed Abstract \| Publisher Full Text\n\nClewell DB, An FY, Flannagan SE, et al.: Enterococcal sex pheromone precursors are part of signal sequences for surface lipoproteins. Mol Microbiol. 2000; 35(1): 246–247. PubMed Abstract \| Publisher Full Text\n\nKolodkin-Gal I, Engelberg-Kulka H: The extracellular death factor: physiological and genetic factors influencing its production and response in Escherichia coli. J Bacteriol. 2008; 190(9): 3169–3175. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris KD, Zahavi A: The evolution of ACh and GABA as neurotransmitters: A hypothesis. Med Hypotheses. 2013; 81(5): 760–762. PubMed Abstract \| Publisher Full Text\n\nZahavi A: The testing of a bond. Anim Behav. 1977; 25(1): 246–247. Publisher Full Text\n\nHarris KD, Weiss M, Zahavi A: Why are neurotransmitters neurotoxic? An evolutionary perspective [v2; ref status: indexed, http://f1000r.es/4sz]. F1000Res. 2014; 3: 179. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWessler I, Kirkpatrick CJ, Racké K: Non-neuronal acetylcholine, a locally acting molecule, widely distributed in biological systems: expression and function in humans. Pharmacol Ther. 1998; 77(1): 59–79. PubMed Abstract \| Publisher Full Text\n\nCalissano P, Matrone C, Amadoro G: Nerve growth factor as a paradigm of neurotrophins related to Alzheimer's disease. Dev Neurobiol. 2010; 70(5): 372–383. PubMed Abstract \| Publisher Full Text\n\nScalzo P, Kümmer A, Bretas TL, et al.: Serum levels of brain-derived neurotrophic factor correlate with motor impairment in Parkinson’s disease. J Neurol. 2010; 257(4): 540–545. PubMed Abstract \| Publisher Full Text\n\nHarandi VM, Lindquist S, Kolan SS, et al.: Analysis of neurotrophic factors in limb and extraocular muscles of mouse model of amyotrophic lateral sclerosis. PloS One. 2014; 9(10): e109833. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsunemi A, Utsuyama M, Seidler BK, et al.: Age-related decline of brain monoamines in mice is reversed to young level by Japanese herbal medicine. Neurochem Res. 2005; 30(1): 75–81. PubMed Abstract \| Publisher Full Text" }	[ { "id": "10173", "date": "01 Sep 2015", "name": "Uzi Motro", "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\nA stimulating paper, suggesting a new perspective on the nature of signaling peptides. The main idea can be summarized as follows:The mere structure of mature signaling peptides does not necessarily convey the information they carry, but rather their sheer existence, which is due to successful processes undergone by their propeptides, reliably indicates the phenotypic state of the signaling cell. The information given by signaling peptides is heard by the receiving cell because of the toxic nature of the signaling peptides – their toxicity cannot be ignored by the receiving cell.This is a truly novel idea, which seems to be a solid consequence of natural selection. The authors present a few examples, yet – as they point out – sound experimental evidence is necessary. As a theoretician of evolutionary biology (but by no means an expert in cell biology!) I am not in a position to comment or to suggest such an experimentation. If this idea is valid, the consequences to some medical treatments can be substantial, as indicated in the paper's last section.", "responses": [] }, { "id": "10773", "date": "12 Oct 2015", "name": "David Granot", "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 opinion article addresses basic questions about the nature of signals in biological systems. The underlying assumption is the brilliant handicap theory of Amotz Zahavi that signals must have some burden on the signaling cell to be perceived as a reliable signal. The central question that this opinion article is dealing with is how very similar (mature) signaling peptides control different metabolic pathways in different species or even within the same organism. The authors suggest that the message (signal) resides in the propeptides that yield the mature signaling peptides. The amount, location and timing of the very similar (yet not identical) signals provide a reliable and specific information on the capability of the producing cells to synthesize, modify, cleave and secrete the mature signaling peptides. They also suggest that the mature signaling peptides might have toxic effects on the receiving cells that enforce them to respond. The authors also offer an evolutionary prediction of how such signaling peptides might have evolved from propeptides that might have functions which are not necessarily related to the signal. Finally they offer a perspective on the potential impact of their theory on evaluation of pathological treatments, in which the current practice is increasing the amount of the signaling peptides rather than dealing with malfunction of the source of the peptides. As usual, the theories stemming from Zahavi’s and his colleagues (Harris and Barzilai this time) are stimulating, but require validation, as emphasized by the authors themselves. Without validation, the appeal of theories is dependent on how reasonable they are, and if they make (new) sense of disturbing questions. The suggested theories should also be evaluated in comparison with alternative theories. It is not clear if alternative theories exist, and if they do, I would suggest discussing them. The major question raised by the authors is how very similar signaling peptides have different functions? Although very similar, even small signaling peptides are not identical, and it is advised not to exclude the option that the specificity resides within the different amino acids. The examples brought by the authors of cross species effects of similar peptides support the authors’ theory, but can also be explained by the modular nature of biological structures of signaling peptides and receptors. As suggested by the authors, shuffling mature peptides between propeptides may provide an answer. An important part of the author’s hypothesis is that signals might (should?) be toxic to the receiving cell to enforce a response. The only example is the E. coli EDF pentapeptide that has a direct toxicity that is crucial to its function. The biological logic of toxicity to the receiving cells as a way to enforce response is not clear. Toxicity means that the outcome could be harmful, such as killing or deactivating the receiving cells. Do the authors suggest that in such cases the signaling peptide has also a selection function? The authors also discuss a case in which neurons within the brain produce a signaling peptide that affects peripheral cells and suggest that although processed from propeptides in the brain it does not provide information relating to the phenotype of the secreting neuron. It is not clear why these neurons are excluded from the “handicap principle”? Furthermore, dealing with pathological conditions related to the administration of overdose of the signaling peptides (in Alzheimer disease), the authors suggest that it might mask the “inability to process the propeptide correctly by the signaling cell” which could be the cause of the disease. Wouldn’t that mean that production of signaling peptide from the propeptide in neurons should provide information about the phenotype of the secreting neuron? As hinted above, brain stimulation by various scientific hypotheses is usually positive, and may indicate the excellent “phenotypic state” of the current “signaling” group of authors. Yet, I hope there is some “toxicity” that may enforce few readers to respond and test the hypotheses", "responses": [] } ]	1	https://f1000research.com/articles/4-512
https://f1000research.com/articles/4-495/v1	06 Aug 15	{ "type": "Review", "title": "Macrophage Gene Therapy: opening novel therapeutic avenues for immune disorders", "authors": [ "Gyanesh Singh", "U.C. Pachouri", "Chirag Chopra", "Preeti Bajaj", "Pushplata Singh", "U.C. Pachouri", "Chirag Chopra", "Pushplata Singh" ], "abstract": "Macrophages are probably the most important cells of the mammalian immune system, and compromised macrophage function is known to cause several diseases. Their involvement in arthritis, cancer, infections, atherosclerosis, diabetes, and autoimmune disorders is well known. There has been a constantly growing need to transfer therapeutic genes into macrophages. Like most non-macrophage gene therapies, in vitro gene transfer has been attempted much more frequently in case of macrophages. However, primary macrophages are still somewhat recalcitrant to transfection. Macrophage-specific synthetic promoters, which were recently used successfully, can have up to 100-fold higher activity than that of native promoters. Adenovirus, lentivirus, and adeno-associated virus are commonly used for macrophage gene therapy. A number of non-viral methods are also popular for the transfer of exogenous DNA into macrophages. Gene transfer to macrophages using naked DNA has also been successful in a few cases. Macrophages have speciﬁc mechanisms to recognize and respond to bacterial DNA because of the presence of unmethylated CpG dinucleotides, which are rare in eukaryotic DNA. With interesting developments in this area, macrophage gene therapy appears to have great potential for immune therapies.", "keywords": [ "Macrophage", "Gene Therapy", "Immunotherapy" ], "content": "Introduction\n\nMacrophages are known to be involved in the development of various diseases, not only because of their key role in host immunity, but also because of their ability to act as a host and reservoir for certain pathogens1,2. Their involvement in rheumatoid arthritis, tumorigenesis, AIDS, atherosclerosis, diabetes, and lupus erythematosus has already been well documented (3–6; Figure 1). Since, like other primary cells, primary macrophages also resist transfection by contemporary methods, efficient transfection methods are required7. A recent breakthrough in macrophage gene therapy is the successful use of macrophage-specific synthetic promoters. These promoters were made by random combination of macrophage and myeloid cis elements8. This promoter was found to have up to 100-fold higher activity than that of a native macrophage promoter with minimal activity in nonmyeloid cells. When this promoter was used with GFP, blood leukocytes showed stable and high levels of CD11b+ specific GFP expression for more than a year. Similarly, hypoxia-response elements (HREs) are being utilized for delivery of therapeutic genes specifically to hypoxic tumour areas9. Self-inactivating retroviral vectors that had regulatory cis elements from the macrophage-specific human CD68 gene, successfully directed cell-specific expression of reporter genes in immune cells in vivo10. Furthermore, a number of gene therapy trials, where macrophage granulocyte-macrophage colony-stimulating factor (GM-CSF) expression was altered, showed promising results11.\n\n\nViral gene therapy\n\nViruses including adenovirus, lentivirus, and adeno-associated virus are commonly used to transfect macrophages12. However, a number of non-viral methods have also been used13. The biggest disadvantage of non-viral methods is that exogenous DNA entering the macrophage cells by endocytosis is often degraded by nucleases. On the other hand, viral methods generally give higher transfection efficiency and sustained transgene expression14. Retroviruses (other than the lentivirus sub-family) are generally ineffective in the transfection of primary macrophages as these cells are of non-dividing nature15. Therefore, macrophages are often generated from retrovirus-infected macrophage precursors that divide more frequently16. Convenient transfection of macrophage precursors offers a greater advantage of re-introduction of transfected cells into a suitable host. An encouraging macrophage-infection rate of 90% was seen, when a nuclear localization signal (NLS) sequence was introduced into the matrix protein of the C type retrovirus17. A common problem with retroviruses is that transgene expression from integrated retroviruses is lost after a short duration, possibly due to transcriptional silencing.\n\nLentiviruses can normally infect non-dividing cells including monocytes. Viral matrix proteins, Vpr, integrase, and pol genes have been shown to be important for the transduction of non-dividing cells in this case18. Lentiviral vectors might also be less susceptible to transcriptional silencing. Adenoviruses can infect non-dividing cells with high efficiency but with moderate stability. 10–80% transfection efficiencies of human macrophages have been reported using adenoviruses19. Interestingly, adenoviruses are less capable of infecting monocytes. However, more than 90% infection rate was reported, where primary human monocytes were pre-treated with macrophage-colony stimulating factor8. Despite a number of advantages, the major drawback of lentiviral vectors (particularly of human lentiviruses) is the extreme safety concerns over its use8. Adeno-associated viruses (AAV) offer some advantages similar to retroviruses; they can insert themself into the host cell chromosome for stable expression of transgenes, and, they can also infect both proliferating and non-proliferating cells20. One of the common drawbacks of using AAV or retroviral vectors for gene therapy is that they offer low size limits (4–8 kb). On the other hand, poxvirus and herpes simplex virus, that have also been used to infect macrophages, have much higher capacity. A major hurdle in viral gene therapy is immune responses generated by viral vectors especially in the case of adenoviruses and AAV. The common solution for these problems is to use a viral vector having a truncated genome and therefore produce fewer viral proteins that can be immunogenic21,22. Importantly, most of the viruses used so far infect a broad range of cell types, and are not highly specific for macrophages. Altering the cell tropism of viruses to enhance specificity has been an important area of research recently23. Some of the interesting reports, where an artificial virus cell-binding receptor was engineered, have shown promising results24. Also, genetic manipulation of the genes coding for the adenoviral fiber protein was helpful to gain higher specificity.\n\n\nNon-viral gene therapy\n\nBoth chemical and physical methods including liposomes, lipoplexes, DNA carriers, diethylaminoethyl (DEAE)-dextran, microinjection, and electroporation have been used with variable success rates25. Some of the drawbacks of non-viral methods are low transfection efficiency and transient transgene expression. The other common problems are: limited systemic application, rapid clearing from the circulation, and transfection of non-target tissues26–28. This can be a problem during gene therapy. Fortunately, this problem can easily be avoided by treating the transfection-ready DNA with a methylases. Although electroporation has been shown to achieve moderate levels of transfection efficiency for human monocytic cell lines and monocyte-derived macrophages, the main disadvantage in this case is increased cell death29. Improved transfection reagents offer benefits like low toxicity, large DNA delivery, and lack of immunogenicity30. However, poor transfection efficiencies were seen in the case of primary macrophages19. LipofectAMINE along with protamine sulfate was found to be the most effective in macrophage cell line RAW 264.7, followed by Lipofectin, DOTAP, and DEAE-dextran31. Cationic liposome/DNA complexes have also been used for transfecting monocytes/macrophages in vivo in the blood, liver, and spleen32. However, a number of non-target cells were transfected in these cases. Several transfection methods have been developed to target specific macrophage cell surface receptors33. Ligands such as mannose and transferrin have been incorporated into gene transfer vehicles to increase the efficacy of transfection for macrophages. Some of the intracellular microbes that infect macrophage cells were also used as a method of transferring DNA constructs. An attenuated derivative of Salmonella typhimurium containing interferon-expressing plasmid has been used for macrophage-specific interferon expression34. Similarly, attenuated mutants of leishmania have been developed that can safely be used for macrophage gene therapy35,36.\n\n\nConclusion\n\nMacrophages are of central importance in immune system, and also have important physiological roles in major organs including brain and liver. Their role in variety of immune, physiological, and pathogen-centric disorders have been repetitively validated. More and more diseases are showing associations with compromised or exaggerated macrophage functions. With the recent developments in the gene therapy, macropahges have become an attractive target to prevent or cure major diseases including cancer, range of blood-vascular disorders, diabetes, and various immune disorders.", "appendix": "Author contributions\n\n\n\nGS and PB conceived the study and prepared the first draft of the manuscript.\n\nUCP and CC did the analysis of the literature. PS did cross-checking and referencing.\n\n\nCompeting 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\nAcknowledgments\n\nThe author is thankful to Dr. Pawan Sharma for their valuable guidance.\n\n\nReferences\n\nCao Q, Zheng D, Wang YP, et al.: Macrophages and dendritic cells for treating kidney disease. Nephron Exp Nephrol. 2011; 117(3): e47–52. PubMed Abstract \| Publisher Full Text\n\nChávez-Galán L, Olleros ML, Vesin D, et al.: Much More than M1 and M2 Macrophages, There are also CD169+ and TCR+ Macrophages. Front Immunol. 2015; 6: 263. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi J, Hsu HC, Mountz JD: Managing macrophages in rheumatoid arthritis by reform or removal. Curr Rheumatol Rep. 2012; 14(5): 445–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArango Duque G, Descoteaux A: Macrophage cytokines: involvement in immunity and infectious diseases. Front Immunol. 2014; 5: 491. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcNelis JC, Olefsky JM: Macrophages, immunity, and metabolic disease. Immunity. 2014; 41(1): 36–48. PubMed Abstract \| Publisher Full Text\n\nWynn TA, Chawla A, Pollard JW: Macrophage biology in development, homeostasis and disease. Nature. 2013; 496(7446): 445–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBooth C, Carmo M, Gaspar HB: Gene therapy for haemophagocytic lymphohistiocytosis. Curr Gene Ther. 2014; 14(6): 437–46. PubMed Abstract \| Publisher Full Text\n\nLevin MC, Lidberg U, Jirholt P, et al.: Evaluation of macrophage-specific promoters using lentiviral delivery in mice. Gene Ther. 2012; 19(11): 1041–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurke B, Giannoudis A, Corke KP, et al.: Hypoxia-induced gene expression in human macrophages: implications for ischemic tissues and hypoxia-regulated gene therapy. Am J Pathol. 2003; 163(4): 1233–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu JS, Burwick JA, Dranoff G, et al.: Gene therapy for metastatic brain tumors by vaccination with granulocyte-macrophage colony-stimulating factor-transduced tumor cells. Hum Gene Ther. 1997; 8(9): 1065–72. PubMed Abstract \| Publisher Full Text\n\nvan de Laar L, Coffer PJ, Woltman AM: Regulation of dendritic cell development by GM-CSF: molecular control and implications for immune homeostasis and therapy. Blood. 2012; 119(15): 3383–93. PubMed Abstract \| Publisher Full Text\n\nMiyake K, Suzuki N, Matsuoka H, et al.: Stable integration of human immunodeficiency virus-based retroviral vectors into the chromosomes of nondividing cells. Hum Gene Ther. 1998; 9(4): 467–75. PubMed Abstract \| Publisher Full Text\n\nSchwendener RA: Liposomes in biology and medicine. Adv Exp Med Biol. 2007; 620: 117–28. PubMed Abstract \| Publisher Full Text\n\nO'Connor DM, Boulis NM: Gene therapy for neurodegenerative diseases. Trends Mol Med. 2015; pii: S1471-4914(15)00118-5. PubMed Abstract \| Publisher Full Text\n\nGough PJ, Raines EW: Gene therapy of apolipoprotein E-deficient mice using a novel macrophage-specific retroviral vector. Blood. 2003; 101(2): 485–91. PubMed Abstract \| Publisher Full Text\n\nGarbitt RA, Bone KR, Parent LJ: Insertion of a classical nuclear import signal into the matrix domain of the Rous sarcoma virus Gag protein interferes with virus replication. J Virol. 2004; 78(24): 13534–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrenstein JM, Meltzer MS, Phipps T, et al.: Cytoplasmic assembly and accumulation of human immunodeficiency virus types 1 and 2 in recombinant human colony-stimulating factor-1-treated human monocytes: an ultrastructural study. J Virol. 1988; 62(8): 2578–2586. PubMed Abstract \| Free Full Text\n\nKalter DC, Nakamura M, Turpin JA, et al.: Enhanced HIV replication in macrophage colony-stimulating factor-treated monocytes. J Immunol. 1991; 146(1): 298–306. PubMed Abstract\n\nZhang X, Edwards JP, Mosser DM: The expression of exogenous genes in macrophages: obstacles and opportunities. Methods Mol Biol. 2009; 531: 123–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZaiss AK, Liu Q, Bowen GP, et al.: Differential activation of innate immune responses by adenovirus and adeno-associated virus vectors. J Virol. 2002; 76(9): 4580–4590. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDuffy MR, Parker AL, Bradshaw AC, et al.: Manipulation of adenovirus interactions with host factors for gene therapy applications. Nanomedicine (Lond). 2012; 7(2): 271–88. PubMed Abstract \| Publisher Full Text\n\nKhare R, Chen CY, Weaver EA, et al.: Advances and future challenges in adenoviral vector pharmacology and targeting. Curr Gene Ther. 2011; 11(4): 241–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerkol T, Perales JC, Mularo F, et al.: Receptor-mediated gene transfer into macrophages. Proc Natl Acad Sci U S A. 1996; 93(1): 101–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKawakami S, Sato A, Nishikawa M, et al.: Mannose receptor-mediated gene transfer into macrophages using novel mannosylated cationic liposomes. Gene Ther. 2000; 7(4): 292–9. PubMed Abstract \| Publisher Full Text\n\nBurke B, Sumner S, Maitland N, et al.: Macrophages in gene therapy: cellular delivery vehicles and in vivo targets. J Leukoc Biol. 2002; 72(3): 417–428. PubMed Abstract\n\nMehier-Humberta S, Guya RH: Physical methods for gene transfer: improving the kinetics of gene delivery into cells. Adv Drug Deliv Rev. 2005; 57(5): 733–753. PubMed Abstract \| Publisher Full Text\n\nLaube BL: Aerosolized Medications for Gene and Peptide Therapy. Respir Care. 2015; 60(6): 806–24. PubMed Abstract \| Publisher Full Text\n\nJain S, Tran TH, Amiji M: Macrophage repolarization with targeted alginate nanoparticles containing IL-10 plasmid DNA for the treatment of experimental arthritis. Biomaterials. 2015; 61: 162–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeller LC, Heller R: In vivo electroporation for gene therapy. Hum Gene Ther. 2006; 17(9): 890–897. PubMed Abstract \| Publisher Full Text\n\nXu H, Li Z, Si J: Nanocarriers in gene therapy: a review. J Biomed Nanotechnol. 2014; 10(12): 3483–507. PubMed Abstract \| Publisher Full Text\n\nKuo JH, Jan MS, Chiu HW: Mechanism of cell death induced by cationic dendrimers in RAW 264.7 murine macrophage-like cells. J Pharm Pharmacol. 2005; 57(4): 489–495. PubMed Abstract \| Publisher Full Text\n\nHattori Y, Kawakami S, Nakamura K, et al.: Efficient gene transfer into macrophages and dendritic cells by in vivo gene delivery with mannosylated lipoplex via the intraperitoneal route. J Pharmacol Exp Ther. 2006; 318(2): 828–34. PubMed Abstract \| Publisher Full Text\n\nFioravanti J, Medina-Echeverz J, Berraondo P: Scavenger receptor class B, type I: a promising immunotherapy target. Immunotherapy. 2011; 3(3): 395–406. PubMed Abstract \| Publisher Full Text\n\nLoessner H, Weiss S: Bacteria-mediated DNA transfer in gene therapy and vaccination. Expert Opin Biol Ther. 2004; 4(2): 157–68. PubMed Abstract \| Publisher Full Text\n\nSheel M, Engwerda CR: The diverse roles of monocytes in inflammation caused by protozoan parasitic diseases. Trends Parasitol. 2012; 28(10): 408–16. PubMed Abstract \| Publisher Full Text\n\nde Menezes JP, Guedes CE, Petersen AL, et al.: Advances in Development of New Treatment for Leishmaniasis. Biomed Res Int. 2015; 2015: 1–11, 815023. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "10724", "date": "22 Oct 2015", "name": "Natalia lago", "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\nThe present article attempts to review a very important topic in the immunology and gene therapy fields regarding the available tools to target and modulate macrophages using gene therapy. The field of gene therapy has evolved tremendously in the last 10-15 years, but the attempt to target macrophage still remains challenging. However, this reviewer has some major concerns about the article: A review should conform a critical view of the classical information and the latest developments, carefully cited to acknowledge the original authors of the experiments, and importantly, provide with a deep analysis and sum-ups with the personal interpretation or insights of the authors. It should suggest the future directions in the field and possible experiments or models to be used for a significant further advance. This review does not fulfill these requirements. The presented text is not well organized in a focused vision, which makes it difficult to go through. Moreover, the text suffers from a lack of deep insights in the gene therapy field. Carefully designed figures and tables should be presented to contribute to the establishment of a central idea of the article. For instance, the reviewer feels that a table summarizing the viral and non-viral vectors that have been used for targeting macrophages and their corresponding references would help to a better understanding of the article. Moreover, the only figure present in this review does not involve gene therapy and has no figure legend. Authors should be careful with the bibliography. There seems to be a mismatch between the citations in the text and the list of references. Finally, and most importantly, the reviewer feels that authors do not have experience in this field to be able to write this type of review. In conclusion, the reviewer does not approve this article.", "responses": [] }, { "id": "11167", "date": "11 Nov 2015", "name": "Zhuling Guo", "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 article deals with a novel approach to immune disorders via the macrophage gene therapy. I personally think that it has reached to the threshold level of a short review. It is relatively easy to gain the main points that the authors try to summarize here. Therefore, a brief comment attached for further modification. It is better to elaborate in-depth to support the current topic, or at least make your focal claims stand out. Another concern is that the authors are expected to point out some experimental directions or perspectives in this field to facilitate the next original works. Lastly, there are numerous abbreviations should be presented or listed throughout the text.", "responses": [] } ]	1	https://f1000research.com/articles/4-495
https://f1000research.com/articles/4-182/v1	06 Jul 15	{ "type": "Review", "title": "Understanding the direct and indirect costs of patients with schizophrenia", "authors": [ "Kazuhiro Tajima-Pozo", "María Jesús de Castro Oller ", "Adrian Lewczuk", "Francisco Montañes-Rada", "María Jesús de Castro Oller ", "Adrian Lewczuk", "Francisco Montañes-Rada" ], "abstract": "Background: Schizophrenia is a disabling mental disorder with high prevalence and that usually requires long-term follow-up and expensive lifelong treatment. The cost of schizophrenia treatment consumes a significant amount of the health services' budget in western countries.Objective: The aim of the study was to find out about the costs related to schizophrenia across different european countries and compare them.Results: Schizophrenia treatment costs an estimated 18 billion euros annually worldwide. The direct costs associated with medical help are only part of the total expenditure. The indirect costs are an equally (or even more)important part of the total cost. These expenses are related to the lack of productivity of schizophrenic patients and the cost that relatives have to bear as a result of taking care of their affected relatives.Conclusions: Although data on the cost of schizophrenia may vary slightly between different european countries, the general conclusion that can be drawn is that schizophrenia is a very costly disorder. Not only because of direct costs related to medical procedures, but also due to the non-medical (indirect) costs. Together this suggests the need to investigate cost-efficient strategies that could provide a better outcome for schizophrenic patients, as well as the people who care for them.", "keywords": [ "Schizophrenia costs", "Europe", "indirect and direct costs", "disability", "antipsychotic" ], "content": "Introduction\n\nSchizophrenia is a chronic psychiatric condition that affects around 1% of the population worldwide1–4. It is one of the most stigmatizing diseases of all time3,5. Schizophrenic disorders present with a wide range of symptoms, both positive and negative, leading to cognitive, social and functional impairment. Therefore most schizophrenic patients are considered disabled and claim benefits6. The debilitating nature of the disorder means that patients receive lifelong treatment, and a large proportion of them need to be admitted to a hospital inpatient unit on multiple occasions throughout their lifetime1–3.\n\nAll of the above factors lead to the high costs associated not only with the treatment of schizophrenia but also related to social impairment3,7. By these we mean the inability to work and also the way that schizophrenia affects the patients´ environment. The need to take care of and support schizophrenic relatives is a major reason for members of their families to take sick leave or even sacrifice their own career3,8. All of this, plus the high expense associated with the newest antipsychotic drugs, makes the costs of schizophrenia management as high as 3% of the total healthcare budget of western countries1.\n\nThe aim of this paper is to give a global view on the problem and to emphasize certain cost-inducing aspects of schizophrenia management by reviewing past research on the costs of schizophrenia management. Only some of the data are comparable, e.g. annual expenses on schizophrenia treatment, expenses per capita on schizophrenia treatment, while we can not compare the other data we found in such a manner. Nevertheless we decided to include them to give the reader a broader view on the matter. As such, taking into account these possible and obvious differences concerning data collecting methods used by the particular researchers, as well as different years in which the researchers conducted their studies, the comparison can only estimate the true cost of schizophrenia treatment.\n\nRegardless of the authors’ origin, they all agree unanimously that the costs associated with the treatment and care of schizophrenic patients can be divided into two important groups: direct and indirect costs2–4,8,9. The direct costs of treating schizophrenia include cost of hospitalization (short- and long-term), outpatient follow-up, residential and day care, pharmaceutical interventions, laboratory testing and social security payments, whereas the indirect costs are mainly related to the loss of productivity1,3,10,11. The age of onset of the disorder, usually in the late teens or early 20s, can preclude patients from even starting to work12. Later on, most of the patients receive benefits for incapacity for work due to disability6. Nowadays, most schizophrenic patients receive a disability certificate and eventually do not work. Up to 80% of schizophrenic patients in the UK do not have paid employment3. In Italy and Spain, three out of four patients with schizophrenia are excluded from the job market9,13. Thus, some authors consider the loss of productivity as accounting for the majority of the indirect costs4.\n\nSpeaking of indirect costs, it is also important to consider the indirect cost associated with caregivers to schizophrenic patients, who contribute with their time and in-kind services (Table 1). Therein lies the issue. The real number of people affected by schizophrenia is much bigger than just the number of the patients. According to some authors, direct and indirect costs are approximately equal14, whereas others suggest that indirect costs can outnumber the direct ones up to three or four times11,15.\n\nSome authors also distinguish a third group of costs, called “intangible costs”3. These are expenses of a non-financial nature. They try to accomplish the hard task of reflecting the patients´ quality of life, including side-effects of pharmaceutical interventions and stress and anxiety, both caused by the disease itself and also the treatment process. Although this group might have no direct financial impact, it is worth considering these factors as they probably affect the cooperation between patient and health providers16,17. Furthermore, we can also take into account the intangible costs of the caregivers of schizophrenic patients3. Daily care of a schizophrenic relative can be a very challenging and exhausting experience3,18. Social stigma and the lack of sympathy and understanding may lead to anxiety and depression in caregivers as well as sufferers5. This could damage the relationship between the caregiver and the caretaker, which may lead to an increased rate of patient deterioration and worse prognosis in the long-term18.\n\nThe main part of data used in this article come from psychiatric wards in Spain, France, Sweden, Poland, United Kingdom and Ukraine (Table 2, Table 3 and Table 4). Data from recent USA research have also been included for comparison purposes. Although some of the methods used for data collection vary, depending on the country and researchers, the general idea of this article is to get a global view on the subject. Thus, some estimations can be made and their legitimacy is consistent as shown by the similarities between the results.\n\nDirect, indirect and total cost are defined by authors across different articles with similar criteria which have been discussed in the Introduction section. Total cost is defined as a sum of the direct and indirect cost. The table also contains the proportion of the cost of pharmaceutical treatment (Drug cost) in relation to the total cost.\n\nThe data coming from these two eastern European countries only includes direct costs (total cost=total direct cost) 50 patients were included in the search in Poland and 58 patients in Ukraine. (Tomasz Zaprutko et al., 2014; data comes from years 2010–2011)4.\n\nData from a study that lasted 5 years4. Mean values of indirect, direct and total costs per patient. Percentage of drugs participation in the total cost is included.\n\n\nMethods\n\nThe research data for this article was collected by the use of the PubMed database in April 2015, having used key words: “schizophrenia costs in Europe”, “(indirect and direct costs) schizophrenia”, “schizophrenia costs worldwide”, “schizophrenia costs United States”, “schizophrenia and disability”, “antipsychotic treatment in Europe” as a part of the abstract, title or included anywhere in the whole paper. We analysed 41 articles that we managed to find according to the criteria we adopted.\n\nWe wanted to include only the latest data coming from research conducted after year 2000. Due to the fact that there has been very little research done in this field in general we decided to analyse papers from the whole PubMed dataset. We did however exclude the earliest data (Australia 1976; USA 1975, 1985; Netherlands 1989) since we found them irrelevant (for example, at that time second-generation antipsychotic drugs were not used and they account for a significant part of medication costs).\n\nThe nature of data varies between different authors. Some articles give total amounts of money in relation to the direct and indirect schizophrenia treatment costs in a particular country. The others give numbers per patient. The percentage approximations of pharmacological costs are also present, directly obtained from papers.\n\nWe included only those articles showing general information about costs of schizophrenia across different countries, and excluded those ones related only to specific services, like acute inpatient units, where stated costs did not include rehabilitation expenses or full treatment options, what could lead to important bias when comparing pharmacological expenses or indirect costs.\n\n\nResults and discussion\n\nThis paper points at the magnitude of the problem of schizophrenia treatment costs and estimates the huge impact that this mental disorder has on patients´ environment and society at a financial level across several different countries3,7,16.\n\nWhat definitely strikes attention when it comes to available data is that the estimated indirect cost represents a significant part of the total cost of schizophrenia. It is particularly important to bear that in mind in order to manage schizophrenia efficiently.\n\nThe other conclusion that can be drawn from our research is that the cost of pharmaceutical treatment doesn’t contribute significantly to the total cost of treatment7,16. This statement is equally consistent for both Western (Table 2 shows percentages which range from 4% in UK to 16,1% in France; Table 4 shows the percentage in Sweden – 3,8%) and Eastern European countries such as Poland and Ukraine (Table 3). These results may suggest important differences on the cost of non-pharmaceutical care provided across countries1, which need further investigation.\n\nCollected data across different studies vary in terms of the number of patients included or hospitals involved across different investigations. An exact comparison between all papers is obviously not possible, but certain estimations can be made.\n\nIn some countries, like France and UK, the indirect costs outnumber the direct ones (Table 2)4,15. On the other hand, results coming from Spain and USA suggest that both types of cost are equally important in total12,14. To better understand this result we need to consider the differences across particular studies regarding design and methods used. For instance, the approach to calculate the cost of lost workforce varies between the countries, which could lead to some of the differences observed. In Spain, the official registries do not reflect the work force lost by people who have never even started a job – and this group of people accounts for a large proportion of schizophrenic patients since the onset of the disorder (and so the problems with getting or maintaining employment) may occur early in life, before young people begin their career14.\n\nTherefore, it could be possible that the indirect costs are even higher in the countries who register its workforce in this way.\n\nMany authors raise the issue of patients´ adherence to prescribed therapy. An optimal control of schizophrenic symptoms is proven to lead to fewer number of hospitalizations and less need to use other approaches of formally organized patient care19. Although good adherence to treatment means no reduction in the budget for pharmaceutical interventions, this cost does not seem to represent a significant percentage when compared to total cost of the illness (drugs costs vs. total costs – see Table 2, Table 3 and Table 4). Good psychoeducational programs and building insight about the disease and its management makes patients feel more secure and thus more cooperative1,3.\n\nWhen the symptoms of schizophrenia remain under control, patients experience a smaller risk of mental impairment and social exclusion which extends to participation in work opportunities20. So it is of crucial importance to a investigate whether better control of symptoms could allow patients to get a chance to attain and keep a job, and how this fact could affect their quality of life. Moreover, we should consider how this could be used to lighten the burden of care that relatives and caregivers experience and thus reduce the indirect costs which make up a significant part of the costs of this illness.\n\nTherefore, it is our suggestion that the future of schizophrenia treatment should address more carefully important elements of the financial aspects of the disease, such as cost-efficacy of treatment, including psychological therapies and psychoeducational approaches for both patients and their families. A wider view on the matter is needed.", "appendix": "Author contributions\n\n\n\nThe manuscript was written by Adrian Lewczuk and Dr. Tajima-Pozo. Dr. Castro and Dr. Montañes-Rada contributed to analysis of the data and identification of suitable references. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declared no competing interests.\n\n\nGrant information\n\nNo funding was involved in supporting this research.\n\n\nReferences\n\nKnapp M: Schizophrenia costs and treatment cost-effectiveness. Acta Psychiatr Scand Suppl. 2000; 407: 15–8. PubMed Abstract \| Publisher Full Text\n\nSarlon E, Heider D, Millier A, et al.: A prospective study of health care resource utilisation and selected costs of schizophrenia in France. BMC Health Serv Res. 2012; 12: 269–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKnapp M: Costs of schizophrenia. Br J Psychiatry. 1997; 171(6): 509–18. PubMed Abstract \| Publisher Full Text\n\nLindström E, Eberhard J, Neovius M, et al.: Costs of schizophrenia during 5 years. Acta Psychiatr Scand Suppl. 2007; 116(435): 33–40. PubMed Abstract \| Publisher Full Text\n\nPhelan JC, Bromet EJ, Link BG: Psychiatric illness and family stigma. Schizophr Bull. 1998; 24(1): 115–26. PubMed Abstract \| Publisher Full Text\n\nJonsson U, Alexanderson K, Kjeldgård L, et al.: Psychiatric diagnoses and risk of suicidal behaviour in young disability pensioners: prospective cohort studies of all 19-23 year olds in Sweden in 1995, 2000, and 2005, respectively. PLoS One. 2014; 9(11): e111618. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSalize HJ, McCabe R, Bullenkamp J, et al.: Cost of treatment of schizophrenia in six European countries. Schizophr Res. 2009; 111(1–3): 70–7. PubMed Abstract \| Publisher Full Text\n\nBecker T, Knapp M, Knudsen HC, et al.: The EPSILON study of schizophrenia in five European countries. Design and methodology for standardising outcome measures and comparing patterns of care and service costs. Br J Psychiatry. 1999; 175: 514–21. PubMed Abstract \| Publisher Full Text\n\nOliva-Moreno J, López-Bastida J, Osuna-Guerrero R, et al.: The costs of schizophrenia in Spain. Eur J Health Econ. 2006; 7(3): 182–88. PubMed Abstract \| Publisher Full Text\n\nVázquez-Polo FJ, Negrín M, Cabasés JM, et al.: An analysis of the costs of treating schizophrenia in Spain: a hierarchical Bayesian approach. J Ment Health Policy Econ. 2005; 8(3): 153–65. PubMed Abstract\n\nDavies LM, Drummond MF: Economics and schizophrenia: the real cost. Br J Psychiatry Suppl. 1994; 25: 18–21. PubMed Abstract\n\nMcEvoy JP: The costs of schizophrenia. J Clin Psychiatry. 2007; 68(Suppl 14): 4–7. PubMed Abstract\n\nGarattini L, Barbui C, Clemente R, et al.: Direct costs of schizophrenia and related disorders in Italian community mental health services: a multicenter, prospective 1-year followup study. Schizophr Bull. 2004; 30(2): 295–302. PubMed Abstract\n\nRupp A, Keith SJ: The costs of schizophrenia. Assessing the burden. Psychiatr Clin North Am. 1993; 16(2): 413–23. PubMed Abstract\n\nAndrews G, Sanderson K, Corry J, et al.: Cost-effectiveness of current and optimal treatment for schizophrenia. Br J Psychiatry. 2003; 183: 427–35; discussion 436. PubMed Abstract \| Publisher Full Text\n\nKnapp M, Mangalore R, Simon J: The global costs of schizophrenia. Schizophr Bull. 2004; 30(2): 279–93. PubMed Abstract \| Publisher Full Text\n\nBalak N, Elmaci I: Costs of disorders of the brain in Europe. Eur J Neurol. 2007; 14(2): e9. PubMed Abstract \| Publisher Full Text\n\nMari JJ, Streiner D: The effects of family intervention on those with schizophrenia. Schizophrenia module, Cochrane Database of Systematic Reviews. 1996.\n\nLindström E, Bingefors K: Patient compliance with drug therapy in schizophrenia. Economic and clinical issues. Pharmacoeconomics. 2000; 18(2): 106–24. PubMed Abstract \| Publisher Full Text\n\nZaprutko T, Kus K, Bilobryvka R, et al.: Schizophrenia and Employment: Evaluation From Professionals Point of View. Psychiatr Q. 2015. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9406", "date": "16 Jul 2015", "name": "Lucas Giner", "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\nBrief and nice review of an important topic related to economic cost of schizophrenia. Results are presented clear. However, I think that tables with examples are not necessary and should be considered to be deleted. The text does not need any change in my consideration.", "responses": [ { "c_id": "1465", "date": "16 Jul 2015", "name": "Kazuhiro Tajima-Pozo", "role": "Author Response", "response": "Thanks for your quick review and your comments." }, { "c_id": "1474", "date": "06 Aug 2015", "name": "Kazuhiro Tajima-Pozo", "role": "Author Response", "response": "According to the reviewer comments, we have deleted table 4, and we have maintained table 1, 2 and 3. I hope that this changes could contribute to the final acceptance of the paper." } ] }, { "id": "9352", "date": "03 Aug 2015", "name": "Mayumi Okuda", "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 article offers a concise, brief summary of an important topic. The following are minor revisions that will improve the manuscriptChanging the terminology \"schizophrenic patients\" to \"patients with schizophrenia\" or \"individuals with schizophrenia\". In the abstract the objective states the aim is to report on the costs of schizophrenia across different European countries, yet the results states \"18 billion euros worldwide.\" Given that the manuscript reviews economic studies in Europe, it should be modified to focus on European findings. The last few sentences in paragraph 3 of the introduction describe limitations of this review; these sentences should be described in the methods section. It is not completely clear to me what the authors are referring to when they mention \"other data\" that could not be compared. A paragraph following the introduction or within the introduction detailing the types of costs that are typically measured in these type of studies could replace table 1. Detailing the types of costs that were measured in the different studies would be a valuable contribution to Table 2.", "responses": [ { "c_id": "1490", "date": "06 Aug 2015", "name": "Kazuhiro Tajima-Pozo", "role": "Author Response", "response": "Thanks for your review, we have made the following changes in our latest versionChange the terminology \"schizophrenic patients\" to \"patients with schizophrenia\" or \"individuals with schizophrenia\". The last few sentences in paragraph 3 have been moved to the methods section. A brief paragraph following the introduction of table 1 have been included. A brief paragraph following the introduction of table 2 have been included.Best regards and thanks for your quick and good review of our paper" } ] } ]	1	https://f1000research.com/articles/4-182
https://f1000research.com/articles/4-478/v1	05 Aug 15	{ "type": "Software Tool Article", "title": "CyREST: Turbocharging Cytoscape Access for External Tools via a RESTful API", "authors": [ "Keiichiro Ono", "Tanja Muetze", "Georgi Kolishovski", "Paul Shannon", "Barry Demchak", "Keiichiro Ono", "Tanja Muetze", "Georgi Kolishovski", "Paul Shannon" ], "abstract": "As bioinformatic workflows become increasingly complex and involve multiple specialized tools, so does the difficulty of reliably reproducing those workflows. Cytoscape is a critical workflow component for executing network visualization, analysis, and publishing tasks, but it can be operated only manually via a point-and-click user interface. Consequently, Cytoscape-oriented tasks are laborious and often error prone, especially with multistep protocols involving many networks.In this paper, we present the new cyREST Cytoscape app and accompanying harmonization libraries. Together, they improve workflow reproducibility and researcher productivity by enabling popular languages (e.g., Python and R, JavaScript, and C#) and tools (e.g., IPython/Jupyter Notebook and RStudio) to directly define and query networks, and perform network analysis, layouts and renderings. We describe cyREST’s API and overall construction, and present Python- and R-based examples that illustrate how Cytoscape can be integrated into large scale data analysis pipelines.cyREST is available in the Cytoscape app store (http://apps.cytoscape.org) where it has been downloaded over 1900 times since its release in late 2014.", "keywords": [ "Workflow", "Reproducibility", "Cytoscape", "Service Oriented Architecture", "Interoperability", "REST", "Resource Oriented Development", "Microservice" ], "content": "Introduction\n\nBecause of its robust network analysis and visualization capabilities coupled with its vibrant user and developer community, Cytoscape 3 has become a tool of choice for studying large network-oriented omics data sets on common workstations and for publishing results. However, even as Cytoscape1 is well positioned to handle customized omics workflows, bioinformaticians’ need to quickly and efficiently create complex, varied, and repeatable workflows exceeds the capabilities of Cytoscape’s existing automation features. At the same time, bioinformaticians have embraced a class of highly flexible tools consisting of fully fledged programming environments (e.g., IPython/Jupyter Notebook2, RStudio, and MATLAB) coupled with programming languages (e.g., Python and R) and highly capable and flexible bioinformatic libraries.\n\nInasmuch as these tools address the data collection and analysis portions of typical bioinformatic workflows, Cytoscape complements them by addressing visualization, additional analysis, and network publication. To date, combining these tools with Cytoscape has seen only limited success, largely because of the limitations of Cytoscape’s automation interfaces and its point-and-click user interface. Consequently, this integration has been labor intensive, inconvenient, and often unrepeatable, particularly as the complexity of analysis and visualization processing increases.\n\nWe created the cyREST Cytoscape app to enable automated access to the Cytoscape network and visualization models directly from within these tools, thereby exposing Cytoscape visualization, analysis, and publishing features in complex, varied, and reproducible bioinformatic workflows as shown in Figure 1.\n\ncyREST transforms Cytoscape into a REST-based microservice3 easily callable by workflows coded in REST-enabled languages such as Python, R, and Java. It is complemented by language-specific libraries that simplify Cytoscape access and harmonize native data models with Cytoscape’s network model as shown in Figure 2. (REST4 is short for Representational State Transfer.)\n\nHarmonization libraries are green and blue, representing an interface between workflow code and cyREST.\n\ncyREST complements Cytoscape’s existing Command Line Tool, where cyREST operates on Cytoscape’s data and Command executes Cytoscape commands. Since its original beta release date in late 2014, cyREST has been downloaded over 1900 times.\n\nIn this paper, we explain how cyREST relates to existing Cytoscape-oriented automation solutions, and then describe the design and use of cyREST itself. Finally, we give examples of cyREST’s use from applications written in Python (using IPython Notebook) and R (using RStudio).\n\nSeveral Cytoscape apps and plugins tackle tool interoperability and workflow reproducibility challenges (Table 1), most notably Cytoscape 3’s Command core module, but also including plugins deprecated along with Cytoscape 2.\n\nCytoscape 3’s Command Line Tool5 facilitates task automation via its own domain specific language, which provides access to high-level Cytoscape functions using a separate REST server within Cytoscape. While Command can execute individual commands (e.g., for loading and applying layouts) and sequences of commands (as scripts), it has no provision for accessing the network, style, and visualization information available through cyREST. Command is a complement to cyREST, where the combination greatly improves interoperability between Cytoscape and workflow-capable external tools, which contribute looping and control flow. A workflow can intermix Command and cyREST calls without conflict – they address different capabilities within Cytoscape.\n\nA notable alternative to cyREST is the Cyrface app6, which allows R and Bioconductor7 functions to be executed from Cytoscape 3, with the results returned to Cytoscape 3 – the opposite of a cyREST call. While this approach enables Cytoscape to act as the workflow orchestrator, it requires that a target application act as a server, which often requires idiosyncratic and complex support for each target application. So far, this approach has been taken only for interfacing to R.\n\nNumerous approaches to interoperability were implemented as ScriptingEngine8-based plugins for Cytoscape 2, now deprecated. Such plugins were created for executing scripts written in languages (e.g., JRuby9, Jython10, Groovy11, Clojure12, and JavaScript13). While these scripts had full access to Cytoscape’s comprehensive public API, their tight coupling to the Cytoscape runtime made them difficult to write, debug, and maintain. Because they were built on top of the Java virtual machine (JVM) and shared Cytoscape’s JVM, they had little access to increasingly capable and standardized third party native libraries (e.g., SciPy14 for Python). By contrast, the cyREST approach allows control of Cytoscape by best-of-breed tools and languages running independently as separate processes and leveraging best-of-breed native libraries. Conversely, while plugin implementations could interact with the user via dialog boxes directly within Cytoscape 2, scripts executing in separate processes run within their own windows, disconnected from Cytoscape 3.\n\nSimilar to the cyREST approach, the CytoscapeRPC plugin15 enabled independent scripts (e.g., Python) to create, query, and modify networks and visual styles in Cytoscape 2, but using an XML-RPC16 protocol instead of REST. Given the rapid adoption of REST conventions and supporting infrastructure, use of XML-RPC is becoming less common. Notably, RCytoscape17 is a Bioconductor package that leveraged CytoscapeRPC to enabled R applications to control Cytoscape 2. For Cytoscape 3, RCytoscape has been replaced by the RCy3 package in Bioconductor release 3.2, which leverages cyREST instead and is described below.\n\n\nWhat is cyREST?\n\ncyREST is a Cytoscape app that exposes Cytoscape network-related data structures and publishing functionality as a microservice callable via a REST protocol by external tools and languages. To date, it offers over 113 API calls, as documented at http://idekerlab.github.io/cyREST, where each API call accepts JSON-encoded values18 and returns JSON-encoded results.\n\nGiven that most modern tools and languages can call JSON-oriented REST services either directly or through well-vetted libraries, cyREST enables near-universal access to Cytoscape. However, such tools and languages often define data structures well tuned for use with their own specialized libraries that manipulate network-oriented data. To ease and accelerate the programming process, cyREST provides harmonization libraries designed to make calling cyREST natural and native within a tool or language. Harmonization libraries are described below.\n\nNote that while cyREST enables Cytoscape to act as a service, it is intended to serve only one client application at a time, where the client and Cytoscape run on the same workstation. Cytoscape itself remains capable of working on a single Cytoscape session at a time and maintains a visible window accessible to a user – Cytoscape does not operate in so-called headless mode. As a result, a client application is free to implement a workflow that intentionally sets up a network within Cytoscape so that a user can work further with it.\n\nThis section describes both the cyREST design and implementation and the implementation of harmonization libraries. It then presents example workflows created by combining standard data analysis tools with Cytoscape/cyREST.\n\ncyREST is a Cytoscape app that exposes the Cytoscape network data model to external tools and languages. It presents an API based on principles of REST, as do other popular biology-related data services, including those provided by EBI19. As a result, cyREST leverages REST facilities in existing tools and languages already built and vetted for use with other REST-based services. The definition and packaging of individual API functions takes advantage of lessons learned in building similar interfaces for Cytoscape 2.\n\ncyREST APIs represent all Cytoscape data objects and functions as resources according to principles of Resource-oriented Design (ROD)20. Data objects include networks, tables, and Visual Styles. Functions include applying layout algorithms on networks, updating Visual Styles, and performing statistical analysis. Under REST and ROD, each resource is encoded as a URL where hierarchy is represented as segments within the URL. For example, the URL http://localhost:1234/v1/tables/count can be decomposed into a REST server (http://localhost), port number (1234), an API version (v1), a resource (tables), and an attribute of the resource (count). So, this URL represents the count of global tables maintained by Cytoscape. Table 2 shows a sampling of resources available under the http://localhost:1234/v1 URL, with a more comprehensive list in the cyREST document at http://idekerlab.github.io/cyREST).\n\ncyREST follows ROD recommendations for sensible mappings between CRUD operations (create, read, update, and delete) and HTTP operations (POST, GET, PUT, DELETE) on data objects. Unless otherwise specified in the cyREST documentation, all HTTP operations accept or return values encoded as JSON. For example, GET http://localhost:1234/v1/networks returns a list of networkIds in an array (e.g., [1,2,3]). GET http://localhost:1234/v1/networks/networkId returns all nodes, edges, tables, and other data relating to network networkId in the Cytoscape.js21 JSON format.\n\nFor functions, ROD provides less guidance for CRUD/HTTP mappings or URL encoding. cyREST addresses this by grouping actions under http://localhost:1234/v1/apply (using GET operations) as illustrated in Table 3.\n\ncyREST is implemented as a Cytoscape app written in the Java programming language. It uses the Jersey JAX-RS22 library to implement the RESTful API, and provides access to data object and function operations as calls to public Cytoscape APIs. Under REST, each client request is phrased as an HTTP command (e.g., GET http://localhost:1234/v1/networks HTTP/1.1) and the reply is returned as a JSON structure.\n\ncyREST uses an embedded Grizzly HTTP server to receive and process client requests, where each HTTP request’s URL is mapped to a method in a resource manager class created by cyREST and registered with Grizzly. Each resource method declares the URL it services. When Grizzly receives a REST request, it calls the resource function registered for the URL, which calculates and returns a REST reply. For example, the NetworkResource defines a function that returns the number of Cytoscape networks in the current session, shown in the code fragment below. Note that the fragment defines its associated HTTP command, URL, and JSON output via Java annotations.\n\n\n\nAs with most Cytoscape apps, cyREST is initialized in its cyActivator function, which creates resource classes that reference all Cytoscape APIs to be used in servicing client requests. These include factories and managers for networks, network views, visual mapping, layout algorithms, groups, tables, sessions, and others.\n\nThe default HTTP port for cyREST is 1234, which can be changed by creating or modify the Cytoscape rest.port property (via Cytoscape’s Edit \| Preferences \| Properties dialog). Note that security-conscious workstations should firewall the cyREST port to prevent unintended outside access.\n\nTo test for the availability of a cyREST server, use an Internet browser to view the URL http://localhost:1234/v1/, which returns JSON-formatted version information similar to:\n\n\n\nEach cyREST function is exercised and validated before release by a suite of JUnit-based tests.\n\nWhile most programming languages make calling REST APIs and composing or parsing JSON simple, the data returned by cyREST may not be organized efficiently for ease of use in a particular language or with that language’s libraries. To maximize programmer productivity, we provide harmonization libraries (see Figure 2) to perform efficient cyREST calls on one hand, and present an interface easily used by programmers on the other hand. To date, we provide harmonization libraries for Python and R, and we expect to produce others.\n\nThe Python programming language has become popular among scientists and data analysts because of its rich collection of open source data analysis packages and a large developer community. It is an excellent tool for data cleansing, manipulation, analysis, and visualization; its igraph23, NetworkX24, and graph-tool25 packages are useful components in network data analysis workflows. In a workflow, it functions well as a glue that connects multiple heterogeneous computing resources, public databases, and private data files to build data analysis pipelines on workstations and computing clusters.\n\nWe created the py2cytoscape library to enable Python-based workflows to easily incorporate Cytoscape functionality by wrapping Python calls to cyREST and performing automatic translations between these packages’ data structures and cyREST’s JSON. For example, the following code creates a new Cytoscape network by using py2cytoscape calls, and replaces 16 lines that would be necessary when calling cyREST directly – see https://github.com/idekerlab/py2cytoscape/blob/develop/README.md for the direct cyREST calls.\n\n\n\npy2cytoscape is open source and is available from the PyPI repository (https://pypi.python.org/pypi/py2cytoscape).\n\nNote that py2cytoscape provides a widget that renders a network in cytoscape.js JSON format and then visualizes the network interactively within a Jupyter/IPython Notebook26 document, an example of which is at http://nbviewer.ipython.org/github/idekerlab/py2cytoscape/blob/develop/examples/New_wrapper_api_sample.ipynb.\n\nR is a particularly important platform for biologists because of the complimentary Bioconductor library. We are collaborating with the Bioconductor group to produce the RCy3 harmonization library for R27, which leverages cyREST to realize native R network visualization, analysis, and publishing functions. Its igraph, graph28, and RBGL29 packages are useful components for network data analysis workflows.\n\nA typical workflow performs data acquisition and integration, analysis, network visualization, and publishing. Often, these steps are performed one at a time by humans executing one discreet tool after another, possibly resulting in high labor costs, low throughput, high error rates, and an inability to reproduce the workflow reliably. In contrast, Figure 1 shows a workflow orchestrated by external tools such as Python and R, which interact with Cytoscape to perform parts of the workflow. As supplementary material, we provide downloadable sample workflows that incorporate and demonstrate cyREST functionality using py2cytoscape and RCy3 harmonization libraries.\n\nNote that Cytoscape/cyREST is designed to run on the same workstation as the workflow that calls it – Cytoscape maintains its own application window, and workflows may find advantage in soliciting users directly within Cytoscape.\n\nOur Python-based sample workflows are simple reflections of real world data analysis and visualization pipelines (see Figure 1) and use standard Python packages as much as possible. They are located in https://github.com/idekerlab/cy-rest-python and are viewable using the nbviewer web application (http://nbviewer.ipython.org/) in Jupyter Notebook format.\n\nSome Python packages are more capable or faster than equivalent Cytoscape functions, so the examples use them instead of calling Cytoscape. For example, Pandas30 prepares and analyzes data by using NumPy31 and SciPy library for processor-intensive tasks such as community detection.\n\nThe examples use the py2cytoscape harmonization libraries to demonstrate efficient cooperation between Python workflows by using NetworkX, igraph, and Cytoscape to integrate and visualize data generated in external tools. They show:\n\nData import from multiple data sources (remote/local)\n\nReformat and integration\n\nStatistical network analysis\n\nVisualization\n\nFor instance, the “Import KEGG pathways from web service” example demonstrates a typical biological data integration and visualization process involving KEGG databases32:\n\nSend a disease name query to the KEGG API\n\nFilter the result and reformat it\n\nImport disease pathway data directly from the KEGG database\n\nVisualize pathway data in Cytoscape\n\nEmbed the result as an interactive network diagram in the Jupyter Notebook\n\nThis workflow is simple to do with Cytoscape – the alternative would be a custom program or manual, file based operations that are hard to reproduce. With this workflow script, collaborators or reviewers can easily execute the same process on their environment, which is essential for reproducible scientific research.\n\nFor network analysis and visualization, igraph is an important and much used package by R users, and our sample R workflows (https://github.com/idekerlab/cy-rest-R) use it to complement the graph analysis features in Cytoscape.\n\nIn our Workflow 1 example, we scripted typical network visualization techniques using igraph’s graph analysis functions and Cytoscape’s data visualization features. First, we used igraph to detect community structure using a fast greedy modularity optimization algorithm33, and we calculated basic statistics of the network, including PageRank34 and betweenness centrality35. Our R code calls Cytoscape to create the resulting network, set properties for both layout and visual mapping, and generate an interactive network visualization. Output of this workflow helps users to visually understand the basic structure of the network (see Figure 3, which shows community structures color coded and used as weights for the Kamada-Kawai layout algorithm36).\n\nIn many cases, users apply an automatic layout algorithm early in a workflow to visually check the overall structure of a network. However, such layouts are often based on a simple force simulation and tend to produce uninformative “hairballs.” Our example illuminated network sub structures by using a community detection algorithm and igraph’s statistical analysis algorithms and visual styling. The alternative would be manual operation of both R and Cytoscape, which is laborious and error prone even for proficient users, and which is not reusable for subsequent networks.\n\n\nFuture development\n\nIn this paper, we demonstrated Python- and R-based workflow examples. In the near future, we expect to demonstrate cyREST usage in MATLAB and JavaScript (via Node.js38). While the cyREST and Command apps implement different automation features, we expect to unify the two APIs through a common implementation library in the Cytoscape core in the next Cytoscape release. Existing Cytoscape implementations manage a single Cytoscape session on behalf of a single user, can produce screen visualizations, and can potentially solicit user input even while under the control of cyREST. Future versions of Cytoscape will run headlessly and service multiple sessions simultaneously.\n\n\nSummary\n\nCytoscape is a highly popular desktop application for network biology analysis, visualization, and publication. The cyREST app extends Cytoscape into the realm of reproducible and high volume bioinformatic workflows by exposing a RESTful API that recasts Cytoscape as a visualization and rendering microservice. Using cyREST, data acquisition and analysis workflows previously limited to low quality (if any) visualizations can now leverage Cytoscape’s substantial library of network layouts, visualization features, and rendering options. Similarly, cyREST’s seamless integration with research and publication tools such as IPython/Jupyter Notebook improves individual researcher productivity by avoiding the need to manually operate Cytoscape.\n\nBecause it presents a RESTful interface, cyREST benefits can be realized in workflows built in most modern programming languages, and represents a significant contribution to productivity and reproducibility in data driven biology.\n\n\nSoftware availability\n\nCyREST software is available from the Cytoscape App Store: http://apps.cytoscape.org/apps/cyrest\n\nLatest source code of cyREST: https://github.com/idekerlab/cyREST\n\nFull REST API v1 document: http://idekerlab.github.io/cyREST/\n\nPy2cytoscape is in beta and is installable from PyPI repository: https://pypi.python.org/pypi/py2cytoscape\n\nPy2cytoscape source code: https://github.com/idekerlab/py2cytoscape\n\nPython sample workflows in Jupyter Notebook format: https://github.com/idekerlab/cy-rest-python\n\nR sample workflows: https://github.com/idekerlab/cy-rest-R\n\nLicense for cyREST, py2cytoscape, and all example workflows: MIT: http://opensource.org/licenses/MIT\n\nRCy3 source code: https://github.com/tmuetze/Bioconductor_RCy3_the_new_RCytoscape", "appendix": "Author contributions\n\n\n\nKO designed and implemented the software. KO and BD wrote this manuscript. TM and GK helped design and implement the RCy3 harmonization library and sample workflows. PS supervised TM and GK, and reviewed this manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported with funding from the National Resource for Network Biology (NRNB) under award numbers P41 RR031228 and GM103504 assigned to Trey Ideker.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nSome of the basic Python and R sample workflows are based on material developed by Dr. Kazuhiro Takemoto. Some of the basic Python workflows were written by Kozo Nishida. We incorporated numerous valuable editorial suggestions from Dr. Christian Zmasek and William Longabaugh into this paper.\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\nPérez F, Granger BE: IPython: a system for Interactive Scientific Computing. Comput Sci Eng. 2007; 9(3): 21–29. Publisher Full Text\n\nLewis J, Fowler M: Microservices. 2014. Reference Source\n\nFielding RT, Taylor RN: Principled design of the modern web architecture. ACM Trans Internet Technol. 2002; 2(2): 115–150. Publisher Full Text\n\nCommand Tool. 2013. Reference Source\n\nGonçalves E, Saez-Rodriguez J: Cyrface: An interface from Cytoscape to R that provides a user interface to R packages. [v1; ref status: indexed, http://f1000r.es/1tv]. F1000Res. 2013; 2: 192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGentleman RC, Carey VJ, Bates DM, et al.: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004; 5(10): R80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScriptingPlugins. December 2009. Reference Source\n\nRubyScriptingEngine. October 2012. Reference Source\n\nPythonScriptingEngine. October 2012. Reference Source\n\nGroovyEngine. August 2011. Reference Source\n\nClojureEngine. August 2011. Reference Source\n\nJavaScriptEngine. August 2011. Reference Source\n\nJones E, Oliphant T, Peterson P: {SciPy}: Open source scientific tools for {Python}. 2001.\n\nCytoscapeRPC. October 2011. Reference Source\n\nCerami E: Web services essentials: distributed applications with XML-RPC, SOAP, UDDI & WSDL. \"O’Reilly Media, Inc.\", 2002. Reference Source\n\nShannon PT, Grimes M, Kutlu B, et al.: RCytoscape: tools for exploratory network analysis. BMC Bioinformatics. 2013; 14(1): 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJson. Reference Source\n\nEMBL-EBI Web Services. April 2015. Reference Source\n\nSletten B: Resource-Oriented Architecture: The Rest of Rest. December 2009. Reference Source\n\nCytoscape.js. Reference Source\n\nApache CXF. Reference Source\n\nCsardi G, Nepusz T: The igraph software package for complex network research. InterJournal, Complex Systems. 2005; 1695(5): 1–9. Reference Source\n\nAA Hagberg, Schult DA, Swart P: Exploring Network Structure, Dynamics, and Function using NetworkX. In Proceedings of the 7th Python in Science Conferences (SciPy 2008). 2008; 2008: 11–16. Reference Source\n\ngraph-tool: Efficient network analysis. Reference Source\n\nJupyter. Reference Source\n\nThe R Project for Statistical Computing. Reference Source\n\ngraph. Reference Source\n\nRBGL. Reference Source\n\nMcKinney W: pandas: a Python data analysis library. 2015. Reference Source\n\nMcKinney W: Python for data analysis: Data wrangling with Pandas, NumPy, and IPython. \" O’Reilly Media, Inc.\", 2012. Reference Source\n\nKawashima S, Katayama T, Sato Y, et al.: KEGG API A web service using SOAP/WSDL to access the KEGG system. Genome Informatics. 2003; 14: 673–674. Reference Source\n\nClauset A, Newman MEJ, Moore C: Finding community structure in very large networks. Phys Rev E Stat Nonlin Soft Matter Phys. 2004; 70(6 Pt 2): 66111. PubMed Abstract \| Publisher Full Text\n\nPage L, Brin S, Motwani R, et al.: The PageRank Citation Ranking: Bringing Order to the Web. 1999. Reference Source\n\nFreeman LC: Centrality in social networks conceptual clarification. Social Networks. 1978–1979; 1(3): 215–239. Publisher Full Text\n\nKamada T, Kawai S: An algorithm for drawing general undirected graphs. Inf Process Lett. 1989; 31(1): 7–15. Publisher Full Text\n\nLee TI, Rinaldi NJ, Robert F, et al.: Transcriptional regulatory networks in Saccharomyces cerevisiae. Science. 2002; 298(5594): 799–804. PubMed Abstract \| Publisher Full Text\n\nTilkov S, Vinoski S: Node.js: Using Javascript to Build High-Performance Network Programs. IEEE Internet Computing. 2010; 6: 80–83. Publisher Full Text" }	[ { "id": "9848", "date": "10 Aug 2015", "name": "Sergey Nepomnyachiy", "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 present a Cytoscape app that turns it into a REST server. Most core functionality of Cytoscape can now be triggered by external agents using HTTP requests, rather than by point-and-clicking UI. This enables workflow automation and distribution that was hardly possible before. The fact that the protocol is HTTP based, allows working with it programmatically using any modern language. CyREST comes with libraries for Python and R that make the communication with Cytoscape through REST API even simpler, by enveloping the calls in methods and wrapping the returning JSON in objects.Perhaps in the future work section the authors could shed some light on the plans to allow users extending the set of REST handlers. The virtues of Cytsocape are beyond the core functionality and currently there is no way for a Cyto-app developer to map her own functions to REST (those are hard-wired in the code of CyREST app). A simple registry for binding 3rd party Cyto-app functions to URLs at runtime could bring the community to bring their on plug-in REST. Possible typospage 6:The default HTTP port for cyREST is 1234, which can be changed by creating or modify the Cytoscape rest.port propertypage 7:This workflow is simple to do with Cytoscape – the alternative would be a custom program or manual, file based operations that are hard to reproduce.", "responses": [] }, { "id": "9846", "date": "20 Aug 2015", "name": "Guanming Wu", "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 describe a Cytoscape app, CyREST, which exposes core Cytoscape functions as REST APIs for external software components to process network related data sets in automatic and reproducible workflows built using almost any programming languages. Users of workflows can visualize network data in Cytoscape via its powerful visualization features. The accompanied harmonization libraries for Python and R make the use of CyREST much easier and simpler. The manuscript is well organized, and the described app should be highly valuable for users working with big data related to networks for analysis and visualization. Minor comments: Figure 1: The difference between R/Bioconductor and Python is not clear to me. The double-arrowed line for Python is longer than R/Bioconductor, but I am not sure what extra is covered by Python. Python is a full-fledged programming language and capable for anything. Should some specific packages listed for Python? Further, probably the line for Cytoscape should be extended to cover two boxes for Curated Data Sets and Publications. Figure 2: Will it be better to call “Network (REST protocol)” just as “REST Protocol” to avoid confusing with other uses of \"network\" in the manuscript? What is the development status of RCy3? It will be nicer to indicate its status somewhere. From the manuscript, it seems that this is still in work-in-progress?Typos: In abstract: e.g. Python and R, JavaScript, and C# There are several typos in this page: https://github.com/idekerlab/cy-rest-R, e.g. Cytoscaep, Lesons, etc.", "responses": [] } ]	1	https://f1000research.com/articles/4-478
https://f1000research.com/articles/4-477/v1	05 Aug 15	{ "type": "Software Tool Article", "title": "SLiMScape 3.x: a Cytoscape 3 app for discovery of Short Linear Motifs in protein interaction networks", "authors": [ "Emily Olorin", "Kevin T. O'Brien", "Nicolas Palopoli", "Åsa Pérez-Bercoff", "Denis C. Shields", "Richard J. Edwards", "Emily Olorin", "Kevin T. O'Brien", "Nicolas Palopoli", "Åsa Pérez-Bercoff", "Denis C. Shields" ], "abstract": "Short linear motifs (SLiMs) are small protein sequence patterns that mediate a large number of critical protein-protein interactions, involved in processes such as complex formation, signal transduction, localisation and stabilisation. SLiMs show rapid evolutionary dynamics and are frequently the targets of molecular mimicry by pathogens. Identifying enriched sequence patterns due to convergent evolution in non-homologous proteins has proven to be a successful strategy for computational SLiM prediction. Tools of the SLiMSuite package use this strategy, using a statistical model to identify SLiM enrichment based on the evolutionary relationships, amino acid composition and predicted disorder of the input proteins. The quality of input data is critical for successful SLiM prediction. Cytoscape provides a user-friendly, interactive environment to explore interaction networks and select proteins based on common features, such as shared interaction partners. SLiMScape embeds tools of the SLiMSuite package for de novo SLiM discovery (SLiMFinder and QSLiMFinder) and identifying occurrences/enrichment of known SLiMs (SLiMProb) within this interactive framework. SLiMScape makes it easier to (1) generate high quality hypothesis-driven datasets for these tools, and (2) visualise predicted SLiM occurrences within the context of the network. To generate new predictions, users can select nodes from a protein network or provide a set of Uniprot identifiers. SLiMProb also requires additional query motif input. Jobs are then run remotely on the SLiMSuite server (http://rest.slimsuite.unsw.edu.au) for subsequent retrieval and visualisation. SLiMScape can also be used to retrieve and visualise results from jobs run directly on the server. SLiMScape and SLiMSuite are open source and freely available via GitHub under GNU licenses.", "keywords": [ "SLiM", "Short Linear Motif", "protein-protein interaction", "domain-motif interaction", "minimotif", "SLiMSuite", "SLiMFinder" ], "content": "Introduction\n\nMany protein-protein interactions (PPIs) are mediated by a short linear motif (SLiM) in one protein interacting with globular domains in another1. The past decade has seen the development of many computational methods and tools for predicting SLiMs from protein sequences and/or PPI data2,3. SLiMs are short (2–15 amino acids in length) and degenerate (with few residues determining specificity)4, which makes them hard to identify against a backdrop of highly conserved structural domains. These features also impart remarkable evolutionary plasticity, and convergent (i.e. independent) evolution of new SLiM occurrences is common5,6. Some of the best examples of this come from viruses, which often hijack host cellular processes via molecular mimicry of host SLiMs7. An effective approach for de novo SLiM discovery is to explicitly model this convergent evolution and look for enriched sequence patterns in non-homologous proteins5,8,9. SLiMFinder combines this approach with a robust statistical model, which enables a good estimation of the probability that an observed enrichment is by chance6,9–11. “Query” SLiMFinder (QSLiMFinder) extends this approach to define the motif search space on a specific protein, such as a viral interactor, and look for enrichment in the remaining non-homologous proteins in the dataset11.\n\nCombining motif discovery tools with biological knowledge has recently identified a new motif (“ABBA”) that binds the anaphase-promoting complex or cyclosome (APC/C) ubiquitin ligase12. Nevertheless, despite the improved performance and potential of these methods, they are yet to deliver the promised windfall of new motifs4. In large part, this is due to the difficulty in constructing appropriate datasets for SLiM discovery2,6. SLiM discovery relies on a careful balance of maximising the SLiM-containing signal in the data whilst removing noise by reducing the search space, either in terms of PPI or protein regions11. Cytoscape is a well developed platform for the interactive generation and exploration of PPI datasets13. Cytoscape is a useful resource for SLiM discovery, enabling visual groupings of proteins that share a common interaction partner and may also share a SLiM-mediated mechanism of binding. Such proteins can either be used as input for de novo SLiM discovery approaches5,6 or explored for enrichment of known motifs14.\n\nSLiMScape brings these tools together in a friendly environment that allows the user to browse, define and explore protein nodes whose interactors are enriched for over-represented motifs. The previously developed SLiMScape plugin for Cytoscape 2.x enabled the user to interactively run SLiMFinder9 for de novo SLiM discovery, or SLiMSearch15 for predicting novel occurrences of known SLiMs16. To take advantage of the recent developments and features of Cytoscape, we have developed SLiMScape 3.x, a redesigned and updated app for Cytoscape 3.x. The SLiM discovery functions of SLiMScape have also been extended through the incorporation of QSLiMFinder11 and enrichment statistics for known SLiMs using SLiMProb (formerly called SLiMSearch 1.x)14. SLiMScape 3.x is built on a new set of SLiMSuite servers that permit any of the commandline options of the standalone programs to be called via the Cytoscape app. Alternatively, SLiMSuite server jobs can be executed online (http://www.slimsuite.unsw.edu.au/servers.php) and the results imported and visualised using SLiMScape.\n\n\nMethods\n\nSLiMScape calls on the original Python implementations of programs in SLiMSuite; namely SLiMFinder9, QSLiMFinder11 and SLiMProb14. These are run remotely on the SLiMSuite servers at the University of New South Wales (UNSW) through a RESTful API interface (http://rest.slimsuite.unsw.edu.au) built on webpy 0.37 (https://github.com/webpy). SLiMScape also includes visualisation classes that utilise the Cytoscape interface to provide a graphical representation of results.\n\nSLiMScape was developed for Cytoscape 3, which substantially differs from previous Cytoscape versions. Because of this, it was necessary that the predecessor app, SLiMScape 1.x, be rewritten. The new version of SLiMScape was written on top of Open Service Gateway Initiative (OSGi)17, a software framework of pluggable modules, using the Maven project management tool (http://maven.apache.org)27. The language used has been Java SE Runtime Environment 7 (Java 7). An active internet connection is required to submit or retrieve server jobs, although subsequent analysis can be performed offline. Cytoscape may be closed whilst jobs are queued and/or running on the server.\n\nSLiMScape is designed to be run directly from within Cytoscape, or to visualise the results of a previous SLiMSuite server job. As such, all three SLiMSuite programs will accept three different inputs to identify the primary dataset of proteins for analysis (Figure 1):\n\n1. A selection of nodes from an existing Cytoscape network view. Node ‘name’ attributes must be Uniprot identifiers or accession numbers.\n\n2. A list of Uniprot identifiers or accession numbers, separated by commas, whitespace and/or new lines.\n\n3. The Job ID of a previous SLiMSuite server job. This may have been submitted via SLiMScape or run directly on the server.\n\nA. SLiMFinder run panel. B. SLiMFinder options panel. C. Results panel containing sequence and motif information. D. Results graph. Nodes containing SLiMs are indicated as red diamonds, where dark red indicates 2+ SLiMs. Nodes without SLiMs are displayed with default settings (blue circles in this case).\n\nSLiMFinder. A set of proteins is the only required input for SLiMFinder.\n\nQSLiMFinder. QSLiMFinder needs an additional query protein input, which is used to build the motif space11. This should be the Uniprot accession number of one of the input proteins. If no accession number is given, the first protein returned by Uniprot will be used. This is not necessarily the same as the first accession number provided and should therefore be avoided. The query protein(s) used will be reported in the job’s log output at the server.\n\nSLiMProb. SLiMProb needs one or more SLiMs to search within the protein dataset. These should be provided as SLiM regular expressions, e.g. DSG.{2,3}[ST] (where .{2,3} indicates two or three “wildcards” are permitted and [ST] is a serine or threonine). (See the Eukaryotic Linear Motif (ELM) database18 or SLiMScape documentation for more examples.) Multiple motifs can be provided, separated by commas. Whitespace is not permitted.\n\nData aliases. The SLiMSuite REST servers also feature a number of input aliases (http://rest.slimsuite.unsw.edu.au/alias). These include Uniprot ID lists and motif definitions for ELM and their occurrences from the ELM database18.\n\nThe main input parameters for the SLiM programs are specified in the “Settings” panel (Figure 1B):\n\nDisorder masking: masks residues with an IUPred19 disorder score < 0.2. This threshold can be modified by adding an iucut=X option to the custom parameters box.\n\nConservation masking: masks residues with low relative local conservation20. By default, this uses GOPHER21 to generate a Clustal Omega22 alignment of predicted eukaryotic orthologues from the April 2015 release of the Quest For Orthologues reference proteomes23. Different protein databases for GOPHER can be selected by adding an orthdb=X command to the custom parameters box. (See GOPHER server for details.)\n\nFeature masking: masks residues which occur in Uniprot-annotated domain and transmembrane features. A different set of Uniprot features may be masked by adding ftmask=LIST to the custom parameters box.\n\nSLiMBuild settings: the maximum motif length (number of defined positions and maximum wildcard spacer length) and whether to return ambiguous motifs. Amino acid equivalences for motif ambiguity can be set using the custom parameters box.\n\nCustom parameters: additional commandline options can be provided as “Custom parameters” (Figure 1B). These may be used to modify/supersede the options above, or to use SLiMSuite features that have been left out of the dialogue box for clarity. Please see the SLiMSuite documentation for a full list of commandline options. New features added to the SLiMSuite servers are instantly available through the custom parameters box.\n\nOnce all required inputs are provided and parameters are set, the desired program can be executed by clicking on the “Run X” button (where X is the name of the program being run). A popup window indicating processing will appear if a new server run is commencing (Figure 2). Jobs typically take a few minutes to run, although larger jobs (>100 proteins) may take several hours to complete depending on server load and the program settings. The server Job ID for this run will be displayed and also populate the Job ID box in the input panel (Figure 1A). This popup gives three progress options:\n\nStop and return to Cytoscape. The job will continue running on the server but Cytoscape can be used as usual in the meantime. Additional jobs may be sent to the server whilst waiting for one to complete.\n\nMonitor job progress at the SLiMSuite server. This will open the job’s status page at the SLiMSuite server in the user’s default web browser.\n\nCheck for job completion. If complete, this will load the results for visualisation. Otherwise, the popup will reload.\n\nRunning jobs will report the server job ID for future recall and provide options to: Stop and return to Cytoscape; Monitor job progress at the SLiMSuite server; or Check for job completion.\n\nAlternatively, a previous server Job ID can be entered in the Job ID box and loaded by clicking the “Retrieve” button. If complete, this will load the data into Cytoscape for visualisation. Otherwise, the running popup will appear. If an inappropriate Job ID is provided, or a job has crashed, an appropriate message should appear. If in doubt, the Monitor button in the progress popup (Figure 2) can be used to check that a job has executed correctly.\n\nOnce finished, tables showing the discovered SLiMs are presented in the Cytoscape control panel, in a new tab named after the SLiMSuite Job ID (Figure 1C). In each case, two tables are provided: (1) a table summarising overall statistics for each motif in the dataset, and (2) details of the individual occurrences (if any) of motifs in the input proteins. Table fields are identical for SLiMFinder and QSLiMFinder, whereas SLiMProb has slightly different output (Table 1). The Job ID produced by the server is presented in the “Job ID” box. The SLiMScape panel only shows a subset of the (Q)SLiMFinder/SLiMProb results fields; full output can be accessed at the server by clicking the “Full results” button and viewing the “main” or “occ” tabs. Results can also be accessed by entering the Job ID directly at the SLiMSuite server (http://www.slimsuite.unsw.edu.au/servers.php), enabling results to be viewed and shared independently of Cytoscape.\n\nIf a Job ID is retrieved without any nodes selected, a new network is created. By default, this will be named “SLiMOutput”; it is recommended to rename it in the network tab if multiple networks are to be analysed. Edges in this new network represent homology as detected with BLAST+ (E < 1e−4)24. The subnetworks defined by these edges correspond to the “Unrelated Protein Clusters (UPC)” used by the SLiMChance statistics to correct for evolutionary relationships and explicitly model convergent evolution.\n\nWhen nodes are selected in the Cytoscape network graph, the visual representation of nodes will be updated upon results retrieval. Due to the wide range of possible applications and user requirements, SLiMScape node formatting has been kept deliberately simple to avoid confusion. SLiM presence in a node is indicated by a change in colour and shape; from the native settings to a red diamond (Figure 1C). A darker shade of red indicates multiple SLiMs being present in that node. Only selected nodes will be altered and any nodes without SLiM occurrences remain as they were; if a network has already been formatted (e.g. by an earlier SLiMScape run), the old formatting is not removed first. To remove SLiMScape formatting, select the appropriate nodes, visit the Node tab of the Style control panel and “Remove Bypass” for the relevant properties. Users can also apply additional bypass or mapping styles to manually combine results from different runs. Missing nodes will not be added to an existing network; if nodes need to be added, users should create a new network by retrieving results without any nodes selected, and then merge the networks. Clearly, this will only happen when importing results from a SLiMSuite run that was created directly on the server, or with a different network.\n\nIt should be noted that although modified node attributes will be retained, SLiMScape tabs are not saved with a Cytoscape session. It is therefore recommended to rename modified networks with the Job ID if multiple runs have been performed on the same data.\n\nThe app relies on the Apache HTTP Client library to make HTTP requests to the SLiMSuite RESTful API server. The Java built in class (java.net.HttpURLConnection) was avoided as it does not support cancellation; an element important to a responsive user interface, particularly with large data sets and the substantial processing times these require.\n\n\nUse cases\n\nCytoscape can be used to import PPI data from a number of databases. The IntAct database at the European Bioinformatics Institute (EBI)25 is particularly suitable for SLiMScape analyses because the majority of nodes are mapped to Uniprot accession numbers. Other databases can be used but some additional mapping on to Uniprot identifiers might be required.\n\nThe proteins ‘F-box and WD repeat domain containing 11’ (Gene Symbol: FBXW11; Uniprot: Q9UKB1) and ‘beta-transducin repeat containing E3 ubiquitin protein ligase’ (Gene Symbol: BTRC; Uniprot: Q9Y297) are two proteins of the SCF(beta-TRCP)-ubiquitin ligase complex, which recognise phosphodegron motifs (ELM DEG_SCF_TRCP1_118) via their WD40 domains. We used the File » Import » Network » Public Databases function of Cytoscape to search for records matching FBXW11 and BTRC and imported the 184 records (184 edges; 86 nodes) from IntAct (Figure 3A). This network was then reduced to direct human and viral PPI of BTRC and FBXW11 and duplicate edges compressed.\n\nA. PPI network imported from IntAct. B. Human and viral proteins that interact with both FBXW11 and BTRC. C. SLiMFinder results for de novo SLiM prediction in the shared interactors (Job ID 15061100050). D. QSLiMFinder server results for de novo SLiM prediction using HIV Vpu as a query (Job ID 15061200029).\n\nWorking on the principle that WD40-interaction motifs are most likely to be found in proteins that interact with both WD40 proteins, Cytoscape was used to reduce the network further to nodes interacting with both proteins (Figure 3B). These interactors were then input to SLiMFinder for de novo SLiM prediction, using disorder and Uniprot feature masking. SLiMFinder returned a variant of the phosphodegron motif, DSG (P < 0.05), which was found in 14 of the 21 proteins (Figure 3C).\n\nThe HIV Vpu protein interacts with both proteins and is therefore of particular interest as a potential molecular mimic. We repeated the analysis using QSLiMFinder with Vpu as the query. As previously observed11, restricting the motif search space with QSLiMFinder increases the sensitivity of the search and returns the DSG motif with greater significance (P < 10−5) in addition to a number of different variants of the same motif (Figure 3D).\n\nThese data were clearly enriched for a DSG motif, which is a more degenerate variant of the annotated ELM DEG_SCF_TRCP1_1 motif, DSG.{2,3}[ST]. To investigate this further, the full set of BTRC and/or FBXW11 interactors were subject to a SLiMProb search of both the DSG and ELM (DSG.{2,3}[ST]) motifs with the same disorder and Uniprot feature masking (Job ID 15061200035). Using Cytoscape, proteins were arranged into three PPI sets of BTRC-only, FBXW11-only and shared interactors, arranged with DSG-containing proteins at one end and DSG-free proteins at the other (Figure 4). The SLiMProb run was opened up as new network to identify homology between the proteins (Figure 4A) and the two networks merged (Figure 4B). The SLiMProb search was repeated with additional conservation masking (Job ID 15061200036). Proteins with conserved motif occurrences were manually changed to circles (conserved DSG) or hexagons (conserved DSG.{2,3}[ST]) in the merged network (Figure 4B).\n\nA. Protein network generated from SLiMProb results labelled using Gene names extracted from Uniprot IDs. Edges in this network represent sequence homology. Red nodes indicate proteins with motif occurrences. Proteins with both motifs are in dark red. Proteins without either motif are blue circles. B. Merged PPI and homology network. Blue dotted lines represent sequence homology. Nodes are coloured and shaped according to motif occurrences: DSG, orange; DSG.{2,3}[ST], dark red; conserved DSG, circles; conserved DSG.{2,3}[ST], hexagons; no motifs, blue rectangles. The dark red circles (WWTR1 and TRIM9) have conserved DSG occurrences and unconserved DSG.{2,3}[ST] occurrences.\n\nVisual inspection of the motif distribution suggested that the DSG motif is actually more specific for the interaction than the defined ELM, showing a greater enrichment in the proportion of joint interactors versus interaction partners of only BTRC or FBXW11. This observation was confirmed by subsequent SLiMProb analysis of six different groups of proteins (Figure 5). However, when the sequence composition of the proteins was taken into consideration, it became clear that the defined ELM was considerably more enriched across all BTRC/FBXW11 interactors than the simpler DSG motif. This was particularly pronounced when looking at evolutionarily conserved instances (i.e. results from SLiMProb analyses with conservation masking) (Figure 5C). This demonstrates the power of combining Cytoscape and SLiMSuite to get insights that are not obvious from either tool in isolation.\n\nA. Schematic of the different protein subsets analysed. B. Proportions of unrelated proteins with motif occurrences. ELM indicates DSG.2,3[ST]. DSGcons and ELMcons are conserved occurrences. C. Observed/expected number of unrelated proteins with motif occurrences (N_UPC/E_UPC, see Table 1).\n\nAlso of interest is the other viral protein in the dataset, the NSP1 protein of ribovirus A (Uniprot: Q84940), which is reported to interact with FBXW11 but not BTRC. It has a DSG.SD sequence, which is intriguingly similar to the DEG_SCF_TRCP1_1 motif. Indeed, this motif was recently reported to be another case of molecular mimicry, targeting the beta-TrCP subunit26. The evolutionary dynamics of SLiMs are complex; further work will need to be done to establish whether any of the other DSG instances that fail to match the more specific ELM definition represent hitherto undescribed variants of the motif or non-functional background sequence patterns.\n\n\nSummary\n\nSLiM discovery is a challenging task that requires high quality data in addition to appropriate bioinformatics tools. There is often a disconnect between users with the biological knowledge to construct the former and those with the computational experience to run the latter. Embedding the SLiM discovery tools of SLiMSuite within the interactive environment of Cytoscape will help to bridge that gap, enable new patterns to be identified and new questions to be formulated.\n\n\nData availability\n\nSLiMSuite results for the figures in this manuscript can be retrieved by entering the Job ID indicated in the text through the SLiMScape app, or at: http://www.slimsuite.unsw.edu.au/servers.php. Job IDs for Figure 5 are given in Table 2.\n\n\nSoftware availability\n\nhttp://apps.cytoscape.org/apps/slimscape\n\nhttps://github.com/slimsuite/SLiMScape\n\nhttps://github.com/F1000Research/SLiMScape\n\nhttp://dx.doi.org/10.5281/zenodo.1983528\n\nGNU Lesser General Public License (http://www.gnu.org/licenses/gpl.html).\n\nSLiMSuite is available via GitHub (https://github.com/slimsuite/SLiMSuite) under a GNU General Public License (DOI: 10.5281/zenodo.1948029).", "appendix": "Author contributions\n\n\n\nRJE, DCS and KTO’B conceived of the SLiMScape app. EO, RJE, ÅP-B and NP wrote the paper. KTO’B and DCS commented on the paper. EO coded the SLiMScape app with advice from KTO’B. RJE coded the SLiMSuite REST servers. RJE, DCS, NP and ÅP-B developed the SLiMSuite tools used by the app. KTO’B, NP, DCS and RJE tested the SLiMScape app and SLiMSuite servers.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nEO and ÅP-B were funded from a UNSW start up grant awarded to RJE. Development of some of the tools by NP and RJE was funded by a Biotechnology and Biological Sciences Research Council (BBSRC) New Investigator Award (BB/I006230/1) to RJE.\n\nI confirm that the 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 the UNSW Science Faculty Computing Unit for all their assistance in setting up the SLiMSuite REST servers. In particular, RJE would like to thank Adrian Plummer, without whom there would be no functional server.\n\n\nSupplementary material\n\nCytoscape session file for network figures in SLiMScape manuscript. This file contains a Cytoscape 3 session file (*.cys) containing five networks used to generate Figure 3–Figure 5.\n\nClick here to access the data.\n\n\nReferences\n\nTompa P, Davey NE, Gibson TJ, et al.: A million peptide motifs for the molecular biologist. Mol Cell. 2014; 55(2): 161–9. PubMed Abstract \| Publisher Full Text\n\nEdwards RJ, Palopoli N: Computational prediction of short linear motifs from protein sequences. Methods Mol Biol. 2015; 1268: 89–141. PubMed Abstract \| Publisher Full Text\n\nKelil A, Dubreuil B, Levy ED, et al.: Fast and accurate discovery of degenerate linear motifs in protein sequences. PLoS One. 2014; 9(9): e106081. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavey NE, Van Roey K, Weatheritt RJ, et al.: Attributes of short linear motifs. Mol Biosyst. 2012; 8(1): 268–81. PubMed Abstract \| Publisher Full Text\n\nNeduva V, Linding R, Su-Angrand I, et al.: Systematic discovery of new recognition peptides mediating protein interaction networks. PLoS Biol. 2005; 3(12): e405. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdwards RJ, Davey NE, O’Brien KT, et al.: Interactome-wide prediction of short, disordered protein interaction motifs in humans. Mol Biosyst. 2012; 8(1): 282–95. PubMed Abstract \| Publisher Full Text\n\nDavey NE, Travé G, Gibson TJ: How viruses hijack cell regulation. Trends Biochem Sci. 2011; 36(3): 159–69. PubMed Abstract \| Publisher Full Text\n\nDavey NE, Shields DC, Edwards RJ: SLiMDisc: short, linear motif discovery, correcting for common evolutionary descent. Nucleic Acids Res. 2006; 34(12): 3546–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdwards RJ, Davey NE, Shields DC: SLiMFinder: a probabilistic method for identifying over-represented, convergently evolved, short linear motifs in proteins. PLoS One. 2007; 2(10): e967. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavey NE, Edwards RJ, Shields DC: Estimation and efficient computation of the true probability of recurrence of short linear protein sequence motifs in unrelated proteins. BMC Bioinformatics. 2010; 11: 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPalopoli N, Lythgow KT, Edwards RJ: QSLiMFinder: improved short linear motif prediction using specific query protein data. Bioinformatics. 2015; 31(14): 2284–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDi Fiore B, Davey NE, Hagting A, et al.: The ABBA motif binds APC/C activators and is shared by APC/C substrates and regulators. Dev Cell. 2015; 32(3): 358–72. 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\nDavey NE, Haslam NJ, Shields DC, et al.: SLiMSearch: a webserver for finding novel occurrences of short linear motifs in proteins, incorporating sequence context. In TMH Dijkstra, E Tsivtsivadze, E Marchiori, T Heskes, editors, Pattern Recognition in Bioinformatics. 5th IAPR International Conference, PRIB 2010, Nijmegen, The Netherlands, September 22–24, 2010. Proceedings, volume 6282 of Lecture Notes in Computer Science, Berlin, Springer-Verlag. 2010; 50–61. Publisher Full Text\n\nDavey NE, Haslam NJ, Shields DC, et al.: SLiMSearch 2.0: biological context for short linear motifs in proteins. Nucleic Acids Res. 2011; 39(Web Server issue): W56–W60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nO'Brien KT, Haslam NJ, Shields DC: SLiMScape: a protein short linear motif analysis plugin for Cytoscape. BMC Bioinformatics. 2013; 14: 224. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOSGi Alliance: OSGi service platform. Release 3. IOS Press, Inc. 2003. Reference Source\n\nDinkel H, Van Roey K, Michael S, et al.: The eukaryotic linear motif resource ELM: 10 years and counting. Nucleic Acids Res. 2014; 42(Database issue): D259–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDosztányi Z, Csizmok V, Tompa P, et al.: IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content. Bioinformatics. 2005; 21(16): 3433–4. PubMed Abstract \| Publisher Full Text\n\nDavey NE, Shields DC, Edwards RJ: Masking residues using context-specific evolutionary conservation significantly improves short linear motif discovery. Bioinformatics. 2009; 25(4): 443–50. PubMed Abstract \| Publisher Full Text\n\nDavey NE, Edwards RJ, Shields DC: The SLiMDisc server: short, linear motif discovery in proteins. Nucleic Acids Res. 2007; 35(Web server issue): w455–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSievers F, Wilm A, Dineen D, et al.: Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol. 2011; 7: 539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDessimoz C, Gabaldón T, Roos DS, et al.: Toward community standards in the quest for orthologs. Bioinformatics. 2012; 28(6): 900–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCamacho C, Coulouris G, Avagyan V, et al.: BLAST+: architecture and applications. BMC Bioinformatics. 2009; 10: 421. 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\nMorelli M, Dennis AF, Patton JT: Putative E3 ubiquitin ligase of human rotavirus inhibits NF-κB activation by using molecular mimicry to target β-TrCP. MBio. 2015; 6(1): e02490–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nApache Software Foundation. Apache Maven Project. 2015. Reference Source\n\nOlorin E, O’brien KT, Palopoli N, et al.: SLiMScape: SLiMScape v3.0.1 (Publication release). Zenodo. 2015. Data Source\n\nEdwards Lab: SLiMSuite: SLiMSuite v1.0.0 (2015-07-06). Zenodo. 2015. Data Source" }	[ { "id": "9827", "date": "14 Aug 2015", "name": "Ulrich Stelzl", "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\nSLimfinder is one of the state of the art linear sequence motif discovery tool. The idea of short linear sequence motifs is intuitive, has been demonstrated as a binding mode for many domains and may govern PTMs to some extent. However, because of low sequence complexity, it is very hard to find motifs de novo in a reliable manner, i.e. statistically sound manner, and it is thus beyond the scope of non-motif expert biologist. Cytoscape 3 was a major change in network analysis. APPs are required that run easily and reliably and on a solid algorithmic basis, which are often computer intensive.This manuscript reports the update of the SLIMfinder plugin to a SLiMScape APP and provides a very detailed description how to use it and a test case searching for FBXW11 and BTRC WD40 domain binding motifs “de novo” (within the space of their interaction partners from INTACT). This is a very useful report for a very useful tool and the nice documentation, including the available cys sessions, will stimulate its use.Two minor points:Praxis of storage of data on the external server is a little unclear, I guess quite some potential use may find an option to make sure data are removed after analysis / after some time quite important. In practice output files will be renamed by the user, but it is always nice to give unique name to files that are exported e.g. coupled to jobID or something systematic, so that many outputs can be saved e.g. as batch and then processed with a script. If all are “SLiMOutput” than this may get tedious.", "responses": [ { "c_id": "1537", "date": "27 Aug 2015", "name": "Richard Edwards", "role": "Author Response", "response": "These are excellent points and we will update the documentation accordingly. The quick/simple answers are:1. Server data is currently stored for at least one week but longer in reality. It can be deleted sooner on request. Runs can be password protected using password=X in the “Custom parameters” box (or &password=X directly at the server). This does not encrypt the data (i.e. we can still access it directly on the server) but it will stop random access by others.2. The SLiMSuite download includes a program, SLiMParser, that can be used to download and parse data from server runs. Basic documentation can be found here: http://rest.slimsuite.unsw.edu.au/slimparser. To parse and locally output job files, run with the commands: restin=JOBID restout=T [password=X] [restbase=X]. If restbase=X is not given, the jobid will be used for output files (JOBID.) else they will be named X. as set by restbase=X. Additional SLiMParser documentation will be posted on the SLiMSuite Blog." } ] }, { "id": "10973", "date": "09 Nov 2015", "name": "Srikrishna Subramanian", "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\nIn their manuscript entitled “SLiMScape 3.x: a Cytoscape 3 app for discovery of Short Linear Motifs in protein interaction networks”, Olorin et al. have described the use of their updated SLiMScape plugin (Cytoscape 2.x) as a SLiMScape 3.x APP (Cytoscape 3.x). The effort made by the research group to actually update the plugin for use with the latest version of Cytoscape is commendable. They have also added new tools to the updated APP.The integration of some of the SLiMSuite tools with the Cytoscape network analysis platform is likely to be of potential importance in the discovery of SLiMs. We believe that the manuscript is suitable for indexing. However, we have the following general queries:Are there any other APPs or plugins available with Cytoscape with similar utility? If yes, then has the performance of SLiMScape been compared with those? On pg.4 of the version 1 of the paper, the authors refer to the “SLiMSuite documentation” for the complete list of commandline options. It would be good if the link to the documentation be added somewhere in the manuscript.", "responses": [ { "c_id": "1707", "date": "30 Nov 2015", "name": "Richard Edwards", "role": "Author Response", "response": "We are not aware of any other apps or plugins with similar utility, with the exception of the SLiMScape plugin for Cytoscape 2.x. The SLiMSuite tools made available through SLiMScape have themselves been benchmarked against other tools where possible. We do not currently have a single page that lists all of the SLiMSuite documentation, as each tool has its own documentation. This documentation is available via the existing links in the paper: http://www.slimsuite.unsw.edu.au/servers.php and http://rest.slimsuite.unsw.edu.au/. It can also be accessed via the SLiMSuite Blog. Finished jobs at the REST server also link directly to the relevant tool's documentation, which include commandline options." } ] } ]	1	https://f1000research.com/articles/4-477
https://f1000research.com/articles/3-149/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "setsApp: Set operations for Cytoscape Nodes and Edges", "authors": [ "John H. Morris", "Allan Wu", "Nadezhda T. Doncheva", "Mario Albrecht", "Thomas E Ferrin", "Allan Wu", "Nadezhda T. Doncheva", "Mario Albrecht", "Thomas E Ferrin" ], "abstract": "setsApp (http://apps.cytoscape.org/apps/setsapp) is a relatively simple Cytoscape 3 app for users to handle groups of nodes and/or edges. It supports several important biological workflows and enables various set operations. setsApp provides basic tools to create sets of nodes or edges, import or export sets, and perform standard set operations (union, difference, intersection) on those sets. The sets functionality is also exposed to users and app developers in the form of a set of commands that can be used for scripting purposes or integrated in other Cytoscape apps.", "keywords": [ "Cytoscape1", "2 provides an environment for the visualization and analysis of networks and associated annotations. The primary audience for Cytoscape is the biological community and Cytoscape supports a number of standard use cases for analyzing and visualizing biological data. Many of these use cases involve the selection of a number of nodes or edges based on some analysis or annotation and either performing an action on that selection or comparing those nodes or edges to a different set of nodes or edges that resulted from alternative analyses or analyses based on alternative annotations. The core capabilities for Cytoscape provide some tools to facilitate these types of comparisons but they can be counterintuitive or complicated to use. setsApp is a Cytoscape 3 application that provides a general set of tools for users and developers to define and maintain sets of nodes or edges and compare those sets using the standard set operations of union", "intersection", "and difference." ], "content": "Introduction\n\nCytoscape1,2 provides an environment for the visualization and analysis of networks and associated annotations. The primary audience for Cytoscape is the biological community and Cytoscape supports a number of standard use cases for analyzing and visualizing biological data. Many of these use cases involve the selection of a number of nodes or edges based on some analysis or annotation and either performing an action on that selection or comparing those nodes or edges to a different set of nodes or edges that resulted from alternative analyses or analyses based on alternative annotations. The core capabilities for Cytoscape provide some tools to facilitate these types of comparisons but they can be counterintuitive or complicated to use. setsApp is a Cytoscape 3 application that provides a general set of tools for users and developers to define and maintain sets of nodes or edges and compare those sets using the standard set operations of union, intersection, and difference.\n\nIn this paper, we present the implementation of setsApp, in particular, how setsApp integrates with the Cytoscape command system, and then present a sample biological workflow using setsApp.\n\n\nImplementation\n\nCytoscape provides two approaches to implementing apps: a simple app and a bundle app. Simple apps are implemented using the same general approach as in Cytoscape 2.8, but at the cost of significant flexibility. Bundle apps utilize Open Service Gateway Initiative (OSGi)3 interfaces through APIs provided by Cytoscape to interact with the Cytoscape core functionality. setsApp is implemented as a bundle app and utilizes the Cytoscape 3.1.0 API. There are three main components to the setsApp implementation: the user interface, the command interface, and the underlying data model for maintaining sets of nodes and edges.\n\nThe setsApp user interface consists of menu items in the main Apps menu, node and edge context menus, and a panel added to the Control Panel (left or west) section of the Cytoscape user interface. The main feature of the Sets panel is the list of currently defined sets. Each set can be expanded to see all of the nodes or edges within that set, and context menus provide the ability to select, deselect, rename, or remove sets. When multiple sets are selected, the Set Operations buttons are enabled. This allows users to create new sets based on the union, intersection, or difference of other sets. Note that the results of a union or intersection are well-defined for multiple sets, but the difference operation is order dependent. If only two sets are selected, the order of selection is preserved. If more than two sets are selected, the order is the order of selection, so care must be taken when attempting to create a difference set of more than two sets. Sets can be created from the currently selected nodes or edges, or based on a particular node or edge attribute. When creating sets from attributes, the user will need to supply a prefix for the sets to be created and choose the attribute (currently only String attributes are supported) from a list. Sets can also be created by importing them from a simple text file. Each set can be individually exported to a text file.\n\nsetsApp provides a context menu for sets and set members in the control pane. In addition to having menu items to manage sets, the user may select all set members in the network, or if the set is expanded, individual members. This functionality presents an easy way for users to visualize the results of set operations and to perform interactive exploratory analysis.\n\nThe menus provided through the top-level Apps menu offer the same functionality as the Create set from menu in the setsApp control panel with the addition of a menu to import a set from a file. Node and edge context menus provide the user with the ability to add or remove the corresponding node or edge from sets.\n\nIn addition to the standard user interface described above, setsApp provides a number of “commands”. These commands may be used for scripting purposes or by other Cytoscape apps that wish to take advantage of the setsApp functionality. Table 1 provides a list of commands and the arguments.\n\nArguments with asterisks are required.\n\nA command is made available to Cytoscape by creating a standard Cytoscape TaskFactory with two new properties in the org.cytoscape.work package: ServiceProperties.COMMAND_NAMESPACE, which is always set to “setsApp”; and ServiceProperties.COMMAND, which is the command name (e.g. “addTo”). The command arguments are implemented as Tunables within the Task called by the designated TaskFactory. Because there is no guarantee that the Task will be executed within the context of a GUI, care should be taken to make sure that the appropriate Tunable types are used. For example, the NodeList Tunable allows the command-line user to enter a list of nodes rather than assuming that the user will have selected nodes interactively.\n\nFor another Cytoscape app to use any of these commands, it would need to call one of the Cytoscape TaskManagers and provide it org.cytoscape.command.CommandExecutorTaskFactory’s createTaskIterator method with the appropriate argument map, command, and command namespace. The TaskObserver method may be used if the command returns any values. For example:\n\nListing 1. Example Command Usage.\n\n\n\nThe main model object for a node or edge set is the Set object, which stores a map of all of the nodes or edges in this set. A SetsManager provides the methods to create and destroy sets. The SetsManager also serves the critical function of serializing the information about Sets to the default hidden table (see CyNetwork.HIDDEN_ATTRS) for nodes or edges (depending on the type of the Set). Each Set is created as a boolean column in the hidden table which is set to true if the corresponding node or edge is in that set. By storing values in the default hidden tables, the information about sets is automatically saved in Cytoscape sessions and restored when sessions are reloaded. SetsManager implements SessionReloadedListener and recreates the Sets from the information stored in the hidden table columns.\n\n\nResults\n\nA simple example use case might be the exploration of the data set from Ideker et al., 20014, which measured the change in expression for 331 genes after a systematic deletion of genes known to be involved in the Saccharomyces cerevisiae switch to galactose metabolism. This data was combined with known protein-protein and protein-DNA interactions to explore the biological response to deletions in the presence (the G in the column names) or absence of galactose in the medium. This data set is now included as a sample with Cytoscape downloads (galFiltered.cys). In our workflow we use Cytoscape’s Select panel to select all proteins that are underexpressed (gal1RGexp < -0.5 fold change) in the deletion of GAL1 (Figure 2). That selection is saved as a set named GAL1- (Apps→SetsApp→Create node set). We then select all of the proteins that are overexpressed (gal1RGexp > 0.5 fold change) in the deletion of GAL1 and save that selection as a set named GAL1+. Repeating this for GAL4 (column gal4RGexp) and GAL80 (column gal80Rexp) results in 6 sets altogether: GAL1+, GAL1-, GAL4+, GAL4-, GAL80+, GAL80- (Figure 3). Note that the data for GAL1 and GAL4 is in the presence of galactose, but the data for GAL80 is in the absence of galactose since GAL80 is a known repressor of GAL4.\n\nGiven those six sets of genes, we can explore the data sets by looking at combinations of the sets. For example, we could look at the intersection of all of the underexpressed proteins by selecting each of GAL1-, GAL4-, and GAL80- in the Sets panel and pressing the Intersection button in the Set Operations box near the bottom of the panel. If we name the resulting set GAL- we see that it contains a single gene: ARG1. In this data set of 331 genes, only this one gene is repressed for all three of the deleted GAL genes. In the absence of galactose when GAL80 is deleted, ARG1 is underexpressed, and in the presence of galactose when either GAL1 or GAL4 are deleted the gene is also underexpressed. Looking at the expression significance values in the Node Table Panel of Cytoscape (gal1RGsig, gal4RGsig, and gal80Rsig) this is a highly significant result, although there is no direct correlation between the galactose switch and arginine biosynthesis regulation that we were able to find in the literature. On the other hand, ARG1 is regulated by the GCN4 activator which is known to repress protein synthesis during periods of stress or starvation5, which explains the significant down-regulation of ARG1. We can perform a similar analysis to understand the consistent up-regulation of the five genes in the GAL+ set: GIP1, NCE103, YIG1, POT1, and ICL1. Figure 1 shows the results of the intersection operations.\n\nGAL+ and GAL- were created by performing the intersection of all of the + and - sets, respectively.\n\n\nConclusions\n\nThere are many Cytoscape workflows that could take advantage of the features of setsApp. Any workflow that might want to look for groups of nodes or edges that share multiple traits, or that explicitly do not share those traits. While it is possible to duplicate many of the final results enabled by setsApp by using Cytoscape 3.1’s new Select panel, a user would need to know in advance exactly the combination of features that were biologically interesting. setsApp provides an alternative that allows users to explore various combinations of nodes and edges and to save such selections for later uses.\n\nIn the workflow we developed above, we combined the functionality of Cytoscape’s Select panel with setsApp to explore combinations of sets of genes based on shared properties. There are many more sophisticated apps available to Cytoscape users that could be used to do a more thorough analysis of this data set including jActiveModules6, clusterMaker7 and RINalyzer8, however, the workflow above demonstrates the utility of a simple set-oriented approach to exploring networks.\n\n\nSoftware availability\n\nSoftware available from: http://apps.cytoscape.org/apps/setsApp\n\nLatest source code: https://github.com/RBVI/setsApp.\n\nSource code as at the time of publication: https://github.com/F1000Research/setsApp\n\nArchived code as at the time of publication: http://www.dx.doi.org/10.5281/zenodo.104249\n\nLicense: Lesser GNU Public License 3.0: https://www.gnu.org/licenses/lgpl.html", "appendix": "Author contributions\n\n\n\nJHM wrote the manuscript and enhanced the app. AW ported the initial version of the app from Cytoscape 2.8 to Cytoscape 3. NTD wrote the initial version of the app. MA and TF supervised app development and provided input on the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nJHM was funded by NIGMS grants P41-GM103504 and P41-GM103311. AW and TF were funded by NIGMS grant P41-GM103311. NTD was partially funded by a Boehringer Ingelheim Fonds travel grant, and her research was conducted in the context of the DFG-funded Cluster of Excellence for Multimodal Computing and Interaction. MA was financially supported by the projects GANI_MED and BioTechMed-Graz.\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): 2498–504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCline MS, Smoot M, Cerami E, et al.: Integration of biological networks and gene expression data using Cytoscape. Nat Protoc. 2007; 2(10): 2366–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOSGi Alliance. OSGi service platform: release 3, March 2003. IOS Press: Ohmsha, Amsterdam; Washington, DC, 2003. Reference Source\n\nIdeker T, Thorsson V, Ranish JA, et al.: Integrated genomic and proteomic analyses of a systematically perturbed metabolic network. Science (New York, N.Y). 2001; 292(5518): 929–34. PubMed Abstract \| Publisher Full Text\n\nHinnebusch AG: Translational regulation of yeast GCN4. A window on factors that control initiator-trna binding to the ribosome. J Biol Chem. 1997; 272(35): 21661–4. PubMed Abstract \| Publisher Full Text\n\nIdeker T, Ozier O, Schwikowski B, et al.: Discovering regulatory and signalling circuits in molecular interaction networks. Bioinformatics. 2002; 18(Suppl 1): S233–40. PubMed Abstract \| Publisher Full Text\n\nMorris JH, Apeltsin L, Newman AM, et al.: clusterMaker: a multi-algorithm clustering plugin for Cytoscape. BMC Bioinformatics. 2011; 12: 436. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoncheva NT, Assenov Y, Domingues FS, et al.: Topological analysis and interactive visualization of biological networks and protein structures. Nat Protoc. 2012; 7(4): 670–85. PubMed Abstract \| Publisher Full Text\n\nMorris JH, Wu A, Doncheva NT, et al.: F1000Research/SettsApp. ZENODO. 2014. Data Source" }	[ { "id": "5300", "date": "14 Jul 2014", "name": "Jiguang Wang", "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 manuscript described a new Cytoscape App, named SetsApp. This app provides very simple function about the operations on node or edge sets. It is easy to install, and easy to use. I have only minor suggestions for revision:The operation \"difference\" on three sets is not well defined. This operation should be disabled when selecting three sets. If a new set has the same name with the old one, there will be a warning. This is good, but when I close the warning, my operation is also closed. It will be more convenient if this app goes back to the window of \"Enter a new name for the new set\". When importing sets from file: if the parameters are not properly set, the app will import nothing. It will be better if warning information tells users the parameters are not properly used.", "responses": [] }, { "id": "5302", "date": "14 Jul 2014", "name": "Thomas Kelder", "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 describe setsApp for Cytoscape, an app that provides a user and developer friendly way to define sets of nodes and edges and perform union, intersect and difference operations on these sets. The features of this app are rather basic (set definition and simple set operations), but nevertheless provide a very useful extension to core Cytoscape functionality. By designing the app so that it can be used through either the GUI or the commands interface makes it a very powerful utility component that can be used in different use cases, workflows, or as building block in other apps. The manuscript is well written and provides a clear and detailed description of the app data model, usage instructions and use case. I tested the app using Cytoscape 3.1.1 and everything worked as described.Minor comments on the manuscript: The paragraph following Table 1 is very technical. A code example to illustrate the use of Tunables would be useful. Also consider placing this paragraph after Listing 1, so the reader is first being shown how to call the setsApp commands, before diving into technical details. Listing 1: The code example is incomplete, please define the “serviceRegistrar” variable, how would I get an instance as developer?Minor comments and suggestions on the App: It doesn't seem possible to compare sets across different networks, the App gives an error when I tried. What is the reasoning behind this? If the networks contain overlapping nodes, wouldn't it be valid to perform the operations across the different networks? In case sets have been defined for multiple networks, it is hard to see in the Sets panel to which network each set belongs. The only way I could find was to click the set and choose “Select” so the nodes get selected in the corresponding network. It would be useful to group the sets by network or indicate the parent network otherwise (i.e. different colors of the red dots). Small GUI tweak proposal: in the dialog where the user needs to specify the set name, it would be handy if the “Enter” key would map to the “Ok” button, so you don’t have to switch to the mouse. It would be more intuitive and speed up the creation of several sets.", "responses": [] }, { "id": "5299", "date": "28 Jul 2014", "name": "Tamara Munzner", "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 short article clearly describes the setsApp plugin for Cytoscape and walks the reader through example analysis. It ends with a useful disclaimer that the goal of the app is to provide users with a simple workflow, rather than completely novel capabilities.The paragraph covering TaskFactory details will probably require multiple passes for readers unfamiliar with Cytoscape internals, but the major point can still be gleaned from the write-up as it stands.Minor issue: It would be easier to read if the Figures were renumbered so that they match the order of discussion in the text; now Figure 1 comes after 2 and 3.", "responses": [] } ]	1	https://f1000research.com/articles/3-149
https://f1000research.com/articles/4-476/v1	05 Aug 15	{ "type": "Software Tool Article", "title": "CySpanningTree: Minimal Spanning Tree computation in Cytoscape", "authors": [ "Faizaan Shaik", "Srikanth Bezawada", "Neena Goveas", "Neena Goveas" ], "abstract": "Simulating graph models for real world networks is made easy using software tools like Cytoscape. In this paper, we present the open-source CySpanningTree app for Cytoscape that creates a minimal/maximal spanning tree network for a given Cytoscape network. CySpanningTree provides two historical ways for calculating a spanning tree: Prim’s and Kruskal’s algorithms. Minimal spanning tree discovery in a given graph is a fundamental problem with diverse applications like spanning tree network optimization protocol, cost effective design of various kinds of networks, approximation algorithm for some NP-hard problems, cluster analysis, reducing data storage in sequencing amino acids in a protein, etc. This article demonstrates the procedure for extraction of a spanning tree from complex data sets like gene expression data and world network. The article also provides an approximate solution to the traveling salesman problem with minimum spanning tree heuristic. CySpanningTree for Cytoscape 3 is available from the Cytoscape app store.", "keywords": [ "minimum spanning tree", "gene expression data", "euclidean distance", "Hamiltonian cycle" ], "content": "Introduction\n\nGraph theory is being widely used for network analysis in various fields1. Extraction of various kinds of subnetworks is one of the ways to identify functional modules within complex networks2. A tree is a subnetwork with minimal connections. Specifically in graph theory, a tree is a graph with only one path between every two nodes. In other words, any connected graph without simple cycles is a tree. Given a connected graph, which is not a tree, one can extract a tree from it by eliminating cyclic edges. A spanning tree contains all the nodes of the graph and has (N-1) edges where N is the number of nodes in the given graph. Extracting a spanning tree gets interesting when edges of the given graph have weights. In finding the minimal/maximal spanning tree, one would ideally extract the tree whose sum of weights is minimum/maximum respectively. The weight of a spanning tree is the sum of weights given to each edge of the spanning tree. There may be several minimum spanning trees of the same weight; in particular, if all the edge weights of a given graph are the same, every spanning tree of that graph is minimal. If each edge has a distinct weight then there will be only one unique minimum spanning tree.\n\nIn this paper, we present CySpanningTree, a Cytoscape3 3 app for extracting a spanning tree from a given graph. Once the user imports a dataset, by clicking the “Create spanning tree” button of the app, a new spanning tree network is created in the network panel of Cytoscape. Historically, spanning trees are used in various applications like constructing a road network between cities with a minimum cost, as a heuristic for the traveling salesman problem (TSP), for the spanning tree network optimization protocol in networking, clustering gene expression data, etc. Three of the mentioned cases have been demonstrated in the use cases section.\n\n\nMethods\n\nCySpanningTree is the Java implementation of Prim’s4 and Kruskal’s algorithms5, using the Cytoscape 3 API and Java 7 for extracting a minimal spanning tree (MST). An MST for a given graph might not be unique, however for a given same Cytoscape session, the tie-breaking approach for selecting edges of equal weights is deterministic. The user gets the same spanning tree in a given Cytoscape session unless he reloads the network.\n\nThis tool also has a “Create Hamiltonian cycle” button which invokes the computation of the Hamiltonian cycle6. For computing this cycle, it first finds an MST using Prim’s algorithm and then performs a pre-order traversal on it. This pre-order traversal is a modified version of the depth-first search algorithm which results in a Hamiltonian path. Later, we connect the last node and the first node of this path to make a cycle. Users are recommended to run the Hamiltonian cycle algorithm on a fully connected graph to avoid missing of the edges while traversing.\n\nTable 1 has the complexities of the algorithms and the uniqueness of the outputs used in the app. Prim’s algorithm runs using adjacency list representation of the graph and thus implemented with a complexity O(V2). Kruskal’s algorithm runs using adjacency matrix of the graph and has a complexity of O(EV2(E+V)). The Hamiltonian cycle first calculates a spanning tree using Prim’s algorithm with a complexity of O(V2) and then runs depth-first search algorithm with a complexity O(E + V).\n\nThe GUI component of CySpanningTree is represented as a tabbed panel in the control panel of Cytoscape. Cytoscape takes care of loading the input network. The CySpanningTree menu (Figure 1) loads in the control panel of Cytoscape by selecting it from App menu. Currently the app runs only on connected networks. When the user tries to execute a spanning tree algorithm on an unconnected graph, an error message pops up. For weighted graphs, the user has to select the edge attribute from the drop down list (which is by default “None” that treats all edges with the same weight).\n\nPrim’s algorithm starts with a root node and hence the user is asked for the same when the Prim’s Spanning Tree button is pressed. If the user enters a node that is not in the network, the user gets an error message and the program terminates.\n\nThe resultant MST or the Hamiltonian cycle network has the same layout as that of the input network with nodes positioned at the same location and edges scaled down. When spanning tree subnetworks are created, the corresponding spanning edges are highlighted in the input network. In Figure 2, the input network is a fully connected graph of capital cities of countries in the world, containing 203 cities and 20503 connections between them. The resultant networks: “Kruskal’s Spanning Tree”, “Prim’s Spanning Tree” and “Hamiltonian Cycle” are connected graphs containing all the 203 cities and only 202, 202 and 203 edges respectively. Spanning trees are extracted as separate Cytoscape networks under the same network collection as shown in Figure 2.\n\n\nUse cases\n\nIn this section, we present the spanning tree results on use cases with datasets in four scenarios: gene expression matrix of gene expression data, building a cost efficient road network when all possible costs are known, an approximate solution to the travelling salesman problem and connecting a 10-home village with phone lines with minimum wiring. In each scenario, the contents of the network are introduced first and then extraction of spanning trees is demonstrated.\n\nThe expression levels of genes when exposed to various environmental conditions are recorded at different times with different samples. This data is called gene expression data and is analyzed to extract the similarities between genes. Gene expression data G(g→1, g→2,…,g→n) for n genes is multi-dimensional data with each g→i=(di1, di2,…,dim) for given m expression levels. Here g→i represents the ith gene and dij represents the jth expression level of this ith gene.\n\n\n\nThis data has been simulated as a graph with nodes being genes and edges being the genetic distance between them. Genetic distance is defined as the measurement of similarity between genes.\n\nEuclidean distance between genes g→i and g→j = (di1−dj1)2+(di2−dj2)2+…+(dim−djm)2\n\nFor each pair of genes, this genetic distance is calculated which gives a fully connected graph. The data set7 has been taken from the Saccharomyces Genome Database and contains expression levels of budding yeast — S. cerevisiae with a total of 6149 genes (http://downloads.yeastgenome.org/expression/microarray/Cho_1998_PMID_9702192/). Typically, it becomes difficult to visualize a large graph of 6149 nodes with each node connected to every other node in the graph. A spanning tree of the gene expression data makes it possible to visualize such a large network as shown in Figure 3.\n\nInput network: A fully connected graph of S. cerevisiae expression data\n\nNodes: Genes of S. cerevisiae\n\nEdges: Euclidean distance between genes calculated using expression levels\n\nOutput network (Figure 3): Kruskal’s spanning tree of the input gene expression data\n\nAlthough a lot of edges are removed from the network during the process of creating a spanning tree, no essential information is lost8. A spanning tree is a better way to visualize large networks compared to fully connected graphs. We observed that genes with similar functionalities are connected closely in the resultant spanning tree. Many clustering algorithms have been applied to gene expression data8,9, we are currently working on clustering using minimum spanning trees for our next release of CySpanningTree.\n\nThis dataset10 consists of nodes which are capital cities of all countries in the world and edges between them representing the distance in kilometers. These distances are measured using latitude and longitude coordinates of the cities (http://privatewww.essex.ac.uk/~ksg/data-5.html). This dataset, when imported into Cytoscape, results in a fully connected graph as the distance is calculated for each pair of capital cities. Prim’s algorithm has been executed on this dataset to produce a MST network as shown in Figure 5\n\nInput network: Fully connected graph of capitals cities as shown in Figure 4\n\nNodes: Capital cities of all countries in the world\n\nEdges: Displacement between cities\n\nOutput minimum spanning tree: Network with minimum cost such that each city is connected. Cities separated with large distances are represented with strong edges as shown in Figure 5\n\nFurthermore, this solution can be used for drawing a Hamiltonian cycle which is an approximation to the Travelling Salesman problem. Drawing a Hamiltonian cycle for a smaller network is discussed in the next subsection.\n\nThe TSP is a well-known combinatorial optimization problem. The goal is to find the shortest tour that visits each city in a given list exactly once and returns to the starting city. Though the problem statement looks simple, TSP is NP-complete11. Even though the problem is computationally difficult, a large number of heuristic solutions12 are known due to the number of applications of this problem13 like planning, logistics, DNA sequencing, predicting protein functions, etc.\n\nPre-order traversal on a minimum spanning tree is one of the heuristic solutions for TSP5,14. In this subsection, a Hamiltonian cycle is drawn for a spanning tree to show that the resultant cycle is a near solution to the TSP. The optimal TSP tour in Figure 9 is about 17% shorter than the Hamiltonian cycle obtained using spanning tree in Figure 8. On executing the Hamiltonian cycle algorithm on the input network, the software will create both Prim’s spanning tree as well as the Hamiltonian cycle. Five nodes from the above capital city network are used for the TSP use case.\n\nInput network: Fully connected graph of 5 capital cities\n\nNodes: Capital cities of countries: USA, Brazil, South Africa, India and Italy\n\nEdges: Displacement between cities shown in kilometers\n\nThis dataset consists of houses depicted as nodes and the edges are the means by which one house can be wired up to another. The weights of the edges dictate the distance between the houses. The task of the telephone company is to wire all houses using the least amount of telephone wiring possible.\n\nInput network: Houses in village depicted as graph as shown in Figure 10\n\nNodes: Houses H1 to H10\n\nEdges: Distance between the houses\n\nOutput MST: Network which connects the houses via wires with least possible wiring. Figure 11 and Figure 12 are the spanning trees obtained using Prim’s (H1 as root node) and Kruskal’s algorithm, respectively.\n\n\nSummary\n\nIn this paper, we present CySpanningTree app for Cytoscape 3. CySpanningTree fills an important need for many Cytoscape users and researchers in obtaining spanning trees across different types of networks. CySpanningTree makes effective use of the Cytoscape 3 API in extracting the subnetwork and creating it as a separate network. In the near future, we will be exploring MST based clustering and we are determined to explore more datasets whose spanning tree evaluation is significant.\n\n\nSoftware availability\n\nCySpanningTree app can be downloaded from the Cytoscape app store.\n\nhttp://apps.cytoscape.org/apps/cyspanningtree\n\nhttps://github.com/smd-faizan/CySpanningTree\n\nhttp://dx.doi.org/10.5281/zenodo.1966815\n\nhttps://www.gnu.org/licenses/lgpl.html", "appendix": "Author contributions\n\n\n\nFS and SB conceived the CySpanningTree app. NG supervised the project. FS contributed to the implementation of Kruskal’s algorithm, Hamiltonian cycle and user interface of the app. SB contributed to the implementation of Prim’s algorithm. FS and SB worked on the use cases. FS and SB wrote the manuscript. NG participated in the design of the app and in the revision of the manuscript.\n\n\nCompeting 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\nAcknowledgments\n\nThe authors would like to thank their professor Bharat.M.Deshpande for shaping and motivating their interests towards Discrete Mathematics, Scooter Morris from Cytoscape open source community for helping with Cytoscape API to extract the subnetwork in an intuitive way.\n\n\nSupplementary material\n\nCytoscape session files for use cases. Cytoscape session files (*.cys) for the TSP, world network, and 10-home village use cases.\n\nClick here to access the data.\n\n\nReferences\n\nPavlopoulos GA, Secrier M, Moschopoulos CN, et al.: Using graph theory to analyze biological networks. BioData Min. 2011; 4(10): 1–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLemetre C, Zhang Q, Zhang ZD: SubNet: a Java application for subnetwork extraction. Bioinformatics. 2013; 29(19): 2509–11. PubMed Abstract \| Publisher Full Text \| Free 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): 2498–2504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrim RC: Shortest connection networks and some generalizations. Bell System Technical Journal. 1957; 36(6): 1389–1401. Publisher Full Text\n\nKruskal JB: On the shortest spanning subtree of a graph and the traveling salesman problem. Proc Am Math Soc. 1956; 7(1): 48–50. Publisher Full Text\n\nWest DB, et al.: Introduction to graph theory, volume 2. Prentice hall Upper Saddle River. 2001. Reference Source\n\nCho RJ, Campbell MJ, Winzeler EA, et al.: A genome-wide transcriptional analysis of the mitotic cell cycle. Mol Cell. 1998; 2(1): 65–73. PubMed Abstract \| Publisher Full Text\n\nXu Y, Olman V, Xu D: Clustering gene expression data using a graph-theoretic approach: an application of minimum spanning trees. Bioinformatics. 2002; 18(4): 536–545. PubMed Abstract \| Publisher Full Text\n\nJiang D, Tang C, Zhang A: Cluster analysis for gene expression data: a survey. IEEE Trans Knowl Data Eng. 2004; 16(11): 1370–1386. Publisher Full Text\n\nGleditsch KS: Distance between capital cities. 2008. Reference Source\n\nPapadimitriou CH: The Euclidean travelling salesman problem is NP-complete. Theor Comput Sci. 1977; 4(3): 237–244. Publisher Full Text\n\nRosenkrantz DJ, Stearns RE, Lewis PM II: An analysis of several heuristics for the traveling salesman problem. SIAM J Comput. 1977; 6(3): 563–581. Publisher Full Text\n\nLenstra JK, Rinnooy Kan AHG: Some simple applications of the travelling salesman problem. J Oper Res Soc. 1975; 26: 717–733. Publisher Full Text\n\nHeld M, Karp RM: The traveling-salesman problem and minimum spanning trees. Operations Research. 1970; 18(6): 1138–1162. Publisher Full Text\n\nShaik F, Bezawada S: CySpanningTree: Hamiltonian. Zenodo. 2015. Data Source" }	[ { "id": "10256", "date": "14 Sep 2015", "name": "Shaillay Dogra", "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 have come up with a useful plug-in for cytoscape. Different algorithms have been implemented to reduced a cluttered network to a more meaningful one. Such efforts are welcome and potentially useful especially for those working in network analysis and visualization.The manuscript can be enhanced by considering the suggestions below:1. Include a schematic figure to illustrate the points mentioned in the Introduction for the benefit of a wider audience or non specialist users like experimental biologists.2. It will be helpful to intended users like experimental biologists if different algorithm choices were explained in terms of what they mean, in which case it is advised to use which particular algorithm etc.3. The author's mention that different sessions may lead to different trees. What are the potential pitfalls of this in generating results and possible different interpretations. Please discuss this aspect.4. How do the authors define genetic distance? It is not clear. Is it based on correlation value of expression of genes? Please elaborate.5. Figure 5, \"MST on world network\" - how to use a weight; for ex., 'effective distance' between cities that is a measure of air-connectivity can be used to depict 'realistic distance' than physical distance.6. More discussion on interpretation of figures 6,7 and figures 8,9 will be helpful to the readers.7. What is a way to verify that the solution is actually what it is 'supposed to be'?", "responses": [] }, { "id": "12115", "date": "29 Mar 2016", "name": "Ankush Sharma", "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 research article entitled -\"CySpanningTree: Minimal Spanning Tree computation in Cytoscape, the authors describe the app for Cytoscape version 3 that creates minimal/maximal spanning tree for a given network using network Prim’s and Kruskal’s algorithms.The CySpanningTree app appears to be useful in approximating the minimum-cost weighted perfect matching, maximum flow problems and other related issues (Supowit et al. 1980; Dahlhaus et al. 2006). The description of the proposed implementation of CySpanningTree app for Cytoscape version 3 is informative and detailed for audience. The article provides sufficient details with appropriate title and well-written abstract. Minor Concerns Some more details on usage on practical applications are strongly suggested to include in this research article as requested by Reviewer 1 in Point 2. The definition of gene expression and generalizing gene expression data in one context is not correct in section MST of gene expression data. It is highly recommended to correct it and cite appropriate research articles defining gene expression and Gene expression data.\n\nGene-gene interaction network reconstruction from gene expression needs to be detailed in methodology sections e.g. how edge weights are calculated and then used for calculation of Euclidean distance between genes. The usage of Genetic distance seems to be inappropriate in this context as it is a measure of the genetic divergence between species or between populations within a species. Please elaborate, if it is used in this context in research article. I would suggest making comprehensive figures for better readability e.g. (figure 1 and figure 2 may be merged into figure 1, Similarly figure 4,5,6,7 into figure 3, figure 8, 9 into figure 4 and figure 10, 11, 12 into figure 5) and brief description of figures in text as well as in legend will make help in better understanding of the examples and usage of the cySpanning trees.", "responses": [] } ]	1	https://f1000research.com/articles/4-476
https://f1000research.com/articles/4-133/v1	28 May 15	{ "type": "Research Note", "title": "A Case Study of the Impediments to the Commercialization of Research at the University of Kentucky", "authors": [ "Nathan L. Vanderford", "Elizabeth Marcinkowski", "Elizabeth Marcinkowski" ], "abstract": "The commercialization of university-based research occurs to varying degrees between academic institutions. Previous studies have found that multiple barriers can impede the effectiveness and efficiency by which academic research is commercialized. This case study was designed to analyze the status of the commercialization activity at the University of Kentucky via a survey and interview with a successful academic entrepreneur in order to determine the impediments the individual perceived during the commercialization process. The study also garnered insight from the individual as to how the commercialization process could be improved. Issues with infrastructure were highlighted as the most significant barrier faced by the individual. The research subject also suggested that commercialization activity may generally increase if a number of factors were mitigated. Such insight can be communicated to the administrative leadership of the commercialization process at the University of Kentucky. Long term, improving university-based research commercialization will allow academic researchers to be more active and successful entrepreneurs such that intellectual property will progress more freely to the marketplace for the benefit of inventors, universities, and society.", "keywords": [ "Research commercialization", "entrepreneurship", "intellectual property", "technology transfer", "disclosures", "patents", "license", "start-ups" ], "content": "Introduction\n\nResearch is a vital component of the mission of universities, and indeed academic institutions conduct a substantial volume of research that is funded by government, industry and philanthropic agencies. Development or the commercialization of research should also be a key component of the research mission such that novel ideas, techniques and products can enter the marketplace for the consumption and benefit of society. In order to facilitate academic-based commercialization, legislation, such as the Bayh-Dole Act, provides universities the legal framework for commercializing the research that is developed within university settings1,2.\n\nIn a commercialization survey conducted by the Association of University Technology Managers (AUTM), in 2013, United States-based institutions obtained over 5,000 new patents, executed over 5,000 licensing agreements, formed over 800 start-up companies and generated $2.75 billion in license income3. Despite this overall success, academic researchers experience many issues that obstruct the commercialization of research within higher education settings. Previous studies at academic institutions have documented the challenges to the commercialization process that include, but are not limited to: risk aversion; constraints on faculty time; lack of financial support; policy/regulation barriers; infrastructure insufficiencies; lack of a common understanding of the value of research commercialization; lack of entrepreneurial thinking among faculty; and lack of interaction and collaboration between universities and industry4–10. A previous study at the University of Kentucky found that expense, time constraints, insufficient infrastructure and lack of industry partnerships were the most common factors experienced by cancer researchers that impede research commercialization11. Ultimately, challenges to the effective and efficient commercialization of research inhibits obtaining the maximum benefit of university research in that such barriers can prevent university-based innovation from progressing to the marketplace for the benefit of inventors, universities and society.\n\nThe University of Kentucky commercializes its research through the Intellectual Property Development and Technology Transfer Office, a unit of the Office of the Vice President for Research. Through this office, the university’s research commercialization activities are historically modest compared to its benchmark institutions. The university currently ranks last among its benchmark institutions in regard to several commercialization metrics including in staffing, invention disclosures, patent applications and license/options executed (Table 1). And, growth in commercialization activity has been relatively flat from 2010–2013 with the exception of a recent increase in license income (Table 2). These data could suggest that the University of Kentucky may experience additional commercialization barriers as compared to its benchmark institutions and/or a higher magnitude of common barriers among institutions.\n\nData obtained from the fiscal year 2013 AUTM report.\n\nData obtained from the fiscal year 2010–2013 AUTM reports.\n\nThe study herein focused on understanding the impediments to commercializing research at the University of Kentucky from the perspective of a single faculty member that has been successful in the continuum of commercialization through successfully obtaining multiple patents, licensing intellectual property and forming multiple start-up companies. The rationale for conducting the study with one successful academic entrepreneur was that we believed that such a serial entrepreneur could provide more insight into the commercialization process versus someone that had more limited or no experience in commercializing research.\n\n\nMethods\n\nThe study herein is modeled closely after a similar, larger scale study conducted at the University of Kentucky specifically among cancer researchers11. The methodology and design of this study was qualitative in nature and was based on two modules: an online survey (included as Supplementary materials S1) followed by a face-to-face interview. The selection criteria for inclusion in the study was that the selected participant must be a faculty member, have an active research program and be a successful academic entrepreneur based on having obtained patents, licensed intellectual property and created start-up companies. The research subject for this study was identified through searches of publically available databases containing information on the selection criteria. For module one, data were collected and managed using the Research Electronic Data Capture (REDCap) tool. REDCap is a secure, Internet-based study-support application12. Module two data were recorded in written format during the face-to-face interview.\n\nThis study was determined to not require review by the University of Kentucky Institutional Review Board. The research subject consented to participate in the study electronically via engagement with the online survey and chose to participate in both modules of the study. The participant chose to remain anonymous beyond interaction with the investigators involved in the study.\n\n\nResults\n\nThe first series of questions aimed to assess the subject’s category of research, professional productivity and the perspective he has on research commercialization. The subject classified his research as “translational;” he felt satisfied with his level of professional productivity in terms of publishing research manuscripts, obtaining grant funding and other means of academic productivity; and he indicated that he intends to commercialize additional research in the future. Despite believing that research commercialization is important in the academic setting and that his research field values such work, he feels that the University of Kentucky places little emphasis on and thus does not greatly value research commercialization (Table 3).\n\nThe second set of questions addressed the research subject’s perceived impediments to commercializing research. The subject responded that risk, lack of investors, commercialization infrastructure, unsupportive university and federal policies, and “other barriers not listed” prohibited his efforts to effectively and efficiently commercialize research (Table 4). In the face-to-face interview, the subject indicated that the “other barriers” included major prohibiting factors such as the lack of university support in areas of market analysis, grant development, and navigating legal matters including conflict of interest and intellectual property ownership issues. Of these “other” items, we had anticipated that such factors could be captured under the commercialization infrastructure and/or policy categories of answer choices in the survey. Ultimately, the subject indicated that infrastructure issues are the most significant factors that impede research commercialization at the University of Kentucky. The subject also discussed how some of these barriers are more challenging and more difficult to overcome and that he felt that the barriers he has encountered are different at other universities. Thus, similar to the previous study among cancer researchers11, these data suggest that this faculty member experiences multiple barriers in the commercialization process. Additionally, in comparison with previous studies4–11, the data may suggest that not all barriers are consistent or common between individual faculty members (for example, expense, time constraints, insufficient infrastructure, and lack of industry partnerships were the most common barriers experienced among University of Kentucky cancer researchers11).\n\nThe final set of questions were meant to determine which impediments would need to be overcome in order to increase faculty participation in research commercialization. Interestingly, the subject indicated that the barriers in the commercialization process do not deter him from attempting to commercialize his research, however, he believed that reducing/mitigating all the potential barriers surveyed, other than addressing royalty pay to inventors, would enhance research commercialization activity at the University of Kentucky (Table 5). The subject also indicated that he would utilize outside (off campus) commercialization resources to lower the barriers he faces in order to improve his commercialization efforts. These data are similar to the feelings reported by cancer researchers11 in which respondents believed that a greater number of mitigating factors compared to the identified barriers in the previous set of questions would presumably increase commercialization activity.\n\n\nConclusion\n\nThis case study investigated the mindset of one successful academic entrepreneur at the University of Kentucky in relation to research commercialization and in context with the university’s general commercialization activity. The general status of the institution’s commercialization activity is modest relative to its benchmark institutions and stagnant in growth over time. The research subject identified several factors that generally impede research commercialization and that mitigating many factors may increase commercialization activity. Infrastructure was pinpointed as the major issue impeding research commercialization at the university. While generally fitting with the impediments found at other universities and among cancer researchers at the University of Kentucky4–11, the results suggest that not all barriers are common or consistent between faculty and that some impediments may be more prohibitive than others.\n\nThese data can be shared with the University of Kentucky’s Intellectual Property Development and Technology Transfer Office and the Office of the Vice President for Research and used as a guide to make changes that will improve the research commercialization process. The research subject’s comments regarding commercialization infrastructure may be particularly important to address in order to enhance commercialization activity at the university. Additionally, similar work could be conducted at and among other institutions. For example, a survey similar to the one herein and that used in the prior study11 could be incorporated into the yearly AUTM licensing survey in order to gauge, on a much broader scale, the impediments to academic research commercialization as well as to understand how other institutions are mitigating such impediments. Understanding how institutions that are highly successful in commercializing research mitigate barriers in the process would be greatly beneficial to institutions that have low to modest commercialization activity.", "appendix": "Author contributions\n\n\n\nNLV and EM conceived and designed the study; conducted the study; analyzed the data; and wrote the paper. This research project was completed, in part, to fulfill the requirements of EM’s Bachelor of Science degree in Agricultural Biotechnology.\n\n\nCompeting interests\n\n\n\nThe authors have no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary materials\n\nSupplementary materials 1\n\nResearch Commercialization Survey\n\nClick here to access the data.\n\n\nReferences\n\nLoise V, Stevens AJ: The Bayh-Dole Act turns 30. Sci Transl Med. 2010; 2(52): 52cm27. PubMed Abstract \| Publisher Full Text\n\nThursby JG, Thursby MC: Intellectual property. University licensing and the Bayh-Dole Act. Science. 2003; 301(5636): 1052. PubMed Abstract \| Publisher Full Text\n\nAssociation of University Technology Managers. AUTM U.S. Licensing Activity Survey: FY2013. Deerfield, IL: Hippenmeyer P, Hawkins S, Mroz MA, Robertson R, Ruey N, and Stevens AJ. 2015. Reference Source\n\nAtlantic Canada Opportunities Agency. University Research Activity Private Sector Collaboration and Commercialization of Research in an Academic Environment: Memorial University of Newfoundland as a Case Study. New Brunswick, Canada: Locke W, Lynch S, Girard B. 2002. Reference Source\n\nERA-Net NEURON. Transferring Technology from Bench to Bedside: Practices, Barriers, Policies. Bonn, Germany: Meyer M, Glod F. 2011. Reference Source\n\nKlein R, Haan UD, Goldberg AI: Overcoming Obstacles Encountered on the way to Commercialize University IP. Journal of Technology Transfer. 2010; 35(6): 671–679. Publisher Full Text\n\nLawson C: Academic Patenting: The Importance of Industry Support. J Technol Transf. 2013; 38(4): 509–535. Publisher Full Text\n\nO’Shea RP, Allen TJ, Chevalier A, et al.: Entrepreneurial orientation, technology transfer and spinoff performance of U.S. universities. Research Policy. 2005; 34(7): 994–1009. Publisher Full Text\n\nPowers JB, Campbell EG: Technology Commercialization Effects on the Conduct of Research in Higher Education. Res High Educ. 2011; 52(3): 245–260. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSwedish Institute for Growth Policy Studies. Commercialization of Life-Science Research at Universities in the United States, Japan and China. Ostersund, Sweden: Nilsson AS. 2006. Reference Source\n\nVanderford NL, Weiss LT, Weiss HL: A survey of the barriers associated with academic-based cancer research commercialization. PLoS One. 2013; 8(8): e72268. 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): 377–381. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8827", "date": "05 Jun 2015", "name": "Jessica Silvaggi", "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 goal of the authors was to document any impediments to the commercialization process at the University of Kentucky. The study was spurred by the findings that UK significantly underperforms when compared to peer institutions of similar size. A previous study was published in 2013 regarding this same topic at UK for the commercialization of cancer research. In this study, similar surveys were conducted with one single entrepreneurial respondent as opposed to a larger cohort of cancer researchers in the previous UK study.They report that the major issue impeding commercialization, according to the respondent, is the lack of appropriate infrastructure at the university. The results were apparently similar to those found in this 2013 paper. I found this topic to be of great interest in comparing the performance to schools of similar benchmarks. The results could be helpful to many institutions that are underperforming and looking to boost commercialization. The survey questions used could be useful for any institution to keep track of the satisfaction of the researchers in regard to commercialization of their technologies, and to assess the adequacy of the performance of the technology transfer office.Major concerns/RevisionsWhile I think the survey questions used were reasonable, I have doubts about the validity of asking a single entrepreneur the survey questions. It comes across as if the same 2013 study was repeated, but with only one person. I don’t feel that one person is a significant number for this type of paper to make conclusions with. N=1 doesn’t seem to be a true scientific sampling. Unfortunately, if more respondents were utilized, then this paper would appear to be a repeat of the earlier study. Perhaps another angle would be to survey the staff at UK as to why they are underperforming. Asking the tech transfer professionals and other research administrators would also be informative and provide a different angle to the original survey of cancer researchers. Or perhaps a broader survey of other departments would be helpful, rather than only focusing on the cancer department. I feel that several other angles could have also been explored in this paper to contrast or follow up on the work in the 2013 paper about UK. After the first study were any changes made based on the conclusions? Did the number of staff change? Did any of the infrastructure change at UK? Did UK do anything to increase the level of commercialization? There are many other interesting questions that could be explored rather than repeating the study over again with one participant. If the university has not responded to the previous study in any fashion, I find that of great interest. Why hasn’t the technology transfer office been expanded and further supported? Why is there no incentive by those in charge of this area to revamp the office? There has been a large push at many universities to further promote entrepreneurship in faculty and students and support commercialization. In some online information it states that in the past UK was pushing to be a top 20 university by 2020. I am curious to know why there has been no change in the recent years.Minor RevisionsWhen comparing benchmarks, another important factor missing in the table is the amount of research dollars. There is a ratio of expected disclosures per millions of dollars of research which varies, but is quoted in several locations at 1 disclosure per $1.5M-$3M and 1 start-up company per ~$100M. Knowing the amount of research dollars going into UK would help to show a lack of productivity with inventions at the university. This information should be available through AUTM if not through the technology transfer office directly. Based on some online information the research expenditures at UK appear to be in the hundreds of millions of dollars. Assuming ~$400M in expenditures, one would expect about 130 disclosures. The UK website cites 84 disclosures in 2014.Summary:I feel that major revisions are necessary for this paper to add some new information and expand the scope of the study. The results here do not seem to add onto those previously found but simply confirm the previous findings with the survey of one entrepreneurial researcher. A different sample of respondents would greatly enhance this paper such as asking researchers from all departments, or focusing on the staff involved in technology transfer, commercialization and entrepreneurship. Or perhaps if changes were made by the tech transfer office and these did not have an effect, this would be good to know for other offices. There appears to be something missing from this story. This paper would be of interest if revamped to add a new twist distinct from the 2013 paper.I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.", "responses": [ { "c_id": "1496", "date": "04 Aug 2015", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Silvaggi, We greatly appreciate the time you dedicated to review our article. We found your comments very helpful as we revised the article. We are happy that you found this topic of interest and applicable to other institutions. Before reading the new version, we would like to address some of your comments and concerns. First, we appreciate your concerns regarding the “n-of-1” design of the study. As suggested by each reviewer, we have added two additional respondents to the study and this did indeed generate several additional important points regarding the issues related to the low commercialization activity at the University of Kentucky. Second, we appreciate your comments regarding how this work could be construed as a repeat of the prior study. We agree with your assertions that other issues are likely involved in the low commercialization activity at the university and we agree that surveying a mix of stakeholders involved in the research commercialization process could aid in uncovering other issues. However, we would like to clarify that this case study was, in fact, specifically designed and conducted as a supplement to the prior work looking at the barriers associated with cancer research commercialization at the University of Kentucky (we have made this clarification in the new version of the article). The prior study’s conclusions were based solely on the responses to the survey and individual interviews were not conducted. The intent of the current study was to collect more specific, supplemental information through interviews. Given this intent and the study’s supplemental nature, we have purposefully not incorporated new research questions into the current case study. We believe that future work would best address many of your comments which add additional research questions that would help understand other issues that may be connected to the low research commercialization activity. You have also made several important comments regarding whether the university has made any changes in the research commercialization process since the 2013 study. In fact, few changes have been made and that is one reason why it was important to conduct this case study; it was important to obtain specific comments from “successful” entrepreneurs so that these individuals could identify specific issues in the system. Finally, to offset additional concerns, we have included a description of the limitations of this case study in the methods sections. Per your minor point, we have now also added research expenditures to Table 1. In closing, we hope that you will review this revised version of the article in light of our changes based on your comments as well as those of the other two reviewers (we hope that you will read the other reviewers’ comments as well as our response to those reviews) and in relation to its intended purpose of being a supplemental component to the prior study at the University of Kentucky. Sincerely,Nathan L. Vanderford and Elizabeth Marcinkowski" } ] }, { "id": "9005", "date": "17 Jun 2015", "name": "Scott Crick", "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 study attempted to identify potential impediments to commercialization of research at the University of Kentucky. The authors point out that, according to data from AUTM, the University of Kentucky ranks near the bottom in a number of key metrics of commercialization when compared to its benchmark institutions. The goal of the study was to identify barriers to commercialization. These findings will then be brought to administration in hopes of rectifying the problem.I think there are several issues with this work that, if addressed properly, will greatly strengthen its impact and utility not only at UK but also at other university technology transfer organizations. The first issue I have with the work is the inclusion of only a single faculty member with entrepreneurial experience. The rationale given was that that person could provide more insight into the process as a whole. Looking at the data referenced in Table 1, the most striking discrepancy between UK and benchmark institutions is the number of invention disclosures. When the data are normalized to the number of invention disclosures per institution, UK appears to be making good use of the inventions that are disclosed to them. It seems to me that one of the major issues at UK is simply getting inventors. While I agree that the entrepreneurial faculty member could have a lot of insight into issues after the first step, I wonder how much insight this person provided with regards to why people are not disclosing inventions. I am certain it would be possible to identify faculty that publicly disclosed potentially valuable assets without ever filing an invention disclosure, and I would suggest understanding those issues are 1) extremely important for improving commercialization at UK, and 2) more easily addressable at an institutional level than some other potential barriers.The second issue with relying on input from a single faculty member is that barriers are variable (as even mentioned in the article) not only between individuals, but between disciplines and sub-disciplines as well. Although the individual faculty member can identify barriers he/she has experienced, it would be disingenuous to suggest these barriers and the relative weight given to each of them is an accurate reflection of the system as a whole. A minor point....It would also be very helpful to know general field of research of the faculty member interviewed. \"Translational\" is very broad. Is he/she in pharma, biotech, medical devices, engineering...?I have two recommendations to strengthen this article. My first recommendation is that the authors include in the case study at least two other faculty members with technology commercialization experience (not necessarily on par with the initial interviewee). I would also suggest that these faculty be from distinct research areas and departments. Although still qualitative, it would be insightful to see if these people with different types of technology, different department makeup, and very likely different experiences and backgrounds still identify the same barriers.My second recommendation is to limit the scope to barriers that are perceived after invention disclosure. It appears as though there is an issue at UK (which I should say is certainly not unique) that I suspect has to do with education of potential inventors such that a number of these people are not even considering that their technology might have commercialization potential. A follow-up study trying to tease apart this issue would be interesting and may have broader appeal.", "responses": [ { "c_id": "1495", "date": "04 Aug 2015", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Crick, We thank you for taking the time to review our case study. Your critique has helped shape our current version of the article. We would like to respond to some of your comments. First, per your suggestion, we have expanded our sample size by two respondents. This expansion lead to the collection of a significant amount of additional crucial comments that are important to understand about the issues related to the low commercialization activity at the University of Kentucky. Second, in regard to your question about the respondents’ field of research, we understand your desire to have more information regarding the respondents’ research areas, but we feel that for confidentiality purposes, we cannot be any more specific; given the small sample size of “successful” academic entrepreneurs at the University of Kentucky, reporting a respondent’s specific research field could potentially allow for the identification of the subject. Next, you made very insightful comments regarding the desire to better understand the issues involved in the university’s low invention disclosure rate. We agree with your points and we would also like to address this issue. With that said, however, we would like to clarify that this case study was designed and conducted as a supplement to the prior study looking at the barriers associated with cancer research commercialization at the University of Kentucky; the sole intent of the current study was to collect more specific, supplemental information through interviews with the respondents (we have now explicitly stated this intent in the new version of the article). Given this intent and the study’s supplemental nature, we have purposefully not incorporated new research questions into the current case study. We feel that expanding the current study would likely change its primary intent. We do hope to address additional research questions, including understanding issues related to the low disclosure rate, in future work. Lastly, to offset additional concerns, we have included a description of the limitations of this case study in the methods sections. In closing, we hope that you will review this revised version of the article in light of our changes based on your comments as well as those of the other two reviewers (we hope that you will read the other reviewers’ comments as well as our response to those reviews) and in relation to its intended purpose of being a supplemental component to the prior study at the University of Kentucky. Sincerely,Nathan L. Vanderford and Elizabeth Marcinkowski" } ] }, { "id": "9270", "date": "01 Jul 2015", "name": "Evan Facher", "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 study by Vanderford and Marcinkowski attempts to identify challenges to the commercialization of innovations discovered at their institute of higher learning, the University of Kentucky. The goal of this work is to improve the sluggish local climate for translation of discoveries by communicating the findings resulting from this effort back to administrative leadership with the hope that the identification of these impediments leads to real change. By deploying a survey and subsequent interviews the authors plan to generate learnings sufficient to form the basis of their recommendations. Movement of research discoveries from academic institutes to the marketplace is important not only for these discoveries to have their societal impact but also to drive economic growth of a region. The juxtaposition of most new jobs created over the last two decades coming from startup companies and much innovation from academia serving as the impetus for these new entities hints at the regional criticality of a university being able to successfully translate its discoveries into products reaching the market. While the Vanderford and Marcinkowski article has a solid thesis and very good intent, it is however limited (in this reviewers mind) by a handful of items. First, the conclusions the authors generate are based on the survey and interview of a single academic entrepreneur. The data achieved from this individual is certainly very valuable but it might not be reflective of the other individuals on campus that have had experiences with commercialization of university-based research. The challenges described by this individual on campus ring true of the commercialization issues found throughout other academic organizations; however due to the small sample size it is hard to draw solid conclusions for the University of Kentucky as a whole. As one of the aims of the study is to report findings back to the administration with the goal of impacting change, I would suggest that the senior leadership, while sympathetic to the challenges expressed by the authors, will not institute changes based on such a small set of data that may not accurately reflect the general experiences of its academic entrepreneurs in totality. I would recommend that the authors attempt to increase the number of study subjects to enhance the power of their research. In addition to surveys/interviews with academics that have successfully commercialized their innovations on campus, I would strongly urge the authors to include a set of individuals that have had unsuccessful experiences as well. Learnings from this cohort might provide an additional set of data to further drive the goals of enhancing translation at their institute. The second item I would suggest the authors to contemplate in their assessments is expanding their perspectives on reasons for an innovation not being commercialized. It seems as if the main focus of the surveys/interviews is on structural elements involved in the workflow for moving a university idea to commercialization. It is often that the main reason for the lack of translation has nothing to do with the internal processes for moving the opportunity forward but rather that the innovation never really addressed a true market need despite the solid academic research. As such, regardless of the view of commercialization risk, the number of investors, the supportiveness of university policies or facilities/staff to advance the process, the idea itself is not commercialized because it does not contain a value proposition for any outside entity to take hold of. It is imperative to identify a product-market fit for an innovation as not all concepts should become companies and not all ideas impact the market. A strong technology translation capability cannot make up for an opportunity without a direct connection to an existing customer pain-point, which is only identified through a process of customer discovery outside of the university’s walls. Lastly, from a practical perspective, I would recommend that as part of the engagement with a larger set of academic entrepreneurs on campus (both successful and unsuccessful), the survey/interview deployed by the authors engage these individuals in soliciting programs, opportunities, efforts, and ideas to improve the existing stagnant innovation culture on campus. I believe by engaging these “customers” on campus in dialogue focused on solving the challenge, administrative support will be more easily achieved. In conclusion, I believe the authors are on the right track and that their efforts have significant merit. I would urge them to continue their work, expand the sample size, examine a bit broader set of reasons why the problems exist and work with their respondents to improve the situation.", "responses": [ { "c_id": "1494", "date": "04 Aug 2015", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Facher,Thank you for your time and comments. Your critique has helped guide us through revising the article. We would like to directly respond to some of your comments. We would first like to clarify that this case study was designed and conducted as a supplement to the prior study looking at the barriers associated with cancer research commercialization at the University of Kentucky (we have made this clarification in the new version of the article). The prior study’s conclusions were based solely on the responses to the survey as individual interviews were not conducted. Thus, the intent of the current study was to collect more specific, supplemental information through interviews with “successful” entrepreneurs. Given this intent and the study’s supplemental nature, we have purposefully not incorporated new research questions into the current case study. We have, however, at your suggestion, expanded the sample size by two respondents and this generated a number of additional significant comments that are important to understand regarding the university’s low commercialization rate. We agree with your assertions that additional reasons for low commercialization activity likely exist and that expanding our research questions and sample size to include a mix of stakeholders involved in the research commercialization process (including faculty that have been unsuccessful at commercializing their research and staff/administrators of the commercialization process, etc.) would aid in uncovering other issues. We would like to note, however, that expansion of this current work would change the intent (described above) and design of our study. Therefore, we have maintained the overall intent and design of the study other than adding the additional respondents. We believe that this design will allow us to bring some additional closure to the prior study. Lastly, to offset additional concerns, we have included a description of the limitations of this case study in the methods sections. In closing, we hope that you will review this revised version of the article in light of our changes based on your comments as well as those of the other two reviewers (we hope that you will read the other reviewers’ comments as well as our response to those reviews) and in relation to its intended purpose of being a supplemental component to the prior study at the University of Kentucky. Sincerely,Nathan L. Vanderford and Elizabeth Marcinkowski" } ] } ]	1	https://f1000research.com/articles/4-133
https://f1000research.com/articles/4-469/v1	04 Aug 15	{ "type": "Case Report", "title": "Case Report: Pheochromocytoma-related catecholamine cardiomyopathy with successful outcomes after orthotopic heart transplantation", "authors": [ "Jagdeesh Ullal", "Wayne Old", "Marc L. Silverberg", "Christine Nguyen-Buckley", "Rebecca E. Miller", "Donald Richardson", "Wayne Old", "Marc L. Silverberg", "Christine Nguyen-Buckley", "Rebecca E. Miller", "Donald Richardson" ], "abstract": "Pheochromocytomas (rare catecholamine-producing neuroendocrine tumors) have many different manifestations, and complications can occasionally include myocardial infarction and cardiomyopathy. In a majority of cases, cardiomyopathy reverses following medical or surgical treatment of the pheochromocytoma. We report a case of a 28-year-old male patient with preoperative diagnosis of pheochromocytoma and for whom a successful adrenalectomy revealed a benign pheochromocytoma. The patient had decompensation of heart failure and subsequent needed heart transplantation for irreversible cardiomyopathy; this gave a good outcome three years post-transplant. The heavy burden of atherosclerosis and fibrosis in a young patient with few cardiac risk factors and the irreversible cardiac damage are unique features of this case. This is also the first report (to our knowledge) of a patient with a pheochromocytoma that was surgically resected but who subsequently needed cardiac transplantation. We conclude that catecholamine-induced cardiomyopathy may be irreversible if there is structural damage to myocytes despite adequate medical and surgical treatment of a pheochromocytoma.", "keywords": [ "Pheochromocytoma", "Heart transplantation", "catecholamine cardiomyopathy" ], "content": "Introduction\n\nPheochromocytomas are rare catecholamine-producing neuroendocrine tumors. The common signs and symptoms of these tumors (headaches, sweating, palpitations and hypertension) can be attributed to the direct effects of catecholamines at various receptor sites throughout the body1–4. Catecholamine-induced cardiomyopathy is a potentially deadly outcome in patients with pheochromocytomas; it is caused by direct injury to the cardiac myocardium by catecholamines2,3,5,6. Histological changes found in catecholamine-induced cardiomyopathy are characterized by progression from diffuse edema and mild changes in the nuclei of myocytes to fibrotic changes with inflammatory infiltrates, granular cytoplasm, and contraction band necrosis7.\n\nDecreased ejection fraction (EF; measurement of how much blood is being pumped out of the left ventricle of the heart with each contraction) in those with catecholamine-induced cardiomyopathy is attributed to both myofibrillar damage and down-regulation of beta 1 and 2 adrenergic receptor, which leads to dilated cardiomyopathy in most cases but rarely to hypertrophic cardiomyopathy7,8. Damage to the myocardium is a result of enhanced lipid mobility leading to increased atherosclerosis, hypoxia during coronary vasospasm, increased calcium influx due to changes in the permeability of the sarcolemmal membrane, and free radical insult by the oxidized products of catecholamines7,8. Additionally, catecholamines are thought to stimulate protein synthesis that may contribute to left ventricular hypertrophy independently of pressure overload8. Changes seen in catecholamine-induced cardiomyopathy are usually reversible by surgical excision of the tumor or medical adrenergic blockade8. In the case presented here, removal of the pheochromocytoma did not result in reversal of the cardiomyopathy. Thus, the patient underwent orthotopic heart transplantation, which gave a successful outcome and stable left ventricular systolic function at three years.\n\n\nPresenting concerns\n\nA 28-year-old Caucasian man presented in July 2008 with dyspnea, cough and chest pain. He had been treated for pneumonia a month before, and computed tomography (CT) had revealed a right adrenal mass. Elevated catecholamines confirmed pheochromocytoma (Table 1). Upon further questioning the patient admitted to right shoulder pain for the past year, paroxysmal headaches for 10 years, and episodes thought to be “panic attacks” with tachycardia, palpitations, diaphoresis and red-purple discoloration of extremities. He had had a myocardial infarction (MI) three years previously, for which he had received a bare metal stent placed to the left anterior descending artery; his EF was 30%. While myocardial infarctions can cause reduction in EF, this was unusually low for the extent of his MI and for his age. Other conditions included hypertension, with blood pressure occasionally exceeding 130/90 mmHg, dyslipidemia and a pyloric stenosis repair in childhood. He used no tobacco, was a vegetarian and had a family history of hypertension and obesity.\n\n\nClinical findings\n\nPhysical examination revealed a lean appearing, young male who was tachycardic with scattered rhonchi in bilateral lung fields, tenderness to palpation in the right upper quadrant of the abdomen and right lower chest, and bilateral minimal edema of ankles. He had no neurofibromata or café-au-lait spots. His extremities were observed to turn bluish and pale during spells of anxiety accompanied by modest hypertension of no more than 160/85 mmHg. There were no Cushingoid features. The thyroid gland was not enlarged and there were no nodules. Vital signs were as follows: temperature 37°C, blood pressure 128/83 mmHg, heart rate 120 beats/min, respirations 20/min, and oxygen saturation of 99% on room air. Brain natriuretic peptide (BNP) levels were 3810 pg/ml (normal range 34–42) pg/ml. Lipid analysis showed total cholesterol levels of 165 mg/dl (normal range 110–200 mg/dl), HDL 30 mg/dl (normal range 40–59 mg/dl), LDL 116 mg/dl (normal range 50–99 mg/dl), and triglycerides 97 mg/dl (normal range 40–149 mg/dl). Catecholamines remained significantly elevated (Table 1). Electrocardiogram showed sinus tachycardia, biatrial enlargement, left ventricular hypertrophy with repolarization abnormalities and old anteroseptal myocardial infarction. Chest X-ray showed bilateral pulmonary parenchymal densities and cardiomegaly. Initial ejection fraction was 10% (Table 2) and cardiac catheterization revealed decreased cardiac index and elevated pulmonary capillary wedge pressure (Table 3).\n\nHe began treatment with phenoxybenzamine 10 mg twice daily, metyrosine 240 mg four times daily and carvedilol 6.25 mg twice daily. A week later, phenoxybenzamine and carvedilol was withheld due to the development of hypotension. He subsequently developed ischemic hepatitis (shock liver) with coagulopathy and renal failure. On day 11 of hospital stay, he experienced a catecholamine release episode with right shoulder pain, tachycardia, diaphoresis, and acrocyanosis of distal extremities. Because his cardiac index had decreased to 1.4 l/min/m2, he began treatment with labetalol drip at a infusion rate of 1mg/min, an intra-aortic balloon pump was placed and he was intubated. With improvement of cardiac indices he was thought to be ready for surgery. On day 15 he underwent right adrenalectomy, which revealed an 8.2 × 8.1 × 4.2 cm retroperitoneal pheochromocytoma with benign pathology, with chromogranin- and synaptophysin-positive cells (Figure 1). He was given fluid and blood resuscitation, and placed on infusions of norepinephrine 0.1 mcg/kg/min, vasopressin 0.1 units/min, epinephrine 1 mcg/min and milrinone 0.75 mcg/kg/min, which were titrated as needed. His ejection fraction improved to 15–20% and cardiac output improved to 5 l/min, allowing removal of the intra-aortic balloon pump and extubation. The post-operative course was complicated by hypotension and respiratory distress. When his ejection fraction had improved to 20–25% (Table 2), he was classified as New York Heart Association Stage IV, class D, his medical management was planned and he was discharged 34 days after initial presentation.\n\nThere was no evidence of any multiple endocrine neoplasia syndromes, and he was negative for the Ret proto-oncogene mutation. Initial endocrine work up done during the first week of his hospital stay indicated that he had mild secondary hyperparathyroidism with a PTH level of 80 pg/ml (12–65 pg/ml), which was treated with cholecalciferol 1000 IU once daily with the first week of his stay and the PTH level normalized when rechecked before discharge. The calcitonin level was normal. The chromogranin level was 40 nmol/l (0–5 nmol/l). The urine-free cortisol level was 297 µg per 24 hours, with a detectible ACTH level of 8 pg/ml. Plasma and urine dopamine levels were normal and there was no elevation of urine 5-hydroxyindole acetic acid (which would indicate a serotonin-secreting tumor).\n\nThe patient was readmitted 16 days after discharge with symptoms of cough, lethargy and oliguria. He received intravenous fluids for vomiting and diarrhea. He was immediately intubated for respiratory distress and placed on dopamine infusion. Temperature was 36.7°C, heart rate was 104 beats/min, blood pressure was 112/72 mmHg and oxygen saturation was 100% on Fraction of inspired oxygen (FiO2) 80%. On examination, he was tachycardic with grade III/VI systolic ejection murmur, hepatomegaly and diminished peripheral pulses. Chest X-ray showed right-sided pleural effusion. Ejection fraction was 20% (see Table 2) and cardiac index was decreased to 1.36 l/min/m2 (Table 3). He was treated with milrinone infusion at a rate of 0.75 mcg/kg/min which was titrated and the patient was weaned from mechanical ventilation. However, he decompensated in the following days, requiring re-intubation and placement of an intra-aortic balloon pump, and was listed as status 1A for heart transplant. Approximately 3 weeks after the second admission a donor heart became available. He tolerated the procedure and was placed on low-dose epinephrine, vasopressin and milrinone. The patient’s explanted heart showed cardiomegaly (mass 380 g), with biventricular cardiac myocyte hypertrophy (Figure 2) and dilation. Septal and left ventricular white-tan fibrous scarring consistent with prior ischemic injury was evident (Figure 3). Most coronary vessels showed various degrees of eccentric atherosclerotic stenosis (Figure 4). No occlusive or thromboembolic lesions were discovered. Post-operative echocardiogram showed normal systolic function, with ejection fraction 65% (Table 2). He has been clinically stable since discharge in October 2008, with an ejection fraction of 60% at 2.5 years (Table 2) and normal cardiac pressures at three-year follow-up (Table 3). Seven years after the initial event, there has been no recurrence of the pheochromocytoma and the transplanted heart remains at optimal function.\n\nDuring each cardiac catheterization, the cardiac index, pulmonary artery pressure and pulmonary capillary wedge pressure was obtained.\n\n\nDiscussion\n\nTo the authors’ knowledge there are no reports to date of a catecholamine-induced cardiomyopathy that has failed to improve following appropriate treatment and surgical removal of a pheochromocytoma. Our case is unique in that the patient had successful surgical excision of the catecholamine-secreting tumor without reversal of left ventricular failure, and because the presence of atherosclerosis and fibrosis in the coronary vessels was substantial. This is the first (to the authors’ knowledge) reported incident of catecholamine-induced cardiomyopathy requiring orthotopic heart transplant following adequate treatment of a known pheochromocytoma. In three previous cases, the diagnosis of pheochromocytoma was not made until after the patients had undergone an orthotopic heart transplant9,10. Two additional reports involved a left ventricular assist device and intra-aortic balloon pump with extracorporeal membrane oxygenation as a bridge to myocardial recovery11,12.\n\nThe patient reported here had a significant degree of atherosclerosis of the coronary arteries, particularly given his age of 28 years. This brings to light the importance of considering the effects of coronary artery disease both independently and in the setting of catecholamine-induced vasospasm. The persistent elevation of catecholamines may accelerate the progression of atherosclerosis and development of fibrosis in the coronary vessels. Additionally, ischemic damage to the myocardium caused by coronary artery disease may worsen the prognosis of catecholamine-induced cardiomyopathy and decrease the chances of improving left ventricular function following resection of a pheochromocytoma.\n\nConsideration of possible residual tumor burden or the presence of metastases is appropriate in this circumstance that left ventricular function fails to improve following surgical resection. However, this patient has had no clinical evidence of elevated catecholamines and there has been no clinically significant elevation of follow-up catecholamine levels.\n\nWe conclude that catecholamine-induced cardiomyopathy may be irreversible if there is structural damage to myocytes despite adequate medical and surgical treatment of a pheochromocytoma. In such cases patients may have a positive long-term outcome with orthotopic heart transplant and sustain normal left ventricular function following transplant. It is also important to consider other contributing factors to myocardial damage, including pre-existing atherosclerosis and how the presence of persistently elevated catecholamines may exacerbate known coronary artery disease.\n\n\nConsent\n\nThe patient has provided written informed consent for the publication of his clinical details and clinical images.", "appendix": "Author contributions\n\n\n\nJU wrote the manuscript, revised and edited the final version; WO wrote the manuscript; DR revised and edited the manuscript; MLS provided images of the pathology and the narrative for the images; CNB wrote the manuscript, reviewed the charts for details of the case; REM wrote the manuscript, reviewed the charts for details of the case.\n\n\nCompeting 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\nReferences\n\nGuerrero MA, Schreinemakers JM, Vriens MR, et al.: Clinical spectrum of pheochromocytoma. J Am Coll Surg. 2009; 209(6): 727–732. PubMed Abstract \| Publisher Full Text\n\nBrilakis ES, Young WF Jr, Wilson JW, et al.: Reversible catecholamine-induced cardiomyopathy in a heart transplant candidate without persistent or paroxysmal hypertension. J Heart Lung Transplant. 1999; 18(4): 376–380. PubMed Abstract \| Publisher Full Text\n\nKassim TA, Clarke DD, Mai VQ, et al.: Catecholamine-induced cardiomyopathy. Endocr Pract. 2008; 14(9): 1137–49. PubMed Abstract \| Publisher Full Text\n\nLenders JW, Eisenhofer G, Mannelli M, et al.: Phaeochromocytoma. Lancet. 2005; 366(9486): 665–675. PubMed Abstract \| Publisher Full Text\n\nBybee KA, Prasad A: Stress-related cardiomyopathy syndromes. Circulation. 2008; 118(4): 397–409. PubMed Abstract \| Publisher Full Text\n\nRoghi A, Pedrotti P, Milasso A, et al.: Adrenergic Myocarditis in Pheochromocytoma. J Cardiovasc Magn Reson. 2011; 13(1): 4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSardesai SH, Mourant AJ, Sivathandon Y, et al.: Phaeochromocytoma and catecholamine induced cardiomyopathy presenting as heart failure. Br Heart J. 1990; 63(4): 234–237. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrejbisz A, Lenders JW, Eisenhofer G, et al.: Cardiovascular manifestations of phaeochromocytoma. J Hypertens. 2011; 29(11): 2049–2060. PubMed Abstract \| Publisher Full Text\n\nDalby MC, Burke M, Radley-Smith R, et al.: Pheochromocytoma presenting after cardiac transplantation for dilated cardiomyopathy. J Heart Lung Transplant. 2001; 20(7): 773–5. PubMed Abstract \| Publisher Full Text\n\nWilkenfeld C, Cohen M, Lansman SL, et al.: Heart transplantation for end-stage cardiomyopathy caused by an occult pheochromocytoma. J Heart Lung Transplant. 1992; 11(2 Pt 1): 363–366. PubMed Abstract\n\nWestaby S, Shahir A, Sadler G, et al.: Mechanical bridge to recovery in pheochromocytoma myocarditis. Nat Rev Cardiol. 2009; 6(7): 482–487. PubMed Abstract \| Publisher Full Text\n\nSuh IW, Lee CW, Kim YH, et al.: Catastrophic catecholamine-induced cardiomyopathy mimicking acute myocardial infarction, rescued by extracorporeal membrane oxygenation (ECMO) in pheochromocytoma. J Korean Med Sci. 2008; 23(2): 350–354. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9802", "date": "01 Sep 2015", "name": "Andreas Moraitis", "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 case report describes an unusual complication of chronic catecholamine excess due to a pheochromocytoma in a young patient. Although the symptoms of catecholamine excess had been present for over a decade, the diagnosis was missed even after he suffered a myocardial infarction 3 years prior to the diagnosis. It addresses an important topic, the necessity of screening for secondary causes of atherosclerosis and hypertension in young individuals.Below are more specific comments by section:Abstract:The conclusion that catecholamine-induced cardiomyopathy may be irreversible if there is structural damage to myocytes despite adequate medical and surgical treatment of a pheochromocytoma is a heavy statement, considering the fact that stress induced cardiomyopathy is usually self-limiting. In the current case the reason for the irreversible damage was primarily the prolonged severe catecholamine excess that had remained undiagnosed for many years.Introduction:More information about the pathophysiology of accelerated atherosclerosis and cardiac injury due to chronic catecholamine excess would be useful.Clinical findings:The authors should explain why long acting alpha adrenergic blockade agents were used instead of other short acting agents. In severe cases of catecholamine induced cardiomyopathy with low EF, use of short acting iv calcium channel blockers (nicardipine), is by far safer (short acting, easy to titrate, fast clearance, etc). The authors should also mention whether screening for other genetic mutations (especially SDH) has been performed.Discussion:The discussion should focus on the following areas:The necessity of close monitoring post adrenalectomy in cases of catecholamine producing tumors associated with cardiac complications. Selection of adrenergic receptor blockers or other anti-hypertensive medications in patients with cardiomyopathy. Review of the recovery time of cardiomyopathy post successful surgical removal of catecholamine producing tumors.", "responses": [] }, { "id": "11082", "date": "08 Dec 2015", "name": "Yukio Hayashi", "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 case report is interesting and educative. This referee would like to ask the authors to emphasize that early diagnosis would have did without heart transplantation. Hypertension is not common in young people, so possibility of pheochromocytoma should be included in the population.", "responses": [] } ]	1	https://f1000research.com/articles/4-469
https://f1000research.com/articles/4-467/v1	04 Aug 15	{ "type": "Case Report", "title": "Case Report: A case report highlighting bilateral EDB wasting as a clinical marker for lumbar canal stenosis", "authors": [ "Bijoy Mohan Kumar", "Sunil Munakomi", "Bijoy Mohan Kumar" ], "abstract": "Herein we discuss a case of a 55 year old male presenting with history suggestive of sciatica on the left leg. Straight leg raising (SLR) test was positive at 45 degrees on the left side. His ankle reflex was absent and the power of extensor hallusus longus (EHL) was 4/5 on the same side. MRI lumbosacral spine revealed left paramedian disc prolapsed on L4/L5 level with spinal canal diameter of 9mm.However since his bilateral extensor digitorm brevis (EDB) were wasted, we suspected associated lumbar canal stenosis and thereby opted for laminectomy and discectomy in this case. Intraoperatively dural wasting, hypertrophied facets and narrow canal were confirmed. Laminectomy, medial facectectomy and discectomy were carried out. The patient recovered uneventfully with resolution of his sciatica-like pain. Bilateral EDB wasting thereby provides a clinical clue to the underlying lumbar canal stenosis and can help in making correct therapeutic decisions.", "keywords": [ "extensor digitorum brevis", "lumbar canal stenosis", "dural wasting" ], "content": "Introduction\n\nLumbar disc herniation mostly causes radicular symptoms but can also lead to lumbar canal stenosis1,2. Tackling only the disc may not suffice in improving the symptomatology in patients and can invariably lead to failed back syndrome. Wasting of the extensor digitorum brevis (EDB) has been previously used as a marker for L5/S1 radiculopathy3,4. Herein we highlight the clinical importance of observing for evidence of bilateral EDB wasting as a marker for underlying lumbar canal stenosis. This simple clinical observation can help decide the correct surgical strategy and thereby prevent failed back syndrome by carrying out laminectomy rather than just tackling the disc by performing minimally invasive discectomy.\n\n\nCase report\n\nA 55 year old male from Lumbini, Nepal presented to us with a history of low back pain for 4 months with recent onset sciatica on his left side. There was no history suggestive of vascular claudication. His bladder and bowel habit was normal. His peripheral pulses in the legs were normal. There was no significant past medical or surgical illnesses. The patient had been taking oral analgesics for his pain that reduced his pain to some extent. Examination revealed straight leg raising (SLR) of 45 degrees on his left leg. Left ankle reflex was absent. The power of the extensor hallusus longus (EHL) on his left leg was 4/5. Pain sensation was diminished on his left first dorsal web space and the lateral part of the foot dorsum. However his bilateral EDB muscles were wasted (Figure 1, Figure 2) and so, clinical diagnosis of L4/L5 disc with canal stenosis was made. MRI lumbar spine revealed L4/L5 left paramedian disc with a canal diameter of 9mm. Dynamic X-ray of the lumbar spine did not show any instability. Because of the presence of bilateral EDB wasting, we opted for laminectomy in the patient rather than minimal access discectomy. Removing only the disc might result in failed back syndrome in such a patient. After detailed counseling regarding the disease process, probable complications, benefits and risks of different modes of surgical management and obtaining both written and verbal consent from the patient’s son akin, we posted the case for surgery. Intraoperatively, hypertrophic facet joints and a narrow canal were confirmed. There was significant dural wasting (Figure 3). We performed discectomy, bilateral medial facetectomy and laminectomy on the corresponding level (Figure 4). Postoperative there was resolution of the sciatica-like pain and the patient was mobilized from the second postoperative day. The patient was started on tablet pregabalin 75 mg and tablet methycobalamine 1500 µg once daily orally for 3 weeks. Patient follow-up one month later revealed no new symptoms. The patient was advised to perform regular back exercises and physiotherapy. Dynamic lumbar spine X-ray did not reveal any instability.\n\n\nDiscussion\n\nWith the increasing longevity and continually climbing proportion of middle-aged and elderly persons, low back ache is surely going to be a ubiquitous and disabling disease of mankind2.\n\nThe diagnosis of spinal stenosis is normally aided by radiological studies5. CT of the lumbar spine can show characteristic trefoil appearance of the canal while MRI can show loss of CSF surrounding the canal. However, in developing countries like ours, radiological studies may be limited due to a lack of patient finances and hospital resources. As a result, doctors are limited to clinical diagnosis.\n\nManagement of lumbar disc disease ranges from conservative6 to epidural steroids injection7,8 and surgery9. However, failure to correctly diagnose and treat canal stenosis may invariably lead to failed back syndrome in patients10.\n\nThe role of EDB as a clinical indicator of the L5 radiculopathy has already been proven3,4. Therefore, simple assessment of the bulk of the EDB muscle on both sides can predict the underlying canal stenosis and thereafter help make correct therapeutic decisions.\n\nIn this era of minimally invasive procedures, this simple bedside marker provides a word of caution for novices in the vast realms of lumbar spine procedures.\n\n\nConclusion\n\nBilateral EDB wasting can be taken as a reliable clinical marker for the diagnosis of lumbar canal stenosis. This simple bedside observation can aid us in deciding on the correct surgical strategy and thereby prevent failed back syndrome if we happen to miss the underlying canal stenosis and instead manage the disc only. It is a clinical pearl for general doctors working in remote areas to correctly assess and refer patients with EBD wasting to tertiary care centres from a subset of patients presenting with low back ache.\n\n\nConsent\n\nBoth written and verbal informed consent for publication of images and clinical data related to this case was sought and obtained from the son of the patient.", "appendix": "Author contributions\n\n\n\nSM reviewed the literature, wrote and formatted the paper. BMK revised and edited the final format.\n\n\nCompeting 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\nReferences\n\nWeinstein PR: Lumbar stenosis. In: Hardy RW Jr, ed. Lumbar disc disease. 2d ed. New York: Raven. 1993: 241–55.\n\nHumphreys SC, Eck JC: Clinical evaluation and treatment options for herniated lumbar disc. Am Fam Physician. 1999; 59(3): 575–82. PubMed Abstract\n\nSinanovic O, Custovic N: Musculus extensor digitorum brevis is clinical and electrophysiological marker for L5/S1 radicular lesions. Med Arh. 2010; 64(4): 223–4. PubMed Abstract\n\nKelly PM, Byrne S, Fleming P, et al.: Wasting of the extensor digitorum brevis - a reliable sign of l5 radiculopathy. A prospective study. J Bone Joint Surg Br. 2004; 86-B(Supp II): 121. Reference Source\n\nModic MT, Masaryk T, Boumphrey F, et al.: Lumbar herniated disk disease and canal stenosis: prospective evaluation by surface coil MR, CT, and myelography. AJR Am J Roentgenol. 1986; 147(4): 757–765. PubMed Abstract \| Publisher Full Text\n\nJacobs WC, van Tulder M, Arts M, et al.: Surgery versus conservative management of sciatica due to a lumbar herniated disc: a systematic review. Eur Spine J. 2011; 20(4): 513–522. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMobaleghi J, Allahdini F, Nasseri K, et al.: Comparing the effects of epidural methylprednisolone acetate injected in patients with pain due to lumbar spinal stenosis or herniated disks: a prospective study. Int J Gen Med. 2011; 4: 875–878. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManchikanti L, Benyamin RM, Falco FJ, et al.: Do Epidural Injections Provide Short- and Long-term Relief for Lumbar Disc Herniation? A Systematic Review. Clin Orthop Relat Res. 2015; 473(6): 1940–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSedighi M, Haghnegahdar A: Lumbar disk herniation surgery: outcome and predictors. Global Spine J. 2014; 4(4): 233–244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLong DM: Failed back surgery syndrome. Neurosurg Clin N Am. 1991; 2(4): 899–919. PubMed Abstract" }	[ { "id": "9970", "date": "24 Aug 2015", "name": "Amit Thapa", "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 Kumar et al. in this article have persuasively highlighted the importance of clinical examination while deciding on a case of low back ache with radiculopathy. This is becoming ever vital as incidence of failed back syndrome is on a rise, a major causative factor being incomplete clinical evaluation of patient 1. However few points need to be highlighted.Since the patient had been complaining of low back ache for 4 months and only recently started having sciatica, it would be prudent to ask patient regarding claudication and relation of position on pain. The first question would tell us about canal stenosis and the later about facet arthropathy. This is important in this case, as the procedure done on the patient would not relieve the facet arthropathy and patient would keep on complaining of axial back ache later. As this case highlights co existence of degenerative lumbar spinal stenosis and disc herniation, more discussion on lumbar spinal stenosis is warranted. Degenerative lumbar spinal stenosis describes a condition in which there is diminished space available for the neural and vascular elements in the lumbar spine secondary to degenerative changes in the spinal canal, presenting with a history of gluteal or lower extremity symptoms exacerbated by walking or standing which improves or resolves with sitting or bending forward. Patients whose pain is not made worse with walking have a low likelihood of stenosis2. We would request more details on the MRI study if possible an axial film of the site of compression. Guideline suggest getting a thin (4-5 mm) MRI sections with a combination of T1, proton density and T2 pulse sequences in both axial and sagittal planes with additional angled and stacked axial sections2. Meta-analysis has shown sensitivity of MRI in the diagnosis of adult spinal stenosis to be 81-97%, of CT 70-100% and myelography 67-78%3. Besides the antero-posterior diameter (< 10 mm) and cross-sectional area (< 70 mm2) of spinal canal, MRI finding of positive sedimentation sign is a good positive sign to rule in lumbar spinal stenosis with high specificity and sensitivity4. Authors have correctly highlighted the importance of dynamic x ray study to see for any instability. In presence of instability based on Posner’s criteria, patient should be offered decompression with fusion if the stenosis is moderate to severe2. However we stress on the need of standing full-length lateral radiographs of the spine to check for sagittal balance of the patient which has bearing of increasing instability after performing procedures like laminectomy. In particular, 3 measures are of vital importance: (1) global sagittal balance (C7 plumb line [C7PL], C7/sacro-femoral distance ratio, and spino-sacral angle), (2) spino-pelvic morphology (pelvic incidence, sacral slope, and pelvic tilt), and (3) spinal parameters (lumbar lordosis and thoracic kyphosis). Jeon et al. have found posterior migration of the C7PL and increase lumbar lordosis following decompressive laminectomy, in their evaluation of 40 patients over 2 years5. Figure 1 and 2 shows EDB located laterally than are expected. Authors must have drawn them to show the visible differences. However to see for EDB, best location would be interdigital spaces over dorsal of foot as EDB helps extend digits 2 though 4 (see figure)6. EDB being a muscle with smallest bulk in foot is clinically very sensitive for L5 radiculopahty. This is a case report, similar to earlier reports in cases of spina bifida or tethered cord syndrome where late manifestation has led to EDB weakness7. However North American Spine Society (NASS) in their recommendation, have found insufficient evidence to make a recommendation for or against certain physical findings for the diagnosis of degenerative lumbar spinal stenosis including an abnormal Romberg test, thigh pain exacerbated with extension, sensorimotor deficits, leg cramps and abnormal Achilles tendon reflexes2.To conclude, the authors have genuinely stress on the need of comprehensive clinical evaluation of spine and neurological function before embarking on surgical management of low back ache or radiculopathy. This paper supports the case with a good summary.", "responses": [] }, { "id": "10565", "date": "28 Sep 2015", "name": "Mehmet Zileli", "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 report tries to show the importance of muscle atrophy of extensor hallusus (EHL) longus in diagnosis of lumbar spinal stenosis. I have following criticisms:They must place MRI images of the patient instead of intraoperative surgical views. They claim that atrophy of EHL on both sides shows the disease is spinal canal stenosis instead of disc herniation. However they do not have any evidence that it cannot be produced by a midline disc herniation. Nor can a case report make us reach a conclusion like that. Atrophy of the EHL muscles should be justified by muscle strength measurements and EMG.", "responses": [] }, { "id": "10763", "date": "12 Oct 2015", "name": "Rully Hanafi Dahlan", "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 have following points that should be clearly described:Fig 3 and 4 show surgical views. Authors must state with symbols (asterisks etc) to show what level of surgical work. We would suggest a detailed axial and sagital view of the MRI for the detail level of lumbar canal compression. EDB is sensitive for L5-S1 radiculopathy. Authors must verify it with EMG modality.", "responses": [] } ]	1	https://f1000research.com/articles/4-467
https://f1000research.com/articles/4-466/v1	04 Aug 15	{ "type": "Method Article", "title": "Creating 3D visualizations of MRI data: A brief guide", "authors": [ "Christopher R. Madan" ], "abstract": "While magnetic resonance imaging (MRI) data is itself 3D, it is often difficult to adequately present the results papers and slides in 3D. As a result, findings of MRI studies are often presented in 2D instead. A solution is to create figures that include perspective and can convey 3D information; such figures can sometimes be produced by standard functional magnetic resonance imaging (fMRI) analysis packages and related specialty programs. However, many options cannot provide functionality such as visualizing activation clusters that are both cortical and subcortical (i.e., a 3D glass brain), the production of several statistical maps with an identical perspective in the 3D rendering, or animated renderings. Here I detail an approach for creating 3D visualizations of MRI data that satisfies all of these criteria. Though a 3D ‘glass brain’ rendering can sometimes be difficult to interpret, they are useful in showing a more overall representation of the results, whereas the traditional slices show a more local view. Combined, presenting both 2D and 3D representations of MR images can provide a more comprehensive view of the study’s findings.", "keywords": [ "MRI visualization", "brain mapping", "fMRI", "image processing", "neuroanatomy", "ROI" ], "content": "Introduction\n\nWhen presenting and publishing findings of magnetic resonance imaging (MRI) studies, sometimes it is difficult to adequately present the results because they are 3D, while papers and slides can inherently only be 2D. A solution is to create figures that include perspective and can convey 3D information, and the creation of such figures can be readily produced using standard functional magnetic resonance imaging (fMRI) analysis packages, such as SPM (http://www.fil.ion.ucl.ac.uk/spm/), AFNI (http://afni.nimh.nih.gov/afni/; with SUMA), and FreeSurfer (https://surfer.nmr.mgh.harvard.edu), as well as some more specialty programs, such as MRIcroGL (http://www.mccauslandcenter.sc.edu/mricrogl/), 3DSlicer (http://www.slicer.org), and Mango (http://ric.uthscsa.edu/mango/). While these numerous options can provide 3D renderings of MRI data, many of them are unable to provide useful functionality such as visualizing activation clusters that are both cortical and subcortical, i.e., a 3D glass brain. This difficulty increases further if one wants to produce 3D renderings of several activation maps with an identical perspective (e.g., camera angle) or animated renderings (e.g., a rotating 3D glass brain). Here I briefly detail a straightforward approach for creating 3D visualizations of MRI data that work in these scenarios, as well as readily generalize to most other instances. An illustration of this processing workflow is shown in Figure 1. An additional example of making a 3D rendering of traced regions of interest (ROIs) is also outlined.\n\nThe guide will primarily utilize two programs, ITK-SNAP (v. 3.0.0; Figure 2A; http://www.itksnap.org; Yushkevich et al., 2006) and ParaView (v. 4.3.1; Figure 2B; http://www.paraview.org; Ayachit, 2015). Both programs are available for both Windows and Mac operating systems and are freely available. Data files produced in the examples are provided in the Supplementary material (see Appendix A).\n\nScreenshots of (a) ITK-SNAP and (b) ParaView.\n\n\nMethods\n\nAs a first exercise in visualizing MRI data in 3D, we will start with a statistical map. Depending on where your maps are coming from, you may need to apply a height threshold (i.e., t- or Z-critical) and/or a minimum cluster extend threshold (k). As a starting point and to make this guide more general and more reproducible, I will start with a statistical map obtained from NeuroSynth (http://www.neurosynth.org; Yarkoni et al., 2011), which will be in NIfTI format (Neuroimaging Informatics Technology Initiative; http://nifti.nimh.nih.gov). Briefly, NeuroSynth conducts automated meta-analyses across thousands of fMRI studies by calculating a frequency metric for how often specific terms are mentioned in the paper (e.g., “memory”, “emotion”) in relation to voxels reported in the results tables. See Yarkoni et al. (2011) for further details. As an example of how to obtain thresholded statistical maps from SPM analyses, see Appendix B.\n\nFor this example I searched the online version of NeuroSynth for the term “memory” (http://www.neurosynth.org/analyses/terms/memory/; see Figure 3A). I used the “forward inference” map as the example statistical map, along with the anatomical volume provided (obtained by clicking the download buttons displayed to the right of the layer names). For these examples, un-gzip the NIfTI volumes from NeuroSynth. Rename the forward inference map file from memory_pAgF_z_FDR_0̷.0̷1.nii to statmap.nii.\n\n(a) Obtaining the statistical map from NeuroSynth.org. (b) Coronal slices of the thresholded activation clusters. (c) 3D renderings of the clusters from two different perspectives. (d) Stereoscopic anaglyph 3D rendering of the first perspective shown in panel C, to be viewed using red-blue 3D glasses.\n\nObtaining the anatomical ‘glass brain’ image. Since the anatomical 3D surface meshes are generally usable, in addition to outlining the steps for creating this glass brain volume, the resulting surface mesh file is also provided as Supplementary material (see Appendix A). While there is an abundance of anatomical volumes in normalized template space, here we will use the one provided on NeuroSynth (click “anatomical” where shown in Figure 2A).\n\nObtaining the thresholded cluster image. Before the map can be rendered in 3D, both the height and cluster-extent thresholds should be applied. In some fMRI analysis packages this can be output directly (e.g., see Appendix B). If this is the case, export the threshold cluster image and skip to section 1.3; if this is not the case, we will manually apply these thresholds ourselves. Here I will use examples of how to manually apply these thresholds using MATLAB (R2013a; The MathWorks Inc., Natick, MA) and SPM8 (http://www.fil.ion.ucl.ac.uk/spm/; Wellcome Department of Cognitive Neurology, UCL, London, UK), though other packages are able to do this as well. (See Madan, 2014, for an introductory guide to MATLAB.)\n\nManually applying the height and cluster-extent threshold is a bit cumbersome. Using the imcalc function in SPM, we can easily apply a height threshold to our volume by outputting a binary volume, where the voxel intensity statistic (i.e., from a t-, F-, or Z-statistic map) is above the threshold. Since our NeuroSynth memory image has a large number of highly significant clusters, we will threshold our statmap.nii to isolate the voxels where the statistic (Z-value) is above 12 (MATLAB code shown below). In the current case, lower thresholds yielded large clusters, which made the figure less interpretable (i.e., many regions comprised a single cluster, making it difficult to view the topology of the regions from the 3D view). When plotting results from your own fMRI study, you would likely use a threshold around 3 for the t- or Z-statistical map (corresponding to approximately p<.001).\n\n\n\nTo apply the cluster-extent threshold, we will use the nii_threshreslicecluster function (freely available from http://www.mccauslandcenter.sc.edu/CRNL/tools/spm8-scripts) to isolate clusters of voxels of at least a minimum volume of 400 mm3. Again, this value can be adjusted, and usually would be set higher than you would use for your statistical analyses, as the 3D rendering is intended more to provide a global view of the significant clusters, and is encumbered by the inclusion of many small clusters. A volume of 400 mm3 corresponds to 50 voxels where the voxel size is 2 mm-isotropic. The function can also apply height thresholds, but it thresholds rather than binarizes the image (i.e., converting it to a mask), which is not as useful for our current purposes.\n\n\n\nThe output file from this command will automatically be named rstatmapH.nii, rename it to statmapThresh.nii.\n\nThe height threshold should also be applied to the anatomical volume, anatomical.nii, but the cluster threshold is unnecessary. The resulting output file will be named anatomicalH.nii, rename it to glassbrain.nii.\n\nConvert to VTK. To visualize the NIfTI volumes in 3D, we need to convert the voxel data into a 3D surface mesh in the VTK (Visualization ToolKit) format. Designed for anatomical tracing, ITK-SNAP includes this functionality. The simplest way to do so is to load each volume as both the main volume and as the segmentation volume. If you use the structural volume as the main volume and the statistical map as the segmentation, you may have issues with the bounding boxes not matching. Since we will move to another program with our 3D surfaces, it does not matter if the bounding boxes match or not.\n\nMake sure that the volume is loaded correctly, as shown in Figure 2A. While ITK-SNAP can render 3D volumes, as shown in the bottom left portion of the screenshot shown in Figure 2A, its rendering options are limited. For instance, if you want to render several volumes in 3D from a consistent perspective/camera angle, ITK-SNAP is unable to accommodate; while ITK-SNAP can temporarily store camera information, this perspective information is lost if the program is closed or crashes, and it cannot be saved for later use nor can it be manually specified. As a result, it will be impossible to obtain the exact same camera angle. To rectify this shortcoming, we will make our 3D renderings in ParaView, which also has additional useful features. To export the meshes in VTK format from ITK-SNAP, use the menus to navigate to Segmentation, then “Export as Surface Mesh…”. Next, choose “Export meshes for all labels as a single scene” and save the file as a “VTK PolyData File”. In the current example there is only one surface mesh in each volume, but this is not always the case, such as in the ROI example discussed later. Note, it is possible to export volume data, rather than surface mesh data, as a VTK file in ITK-SNAP, but these files will not work with ParaView in the next step. If your VTK file does not work, double check that it was correctly exported as a surface mesh.\n\nRepeat these steps for both the statistical map and anatomical volume.\n\nRender in 3D. Start ParaView and open your two new VTK files within the same scene. ParaView can be a bit overwhelming at first, but it has many useful features for rendering and setting up the camera. With some adjustment of the colors and opacity for the two surfaces, it should be fairly easy to produce a set up in ParaView similar to Figure 2B. You can rotate the camera manually using the mouse, and can reset the camera position with the buttons labelled “+X” through “−Z”. When the scene state is saved, the camera position is preserved in the scene file, allowing you to easily load another statistical map at a later time. The scene as a whole can be saved by selecting “File” then “Save State…” (PVSM format). The final renderings produced here are shown in Figure 3C, corresponding to the series of coronal slices shown in Figure 3B. Renderings can be saved using either “File” then “Save Screenshot…” or “Export Scene…”. Screenshots will always be exported as raster (i.e., pixel) images, while ‘exported scenes’ are vector/polygon based. Note that exported PDFs can also be based on “rasterize 3D geometry” (there is a checkbox). If you are unsure what you require, a screenshot is likely sufficient, but do try and experiment to find out what settings best meet your needs, as this overview of ParaView’s functionality is far from comprehensive.\n\nParaView can also render stereoscopic 3D figures (e.g., anaglyph [red-blue], side-by-side) with a variety of 3D-compatible glasses options. An example of a red-blue stereoscopic render is shown in Figure 3D.\n\nObtain ROI volume. For this example, I extracted several regions of the medial temporal cortex (hippocampus, amygdala, parahippocampal gyrus, fusiform gyrus) from the right hemisphere of the Hammers et al. (2003) maximum probability atlas (n30r83; http://biomedic.doc.ic.ac.uk/brain-development/index.php?n=Main.AdultMaxProb). Regions were extracted using the imcalc tool included in SPM8, such that each ROI corresponded to a unique intensity value (1=hippocampus, 2=amygdala, 3=parahippocampal gyrus, 4=fusiform gyrus):\n\n\n\nThe ROI is shown plotted over a structural volume in Figure 4A.\n\n(a) Coronal slices of the anatomical ROIs. Panels B-E depict different 3D rendering settings of the ROIs, (b) points along the surface of the ROIs, (c) wireframe, (d) surface, and (e) surface rendering with specular. Note that the structural image used in panel A is from a different source than the traced ROIs, and thus they do not perfectly align.\n\nConvert to VTK. As before, use ITK-SNAP to load the NIfTI volume and convert it to a VTK surface mesh. If you have multiple surfaces in the same volume, as we do here, be sure to select “Export meshes for all labels as a single scene” when exporting the surfaces.\n\nRender in 3D. Start ParaView and load the VTK file, as done previously. As shown in Figure 4B–E, the volumes can be rendered as points, wireframes, and surfaces. Furthermore, many settings can be customized to adjust the rendering properties, such as the lighting/reflectance properties shown in Figure 4D and 4E.\n\nParaView can also create cameras that move over time, allowing for the generation of animations of the structures rotating. This can be done using the “Animation View” panel in the bottom-center of ParaView: select “Camera”, “Orbit”, and then “+”. The default settings for the camera positions are usually sufficient. If desired, the camera path can also be edited afterwards by inputting specific coordinates (the best way to preview the path is to simply press ‘play’ at the top and see how it looks). Even without rendering the animation itself, having a camera path allows for later reproduction of 3D renderings from the same camera positions.\n\nUsing a camera path, an animation can be rendered by going “File”, “Save Animation”. An example rendered video is shown in Movie 1. (Note, videos here were re-compressed with Handbrake [https://www.handbrake.fr; freely available for Windows and Mac] to reduce their file size).\n\n\n\n\nAdditional examples\n\nUsing the techniques discussed thus far, it is possible to create an image such as that shown in Figure 5, where the hippocampus is shown within a glass brain for a number of different species, using freely available brain atlases. Each panel was rendered separately, but all of the surface meshes were loaded into the same scene in ParaView. By additionally adding a plane with a checkerboard texture, it is also easy to present the scale of the structures. See Appendix C for details regarding each of the brain atlases.\n\nThe square grid included in each panel measures 20 mm across, with each grid square subtending 2 mm. See Appendix C for details and references describing each brain atlas.\n\nWith a few additional steps, more intricate 3D renderings can also be produced. For instance, if the anatomical volume is down-sampled while in NIfTI format, the resulting surface mesh is less dense and can be rendered as a wireframe, as shown in Figure 6A–C. For demonstration purposes, if the lower-resolution anatomical volume was subsequently up-sampled, the resulting high-density mesh is ‘blocky’ (Figure 6D). If a combination of different densities of anatomical surface meshes are used together, e.g., the meshes from Figure 6A–C, along with a ROI, a rendering such as Figure 6E can be produced.\n\n(a) Wireframe rendering of brain surface mesh produced from the Hammers et al. (2003) atlas, which originally has voxel size of 1 mm-isotropic. Panels B-C show the wireframes of surfaces meshes made after first downsampling the volume to 5 or 10 mm-isotropic, respectively, resulting in less dense wireframes. (d) For demonstration purposes, the result of upsampling the 10 mm-isotropic mesh (panel C) back to 1mm-isotropic. (e) Rendering produced by combining the surface meshes used in panels A-C, along with an anatomical ROI of the hippocampus.\n\nFigure 7 and Movie 2 show a few additional rendering examples from freely available data. fMRI activity related to finger tapping is shown in Figure 7A, with data obtained from Gorgolewski et al. (2013; http://www.neurovault.org/collections/63/; full dataset available at: http://www.openfmri.org/dataset/ds000114). Striatal anatomy is shown in Figure 7B (Oxford-GSK-Imanova Structural Striatal Atlas, from FSL; Tziortzi et al., 2011; http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Atlases/striatumstruc). DTI tractography showing 20 structures at varying levels of probability estimates is shown in Figure 7C and Movie 2 (JHU white-matter tractography atlas, from FSL; Hua et al., 2008).\n\n(a) Glass brain rendering of fMRI clusters associated with finger tapping, based on data from Gorgolewski et al. (2013), from three perspectives. (b) Glass brain rendering of the anatomical ROIs included in the Oxford-GSK-Imanova Structural Striatal Atlas (c) 3D rendering of the structures included in the JHU white-matter tractography atlas, with different mesh properties used for the 0%, 25%, and 50% probability estimates from the maximum probability volumes. See main text for additional details on the sources of the MRI data.\n\n\n\n\nConclusion\n\nThough a 3D ‘glass brain’ rendering of fMRI activations can sometimes be difficult to interpret, they are useful in showing a more overall representation of which regions are activated, whereas the traditional slices show a more local view of the results. When the goal is to show anatomical structures, 3D figures are definitively more useful in conveying the 3D structure of the regions, as shown in the examples here. Combined, 2D and 3D representations of MR images can provide a more comprehensive view of the results, particularly when at least two 3D perspectives are shown, allowing for some ability to provide depth information.\n\n\nData availability\n\nFigshare: Movie 1. Rotating 3D animation of the anatomical ROI with the same render settings as used in Figure 4E. doi: 10.6084/m9.figshare.1499152 (Madan, 2015a).\n\nFigshare: Movie 2. Rotating 3D animation of the white-matter tractography atlas with the same render settings as used in Figure 8C. doi: 10.6084/m9.figshare.1499153 (Madan, 2015b).", "appendix": "Competing interests\n\n\n\nThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\n\n\nGrant information\n\nThe authors declare that no grants contributed to this article.\n\n\nSupplementary material\n\nData files for ‘Creating 3D visualizations of MRI data: A brief guide’. The supplementary files are divided into two folders, “neurosynth” and “hammers”; see Appendix A.\n\nClick here to access the data\n\n\nReferences\n\nAyachit U: The ParaView Guide: A Parallel Visualization Application. Clifton Park, NY: Kitware. 2015. Reference Source\n\nBakker R, Tiesinga P, Kötter R: The Scalable Brain Atlas: Instant Web-Based Access to Public Brain Atlases and Related Content. Neuroinformatics. 2015; 13(3): 353–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalabrese E, Badea A, Watson C, et al.: A quantitative magnetic resonance histology atlas of postnatal rat brain development with regional estimates of growth and variability. Neuroimage. 2013; 71: 196–206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGorgolewski KJ, Storkey A, Bastin ME, et al.: A test-retest fMRI dataset for motor, language and spatial attention functions. Gigascience. 2013; 2(1): 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGüntürkün O, Verhoye M, De Groof G, et al.: A 3-dimensional digital atlas of the ascending sensory and the descending motor systems in the pigeon brain. Brain Struct Funct. 2013; 218(1): 269–281. PubMed Abstract \| Publisher Full Text\n\nHammers A, Allom R, Koepp MJ, et al.: Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe. Hum Brain Mapp. 2003; 19(4): 224–247. PubMed Abstract \| Publisher Full Text\n\nHua K, Zhang J, Wakana S, et al.: Tract probability maps in stereotaxic spaces: analyses of white matter anatomy and tract-specific quantification. Neuroimage. 2008; 39(1): 336–347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnson GA, Badea A, Bradenburg J, et al.: Waxholm Space: an image-based reference for coordinating mouse brain research. Neuroimage. 2010; 53(2): 365–372. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMadan CR: An Introduction to MATLAB for Behavioral Researchers. Thousand Oaks, CA: Sage. 2014. Reference Source\n\nMadan CR: Movie 1. Rotating 3D animation of the anatomical ROI with the same render settings as used in Figure 4E. Figshare. 2015a. Data Source\n\nMadan CR: Movie 2. Rotating 3D animation of the white-matter tractography atlas with the same render settings as used in Figure 8C. Figshare. 2015b. Data Source\n\nMajka P, Kublik E, Furga G, et al.: Common atlas format and 3D brain atlas reconstructor: Infrastructure for constructing 3D brain atlases. Neuroinformatics. 2012; 10(2): 181–197. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuñoz-Moreno E, Arbat-Plana A, Batalle D, et al.: A magnetic resonance image based atlas of the rabbit brain for automatic parcellation. PLoS One. 2013; 8(7): e67418. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRohlfing T, Kroenke CD, Sullivan EV, et al.: The INIA19 Template and NeuroMaps Atlas for Primate Brain Image Parcellation and Spatial Normalization. Front Neuroinform. 2012; 6: 27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTziortzi AC, Searle GE, Tzimopoulou S, et al.: Imaging dopamine receptors in humans with [11C]-(+)-PHNO: dissection of D3 signal and anatomy. Neuroimage. 2011; 54(1): 264–77. PubMed Abstract \| Publisher Full Text\n\nYarkoni T, Poldrack RA, Nichols TE, et al.: Large-scale automated synthesis of human functional neuroimaging data. Nat Methods. 2011; 8(8): 665–670. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYushkevich PA, Piven J, Hazlett HC, et al.: User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage. 2006; 31(3): 1116–1128. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9794", "date": "06 Aug 2015", "name": "Jens Foell", "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 manuscript in question describes different methods to visualize data acquired through MRI/fMRI scans in a three-dimensional manner. This is something that is sometimes done in current neuroimaging research, but that is rarely done in a standardized manner, which makes this guide timely and relevant. In many cases, researchers choose to use 2D images instead, which can sometimes distort or omit information, as fMRI depictions are derived from an inherently 3-dimensional signal. The current manuscript separately describes ways to visualize clusters of activation (i.e. activation as it would be found when running an fMRI experiment) and anatomical regions of interest. It also provides hyperlinks to download relevant visualization software. The author goes into sufficient detail to include, for example, information on price and OS compatibility of different software packages. Also, the text provides details about how to create the images within a particular software package, or functions that increase user efficiency. Information like this, in addition to several informative illustrations in the manuscript, will make this text particularly useful for many people working in neuroimaging, and I am convinced that the publication of this manuscript will lead to a fruitful online discussion about the best ways to visualize and report 3D brain data.The title, abstract, and structuring of the manuscript are well-written and appropriate for its purpose as a brief guide.Overall, this concise and informative guide is useful, interesting, and well-written. I recommend its indexing after some very minor comments (listed below) have been addressed to increase the readability of the manuscript. Minor suggestions:While the term ‘3D’ could be considered to be a household word, I would still recommend to spell it out as ‘three-dimensional (3D)’ or ‘3-dimensional (3D)’ the first time the term is used in the text. Likewise, the term ‘glass brain’ is intuitive, but not always used in the same way by all researchers. A quick description of the concept at the first mention of the term in the text would make the manuscript more accessible to the general reader.", "responses": [] }, { "id": "9899", "date": "10 Aug 2015", "name": "Anders Eklund", "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 that this is a useful paper. Here are some minor commentsYou do not mention anything about multiple comparisons for the thresholding. I understand that these visualizations are mainly for obtaining a better understanding of the brain activation, but it would still be nice to mention the problems of multiple testing. For cluster level inference, I prefer if a cluster p-value threshold is used, and not an arbitrary cluster size like 400 mm3 or 50 voxels. Cluster p-values can be obtained through parametric methods (Gaussian random field theory, available in SPM and FSL) or non-parametric methods (permutation testing, available in SnPM, FSL and BROCCOLI). I know that a very common approach is to use a cluster defining threshold of p = 0.001 or p = 0.005 (uncorrected for multiple comparisons), combined with an arbitrary cluster size threshold of 10 voxels. Such approaches should in my opinion be avoided, since the method is ad-hoc; it is impossible to know what the (corrected) p-value is for the combined procedure.The following paper may be of interest:Choong-Wan Woo, Anjali Krishnan, Tor D. Wager, Cluster-extent based thresholding in fMRI analyses: Pitfalls and recommendations, NeuroImage, Volume 91, 1 May 2014, Pages 412-419, ISSN 1053-8119, http://dx.doi.org/10.1016/j.neuroimage.2013.12.058-------You may mention two additional pieces of software, pysurfer and MevisLab.Pysurfer is a python tool for visualizing cortical surface representationshttps://pysurfer.github.io/MevisLab is a free software that can be used for image processing and visualization. MevisLab includes functions from the libraries VTK and ITK, and it is easy to setup more advanced volume rendering pipelines, where you for example have several volume renderers, clip planes and more advanced transfer functions.http://www.mevislab.de/-------You do not mention anything about visualization research regarding fMRI. A more advanced way to visualize brain activation is to treat the activation as a light source in the anatomical volume, making the activity \"glow\" from the inside. You could include some of the following papers.Nguyen, T. K., Eklund, A., Ohlsson, H., Hernell, F., Ljung, P., Forsell, C., Andersson, M., Knutsson, H., Ynnerman, A., Concurrent Volume Visualization of Real-time fMRI, Proceedings of the 8th IEEE/EG International Conference on Volume Graphics, 53-60, 2010, http://dx.doi.org/10.2312/VG/VG10/053-060Janoos, F., Nouanesengsy, B., Machiraju, R., Shen, H. W., Sammet, S., Knopp, M. and Mórocz, I. Á. (2009), Visual Analysis of Brain Activity from fMRI Data. Computer Graphics Forum, 28: 903–910. doi: 10.1111/j.1467-8659.2009.01458.xJainek, W. M., Born, S., Bartz, D., Straßer, W. and Fischer, J. (2008), Illustrative Hybrid Visualization and Exploration of Anatomical and Functional Brain Data. Computer Graphics Forum, 27: 855–862. doi: 10.1111/j.1467-8659.2008.01217.xRieder, C., Ritter, F., Raspe, M. and Peitgen, H.-O. (2008), Interactive Visualization of Multimodal Volume Data for Neurosurgical Tumor Treatment. Computer Graphics Forum, 27: 1055–1062. doi: 10.1111/j.1467-8659.2008.01242.x", "responses": [] }, { "id": "9792", "date": "07 Sep 2015", "name": "Matthew Wall", "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 very useful guide to an important issue that is currently largely overlooked in the literature; producing high-quality presentations of brain imaging results that are informative, clear, and useful. The article is comprehensive and easy to follow, and the examples provided are appropriate, and produce very attractive images. This is an extremely useful paper that deserves wide readership in the field.While I agree with the author that ‘glass-brain’ visualisations are extremely useful for providing a comprehensive overview of patterns of brain activity in fMRI experiments, that doesn’t mean that conventional 2D slice views are not also useful. In fact, 2D views of particular activation clusters are really the only way to get a good idea of the precise position of a cluster, in relation to the sulcal/gyral anatomy, which is often important. An optimal strategy for comprehensive visualisation and localisation might then be to combine 2D and 3D views of results in the same figure. The author has done this more-or-less in Figure 3 (which includes coronal slices), but I wonder if perhaps an additional example figure which combines 2D and 3D views might be helpful? Perhaps as an example of the kinds of ‘real’ figures that could be produced for publications and presentations.Minor points of grammar, etc.:Abstract:\"they are useful in showing a more overall representation of the results\" More overall? Somewhat clumsy; replace with \"more general\" or just \"overall\".Page 2 first paragraph: \"Here I briefly detail a straight- forward approach for creating 3D visualizations of MRI data that work in these scenarios, as well as readily generalize to most other instances.\" Something wrong with the tenses here; would suggest: \"Here I briefly detail a straight- forward approach for creating 3D visualizations of MRI data that works in these scenarios, and also readily generalizes to most other instances.\"Page 4. Section on obtaining and thresholding the images. Fine, but the procedure outlined here is pretty cumbersome, as the author admits! This procedure might be optimal for those who use SPM as their primary analysis tool, but the 'fslmaths' function included with FSL could achieve this in a single command-line entry. Maybe include a sentence saying something like \"Other options for thresholding are available, such as the basic functions included with FSL.\"", "responses": [] } ]	1	https://f1000research.com/articles/4-466
https://f1000research.com/articles/4-460/v1	03 Aug 15	{ "type": "Research Article", "title": "Deep Sequencing of the T-cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis", "authors": [ "Jamie L. Harden", "David Hamm", "Nicholas Gulati", "Michelle A. Lowes", "James G. Krueger", "David Hamm", "Nicholas Gulati", "Michelle A. Lowes", "James G. Krueger" ], "abstract": "It is well known that infiltration of pathogenic T-cells plays an important role in psoriasis pathogenesis. However, the antigen specificity of these activated T-cells is relatively unknown. Previous studies using T-cell receptor polymerase chain reaction technology (TCR-PCR) have suggested there are expanded T-cell receptor (TCR) clones in psoriatic skin, suggesting a response to an unknown psoriatic antigen. Here we describe the results of high-throughput deep sequencing of the entire αβ- and γδ- TCR repertoire in normal healthy skin and psoriatic lesional and non-lesional skin. From this study, we were able to determine that there is a significant increase in the abundance of unique β- and γ- TCR sequences in psoriatic lesional skin compared to non-lesional and normal skin, and that the entire T-cell repertoire in psoriasis is polyclonal, with similar diversity to normal and non-lesional skin. Comparison of the αβ- and γδ- TCR repertoire in paired non-lesional and lesional samples showed many common clones within a patient, and these close were often equally abundant in non-lesional and lesional skin, again suggesting a diverse T-cell repertoire. Although there were similar (and low) amounts of shared β-chain sequences between different patient samples, there was significantly increased sequence sharing of the γ-chain in psoriatic skin from different individuals compared to those without psoriasis. This suggests that although the T-cell response in psoriasis is highly polyclonal, particular γδ- T-cell subsets may be associated with this disease. Overall, our findings present the feasibility of this technology to determine the entire αβ- and γδ- T-cell repertoire in skin, and that psoriasis contains polyclonal and diverse αβ- and γδ- T-cell populations.", "keywords": [ "T-cell receptor (TCR)", "TCR deep-sequencing", "psoriasis", "T-cell", "polyclonal", "αβ-T-cells", "γδ- T-cells" ], "content": "Introduction\n\nPsoriasis vulgaris (plaque psoriasis) is a chronic, inflammatory skin disease characterized by raised, red scaly plaques (Lowes et al., 2014). It has been well established that infiltration of pathogenic T-cells plays an important role in disease pathology, particularly T-cells with the ability to produce interleukin (IL)-17 and IFN-γ (Lowes et al., 2014; Perera et al., 2012). T-cell targeting and specifically biologics targeting the TH17 axis, have proven to be extremely efficacious in treatment of psoriasis (Lowes et al., 2014). However, despite the established role of T-cells in psoriasis pathogenesis, the potential auto-antigens eliciting disease or the antigen specificity of T-cells contributing to psoriasis are relatively unknown.\n\nHigh-throughput sequencing of the T-cell receptor (TCR) repertoire can help elucidate whether a narrow subset of T-cells have undergone clonal expansion, potentially elicited by a particular antigen. Previous evaluation of TCR repertoire in psoriasis has shown preferential usage of particular β-chain variable gene segments (Chang et al., 1995; Menssen et al., 1995). However in these studies, a small number of TCR sequences was obtained from a potentially vast underlying diversity of TCR sequences (Robins et al., 2009) and clonal expansion was determined only in the context of particular Vβ-chain usages. Recent advances in next generation deep sequencing have enabled the quantitative sampling of a significantly larger enough fraction of the T-cell repertoire to make strong inferences regarding the diversity of the T-cell repertoire and measure the degree of oligoclonality based on individual clone lineages (Robins, 2013).\n\nHere we present the first exploration of TCR deep sequencing in psoriatic (both lesional and non-lesional) and normal skin using the ImmunoSEQ assay (Adaptive Biotechnologies, Seattle, WA, USA). Genomic DNA (gDNA) and complementary DNA (cDNA) from skin of normal volunteers, as well as non-lesional and lesional skin from psoriatic patients, were used for deep sequencing of the β-chain and γ-chain, respectively. From these initial studies, we were able to fully characterize the TCR repertoire of each sample, and compare diseased and healthy skin both between and within an individual patient.\n\n\nMethods\n\nNormal skin biopsies from healthy volunteers (n=7), and non-lesional (n=5) and lesional skin (n=8) from patients with psoriasis, were obtained under Rockefeller University Institutional Review Board-approved protocols (IRB numbers MLO-0651 and JKR-0742), and stored in a cryovial in liquid nitrogen until use. Written informed consent was obtained and the study was performed in adherence with the Declaration of Helsinki. Data regarding patient demographics is found in Table S1. All psoriasis samples were obtained from patients with moderate-to-severe disease, which was determined by a psoriasis area severity index (PASI score) of more than 12 (moderate-to-severe psoriasis vulgaris with >10% body surface area involvement).\n\nFrozen skin biopsies were processed in 1mL of RLT buffer (Qiagen, Venlo, Limburg) with 1% β-mercaptoethanol, through several rounds of homogenation (Polytron PT-10-35 GT-Kinematica) and sonication (Sonics Vibra cell; model VCX130). In brief, samples were homogenized for several 30-second intervals until the sample was significantly disrupted; samples were then sonicated for 6 pulses of 1–2 seconds (at an amplitude of 80–100) to fully lyse the cells. Genomic DNA and RNA was isolated using the Qiagen All-Prep DNA/RNA/Protein kit (Qiagen, Venlo, Limburg) (Catalog number 80004), following the manufacturer’s instructions. In brief, cell lysate was added to a DNA binding column and centrifuged. The flow-through containing RNA and protein was then added to a separate RNA binding column and centrifuged. Each column containing either the DNA or RNA from a sample was then washed and the appropriate nucleic acid was eluted, according to manufactures instructions.\n\n6–8μg of genomic DNA (at a concentration of 50ng/μl), was used for survey level deep sequencing of the β-chain, using the ImmunoSEQ platform (Adaptive Biotechnologies, Seattle, WA, USA). 2μg of RNA was used for reverse transcription using the High Capacity cDNA Reverse Transcriptase Kit (Applied Biosystems, Carlsbad, CA) (Catalog number 4368814), according to manufacturer’s protocols, and the entirety of the cDNA reaction was sent to Adaptive biotechnologies for survey level deep sequencing of the γ-chain using the ImmunoSEQ platform. Details of the deep sequencing assay are as follows: the TCRβ and TCRγ CD3 region was amplified and sequenced using the ImmunoSEQ assay (Adaptive Biotechnologies, Seattle, WA). In this assay, a multiplex PCR system was used to amplify the rearranged CDR3β and CDR3γ sequences from sample DNA or cDNA, respectively. The 87-base-pair fragment is sufficient to identify the VDJ region spanning each unique CDR3β. Amplicons were sequenced using the Illumina platform. TCRβ and TCRγ V, D and J gene definitions were provided by the IMGT database (www.imgt.org). The assay is quantitative, having used a complete synthetic repertoire of TCRs to establish an amplification baseline and adjust the assay chemistry to correct for primer bias. In addition, barcoded, spiked-in synthetic templates were used to measure the degree of sequencing coverage and residual PCR bias. This information was used for further PCR bias correction and to estimate the abundance of sequenceable templates in each sample. The resulting data is filtered and clustered using both the relative frequency ratio between similar clones and a modified nearest-neighbor algorithm, to merge closely related sequences and remove both PCR and sequencing errors. Data was analyzed using the ImmunoSEQ analyzer toolset, as further described below.\n\nSequencing data were analyzed using the immunoSEQ Analyzer (https://clients.adaptivebiotech.com/login). Entropy H (i.e. Shannon’s Entropy), is a measure of the richness and uniformity of the TCR repertoire’s frequency distribution\n\nH=−∑i=1NPilog2Pi,\n\nin which N is the number of unique clones and Pi is the frequency of clone i. Entropy ranges from 0 in a sample with only one clone, to Hmax = log2N for a sample with a uniform distribution of clone frequencies. Monoclonal or oligoclonal samples have low entropy, and polyclonal highly diverse samples have entropy just under log2N. The measurement of entropy is sensitive to sampling depth, as sample size is the primary driver of the number of unique sequences in a sample. To account for variation in sequencing depth, entropy is normalized by its maximum value to give a metric that measure just sample evenness.\n\nHN=HHmax.\n\nClonality (Sherwood et al., 2013) is defined as C = 1 - HN. Clonality equals 0 when all sequences are equally abundant, and equals 1 when a single sequence makes up the entire sample. Overlap scores (Figure 3) excluded all paired non-lesional (NL) and lesional (LS) psoriatic skin comparisons were removed to prevent artificial inflation of the NL-LS comparison group. All data was analyzed using an unpaired students t-test or ANOVA, using the GraphPad Prism5 software. Error bars represent the standard error of the mean. P-values < 0.05 were considered significant.\n\n\nResults\n\nDeep sequencing identified an average of 2.7 × 103 unique TCRβ-CDR3 (complementarity-determining region 3) sequences in normal skin, 3.2 × 103 unique TCRβ-CDR3 in non-lesional skin, and an average of 10.9 × 103 unique TCRβ-CDR3 sequences in psoriatic lesional skin, a significant three-fold increase compared to both normal (p = 0.0127) and non-lesional (p = 0.0295) skin (Figure 1a). The number of unique sequences was also significantly increased (p = 0.0159) approximately three-fold for the γ-chain, with an average of 397 unique TCRγ-CDR3 sequences in normal skin, 558 unique TCRγ-CDR3 in non-lesional skin, and an average of 1.57 × 103 unique TCRγ-CDR3 sequences in psoriatic lesional skin (Figure 1b). The observation that there were more productive unique sequences in psoriatic skin is most likely a consequence of the increased numbers of T-cells well-known to occur in this disease (Lowes et al., 2014; Zaba et al., 2007).\n\nThe total number of unique (a) β- and (b) γ-chain CDR3 sequences, and the clonality of the entire (c) β- and (d) γ-chain repertoire (y-axis) in normal and psoriatic lesional and non-lesional skin (x-axis). Error bars represent standard error of the mean. Statistics were performed using an unpaired t-test. * p < 0.05.\n\nClonality was used to compare the degree of clonal expansion across groups of samples. Psoriatic skin (both non-lesional and lesional) had slightly less clonality in the β-chain (i.e. more diverse and polyclonal population) than normal skin; however this finding was not statistically significant (p = 0.1255 for normal versus lesional; p = 0.1419 for non-lesional versus lesional) (Figure 1c). The clonality of the γ-chain repertoire was very similar between normal, non-lesional and lesional psoriatic skin (Figure 1d). Therefore, in combination with a clear signal of increased numbers of T-cells in psoriatic tissue, this suggests that the increase in T-cells is not due to the clonal expansion of a limited number of disease specific T-cells but the infiltration of large and diverse numbers of both αβ- and γδ- T-cells responding to an immune signaling cascade.\n\nThe variable (V) and joining (J) gene usage of both the β- and γ-chains was compared across the normal, non-lesional, and lesional psoriatic skin groups (Figure S1). Previous studies have suggested significant variations in V- and J- gene usage between normal and psoriatic skin. Although we found trends supporting these previous observations, such as TCRVβ02.1 preferentially utilized in normal skin (Chang et al., 1995; Menssen et al., 1995), much larger sample sizes would be required to verify potential gene usage differences.\n\nTo determine if the most abundant clones in lesional skin may expand from a rare subset in non-lesional skin, we compared the β- and γ-chain TCR repertoire in paired lesional and non-lesional samples (Figure 2). Although there were many clones uniquely present in only non-lesional and lesional skin, in all patients the most abundant clones in lesional skin were also the most abundant in non-lesional skin. This suggests that the abundance of clones in lesional skin is not the result of the expansion of a rare-subset in non-lesional skin, but is rather a general expansion and/or influx of many clones.\n\nComparison of TCR (left) β- and (right) γ-chain CDR3 sequences in paired non-lesional and lesional psoriasis samples. Each dot represents one clone and the percent of total reads of a given clone in non-lesional (y-axis) and lesional (x-axis) skin are shown. Clones uniquely in non-lesional skin are located on the y-axis in red; clones uniquely in lesional skin are located on the x-axis in green; and clones found in both non-lesional and lesional skin are in blue. The absolute number of sequences in only non-lesional (red), only lesional (green), or both (blue) are located under each graph.\n\nAlthough there is a highly diverse and polyclonal T-cell response in psoriasis, with no clear enrichment for particular variable (V) and joining (J) genes, there may still be clones specifically associated with psoriatic skin. If this is correct, it would be expected then that the degree of pair-wise sequence sharing would be greater within the psoriatic group. To test this hypothesis, the fraction of shared sequences between all pair-wise sets of samples was calculated (Figure 3).\n\nSequence similarity overlap scores (y-axis) obtained from the comparisons, normalized to total number of sequences and pooled into common groups (x-axis) for the (a) β- and (b) γ-chain. (c) Common clones found within all lesional psoriatic skin samples (top), and one clone found in all psoriatic lesional and non-lesional skin (bottom). Error bars represent standard error of the mean. Statistics were performed using an unpaired t-test. * p < 0.05.\n\nThere was no significant difference (ANOVA p = 0.1114) in the overlap score for any comparisons of the β-chain (Figure 3a). However, there were significant differences in overlap score comparisons for the γ-chain (ANOVA p = 0.0039) (Figure 3b). The psoriatic lesional skin γ-chain repertoire exhibited significantly more pair-wise overlap than healthy controls. Additionally, the γ-chain repertoire of non-lesional skin compared to psoriatic skin contained significantly more overlap than comparison of non-lesional skin to normal skin. This finding suggests that unique populations of cutaneous γδ- T-cells may be present in psoriatic patients.\n\nTo further assess this hypothesis, we explored if there were any common TCRγ sequences between all of the samples within a group. There were no common γ-chain sequences among all the normal samples. However, there were 3 common γ-chain sequences found among all psoriatic lesional skin samples (Figure 3c). Additionally, one of those γ-chain sequences was also found in all non-lesional skin samples as well. Regarding the β-chain, no common sequences were found in any group.\n\n\nDiscussion\n\nWe have utilized the recent advances in next-generation sequencing to provide evidence for diverse and polyclonal αβ- and γδ- T-cell populations in psoriasis lesional skin. Although several previous studies have focused on the clonality within a range of Vβ- usages (Chang et al., 1995; Menssen et al., 1995; Vollmer et al., 2001), this is the first time sufficient numbers of sequences have been obtained to fully characterize the diversity of the TCR repertoire in psoriasis. We found that although lesional skin contained significantly more unique β- and γ- chain sequences than normal and non-lesional skin, lesional skin was highly polyclonal with no dominant T-cell clones. This finding supports a diverse and non-specific polyclonal T-cell infiltrate in psoriasis lesional skin.\n\nBased on previous work, it may have been anticipated that there would be higher clonality compared to control samples, due to expansion of T-cells to an unknown antigen. The first gene associated with psoriasis (located at PSORS1) is HLA-Cw6 (Perera et al., 2012); it was thought that HLA-Cw6 likely predisposes to psoriasis by presentation of an unknown autoantigen. Mutations in other immune-related genes have recently been found to impose psoriasis-susceptibility, such as IL-36RN and CARD14 (Jordan et al., 2012; Marrakchi et al., 2011); however the immunological activities of these latter genes are not solely at the adaptive immune level, and play important roles in general inflammation.\n\nSeveral previous studies utilizing TCR-PCR technology found the same clones present in lesional skin of a patient over time (Chang et al., 1995; Menssen et al., 1995; Vollmer et al., 2001), and these clones were absent in non-lesional skin. In our study, we found that the most prevalent clones in lesional skin were present at a similar amount (percent of reads) in non-lesional skin. Our data suggests that it is not an expansion of common non-lesional/lesional clone as the driver of psoriasis, but rather a general polyclonal T-cell expansion in psoriatic lesional skin.\n\nAlthough αβ-T-cells are more prevalent than γδ T-cells in human skin (Elbe et al., 1996), it has recently been appreciated that γδ T-cells may contribute to psoriatic inflammation, as they can be major producers of IL-17, a key cytokine in psoriasis pathogenesis (Cai et al., 2012). We found minimal similarities between the TCRβ-repertoire of different patient samples despite the much greater sample sizes of TCRβ clones. However, three common TCRγ clones were found in all lesional skin samples, and one clone was found in all non-lesional and lesional skin, but was absent in normal skin. This finding merits further investigation with larger samples sizes to determine if particular γδ- T-cells are common among psoriasis patients and may represent a population(s) responding to a similar antigen(s).\n\nIn conclusion, we have provided the first deep sequencing results of the entire β- and γ- T-cell repertoire in normal, non-lesional, and lesional human skin. Our findings demonstrate highly polyclonal αβ- and γδ- T-cell populations in psoriasis lesional skin, with the most common clones being present in both non-lesional and lesional skin. Lastly, there may be possible contributions of specific γδ- T-cell in psoriasis, as evidenced by common CDR3 sequences between patients.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for \"Deep Sequencing of the T-cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis\", 10.5256/f1000research.6756.d97231 (Harden et al., 2015).\n\nThis data is also available from the Adaptive Biotechnology ImmunoSEQ site (http://adaptivebiotech.com/pub/Harden-2015-F1000Res) which provides access to ImmunoSEQ Analyzer and other tools used for analyses.", "appendix": "Author contributions\n\n\n\nJLH, MAL, and JGK conceived the study and designed the experiments. JLH and NG carried out the research. DH provided essential expertise in data analysis. JLH, MAL and DH prepared the first draft of the manuscript. All authors were involved in the revision of the manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nAll authors state no conflict of interest. DH has employment and stock options at Adaptive Biotechnologies.\n\n\nGrant information\n\nThis research was made possible by The American Skin Association Research Grant and Adaptive Biotechnologies Young Investigator Award, both awarded to JLH. MAL and JLH were supported by NIH 1R01AR060222.\n\nI confirm that the 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. Catherine Sanders (Adaptive Biotechnologies) for project consultation, Dr. Dáibhid Ó Maoiléidigh for assistance in manuscript preparation and interpretation of data, and Mary M. Sullivan-Whalen for acquiring biopsies for this study.\n\n\nSupplementary material\n\nSupplementary material for \"Deep Sequencing of the T-cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis\".\n\nClick here to access the data\n\n\nReferences\n\nCai Y, Fleming C, Yan J: New insights of T cells in the pathogenesis of psoriasis. Cell Mol Immunol. 2012; 9(4): 302–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChang JC, Smith LR, Froning KJ, et al.: CD8+ T-cells in psoriatic lesions preferentially use T-cell receptors V beta 3 and/or V beta 13.1 genes. Ann N Y Acad Sci. 1995; 756: 370–81. PubMed Abstract \| Publisher Full Text\n\nElbe A, Foster CA, Stingl G: T-cell receptor alpha beta and gamma delta T cells in rat and human skin--are they equivalent? Semin Immunol. 1996; 8(6): 341–9. PubMed Abstract \| Publisher Full Text\n\nHarden J, David H, Gulati N, et al.: Dataset 1 in: Deep Sequencing of the T-cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis. F1000Research. 2015. Data Source\n\nJordan CT, Cao L, Roberson ED, et al.: PSORS2 is due to mutations in CARD14. Am J Hum Genet. 2012; 90(5): 784–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLowes MA, Suárez-Fariñas M, Krueger JG: Immunology of psoriasis. Annu Rev Immunol. 2014; 32: 227–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarrakchi S, Guigue P, Renshaw BR, et al.: Interleukin-36-receptor antagonist deficiency and generalized pustular psoriasis. N Engl J Med. 2011; 365(7): 620–8. PubMed Abstract \| Publisher Full Text\n\nMenssen A, Trommler P, Vollmer S, et al.: Evidence for an antigen-specific cellular immune response in skin lesions of patients with psoriasis vulgaris. J Immunol. 1995; 155(8): 4078–83. PubMed Abstract\n\nPerera GK, Di Meglio P, Nestle FO: Psoriasis. Annu Rev Pathol. 2012; 7: 385–422. PubMed Abstract \| Publisher Full Text\n\nRobins H: Immunosequencing: applications of immune repertoire deep sequencing. Curr Opin Immunol. 2013; 25(5): 646–52. PubMed Abstract \| Publisher Full Text\n\nRobins HS, Campregher PV, Srivastava SK, et al.: Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta T cells. Blood. 2009; 114(19): 4099–107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSherwood AM, Emerson RO, Scherer D, et al.: Tumor-infiltrating lymphocytes in colorectal tumors display a diversity of T cell receptor sequences that differ from the T cells in adjacent mucosal tissue. Cancer Immunol Immunother. 2013; 62(9): 1453–61. PubMed Abstract \| Publisher Full Text\n\nVollmer S, Menssen A, Prinz JC: Dominant lesional T cell receptor rearrangements persist in relapsing psoriasis but are absent from nonlesional skin: evidence for a stable antigen-specific pathogenic T cell response in psoriasis vulgaris. J Invest Dermatol. 2001; 117(5): 1296–301. PubMed Abstract \| Publisher Full Text\n\nZaba LC, Cardinale I, Gilleaudeau P, et al.: Amelioration of epidermal hyperplasia by TNF inhibition is associated with reduced Th17 responses. J Exp Med. 2007; 204(13): 3183–94. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "10141", "date": "08 Sep 2015", "name": "Liv Eidsmo", "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\nHarden et al., provide the first deep sequence analysis of the T cell receptor in lesional and non-lesional skin collected from psoriasis patients. The authors simultaneously track and quantify of the V-β (from DNA) and V-γ (from RNA) T cell receptor within the same skin tissue samples. The data suggest that the T cell infiltrate in full-thickness psoriasis lesions is clonally diverse and the authors could not reproduce previous results showing oligoclonal expansion of TCR ab within the skin.The authors do however suggest clonal enrichment of TCR γ in psoriasis patients, both in lesional and non-lesional skin, in comparison to healthy skin. The manuscript is straightforward but there are several points that the authors could comment or discuss further as listed below. Major concerns:The majority of infiltrating T cells in psoriasis are found in the dermal compartment of the skin with only a small population of T cells infiltrating epidermis. However, an enrichment of pathogenic IL-17 and IL-22 producing T cells in epidermis as compared to dermis of psoriasis has been shown in a number of reports (Res et al., 2010; Hijnen et al., 2013; Cheuk et al., 2014). Of interest, the oligoclonal T cell expansion shown by Menssen et al. (1995) in the same psoriasis plaques over time were detected in epidermis and intraepithelial oligoclonal T cell expansion was also suggested by Lin et al. (2001). Without dissecting epidermis and dermis, it is difficult to draw the conclusion that psoriasis is a “highly polyclonal” response lacking oligo-clonal expansion. Intraindividual comparison of two lesional and two non-lesional samples would be interesting to determine the heterogeneity of T cell infiltration in different sites of inflamed and non-affected skin. Please comment on the skewed ethnicity between the healthy controls and the patient group analyzed. Could this affect the interpretation of the γδ data, in which the authors showed overlap among NL and LS as compared to the controls? Did quantification of TCRg analysed in RNA or DNA give identical results or was a bias in the quantification introduced when analysing cDNA? Figure 3: How the overlap score is calculated is unclear. Is that equivalent to “ the fraction of shared sequences between all pair-wise sets of samples”? A more detail description will be appreciated. Minor points:Abstract row 14: “...these close were often...” should probably read “...these clones were often...”. Figure 2: Out of 5 donors, 2 (or 3) did not show much difference in the total number of unique sequences between NL and LS. That was inconsistent with data from figure 1. Please comment on that. The sharing of specific TCR γ clones among different patients but absent in healthy controls is intriguing. Are these clones consistent with the infiltrating gd T cells in psoriasis described by Laggner et al. (2011) Figure 3: Were the pair-wised comparison from the same donor omitted or included in the calculation since some of the LS and NL were collected from the same donor? That would affect the interpretation statistical testing between the groups concerning the healthy control and the groups concerning only psoriasis patients. Please comment on that. Figure 3: The overlapping score or fraction of shared sequences between different donors in the case of TCR-β is difficult to interpret since the donor HLA type is not shown.", "responses": [] }, { "id": "10142", "date": "09 Sep 2015", "name": "Jun Yan", "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 manuscript by Harden et al. entitled “Deep Sequencing of the T cell Receptor Repertoire Demonstrates Polyclonal T-cell Infiltrates in Psoriasis” describes the results of high-throughput deep sequencing of the TCR beta and gamma chains in normal healthy skin and psoriatic lesional and nonlesional skin tissues. The authors found that although there is a significant increase in the abundance of unique beta and gamma TCR sequences in poriatic lesional skin, the entire T cell repertoire in psoriasis is polyclonal. More strikingly, they found three common sequences of gamma chain repertoire in psoriatic skin. These are really exciting findings and provide novel insights into understanding psoriasis immunopathogenesis. Minor comments:The rationale for using genomic DNA (gDNA) for TCR beta chain while cDNA for TCR gamma chain sequencing needs to be clearly delineated. This is particularly important given the fact that different target enrichment solutions have its own pros and cons. The authors conclude that the entire T cell repertoire in psoriasis is rather polyclonal with similar diversity in normal vs non-lesional skin, which is contradictory to previous reports. The authors may want to discuss this further, e.g. whether this discrepancy is due to different technology used or other factors. Were psoriatic lesional and non-lesional skin tissues paired?", "responses": [] } ]	1	https://f1000research.com/articles/4-460
https://f1000research.com/articles/4-456/v1	03 Aug 15	{ "type": "Opinion Article", "title": "Is suvorexant a better choice than alternative hypnotics?", "authors": [ "Daniel F. Kripke" ], "abstract": "Suvorexant is a novel dual orexin receptor antagonist (DORA) newly introduced in the U.S. as a hypnotic, but no claim of superiority over other hypnotics has been offered. The manufacturer argued that the 5 and 10 mg starting doses recommended by the FDA might be ineffective. The manufacturer's main Phase III trials had not even included the 10 mg dosage, and the 5 mg dosage had not been tested at all in registered clinical trials at the time of approval. Popular alternative hypnotics may be similarly ineffective, since the FDA has also reduced the recommended doses for zolpidem and eszopiclone. The \"not to exceed\" suvorexant dosage of 20 mg does slightly increase sleep. Because of slow absorption, suvorexant has little effect on latency to sleep onset but some small effect in suppressing wakening after sleep onset and in improving sleep efficiency. The FDA would not approve the manufacturer's preferred 40 mg suvorexant dosage, because of concern with daytime somnolence, driving impairment, and possible narcolepsy-like symptoms. In its immediate benefits-to-risks ratio, suvorexant is unlikely to prove superior to currently available hypnotics—possibly worse—so there is little reason to prefer over the alternatives this likely more expensive hypnotic less-tested in practice. Associations are being increasingly documented relating hypnotic usage with incident cancer, with dementia risks, and with premature death. There is some basis to speculate that suvorexant might be safer than alternative hypnotics in terms of cancer, dementia, infections, and mortality. These safety considerations will remain unproven speculations unless adequate long-term trials can be done that demonstrate suvorexant advantages.", "keywords": [ "suvorexant", "Belsomra®", "zolpidem", "eszopiclone", "melatonin", "sleep", "hypnotic", "mortality", "cancer" ], "content": "A new kind of hypnotic drug\n\nThe manufacturer has begun U.S. marketing for suvorexant (Belsomra®), a dual orexin receptor antagonist (DORA) offered as a new hypnotic for treatment of insomnia (See Table 1 for abbreviations). The manufacturer's information emphasizes that the drug is novel and acts by a mechanism distinct from the benzodiazepine agonists and antihistamines commonly marketed as hypnotics. The prescribing information does not claim that suvorexant has greater benefits or fewer risks than other drugs marketed for insomnia. Indeed, a search of PubMed (www.PubMed.gov), ClinicalTrials.gov (www.ClinicalTrials.gov), and the International Clinical Trials Registry Platform multinational clinical trials registries (http://www.who.int/ictrp/) found no trials comparing suvorexant with other hypnotics for treatment of insomnia (searched July 17, 2015). Some small comparative trials have been done focused on specific adverse risks such as middle-of-the night impairment and driving impairment1. Physicians and their patients may thus wonder whether they should switch from familiar hypnotics to suvorexant that may have higher costs than popular generics. This discussion presents a clinician's opinions about the choice of hypnotics. Not discussed here are the much more complex issues of when insomnia should be treated with hypnotics and when new developments such as the cognitive-behavioral treatment of insomnia or bright light treatment should be seen as better choices than any hypnotic.\n\nOrexins are excitatory neurotransmitters, secreted primarily by a small number of cells in the lateral hypothalamus2–4. Orexins have many actions in the brain2,4,5, but the current interest is in orexin actions in maintaining wakefulness, for example, through activating tuberomammillary histamine neurons that secrete wake-maintaining histamine throughout many brain areas6,7. Suvorexant blocks orexin's stimulation of histaminergic neurons. Suvorexant advocates suggests that there is a qualitative difference between suvorexant antagonizing wakefulness whereas in contrast, competitive hypnotics promote sleep, but I cannot conceptualize this distinction clearly. For example, benzodiazepine receptor agonists and histamine receptor antagonists (antihistamines) also suppress histaminergic alerting, besides diverse other actions8. Sleep-wake regulation has been conceptualized as a “flip-flop switch”9 in which a stronger flip or a weaker flop might produce equivalent switching.\n\nWhen orexin-secreting neurons or orexin receptors are destroyed by autoimmune reactions, narcolepsy may result10–13. Narcolepsy is an illness characterized by sleep attacks and daytime somnolence, as well as cataplexy (sudden transient weakness or paralysis), sleep paralysis, and hallucinations. The suvorexant inspiration is to help insomnia patients to sleep better by reducing orexigenic maintenance of wakefulness, perhaps similar to what occurs among narcoleptics14,15, but this idea has limitations. A characteristic of narcolepsy is disturbed nocturnal sleep16,17. Also, many insomnia patients arise out of bed during the night, and if treated with an orexin receptor antagonist, they might experience certain peculiar narcoleptic symptoms--more about this later. Narcoleptics may not experience more total 24-hour sleep than unaffected people, but more of their sleepiness and sleep tend to occur during the day16,17. Indeed, narcoleptics suffer daytime somnolence as characterized by a daytime “multiple sleep latency test.” Accordingly, narcolepsy is not usually characterized by a daytime feeling of being well-rested. Because of the relatively long half-life of suvorexant and its day-by-day accumulation, suvorexant might sometimes produce effects like narcolepsy symptoms during the day as well as at night.\n\nSome physicians advise against trying new drugs without proven advantages, until several years of long-term Phase IV monitoring has allowed more experience with the benefits and adverse effects. Let us review some of what is currently known about suvorexant immediate benefits and risks, to offer matters worth considering in making clinical choices in comparison with alternative hypnotics. I shall also emphasize what is unknown, concluding with issues of long-term benefits and risks that may ultimately be far more important than the immediate benefit/risk ratio.\n\n\nImmediate benefits of suvorexant and alternative hypnotics\n\nSince we do not have comparative controlled trials of suvorexant versus competing hypnotics given for insomnia, the best we can do is to review the evidence of suvorexant benefits versus placebo in randomized double-blind controlled trials. Then we can discuss whether these benefits are likely to be superior or equal to those of popular alternatives, even though randomized unbiased comparative trials are not available.\n\nMany insomnia patients consume hypnotics at bedtime hoping to benefit by better function on the following day. In some studies, suvorexant on average made various kinds of objectively-measured performance such as word recall and driving worse the next morning1, but no significant areas of improved objective function were documented1. If the primary hypnotic benefit desired is to improve next-day performance (measured objectively), suvorexant does not seem to offer that benefit. Quite the opposite. Note that many of the competitive popular hypnotics likewise make an insomnia patient's next-morning performance worse, not better18–20. It is conceivable that once a hypnotic is fully metabolized (often a variable number of hours after wake-up time), sedation would dissipate and objective performance might rebound. Moreover, considering that insomnia patients sometimes experience increased anxiety after taking a short-acting hypnotic21, and some hypnotics cause increased insomnia on the following night22, afternoon-evening rebound activation and accompanying performance enhancements might conceivably result from some short-acting hypnotics, but this enhancement has not been proven with statistical rigor23 and certainly not with suvorexant. Indeed, I know of no objective evidence that any hypnotic (approved in the U.S.) taken at bedtime improves the next-day performance of insomnia patients. I emphasize objective performance because (like alcoholics), intoxicated hypnotic patients commonly subjectively assert that their performance is enhanced when objective testing shows that it is not.\n\nProlonged-release melatonin (Circadin®), though not FDA-approved in the U.S., may be an exception to the general failure of sedative-hypnotics to improve next-day performance. Manufacturer-sponsored studies have reported several kinds of performance enhancement24,25, and there are some reported sleep and behavioral improvements among children with ADHD given ordinary melatonin26.\n\nSleep induction strengthens as the suvorexant dosage increases1,27. Although the manufacturer requested an initial suvorexant dosage ranging from 40 mg down to 15 mg, the FDA would only allow a recommended dose of 10 mg “not to exceed 20 mg daily”28, concluding that a lowered dosage was necessary to reduce the excessive risks produced by higher dosages29. The company's scientists were quoted as telling an FDA committee that “ten milligrams is ineffective,” from a patient’s point of view4,14. My opinion that 10 mg is generally ineffective agrees with that expressed at that time by the manufacturer. However, desperate to sleep, insomnia patients often take more than the recommended starting dose. Among the first 21 User Reviews of suvorexant listed at the popular WebMD internet site (www.webmd.com), 2 reported satisfaction with the recommended 10 mg starting dose, 13 reported taking more than 10 mg (as much as 40 mg, sometimes combined with other sedatives), and the others did not report their dosage information30. The FDA authorizes the nocturnal dosage to be increased to 20 mg if 10 mg proves well-tolerated but ineffective. It will be interesting to learn what dosages representative suvorexant patients actually choose to consume.\n\nIn the first night of polysomnographic data, 10 mg suvorexant decreased the latency to persistent sleep 3.4 min. (-15.6, 8.7, 95% Confidence Interval) more than placebo, i.e., there was no statistically significant benefit1,27,31. Likewise, 20 mg reduced the sleep latency by 9.4 min. (-21.5, 2.9) more than placebo, also not statistically significant, and not clinically significant compared to an initial sleep latency of about 70 minutes. At the end of week 4, 10 mg decreased latency to persistent sleep by 2.3 min. (still not significant), but 20 mg produced 22.3 min. (32.3,12.3) improvement compared to placebo, a statistically significant benefit27. The effects of 10 mg and 20 mg on polysomnographic latency to persistent sleep were found to somewhat greater (and entirely statistically significant) if the preplanned cross-over-phase data of the study were retrospectively excluded31. Also, the 10 mg dose reduced wake after sleep onset (WASO) by about 21 minutes at night 1 and after 4 weeks, which was statistically significant, and the 20 mg dose similarly decreased WASO by 24.7 and 28.1 min. respectively, both significant statistically31. Consequently, the 10 mg dose improved sleep efficiency (percent of in-bed time asleep) by 5.2% on night 1 and 4.7% at the end of week 4, and the 20 mg dose improved sleep efficiency 7.6% and 10.4% respectively, all of which were statistically significant but of uncertain clinical significance, considering that the starting sleep efficiencies were 65%–66%27,31. By patient self-report, moreover, with 10 mg and 20 mg doses given at night 1 and ending the 4th week, neither the subjective sleep latency nor the subjective total sleep time were improved with statistical or clinical significance27. These patients tended to underestimate the modest objective benefits of suvorexant, so many patients will not be satisfied with either the recommended or the “not to exceed” dosage.\n\nOddly enough, whereas the patients fairly consistently reported more subjective benefits at the 40 mg dose of suvorexant, that were both statistically and possibly clinically significant benefits (30 mg if age≥65 years), the polysomnographic data for 40 mg showed unimpressive advantages at the end of 4 weeks compared to the lower doses, and the adverse effects were distinctly more common1,27. This may have been one reason why the FDA insisted on the lower starting dosage.\n\nIt is important to keep in mind that the three-month studies described at length in the current Belsomra Prescribing Information28,32 supported the small-magnitude efficacy of the \"not to exceed\" dosage of 20 mg (15 mg for age≥65), not the efficacy of the recommended starting dose. The modest efficacies were similar in the three-month studies to those for the 20/15 mg group described in the multiple-dosage study described above. Though statistical significance was more robust in the three-month studies because of the larger group sizes, some of the outcomes still failed to achieve statistical significance at some time points. The recommended 10 mg dose had not been included in the Phase III studies, perhaps another indication that 10 mg was regarded as ineffective. The Phase IIB study described in the two previous paragraphs was the only randomized study reported that compared the 10 mg, 20 mg, and 40 mg doses along with placeboes27.\n\nOverall, comparing suvorexant augmentations of sleep with those reported for the benzodiazepines and benzodiazepine agonists in an authoritative meta-analysis33, all of the hypnotic categories seemed to produce benefits (or lack of benefits) in a similar range. That meta-analysis even questioned whether the “z” hypnotics significantly increased objective total sleep time33. After that meta-analysis, the FDA lowered the recommended doses for zolpidem and eszopiclone, but as with suvorexant, there are few controlled-trial results for the new lower recommended dosages. We do not know if the benefits of low-dose zolpidem and eszopiclone are as minimal as those of suvorexant. For example, the now-recommended 1 mg dosage of eszopiclone was ineffective in many PSG contrasts34,35. Without randomized comparative trials, one cannot rationally determine whether suvorexant produces as much benefit as the recently-popular hypnotics at currently-recommended doses, since the participants' ages, baseline sleep characteristics, and other factors varied among separate trials, as did elements of the trial designs. One can imagine that suvorexant would be particularly effective for the subgroup of insomnia patients with daytime hyperarousal, but so far no evidence has been produced. I suspect that suvorexant produces better reduction of WASO than popular short-acting hypnotics (although less reduction of sleep latency), but medium-half-life hypnotics such as temazepam and low-dose doxepin might have similar WASO efficacy, and low-dose doxepin may have comparatively fewer adverse effects36,37. To summarize, for suvorexant, greater overall efficacy than generic competitors at the recommended dosages does not appear likely.\n\nSuvorexant increases nocturnal sleep mainly by reducing WASO, similar to some alternative hypnotics, but unlike short-acting zaleplon, triazolam, or the standard-release zolpidem formulation. The suvorexant effect on the latency to fall asleep is quite weak at the recommended or “not to exceed” dosages due to slow absorption. Accordingly, suvorexant will be particularly unsatisfactory for patients primarily concerned with trouble falling asleep, but suvorexant may be preferred to the shortest-half-life hypnotics for patients who mainly complain of trouble staying asleep and early awakening i.e., WASO. Trouble staying asleep is more common than trouble falling asleep for patients over age 40, probably because circadian rhythms tend to peak progressively earlier from adolescence to old age unless dementia begins.\n\n\nImmediate risks of suvorexant and alternative hypnotics\n\nSuvorexant has some distinguishing risks, as well as most of the same immediate risks as the alternative hypnotics. Because suvorexant is not very rapidly absorbed (median Tmax of 2 hours, range 30 min. to 6 hours, with further delay of approximately 1.5 hours after a high-fat meal) and has an average half-life of approximately 12 hours28, a meaningful blood concentration usually persists throughout the day after prior-evening administration, and there is “an accumulation of approximately 1- to 2-fold with once-daily dosing, leading to an estimated 20% increase in the concentration after repeated dosing1,28. After 7 nights of administration, the suvorexant blood concentration remained so substantial during the day that just before the next evening dose, the lowest daytime concentration on the 7th day was more than half the maximal concentration achieved at Tmax during the first night38. Since receptor binding and release of orexins is quite indolent, the actions of suvorexant on neurons perhaps lag even later than the plasma Tmax and the stated half-life might suggest4,39. Moreover, since suvorexant is mainly metabolized by CYP3A and CYP2C19 enzymes28, the actions of which may be augmented or reduced by common genetic variants40 and other drugs, half-life and daytime accumulation may be quite variable or idiosyncratic. In healthy young adults, the maximum first-night concentrations can vary two-fold, obese females have an approximate 20% increase in morning-after blood levels, and strong CYP3A inhibitors result in three times the drug area under the curve1,38. Also, suvorexant might influence the metabolism of other drugs through CYP3A. The Prescribing Information recommends against use of suvorexant with strong CYP3A inhibitors28, but one may be skeptical how universally that caution can be observed.\n\nEvidently, the FDA intends that the 5 mg dosage be chosen for those using moderate CYP3A inhibitors or for patients who appear not to tolerate 10 mg well, whereas other patients may need the 20 mg dosage29. Above a 20 mg dosage, the FDA analysis concluded that benefits did not increase in proportion to the strong increase in disturbing adverse effects at the higher dosages. In 30–40 mg dosage groups, 2.8% of patients discontinued use within 3 months due to somnolence, fatigue, sedation, and lethargy combined, and additionally 0.2% also discontinued due to each of the following: nightmares, sleep paralysis, memory impairment, and depression1. In the 15–20 mg groups, the discontinuation rate for adverse events was only 0.6% as compared to 0.4% for placebo1.\n\nAs an orexin receptor antagonist, suvorexant appears to produce occasional narcolepsy-like symptoms, especially in the not-recommended 40 mg dosage, such as rare cataplexy (sudden weakness or paralysis), sleep paralysis, hypnagogic or hypnopompic hallucinations, and disturbing dreams1. Suvorexant seems unique among approved hypnotics in its narcolepsy-like adverse effects that can be frightening or temporarily disabling for a very small percentage of patients.\n\nLike most hypnotics with half-lives exceeding 3–6 hours, suvorexant causes daytime somnolence and fatigue among a percentage of users, but suvorexant in recommended doses did not appear to cause reported daytime somnolence more often than alternative hypnotics. In the Phase III trials, some patients suffered disabling sleepiness while driving the following morning. Driving impairments tended to be more severe with zopiclone 7.5 mg than with suvorexant 20 mg or 40 mg (30 mg if age≥65 years), but it was estimated that suvorexant might impair 10%–20% of adult patients on a driving test as much as would a blood alcohol level of 0.05–0.081. Note that zopiclone 7.5 mg contains about 3.75 mg eszopiclone, and patients and their physicians approaching such doses must be cautious of potential driving impairment. As with other hypnotics, it may be assumed that this daytime somnolence and these performance impairments can be augmented by combinations of suvorexant with other sedative drugs, narcotics, or alcohol38 that were generally avoided by participants selected for controlled trials. According to the Prescribing Information28, a variety of mental and behavioral impairments may occasionally occur among patients taking suvorexant such as amnesia, anxiety, hallucinations, and complex sleep behaviors. Symptoms of this kind, of which “zombie driving” is an example, occur with other hypnotics and have become somewhat notorious with triazolam and zolpidem41–44.\n\nIn the preapproval trials, suicidal ideation appeared to be a distinct risk of suvorexant, almost entirely at the 30–40 mg dosage level (0.6%)1. That should not be surprising, since suvorexant causes short REM sleep latency1, as is also associated with narcolepsy and depression, and narcolepsy is often treated with antidepressants45. Considering that orexin is increased during pleasure and inhibited during pain, one theory is that a link between narcolepsy and depression results from a changed balance of dynorphin and orexin7. Depression and suicide are likewise associated with many other hypnotics, based on both controlled trials demonstrating causality and epidemiologic studies46,47.\n\nIn a one-year controlled trial of suvorexant 30–40 mg versus placebo, those randomized to suvorexant experienced a dramatic increase in time to sleep onset, once the drug was withdrawn, so that even at the end of two months' drug-free follow-up, the sleep latency of suvorexant-withdrawn patients was subjectively 10–12 min. worse than that of patients who had previously received placebo throughout48,49. Simply comparing the subjectively-reported sleep of participants while receiving suvorexant vs placebo to the drug-free follow-ups, this withdrawal effect was glaringly apparent. Clinical trial investigators denied that “rebound” was a problem, having relied on a drug “rebound” criterion biased against demonstrating withdrawal effects and lacking statistical power50. Nevertheless, it is to the investigators' credit that they obtained a long two-month post-drug follow-up. This was the longest-lasting randomized, controlled demonstration of hypnotic-withdrawal insomnia of which I am aware48. Certainly, popular alternative hypnotics also produce drug-withdrawal insomnia22,49, but their withdrawal liabilities have not been studied with equivalent designs. Zolpidem 10 mg caused no appreciable problem in a 1-year study of somewhat different design with a somewhat anomalous outcome51. We do not know which drugs would cause more withdrawal distress given at the recommended dosages.\n\nSome hypnotics cause increased infections in randomized controlled trials, supported by extensive epidemiology52–55. When given suvorexant, patient infections and infestations overall were about equal with placebo, but there was a dose-response trend for more common \"URI\" reports among participants receiving suvorexant than placebo1. The controlled trial evidence for causing infections would appear stronger for alternative hypnotics than for suvorexant.\n\nTo examine effects of suvorexant on nocturnal respiration, patients with COPD and \"moderate\" sleep apnea were randomized to suvorexant 40 mg (30 mg if age≥65 years) or placebo for 4-night sleep recordings. Participants had a mean SpO2 of >94% awake and >93% during sleep, and mean BMI of 25.9. Nevertheless, suvorexant produced significantly reduced SpO2 both during wake and during sleep, and increased time below 85% SpO2, though these effects were quite small and were not considered clinically significant56. In a study of participants with \"mild or moderate\" sleep apnea and with average age 49, night 4 AHI (apnea-hypopnea index) was increased from 14.41 to 17.07 events per hour with suvorexant versus placebo, a difference of 2.66 (0.22 to 5.09), therefore significant57. Though these adverse effects did not appear clinically significant on average, in both studies suvorexant did impair nocturnal breathing. These were not the sorts of patients whose nocturnal breathing would be most vulnerable to a hypnotic and of greatest concern, e.g., those with marked nocturnal oxygen desaturation, obesity, and concomitant use of narcotics58 or other sedatives. Since alternative hypnotics also depress nocturnal respiration, it is unclear if suvorexant causes more respiratory risk than the alternatives.\n\nFalls are strongly associated with use of many hypnotics59–61, but falls among patients randomized to suvorexant were no more common than those among participants randomized to placebo1.\n\nLike most benzodiazepine-agonist hypnotics, suvorexant is thought to have some addiction potential and is rated Schedule IV by the DEA62. In contrast, doxepin, antihistamines, and melatonin are not controlled by the DEA.\n\nTo summarize, it seems unlikely that suvorexant could prove superior to alternative hypnotics in comparative trials focusing on the immediate benefits/risks ratios, because of 1) weak subjective benefit at low doses, 2) weak polysomnographic benefit for reducing sleep latency, 3) a relatively long half-life resulting in accumulation and daytime sedation, 4) particularly variable rates of absorption and CYP3A metabolism making dosing unpredictable, and 5) relatively unique narcolepsy-like symptoms with more-than-recommended doses. On the other hand, the long-term effects of hypnotics might be more important than their immediate effects.\n\n\nLong-term benefits and risks of suvorexant and alternative hypnotics\n\nMost patients who receive a prescription for a hypnotic consume the drug for only a brief time. However, the unusual patient who consumes a hypnotic nightly or several times a week for years receives so many prescriptions, that these heavy users consume most of the hypnotic drug market63,64. Among long-term habitual hypnotic consumers, there is a need to consider hypnotic benefits and risks not only for sleep but also for the long-term risks of dementia, cancer, and mortality. Unfortunately, there have been no long-term controlled trials assessing years of hypnotic usage by contrasting samples randomized to a hypnotic versus placebo, or comparing different hypnotics randomly assigned. Trials of cardiology drugs such as statins or the Women's Health Initiative long-term trial of estrogens assessed years of drug usage among tens of thousands of participants, but we have no comparable controlled trials of hypnotics.\n\nThe body tends to clear amyloid-β from brain intercellular regions during sleep, a process that may be inhibited during wakefulness by orexin65–69. This has led to speculation that orexin antagonists, such as suvorexant, might hypothetically reduce risks of Alzheimer's disease. However, one small study found evidence of an average amount of Alzheimer's amyloid plaque accumulation in brains of aged narcoleptics70. Further, suvorexant in the recommended dosages increases total sleep rather little, and the increment is mainly REM sleep rather than deep sleep4. Since it appears to be non-REM sleep that is associated with amyloid-β clearance rather than orexin itself68, any idea that suvorexant would have a beneficial effect on amyloid-β may be wishful thinking70. In contrast, there is more persuasive evidence that prior use of benzodiazepine-agonist hypnotics is associated with future Alzheimer's dementia71–73. Causality has not been proven.\n\nThere is suggestive evidence from small controlled trials that benzodiazepine-agonist hypnotics cause cancer. In a group of rather small controlled trials reviewed by the FDA, 13 incident cancer cases (mainly skin cancers) were found among patients randomized to hypnotics, but none were found among the sometimes-smaller randomized placebo groups74. Further, epidemiologic studies have supported an association of prior hypnotic use with cancer incidence64,75,76. It is controversial whether the presence of epidemiologic association might imply that benzodiazepine-agonist hypnotics cause cancer77,78. In the distinct case of suvorexant, my tabulation of the suvorexant randomized controlled trials reported to ClinicalTrials.gov (www.ClinicalTrials.gov) indicated no incident malignancies among 493 participants receiving suvorexant 20 mg (15 mg if age≥65 years), 9 incident malignancies among 1291 participants receiving suvorexant 40 mg (30 mg if age≥65 years), and 9 malignancies among 1025 participants randomized to placebo. Thus, incident cancers were less frequent among those randomized to suvorexant, particularly less than 30 mg as compared to placebo. These differences in cancer incidence were not statistically reliable in the suvorexant trials for these very infrequent cancer events. The FDA's approach to enumerating the incident neoplasms in suvorexant controlled trials produced slightly different tabulations, but essentially similar trends were described1. In summary, more data are needed, but there is a possibility that benzodiazepine agonist hypnotics are carcinogenic whereas suvorexant is not. Even the possibility that suvorexant is anti-neoplastic cannot be excluded.\n\nFinally, there are now more than 20 epidemiologic studies showing that use of benzodiazepine-agonists and diphenhydramine has been significantly associated with excess mortality, with hazard ratios as high as 3 to 564,79–81. A much smaller number of studies has observed no significant survival risk associated with hypnotics use, but no published studies yet have suggested any evidence that use of hypnotics improves survival. Indeed, some studies suggest that hypnotics have posed as much mortality risk as cigarettes64,79,82. Despite various efforts of many investigators to control for potential confounding in epidemiologic studies, it remains possible that this strong risk association is entirely due to statistical confounding, reflecting no causality83. A persuasive demonstration of mortality causation could only come from long-term randomized controlled trials or perhaps Mendelian randomization studies. At present, there is no epidemiologic evidence whether suvorexant use is associated with increased or decreased survival. There is as yet no evidence base permitting a guess about how suvorexant compares with alternative hypnotics for association with mortality.\n\n\nConclusions\n\nWith the limited available evidence, we can only guess about whether suvorexant is more or less effective and more or less safe than popular prescription hypnotics. However, none of them are very effective for increasing objective sleep or for improving daytime performance. Apart from some very small unpublished Phase II trials focused on special risks, there have been no randomized comparative trials examining whether the suvorexant benefits/risks ratio is better or worse than that of popular alternatives, so there are no bases for a clear preference save the amount of prior experience and costs. One might suppose that had the developers thought that suvorexant was superior in its immediate effects, they would have performed comparative clinical trials to highlight the advantages. Currently, suvorexant appears more expensive than many popular generic hypnotics. Suvorexant has so-far undergone little Phase IV safety surveillance. It appears that the overall balance of immediate benefits and risks with suvorexant most likely would be comparable or inferior to alternative hypnotics such as zolpidem (Table 2). As to long-term benefits and risks, we will have to hope that the industry conducts the necessary long-term comparative trials to assess which hypnotic compounds are safest and most beneficial.\n\nTable 2 summarizes some evidence concerning how suvorexant might compare with zolpidem, currently the most popular hypnotic in the United States. The comparisons for long-term risks and benefits are pure speculation.\n\nX Based on somewhat-parallel placebo-controlled-trials studies but no comparative-trials studies\n\n? Based only on non-comparative epidemiology and scientific speculation\n\n\nThe future choice of the best hypnotics\n\nThere are other orexin receptor antagonists in development. Possibly, a new orexin receptor antagonist will become available as a hypnotic with more reliable pharmacokinetics than suvorexant and a shorter Tmax and half-life. Perhaps such a drug might safely be given in a more effective bedtime dosage, with less danger of daytime adverse effects. A focus of future study will be how orexin receptor antagonists compare with alternative treatments of disturbed sleep such as cognitive behavioral treatment of insomnia or bright light treatment. There is a growing consensus that current data favor cognitive-behavioral therapy over hypnotics, and bright light may be superior for particular circadian rhythm sleep disorders manifesting as insomnia.\n\nLooking forward, I anticipate that better-organized exploitation of electronic medical records will produce increasing scrutiny of effects of hypnotics on inpatient falls, infections, and length of stay. Likewise, there will be increasing examination of the association of hypnotics with outpatient readmissions, infections, dementia, cancer, and mortality. Increasing use of genome-wide-association studies, exome sequencing, and whole-genome sequencing will make it possible to do Mendelian randomization studies that assess the causality of hypnotic associations with excess depression, infection, cancer, and mortality. Unfortunately, it will be many years before experience will be collected sufficient to apply Mendelian randomization strategies to orexin receptor antagonists. It is only speculation that in regard to long-term risks, particularly dementia, cancer, and mortality, suvorexant might be found safer than alternatives or even beneficial. Unless and until the industry can provide us with long-term-trials evidence of more distinct suvorexant advantages, cautious and cost-conscious physicians and their patients may prefer the alternatives.", "appendix": "Competing interests\n\n\n\nSince 1979 publication of hypnotics epidemiology from the American Cancer Society CPSI study, the author has been a frequent critic of hypnotics risks and benefits, especially through his non-profit internet web site, www.DarkSideOfSleepingPills.com, that provides readers with more extensive information and references about risks of hypnotics. Dr. Kripke's family owns stock and options in a large conglomerate that in turn invested a tiny percentage of its capital in Sanofi-Aventis stock. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, other stock ownership or options, expert testimony, grants or patents received or pending or royalties. No writing assistance was utilized in the production of this manuscript.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nSeveral boarded sleep specialists at the Scripps Clinic Viterbi Family Sleep Center kindly read and provided comments on a previous draft of this manuscript.\n\n\nReferences\n\nFarkas RH, Katz R, Illoh K, et al.: Application Number 204569Orig1s000: Medical Review(s). 2013. Reference Source\n\nMignot E: Sleep, sleep disorders and hypocretin (orexin). Sleep Med. 2004; 5(Suppl 1): S2–S8. PubMed Abstract \| Publisher Full Text\n\nMieda M, Sakurai T: Orexin (hypocretin) receptor agonists and antagonists for treatment of sleep disorders. Rationale for development and current status. CNS Drugs. 2013; 27(2): 83–90. PubMed Abstract \| Publisher Full Text\n\nJacobson LH, Callander GE, Hoyer D: Suvorexant for the treatment of insomnia. Expert Rev Clin Pharmacol. 2014; 7(6): 711–30. PubMed Abstract \| Publisher Full Text\n\nAlexandre C, Andermann ML, Scammell TE: Control of arousal by the orexin neurons. Curr Opin Neurobiol. 2013; 23(5): 752–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNishino S, Sakurai E, Nevsimalova S, et al.: Decreased CSF histamine in narcolepsy with and without low CSF hypocretin-1 in comparison to healthy controls. Sleep. 2009; 32(2): 175–80. PubMed Abstract \| Free Full Text\n\nShan L, Dauvilliers Y, Siegel JM: Interactions of the histamine and hypocretin systems in CNS disorders. Nat Rev Neurol. 2015; 11(7): 401–13. PubMed Abstract \| Publisher Full Text\n\nLin JS: Brain structures and mechanisms involved in the control of cortical activation and wakefulness, with emphasis on the posterior hypothalamus and histaminergic neurons. Sleep Med Rev. 2000; 4(5): 471–503. PubMed Abstract \| Publisher Full Text\n\nSaper CB: The neurobiology of sleep. Continuum (Minneap Minn). 2013; 19(1 Sleep Disorders): 19–31. PubMed Abstract\n\nOllila HM, Ravel JM, Han F, et al.: HLA-DPB1 and HLA class I confer risk of and protection from narcolepsy. Am J Hum Genet. 2015; 96(1): 136–46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMahlios J, De la Herrán-Arita AK, Mignot E: The autoimmune basis of narcolepsy. Curr Opin Neurobiol. 2013; 23(5): 767–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaraco J, Lin L, Kornum BR, et al.: ImmunoChip study implicates antigen presentation to T cells in narcolepsy. PLoS Genet. 2013; 9(2): e1003270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAhmed SS, Volkmuth W, Duca J, et al.: Antibodies to influenza nucleoprotein cross-react with human hypocretin receptor 2. Sci Transl Med. 2015; 7(294): 294ra105. PubMed Abstract \| Publisher Full Text\n\nParker I: THE BIG SLEEP: insomnia drugs like Ambien are notorious for their side effects. Has Merck created a blockbuster replacement? The New Yorker. 2013: 50–63. Reference Source\n\nMignot E: Physiology. The perfect hypnotic? Science. 2013; 340(6128): 36–8. PubMed Abstract \| Publisher Full Text\n\nRogers AE, Aldrich MS, Caruso CC: Patterns of sleep and wakefulness in treated narcoleptic subjects. Sleep. 1994; 17(7): 590–7. PubMed Abstract\n\nThorpy M: Current concepts in the etiology, diagnosis and treatment of narcolepsy. Sleep Med. 2001; 2(1): 5–17. PubMed Abstract \| Publisher Full Text\n\nJohnson LC, Chernik DA: Sedative-hypnotics and human performance. Psychopharmacology (Berl). 1982; 76(2): 101–13. PubMed Abstract \| Publisher Full Text\n\nVerster JC, Veldhuijzen DS, Patat A, et al.: Hypnotics and driving safety: meta-analyses of randomized controlled trials applying the on-the-road driving test. Curr Drug Saf. 2006; 1(1): 63–71. PubMed Abstract \| Publisher Full Text\n\nBoyle J, Groeger JA, Paska W, et al.: A method to assess the dissipation of the [corrected] residual effects of [corrected] hypnotics: eszopiclone versus zopiclone. J Clin Psychopharmacol. 2012; 32(5): 704–9. PubMed Abstract \| Publisher Full Text\n\nTan TL, Bixler EO, Kales A, et al.: Early morning insomnia, daytime anxiety, and organic mental disorder associated with triazolam. J Fam Pract. 1985; 20(6): 592–4. PubMed Abstract\n\nWalsh JK, Roth T, Randazzo A, et al.: Eight weeks of non-nightly use of zolpidem for primary insomnia. Sleep. 2000; 23(8): 1087–96. PubMed Abstract\n\nZemlan FP, Mulchahey J, Scharf MB, et al.: The efficacy and safety of the melatonin agonist beta-methyl-6-chloromelatonin in primary insomnia: a randomized, placebo-controlled, crossover clinical trial. J Clin Psychiatry. 2005; 66(3): 384–90. PubMed Abstract \| Publisher Full Text\n\nLuthringer R, Muzet M, Zisapel N, et al.: The effect of prolonged-release melatonin on sleep measures and psychomotor performance in elderly patients with insomnia. Int Clin Psychopharmacol. 2009; 24(5): 239–49. PubMed Abstract \| Publisher Full Text\n\nWade AG, Farmer M, Harari G, et al.: Add-on prolonged-release melatonin for cognitive function and sleep in mild to moderate Alzheimer’s disease: a 6-month, randomized, placebo-controlled, multicenter trial. Clin Interv Aging. 2014; 9: 947–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRossignol DA, Frye RE: Melatonin in autism spectrum disorders: a systematic review and meta-analysis. Dev Med Child Neurol. 2011; 53(9): 783–92. PubMed Abstract \| Publisher Full Text\n\nHerring WJ, Snyder E, Budd K, et al.: Orexin receptor antagonism for treatment of insomnia: a randomized clinical trial of suvorexant. Neurology. 2012; 79(23): 2265–74. PubMed Abstract \| Publisher Full Text\n\nMerck Sharp & Dohme Corp.: BELSOMRA Prescribing Information. Whitehouse Station, NJ, Merck Sharp & Dohme. 2014. Reference Source\n\nUnger EF: Office Director Decisional Memo. 2014. Reference Source\n\nvarious. WebMD User Reviews & Ratings - Belsomra orgal. 2015. Reference Source\n\nMerck Sharp & Dohme Corp.: Phase IIB 2-Period Crossover Polysomnography Study in Participants With Primary Insomnia (MK-4305-006). Bethesda, MD, U.S. National Institutes of Health. 2015. Reference Source\n\nHerring WJ, Connor KM, Ivgy-May N, et al.: Suvorexant in Patients with Insomnia: Results from Two 3-Month Randomized Controlled Clinical Trials. Biol Psychiatry. 2014, in press; pii: S0006-3223(14)00762-8. PubMed Abstract \| Publisher Full Text\n\nBuscemi N, Vandermeer B, Friesen C, et al.: The efficacy and safety of drug treatments for chronic insomnia in adults: a meta-analysis of RCTs. J Gen Intern Med. 2007; 22(9): 1335–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRosenberg R, Caron J, Roth T, et al.: An assessment of the efficacy and safety of eszopiclone in the treatment of transient insomnia in healthy adults. Sleep Med. 2005; 6(1): 15–22. PubMed Abstract \| Publisher Full Text\n\nScharf M, Erman M, Rosenberg R, et al.: A 2-week efficacy and safety study of eszopiclone in elderly patients with primary insomnia. Sleep. 2005; 28(6): 720–7. PubMed Abstract\n\nRojas-Fernandez CH, Chen Y: Use of ultra-low-dose (≤6 mg) doxepin for treatment of insomnia in older people. Can Pharm J (Ott). 2014; 147(5): 281–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatz R, Farkas R, Cai J: Application Number: 022036Orig1s00: Medical Reviews (FDA). Accessed 12-4-2009.\n\nDimova H, Brar S, Men A: Application Number: 204569Orig1s000: Clinical Pharmacology And Biopharmaceutics Review(s). 2014. Reference Source\n\nCallander GE, Olorunda M, Monna D, et al.: Kinetic properties of \"dual\" orexin receptor antagonists at OX1R and OX2R orexin receptors. Front Neurosci. 2013; 7: 230. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShen M, Shi Y, Xiang P: CYP3A4 and CYP2C19 genetic polymorphisms and zolpidem metabolism in the Chinese Han population: a pilot study. Forensic Sci Int. 2013; 227(1–3): 77–81. PubMed Abstract \| Publisher Full Text\n\nTsai JH, Yang P, Chen CC, et al.: Zolpidem-induced amnesia and somnambulism: rare occurrences? Eur Neuropsychopharmacol. 2009; 19(1): 74–6. PubMed Abstract \| Publisher Full Text\n\nPressman MR: Sleep driving: Sleepwalking variant or misuse of z-drugs? Sleep Med Rev. 2011; 15(5): 285–92. PubMed Abstract \| Publisher Full Text\n\nPoceta JS: Zolpidem ingestion, automatisms, and sleep driving: a clinical and legal case series. J Clin Sleep Med. 2011; 7(6): 632–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorgenthaler TI, Silber MH: Amnestic sleep-related eating disorder associated with zolpidem. Sleep Med. 2002; 3(4): 323–7. PubMed Abstract \| Publisher Full Text\n\nVignatelli L, D'Alessandro R, Candelise L: Antidepressant drugs for narcolepsy. Cochrane Database Syst Rev. 2008; (1): CD003724. PubMed Abstract \| Publisher Full Text\n\nKripke DF: Greater incidence of depression with hypnotic use than with placebo. BMC Psychiatry. 2007; 7: 42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGunnell D, Chang SS, Tsai MK, et al.: Sleep and suicide: an analysis of a cohort of 394,000 Taiwanese adults. Soc Psychiatry Psychiatr Epidemiol. 2013; 48(9): 1457–65. PubMed Abstract \| Publisher Full Text\n\nMichelson D, Snyder E, Paradis E, et al.: Safety and efficacy of suvorexant during 1-year treatment of insomnia with subsequent abrupt treatment discontinuation: a phase 3 randomised, double-blind, placebo-controlled trial. Lancet Neurol. 2014; 13(5): 461–71. PubMed Abstract \| Publisher Full Text\n\nKripke DF: Hypnotics cause insomnia: evidence from clinical trials. Sleep Med. 2014; 15(9): 1168–9. PubMed Abstract \| Publisher Full Text\n\nKripke DF: \"Rebound\" is not an appropriate criterion for withdrawal insomnia. Sleep Med. 2014; 15(12): 1594. PubMed Abstract \| Publisher Full Text\n\nRoehrs TA, Randall S, Harris E, et al.: Twelve months of nightly zolpidem does not lead to rebound insomnia or withdrawal symptoms: a prospective placebo-controlled study. J Psychopharmacol. 2012; 26(8): 1088–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoya FL, Kripke DF, Loving RT, et al.: Meta-analyses of hypnotics and infections: eszopiclone, ramelteon, zaleplon, and zolpidem. J Clin Sleep Med. 2009; 5(4): 377–83. PubMed Abstract \| Free Full Text\n\nObiora E, Hubbard R, Sanders RD, et al.: The impact of benzodiazepines on occurrence of pneumonia and mortality from pneumonia: a nested case-control and survival analysis in a population-based cohort. Thorax. 2013; 68(2): 163–70. PubMed Abstract \| Publisher Full Text\n\nIqbal U, Syed-Abdul S, Nguyen PA, et al.: The impact of benzodiazepines on occurrence of pneumonia and mortality from pneumonia: a nested case-control and survival analysis in a population-based cohort. Thorax. 2013; 68(6): 591–2. PubMed Abstract \| Publisher Full Text\n\nHuang CY, Chou FH, Huang YS, et al.: The association between zolpidem and infection in patients with sleep disturbance. J Psychiatr Res. 2014; 54(7): 116–20. PubMed Abstract \| Publisher Full Text\n\nSun H, Palcza J, Rosenberg R, et al.: Effects of suvorexant, an orexin receptor antagonist, on breathing during sleep in patients with chronic obstructive pulmonary disease. Respir Med. 2015; 109(3): 416–26. PubMed Abstract \| Publisher Full Text\n\nMerck Sharp & Dohme Corp.: Effects of suvorexant in participants with obstructive sleep apnea (MK-4305-036). Bethesda, MD, National Institutes of Health. ClinicalTrials.gov NCT01300455. 2015. Reference Source\n\nWebster LR, Choi Y, Desai H, et al.: Sleep-disordered breathing and chronic opioid therapy. Pain Med. 2008; 9(4): 425–32. PubMed Abstract \| Publisher Full Text\n\nDiem SJ, Ewing SK, Stone KL, et al.: Use of non-benzodiazepine sedative hypnotics and risk of falls in older men. J Gerontol Geriatr Res. 2014; 3(3): 158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerry SD, Lee Y, Cai S, et al.: Nonbenzodiazepine sleep medication use and hip fractures in nursing home residents. JAMA Intern Med. 2013; 173(9): 754–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKolla BP, Lovely JK, Mansukhani MP, et al.: Zolpidem is independently associated with increased risk of inpatient falls. J Hosp Med. 2013; 8(1): 1–6. PubMed Abstract \| Publisher Full Text\n\nHarrigan TM: Schedules of controlled substances: Placement of Suvorexant into Schedule IV. 21 CFR Part 1308 [Docket No. DEA-381], Washington, D.C., Federal Register. Accessed 8-28-2014; 79(167); 51243–51247. Reference Source\n\nKripke DF, Garfinkel L, Wingard DL, et al.: Mortality associated with sleep duration and insomnia. Arch Gen Psychiatry. 2002; 59(2): 131–6. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Langer RD, Kline LE, et al.: Hypnotics’ association with mortality or cancer: a matched cohort study. BMJ Open. 2012; 2(1): e000850. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKang JE, Lim MM, Bateman RJ, et al.: Amyloid-beta dynamics are regulated by orexin and the sleep-wake cycle. Science. 2009; 326(5955): 1005–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiguori C, Romigi A, Nuccetelli M, et al.: Orexinergic system dysregulation, sleep impairment, and cognitive decline in Alzheimer disease. JAMA Neurol. 2014; 71(12): 1498–505. PubMed Abstract \| Publisher Full Text\n\nXie L, Kang H, Xu Q, et al.: Sleep drives metabolite clearance from the adult brain. Science. 2013; 342(6156): 373–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoh JH, Jiang H, Finn MB, et al.: Potential role of orexin and sleep modulation in the pathogenesis of Alzheimer’s disease. J Exp Med. 2014; 211(13): 2487–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDauvilliers YA, Lehmann S, Jaussent I, et al.: Hypocretin and brain β-amyloid peptide interactions in cognitive disorders and narcolepsy. Front Aging Neurosci. 2014; 6: 119. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScammell TE, Matheson JK, Honda M, et al.: Coexistence of narcolepsy and Alzheimer’s disease. Neurobiol Aging. 2012; 33(7): 1318–9. PubMed Abstract \| Publisher Full Text\n\nChen PL, Lee WJ, Sun WZ, et al.: Risk of dementia in patients with insomnia and long-term use of hypnotics: a population-based retrospective cohort study. PLoS One. 2012; 7(11): e49113. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBillioti de Gage S, Bégaud B, Bazin F, et al.: Benzodiazepine use and risk of dementia: prospective population based study. BMJ. 2012; 345: e6231. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBillioti de Gage S, Moride Y, Ducruet T, et al.: Benzodiazepine use and risk of Alzheimer’s disease: case-control study. BMJ. 2014; 349: g5205. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF: Possibility that certain hypnotics might cause cancer in skin. J Sleep Res. 2008; 17(3): 245–50. PubMed Abstract \| Publisher Full Text\n\nKao CH, Sun LM, Liang JA, et al.: Relationship of zolpidem and cancer risk: a Taiwanese population-based cohort study. Mayo Clin Proc. 2012; 87(5): 430–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKao CH, Sun LM, Su KP, et al.: Benzodiazepine use possibly increases cancer risk: a population-based retrospective cohort study in Taiwan. J Clin Psychiatry. 2012; 73(4): e555–e560. PubMed Abstract \| Publisher Full Text\n\nPottegård A, Friis S, Andersen M, et al.: Use of benzodiazepines or benzodiazepine related drugs and the risk of cancer: a population-based case-control study. Br J Clin Pharmacol. 2013; 75(5): 1356–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF, Langer RD: Evidence for harm, comment on ‘Use of benzodiazepines or benzodiazepine related drugs and the risk of cancer: a population-based case-control study’. Br J Clin Pharmacol. 2014; 78(1): 186–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMallon L, Broman JE, Hetta J: Is usage of hypnotics associated with mortality? Sleep Med. 2009; 10(3): 279–86. PubMed Abstract \| Publisher Full Text\n\nWeich S, Pearce HL, Croft P, et al.: Effect of anxiolytic and hypnotic drug prescriptions on mortality hazards: retrospective cohort study. BMJ. 2014; 348: g1996. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen HC, Su TP, Chou P, et al.: A nine-year follow-up study of sleep patterns and mortality in community-dwelling older adults in Taiwan. Sleep. 2013; 36(8): 1187–98. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF, Klauber MR, Wingard DL, et al.: Mortality hazard associated with prescription hypnotics. Biol Psychiatry. 1998; 43(9): 687–93. PubMed Abstract \| Publisher Full Text\n\nLevine M: ACP Journal Club. Hypnotic drugs were associated with increased risk for mortality. Ann Intern Med. 2012; 156(12): JC6–13. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Simons RN, Garfinkel L, et al.: Short and long sleep and sleeping pills. Is increased mortality associated? Arch Gen Psychiatry. 1979; 36(1): 103–16. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9764", "date": "05 Aug 2015", "name": "Jerome M. Siegel", "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\nKripke presents a thorough analysis of the risks and possible benefits of suvorexant for the treatment of insomnia. Suvorexant is an antagonist for the two receptors for hypocretin (also called orexin), a peptide released by a small group of neurons in the hypothalamus. Soon after the peptide was identified, it was found that 90% of neurons containing this peptide are lost in human narcolepsy1,2. Since one of the main symptoms of narcolepsy is sleepiness, it seemed plausible that a drug blocking hypocretin receptors would cause sleepiness, an effect that might be useful in treating insomnia. Suvorexant is the first such drug to hit the market. Human insomnia is a complaint about inadequate sleep, but is not necessarily correlated with low sleep duration or with decreased lifespan3,4. Current insomnia treatments act on GABA receptors, particularly on the benzodiazepine type of GABA receptor. An obvious problem with manipulation of systemic GABA levels is the very large number of GABA neurons and GABA receptors in the brain. GABA receptors exist not only in regions such as the anterior hypothalamus and adjacent forebrain regions implicated in sleep induction, but throughout the brain. In some regions nearly 90% of neurons contain GABA5. Benzodiazepine receptors also exist in large numbers in bodily organs including the heart6,7, gall bladder, urinary bladder8, thyroid, liver9, lung, stomach10,11, testes11, pancreas10 and kidneys10,12 and are activated by many commonly used sleeping pills13,14. Benzodiazepine receptors are present on red blood cells, on tumors, as well as on cells of the immune system15-19. Increased rates of infection have been reported with the use of hypnotics20. In contrast to GABA, there only about 75,000 hypocretin neurons in the human brain2,21,22, a tiny fraction of the 75,000,000,000 neurons estimated to be in the human brain. They are distributed from the most medial portions of the hypothalamus adjacent to the 3rd ventricle, to the far lateral hypothalamus. Although initial reports suggested that there were orexin neurons in the gut, these reports have not been replicated23. Hypocretin neurons have widespread projections, directly innervating and activating cortical, subcortical and brainstem neurons24.\n\nSome work has suggested potential problems with dual orexin receptor antagonists. Humans who have attempted suicide have reduced levels of hypocretin-1 in their cerebrospinal fluid25. In a study of human patients with electrodes implanted in the amygdala for diagnostic purposes, we found that hypocretin release was maximal during pleasure and was minimal when they reported feeling sad or when they were in pain, despite a high level of arousal26. Allowing for species differences, these human data bear considerable resemblance to data on hypocretin neuron activity in animals. In normal mice, we found that hypocretin neurons are maximally active during performance of rewarded behaviors27 and that hypocretin knockout mice were strikingly deficient in staying awake to perform rewarded behaviors. Our studies of Fos expression in wild type mice also showed that hypocretin neurons were not activated beyond baseline levels during foot shock, or foot shock avoidance behavior, despite high levels of EEG arousal27. We have also reported that hypocretin neuronal activity in rats is suppressed in novel situations eliciting withdrawal, despite maximal levels of EEG activation. In contrast, hypocretin neuron activity is high during grooming and exploration28. The animal and human studies indicate that hypocretin cells are not simply related to arousal, but are strongly related to positive emotions. This is consistent with the evidence that human depression and reported difficulties with social interaction in narcolepsy may result from the loss of hypocretin function, as would occur with receptor antagonists29-34. These considerations suggest that depression and even suicide might be a risk from the use of orexin receptor antagonists. However, this need not be the case if the drug induces sleep rapidly, and does not persist in the brain. Kripke reviews evidence suggesting that chronic use produces longterm inactivation of hypocretin receptors, highlighting this risk.", "responses": [] }, { "id": "10375", "date": "18 Sep 2015", "name": "Børge Sivertsen", "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\nTitle and Abstract: The title and abstract is well formulated and indeed representative for the rest of the full text paper. Being an opinion article, it is made clear from the beginning – including the abstract - that no new original data are presented; rather an expert’s summarization of available evidence regarding the use of suvorexant. Article content: Dr. Kripke provides a very thorough review of available literature, both published and unpublished reports, with regards to various aspects of suvorexant. Existing evidence (or lack thereof) of both efficacy, side-effects, risk/benefit ratio, comparisons studies etc., are clearly presented, and the conclusions drawn from these reports are well-balanced. Conclusions: The conclusions are very clearly stated, yet sensible, balanced as well as justified on the basis of the available data regarding suvorexant. I consider this an important contribution that will improve our understanding, and provide a solid scientific foundation useful for both clinicians and researchers alike.", "responses": [] } ]	1	https://f1000research.com/articles/4-456
https://f1000research.com/articles/4-100/v1	28 Apr 15	{ "type": "Research Note", "title": "Assessing the bipotency of in vitro-derived neuromesodermal progenitors", "authors": [ "Anestis Tsakiridis", "Valerie Wilson" ], "abstract": "Retrospective clonal analysis in the mouse has demonstrated that the posterior spinal cord neurectoderm and paraxial mesoderm share a common bipotent progenitor. These neuromesodermal progenitors (NMPs) are the source of new axial structures during embryonic rostrocaudal axis elongation and are marked by the simultaneous co-expression of the transcription factors T(Brachyury) (T(Bra)) and Sox2. NMP-like cells have recently been derived from pluripotent stem cells in vitro following combined stimulation of Wnt and fibroblast growth factor (FGF) signaling. Under these conditions the majority of cultures consist of T(Bra)/Sox2 co-expressing cells after 48-72 hours of differentiation. Although the capacity of these cells to generate posterior neural and paraxial mesoderm derivatives has been demonstrated at the population level, it is unknown whether a single in vitro-derived NMP can give rise to both neural and mesodermal cells. Here we demonstrate that T(Bra) positive cells obtained from mouse epiblast stem cells (EpiSCs) after culture in NMP-inducing conditions can generate both neural and mesodermal clones. This finding suggests that, similar to their embryonic counterparts, in vitro-derived NMPs are truly bipotent and can thus be exploited as a model for studying the molecular basis of developmental cell fate decisions.", "keywords": [ "Neuromesodermal progenitors", "Axis elongation", "Pluripotent stem cells", "Paraxial mesoderm", "Neurectoderm", "In vitro differentiation", "Primitive streak" ], "content": "Introduction\n\nAxis elongation in vertebrate embryos proceeds in a rostral-to-caudal sequence and involves the coordinated production of spinal cord neurectoderm and paraxial mesoderm/somites from a population of neuromesodermal progenitors (NMPs) (for a review see1). The bipotent status of these axial stem cells was demonstrated in the mouse by retrospective clonal analysis2. NM-potent cells are located in the node-streak border and the adjacent caudal lateral epiblast of early somite stage embryos and in the chordoneural hinge (CNH) region of the tail bud of later stage embryos3–5 i.e. in areas exhibiting high levels of Wnt and FGF signaling1. The main hallmark of these cells is the co-expression of the mesodermal transcription factor T (Bra) together with the neural marker Sox26–9. NMPs are not only an excellent model for deciphering the mechanisms controlling cell fate choice (neuroectoderm vs mesoderm), but also comprise an attractive source for generating trunk spinal cord neurectoderm cells and skeletal muscle in vitro.\n\nWe and others have recently shown that mouse and human pluripotent stem cells cultured for 48–72 hours in the presence of FGF2 and the Wnt signaling agonist CHIRON99021 (CHIR) yield a high percentage of T(Bra)+Sox2+ double-positive cells that transcriptionally resemble embryonic NMPs10,11. These NMP-like cells were also shown to efficiently differentiate exclusively into paraxial mesoderm and posterior neurectoderm both in vitro and in vivo upon grafting into cultured mouse and chick embryos10 suggesting an NM bipotent character. However, these studies were carried out at the population level and it would thus be important to test the NM potency of single cells. Here we address this issue by showing, through the clonal plating of T(Bra)+ cells generated after culture of epiblast stem cells (EpiSCs)12,13 in NMP-inducing conditions, that individual in vitro-derived NMPs are truly bipotent as they give rise to colonies consisting of both neural and mesodermal cells.\n\n\nMethods\n\nT(Bra)-green fluorescent protein (GFP) reporter (TGFP) EpiSCs were derived from TGFP embryonic stem (ES) cells (sourced from 14) and cultured routinely in fibronectin-treated plates in N2B27 medium containing 10 ng/ml FGF2 (R&D Systems) and 20 ng/ml Activin A (Peprotech) as previously described in15. For NMP differentiation TGFP EpiSCs were plated at a density of approximately 1500–2000/cm2 in N2B27 medium containing 20 ng/ml FGF2 and 3 µM CHIRON99021 (Stemgent) on fibronectin for 48–72 hrs10. For clonal plating experiments in vitro-derived NMPs were pre-treated with 10 µM ROCK inhibitor Y-27632 (Calbiochem) for 1 hr prior to fluorescence-activated cell sorting (FACS). After this they were re-plated at a density of 4,000 cells/well in 12-well plates in medium containing either FGF2, or FGF2/CHIR alongside Y-27632 for the first 8 hours. We have previously found that when 1:1 mixtures of GFP+ and GFP- EpiSCs are plated at a total of 5,000 cells/well in 12-well plates (or 10,000 cells/well in 6-well plates) then 95% of the resulting colonies between 2–8 cells are of monoclonal origin. Here we also included for scoring colonies of up to 10 cells since we employ a smaller initial plating density (4,000 cells/well)6. For non-clonal plating of in vitro-derived NMPs, approximately 40,000 cells/well (12-well plate) were used. Cell sorting was performed using a FACSAria (BD Biosciences).\n\nFor immunocytochemistry cells were fixed with 4% paraformaldehyde, washed with PBS/0.1% Triton X-100 (PBST), treated with 0.5 M Glycine and blocked in PBST/3% donkey serum/7.5% bovine serum albumin (BSA). Primary antibody incubations were performed overnight at 4°C, followed by PBST washes the following day, incubation with secondary donkey Alexafluor antibodies (Life Technologies) for 2–3 hrs at room temperature and further washes in PBST. The primary antibodies used were: donkey polyclonal anti-T(Bra), 1 μg/ml (RRID: R&D Systems Cat# AF2085 RRID:AB_2200235), rabbit monoclonal anti-Sox2, 0.5 μg/ml (RRID: Abcam Cat# ab92494 RRID:AB_10585428) and goat polyclonal anti-Tbx6, 0.5 μg/ml (RRID: R&D Systems Cat# AF4744 RRID:AB_2200834). Fluorescent images were captured using an Olympus IX51 inverted microscope (Olympus) using a x20 objective and the Volocity software (PerkinElmer). Nuclear segmentation followed by single cell fluorescence quantification was performed as described previously16. T(Bra) and Sox2 protein positivity scoring of individual clones was carried out manually.\n\n\nResults\n\nTo track the emergence of NMPs in vitro we employed a T(Bra) reporter EpiSC line (TGFP) generated from ES cells carrying a GFP transgene knocked into the T(Bra) locus14. This reporter line has been shown to faithfully recapitulate endogenous T(Bra) expression. In line with our previous findings10, culture of TGFP EpiSCs in the presence of FGF2/CHIR for 48 or 72 hours gave rise to a significant number of TGFP+ cells, many of which were also positive for Sox2 expression (55% of the total TGFP+ population at 48 hours and 65% at 72 hours) as revealed by antibody staining and image analysis (Figure 1). This result indicates that at least half of the TGFP+ cells emerging in the presence of FGF2/CHIR are NMP-like and thus we used TGFP expression under these conditions to enrich for cells with NMP identity.\n\nLeft: Fluorescence analysis of TGFP and Sox2 expression in TGFP EpiSCs cultured for 48 hours in FGF2/CHIR following antibody staining against Sox2. Right: Quantification of TGFP+Sox2+ and TGFP+Sox2- expressing cells in TGFP EpiSCs differentiated in NMP-inducing conditions after 2 (d2) or 3 (d3) days following immunocytochemistry and image analysis.\n\nWe have previously found that prolonged (i.e. more than 72 hours) culture in FGF2/CHIR mediates further differentiation of NMPs into mutually exclusive paraxial mesoderm and neurectoderm cells10. Therefore apart from promoting an NMP state these conditions simultaneously provide an environment for the production of the natural differentiation products of NMPs. We thus utilized culture in FGF2/CHIR in order to test the NM potency of TGFP+ NMPs at the population level. TGFP EpiSCs were cultured in NMP-promoting conditions for 48 hours and the resulting GFP+ cells were sorted by flow cytometry and re-plated at high density for a further 48–72 hours in the presence of FGF2/CHIR (Figure 2A). We have previously shown that under these conditions hardly any pluripotent cells persist in the differentiating cultures as evidenced by analysis of Nanog/Oct4 expression and grafting into the pluripotency-permissive environment of cultured E7.5 embryos10. Immunofluorescence analysis of the final cultures showed that sorted TGFP+ cells generated predominantly mutually exclusive single T(Bra) positive mesoderm and single Sox2+ neurectoderm (Figure 2B). The cultures also contained clusters of Tbx6+ cells which were distinct from the T(Bra)+ and Sox2+ domains (Figure 2B) and, since this gene uniquely marks emergent paraxial mesoderm, these cells probably arose from the T(Bra)-expressing population. Together these data confirm that the TGFP-expressing cells produced in NMP inducing conditions possess the ability to generate both neural and mesodermal cells upon further differentiation.\n\n(A) Scheme depicting the differentiation and re-plating of in vitro induced NMPs at high density after flow sorting. (B) Fluorescence analysis and immunocytochemistry of TGFP, Sox2 and Tbx6 expression of in vitro-derived NMPs sorted at day 2 of differentiation and re-plated at high density in the presence of FGF2/CHIR for 2 days. In all cases cell nuclei were visualized using DAPI. IF: immunofluorescence.\n\nWe next examined the behaviour of TGFP+ NMPs at the single cell level. TFGP+ cells induced after 48 or 72 hrs of FGF2/CHIR treatment were flow sorted (purity >99%) and re-plated at clonal density in FGF2/CHIR-containing medium (Figure 3A, B). After 48 hours the resulting colonies were analysed by immunofluorescence and categorized based on their composition (Figure 3C). Strikingly, most (55–60% of total) clones obtained from both day 2 and day 3 FGF2/CHIR-induced TGFP+ cells were composed exclusively of single Sox2+ neurectodermal cells indicating a strong neurogenic capacity (Figure 3D, E). The proportion of single Sox2+ colonies was significantly enhanced to 76% (p value<0.05 based on a two-tailed z test) with a concomitant decrease in the proportion of T(Bra)+ cells when isolated single TGFP+ cells produced after 2 days in FGF2/CHIR medium were re-plated in the presence of FGF2 alone for 48 hrs prior to clone scoring (Figure 3D) confirming the pro-mesodermal effect of Wnt activity on NMPs8,10. We also observed purely mesodermal clones consisting of T(Bra)+ cells which were particularly prominent in the case of sorted day two TGFP+ NMPs (Figure 3D, E). These data suggest that many in vitro-derived NMP cells are biased by the signaling environment towards unilinear differentiation into either neurectoderm or mesoderm. However, we did observe clones which comprised combinations of single positive T(Bra)+ and Sox2+ cells (7% for day 2 and 12% for day 3 TGFP+ NMPs) and were thus indicative of neuromesodermal potency. A few clones were found to contain only T(Bra)+Sox2+ double positive cells (Figure 3D, E) possibly reflecting NMP self-renewal. Finally, a small number of colonies were composed only of T(Bra)-Sox2- negative cells (Figure 3D, E) which are likely to represent more differentiated NMP derivatives. Interestingly, we detected no Tbx6+ cells present in the clones (Representative, raw images shown in Dataset 5) despite their presence in cultures derived from sorted day 2 FGF2/CHIR-induced TGFP+ cells plated at high density under the same conditions. This suggests that maturation of T(Bra)+ cells into Tbx6-positive paraxial mesoderm depends on paracrine signaling effects which are absent from the low density, clonally-derived cultures.\n\n(A) Scheme depicting the differentiation and re-plating of in vitro induced NMPs at clonal density after flow sorting. (B) FACS plots depicting analysis of TGFP expression in day 3 FGF2/CHIR-treated TGFP EpiSCs (middle). The purity of the GFP+ sorted population and a negative control (wild-type EpiSCs) are also shown. (C) Representative examples of the clones obtained after culture of single sorted TGFP+ NMPs in FGF2/CHIR medium following immunofluorescence analysis of T(Bra) and Sox2 expression. The colour-coded bars on the right correspond to the scoring groups shown at the top of panel 3D. (D–E) Composition of colonies obtained after clonal plating of TGFP+ NMPs sorted at day 2 (D) or day 3 (E) for a further 48 hrs in FGF2/CHIR or FGF2 only. Pie charts: overall percentages of clones of each phenotype. Total numbers of clones scored are shown below each pie chart.\n\n\n\n\nDiscussion\n\nThe production of axial tissues during embryonic elongation is driven by posteriorly-located progenitors emerging round the end of gastrulation. A long-standing question in the field has been whether this cell population represents a mixture of separate unipotent neural and mesoderm-committed precursors or consists of bipotent progenitors. Genetic marking of single cells and their derivatives using the LaacZ system in mouse embryos shed light on this problem by revealing that spinal cord neurectoderm and paraxial mesoderm originate from bipotent neuromesodermal progenitors2. These NMPs have also recently been captured in vitro through the culture of pluripotent stem cells in Wnt and FGF signaling agonists10,11. However, the bipotent status of these cells had not been previously demonstrated at the clonal level. Here we show that single in vitro-derived NMPs can give rise to mixed clones containing both neural (Sox2+T(Bra)-) and mesodermal (Sox2-T(Bra)+) cells, a finding which indicates that FGF2/CHIR-induced cultures contain bona fide NM bipotent cells.\n\nInterestingly, a considerable fraction of individual sorted NMPs produced exclusively neurectodermal or mesodermal clones suggesting that a proportion of the Sox2+T(Bra)+ cells induced from EpiSCs after 2–3 days in the presence of FGF2/CHIR may already be biased towards adopting a neural or mesodermal fate under conditions promoting both lineages. This may be a reflection of heterogeneity in the relative levels of Sox2 and T(Bra) protein/transcript within the in vitro-derived Sox2+T(Bra)+ population with double-positive cells exhibiting higher levels of Sox2 showing a pro-neural bias while T(Bra)High cells are predisposed to mesoderm differentiation. Indeed such heterogeneity in Sox2 and T(Bra) levels (as well as other mesodermal and neural transcripts) has been shown by single cell transcriptomic analysis of mouse ES cell-derived cultures resembling our in vitro-generated NMPs11. Nevertheless, the clonal-based assay we employed here establishes bipotency of in vitro-derived NMPs and reveals the responsiveness of individual cells to environmental signals.\n\n\nData availability\n\nFigshare: Supplementary data for ‘Assessing the bipotency of in vitro-derived neuromesodermal progenitors’ doi: 10.6084/m9.figshare.137100117", "appendix": "Author contributions\n\n\n\nAT designed and performed experiments. AT, VW prepared the manuscript. All authors have agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the Medical Research Council (MR/K011200).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe wish to thank Fiona Rossi and Clair Cryer (MRC Centre for Regenerative Medicine FACS facility) for their assistance with flow cytometry.\n\n\nReferences\n\nWilson V, Olivera-Martinez I, Storey KG: Stem cells, signals and vertebrate body axis extension. Development. 2009; 136(10): 1591–1604. PubMed Abstract \| Publisher Full Text\n\nTzouanacou E, Wegener A, Wymeersch FJ, et al.: Redefining the progression of lineage segregations during mammalian embryogenesis by clonal analysis. Dev Cell. 2009; 17(3): 365–376. PubMed Abstract \| Publisher Full Text\n\nCambray N, Wilson V: Two distinct sources for a population of maturing axial progenitors. Development. 2007; 134(15): 2829–2840. PubMed Abstract \| Publisher Full Text\n\nCambray N, Wilson V: Axial progenitors with extensive potency are localised to the mouse chordoneural hinge. Development. 2002; 129(20): 4855–4866. PubMed Abstract\n\nMcGrew MJ, Sherman A, Lillico SG, et al.: Localised axial progenitor cell populations in the avian tail bud are not committed to a posterior Hox identity. Development. 2008; 135(13): 2289–2299. PubMed Abstract \| Publisher Full Text\n\nTsakiridis A, Huang Y, Blin G, et al.: Distinct Wnt-driven primitive streak-like populations reflect in vivo lineage precursors. Development. 2014; 141(6): 1209–1221. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOlivera-Martinez I, Harada H, Halley PA, et al.: Loss of FGF-dependent mesoderm identity and rise of endogenous retinoid signalling determine cessation of body axis elongation. PLoS Biol. 2012; 10(10): e1001415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartin BL, Kimelman D: Canonical Wnt signaling dynamically controls multiple stem cell fate decisions during vertebrate body formation. Dev Cell. 2012; 22(1): 223–232. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelfino-Machin M, Lunn JS, Breitkreuz DN, et al.: Specification and maintenance of the spinal cord stem zone. Development. 2005; 132(19): 4273–4283. PubMed Abstract \| Publisher Full Text\n\nGouti M, Tsakiridis A, Wymeersch FJ, et al.: In vitro generation of neuromesodermal progenitors reveals distinct roles for Wnt signalling in the specification of spinal cord and paraxial mesoderm identity. PLoS Biol. 2014; 12(8): e1001937. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTurner DA, Hayward PC, Baillie-Johnson P, et al.: Wnt/β-catenin and FGF signalling direct the specification and maintenance of a neuromesodermal axial progenitor in ensembles of mouse embryonic stem cells. Development. 2014; 141(22): 4243–4253. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrons IG, Smithers LE, Trotter MW, et al.: Derivation of pluripotent epiblast stem cells from mammalian embryos. Nature. 2007; 448(7150): 191–195. PubMed Abstract \| Publisher Full Text\n\nTesar PJ, Chenoweth JG, Brook FA, et al.: New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature. 2007; 448(7150): 196–199. PubMed Abstract \| Publisher Full Text\n\nFehling HJ, Lacaud G, Kubo A, et al.: Tracking mesoderm induction and its specification to the hemangioblast during embryonic stem cell differentiation. Development. 2003; 130(17): 4217–4227. PubMed Abstract \| Publisher Full Text\n\nGuo G, Yang J, Nichols J, et al.: Klf4 reverts developmentally programmed restriction of ground state pluripotency. Development. 2009; 136(7): 1063–1069. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOsorno R, Tsakiridis A, Wong F, et al.: The developmental dismantling of pluripotency is reversed by ectopic Oct4 expression. Development. 2012; 139(13): 2288–2298. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsakiridis A, Wilson V: Supplementary data for ‘Assessing the bipotency of in vitro-derived neuromesodermal progenitors’. Figshare. 2015. Data Source" }	[ { "id": "8486", "date": "07 May 2015", "name": "Patrick P. L. Tam", "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 study on the differentiation of single NMP progenitors generated by FGF2/CHIR treatment of EpiSC is a sequel to the previous study on lineage differentiation of these cells at the population level. Findings of this study are consistent with the assertion that individual T+/Sox2+ cells are likely to possess dual potential for differentiation into mesoderm (T+ only) and neuroectodermal (Sox2+ only) cells. This provides the requisite experimental evidence that some cells in vitro may have acquire the attributes of the bipotent NMPs, which are presumed to exist in the node-streak interface and the chordoneural hinges in vivo.Issues to be clarified:Inconsistency of experimental dataData shown in Fig 3D indicates that T+ clones contain 2 and 4 cells (the majority) with a few having up to 6 cells. The example of T+ve clone shown in Fig 3C contains at least 9 cells. This result is not included the dataset of Fig 3D (Day 2 sorted) or E (Day 3 sorted). Were all or only subsets of clones scored for these clonal culture experiments? Quality of the image dataa. It is difficult to discern the co-expression, or otherwise, of T-GFP and Sox2 in individual cells at the resolution of Fig 1 and 2. b. It appears that cells with mixed gene expression are only found in some colonies (Fig 1), rather than in a salt and pepper manner in every colony. This may require an explanation in the context of clonal development. c. It can be confusing when different colour schemes were used to show the fluorescence results, e.g. T-GFP signals are shown variously in green (Fig 1), white (Fig 2B) and red (Fig 3C), and Sox2 is shown in red (Fig 1) and green (Fig 2B, 3C), rather than red (which is for Tbx6, Fig 2B). Given that the FGF2/CHIR treated cells were sorted based on GFP activity, T-GFP signal should consistently be displayed- in green for all figures. Additional data / information may help:a. While it is plausible that the Tbx6+ cells might be descendents of the mesoderm progenitor, the results do not unequivocally show that they are derived from the T+ve cells.b. What is the evidence for that T-/Sox2- cells (which also did not expressing Tbx6) were “more differentiated” NMP derivatives?c. Absence of Tbx6 cells in low density culture is an intriguing result. Is there any precedence that the differentiation of Tbx6-expressing cells is dependent on any “paracrine” signals?d. Were the Sox2 and Tbx6 immunofluorescence signals captured in emission channel other than for green fluorescence? If not, would the IF results for these two markers be confounded by the T-GFP background?e. Which are the examples of two types of mixed clones in the legend (“T+ mixed with Sox2+” and “T+/Sox+ with T-/Sox2-“) in the FGF2/CHIR and FGF experiments (Fig. 3D)?f. The bottom panel of Fig 3C: The “T+/Sox2+ mixed with T-/Sox2-” clone showed no T+/Sox2+ cells among the four cells in this figure.g. Is there any difference in the clonal types between “Sorted at D2-IF at D4” and “Sorted at D3-IF at Day 5 FGF2/CHIR” groups? What is the rationale for testing the effect of an extended culture to Day 3 before sorting, and was there a parallel culture of “FGF2 only” to Day 3?", "responses": [ { "c_id": "1482", "date": "31 Jul 2015", "name": "Anestis Tsakiridis", "role": "Author Response", "response": "Our responses below are shown in italics while reviewer comments are in bold.Issues to be clarified:Inconsistency of experimental dataData shown in Fig 3D indicates that T+ clones contain 2 and 4 cells (the majority) with a few having up to 6 cells. The example of T+ve clone shown in Fig 3C contains at least 9 cells. This result is not included the dataset of Fig 3D (Day 2 sorted) or E (Day 3 sorted). Were all or only subsets of clones scored for these clonal culture experiments?We wish to thank the reviewers for spotting the mistake. All clone examples shown in Fig. 3C are taken from the culture experiments described in this study and are included in the scoring graphs in Fig. 3D-E. The T(Bra)+Sox2- clone the reviewers are referring to was mistakenly shown in Fig. 3E as consisting of 7 instead of 9 cells. We have rectified the mistake in the new version (3rd clone from the left, depicted in green). We also re-examined all data and found two further errors:(1) a T(Bra)+Sox2+ double positive clone (also shown as an example in Fig. 3C, top row) was accidentally omitted from the top graph in Fig. 3D(2)a 3-cell clone in the same graph was wrongly depicted as containing two Sox2+ cells and a double negative instead of two Sox2+ cells and a T(Bra)+ cell.These changes have now been incorporated into the new figure version and the data is now correct to the best of our knowledge. We would like to apologise for these errors. Quality of the image dataa. It is difficult to discern the co-expression, or otherwise, of T-GFP and Sox2 in individual cells at the resolution of Fig 1 and 2. We have now included higher magnification images in figures 1 and 2 to address this issue.b. It appears that cells with mixed gene expression are only found in some colonies (Fig 1), rather than in a salt and pepper manner in every colony. This may require an explanation in the context of clonal development.This is an intriguing observation supporting the idea that NMPs segregate separately from other mesodermal/neural precursors emerging simultaneously in culture conditions promoting a late primitive streak-like environment. It is likely to reflect the fact that the starting EpiSC population is heterogeneous in self-renewing conditions, consisting of cells with differential capacity to generate separate distinct lineages. For example we have previously reported that T(Bra)+Sox2+ and T(Bra)+Foxa2+ mesodermal precursors emerge in a mutually exclusive manner after Wnt signaling stimulation of EpiSCs (Tsakiridis et al., 2014). It may also be true that the generation of NMP-like, T(Bra)+Sox2+ cells is non-synchronous. We have added a sentence in the first paragraph of the results section to comment on this.c. It can be confusing when different colour schemes were used to show the fluorescence results, e.g. T-GFP signals are shown variously in green (Fig 1), white (Fig 2B) and red (Fig 3C), and Sox2 is shown in red (Fig 1) and green (Fig 2B, 3C), rather than red (which is for Tbx6, Fig 2B). Given that the FGF2/CHIR treated cells were sorted based on GFP activity, T-GFP signal should consistently be displayed- in green for all figures.We have adopted the reviewers’ recommendation and in the new figure versions T(Bra) expression appears in green while Sox2 is depicted in the red. Additional data / information may help:a. While it is plausible that the Tbx6+ cells might be descendents of the mesoderm progenitor, the results do not unequivocally show that they are derived from the T+ve cells.We agree that the data presented here do not show directly that Tbx6+ cells are derived from T(Bra)+ cells. To indicate this we added the word “presumed” before the sentence: “maturation of T(Bra)+ cells into Tbx6-positive paraxial mesoderm” at the end of the Results section. However, we believe that our hypothesis is reasonable since it is based on published studies demonstrating that a considerable fraction of paraxial mesoderm arises from T(Bra)+ late primitive streak precursors (e.g. Anderson et al. 2013; Cambray and Wilson, 2007).b. What is the evidence for that T-/Sox2- cells (which also did not expressing Tbx6) were “more differentiated” NMP derivatives?We are interested in characterizing further the T-/Sox2- clones (see also our response to Referees 1 above). We speculate that these are differentiated NMP-derivatives based on the observation that EpiSCs generating NMPs do not (as far as we can tell) lose Sox2 expression during that process and thus its subsequent loss is likely to indicate further differentiation to downstream NMP products. However, in the absence of any solid data on this point we decided to tone down our statements and expand our hypotheses:1) We modified our statement related to the double negative clones (last paragraph of results section) by replacing the phrase “are likely to represent” with “may represent”2) We raised the alternative possibility that these clones may arise from sorted single T(Bra)+ mesodermal precursors.c. Absence of Tbx6 cells in low density culture is an intriguing result. Is there any precedence that the differentiation of Tbx6-expressing cells is dependent on any “paracrine” signals?Emergence of Tbx6+ cells is likely to depend on FGF signaling (Ciruna and Rossant, 2001)-we have now incorporated this possibility into the text together with the speculation that Notch-based regulation of Tbx6 transcription may also be critical for cell density effects (White et al., 2005).We have also shown previously that paracrine Wnt signaling promotes pioneer T(Bra)+ mesodermal precursors in EpiSCs cultured in self-renewal conditions (Tsakiridis et al., 2014).d. Were the Sox2 and Tbx6 immunofluorescence signals captured in emission channel other than for green fluorescence? If not, would the IF results for these two markers be confounded by the T-GFP background?We find this possibility unlikely especially since the Sox2+, Tbx6+ and TGFP+ expression domains shown in Fig. 2 were in most cases mutually exclusive from each other and very few or no double or triple positive cells (which could potentially represent a background artefact) were observed.e. Which are the examples of two types of mixed clones in the legend (“T+ mixed with Sox2+” and “T+/Sox+ with T-/Sox2-“) in the FGF2/CHIR and FGF experiments (Fig. 3D)?The class of clones defined as “T+ mixed with Sox2+” refers to colonies containing T(Bra)+ single positive cells mixed with Sox2+ single positive cells (example shown in the 5th row in Fig. 3C) while the group defined as “T+/Sox+ mixed with T-Sox2- “ describes clones containing T(Bra)+ OR Sox2+ single positive cells mixed with double negatives (example shown in bottom row in Fig. 3C). We would like to apologise for the lack of clarity. We tried to address this issue by making the legend of Fig. 3D-E more detailed.f. The bottom panel of Fig 3C: The “T+/Sox2+ mixed with T-/Sox2-” clone showed no T+/Sox2+ cells among the four cells in this figure.We use the slash (“/”) to indicate “OR” (see our response directly above). Again we apologise for the confusion.g. Is there any difference in the clonal types between “Sorted at D2-IF at D4” and “Sorted at D3-IF at Day 5 FGF2/CHIR” groups? What is the rationale for testing the effect of an extended culture to Day 3 before sorting, and was there a parallel culture of “FGF2 only”to Day 3?We observed a significant number of T(Bra)+Sox2+ cells both at 48 and 72 hrs of culture in FGF2/CHIR and thus we wished to examine whether these two populations are equivalent in their capacity to clonally generate both neural and mesodermal cells. Since the goal was to compare these two populations, we did not carry out an experiment involving a 3 day culture FGF2 prior to sorting and re-plating." } ] }, { "id": "8488", "date": "12 May 2015", "name": "Jacqueline Deschamps", "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\nTsakiridis and Wilson present the evidence that single cells from a neuromesodermal axial progenitor (NMP) population produced in vitro from mouse epiblast stem cells (EpiSCs) can give rise to both neural and mesodermal cell descendants, and are therefore truly bipotent.In earlier work they had demonstrated the existence of NMPs in the posterior aspect of the developing early somite embryos, and characterized these NMPs as being T Brachyury/Sox2 double positive. In more recent experiments they and others succeeded in defining high Wnt and high Fgf signaling conditions to culture EpiSCs into a cell population wherein more than half the cells are NMP-like. However, it remained to be proven that an individual NMP-like cell expressing both T Brachyury and Sox2 is able to generate neural derivatives (expressing exclusively Sox2), and mesodermal derivatives (expressing exclusively T Brachyury). It is what the authors achieved in this report, by elegantly making use of EpiSCs derived from T Bra Gfp transgenic embryos. By fluorescence activated cell sorting applied to these EpiSCs cultured in NMP-promoting conditions, they could purify the T Bra positive NMPs and analyze their clonal descendants. They clearly obtained neural and mesodermal descendants from single NMP clones, and could demonstrate that the environmental culture conditions influence the bias of individual NMPs to differentiate into cells with a neural versus mesodermal fate.The experiments are well designed and executed. The methods are well described, and the results are clear, thoroughly analyzed and discussed appropriately. The data in this manuscript convey a clear message that represents an advance in the field.", "responses": [] }, { "id": "8552", "date": "12 May 2015", "name": "Kate G. Storey", "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\nFollowing a study on the differentiation of dual-fated Neuromesodermal Progenitors (NMPs) at the population level, Tsakiridis and Wilson describe here the ability of a single NMP cell to generate both neural and mesodermal derivatives. Using T(Bra)-GFP expressing EpiSCs, the authors performed a clonal analysis of NMP differentiation driven by a FGF2/CHIR regime. They show that single T(Bra)-GFP expressing cells can generate both mesodermal T(Bra)-GFP and neural Sox2 expressing cells in the same clone. This main finding demonstrates the neuromesodermal bipotency of in vitro derived NMP at the single cell level, recapitulating the behaviour of NMPs as identified by retrospective clonal analysis in the mouse embryo in Tzouanacou et al. 2009. Overall, the experiments presented are well designed and the results are carefully analyzed. However, the paper would benefit from improvement of specific points. Main comments: MethodsThe authors should comment on the use of the ROCK inhibitor Y-27632 in their protocol for cell sorting and during subsequent plating at low density.Experimental approaches Figure 2 describes the co-appearance of distinct T(Bra)+ and Sox2+ cells from an NMP population treated by FGF2/CHIR as a way to induce simultaneously neural and mesodermal lineages. Why did the authors not perform neural and mesodermal differentiation in parallel using two distinct protocols? In this study, we are dependent on the previous finding that cells plated at low density give rise mainly (95%) to monoclonal colonies in the time frame of the experiment. It would be reassuring here if the authors had demonstrated that sorted individual T-GFP +ve cells were also Sox2 co-expressing at the time of plating – even if this required immuno-cytochemistry, it would at least establish the proportion of bra/sox2 co-expressing cells in the starting conditions. Does the negative control in Fig3B represent WT EpiSCs treated with the FGF2/CHIR regime, or are they just undifferentiated cells? A good negative control would be to analyse T(Bra)-GFP undifferentiated EpiSCs as they will have the same genetic background as the NMPs analyzed later but won’t express T(Bra). This control should be included as well. Figure 3C shows clones of various sizes in terms of cell number. Can the size of the clone have an impact on cell lineage identity? It would be good to standardize the analysis by looking at the different expression patterns in colonies with comparable cell numbers. Figure 3D and E, the authors comment on the appearance of T(Bra)/Sox2 double negative colonies, and suggest that they are likely to represent further differentiated derivatives. The authors should address this by looking at other neural and mesodermal markers, such as Pax6 or Sox1 and Paraxis. It would be very informative to know what those cells become. Indeed, the suggestion that a higher density of cells is needed for paraxial mesoderm differentiation might indicate that those negative clones are not mesodermal derivatives.Minor comments:In figure 1, the authors describe the establishment of an NMP population. Using a FGF2/CHIR differentiation protocol, they only obtain around 38% of T(Bra)-GFP/Sox2 coexpression, with maximum 60% of T(Bra)+ cells in the whole population. However, other studies show up to 80% of coexpression in the same conditions. The authors should comment on that. Figure 2B: the figure legend should indicate “all nuclei” instead of “ell nuclei”. The legend for figure 3 (D-E) is confusing and should be clarified.", "responses": [ { "c_id": "1481", "date": "31 Jul 2015", "name": "Anestis Tsakiridis", "role": "Author Response", "response": "We would like to thank all three reviewers both for their positive comments and constructive suggestions which improved significantly the quality of our manuscript. Our responses below are shown in italics while reviewer comments are in bold.Main comments: MethodsThe authors should comment on the use of the ROCK inhibitor Y-27632 in their protocol for cell sorting and during subsequent plating at low density.The ROCK inhibitor Y-27632 is a well-established reagent in human embryonic (hES) and epiblast stem cell (EpiSC) culture used to enhance survival of single cells (Watanabe et al., 2007) and thus we routinely employ it in FACS sorting experiments as a means of counteracting dissociation-induced apoptosis.Experimental approachesFigure 2 describes the co-appearance of distinct T(Bra)+ and Sox2+ cells from an NMP population treated by FGF2/CHIR as a way to induce simultaneously neural and mesodermal lineages. Why did the authors not perform neural and mesodermal differentiation in parallel using two distinct protocols?Clonal analysis is required because pluripotent stem cell differentiation is never 100% efficient at the population level. We thus aimed to assess the ability of single NMP cells to generate both neural and mesodermal derivatives at the same time using conditions promoting the simultaneous emergence of both lineages (i.e. FGF/CHIR treatment). In this study, we are dependent on the previous finding that cells plated at low density give rise mainly (95%) to monoclonal colonies in the time frame of the experiment. It would be reassuring here if the authors had demonstrated that sorted individual T-GFP +ve cells were also Sox2 co-expressing at the time of plating – even if this required immuno-cytochemistry, it would at least establish the proportion of bra/sox2 co-expressing cells in the starting conditions.We agree that ideally the extent of T(Bra)/Sox2 co-expression should be assessed at the time of low density plating. However, in our opinion, its determination can only be achieved using a T/Sox2 double reporter cell line. The alternative option suggested by the reviewers involving the use of immunocytochemistry would be technically challenging given that freshly sorted and plated TGFP+ cells require a few hours to attach properly thus precluding antibody staining at the very start of the experiment. Does the negative control in Fig3B represent WT EpiSCs treated with the FGF2/CHIR regime, or are they just undifferentiated cells? A good negative control would be to analyse T(Bra)-GFP undifferentiated EpiSCs as they will have the same genetic background as the NMPs analyzed later but won’t express T(Bra). This control should be included as well.The control used in Fig. 3B was indeed differentiated E14tg2a EpiSCs. This is the same genetic background as the TGFP+ ES cells (E14.1, 129/Ola; Fehling et al., 2003) we used to derive the EpiSC line employed in this study. It is not possible to utilize undifferentiated TGFP EpiSCs as a negative control for FACS as they also express significant levels of both T(BRA) protein and the TGFP reporter under self-renewing conditions (i.e. in FGF2 and Activin A) in line with previous reports (Tsakiridis et al., 2014; Kurek et al., 2015). Figure 3C shows clones of various sizes in terms of cell number. Can the size of the clone have an impact on cell lineage identity? It would be good to standardize the analysis by looking at the different expression patterns in colonies with comparable cell numbers.Splitting the data in Fig. 3C-D based on clone size would be a good way to decipher a link between colony cell number and lineage identity acquisition, a possibility which is indeed very interesting. However, this type of representation would be more suitable for a larger dataset. Our clone numbers are too small to support any solid conclusions on this issue and we believe that the representation we opted for is the best way to illustrate graphically both parameters (i.e. clone size and lineage composition). We will be happy to incorporate any specific suggestions and or/consider alternative ways of depicting the results. Figure 3D and E, the authors comment on the appearance of T(Bra)/Sox2 double negative colonies, and suggest that they are likely to represent further differentiated derivatives. The authors should address this by looking at other neural and mesodermal markers, such as Pax6 or Sox1 and Paraxis. It would be very informative to know what those cells become. Indeed, the suggestion that a higher density of cells is needed for paraxial mesoderm differentiation might indicate that those negative clones are not mesodermal derivatives.This is a good point which deserves further investigation. Our preliminary data indicate that T(Bra)-Sox2- colonies are also negative for Sox1. However, a thorough analysis of these clones will require significant effort and is beyond the scope of this short research note. We have added a sentence in the last paragraph of the results section raising the possibility that double negative clones may also comprise differentiated derivatives of single sorted TGFP+Sox2- cells which are probably precursors of mesodermal cell types other than paraxial, also emerging upon culture in FGF2/CHIR.Minor comments:In figure 1, the authors describe the establishment of an NMP population. Using a FGF2/CHIR differentiation protocol, they only obtain around 38% of T(Bra)-GFP/Sox2 coexpression, with maximum 60% of T(Bra)+ cells in the whole population. However, other studies show up to 80% of coexpression in the same conditions. The authors should comment on that.Line-to-line variation in terms of differentiation potential is a common phenomenon in pluripotent stem cell cultures (e.g. see Osafune et al., 2008) and in our hands we also observe some variation between different EpiSC lines both in terms of the extent of induction of NMP-like cells upon culture in FGF2/CHIR as well as the timing of their emergence. The lower numbers of T(Bra)+Sox2+ double positive cells we observed in this study when compared to the high efficiency of induction exhibited by the in vivo derived EpiSC line R04-GFP (Gouti et al., 2014) is an example of such variation. One interesting possibility is raised by the fact that the T(Bra) reporter line we employ here contains a GFP cassette knocked into the first exon of the T(Bra) gene (Fehliing et al., 2003). The resulting heterozygosity may lead to lower efficiency of T(Bra)+Sox2+ cell generation.Figure 2B: the figure legend should indicate “all nuclei” instead of “ell nuclei”.We cannot locate the phrase the reviewers are referring to. The exact figure legend wording in Fig. 2B is “In all cases cell nuclei were visualized using DAPI”. The legend for figure 3 (D-E) is confusing and should be clarified.We have addressed this point in the new version." } ] } ]	1	https://f1000research.com/articles/4-100
https://f1000research.com/articles/4-446/v1	31 Jul 15	{ "type": "Case Report", "title": "Case Report: Whole exome sequencing helps in accurate molecular diagnosis in siblings with a rare co-occurrence of paternally inherited 22q12 duplication and autosomal recessive non-syndromic ichthyosis.", "authors": [ "Aayush Gupta", "Yugal Sharma", "Kirti Deo", "Shamsudheen Vellarikkal", "Rijith Jayarajan", "Vishal Dixit", "Ankit Verma", "Vinod Scaria", "Sridhar Sivasubbu", "Aayush Gupta", "Yugal Sharma", "Kirti Deo", "Shamsudheen Vellarikkal", "Rijith Jayarajan", "Vishal Dixit", "Ankit Verma" ], "abstract": "Lamellar ichthyosis (LI), considered an autosomal recessive monogenic genodermatosis, has an incidence of approximately 1 in 250,000. Usually associated with mutations in the transglutaminase gene (TGM1), mutations in six other genes have, less frequently, been shown to be causative. Two siblings, born in a collodion membrane, presented with fish like scales all over the body. Karyotyping revealed duplication of the chromosome arm on 22q12+ in the father and two siblings. Whole exome sequencing revealed a homozygous p.Gly218Ser variation in TGM1; a variation reported earlier in an isolated Finnish population in association with autosomal recessive non-syndromic ichthyosis. This concurrence of a potentially benign 22q12+ duplication and LI, both rare individually, is reported here likely for the first time.", "keywords": [ "Whole Exome Sequencing", "Lamellar ichthyosis", "chr22q12", "genodermatosis", "TGM1" ], "content": "Introduction\n\nAutosomal recessive congenital ichthyosis (ARCI), a heterogeneous disorder of cornification of skin, encompasses three clinical subtypes: lamellar ichthyosis (LI; OMIM 242300); congenital ichthyosiform erythroderma (CIE; OMIM 242100); and harlequin ichthyosis (HI; OMIM 242500)1. LI has an incidence of approximately 1 in 250,000. Over 115 mutations in TGM1 and less frequent ones in six other genes (NIPAL4z, ALOX12B, CGI-58, FLJ39501, ICHYN and ABCA12) have been associated with the LI/CIE phenotypic spectrum worldwide2,3. Overlapping phenotypes and the non-specificity of the conventional histopathology, makes clinical diagnosis challenging in many cases and inaccurate in some4. Whole exome sequencing has become a useful diagnostic aid for genetic disorders including multigene dermatoses such as epidermolysis bullosa5,6 and acrokeratosis verruciformis7.\n\n\nCase report\n\nTwo, 8 and 6-year-old, siblings born out of a non-consanguineous marriage (Figure 1a) presented with hyperpigmented fish-like scales all over the body including face and flexures ectropion, loss of lateral half of eyebrows and alopecia along the scalp margins (Figure 1b–1e). Both siblings were heat intolerant, photosensitive and hypohidrotic. Born uneventfully vaginally they were encased in a collodion membrane which was shed within a week of birth. There was no family history of any dermatoses. Slit lamp examination revealed bilateral keratitis. Karyotyping of their parents and the siblings performed previously revealed duplication of the chromosome arm on 22q12+ in the father and two siblings. The patients were put on daily oral (5 mg) isotretinoin after analysing their lipid profile.\n\na) Pedigree of the family; (b), (c) and (d), (e) correspond to the ventral and dorsal views of siblings II:1, and II:2 respectively and shows hyperpigmented fish-like scales all over the body including face and flexures, ectropion, loss of lateral half of eyebrows and hair along scalp margins. Panel (f) shows the chromatogram from capillary sequencing for the parents and siblings, while panel (g) shows the domain organization of the protein and the location of the p.Gly218Ser variation with respect to the protein domains.\n\nConsidering the diagnosis of LI, whole exome sequencing was attempted. Genomic DNA (gDNA) was isolated from 5 ml of blood8 of each of the affected children after obtaining written informed consent conforming to the institutional ethical committee approvals (Dr. D.Y. Patil Vidyapeeth, Pune. Approval number DYPV/EC/178/14 ). The whole exome capture and library preparation (Nextera expanded exome, Illumina Inc., USA) were carried out according to the manufacturer’s instructions and followed by high throughput sequence generation on Hiseq 2500 with default 101 paired end single index sequence by synthesis chemistry (Illumina Inc., USA). The raw sequence reads were trimmed at a Phred score of 30 leaving over 44.9 and 33.45 million reads respectively for the two siblings. The variations were called against the hg19 version of the human genome using standard GATK-Picard pipeline with Burrows-Wheeler Alignment according to GATK best practice9. The variants from the genomes of both siblings were further analysed using ANNOVAR10 for coding region and also screened using the NCBI-Clinvar database (http://www.ncbi.nlm.nih.gov/clinvar/). Analysis revealed a homozygous p.Gly218Ser variation in TGM1 previously reported to be associated with autosomal recessive non-syndromic ichthyosis in an isolated Finnish population11. The variant mapped to the transglutaminase domain in the protein (Figure 1g) and was also predicted to be pathogenic by both SIFT (Sorts Intolerant From Tolerant)15 and PolyPhen216 (Polymorphism Phenotyping v2). This variation was further validated in parents and both siblings by site-specific PCR using a forward primer CTTCTCCTGGGGTCAGGCA and reverse primer GAGAAGTCCCAGGCTCCATC (Sigma Aldrich). The PCR was done using taq polymerase (Invitrogen, USA, Cat. No. 10342053) according to the manufacturer supplied protocol with a Tm of 60.5°C. The PCR products were size selected and gel purified (2% agarose) using qiaquick gel extraction kit (QIAGEN, NL) and performed capillary sequencing (Applied Biosystems) performed using manufacturer instruction. Analysis revealed the variant was heterozygous in parents, while homozygous in both affected siblings (Figure 1f).\n\nFollow up after two years of low dose isotretinoin, titrated intermittently, revealed complete subsidence of ectropion, eclabium and alopecia with residual fine scales.\n\n\nDiscussion\n\nARCI is a rare disorder with an estimated prevalence of 1 per 200,000 population in Europe and 1 per 200,000–300,000 population in the United States. Neonates with LI typically present with a collodion membrane which dries and peels away and is replaced by brown, plate-like scales over the entire body. Disease course ranges from very mild to severe, latter entailing ectropion, eclabium, scarring supraciliary and scalp alopecia, and palmoplantar hyperkeratosis11.\n\nDNA based molecular diagnosis is crucial in ichthyosis as it provides a firm basis for genetic counseling of affected individuals and families, and also permits prenatal diagnosis. In a cohort of 520 independent families with ARCI, mutations were identified by direct sequencing of the 6 ARCI genes identified to date in 78% of patients: 32% harbored mutations in TGM1, 16% in NIPAL4, 12% in ALOX12B, 8% in CYP4F22, 5% in ABCA12, and 5% in ALOXE3. Whole exome sequencing may fill in the diagnostic lacuna of at least 22% of the patients who failed in this study to exhibit mutations in any of the known ARCI genes, indicating the existence of additional loci, such as 2 loci on chromosome 12p11.2-q1312. The 22q12+ duplication is known to cause cat eye syndrome, which has a range of potential morbidities with the occurrence of characteristic triad of iris coloboma, aural tags and/or pits and anal atresia14, though none of these features were present in the father or the children.\n\nTo the best of our knowledge, this is the first reported concurrence of a potentially benign 22q12+ duplication and LI, both of which are extremely rare individually. The mother of the siblings is now pregnant and the present finding will be used to help screen the foetus prenatally.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the parent of the patients.\n\n\nData availability\n\nThe raw exome sequencing data are available at the NCBI Sequence Read Archive (http://www.ncbi.nlm.nih.gov/sra), accession numbers SRX1096915 (II:1) and SRX1096920 (II:2).", "appendix": "Author contributions\n\n\n\nAG, YKS and KD identified the patient, took the biopsies for histopathology and sent the blood for DNA analyses. They helped in writing the initial case report as well as editing and formatting the manuscript. SKV, RJ, VD and AV isolated the DNA, performed the quality checks, prepared the exome capture and sequencing library, performed the exome sequencing. SKV performed data quality checks on the reads, reference alignments, variant call and computational prioritisation of the variants, designed and performed the validation experiments. SS and VS conceptualised and oversaw the DNA isolation, quality checks, exome sequencing, exome sequence analysis and validation and contributed to writing the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nAuthors acknowledge funding from the Council of Scientific and Industrial Research (CSIR), India through Grant No. BSC0122 (CARDIOMED).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nAuthors acknowledge Dr. Vamsi Y Krishna for critical comments and members of VS, SS labs and the GUaRDIAN Consortium for their help and support.\n\n\nReferences\n\nOji V, Tadini G, Akiyama M, et al.: Revised nomenclature and classification of inherited ichthyoses: results of the First Ichthyosis Consensus Conference in Sorèze 2009. J Am Acad Dermatol. 2010; 63(4): 607–641. PubMed Abstract \| Publisher Full Text\n\nHerman ML, Farasat S, Steinbach PJ, et al.: Transglutaminase-1 gene mutations in autosomal recessive congenital ichthyosis: summary of mutations (including 23 novel) and modeling of TGase-1. Hum Mutat. 2009; 30(4): 537–547. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCao X, Lin Z, Yang H, et al.: New mutations in the transglutaminase 1 gene in three families with lamellar ichthyosis. Clin Exp Dermatol. 2009; 34(8): 904–909. PubMed Abstract \| Publisher Full Text\n\nAkiyama M, Sawamura D, Shimizu H: The clinical spectrum of nonbullous congenital ichthyosiform erythroderma and lamellar ichthyosis. Clin Exp Dermatol. 2003; 28(3): 235–240. PubMed Abstract \| Publisher Full Text\n\nTakeichi T, Liu L, Fong K, et al.: Whole-exome sequencing improves mutation detection in a diagnostic epidermolysis bullosa laboratory. Br J Dermatol. 2015; 172(1): 94–100. PubMed Abstract \| Publisher Full Text\n\nVellarikkal SK, Patowary A, Singh M, et al.: Exome sequencing reveals a novel mutation, p.L325H, in the KRT5 gene associated with autosomal dominant Epidermolysis Bullosa Simplex Koebner type in a large family from Western India. Human Genome Variation. 2014; 1: 14007. Publisher Full Text\n\nGupta A, Sharma YK, Vellarikkal SK, et al.: Whole-exome sequencing solves diagnostic dilemma in a rare case of sporadic acrokeratosis verruciformis. J Eur Acad Dermatol Venereol. 2015. PubMed Abstract \| Publisher Full Text\n\nMiller SA, Dykes DD, Polesky HF: A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 1988; 16(3): 1215. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDePristo MA, Banks E, Poplin R, et al.: A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet. 2011; 43(5): 491–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang K, Li M, Hakonarson H: ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010; 38(16): e164. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaiho E, Ignatius J, Mikkola H, et al.: Transglutaminase 1 mutations in autosomal recessive congenital ichthyosis: private and recurrent mutations in an isolated population. Am J Hum Genet. 1997; 61(3): 529–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOji V, Tadini G, Akiyama M, et al.: Revised nomenclature and classification of inherited ichthyoses: results of the First Ichthyosis Consensus Conference in Sorèze 2009. J Am Acad Dermatol. 2010; 63(4): 607–41. PubMed Abstract \| Publisher Full Text\n\nFischer J: Autosomal recessive congenital ichthyosis. J Invest Dermatol. 2009; 129(6): 1319–21. PubMed Abstract \| Publisher Full Text\n\nRosias PR, Sijstermans JM, Theunissen PM, et al.: Phenotypic variability of the cat eye syndrome. Case report and review of the literature. Genet Couns. 2001; 12(3): 273–82. PubMed Abstract\n\nSim NL, Kumar P, Hu J, et al.: SIFT web server: predicting effects of amino acid substitutions on proteins. Nucleic Acids Res. 2012; 40(Web Server issue): W452–7. 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" }	[ { "id": "9750", "date": "10 Aug 2015", "name": "Mohamed Badawy Hassan Tawfik Abdel-Naser", "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 reported 2 siblings with lamellar ichthyosis (LI) and the family pedigree shows the autosomal pattern of inheritance. In addition, duplicate 22q12+ has been shown in the father and the 2 affected siblings. It seems, however, that this chromosomal abnormality is unrelated to the LI as the father is phenotypically normal and the 2 siblings do not show any of the manifestations of this chromosomal abnormality. The significance of this concurrence is not clear. The authors may amend the manuscript by naming the site of the TGM1 mutation at the nucleotide level and the name of the gene transcript. It is not clear whether the authors examined the father and the affected siblings for manifestations of the duplicate 22q12+ (apart from the cat eye syndrome), such as learning difficulties, growth retardation, minor genital abnormalities of the boy etc. There are few typo errors. For instance, the \"transglutaminase gene\" is better given as \"transglutaminase gene1\" and \"Ichyn\" is \"Ichthyin\" etc. Ref.13 is not cited in the text and there seems to be a problem with enumeration of the references in the text. ", "responses": [] }, { "id": "10062", "date": "24 Aug 2015", "name": "Regina Fölster-Holst", "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\nGupta et al. focused on two siblings with ARCI with confirmed TGM1 mutation. In addition they found 22q12 duplication in these siblings and in their father. The title and abstract are both appropriate for the article, and the abstract is a suitable summary. The diagnosis has been sufficiently described, however it is a little bit unclear whether the 22q12 duplication is associated with the ARCI. The authors mention the cat eye syndrome which is related to the duplication of the chromosome 22q12. What about the relevance of this finding? It should be worked out in more detail.", "responses": [] } ]	1	https://f1000research.com/articles/4-446
https://f1000research.com/articles/4-438/v1	30 Jul 15	{ "type": "Opinion Article", "title": "Villain of Molecular Biology: Why are we not reproducible in research?", "authors": [ "Vikash Bhardwaj" ], "abstract": "Worldwide, there is an issue of irreproducibility in life science research. In the USA alone $28 billion per year spent on preclinical research is not reproducible. Within this opinion article, I provide a brief historical account of the discovery of the Watson-Crick DNA model and introduce another neglected model of DNA. This negligence may be one of the fundamental reasons behind irreproducibility in molecular biology research.", "keywords": [ "Reproducibility", "Irreproducibility", "Molecular biology", "DNA", "Research" ], "content": "Introduction\n\nEvery year billions of dollars are invested in research worldwide to find solutions for deadly diseases like cancer, AIDS, TB etc. Much research is now focused on DNA and everyone is trying to understand what is happening within DNA at the molecular level. Whenever I ask my students within a molecular biology class, “Who discovered DNA?” many of my students give a very quick response “Watson and Crick”. It reminds me of a statement made by L. Pray that “Many people believe that American biologist James Watson and English physicist Francis Crick discovered DNA in the 1950s”1. Then a few students respond, “No sir” Watson and Crick uncovered the structure of DNA. But still it remains a question to my students: who discovered DNA? When there is no answer, I start explaining history of DNA discovery. I remind them about Friedrich Miescher who isolated DNA for the first time in 1869 and how Watson and Crick deciphered the structure of DNA in 1953. It took almost 85 years to discover the molecular details of DNA structure. Friedrich Miescher was a brilliant scientist but he suffered when his findings were not published immediately and his boss published his results after repeating his experiments on his own. It took almost two years to get his results published. Even Miescher had the opinion that the new substance (“nuclein”) could be thought of as genetic material but various established theories stopped his journey. Many scientists at that time were of the opinion that some proteins could be genetic material and they were busy finding novel proteins which could act as genetic material. Friedrich Miescher died in 1895 without getting credit for his discovery2. Later in 1919, the “Tetra nucleotide hypothesis” proposed by Levene become an obstacle in DNA structure discovery as he proposed DNA as an inert molecule having four nucleotides repeatedly arranged. Levene was also a brilliant scientist and had published more than 700 papers and many scientists followed his opinions3,4. In the coming years scientists were not interested in doing work on DNA until 1928 when Griffith gave some evidence that it is not the proteins which acts as a genetic material5. His experimental findings were worked out further in 1944 by Avery et al. and they clearly demonstrated that it is DNA which acts as a transforming agent but still they faced opposition. Still, many scientists were slow to accept this clear proof that DNA, not protein, is the genetic molecule6,7. From 1950 onwards Chargaff used to meet and discuss with scientists that he had different results showing that DNA cannot be an inert molecule thus clearly rejecting the “Tetra nucleotide hypothesis”. But his views were not given much importance (http://www.dnai.org)8. In 1952, Hershey and Chase did a classical experiment which proved without doubt that DNA is genetic material9. It created interest in others to solve out molecular details of DNA. In 1953 Linus Pauling (one of the famous Nobel laureates) proposed a triple helical model of DNA in which they proposed phosphate groups are inside while nitrogenous bases are outside10. Really Pauling’s authority in science might have become another obstacle in the discovery of the correct DNA model, if Watson and Crick had not realized that phosphate groups cannot be inside, as this would destabilize DNA due to high negative charge. Watson, Crick and Wilkinson proposed the double helical structure of DNA based on work done by Rosalin Franklin11–13. Linus Pauling himself visited Watson and Crick and was convinced about their proposed model (https://paulingblog.wordpress.com/2009/04/30/the-watson-and-crick-structure-of-dna/). Watson, Crick and Wilkinson received the Nobel Prize in 1963 for solving the structure of DNA. Yes, truly their findings have changed molecular biology research worldwide. Many scientists started developing new molecular techniques and the fundamentals of biology are based on it. In 1973 E.M. Southern developed the Southern hybridization technique to detect DNA14 and later in 1977, northern hybridization was developed by Alwine et al. to detect levels of RNA15. In 1983, Dr Kary Mulis developed PCR (Polymerase Chain Reaction) for DNA amplification, which is also based on antiparallel complementary hybridization of DNA and for this he was awarded Nobel Prize in 199316. In 1973, Boyer, Cohen and Chang developed cloning techniques which has allowed the production of recombinant protein and a whole new science of recombinant DNA technology has been developed based on this17. On the basis of Watson and Crick model of DNA and using Sanger DNA sequencing chemistry, scientists throughout the world invested billions of dollars and developed the Human Genome Project18,19. Today efforts are being taken to sequence the genome of each organism. On the basis of homologous DNA sequence, gene knockout technology was developed and many scientists have tried to characterize functions of various proteins and DNA elements on the basis of gene knockouts20–22. Capecchi, Evans and Smithies, were awarded the Nobel Prize for developing gene knockout technology in 2007. In 1998, siRNA technology was developed by Andrew Z. Fire and Craig C. Mello for which they were awarded Nobel Prize in 200623 and Microarray technology was developed by Pat Brown which provides powerful tools for global characterization of gene expression24. Yes, truly whole molecular biology has flourished with much new information and technologies in last 60 years based on the Watson and Crick model of DNA and it will not be possible for me to write about all techniques.\n\n\nVillain of molecular biology\n\nThere are many publications which have reported huge errors of various molecular techniques (for more details see 25–36). You may have experienced, non-specific amplification of DNA in a PCR reaction, non-specific hybridization in Southern and northern hybridization and then have tried harder to find out conditions which give you better results. You may have experienced, non-specific cloning reactions and then must have tried to screen out a specific clone out of non-specific ones. Even today we do not have answer as to why petunia flowers turn white on overexpression of a gene which should have made it more purple37. None of our gene knockout technology explains whether they have taken out one and only one gene and the remainder of the genes have not been affected. To the best of my knowledge, there is no whole genome sequence information available for knock out organisms. I wish to inform readers that the Human Genome Project is not yet complete even though its first draft was announced 15 years ago38. All over the world, billions of dollars are still invested in a hope to find solutions for various diseases. How can we find solutions if the molecular techniques used show errors and many times we are unable to reproduce the same findings in different labs. In 2012 the Biotechnology Company Amgen with a team of 100 scientists found that only 10% (6 out of 53) of research published by reputable labs in top journals is reproducible and 90% of money ($28 billion) is wasted. It looks like that even after development of high throughput techniques and instruments, research worldwide is losing accuracy and precision. It is a worse situation for biotechnology/pharmaceutical industries who are going to invest or have invested millions of dollars for their new drug development programme. It’s again a far worse situation for the public who are looking forward to scientists one day finding solutions to deadly diseases and producing cheaper drugs and the best treatments soon39,40. Recently Professor Eric Lander (one of the leaders of the Human Genome Project, and a member of US President Barack Obama’s scientific advisory panel) visited India and gave an exclusive interview stating that we will have a solution for most cancers in the next 25–30 years (http://www.ndtv.com/video/player/ndtv-special-ndtv-24x7/mapping-the-human-genome-the-eric-lander-interview/358410). I totally disagree with his statements as with the current ways of doing research, it may take many thousand years to find ultimate solutions for mankind’s problems. A recent report by John Arrowsmith revealed that the Phase II success rate for new development projects has decreased by 10% in the last few years. It will definitely increase the cost of new drugs in the future. It will also decrease the trust of the public Government and funding agencies in scientific activities41. But still a question arises, who is the ‘villain’ behind these problems? Yes, the B- form of DNA is a ‘hero’ of molecular biology but there is also a ‘villain’ of molecular biology. It’s a form of DNA which is actually much less studied, discussed and used in designing molecular techniques. It is “parallel stranded duplex DNA” which was first reported by Ramsing & Jovin and Sande et al. in 1988. There are few reports in favour of parallel stranded DNA which summarize that there is no drastic difference in parallel and antiparallel DNA having mixed AT/GC composition42–47. Recently we have developed a PD-PCR technology based on parallel stranded DNA and we have concluded that two PCR products can be synthesized from a single stranded template DNA, one by conventional PCR and another by our approach48. In 2008, Lestienne et al. reported a novel property of TFO (Triplex forming oligonucleotides-known for transcription inhibition) that Triple helix primer (THP) bounded to the duplex DNA in a parallel orientation can initiate DNA synthesis by various DNA polymerases of phage, retrovirus, bacteria and humans49,50. There are reports which state that Southern hybridization reaction can be performed using parallel complementary probe and gene silencing can be applied using parallel complementary RNA51–53. It also makes me think whether earlier scientists have developed a 100% accurate genome sequence of Human which has only been developed on the basis of antiparallel complementarity in DNA. I strongly believe technical errors observed in various molecular techniques can be ruled out by considering both parallel and antiparallel complementarity of DNA. A probe for Southern blotting/northern blotting can be designed such that it binds to its target only in an antiparallel manner. Primers for PCR can be designed in a similar way. There is a need to develop siRNA and microarray chips keeping in mind parallel and antiparallel hybridization of DNA. Science without errors will increase reproducibility in research worldwide (Figure 1).\n\nUsed with permission from Macmillan Publishers Ltd. Nature ©201354.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nReferences\n\nPray LA: Discovery of DNA structure and function: Watson and Crick. Nature Education. 2008; 1(1): 100. Reference Source\n\nDahm R: Discovering DNA: Friedrich Miescher and the early years of nucleic acid research. Hum Genet. 2008; 122(6): 565–581. PubMed Abstract \| Publisher Full Text\n\nLevene PA: The structure of yeast nucleic acid. Studies from the Rockefeller Institute for Medical Research. 1919; 30: 221. Reference Source\n\nHargittai I: The tetranucleotide hypothesis: a centennial. Struct Chem. 2009; 20(5): 753–756. Publisher Full Text\n\nGriffith F: The Significance of Pneumococcal Types. J Hyg (Lond). 1928; 27(02): 113–159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAvery OT, MacLeod CM, McCarty M: Studies on the chemical nature of the substance inducing transformation of pneumococcal types : induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus type III. J Exp Med. 1944; 79(2): 137–158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeichmann U: Early responses to Avery et al.’s paper on DNA as hereditary material. Hist Stud Phys Biol Sci. 2004; 34(2): 207–232. Publisher Full Text\n\nDahm R: From discovering to understanding. Friedrich Miescher’s attempts to uncover the function of DNA. EMBO Rep. 2010; 11(3): 153–160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHershey AD, Chase M: Independent functions of viral protein and nucleic acid in growth of bacteriophage. J Gen Physiol. 1952; 36(1): 39–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPauling L, Corey RB: A Proposed Structure For The Nucleic Acids. Proc Natl Acad Sci U S A. 1953; 39(2): 84–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWatson JD, Crick FH: Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature. 1953; 171(4356): 737–738. PubMed Abstract \| Publisher Full Text\n\nWilkins MHF, Stokes AR, Wilson HR: Molecular structure of deoxypentose nucleic acids. Nature. 1953; 171(4356): 738–740. PubMed Abstract \| Publisher Full Text\n\nFranklin RE, Gosling RG: Molecular configuration in sodium thymonucleate. Nature. 1953; 171(4356): 740–741. PubMed Abstract \| Publisher Full Text\n\nSouthern EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol. 1975; 98(3): 503–517. PubMed Abstract \| Publisher Full Text\n\nAlwine JC, Kemp DJ, Stark GR: Method for detection of specific RNAs in agarose gels by transfer to diazobenzyloxymethyl-paper and hybridization with DNA probes. Proc Natl Acad Sci U S A. 1977; 74(12): 5350–5354. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMullis KB: Process for amplifying nucleic acid sequences. U.S. Patent No. 4,683,202 1987. Reference Source\n\nCohen SN, Chang ACY, Boyer HW, et al.: Construction of biologically functional bacterial plasmids in vitro. Proc Natl Acad Sci U S A. 1973; 70(11): 3240–3244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLander ES, Linton LM, Birren B, et al.: Initial sequencing and analysis of the human genome. Nature. 2001; 409(6822): 860–921. PubMed Abstract \| Publisher Full Text\n\nSchuler GD, Boguski MS, Stewart EA, et al.: A gene map of the human genome. Science. 1996; 274(5287): 540–546. PubMed Abstract \| Publisher Full Text\n\nThomas KR, Capecchi MR: Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell. 1987; 51(3): 503–512. PubMed Abstract \| Publisher Full Text\n\nEvans MJ, Kaufman MH: Establishment in culture of pluripotential cells from mouse embryos. Nature. 1981; 292(5819): 154–156. PubMed Abstract \| Publisher Full Text\n\nDoetschman T, Gregg RG, Maeda N, et al.: Targetted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature. 1987; 330(6148): 576–8. PubMed Abstract \| Publisher Full Text\n\nFire A, Xu S, Montgomery MK, et al.: Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature. 1998; 391(6669): 806–811. PubMed Abstract \| Publisher Full Text\n\nBrown P: You say you want a revolution: an interview with Pat Brown. Interview by Jane Gitschier. PLoS Genet. 2009; 5(7): e1000560. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKothapalli R, Yoder SJ, Mane S, et al.: Microarray results: how accurate are they? BMC Bioinformatics. 2002; 3(1): 22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDraghici S, Khatri P, Eklund AC, et al.: Reliability and reproducibility issues in DNA microarray measurements. Trends Genet. 2006; 22(2): 101–109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIrizarry RA, Warren D, Spencer F, et al.: Multiple-laboratory comparison of microarray platforms. Nat Methods. 2005; 2(5): 345–350. PubMed Abstract \| Publisher Full Text\n\nXu J, Deng X, Chan V, et al.: Variability of DNA microarray gene expression profiles in cultured rat primary hepatocytes. Gene Regul Syst Bio. 2007; 1: 235–249. PubMed Abstract \| Free Full Text\n\nKoltai H, Weingarten-Baror C: Specificity of DNA microarray hybridization: characterization, effectors and approaches for data correction. Nucleic Acids Res. 2008; 36(7): 2395–2405. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCastanotto D, Rossi JJ: The promises and pitfalls of RNA-interference-based therapeutics. Nature. 2009; 457(7228): 426–433. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJackson AL, Linsley PS: Noise amidst the silence: off-target effects of siRNAs? Trends Genet. 2004; 20(11): 521–524. PubMed Abstract \| Publisher Full Text\n\nJackson AL, Bartz SR, Schelter J, et al.: Expression profiling reveals off-target gene regulation by RNAi. Nat Biotechnol. 2003; 21(6): 635–637. PubMed Abstract \| Publisher Full Text\n\nScacheri PC, Rozenblatt-Rosen O, Caplen NJ, et al.: Short interfering RNAs can induce unexpected and divergent changes in the levels of untargeted proteins in mammalian cells. Proc Natl Acad Sci U S A. 2004; 101(7): 1892–1897. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTschuch C, Schulz A, Pscherer A, et al.: Off-target effects of siRNA specific for GFP. BMC Mol Biol. 2008; 9(1): 60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJackson AL, Burchard J, Schelter J, et al.: Widespread siRNA “off-target” transcript silencing mediated by seed region sequence complementarity. RNA. 2006; 12(7): 1179–1187. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGumienny R, Zavolan M: Accurate transcriptome-wide prediction of microRNA targets and small interfering RNA off-targets with MIRZA-G. Nucleic Acids Res. 2015; 43(3): 1380–1391. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNapoli C, Lemieux C, Jorgensen R: Introduction of a Chimeric Chalcone Synthase Gene into Petunia Results in Reversible Co-Suppression of Homologous Genes in trans. Plant Cell. 1990; 2(4): 279–289. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlow N: Decoding the unsequenceable. Biotechniques. 2015; 58(2): 52–58. PubMed Abstract\n\nBegley CG, Ellis LM: Drug development: Raise standards for preclinical cancer research. Nature. 2012; 483(7391): 531–533. PubMed Abstract \| Publisher Full Text\n\nFreedman LP, Cockburn IM, Simcoe TS: The Economics of Reproducibility in Preclinical Research. PLoS Biol. 2015; 13(6): e1002165. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArrowsmith J: Trial watch: Phase II failures: 2008–2010. Nat Rev Drug Discov. 2011; 10(5): 328–329. PubMed Abstract \| Publisher Full Text\n\nRamsing NB, Jovin TM: Parallel stranded duplex DNA. Nucleic Acids Res. 1988; 16(14A): 6659–6676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan de Sande JH, Ramsing NB, Germann MW, et al.: Parallel stranded DNA. Science. 1988; 241(4865): 551–557. PubMed Abstract \| Publisher Full Text\n\nShchyolkina AK, Borisova OF, Livshits MA, et al.: Parallel-stranded DNA with mixed AT/GC composition: role of trans G.C base pairs in sequence dependent helical stability. Biochemistry. 2000; 39(33): 10034–44. PubMed Abstract \| Publisher Full Text\n\nShchyolkina AK, Borisova OF, Livshits MA, et al.: Parallel-stranded DNA with natural base sequences. Mol Biol (Mosk). 2003; 37(2): 255–265. PubMed Abstract \| Publisher Full Text\n\nBorisova OF, Shchyolkina AK, Chernov BK, et al.: Relative stability of AT and GC pairs in parallel DNA duplex formed by a natural sequence. FEBS Lett. 1993; 322(3): 304–306. PubMed Abstract \| Publisher Full Text\n\nLi H, Peng X, Leonard P, et al.: Binding of actinomycin C1 (D) and actinomin to base-modified oligonucleotide duplexes with parallel chain orientation. Bioorg Med Chem. 2006; 14(12): 4089–4100. PubMed Abstract \| Publisher Full Text\n\nBhardwaj V, Sharma K: Parallel DNA polymerase chain reaction: Synthesis of two different PCR products from a DNA template [v1; ref status: indexed, http://f1000r.es/4sm]. F1000Res. 2014; 3: 320. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLestienne PP, Boudsocq F, Bonnet JE: Initiation of DNA Replication by a Third Parallel DNA Strand Bound in a Triple-Helix Manner Leads to Strand Invasion. Biochemistry. 2008; 47(21): 5689–5698. PubMed Abstract \| Publisher Full Text\n\nLestienne PP: Priming DNA replication from triple helix oligonucleotides: Possible threestranded DNA in DNA polymerases. Mol Biol Int. 2011; 2011: 562849. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTchurikov NA, Shchyolkina AK, Borissova OF, et al.: Southern molecular hybridization experiments with parallel complementary DNA probes. FEBS Lett. 1992; 297(3): 233–6. PubMed Abstract \| Publisher Full Text\n\nTchurikov NA, Chistyakova LG, Zavilgelsky GB, et al.: Gene-specific silencing by expression of parallel complementary RNA in Escherichia coli. J Biol Chem. 2000; 275(34): 26523–9. PubMed Abstract \| Publisher Full Text\n\nLiu YS, Zhang YQ, Yang LX, et al.: Gene-specific silencing induced by parallel complementary RNA in Pseudomonas aeruginosa. Biotechnol Lett. 2009; 31(10): 1571–1575. PubMed Abstract \| Publisher Full Text\n\nBissell M: Reproducibility: The risks of the replication drive. Nature. 2013; 503(7476): 333–334. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9734", "date": "03 Aug 2015", "name": "Sukrat Sinha", "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 article written by Dr. Vikas Bharadwaj throws light on the issue of non reproducibility of research citing the example of DNA structure elucidated by Watson and Crick. Every year billions of dollars are being spent on this part of research. The article envisages the very valid point that there is a dire need to revise the concepts so that our research could be more fruitful and result oriented. I highly appreciate Dr. Bhardwaj for bringing this issue on forefront.", "responses": [] }, { "id": "9738", "date": "10 Aug 2015", "name": "Nirajkumar Singh", "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 article presented here “Villain of Molecular Biology: Why are we not reproducible in research?” deals mainly with the issue of non-reproducibility of research citing many examples including the example of DNA structure elucidated by Watson and Crick.I feel this is very serious issue throughout the world. The author claims only about 10% research paper work is reproducible. However, keeping the recent development of genetic engineering and its product (used in day to day life e.g., Insulin, BT Cotton, and many more..) in mind, I think this figure should be more than that mentioned by authors, even though the authors are very much correct about the reproducibility of scientific work.Overall the manuscript is written in a very systematic way which fully justifies the title of the manuscript. In my opinion the manuscript is adequate for indexing.", "responses": [] } ]	1	https://f1000research.com/articles/4-438
https://f1000research.com/articles/2-278/v1	17 Dec 13	{ "type": "Correspondence", "title": "Identical twins and Bayes' theorem in the 21st century", "authors": [ "Valentin Amrhein", "Tobias Roth", "Fränzi Korner-Nievergelt", "Tobias Roth", "Fränzi Korner-Nievergelt" ], "abstract": "In a recent article in Science on \"Bayes' Theorem in the 21st Century\", Bradley Efron uses Bayes' theorem to calculate the probability that twins are identical given that the sonogram shows twin boys. He concludes that Bayesian calculations cannot be uncritically accepted when using uninformative priors. We argue that this conclusion is problematic because Efron's example on identical twins does not use data, hence it is not Bayesian statistics; his priors are not appropriate and are not uninformative; and using the available data point and an uninformative prior actually leads to a reasonable posterior distribution.", "keywords": [ "Efron1 provides four examples of Bayesian analyses", "two of which underline the remarkable potential of Bayesian methods. Based on one of the other examples", "however", "Efron ultimately concludes that Bayesian analyses using uninformative priors cannot be uncritically accepted and should be checked by frequentist methods. While we wholeheartedly agree that statistical results should not be uncritically accepted", "we find Efron’s example ineffective in showing that Bayesian statistics require more careful checking than any other kind of statistics." ], "content": "Correspondence\n\nEfron1 provides four examples of Bayesian analyses, two of which underline the remarkable potential of Bayesian methods. Based on one of the other examples, however, Efron ultimately concludes that Bayesian analyses using uninformative priors cannot be uncritically accepted and should be checked by frequentist methods. While we wholeheartedly agree that statistical results should not be uncritically accepted, we find Efron’s example ineffective in showing that Bayesian statistics require more careful checking than any other kind of statistics.\n\nIn his example on uninformative priors, Efron uses Bayes’ theorem to calculate the probability that twins are identical given that the sonogram shows twin boys. Efron finds this probability to be 2/3 when using an uninformative prior versus 1/2 with an informative prior and thereby concludes that an uninformative prior does not have the desired neutral effects on the output of Bayes’ rule. We argue that this example is not only flawed, but useless in illustrating Bayesian data analysis because it does not rely on any data. Although there is one data point (a couple is due to be parents of twin boys, and the twins are fraternal), Efron does not use it to update prior knowledge. Instead, Efron combines different pieces of expert knowledge from the doctor and genetics using Bayes’ theorem. While certainly an impeccable probability law, Bayes’ theorem is a mathematical equation, not a statistical model describing how data may be produced. In essence, Efron uses this equation to show that the value on the left side of the equation changes when a term on the right side is changed, which is trivial and could be shown with any mathematical equation also in a non-Bayesian context. Without new data, our knowledge is by definition determined by prior information; thus, showing that the outcome of a Bayesian analysis with no new data is heavily influenced by the prior would not argue against Bayesian methods. Indeed, without data, Efron’s example is not Bayesian statistics and his conclusion about Bayesian statistics based on this example is unjustified.\n\nWe also have other more technical issues with Efron’s example. Efron interprets the term P(A) on the right side of the equation (see sidebar in Efron 2013a1) as the prior on the probability that twins are identical. To make this prior uninformative, it is assigned a value of P(A) = 0.5 (see Efron 2013b2; although this is not stated in Efron 2013a1). This uninformative prior is set in contrast to the informative “doctor’s prior” of P(A) = 1/3. First, however, the parameter of interest is P(A\|B) rather than P(A) according to Efron’s study question (see sidebar in Efron 2013a1), thus the focus should be on the appropriate prior for P(A\|B). Second, for the uninformative prior, Efron mentions erroneously that he used a uniform distribution between zero and one, which is clearly different from the value of 0.5 that was used. Third, we find it at least debatable whether a prior can be called an uninformative prior if it has a fixed value of 0.5 given without any measurement of uncertainty. For example, if we knew that our chance of winning the next million-dollar jackpot were 50:50, would we really call this uninformative?\n\nIf we use the data point together with an uninformative uniform prior on P(A\|B) to determine the probability of identical twins given the twins are two boys (see Box 1), we obtain, with 95% certainty, a probability of between 0.01 and 0.84; if we use a highly informative prior based on information from the doctor and genetics, we obtain a probability of between 0.49 and 0.51. This looks completely reasonable to us, although of course we do not know much more than we knew before because we had only a single data point.\n\nWe would very much like to check our calculations using frequentist methods; however, this is impossible because there is only one data point, and frequentist methods generally cannot handle such situations. Although we agree with Efron1 that the choice of the prior is essential, we conclude that his article gives a biased impression of the influence of uninformative priors. In his example using Bayes’ theorem, we found no reliable support for his main conclusion that Bayesian calculations cannot be uncritically accepted when using uninformative priors.\n\nData: One pair of twin boys is fraternal.\n\nData model: x~Binomial(θ, n), where θ is the probability of identical twins given the twins are two boys, x is the number of identical twins in the data, and n is the total number of pairs of twin boys; in our case: x = 0 and n = 1.\n\nThe posterior distribution p(θ\|x) is obtained using Bayes’ theorem\n\np(θ\|x) = p(x\|θ)p(θ)/p(x)\n\nWe use two different priors p(θ):\n\n1) Uninformative prior: p(θ) = Unif(0,1) = Beta(1,1).\n\n2) Informative prior: using the information from the doctor and from genetics, we are quite sure that θ must be around 0.51. Transforming this information into a statistical distribution yields p(θ) = Beta(10000, 10000), which has a mean of 0.5 and a 95% interval of 0.49307 – 0.50693. [Note that we had to choose the 95% interval arbitrarily because we are not informed about the certainty of the information provided by the doctor and by genetics].\n\nGiven the single parameter Binomial model, x~Binomial(θ, n), and the prior p(θ) = Beta(α,β), the solution of the Bayesian analysis is given by the posterior distribution p(θ\|x) = Beta(α+x,β+n-x) [see any Bayesian textbook, e.g. Gelman et al. 20043, p. 34].\n\nThe probability of identical twins given the twins are two boys:\n\n1) Uninformative prior: p(θ\|x) = Beta(1+x,1+n-x) = Beta(1+0,1+1-0) = Beta(1, 2), which has an expected value of 0.33 and a 95% interval of 0.013 – 0.84.\n\n2) Informative prior: p(θ\|x) = Beta(10000+x,10000+n-x) = Beta(10000+0,10000+1-0) = Beta(10000, 10001), which has an expected value of 0.49998 and a 95% interval of 0.49305 – 0.50690.", "appendix": "Author contributions\n\n\n\nFK-N reanalyzed the data point. VA wrote the first draft of the manuscript. All authors contributed to the discussion and approved the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\n\n\n\nAcknowledgements\n\nThe authors would like to thank Yves-Laurent Grize and Pius Korner for discussions.\n\n\nReferences\n\nEfron B: Bayes’ Theorem in the 21st Century. Science. 2013; 340(6137): 1177–1178. PubMed Abstract \| Publisher Full Text\n\nEfron B: A 250-year argument: Belief, behavior, and the bootstrap. Bull Amer Math Soc. 2013; 50: 129–146. Publisher Full Text\n\nGelman A, Carlin JB, Stern HS, et al.: Bayesian Data Analysis. Chapman & Hall, New York. 2004. Publisher Full Text" }	[ { "id": "2816", "date": "24 Dec 2013", "name": "Michael McCarthy", "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\nThis paper by Amrhein et al. criticizes a paper by Bradley Efron that discusses Bayesian statistics (Efron, 2013a), focusing on a particular example that was also discussed in Efron (2013b). The example concerns a woman who is carrying twins, both male (as determined by sonogram and we ignore the possibility that gender has been observed incorrectly). The parents-to-be ask Efron to tell them the probability that the twins are identical. This is my first open review, so I'm not sure of the protocol. But given that there appears to be errors in both Efron (2013b) and the paper under review, I am sorry to say that my review might actually be longer than the article by Efron (2013a), the primary focus of the critique, and the critique itself. I apologize in advance for this. To start, I will outline the problem being discussed for the sake of readers.This problem has various parameters of interest. The primary parameter is the genetic composition of the twins in the mother’s womb. Are they identical (which I describe as the state x = 1) or fraternal twins (x = 0)? Let y be the data, with y = 1 to indicate the twins are the same gender. Finally, we wish to obtain Pr(x = 1 \| y = 1), the probability the twins are identical given they are the same gender1. Bayes' rule gives us an expression for this:Pr(x = 1 \| y = 1) = Pr(x=1) Pr(y = 1 \| x = 1) / {Pr(x=1) Pr(y = 1 \| x = 1) + Pr(x=0) Pr(y = 1 \| x = 0)} Now we know that Pr(y = 1 \| x = 1) = 1; twins must be the same gender if they are identical. Further, Pr(y = 1 \| x = 0) = 1/2; if twins are not identical, the probability of them being the same gender is 1/2.Finally, Pr(x = 1) is the prior probability that the twins are identical. The bone of contention in the Efron papers and the critique by Amrhein et al. revolves around how this prior is treated. One can think of Pr(x = 1) as the population-level proportion of twins that are identical for a mother like the one being considered.However, if we ignore other forms of twins that are extremely rare (equivalent to ignoring coins finishing on their edges when flipping them), one incontrovertible fact is that Pr(x = 0) = 1 − Pr(x = 1); the probability that the twins are fraternal is the complement of the probability that they are identical.The above values and expressions for Pr(y = 1 \| x = 1), Pr(y = 1 \| x = 0), and Pr(x = 0) leads to a simpler expression for the probability that we seek - the probability that the twins are identical given they have the same gender:Pr(x = 1 \| y = 1) = 2 Pr(x=1) / [1 + Pr(x=1)]\n\n(1)We see that the answer depends on the prior probability that the twins are identical, Pr(x=1). The paper by Amrhein et al. points out that this is a mathematical fact. For example, if identical twins were impossible (Pr(x = 1) = 0), then Pr(x = 1\| y = 1) = 0. Similarly, if all twins were identical (Pr(x = 1) = 1), then Pr(x = 1\| y = 1) = 1. The “true” prior lies somewhere in between. Apparently, the doctor knows that one third of twins are identical2. Therefore, if we assume Pr(x = 1) = 1/3, then Pr(x = 1\| y = 1) = 1/2.Now, what would happen if we didn't have the doctor's knowledge? Laplace's “Principle of Insufficient Reason” would suggest that we give equal prior probability to all possibilities, so Pr(x = 1) = 1/2 and Pr(x = 1\| y = 1) = 2/3, an answer different from 1/2 that was obtained when using the doctor's prior of 1/3.Efron (2013a) highlights this sensitivity to the prior, representing someone who defines an uninformative prior as a “violator”, with Laplace as the “prime violator”. In contrast, Amrhein et al. correctly points out that the difference in the posterior probabilities is merely a consequence of mathematical logic. No one is violating logic – they are merely expressing ignorance by specifying equal probabilities to all states of nature. Whether this is philosophically valid is debatable (Colyvan 2008), but this example does not lend much weight to that question, and it is well beyond the scope of this review. But setting Pr(x = 1) = 1/2 is not a violation; it is merely an assumption with consequences (and one that in hindsight might be incorrect2).Alternatively, if we don't know Pr(x = 1), we could describe that probability by its own probability distribution. Now the problem has two aspects that are uncertain. We don't know the true state x, and we don't know the prior (except in the case where we use the doctor's knowledge that Pr(x = 1) = 1/3). Uncertainty in the state of x refers to uncertainty about this particular set of twins. In contrast, uncertainty in Pr(x = 1) reflects uncertainty in the population-level frequency of identical twins. A key point is that the state of one particular set of twins is a different parameter from the frequency of occurrence of identical twins in the population.Without knowledge about Pr(x = 1), we might use Pr(x = 1) ~ dunif(0, 1), which is consistent with Laplace. Alternatively, Efron (2013b) notes another alternative for an uninformative prior: Pr(x = 1) ~ dbeta(0.5, 0.5), which is the Jeffreys prior for a probability.Here I disagree with Amrhein et al.; I think they are confusing the two uncertain parameters. Amrhein et al. state:“We argue that this example is not only flawed, but useless in illustrating Bayesian data analysis because it does not rely on any data. Although there is one data point (a couple is due to be parents of twin boys, and the twins are fraternal), Efron does not use it to update prior knowledge. Instead, Efron combines different pieces of expert knowledge from the doctor and genetics using Bayes’ theorem.”This claim might be correct when describing uncertainty in the population-level frequency of identical twins. The data about the twin boys is not useful by itself for this purpose – they are a biased sample (the data have come to light because their gender is the same; they are not a random sample of twins). Further, a sample of size one, especially if biased, is not a firm basis for inference about a population parameter. While the data are biased, the claim by Amrheim et al. that there are no data is incorrect.However, the data point (the twins have the same gender) is entirely relevant to the question about the state of this particular set of twins. And it does update the prior. This updating of the prior is given by equation (1) above. The doctor's prior probability that the twins are identical (1/3) becomes the posterior probability (1/2) when using information that the twins are the same gender. The prior is clearly updated with Pr(x = 1\| y = 1) ≠ Pr(x = 1) in all but trivial cases; Amrheim et al.'s statement that I quoted above is incorrect in this regard.This possible confusion between uncertainty about these twins and uncertainty about the population level frequency of identical twins is further suggested by Amrhein et al.'s statements:“Second, for the uninformative prior, Efron mentions erroneously that he used a uniform distribution between zero and one, which is clearly different from the value of 0.5 that was used. Third, we find it at least debatable whether a prior can be called an uninformative prior if it has a fixed value of 0.5 given without any measurement of uncertainty.”Note, if the prior for Pr(x = 1) is specified as 0.5, or dunif(0,1), or dbeta(0.5, 0.5), the posterior probability that these twins are identical is 2/3 in all cases. Efron (2013b) says the different priors lead to different results, but this result is incorrect, and the correct answer (2/3) is given in Efron (2013a)3. Nevertheless, a prior that specifies Pr(x = 1) = 0.5 does indicate uncertainty about whether this particular set of twins is identical (but certainty in the population level frequency of twins). And Efron’s (2013a) result is consistent with Pr(x = 1) having a uniform prior. Therefore, both claims in the quote above are incorrect.It is probably easiest to show the (lack of) influence of the prior using MCMC sampling. Here is WinBUGS code for the case using Pr(x = 1) = 0.5.model{ pr_ident_twins <- 0.5\n\n# prior probability that the twins are identical x ~ dbern(pr_ident_twins)\n\n# are they identical? If so, x = 1, and 0 otherwise\n\npr_same_gender <- x + (1-x)*0.5\n\n# the probability that the twins have the same gender. It equals 1 if x = 1, and 0.5 otherwise (i.e., if x = 0)\n\nsame_gender <- 1 # the single data point - the twins are the same gender same_gender ~ dbern(pr_same_gender)\n\n# those data arise as a Bernoulli sample with probability pr_same_gender} Running this model in WinBUGS shows that the posterior mean of x is 2/3; this is the posterior probability that x = 1.Instead of using pr_ident_twins <- 0.5, we could set this probability as being uncertain and define pr_ident_twins ~ dunif(0,1), or pr_ident_twins ~ dbeta(0.5,0.5). In either case, the posterior mean value of x remains 2/3 (contrary to Efron 2013b, but in accord with the correction in Efron 2013a).Note, however, that the value of the population level parameter pr_ident_twins is different in all three cases. In the first it remains unchanged at 1/2 where it was set. In the case where the prior distribution for pr_ident_twins is uniform or beta, the posterior distributions remain broad, but they differ depending on the prior (as they should – different priors lead to different posteriors4). However, given the biased sample size of 1, the posterior distribution for this particular parameter is likely to be misleading as an estimate of the population-level frequency of twins.So why doesn’t the choice of prior influence the posterior probability that these twins are identical? Well, for these three priors, the prior probability that any single set of twins is identical is 1/2 (this is essentially the mean of the prior distributions in these three cases).If, instead, we set the prior as dbeta(1,2), which has a mean of 1/3, then the posterior probability that these twins are identical is 1/2. This is the same result as if we had set Pr(x = 1) = 1/3. In both these cases (choosing dbeta(1,2) or 1/3), the prior probability that a single set of twins is identical is 1/3, so the posterior is the same (1/2) given the data (the twins have the same gender).Further, Amrhein et al. also seem to misunderstand the data. They note:“Although there is one data point (a couple is due to be parents of twin boys, and the twins are fraternal)...”This is incorrect. The parents simply know that the twins are both male. Whether they are fraternal is unknown (fraternal twins being the complement of identical twins) – that is the question the parents are asking. This error of interpretation makes the calculations in Box 1 and subsequent comments irrelevant.Box 1 also implies Amrhein et al. are using the data to estimate the population frequency of identical twins rather than the state of this particular set of twins. This is different from the aim of Efron (2013a) and the stated question.Efron suggests that Bayesian calculations should be checked with frequentist methods when priors are uncertain. However, this is a good example where this cannot be done easily, and Amrhein et al. are correct to point this out. In this case, we are interested in the probability that the hypothesis is true given the data (an inverse probability), not the probabilities that the observed data would be generated given particular hypotheses (frequentist probabilities). If one wants the inverse probability (the probability the twins are identical given they are the same gender), then Bayesian methods (and therefore a prior) are required. A logical answer simply requires that the prior is constructed logically. Whether that answer is “correct” will be, in most cases, only known in hindsight.However, one possible way to analyse this example using frequentist methods would be to assess the likelihood of obtaining the data for each of the two hypothesis (the twins are identical or fraternal). The likelihood of the twins having the same gender under the hypothesis that they are identical is 1. The likelihood of the twins having the same gender under the hypothesis that they are fraternal is 0.5. Therefore, the weight of evidence in favour of identical twins is twice that of fraternal twins. Scaling these weights so they sum to one (Burnham and Anderson 2002), gives a weight of 2/3 for identical twins and 1/3 for fraternal twins. These scaled weights have the same numerical values as the posterior probabilities based on either a Laplace or Jeffreys prior. Thus, one might argue that the weight of evidence for each hypothesis when using frequentist methods is equivalent to the posterior probabilities derived from an uninformative prior. So, as a final aside in reference to Efron (2013a), if we are being “violators” when using a uniform prior, are we also being “violators” when using frequentist methods to weigh evidence? Regardless of the answer to this rhetorical question, “checking” the results with frequentist methods doesn't give any more insight than using uninformative priors (in this case). However, this analysis shows that the question can be analysed using frequentist methods; the single data point is not a problem for this. The claim in Armhein et al. that a frequentist analyis \"is impossible because there is only one data point, and frequentist methods generally cannot handle such situations\" is not supported by this example.In summary, the comment by Amrhein et al. raises some interesting points that seem worth discussing, but it makes important errors in analysis and interpretation, and misrepresents the results of Efron (2013a). This means the current version should not be approved.ReferencesBurnham, K.P. & D.R. Anderson. 2002. Model Selection and Multi-model Inference: a Practical Information-theoretic Approach. Springer-Verlag, New York.Colyvan, M. 2008. Is Probability the Only Coherent Approach to Uncertainty? Risk Anal. 28: 645-652.Efron B. (2003a) Bayes' Theorem in the 21st Century. Science 340(6137): 1177-1178.Efron B. (2013b) A 250-year argument: Belief, behavior, and the bootstrap. Bull Amer. Math Soc. 50: 129-146.Footnotes1. The twins are both male. However, if the twins were both female, the statistical results would be the same, so I will simply use the data that the twins are the same gender.2. In reality, the frequency of twins that are identical is likely to vary depending on many factors but we will accept 1/3 for now.3. Efron (2013b) reports the posterior probability for these twins being identical as “a whopping 61.4% with a flat Laplace prior” but as 2/3 in Efron (2013a). The latter (I assume 2/3 is “even more whopping”!) is the correct answer, which I confirmed via email with Professor Efron. Therefore, Efron (2013b) incorrectly claims the posterior probability is sensitive to the choice between a Jeffreys or Laplace uninformative prior.4. When the data are very informative relative to the different priors, the posteriors will be similar, although not identical.", "responses": [ { "c_id": "1485", "date": "29 Jul 2015", "name": "Valentin Amrhein", "role": "Author Response", "response": "We would like to sincerely thank Michael McCarthy for his thorough review, and we revised our paper accordingly. McCarthy's main point is that Efron's calculations and our approach differ because Efron's calculations are about one particular set of twins boys, while our analysis aims at making inference on the frequency of occurrence of identical twins in the larger population of twin boys. Indeed, we did not present a re-analysis of Efron's calculations but instead we used a data point that Efron casually cited, namely that the twin boys turned out to be fraternal. Our aim was to show that a Bayesian data analysis is not the same thing as solving a mathematical equation such as Bayes' theorem. In our manuscript, we now clarified that our approach is different from the calculations provided by Efron. We also shortened the manuscript and removed statements that were criticized by Michael McCarthy.Some other responses to McCarthy:\"This claim might be correct when describing uncertainty in the population-level frequency of identical twins. The data about the twin boys is not useful by itself for this purpose – they are a biased sample (the data have come to light because their gender is the same; they are not a random sample of twins).\"We agree that the sample would be biased if we were interested in the population of twins. If our population of interest are twin boys, however, the data are not biased.\"Further, a sample of size one, especially if biased, is not a firm basis for inference about a population parameter.\"We agree this would be a low sample size if this were an empirical study. We only used the data point to illustrate how to update the knowledge about the probability of identical twins among twin boys.\"Note, if the prior for Pr(x = 1) is specified as 0.5, or dunif(0,1), or dbeta(0.5, 0.5), the posterior probability that these twins are identical is 2/3 in all cases.\"In our view, the fixed posterior probability of 2/3 applies only to the prior specified as a fixed value of 0.5, while the other two prior distributions each produce posterior distributions of different shape. We thus would not agree to the notion that the posterior probabilities are identical in all cases (but we deleted the respective paragraph from our paper)." } ] } ]	1	https://f1000research.com/articles/2-278
https://f1000research.com/articles/3-303/v1	11 Dec 14	{ "type": "Opinion Article", "title": "Rampant software errors undermine scientific results", "authors": [ "David A. W. Soergel" ], "abstract": "Errors in scientific results due to software bugs are not limited to a few high-profile cases that lead to retractions and are widely reported. Here I estimate that in fact most scientific results are probably wrong if data have passed through a computer, and that these errors may remain largely undetected. The opportunities for both subtle and profound errors in software and data management are boundless, yet they remain surprisingly underappreciated.", "keywords": [ "Perhaps because of ingrained cultural beliefs about the infallibility of computation1", "people show a level of trust in computed outputs that is completely at odds with the reality that nearly zero provably error-free computer programs have ever been written2", "3." ], "content": "Computational results are particularly prone to misplaced trust\n\nPerhaps because of ingrained cultural beliefs about the infallibility of computation1, people show a level of trust in computed outputs that is completely at odds with the reality that nearly zero provably error-free computer programs have ever been written2,3.\n\nIt has been estimated that the industry average rate of programming errors is “about 15 – 50 errors per 1000 lines of delivered code”4. That estimate describes the work of professional software engineers—not of the graduate students who write most scientific data analysis programs, usually without the benefit of training in software engineering and testing5,6. The recent increase in attention to such training is a welcome and essential development7–11. Nonetheless, even the most careful software engineering practices in industry rarely achieve an error rate better than 1 per 1000 lines. Since software programs commonly have many thousands of lines of code (Table 1), it follows that many defects remain in delivered code–even after all testing and debugging is complete.\n\nSoftware errors and error-prone designs are compounded across levels of design abstraction. Defects occur not only in the top-level program being run but also in compilers, system libraries, and even firmware and hardware–and errors in such underlying components are extremely difficult to detect12.\n\n\nHow frequently are published results wrong due to software bugs?\n\nOf course, not every error in a program will affect the outcome of a specific analysis. For a simple single-purpose program, it is entirely possible that every line executes on every run. In general, however, the code path taken for a given run of a program executes only a subset of the lines in it, because there may be command-line options that enable or disable certain features, blocks of code that execute conditionally depending on the input data, etc. Furthermore, even if an erroneous line executes, it may not in fact manifest the error (i.e., it may give the correct output for some inputs but not others). Finally: many errors may cause a program to simply crash or to report an obviously implausible result, but we are really only concerned with errors that propagate downstream and are reported.\n\nIn combination, then, we can estimate the number of errors that actually affect the result of a single run of a program, as follows:\n\n\n\nFor these purposes, using a formula to compute a value in Excel counts as a “line of code”, and a spreadsheet as a whole counts as a “program”—so many scientists who may not consider themselves coders may still suffer from bugs13.\n\nAll of these values may vary widely depending on the field and the source of the software. For a typical analysis in bioinformatics, I’ll speculate at some plausible values:\n\n100,000 total LOC (neglecting trusted components such as the Linux kernel).\n\n20% executed\n\n10 errors per 1000 lines\n\n10% chance that a given error meaningfully changes the outcome\n\n10% chance that a consequent erroneous result is plausible\n\nSo, we expect that two errors changed the output of this program run, so the probability of a wrong output is effectively 100%. All bets are off regarding scientific conclusions drawn from such an analysis.\n\nLet’s imagine a more optimistic scenario, in which we write a simple, short program, and we go to great lengths to test and debug it. In such a case, any output that is produced is in fact more likely to be plausible, because bugs producing implausible outputs are more likely to have been eliminated in testing.\n\n1000 total LOC\n\n100% executed\n\n1 error per 1000 lines\n\n10% chance that a given error meaningfully changes the outcome\n\n50% chance that a consequent erroneous result is plausible\n\nHere the probability of a wrong output is 5%.\n\nThe factors going into the above estimates are rank speculation, and the conclusion varies widely depending on the guessed values. Measuring such values rigorously in different contexts would be valuable but also tremendously difficult. Regardless, it is sobering that some plausible values indicate total wrongness all the time, and that even conservative values suggest that an appreciable proportion of results are erroneous due to software defects–above and beyond those that are erroneous for more widely appreciated reasons.\n\n\nSoftware is exceptionally brittle\n\nA response to concerns about software quality that I have heard frequently—particularly from wet-lab biologists—is that errors may occur but have little impact on the outcome. This may be because only a few data points are affected, or because values are altered by a small amount (so the error is “in the noise”). The above estimates account for this by including terms for “meaningful changes to the result” and “the outcome is plausible”. Nonetheless, in the context of physical experiments, it is tempting to believe that small errors tend to reduce precision but have less effect on accuracy–i.e. if the concentration of some reagent is a bit off then the results will also be just a bit off, but not completely unrelated to the correct result.\n\nBut software is different. We cannot apply our physical intuitions, because software is profoundly brittle: “small” bugs commonly have unbounded error propagation. A sign error, a missing semicolon, an off-by-one error in matching up two columns of data, etc. will render the results complete noise. It is rare that a software bug would alter a small proportion of the data by a small amount. More likely, it systematically alters every data point, or occurs in some downstream aggregate step with effectively global consequences. In general, software errors produce outcomes that are inaccurate, not merely imprecise.\n\n\nMany erroneous results are plausible\n\nBugs that produce program crashes or completely implausible results are more likely to be discovered during development, before a program becomes “delivered code” (the state of code on which the above errors-per-line estimates are based). Consequently, published scientific code often has the property that nearly every possible output is plausible. When the code is a black box, situations such as these may easily produce outputs that are simply accepted at face value:\n\nAn indexing off-by-one error associates the wrong pairs of X’s and Y’s14.\n\nA correlation is found between two variables where in fact none exists, or vice versa.\n\nA sequence aligner reports the “best” match to a sequence in a genome, but actually provides a lower-scoring match.\n\nA protein structure produced from x-ray crystallography is wrong, but it still looks like a protein15.\n\nA classifier reports that only 60% of the data points are classifiable, when in fact 90% of the points should have been classified (and worse, there is a bias in which points were classified, so those 60% are not representative).\n\nAll measured values are multiplied by a constant factor, but remain within a reasonable range.\n\n\nSoftware errors and statistical significance are orthogonal issues\n\nA software error may produce a spurious result that appears significant, or may mask a significant result.\n\nIf the error occurs early in an analysis pipeline, then it may be considered a form of measurement error (i.e., if it systematically or randomly alters the values of individual measurements), and so may be taken into account by common statistical methods.\n\nHowever: typically the computed portion of a study comes after data collection, so its contribution to wrongness may easily be independent of sample size, replication of earlier steps, and other techniques for improving significance. For instance, a software error may occur near the end of the pipeline, e.g. in the computation of a significance value or of other statistics, or in the preparation of summary tables and plots.\n\nThe diversity of the types and magnitudes of errors that may occur16–19 makes it difficult to make a general statement about the effects of such errors on apparent significance. However it seems clear that, a substantial proportion of the time (based on the above scenarios, anywhere from 5% to 100%), a result is simply wrong—rendering moot any claims about its significance.\n\n\nWhat can be done?\n\nAll hope is not lost; we must simply take the opportunity to use technology to bring about a new era of collaborative, reproducible science20–22. Open availability of all data and source code used to produce scientific results is an incontestable foundation23–27. Beyond that, we must redouble our commitment to replicating and reproducing results, and in particular we must insist that a result can be trusted only when it has been observed on multiple occasions using completely different software packages and methods. This in turn requires a flexible and open system for describing and sharing computational workflows28. Projects such as Galaxy29, Kepler30, and Taverna31 have made inroads towards this goal, but much more work is needed to provide widespread access to comprehensive provenance of computational results. Perhaps ironically, a shared workflow system must itself qualify as a “trusted component”–like the Linux kernel–in order to provide a neutral platform for comparisons, and so must be held to the very highest standards of software quality. Crucially, any shared workflow system must be widely used to be effective, and gaining adoption is more a sociological and economic problem than a technical one32. The first step is for all scientists to recognize the urgent need.", "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\nAcknowledgements\n\nThanks to Annaliese Beery, Chris Warren, and Eli Dart for helpful comments on the manuscript.\n\n\nReferences\n\nToby SB: Myths about computers. SIGCAS Comput Soc. 1975; 6(4): 3–5. Publisher Full Text\n\nBird J: How many bugs do you have in your code? Java Code Geeks. 2011; 8. Reference Source\n\nFishman C: They write the right stuff. fastcompany. 1996. Reference Source\n\nMcConnell S: Code complete. Microsoft Press, Redmond, Wash. 2004. Reference Source\n\nMerali Z: Computational science: Error, why scientific programming does not compute. Nature. 2010; 467(7317): 775–777. PubMed Abstract \| Publisher Full Text\n\nJoppa LN, McInerny G, Harper R, et al.: Computational science. Troubling trends in scientific software use. Science. 2013; 340(6134): 814–5. PubMed Abstract \| Publisher Full Text\n\nBaxter SM, Day SW, Fetrow JS, et al.: Scientific software development is not an oxymoron. PLoS Comput Biol. 2006; 2(9): e87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeemann T: Ten recommendations for creating usable bioinformatics command line software. Gigascience. 2013; 2(1): 15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStodden V, Miguez S: Best practices for computational science: Software infrastructure and environments for reproducible and extensible research. J Open Res Softw. 2014; 2(1): e21. Publisher Full Text\n\nWilson G: Software carpentry: Getting scientists to write better code by making them more productive. Comput Sci Eng. 2006; 8(6): 66–69. Publisher Full Text\n\nWilson G, Aruliah DA, Brown CT, et al.: Best practices for scientific computing. PLoS Biol. 2014; 12(1): e1001745. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThimbleby H: Heedless programming: ignoring detectable error is a widespread hazard. Software: Practice and Experience. 2012; 42(11): 1393–1407. Publisher Full Text\n\nZeeberg BR, Riss J, Kane DW, et al.: Mistaken identifiers: gene name errors can be introduced inadvertently when using excel in bioinformatics. BMC Bioinformatics. 2004; 5(1): 80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHutson S: Data handling errors spur debate over clinical trial. Nat Med. 2010; 16(6): 618. PubMed Abstract \| Publisher Full Text\n\nChang G, Roth CB, Reyes CL, et al.: Retraction. Science. 2006; 314(5807): 1875–1875. PubMed Abstract \| Publisher Full Text\n\nBeizer B: Software testing techniques. Van Nostrand Reinhold, New York, 1990. Reference Source\n\nKhannur A: Structured Software Testing The Discipline of Discovering. Partridge Pub. 2014. Reference Source\n\nSpinellis D: Code Quality: The Open Source Perspective. Adobe Press, 2006. Reference Source\n\nVipindeep V, Jalote P: List of common bugs and programming practices to avoid them. Electronic, March, 2005. Reference Source\n\nHey T, Tansley S, Tolle K: The fourth paradigm: data-intensive scientific discovery. Microsoft Research, Redmond, Wash. 2009. Reference Source\n\nMesirov JP: Computer science. Accessible reproducible research. Science. 2010; 327(5964): 415–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNielsen MA: Reinventing discovery: the new era of networked science. Princeton University Press, Princeton, N.J. 2012. Reference Source\n\nBarnes N: Publish your computer code: it is good enough. Nature. 2010; 467(7317): 753. PubMed Abstract \| Publisher Full Text\n\nInce DC, Hatton L, Graham-Cumming J: The case for open computer programs. Nature. 2012; 482(7386): 485–8. PubMed Abstract \| Publisher Full Text\n\nLees JM: Open and free: Software and scientific reproducibility. Seismol Res Lett. 2012; 83(5): 751–752. Publisher Full Text\n\nMorin A, Urban J, Adams PD, et al.: Research priorities. Shining light into black boxes. Science. 2012; 336(6078): 159–160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSonnenburg S, Braun ML, Ong CS, et al.: The need for open source software in machine learning. J Mach Learn Res. 2007; 8: 2443–2466. Reference Source\n\nLudäscher B, Altintas I, Bowers S, et al.: Scientific process automation and workflow management. Scientific Data Management: Challenges, Existing Technology, and Deployment, Computational Science Series, 476–508: 2009. Reference Source\n\nGoecks J, Nekrutenko A, Taylor J, et al.: Galaxy: a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences. Genome Biol. 2010; 11(8): R86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltintas I, Berkley C, Jaeger E, et al.: Kepler: an extensible system for design and execution of scientific workflows. In Scientific and Statistical Database Management, 2004. Proceedings. 16th International Conference on, 423–424. IEEE, 2004. Publisher Full Text\n\nDe Roure D, Goble C: Software design for empowering scientists. Software IEEE. 2009; 26(1): 88–95. Publisher Full Text\n\nStodden VC: The scientific method in practice: Reproducibility in the computational sciences. 2010. Reference Source" }	[ { "id": "7019", "date": "16 Dec 2014", "name": "C. Titus Brown", "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\nDavid Soergel's opinion piece applies numerical calculations and common (software engineering) sense to thinking about errors in scientific software. I have seen no other piece that so simply and brutally summarizes the likely problems with current software development approaches in science, and I wholeheartedly agree with his recommendations. I think that the recommendation for a common trusted workflow system is an interesting one; I am particularly impressed by the point that we need separate implementations of important software, as this is often neglected by funding agencies and non-computational scientists.The large majority of the points in the paper are well taken and should not be controversial except perhaps in aggregate!The only major flaw in the paper is an overstatement of the central thesis. For example,The title is too definite; it needs a \"probably\" (which may decrease pithiness);Same with the abstract. One fix might be to eliminate the first sentence and move the last sentence to the top.For scenario 1, it's a pity there are no citations for these numbers, because they are nonintuitive (I found 20% executed to be too low, until I really thought it through, and then I agreed; but I'm not sure many people will believe). Is there any way to either bound or \"suppose\" these numbers a bit more?I would say that if the statements can be softened a bit to indicate that all of this is almost 100% certainly the case but we can't actually say it definitely then the article would be very acceptable.I didn't see the reference to the sign error debacle at the appropriate \"sign error\" point in the paper:http://boscoh.com/protein/a-sign-a-flipped-structure-and-a-scientific-flameout-of-epic-proportions.htmlAnother reference that could be usefully added, given space: http://www.fastcompany.com/28121/they-write-right-stuffIt might be worth adding a reference to the recent SSH debacle, where it turned out that incredibly well used software had a significant flaw. In other words, it's not enough for software to be well used for it to be correct! (Space permitting.)", "responses": [ { "c_id": "1131", "date": "22 Dec 2014", "name": "David Soergel", "role": "Author Response", "response": "Thanks for the kind and helpful comments! The editors prefer to wait for more reviews before issuing a revision, but in the meantime:In the title, I could go with \"...may undermine...\". (The loss in pithiness is indeed a shame, but so be it). Agreed re rearranging and toning down the abstract. I'll do another search for references and justifications for the ballpark estimates of % LOC executed and so on, but I expect this one will be hard because there's so much variation. For now the numbers are just my intuitions based on experience; I can try to clarify that at least. I think the fuzziest one is the plausibility term-- I know of no effort to measure that, and am not even sure how you'd go about it. In the course of code development and data analysis, how often do you look at a result and think \"that's just not right\"? That one varies too with the paranoia level of the scientist. (e.g., I'm a big fan of doing sanity checks that may reveal that some result is not plausible, even if that fact was not immediately obvious).Thanks for the citation suggestions. They're both already in there, actually (15 and 3, respectively), but I'll cite them from additional places as you suggest.And yes, I'll add a para mentioning \"Linus's Law\" (\"Given enough eyeballs, all bugs are shallow\") and the recent counterexamples (Heartbleed, Goto Fail, and Shellshock). These are notable because they're all security vulnerabilities, which (perhaps rightly) get a lot more press than bugs of other kinds. There's also a world of difference between widespread usage and widespread reading the code-- a distinction that is sometimes glossed over in these discussions.Thanks again for the comments!" }, { "c_id": "1472", "date": "29 Jul 2015", "name": "David Soergel", "role": "Author Response", "response": "Thanks again for the comments, and apologies for the long-delayed response. The changes described above are reflected in the new version." } ] }, { "id": "7096", "date": "22 Dec 2014", "name": "Daniel S. Katz", "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 opinion article makes a number of good qualitative points, and while I completely agree that there are errors in most software, I think the chances of those errors leading to incorrect published results are completely unknown, and could potentially be much smaller than the this paper claims. I think the basic claim in the title and the body of the paper may be dramatically overstated. The abstract says \"most scientific results are probably wrong,\" but this itself seems wrong. The author states, \"we must insist that a result can be trusted only when it has been observed on multiple occasions using completely different software packages and methods.\" First, I think this statement is overly focused on software. One method for developing trust in results from a particular code is that they match results from other codes. Another method is that they match results from experiment. A third method might be based on code review. Second, this statement is not only true for software, it is also true for this complete paper. In order to believe the chances for errors claimed here, this paper itself needs to be verified, and not at the level of each assumption made internally (in the \"How frequently ...\" section), but at the level of the overall claim. This is not easy, but it would be worthwhile, similar to the author's statement, \"Measuring such values rigorously in different contexts would be valuable but also tremendously difficult\" (but at a different level). If \"most scientific results are probably wrong,\" the author should be able to select a relatively small number of papers and demonstrate how software errors led to wrong results. I would like to see such an experiment, which would serve to verify this paper, rather than it standing as an unverified claim about verification. Finally, there is the classic problem with verification of a model (software, in this case): that fact that it works well in one domain is no guarantee that it will work well in another domain. Having made these objections to the degree of the illness of the patient, I mostly agree with remedies discussed in the last section. Open available of data and code is clearly good for both trust and reproducibility. Running (computational) experiments multiple ways can help finds any errors in any one of them, assuming they do not use common components (e.g., libraries, tools, or even hardware) that could lead to systematic biases. But how this should be done is less clear. For example, we have enough workflow systems that I don’t see any need for any one of them to be more trusted than the code that runs on them; we can just use different workflow systems with different code as part of the testing. Back to the author’s last point, I agree that \"to recognize the urgent need\" is essential, but to me, the need is verification; I could read this closing comment as saying that the need is widely adopted and widely trusted workflow tools. This should be clarified. In summary, this paper could be better titled and less strongly worded in places, and the paper itself needs to be verified. An alternate title would be one that makes the point, “Software, like other experiments, must be verified to be trusted”", "responses": [ { "c_id": "1473", "date": "29 Jul 2015", "name": "David Soergel", "role": "Author Response", "response": "Thanks very much for your insightful comments, and apologies for the long-delayed response. I believe I have addressed the main point about softening the claims throughout the paper. Some further thoughts follow:\"First, I think this statement is overly focused on software. One method for developing trust in results from a particular code is that they match results from other codes. Another method is that they match results from experiment. A third method might be based on code review.\"I focus on software because I think it is commonly trusted far out of proportion with its level of validation. Everyone understands that physical measurements must be validated, devices must be calibrated, experiments should ideally be reproduced in other labs, etc.-- but code seems to be a cultural blind spot in this regard.When a computational result can be directly compared to an experimental result, then of course agreement should increase trust in both. More commonly, I think, a given result arises from a combination of experiment (\"data collection\") and computation (\"data analysis\"), and comparisons can only be made between attempts incorporating both. Again agreement from multiple attempts should increase trust--but only if the analysis steps lack common components. This is another reason to focus on software: software is typically used downstream of data collection, so a bug can easily mask whatever signal is present in the underlying data, producing spurious agreement or spurious disagreement in the final result. Because software often has the last word in generating a result, then, it demands an even higher level of trust than upstream inputs.Code review is certainly a good thing, but in my view is never sufficient to generate trust. Anecdotally, I've found plenty of bugs in code that was already reviewed. In any case, \"code review\" means many things to many people, and obviously the likelihood of finding bugs varies widely with the skill of the reviewer.\"Second, this statement is not only true for software, it is also true for this complete paper. In order to believe the chances for errors claimed here, this paper itself needs to be verified, and not at the level of each assumption made internally (in the \"How frequently ...\" section), but at the level of the overall claim. This is not easy, but it would be worthwhile, similar to the author's statement, \"Measuring such values rigorously in different contexts would be valuable but also tremendously difficult\" (but at a different level). If \"most scientific results are probably wrong,\" the author should be able to select a relatively small number of papers and demonstrate how software errors led to wrong results. I would like to see such an experiment, which would serve to verify this paper, rather than it standing as an unverified claim about verification.\"This is an opinion piece; I hope the more speculative language now makes clear that I am expressing justifiable anxiety that a serious problem may exist, rather than asserting that it definitely does exist. I certainly agree that verifying my estimates would be a great thing to do (particularly the aggregate error rate, not just the individual factors, as you point out). However I think that would be a major undertaking that is not tractable for me to do in this paper.I do already cite a number of cases where software bugs resulted in wrong results, but these are basically anecdotal, and of course they are the ones that have already been found and reported. The proportion of these to the overall literature is vanishingly small. There are surely many more papers where an author or reader is privately aware of an error. And a still much larger proportion of papers, I believe (but cannot prove), contain errors that remain completely unknown.I can think of only two ways to determine that proportion empirically. The first is to identify existing attempts to reproduce results, confirm that they are not subject to common sources of error, and track down the causes of any disagreement. This method may be subject to selection bias (i.e. in general, only important or controversial results get replication attempts in the first place).The second is to take a random sample of papers and attempt to fully reproduce them, or at least to carefully review the code in search of errors. That would be really a lot of work-- in one example, an independent reproduction of a single computational study took 3 months. A systematic campaign to reproduce computational results would be great, inspired by similar efforts focusing on reproducing experimental results (e.g. the Amgen study and the OSF Reproducibility Project).But I can't take it on alone! Rather I hope this paper helps to demonstrate the need for researchers, funders, and publishers to take code verification more seriously, and to foster the reproduction studies that would be needed to confirm or deny my estimates. Crucially, it's not just a matter of successfully running a study author's code (which, in the example case of in ACM conferences and journals, can be downloaded and compiled for only about half of the papers anyway). Journal policies requiring at least that level of replication would be a good start. But really the point here is to use different code to generate the same result.So I think we agree: I am making an unverified claim about verification, and I too would like to see it verified.\"Finally, there is the classic problem with verification of a model (software, in this case): that fact that it works well in one domain is no guarantee that it will work well in another domain.\"True, we can never make absolute guarantees. But we can do better than the status quo, which all too often provides no verification at all. Also, this point opens the question of the breadth of applicability of a given software artifact: some software does only a very specific thing, and so can be thoroughly verified within its single domain, while other software is very generic and so is much harder to verify across domains. I don't address this in the paper, except to the extent that the factor for \"proportion of lines executed\" is tangentially related (e.g., successful tests exercising some code paths say nothing about runs taking different code paths). That factor could be thought of more abstractly as the likelihood that verification in one domain should generate trust in another.\"Having made these objections to the degree of the illness of the patient, I mostly agree with remedies discussed in the last section. Open available of data and code is clearly good for both trust and reproducibility. Running (computational) experiments multiple ways can help finds any errors in any one of them, assuming they do not use common components (e.g., libraries, tools, or even hardware) that could lead to systematic biases. But how this should be done is less clear. For example, we have enough workflow systems that I don’t see any need for any one of them to be more trusted than the code that runs on them; we can just use different workflow systems with different code as part of the testing.\"I agree that the ideal way to gain trust in a particular result is to observe agreement from two experiments in which everything differs, even the workflow system and the hardware (at some length: http://davidsoergel.com/posts/confirmation-depth-as-a-measure-of-reproducible-scientific-research, but just see Fig. 2 there for the main point).However: if we could agree on a common, trusted workflow system, that would make it much easier both to verify software components and to track down sources of error, simply by swapping out individual components of workflows with purportedly equivalent alternative implementations. When components are reused across workflows (and across labs, etc.), crowdsourced results from such component-swap experiments would quickly reveal which components are most commonly associated with robust results. I'll have to describe that vision more thoroughly elsewhere, but for now I hope it points at one thing I hope we could gain from standardizing workflow and provenance descriptions. Perhaps more simply: researchers are more likely to examine (and perhaps tweak and reuse) a workflow written in a language (or graphical notation, etc.) with which they are already familiar; Balkanization of workflow systems largely defeats their purpose.\"Back to the author’s last point, I agree that \"to recognize the urgent need\" is essential, but to me, the need is verification; I could read this closing comment as saying that the need is widely adopted and widely trusted workflow tools. This should be clarified.\"I did really mean both things--I've tried to clarify that.Thanks again for the very helpful comments!" } ] } ]	1	https://f1000research.com/articles/3-303
https://f1000research.com/articles/4-162/v1	23 Jun 15	{ "type": "Review", "title": "Magnetic Resonance Imaging in Neuropsychiatric Lupus", "authors": [ "Nicolae Sarbu", "Núria Bargalló", "Ricard Cervera", "Nicolae Sarbu", "Núria Bargalló" ], "abstract": "Neuropsychiatric lupus is a major diagnostic challenge, and a main cause of morbidity and mortality in patients with systemic lupus erythematosus (SLE). Magnetic resonance imaging (MRI) is, by far, the main tool for assessing the brain in this disease. Conventional and advanced MRI techniques are used to help establishing the diagnosis, to rule out alternative diagnoses, and recently, to monitor the evolution of the disease. This review explores the neuroimaging findings in SLE, including the recent advances in new MRI methods.", "keywords": [ "neuropsychiatric lupus", "NPSLE", "systemic lupus erythematosus", "SLE", "magnetic resonance imaging", "MRI" ], "content": "Introduction\n\nDespite the fact that the outcome of patients with systemic lupus erythematosus (SLE) has improved considerably over the last decades, neuropshychiatric involvement remains a main cause of morbi-mortality1,2, being responsible for up to 19% of deaths in SLE3,4. The real prevalence of neuropsychiatric SLE (NPSLE) remains unknown, with significant heterogeneity between studies, from 14% to 95% depending on the inclusion criteria; an average of 40–50% is probably widely accepted5–8. Reliable methods for diagnosing NPSLE are also unknown, the clinical judgment remaining the cornerstone for differentiation of these patients9,10. Therefore, NPSLE represents a major diagnostic challenge, being essentially a diagnosis of presumption and exclusion, established after having ruled out other possible causes such as trauma, infection, drug effects, epilepsy, migraine, psychiatric disorders, multiple sclerosis, posterior reversible encephalopathy and previous nervous system disorders5,6,11,12. On the other hand, reaching the correct diagnosis of NPSLE is critical in terms of therapeutic decisions and outcome.\n\nAccording to 1999 American College of Rheumatology (ACR) Case Definitions for NPSLE, 19 neuropsychiatric syndromes are defined, divided into 12 central and 7 peripheral13. The central ones are further divided into neurological (aseptic meningitis, cerebrovascular disease, demyelinating syndrome, headache including migraine and benign intracranial hypertension, movement disorders, myelopathy, epilepsy), and psychiatric (acute confusional states, anxiety disorder, cognitive dysfunction, affective disorder). The peripheral syndromes are acute inflammatory demyelinating polyradiculopathy (Guillain-Barre syndrome), autonomic disorder, mononeuropathy (single/multiplex), myasthenia gravis, cranial neuropathy, plexopathy, and polyneuropathy. The most common syndromes which require neuroimaging studies are headache, cerebrovascular disease, epilepsy and cognitive dysfunction8,14, and also represent four out of five globally most prevalent NPSLE syndromes, as demonstrated by an extensive, recent meta-analysis7. NPSLE is also divided into primary, attributed to SLE specific mechanisms, and secondary, consequence of infections, drugs or metabolic errors, although there are no definitive methods to differentiate between them6,10.\n\nIn spite of outstanding advances and increasing efforts into research, the physiopathology of NPSLE remains still unclear. Neural and vascular injuries related to antibodies and cytokines were incriminated in active NPSLE. The pathological substrate of NPSLE consists of microangiopathic disease (the most frequent neuropathological finding, typically multifocal, due to intimal hyperplasia, erythrocytes extravasation and fibrin thrombi), macroscopic infarcts (partially explained by secondary coagulopathy due to antiphospholipid antibodies or by embolic phenomena due to Libman–Sacks endocarditis), accelerated atherosclerosis (partially due to steroid treatment, vasculitis and microhemorrhages), direct immune mediated alterations, demyelination and microembolisms5,15–18.\n\nMagnetic resonance imaging (MRI) is the gold standard for studying the brain in SLE. The role of other imaging modalities such as computer tomography (CT) is essentially to rule out acute complications such as hemorrhage or large infarcts, or to assess differential diagnoses5,19,20. The large spectrum of clinical presentations, laboratory and pathological findings in NPSLE made the neuroradiological findings nonspecific, a wide range of abnormalities being described8,21. The most frequently reported findings with conventional MRI in large series of NPSLE were multiple small white-matter lesions (30–75%) and cortical atrophy (15–20%), although there is a large percentage of patients (25–60%) with normal MRI scan8,11,22,23. Advanced MRI techniques such diffusion-tensor, magnetization-transfer and volumetric studies, which give microstructural and functional information, could provide evidence of subtle brain changes that allow better understanding of the NPSLE mechanisms. Furthermore, the correlation of the neuroradiological, clinical and immunological biomarkers could give insights into the pathophysiology of the disease. The present review aims to describe the neuroimaging findings in conventional and advanced MRI imaging in NPSLE patients, and their importance from a practical point of view.\n\nAround 50% of the NPSLE patients had normal MRI, especially in diffuse syndromes such as headache, mood disorder, and psychiatric disease8. In the other half of the patients, the most common neuroimaging findings can be classified as vascular diseases (small or large vessel disease), and inflammatory-type lesions (Table 1).\n\nAbbreviations: GCA-Global Cortical Atrophy scale.\n\nVascular disease, although nonspecific and in many forms of manifestation, is the hallmark of NPSLE8. Vascular lesions are ill-defined hyperintensities on T2, and moderately hypointense or isointense on T1. Large vessel disease refers to large infarcts, which have medium-to-large size, are roughly wedge-shaped, occur with a vascular territory distribution, and involve both grey and white matter (Figure 1a). With DWI sequences it is possible to determine if they are in the acute, subacute or chronic stage, including silent infarcts. Large vessel infarcts are one of the most debilitating complications of NPSLE, and are found in 10–15% of patients and at a mean age of 35–40 years8,22,24,25. When infarcts occur in NPSLE, a tendency to multiplicity was noticed, which is translated into a high recurrence of ischemic events8. Middle cerebral artery territory is mainly involved, as in the general population8. Many authors associated antiphospholipid antibodies with infarcts and reported a stroke recurrence of around 50% when these antibodies were present22,24,26. Stroke was also more commonly observed in the presence of hypertension, cerebrovascular syndrome and seizures5.\n\n1a. Large hyperintense cortico-subcortical area consistent with a chronic stroke involving the right middle cerebral artery territory. 1b. Focal bilateral white matter hyperintensities reflecting small vessel disease. Figure origin: Department of Radiology, Hospital Clinic Barcelona.\n\nSmall vessel disease is typically represented by lesions smaller than 1 cm, which follow the distribution of the white matter (periventricular, deep, subcortical) (Figure 1b). Recently, the definitions of neuroimaging findings of small vessel disease have been established and consist in white-matter hyperintensities, recent small subcortical infarcts, lacunes, microbleeds and brain atrophy27. White-matter hyperintensities (WMH) are the most widespread type of small vessel disease seen in SLE patients, and represents the collective term referring to small T2-hyperintensities including the white matter, basal ganglia, cerebellum and brainstem27. They are characterized as hyperintense on T2 and FLAIR sequences, without cavitation, generally small and ill-defined27,28. The differential diagnosis of WMH is very wide, being associated with many conditions including ageing, dyslipidemia, diabetes, hypertension, heart diseases and migraine27. However, many previous reports already proved increased frequency of WMH in SLE and NPSLE21,29–36. WMH had been shown to involve preferentially the frontal and parietal lobes, consistent with an anterior to posterior gradient, similar to other causes of WMH, but different from inflammatory demyelinating etiologies such as multiple sclerosis8,37. In a quantitative cerebral MRI assessment, Appenzeller et al.34 showed that age, duration of neuropsychiatric manifestations and total corticosteroid dosage were independent predictors for WMH in SLE. In a recent study in patients with newly diagnosed SLE, WMH were found in 8% of the patients38. Nevertheless, these lesions were observed more frequently in NPSLE when compared with SLE without neuropsychiatric manifestations, with average ranges from 40 to 60%8,11,19,34,36,38,39. WMH were associated with cerebrovascular disease, cognitive dysfunction, seizures, antiphospholipid antibodies, low complements (C3, C4, CH50), age, disease duration, and total corticosteroid dose8,34. Previous reports demonstrated a significant association between both NPSLE activity (Neuro-SLEDAI) and injury (Neuro-SLICC) scores with the number of WMH (high lesion burden)11,34,36,38,40. Furthermore, new lesions were noticed during onset of new neuropsychiatric manifestations, and resolution of lesions was found with clinical improvement25,41,42. Quantitative methods are increasingly proposed for the quantification and follow-up of the WMH in NPSLE, as they can function as an independent predictor for the NPSLE activity and injury, holding promise to open a new line of follow-up of NPSLE patients and their response to therapy, similar to the monitoring of multiple sclerosis32,34,38.\n\nRecent small subcortical infarcts, commonly known as lacunar infarcts, are infarctions in the territory of perforating arterioles, of less than 20 mm in its maximum diameter in the axial plane, with imaging signs or clinical symptoms consistent with a lesion occurring in the previous few weeks27. Their natural evolution is into lacunes, WMH without cavitation, or they might disappear43. Old lacunar infarcts (lacunes) must be differentiated from perivascular (Virchow-Robin) spaces, which generally are smaller, located mostly around the anterior commissure and usually appear linear when imaged parallel to the course of the vessel. Lacunes were commonly described in elderly, asymptomatic individuals, in the presence of hypertension, and were related to dementia, gait impairment and increased risk of stroke27. Very few studies evaluated lacunes in NPSLE and they were found with a prevalence of 11.5–16%, higher than in the general population8,11. Cerebral microbleeds are small (usually 2–5 mm, but up to 10 mm) round or oval areas of signal void with associated blooming on paramagnetic-sensitive sequences such as T2*-weighted gradient recalled echo (GRE) or susceptibility-weighted images (SWI). Microscopically, hemosiderin-laden macrophages in perivascular tissue are seen, indicating vascular leakage of blood cells, related to bleeding-prone microangiopathy. In the general population, microbleeds are usually located in the cortico-subcortical junction, deep grey and white matter, brainstem and cerebellum. They were associated with hypertension, amyloid angiopathy, cognitive impairment and Alzheimer disease44,45. In NPSLE, microbleeds were found in 14.5% of the patients on GRE/SWI sequences, and were correlated with lupus anticoagulant (antiphospholipid antibodies) and cerebrovascular syndrome8.\n\nCortical atrophy is seen as generalized enlargement of peripheral cerebrospinal fluid spaces and is best evaluated on volumetric 3D-T1 or FLAIR images. In the general population, age related atrophic changes are small prior to age 50 years, as proved by a large study46 or, similarly, by another underlying that brain volumes in females remained stable over a span of 15 to 69 years of age47. There are different scales to unify the radiological language, one of the most known being the global cortical atrophy scale (GCA)48. GCA evaluation at the onset of NPSLE observed cortical atrophy in 18.5% of the subjects, most commonly in a mild grade, and at a mean age of 42.5 years8. Brain atrophy occurs more frequently in the presence of other radiological manifestations consistent with small vessel disease, such as WMH, high lesion burden, lacunes and microbleeds8. Brain atrophy was also correlated with lupus anticoagulant, low complement, longer disease duration, cognitive dysfunction and cerebrovascular disease22,36. Many authors suggested that the atrophy might be the result of the prednisone use36,49, while others found no association30,38,50,51, which suggests that additional mechanisms, probably related to NPSLE itself, seem to be involved52–55.\n\nLess frequently, some NPSLE patients present inflammatory-type lesions which were described as ill defined, hyperintense on T2 and FLAIR, involving the grey and white matter, generally medium or large-sized, some of them with contrast enhancement or diffusion restriction, without vascular territory distribution nor clinical and radiological features of infarcts, which usually resolves after aggressive corticosteroid treatment. They were reported in 5–10% of patients, and were correlated with low complement levels, indicating an immunological damage related to antibodies and cytokines and supporting the immunological pathogenesis of NPSLE6. Yet rarely present, findings related to cerebral vasculitis were described, when angiography exams (MRI or conventional) could show focal beadings and narrowings of large and small arteries15,56–59.\n\nMyelitis, a type of inflammatory involvement of the central nervous system, is one of the most debilitating complications of NPSLE and occurs in 1–5% of SLE patients. It usually develops early in the evolution of the disease and associates a worse outcome. In 39% of the patients with SLE related myelopathy, it constitutes the presenting symptom of SLE, and in another 42% it occurs during the first 5 years after the diagnosis. The most described MRI pattern in SLE myelitis is consistent with transverse myelitis: commonly long affected segment, more than 2–3 vertebral bodies in length and with injury of both halves of the cord (Figure 2). Transverse myelitis associates a variable swelling and focal enlargement. Enhancement is usually absent or poor, patchy in the most active presentations. The outcome of SLE myelitis is variable, ranging from complete recovery to severe disability, but the injury is typically much less extensive on follow up MRI10,60.\n\nFigure origin: Department of Radiology, Hospital Clinic Barcelona.\n\nUp to 40–50% of NPSLE patients had no brain abnormalities on conventional MRI8,11,30,32,35,61. Nonetheless, advanced MRI sequences in NPSLE demonstrated underlying abnormalities in normal appearing white and grey matter, which shows the limitations of conventional sequences51,62,63. Recent studies used advanced MRI techniques in the analysis of NPSLE, as the assessment of tissue-specific atrophy by morphometric methods37,51,64, diffusion-tensor imaging51,63,65,66, magnetization transfer imaging51,65,66, magnetic resonance spectroscopy66,67 and perfusion MRI.\n\nVoxel based morphometry (VBM) is a technique which allows the assessment of the focal differences in brain anatomy and, therefore, the assessment of tissue-specific atrophy. The volume in every voxel is compared across the brain, and VBM is frequently performed for examining differences between populations, although it can also be used to assess asymmetries between brain hemispheres. Morphometric studies showed that decreased whole brain volume with increased lateral ventricle volume and both global gray matter and white matter atrophy are present in SLE patients compared to healthy controls68. Moreover, it was demonstrated that atrophy evolved over a short period of time51,67,69. A number of publications found that selective cortical atrophy was the tissue specific atrophy measure with best correlation with the presence of NPSLE, and suggested that cortical atrophy is more important for mediating brain damage in NPSLE patients than the white matter lesions51,69. From a practical point of view, the macroscopic damage of the cortical gray matter might be more important for identifying NPSLE patients than the micro- or macrostructural damage in the white matter51,69, yet it was reported an association of NPSLE with both cortical and central atrophy51,64,70. Some authors compared cohorts of NPSLE with SLE and controls. The NPSLE group exhibited decreased cortical thickness in left frontal and parietal lobes as well as in right parietal and occipital lobes compared to controls. Both SLE and NPSLE groups exhibited comparable thinning in frontal and temporal lobes71. Automated morphometric methods were also used for segmenting white matter lesions in patients with SLE, which could give a more precise quantification of the focal injuries72.\n\nDiffusion-tensor imaging (DTI) is based on the measurement of water diffusion through cellular compartments, and was demonstrated to provide better resolution than conventional sequences regarding white matter microstructure (Figure 3a–b)73,74. Compared to more isotropic movement of water in gray matter, the diffusion in white matter presents higher anisotropy, with preferential diffusion along the length of the axon. This anisotropy is due to the well-structured axonal membranes and their myelin sheaths. The diffusion can be quantified by the following parameters: apparent diffusion coefficient (ADC), fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD) and axial diffusivity (AD). FA is a measure of myelination and axonal integrity, and MD a measure of molecular motion. High FA and low MD suggest greater myelination and axonal integrity. Previous studies found changes in various DTI indices in SLE and NPSLE patients, in relation to important microscopic injury of the white matter75. In patients with SLE, white matter injury in frontal lobes, corpus callosum, and thalamus has been found67,76–78. FA values were reported to be lower and MD values to be higher in the brain of NPSLE patients than in healthy controls. Increased AD of white matter was also correlated with NPSLE when compared to healthy populations. It was suggested that the underlying pathological substrate of white matter changes in NPSLE may be the selective axonal damage51,65,69. A localized injury of white matter tracts was also demonstrated in the limbic system, internal capsule, corpus callosum, forceps major and corona radiata63,66,77,79,80. Very recent publications underline the role of DTI as an imaging biomarker of NPSLE81.\n\n3a. Normal FA shows the integrity and directionality of the white matter fibers (red: right-left, green: anterior-posterior, blue: craneo-caudal). 3b. Altered (low) FA seen as loss of the normal colors of the left corticospinal tract in the internal capsule and of the left longitudinal fasciculus related to ischemic infarct of the territory of the left middle cerebral artery (arrows) in a patient with neuropsychiatric lupus and stroke. Figure origin: Department of Radiology, Hospital Clinic Barcelona.\n\nMagnetization transfer imaging (MTI) is based on the interaction between free water protons and bound protons. The differences in the proton mobility in various macromolecules and tissues are used to generate differences in image signal. Thus, MTI is used to generate contrast, and it has a variety of clinical applications. Volumetric MTI was used to quantify cerebral lesions in different diseases, mainly in multiple sclerosis. Bosma et al.82 compared MTI histogram parameters in 5 groups of patients: active NPSLE, chronic NPSLE, SLE without NPSLE, multiple sclerosis, and normal control subjects. The magnetization transfer ratio histograms in the group of SLE without NPSLE and the group of healthy controls were similar, whereas those in chronic NPSLE and multiple sclerosis groups were flatter. The active NPSLE group showed also a flattening of the histograms, but with a higher magnetization transfer ratio. This suggests that MTI could be able to differentiate active NPSLE. It is also believed that MTI might be a good method for monitoring treatment trials in NPSLE82. A report combining MTI with magnetic resonance spectroscopy (MRS) found correlation between brain atrophy and MRS markers of axonal and myelin damage66. Studies combining MTI with DWI, MRS and T2 relaxometry data in NPSLE suggest a common pathogenesis in NPSLE in spite of the many differences in the neuropsychiatric presentation51,65.\n\nMRS allows the analysis of brain metabolites. Different proton groups have different magnetic fields in relation to their valence electrons. As a result, they resonate at different frequencies of the magnetic field, which can be demonstrated by MRS, as peaks that correspond to different metabolites. N-acetylaspartate (NAA) is one of the main markers assessed on MRS and is found in higher concentrations in neurons, thus it is a marker of neuronal viability. Glutamate, a non-essential amino acid, is the most important excitatory neurotransmitter, and prolonged neuron excitation by glutamate can be toxic to neurons. NAA and glutamine-glutamate changes were demonstrated in normal-appearing brain in SLE patients, before neurologic and imaging manifestations became apparent, which suggests that these markers might predict the early cerebral involvement of SLE83. Lower NAA ratios were also reported in both SLE and NPSLE patients62, and increased myo-inositol, a marker of gliosis, was suggested as a marker of poor prognosis in NPSLE84.\n\nPerfusion imaging. There are three techniques of perfusion MRI, based on the administration of gadolinium (dynamic susceptibility contrast imaging and dynamic contrast enhanced imaging), or without contrast administration (arterial spin labeled imaging). The main parameters derived from them are mean transit time (MTT), time to peak (TTP), cerebral blood flow (CBF) and cerebral blood volume (CBV). The defined pathological patterns are hypoperfusion (high MTT/TTP, low CBF/CBV) and hyperperfusion (low TTP/MTT, high CBV/CBF)85. Few prospective studies analyzed brain perfusion in SLE patients. Some authors showed that perfusion in SLE patients was not different from healthy controls86, while others reported a pattern of hypoperfusion in both SLE and NPSLE62, or even hyperperfusion in the posterior cingulate gyrus in patients with active disease87.\n\nOverall, advanced MRI techniques seem to be able to detect microstructural brain damage in a very early stage when not visible on conventional sequences. There could be a temporal dissociation between the detection of damage with these sequences and its translation to significant abnormalities on conventional MRI. Advanced MRI is also expected to help to better understand the underlying pathological substrate of cerebral damage in NPSLE. However, the role of advanced MRI techniques in patients with SLE is yet in its infancy and needs to be further investigated. Future longitudinal studies should determine whether early changes of the white and gray matter in NPSLE patients may involve a higher degree of tissue-specific brain atrophy over time and to what extent it would be possible to monitor disease progression and response to therapy.\n\nLooking at all sides of the argument, it is questionable what patients and when should be referred for brain MRI and what is the role of MRI in the clinical management of NPSLE. Syndromes such as cerebrovascular disease, cognitive dysfunction, seizures and myelopathy, as well as the focal symptomatology, were often related with radiological abnormalities, and require to be comprehensively studied; additional sequences such as DWI, GRE/SWI and contrast-enhanced should be included when MRI is performed in these patients. Conversely, MRI is more likely to be unremarkable in some other syndromes such as headache, psychosis and, generally, diffuse neurological presentations rather than focal ones. Additionally, the status of antiphospholipid antibodies, complement and disease activity plays an important role. Therefore, in the current settings, the decision of when imaging a patient remains probably best reached through a case-based clinical judgment.\n\nThe other side of the argument of MRI in NPSLE regards ruling out other causes of neuropsychiatric manifestations rather than diagnosing NPSLE. Despite MRI being the imaging modality of choice and despite significant recent advances in this field, there are neither diagnostic nor specific radiological findings for NPSLE, meaning that MRI cannot confirm nor exclude the diagnosis of NPSLE. Nevertheless, in the absence of alternative diagnoses when imaging a SLE patient, some patterns may be proposed: stroke in young patients, more than one infarct, association between large and small vessel disease, high lesion burden at young age and premature cortical atrophy. All these in a subject meeting criteria for SLE and without other risk factors, probably could suggest either the presence or a possible development of NPSLE in the following period.\n\nIn conclusion, MRI is crucial for both supporting the diagnosis of NPSLE and for ruling out alternative diagnoses. The multimodal approach including conventional and advanced MRI may be an important tool for monitoring the disease activity, progression and treatment response in NPSLE, and may provide fundamental insights into the pathological substrate. To make this possible, a common radiological terminology is a first requirement.", "appendix": "Author contributions\n\n\n\nAll authors were involved in writing and editing the manuscript and have approved the final version.\n\n\nCompeting interests\n\n\n\nNo competing interests are disclosed.\n\n\nGrant information\n\nThis work has been founded by the Sociedad Española de Radiológia Médica (06_NBA_INVESTIGACION _SERAM_2013).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nTo Pilar Toledano, MD, and Gerard Espinosa, MD, PhD, Department of Autoimmune Diseases, Hospital Clinic, Barcelona, Catalonia, Spain, for their collaboration in neuropsychiatric lupus research projects.\n\n\nReferences\n\nCervera R, Khamashta MA, Font J, et al.: Morbidity and mortality in systemic lupus erythematosus during a 10-year period: a comparison of early and late manifestations in a cohort of 1,000 patients. Medicine (Baltimore). 2003; 82(5): 299–308. PubMed Abstract \| Publisher Full Text\n\nO’Neill S, Cervera R: Systemic lupus erythematosus. Best Pract Res Clin Rheumatol. 2010; 24(6): 841–855. PubMed Abstract \| Publisher Full Text\n\nChiewthanakul P, Sawanyawisuth K, Foocharoen C, et al.: Clinical features and predictive factors in neuropsychiatric lupus. Asian Pac J Allergy Immunol. 2012; 30(1): 55–60. PubMed Abstract\n\nZirkzee EJ, Huizinga TW, Bollen EL, et al.: Mortality in neuropsychiatric systemic lupus erythematosus (NPSLE). Lupus. 2014; 23(1): 31–38. PubMed Abstract \| Publisher Full Text\n\nJoseph FG, Scolding NJ: Neurolupus. Pract Neurol. 2010; 10(1): 4–15. PubMed Abstract \| Publisher Full Text\n\nBertsias GK, Boumpas DT: Pathogenesis, diagnosis and management of neuropsychiatric SLE manifestations. Nat Rev Rheumatol. 2010; 6(6): 358–367. PubMed Abstract \| Publisher Full Text\n\nUnterman A, Nolte JE, Boaz M, et al.: Neuropsychiatric syndromes in systemic lupus erythematosus: a meta-analysis. Semin Arthritis Rheum. 2011; 41(1): 1–11. PubMed Abstract \| Publisher Full Text\n\nSarbu N, Alobeidi F, Toledano P, et al.: Brain abnormalities in newly diagnosed neuropsychiatric lupus: systematic MRI approach and correlation with clinical and laboratory data in a large multicenter cohort. Autoimmun Rev. 2015; 14(2): 153–159. PubMed Abstract \| Publisher Full Text\n\nCervera R, Abarca-Costalago M, Abramovicz D, et al.: Systemic lupus erythematosus in Europe at the change of the millennium: lessons from the “Euro-Lupus Project”. Autoimmun Rev. 2006; 5(3): 180–186. PubMed Abstract \| Publisher Full Text\n\nBertsias GK, Ioannidis JP, Aringer M, et al.: EULAR recommendations for the management of systemic lupus erythematosus with neuropsychiatric manifestations: report of a task force of the EULAR standing committee for clinical affairs. Ann Rheum Dis. 2010; 69(12): 2074–2082. PubMed Abstract \| Publisher Full Text\n\nLuyendijk J, Steens SC, Ouwendijk WJ, et al.: Neuropsychiatric systemic lupus erythematosus: lessons learned from magnetic resonance imaging. Arthritis Rheum. 2011; 63(3): 722–732. PubMed Abstract \| Publisher Full Text\n\nRamage AE, Fox PT, Brey RL, et al.: Neuroimaging evidence of white matter inflammation in newly diagnosed systemic lupus erythematosus. Arthritis Rheum. 2011; 63(10): 3048–3057. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nACR Ad Hoc Committee on Neuropsychiatric Lupus Nomenclature. The American College of Rheumatology nomenclature and case definitions for neuropsychiatric lupus syndromes. Arthritis Rheum. 1999; 42(4): 599–608. PubMed Abstract \| Publisher Full Text\n\nNetto TM, Zimmermann N, Rueda-Lopes F, et al.: Neuropsychiatric lupus: classification criteria in neuroimaging studies. Can J Neurol Sci. 2013; 40(3): 284–291. PubMed Abstract\n\nEllis SG, Verity MA: Central nervous system involvement in systemic lupus erythematosus: a review of neuropathologic findings in 57 cases, 1955--1977. Semin Arthritis Rheum. 1979; 8(3): 212–221. PubMed Abstract \| Publisher Full Text\n\nSibbitt WL Jr, Brooks WM, Kornfeld M, et al.: Magnetic resonance imaging and brain histopathology In neuropsychiatric systemic lupus erythematosus. Semin Arthritis Rheum. 2010; 40(1): 32–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSarbu MI, Salman-Monte TC, Rubio Muñoz P, et al.: Differences between clinical and laboratory findings in patients with recent diagnosis of SLE according to the positivity of anti-dsDNA by the Crithidia luciliae method. Lupus. 2015; pii: 0961203315573852. PubMed Abstract \| Publisher Full Text\n\nFanouriakis A, Boumpas DT, Bertsias GK: Pathogenesis and treatment of CNS lupus. Curr Opin Rheumatol. 2013; 25(5): 577–583. PubMed Abstract \| Publisher Full Text\n\nSibbitt WL Jr, Sibbitt RR, Griffey RH, et al.: Magnetic resonance and computed tomographic imaging in the evaluation of acute neuropsychiatric disease in systemic lupus erythematosus. Ann Rheum Dis. 1989; 48(12): 1014–1022. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZardi EM, Taccone A, Marigliano B, et al.: Neuropsychiatric systemic lupus erythematosus: tools for the diagnosis. Autoimmun Rev. 2014; 13(8): 831–839. PubMed Abstract \| Publisher Full Text\n\nCastellino G, Govoni M, Giacuzzo S, et al.: Optimizing clinical monitoring of central nervous system involvement in SLE. Autoimmun Rev. 2008; 7(4): 297–304. PubMed Abstract \| Publisher Full Text\n\nCsépány T, Bereczki D, Kollár J, et al.: MRI findings in central nervous system systemic lupus erythematosus are associated with immunoserological parameters and hypertension. J Neurol. 2003; 250(11): 1348–1354. PubMed Abstract \| Publisher Full Text\n\nToledano P, Sarbu N, Espinosa G, et al.: Neuropsychiatric systemic lupus erythematosus: Magnetic resonance imaging findings and correlation with clinical and immunological features. Autoimmun Rev. 2013; 12(12): 1166–1170. PubMed Abstract \| Publisher Full Text\n\nJoseph FG, Lammie GA, Scolding NJ: CNS lupus: a study of 41 patients. Neurology. 2007; 69(7): 644–654. PubMed Abstract \| Publisher Full Text\n\nSibbitt WL Jr, Sibbitt RR, Brooks WM: Neuroimaging in neuropsychiatric systemic lupus erythematosus. Arthritis Rheum. 1999; 42(10): 2026–2038. PubMed Abstract \| Publisher Full Text\n\nMikdashi J, Handwerger B: Predictors of neuropsychiatric damage in systemic lupus erythematosus: data from the Maryland lupus cohort. Rheumatology (Oxford). 2004; 43(12): 1555–1560. PubMed Abstract \| Publisher Full Text\n\nWardlaw JM, Smith EE, Biessels GJ, et al.: Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol. 2013; 12(8): 822–838. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWahlund LO, Barkhof F, Fazekas F, et al.: A new rating scale for age-related white matter changes applicable to MRI and CT. Stroke. 2001; 32(6): 1318–1322. PubMed Abstract \| Publisher Full Text\n\nRovaris M, Inglese M, Viti B, et al.: The contribution of fast-FLAIR MRI for lesion detection in the brain of patients with systemic autoimmune diseases. J Neurol. 2000; 247(1): 29–33. PubMed Abstract \| Publisher Full Text\n\nSanna G, Piga M, Terryberry JW, et al.: Central nervous system involvement in systemic lupus erythematosus: cerebral imaging and serological profile in patients with and without overt neuropsychiatric manifestations. Lupus. 2000; 9(8): 573–583. PubMed Abstract \| Publisher Full Text\n\nJennings JE, Sundgren PC, Attwood J, et al.: Value of MRI of the brain in patients with systemic lupus erythematosus and neurologic disturbance. Neuroradiology. 2004; 46(1): 15–21. PubMed Abstract \| Publisher Full Text\n\nAppenzeller S, Pike GB, Clarke AE: Magnetic resonance imaging in the evaluation of central nervous system manifestations in systemic lupus erythematosus. Clin Rev Allergy Immunol. 2008; 34(3): 361–366. PubMed Abstract \| Publisher Full Text\n\nKatsumata Y, Harigai M, Kawaguchi Y, et al.: Diagnostic reliability of magnetic resonance imaging for central nervous system syndromes in systemic lupus erythematosus: a prospective cohort study. BMC Musculoskelet Disord. 2010; 11: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAppenzeller S, Vasconcelos Faria A, Li LM, et al.: Quantitative magnetic resonance imaging analyses and clinical significance of hyperintense white matter lesions in systemic lupus erythematosus patients. Ann Neurol. 2008; 64(6): 635–643. PubMed Abstract \| Publisher Full Text\n\nCastellino G, Padovan M, Bortoluzzi A, et al.: Single photon emission computed tomography and magnetic resonance imaging evaluation in SLE patients with and without neuropsychiatric involvement. Rheumatology (Oxford). 2008; 47(3): 319–323. PubMed Abstract \| Publisher Full Text\n\nAiniala H, Dastidar P, Loukkola J, et al.: Cerebral MRI abnormalities and their association with neuropsychiatric manifestations in SLE: a population-based study. Scand J Rheumatol. 2005; 34(5): 376–382. PubMed Abstract \| Publisher Full Text\n\nBenedict RH, Shucard JL, Zivadinov R, et al.: Neuropsychological impairment in systemic lupus erythematosus: a comparison with multiple sclerosis. Neuropsychol Rev. 2008; 18(2): 149–166. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetri M, Naqibuddin M, Carson KA, et al.: Brain magnetic resonance imaging in newly diagnosed systemic lupus erythematosus. J Rheumatol. 2008; 35(12): 2348–54. PubMed Abstract \| Publisher Full Text\n\nSibbitt WL Jr, Schmidt PJ, Hart BL, et al.: Fluid Attenuated Inversion Recovery (FLAIR) imaging in neuropsychiatric systemic lupus erythematosus. J Rheumatol. 2003; 30(9): 1983–1989. PubMed Abstract\n\nPodrazilová L, Peterová V, Olejárová M, et al.: Magnetic resonance volumetry of pathological brain foci in patients with systemic lupus erythematosus. Clin Exp Rheumatol. 2008; 26(4): 604–610. PubMed Abstract\n\nWest SG, Emlen W, Wener MH, et al.: Neuropsychiatric lupus erythematosus: a 10-year prospective study on the value of diagnostic tests. Am J Med. 1995; 99(2): 153–163. PubMed Abstract \| Publisher Full Text\n\nKarassa FB, Ioannidis JP, Boki KA, et al.: Predictors of clinical outcome and radiologic progression in patients with neuropsychiatric manifestations of systemic lupus erythematosus. Am J Med. 2000; 109(8): 628–634. PubMed Abstract \| Publisher Full Text\n\nDoubal FN, Dennis MS, Wardlaw JM: Characteristics of patients with minor ischaemic strokes and negative MRI: a cross-sectional study. J Neurol Neurosurg Psychiatry. 2011; 82(5): 540–542. PubMed Abstract \| Publisher Full Text\n\nCordonnier C, Al-Shahi Salman R, Wardlaw J: Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain. 2007; 130(Pt 8): 1988–2003. PubMed Abstract \| Publisher Full Text\n\nGreenberg SM, Vernooij MW, Cordonnier C, et al.: Cerebral microbleeds: a guide to detection and interpretation. Lancet Neurol. 2009; 8(2): 165–174. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeCarli C, Massaro J, Harvey D, et al.: Measures of brain morphology and infarction in the framingham heart study: establishing what is normal. Neurobiol Aging. 2005; 26(4): 491–510. PubMed Abstract \| Publisher Full Text\n\nCarne RP, Vogrin S, Litewka L, et al.: Cerebral cortex: an MRI-based study of volume and variance with age and sex. J Clin Neurosci. 2006; 13(1): 60–72. PubMed Abstract \| Publisher Full Text\n\nScheltens P, Pasquier F, Weerts JG, et al.: Qualitative assessment of cerebral atrophy on MRI: inter- and intra-observer reproducibility in dementia and normal aging. Eur Neurol. 1997; 37(2): 95–99. PubMed Abstract \| Publisher Full Text\n\nZanardi VA, Magna LA, Costallat LT: Cerebral atrophy related to corticotherapy in systemic lupus erythematosus (SLE). Clin Rheumatol. 2001; 20(4): 245–250. PubMed Abstract \| Publisher Full Text\n\nChinn RJ, Wilkinson ID, Hall-Craggs MA, et al.: Magnetic resonance imaging of the brain and cerebral proton spectroscopy in patients with systemic lupus erythematosus. Arthritis Rheum. 1997; 40(1): 36–46. PubMed Abstract \| Publisher Full Text\n\nZivadinov R, Shucard JL, Hussein S, et al.: Multimodal imaging in systemic lupus erythematosus patients with diffuse neuropsychiatric involvement. Lupus. 2013; 22(7): 675–683. PubMed Abstract \| Publisher Full Text\n\nLauvsnes MB, Omdal R: Systemic lupus erythematosus, the brain, and anti-NR2 antibodies. J Neurol. 2012; 259(4): 622–629. PubMed Abstract \| Publisher Full Text\n\nMarcinko K, Parsons T, Lerch JP, et al.: Effects of prolonged treatment with memantine in the MRL model of CNS lupus. Clin Exp Neuroimmunol. 2012; 3(3): 116–128. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuscal E, Traipe E, de Guzman MM, et al.: Cerebral and cerebellar volume loss in children and adolescents with systemic lupus erythematosus: a review of clinically acquired brain magnetic resonance imaging. J Rheumatol. 2010; 37(8): 1768–1775. PubMed Abstract \| Publisher Full Text\n\nSled JG, Spring S, van Eede M, et al.: Time course and nature of brain atrophy in the MRL mouse model of central nervous system lupus. Arthritis Rheum. 2009; 60(6): 1764–1774. PubMed Abstract \| Publisher Full Text\n\nBöckle BC, Jara D, Aichhorn K, et al.: Cerebral large vessel vasculitis in systemic lupus erythematosus. Lupus. 2014; 23(13): 1417–1421. PubMed Abstract \| Publisher Full Text\n\nKato R, Sumitomo S, Kawahata K, et al.: Successful treatment of cerebral large vessel vasculitis in systemic lupus erythematosus with intravenous pulse cyclophosphamide. Lupus. 2015; pii: 0961203315570163. PubMed Abstract \| Publisher Full Text\n\nAbda EA, Selim ZI, Radwan ME, et al.: Markers of acute neuropsychiatric systemic lupus erythematosus: a multidisciplinary evaluation. Rheumatol Int. 2013; 33(5): 1243–1253. PubMed Abstract \| Publisher Full Text\n\nKizu H, Dobashi H, Kameda T, et al.: Improvement of irregularity of brain vessel walls in systemic lupus erythematosus by tacrolimus. Clin Rheumatol. 2011; 30(5): 715–718. PubMed Abstract \| Publisher Full Text\n\nEspinosa G, Mendizábal A, Mínguez S, et al.: Transverse myelitis affecting more than 4 spinal segments associated with systemic lupus erythematosus: clinical, immunological, and radiological characteristics of 22 patients. Semin Arthritis Rheum. 2010; 39(4): 246–256. PubMed Abstract \| Publisher Full Text\n\nSteup-Beekman GM, Zirkzee EJ, Cohen D, et al.: Neuropsychiatric manifestations in patients with systemic lupus erythematosus: epidemiology and radiology pointing to an immune-mediated cause. Ann Rheum Dis. 2013; 72(Suppl 2): ii76–ii79. PubMed Abstract \| Publisher Full Text\n\nZimny A, Szmyrka-Kaczmarek M, Szewczyk P, et al.: In vivo evaluation of brain damage in the course of systemic lupus erythematosus using magnetic resonance spectroscopy, perfusion-weighted and diffusion-tensor imaging. Lupus. 2014; 23(1): 10–19. PubMed Abstract \| Publisher Full Text\n\nHughes M, Sundgren PC, Fan X, et al.: Diffusion tensor imaging in patients with acute onset of neuropsychiatric systemic lupus erythematosus: a prospective study of apparent diffusion coefficient, fractional anisotropy values, and eigenvalues in different regions of the brain. Acta Radiol. 2007; 48(2): 213–222. PubMed Abstract \| Publisher Full Text\n\nAppenzeller S, Bonilha L, Rio PA, et al.: Longitudinal analysis of gray and white matter loss in patients with systemic lupus erythematosus. Neuroimage. 2007; 34(2): 694–701. PubMed Abstract \| Publisher Full Text\n\nBosma GP, Steens SC, Petropoulos H, et al.: Multisequence magnetic resonance imaging study of neuropsychiatric systemic lupus erythematosus. Arthritis Rheum. 2004; 50(10): 3195–3202. PubMed Abstract \| Publisher Full Text\n\nEmmer BJ, Steup-Beekman GM, Steens SC, et al.: Correlation of magnetization transfer ratio histogram parameters with neuropsychiatric systemic lupus erythematosus criteria and proton magnetic resonance spectroscopy: association of magnetization transfer ratio peak height with neuronal and cognitive dysfunction. Arthritis Rheum. 2008; 58(5): 1451–1457. PubMed Abstract \| Publisher Full Text\n\nFilley CM, Kozora E, Brown MS, et al.: White matter microstructure and cognition in non-neuropsychiatric systemic lupus erythematosus. Cogn Behav Neurol. 2009; 22(1): 38–44. PubMed Abstract \| Publisher Full Text\n\nXu J, Cheng Y, Chai P, et al.: White-matter volume reduction and the protective effect of immunosuppressive therapy in systemic lupus erythematosus patients with normal appearance by conventional magnetic resonance imaging. J Rheumatol. 2010; 37(5): 974–986. PubMed Abstract \| Publisher Full Text\n\nZivadinov R, Havrdová E, Bergsland N, et al.: Thalamic atrophy is associated with development of clinically definite multiple sclerosis. Radiology. 2013; 268(3): 831–841. PubMed Abstract \| Publisher Full Text\n\nAppenzeller S, Carnevalle AD, Li LM, et al.: Hippocampal atrophy in systemic lupus erythematosus. Ann Rheum Dis. 2006; 65(12): 1585–1589. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJung RE, Segall JM, Grazioplene RG, et al.: Cortical thickness and subcortical gray matter reductions in neuropsychiatric systemic lupus erythematosus. PLoS One. 2010; 5(3): e9302. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScully M, Anderson B, Lane T, et al.: An Automated Method for Segmenting White Matter Lesions through Multi-Level Morphometric Feature Classification with Application to Lupus. Front Hum Neurosci. 2010; 4: 27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPierpaoli C, Jezzard P, Basser PJ, et al.: Diffusion tensor MR imaging of the human brain. Radiology. 1996; 201(3): 637–648. PubMed Abstract \| Publisher Full Text\n\nLe Bihan D: Looking into the functional architecture of the brain with diffusion MRI. Nat Rev Neurosci. 2003; 4(6): 469–480. PubMed Abstract \| Publisher Full Text\n\nSchmidt-Wilcke T, Cagnoli P, Wang P, et al.: Diminished white matter integrity in patients with systemic lupus erythematosus. Neuroimage Clin. 2014; 5: 291–297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKozora E, Filley CM: Cognitive dysfunction and white matter abnormalities in systemic lupus erythematosus. J Int Neuropsychol Soc. 2011: 1–8. PubMed Abstract \| Publisher Full Text\n\nZhang L, Harrison M, Heier LA, et al.: Diffusion changes in patients with systemic lupus erythematosus. Magn Reson Imaging. 2007; 25(3): 399–405. PubMed Abstract \| Publisher Full Text\n\nLee SP, Wu CS, Hsieh LC, et al.: Efficacy of magnetic resonance diffusion tensor imaging and three-dimensional fiber tractography in the detection of clinical manifestations of central nervous system lupus. Magn Reson Imaging. 2014; 32(5): 598–603. PubMed Abstract \| Publisher Full Text\n\nJung RE, Caprihan A, Chavez RS, et al.: Diffusion tensor imaging in neuropsychiatric systemic lupus erythematosus. BMC Neurol. 2010; 10: 65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEmmer BJ, Veer IM, Steup-Beekman GM, et al.: Tract-based spatial statistics on diffusion tensor imaging in systemic lupus erythematosus reveals localized involvement of white matter tracts. Arthritis Rheum. 2010; 62(12): 3716–3721. PubMed Abstract \| Publisher Full Text\n\nJones JT, DiFrancesco M, Zaal AI, et al.: Childhood-onset lupus with clinical neurocognitive dysfunction shows lower streamline density and pairwise connectivity on diffusion tensor imaging. Lupus. 2015; pii: 0961203315572718. PubMed Abstract \| Publisher Full Text\n\nBosma GP, Rood MJ, Huizinga TW, et al.: Detection of cerebral involvement in patients with active neuropsychiatric systemic lupus erythematosus by the use of volumetric magnetization transfer imaging. Arthritis Rheum. 2000; 43(11): 2428–2436. PubMed Abstract \| Publisher Full Text\n\nCagnoli P, Harris RE, Frechtling D, et al.: Reduced Insular Glutamine and N-acetylaspartate in systemic lupus erythematosus: a single-voxel (1)H-MR spectroscopy study. Acad Radiol. 2013; 20(10): 1286–1296. PubMed Abstract \| Publisher Full Text\n\nGuillen-Del Castillo A, Alonso J, Martínez-Valle F, et al.: Increased myo-inositol in parietal white and gray matter as a biomarker of poor prognosis in neuropsychiatric lupus: a case report. Lupus. 2014; 23(10): 1073–1078. PubMed Abstract \| Publisher Full Text\n\nSarbu N, López-Rueda A, Chirife O, et al.: CT-perfusion time-maps likely disclose the earliest imaging signs of posterior reversible encephalopathy syndrome (PRES). J Neuroradiol. 2014; 41(2): 147–9. PubMed Abstract \| Publisher Full Text\n\nEmmer BJ, van Osch MJ, Wu O, et al.: Perfusion MRI in neuro-psychiatric systemic lupus erthemathosus. J Magn Reson Imaging. 2010; 32(2): 283–288. PubMed Abstract \| Publisher Full Text\n\nWang PI, Cagnoli PC, McCune WJ, et al.: Perfusion-weighted MR imaging in cerebral lupus erythematosus. Acad Radiol. 2012; 19(8): 965–970. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9358", "date": "07 Jul 2015", "name": "Fabiola Atzeni", "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\nTitle and Abstract: The title is quite generic and perhaps the description of old and new techniques should be emphasized.Article content: Design, methods and analysis are well explained and appropriate, however the use of nanoparticles (e.g. SPIONS) should be mentioned among the new MRI techniques.\n\nConclusions:  The conclusions underline the need for MRI imaging in symptomatic neuropsychiatric SLE, although a standardized consensus still lacks. It should be more evidenced, however, the utility of a complex diagnostic algorithm, including neurophysiologic study of the brain or lab tests, beyond the solely use of MRI.Data: Methods and indications to a specific procedure are well described ; MRI findings are reported according to pathologic or clinical pictures.", "responses": [] }, { "id": "9530", "date": "16 Jul 2015", "name": "Murray B. Urowitz", "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\nSarbu et al. review the utility of MRI imaging in neuropsychiatric lupus (NPL) both what is available today in routine care and for research purposes and what might become useful in clinical care in the future. In NPL clinicians are interested in three major questions. First, is there evidence of NPL at any time in the past; secondly, is there evidence of active NPL currently; and thirdly can you demonstrate change over time with or without treatment in NPL.In terms of the first question Sarbu et al. review the conventional MRI findings in the cortical, subcortical and spinal cord regions. However the nature of the morphologic changes is non specific and many could be compatible with hypertension, amyloid, Alzheimers etc. and thus may not be helpful to the clinician. Even the advanced techniques (not universally available) measuring brain volume and composition are not specific for SLE.In terms of the second question Sarbu et al. review Magnetic Resonance Spectroscopy (MRS) which allows the analysis of brain metabolites. The demonstration of the uptake of specific metabolites might in the future be associated with neuron excitation but no studies are available in NPL.Perfusion imaging such as also done with SPECT and PET scanning can reveal areas of hyper or hypoperfusion but thus far no studies demonstrating active NPL which could change over time exists.In conclusion, this article is a good review of what currently exists and what may be developed over time with MRI imaging. The authors conclude that further clinical association studies with conventional and advanced techniques are required before MRI imaging will become the standard to understand the mechanism, diagnosis and monitoring of NPL.", "responses": [] }, { "id": "9164", "date": "17 Jul 2015", "name": "Gian Domenico Sebastiani", "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 review is well written and give us a comprehensive updating on magnetic resonance imaging in NPSLE. Only few minor suggestions:Introduction - Authors say that \"reliable methods for diagnosing NPSLE are also unknown...\". This is essentially true, but they could quote the recent contribution by Bortoluzzi et al. (2015), dealing with the development of a new algorithm for attribution of NP events in SLE. Conventional MRI, page 2 - \"DWI sequences\". Please specify the meaning of DWI. Conventional MRI, page 3 - \"total corticosteroid\". Cumulative is better.", "responses": [] } ]	1	https://f1000research.com/articles/4-162
https://f1000research.com/articles/2-71/v1	04 Mar 13	{ "type": "Research Article", "title": "Using Akaike’s information theoretic criterion in population analysis: a simulation study", "authors": [ "Erik Olofsen", "Albert Dahan", "Albert Dahan" ], "abstract": "Akaike’s information-theoretic criterion for model discrimination (AIC) is often stated to “overfit”, i.e., it selects models with a higher dimension than the dimension of the model that generated the data. However, when no fixed-dimensional correct model exists, for example for pharmacokinetic data, AIC, or its bias-corrected version (AICc) might be the selection criterion of choice if the objective is to minimize prediction error. The present simulation study was designed to assess the behavior of AICc when applying it to the analysis of population data, for various degrees of interindividual variability. The simulation study showed that, at least in a relatively simple mixed effects modeling context, minimal mean AICc corresponded to best predictive performance even in the presence of large interindividual variability.", "keywords": [ "population model", "pharmacokinetics", "Akaike", "information theoretic criterion" ], "content": "Introduction\n\nPopulation data consist of one or more measurements in two or more individuals. Such data can be characterized by mixed-effects models, where the mixed effects consist of fixed and random effects. Fixed effects are, for example, the times at which the measurements are obtained, and covariates such as demographic characteristics of the individuals. When mixed-effects models are fitted to population data, the question arises how many of those effects should be incorporated in the model. This is the so-called problem of variable selection1.\n\nOne strategy is to observe the change in goodness-of-fit by adding one more parameter and testing the significance of that change2. In the maximum likelihood approach, the objective function value (OFV), being the minus two logarithm of the likelihood function, is minimized. To attain a p-value of e.g., 0.05 or less, the decrease in OFV, when adding one parameter, should be 3.84 or more2.\n\nAnother strategy is to apply Akaike's information theoretic criterion (AIC), which can be written as\n\n\n\nwhere D is the number of parameters in the model1–4. The model with the lowest value of AIC is considered the best one. In the case of just adding one parameter, the OFV needs to decrease only 2 points or more to be incorporated in the model, so the associated p-value > 0.05 seems too high to justify this strategy.\n\nWhen additional model parameters are incorporated, the significance of one model parameter might change, but the interpretation of AIC does not4. However, when multiple significance tests are performed, the significance level of each individual test should be corrected to a lower value, so a decrease of 2 points for one parameter does again seem to be too low.\n\nEven if the strategy of using AIC leads to optimal variable selection, the question arises if this is also the case when using mixed-effects models. In theory, the model that is best according to AIC is the one that minimizes prediction error3,5; and this is also true for a mixed effects model when predicting data for individuals for which no data have been obtained so far5.\n\nIn the literature, simulation studies have assessed the performance of AIC in selecting the model with the lowest prediction error, but to our knowledge these were never done for population data. In this article, we will define a toy pharmacokinetic model and observe the performance of AIC when adding fixed effects to this model, as well as when adding interindividual variability.\n\n\nMethods\n\nConsider the following function y(t), an infinite sum of exponentials, and its relationship with a (negative) power of time6:\n\n\n\nFigure 1A shows that this function looks like a typical pharmacokinetic profile after bolus administration. This model is to be regarded as a toy model, because we do not expect it to adequately describe pharmacokinetic data, although variations of power functions of time have been shown to fit pharmacokinetic data well6. Here we will use the fact that if we approximate y(t) = 1/t by the following sum of M exponentials with K nonzero coefficients α and M fixed parameters λ (as chosen in the next subsections):\n\n\n\nthat with M time instants tj, we would need no less than K = M exponentials to obtain a perfect fit. Moreover, with noisy data, it might be that for K < M an optimal fit is obtained in the sense that the associated prediction error of the model is minimal. Figure 1B shows how eleven (in this case error-free) samples from this function can be approximated by sums of exponentials.\n\nA: function y(t) = 1/t, and B: approximations obtained by fitting six and three exponentials to the depicted eleven samples. Note the log-lin and log-log scales for panels A and B, respectively. Time has arbitrary units.\n\nIn the following, the time instants tj, j = 1, …, M, centered around 1, were chosen within [1/tmax,tmax] according to\n\n\n\nwith γ = log(tmax)/log(M); tmax was set to 100 (see the time axis of Figure 1B for an example with M = 11). Simulated data were generated via\n\n\n\nwhere εj denotes Gaussian measurement noise with variance σ2. The M time constants λ were fixed according to λm = 1/tm, m = 1, …, M. In this setting the model eq. (3) can be fitted to simulated data using weighted linear least squares regression, with weight factors ω(tj) = 1/tj (note that no precaution is needed against ε ≤ –1).\n\nPopulation data consisting of N individuals were simulated via\n\n\n\nwhere ηi denotes interindividual variability with variance ω2. The nonlinear mixed effects model for the population data was then written as:\n\n\n\nNote that with N > 1, a perfect fit is no longer obtained with K = M, because the εi,j are generally different for different i (individuals).\n\nSimulation data were generated via eq. (6), with random generators in R7. Model fitting was also done in R, with function \"lm()\" from package \"stats\", except for nonlinear mixed-effects model fitting for simulated data with ω2 > 0, which was done in NONMEM version 7.3 (beta version a6.5)8. Parameters α (see eq. (7)) were not constrained to be positive, so that it was not possible for parameters to become essentially fixed to zero, reducing the dimensionality of the model. Prediction error (ν2) was calculated with\n\n\n\nusing predictions based on eq. (7) with the random effects ηi = 0, and validation data zi(tj) also generated via eq. (6), but with different realizations of εij and ηi. The objective function OFV was also calculated at the estimated parameters using the validation data, denoted OFVv, which should on average be approximately equal to Akaike's criterion (see Supplementary material). OFVv was compared with AIC and also with Akaike's criterion with a correction for small sample sizes (AICc)4:\n\n\n\nThe above criteria were normalized by dividing them by the number of observations, and averaged over 1000 runs (unless otherwise stated; and runs where NONMEM's minimization was not successful were excluded). For plotting purposes, 95% confidence intervals or confidence regions for means were determined using R's packages \"gplots\" and \"car\", under the assumption that averages over 1000 variables are normally distributed.\n\nSimulation parameters M and σ2 are expected to determine the number of exponentials K; if M increases and/or σ2 decreases, K will increase. Without inter-individual variance, so ω2 = 0, the information in the data increases as N increases, so that K is also expected to increase. With N = 2, M = 11 and σ2 = 0.5, pilot simulations indicated a K ≈ 4. When ω2 > 0, the prediction error will increase, but it is less easy to predict what its effect will be on K. For ω2 values of 0, 0.1, and 0.5 were selected - values that are encountered in practice. Because there is only one random effect in the mixed effects model, the relatively low number of individuals N = 5 was selected.\n\nFor a certain choice of M, there are 2M – 1 possible combinations of λs to choose for the terms exp(–λmtj) in the sum of exponentials (excluding the case of zero exponentials). Because accurate evaluation of all models at different parameter values is not feasible with respect to computer time, the set of possible combinations was reduced to one with evenly spaced λs. Table 1 gives an example for the case M = 11.\n\n\nResults\n\nFigure 2 shows the averaged prediction error versus number of exponentials for all possible choices of λ, with N = 2, M = 11, σ2 = 0.5, and ω2 = 0. From the figure it is clear that prediction error may indeed increase if the number of exponentials selected is too large. The bigger solid circles correspond to the models chosen in Table 1; in general the evenly spaced selection of exponents resulted in models with the smallest prediction error.\n\nMean squared prediction error ν2 (eq. (8)) as a function of the number of exponentials, with 2047 models, averaged over 100 runs, N = 2, M = 11, σ2 = 0.5, ω2 = 0. The dashed line represents the prediction error from the true model, so that ν2 = σ2. The bigger solid circles correspond to the models chosen in Table 1.\n\nFigure 3 shows simulation results using the model set defined in Table 1, starting from K = 4, with parameters N = 5, M = 11, σ2 = 0.5, and ω2 = 0. The model with K = 6 exponentials had both minimal mean OFVv and minimal mean AICc (and minimal prediction error v2 (not shown)). With N = 5, M = 11, there are still visible differences between AICc and AIC; although AIC would in this case also select the optimal model, AIC appears to favor more complex models. Note that the sizes of the confidence intervals and confidence regions can be made arbitrarily small by choosing the number of runs higher than the selected number of 1000 (at the expense of computer time).\n\nMean OFVv as a function minus of two log likelihood (-2LL), the number of exponentials, AIC and AICc (top four panels), and AIC and AICc as a function of the number of exponentials (lower two panels), averaged over 1000 runs, N = 5, M = 11, σ2 = 0.5, ω2 = 0. The dashed lines represent the theoretical values for an infinite amount of data (see Supplementary material). Error bars and ellipses denote 95% confidence intervals and confidence regions, respectively. Each solid line in the middle panels denotes the line of identity.\n\nFigure 4 shows simulation results with ω2 = 0.1; mixed-effects analysis was used to fit the population data. The main difference with the results of data with ω2 = 0 is the overall increase in OFVv and AICc. The optimal number of exponentials remained K = 6.\n\nFigure 5 shows simulation results with ω2 set at the higher value of 0.5. The main differences with the results of data with ω2 = 0.1 are again the overall increase in OFVv, AICc and prediction error, and also in the variability in the prediction error. The optimal number of exponentials remained K = 6, although AICc begins to favor the models with larger K (a simulation with N increased to 7, both OFVv and AICc favored larger models; data not shown).\n\n\nDiscussion\n\nWith the objective of creating a simulation context resembling pharmacokinetic analysis where concentration data are approximated by a sum of exponentials, the toy model y(t) = 1/t was chosen. In this setting, reality - the reality of the toy model - is always underfitted. When mixed effects models were fitted to simulated data, mean AICc was approximately equal to the validation criterion mean OFVv, and their minima coincided. With large interindividual variability, mean expected prediction error (ν2, see eq. (8), with random effects fixed to zero), was less discriminative between models, so that it became less suitable as a validation criterion.\n\nVaida and Blanchard proposed a conditional Akaike information criterion to be used in model selection for the \"cluster focus\"5. It is important to stress that the cluster focus as they defined is the situation where data are to be predicted of a cluster that was also used to build the predictive model. In that case, the random effects have been estimated, and then the question arises how many parameters that required. In our case, a cluster is the data from an individual; AIC was used in the situation of predicting population data consisting of individual data that were not used to build the model. This would seem to be the most common situation in clinical practice. Furthermore, AIC for the population focus is asymptotically equivalent with leave-one-individual-out cross-validation; AIC for the individual focus with leave-one-observation-out cross-validation9.\n\nWe chose to perform simulations using the model given by eq. (2) because approximating data with a sum of exponentials is daily practice in pharmacokinetic analysis where data are obtained from \"infinitely complex\" systems, and we cannot hope to find the \"correct\" model. The Bayesian information criterion (BIC) is consistent in the sense that it selects the correct model, given an infinite amount of data4. The reason that AIC can be used in \"real-life\" problems is that as the amount of data goes to infinity, the complexity, or dimension, of the model that should be applied should also go infinity10. Burnham and Anderson show that it is possible to choose the prior probability distribution for BIC in such a way that it incorporates the knowledge that more complex models should be favored if the amount of data increases, and so that the BIC \"reduces\" to AIC4,10. In the situation that the correct model set belongs to the set of evaluated models, a selection criterion that both finds the correct model and minimizes prediction error would be preferable - but Yang concluded that this may not be possible11.\n\nIn pharmacokinetic analysis, it may not really be appropriate to test (using a hypothesis test assuming a X2 distribution for the objective function) whether an added exponential is statistically significant12. Here the hypothesis H0: the data originate from a K-exponential model (and HA: the data originate from a higher dimensional model) is almost certain to be false. Furthermore, when taking a low p-value, it is also almost certain that the model selected has worse predictive properties. If a model is to be applied in clinical practice, for example for drug administration in a patient never studied before, the model should be as predictive as possible. However, it may be sensible to test whether a certain fixed effect has both a clinically and statistically significant effect, if it is costly to reach a false conclusion, for example in case of increased risks for patients, or in the field of drug development.\n\nIntuitively, predicting data for an individual that cannot be \"individualized\" seems problematic because the data are predicted using a random effect ηi set to zero, instead of the value fitting for that individual. However, AIC is related to the expected model output; and for individual data not used in building the predictive model, the expected model of output is obtained with mixed effects set to zero, although nonlinearities may bias expectation - but this is also true for nonlinear models without mixed effects.\n\nFurthermore, it should be noted that minimizing AIC has a more general interpretation, namely optimally capturing the information contained in the data4. Independent or future population data z are not just predicted by ŷ; also the distributions of the expected random effects ε and η are characterized by σ̂2 and ω̂2. That is why OFVv is the criterion to be used to assess the predictive performance of a model.\n\nThe simulated data were analyzed using weighted (non)linear regression, see eq. (6), where measurement noise was weighted according to the exact function value. In practice, when the weights are unknown, a choice must be made to weight the data according to the measurements or to the model output, depending on which is likely to be the most accurate. To match the latter case, simulated data should be generated (cf. eq. (6)) via\n\n\n\nThe likelihood function and AIC are both still well-defined if the model output ŷi(tj) ≠ 0. Prediction errors are to be calculated with\n\n\n\nwhere ŷ possibly becomes arbitrarily close to zero for less than optimal models, and v2 may be based on long-tailed distributed numbers. To be able to compare prediction errors from different models, the weight factors could be chosen identical for all K to the model output of the largest model - see the Supplementary material for further analysis.\n\nWe recognize the following limitations of our study:\n\nThe model contained only one random effect, and therefore the number of random effect (co)variances was fixed to one. While the number of (co)variance parameters should be counted as ordinary parameters5, at least in well behaved situations13, we did not investigate the process of optimizing this part of a random effects model.\n\nThe nonlinearity in the mixed-effects model was simply due to a multiplicative factor exp(η) in the model output. Usually, random effects in pharmacokinetic models have more complex influence on the model output. However, the lognormal nature of exp(η) is a characteristic property of both our toy model and general pharmacokinetic models.\n\nThe characteristics of the exponentials incorporated in the regression models were evenly spaced, and the values of the rate constants λ were fixed. We expect that with more freedom in the specification of the set of models, prediction errors with overfitted models may be worse. However, the agreement between AICc and prediction error should persist.\n\nWe did not evaluate all possible models within their definition, but only those listed in Table 1, and it makes sense to limit the model set to avoid overfitting the data4,11. We did not address how to optimally select the rate constants λ. Stepwise selection methods have their disadvantages12. With stepwise forward selection, AICc may even perform worse than AIC14.\n\nWe did not evaluate the process of covariate selection. However, the set of exponentials may be viewed as a number of (somewhat correlated) predictors. It is therefore expected that the present findings also hold for other types of covariates.\n\n\nConclusion\n\nIn conclusion, the present simulation study demonstrated that in the presence of inter-individual variability in a relatively simple mixed effects modeling context, minimum mean AICc coincided with best predictive performance.", "appendix": "Author contributions\n\nEO performed the numerical analyses, and EO and AD contributed to the interpretation of the results and the preparation of the manuscript; both authors have agreed to its final content.\n\n\nCompeting interests\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by institutional resources.\n\n\nAcknowledgments\n\nThe authors would like to thank J. de Goede for many fruitful discussions.\n\n\nSupplementary material\n\nIn the following, we summarize theory on the maximum likelihood approach and AIC relevant to this paper. Suppose the model for measured data yj, j = 1, …, M is given by (cf. eq. (5), eq. (6), and eq. (10))\n\n\n\nwhere ŷj is the model output, wj are weight factors, and εj are independent normally distributed with mean zero and variance σ2. The likelihood function L for this data set is then given by\n\n\n\nwhere the set of parameters θ contains σ2 and those needed to calculate ŷ. The objective function value (OFV) is defined as minus two times the natural logarithm of the likelihood:\n\n\n\nNote that in writing \"OFV\", the data and parameters it depends on have been omitted. Now maximum likelihood is obtained when OFV is minimal; constant terms such as M log(2π) may then be discarded (for example, in NONMEM's calculation of the the objective function). The minimum is attained for certain values of parameters of ŷ, and for the parameter value of σ2, when the derivative of OFV with respect to that parameter is zero:\n\n\n\nso the maximum likelihood estimator of σ2 is\n\n\n\nBy subsituting this estimate in eq. (14), we obtain\n\n\n\nBy substituting this result in eq. (1), we have\n\n\n\nThe term 2D arises from the fact that in minimizing the Kullback-Leibler information, i.e., a measure of the distance between reality and the best approximating model, expectations have to be taken over a data space leading to estimates of parameters θ (and hence ŷ, and possibly ω (see below)) and over a second independent data space y4. So AIC as defined above should on average be approximately equal the value of OFV eq. (14), with estimated values for the parameters and validation data zj, denoted OFVv:\n\n\n\nSo when OFV and AIC are both minimized, the latter term - the sum of squared weighted prediction errors - should also be minimal. For the plots in this paper, the measures OFV, OFVv, AIC, and AICc, were normalized by dividing them by the number of data samples. With an infinite amount of data, and σ̂2 = σ2, the normalized criteria should attain the value of log(σ2 )+log(2π)+1.\n\nNote that if the weights wj are taken as stated in subsection \"Data simulation\", the term ∑log(wj2) vanishes (this is a just a curiosity of that choice of weights); if the wj are taken as the measurements yj, the expectation of this term is the same for every K (for every model considered here). However, if the weights are taken as the model output ŷj, the expectation of the term will not vanish for a less than perfect model, and will differ between different models. To compare their v2, the weights for all models could be fixed to the model output of the best model - but since that is unknown at this point - to the output of the largest model.\n\nFor population data, the likelihood function is the product across individual marginal likelihoods where the random effects η contained in eq. (13), when ŷ is given by eq. (6), have been integrated out. Usually, these integrals need to be numerically approximated, e.g., as is done here, by NONMEM. So the context of AIC is then also the one where the ηs have been integrated out (but with the parameters at their estimated values), which is to be done when all data are acquired. So while the characteristics of the set of (validation) data are optimally captured, this context is different from the case where prediction errors are calculated with the random effects set to zero instead of being integrated out. In that case, the above AIC and OFVv criteria do not match, as the components of the likelihood in eq. (13) are no longer independent (they can only be independent if the true values of η for the individuals are also zero). Note however, that from the higher perspective of optimally characterizing a future set of population data, this is a less important case.\n\nFinally, it should be noted that the parameter estimates may not be consistent (i.e., do not converge to their true values when the amount of data goes to infinity if the ŷj do not properly account for heteroscedasticity)15. In the derivation of AIC4, it is only required that the likelihood function is maximized; consistency is not required.\n\n\nReferences\n\nHastie T, Tibshirani RJ, Friedman JH: The elements of statistical learning. Data mining, inference, and prediction (2nd ed), Springer, New York (2009).\n\nBonate PL: Pharmacokinetic-pharmacodynamic modeling and simulation. (2nd ed.), Springer, New York (2011).\n\nAkaike H: A new look at the statistical model identification. IEEE Trans Automat Contr. (1974); 19: 716–23.\n\nBurnham KP, Anderson DR: Model selection and multimodel inference. (2nd ed.), Springer, New York (2002); 488.\n\nVaida F, Blanchard S: Conditional Akaike information for mixed-effects models. Biometrika. (2005); 92: 351–70.\n\nNorwich KH: Noncompartmental models of whole-body clearance of tracers: A review. Ann Biomed Eng. (1997); 25: 421–39.\n\nR Development Core Team, R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria (2012) ISBN 3-900051-07-0.\n\nBeal SL, Sheiner LB, Boeckmann AJ, et al:NONMEM User’s Guides. Icon Development Solutions, Hanover, MD, USA (1989–2012).\n\nFang Y: Asymptotic equivalence between cross-validations and Akaike information criteria in mixed-effects models. J Data Sci. (2011); 9: 15–21.\n\nBurnham KP, Anderson DR: Multimodel inference: Understanding AIC and BIC in model selection. Sociol Meth Res. (2004); 33: 261–304.\n\nYang Y: Can the strengths of AIC and BIC be shared? A conflict between model identification and regression estimation. Biometrika. (2005); 92: 937–50.\n\nSteyerberg EW: Clinical prediction models. A practical approach to development, validating, and updating. Springer, New York (2009); 497.\n\nGreven S, Kneib T: On the behaviour of marginal and conditional AIC in linear mixed effects models. Biometrika. (2010); 97: 773–89.\n\nOlofsen E: The performance of model selection criteria in the absence of a fixed-dimensional correct model.(2007); Page 16, Abstr 1198.\n\nvan Houwelingen JC: Use and abuse of variance models in regression. Biometrics. (1988); 44: 1073–81." }	[ { "id": "809", "date": "05 Mar 2013", "name": "Paul Eilers", "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 an interesting and useful study, written in a relaxed style. I believe we need more studies like this, because there is a lot of folklore around AIC, mainly claiming that is generally over-fits.I have one major objection: the effective dimension (ED) of a mixed model is less than the number of parameters (D), because shrinking takes place. The easiest way to obtain ED is to compute the trace of the “hat” matrix. In linear models this is relatively easy. In non-linear models it is harder. What one needs is (in LaTeX notation) $h_{ii} = \\partial \\hat y_i/ \\partial y_i$. I can provide more details if needed. I don’t know if NONMEM can provide the needed quantities.", "responses": [ { "c_id": "377", "date": "05 Mar 2013", "name": "Erik Olofsen", "role": "Author Response", "response": "Thank you for your comments. Vaida and Blanchard (ref. 5 above) discuss two settings: model focus and cluster focus. In the former setting, the effective number of parameters equals the number of fixed effects parameters and variance components (p.354); in the latter setting, the effective number of parameters needs to be estimated in the way you outlined. The first setting, corresponding to the situation of predicting data of \"new\" subjects, is the one for which the study results should be valid." } ] }, { "id": "928", "date": "07 May 2013", "name": "Frank Harrell", "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\nThe title and abstract do not place the potentially useful study in the right context. 'Population' can have many meanings and the scope is too wide. Consider narrowing the implied scope to mixed effects PK modeling.AIC can be an excellent metric for selecting from among a very limited number of models. If used in a stepwise process it can result in all the severe problems that stepwise variable selection has. The authors need to be much more careful about multiplicity and model uncertainty. This needs to be carefully discussed, and the authors would add to the literature if they can derive the maximum number of models that can be compared with AIC before the method breaks down.", "responses": [] }, { "id": "1693", "date": "21 Oct 2013", "name": "Julie Bertrand", "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 title should indeed specify that this work focuses on pharmacokinetics (PK). However I must add that the model function considered is unusual enough that it seems difficult to extend their conclusions to a real PK study analysis.The abstract is too general and more details should be provided on the simulation study (model function, number of samples, number of subjects, number of random effects) and the results (differences between selection on OFV, AIC and AICc, impact of increasing the random effect variance).The whole methodology is very well described. But one aspect is missing, as underlined by the other reviewer: the (very direct here) link with the best sum of exponential model and the information in the design. I was not much surprised that K=6 (or 5) exponential got the best AIC when you have 11 evenly spaced samples and the candidate models all had evenly spaced rate constants. Also, why not investigate the performance of BIC (with log(N) and log(NxM)) ?Finally, the conclusions are balanced in the sense that the authors have rightly identified the limit of their exercise which is the generalization of their results to a real PK data analysis: only one random effect, no covariance parameters, only slope parameters, etc...", "responses": [] } ]	1	https://f1000research.com/articles/2-71
https://f1000research.com/articles/4-396/v1	27 Jul 15	{ "type": "Research Article", "title": "Emergence of spatial behavioral function and associated mossy fiber connectivity and c-Fos labeling patterns in the hippocampus of rats", "authors": [ "Rachel Comba", "Nicole Gervais", "Dave Mumby", "Matthew Holahan", "Rachel Comba", "Nicole Gervais", "Dave Mumby" ], "abstract": "Improvement on spatial tasks is observed during a late, postnatal developmental period (PND18 – PND24). The purpose of the current work was 1) to determine whether the emergence of spatial-behavioral function was based on the ability to generate appropriate behavioral output; 2) to assess whether mossy fiber connectivity patterns preceded the emergence of spatial-behavioral function; 3) to explore functional changes in the hippocampus to determine whether activity in hippocampal networks occurred in a training-dependent or developmentally-dependent fashion. To these ends, male, Long Evans rats were trained on a spatial water or dry maze task for one day (PND16, PND18 or PND20) then euthanized. Training on these 2 tasks with opposing behavioral demands (swimming versus exploration) was hypothesized to control for behavioral topology. Only at PND20 was there evidence of spatial-behavioral function for both tasks. Examination of synaptophysin staining in the CA3 region (i.e., mossy fiber projections) revealed enhanced connectivity patterns that preceded the emergence of spatial behavior. Analysis of c-Fos labeling (functional changes) revealed developmentally-dependent increases in c-Fos positive cells in the dentate gyrus, CA3 and CA1 regions whereas training-dependent increases were noted in the CA3 and CA1 regions for the water-maze trained groups. Results suggest that changes in mossy fiber connectivity in association with enhanced hippocampal functioning precede the emergence of spatial behavior observed at PND20. The combination of neuroanatomical and behavioural results confirms the hypothesis that this time represents a sensitive period for hippocampal development and modification and the emergence of spatial/ cognitive function.", "keywords": [ "water maze", "object exploration", "memory", "postnatal" ], "content": "Introduction\n\nOne brain region that shows connectivity-based changes during postnatal development is the hippocampus. Neurogenesis in the dentate gyrus (DG) peaks in density between postnatal day 15 (PND15; 7) and PND1812,18,77 and the mossy fibers (MF), forming connections between the DG granule cells and CA3 pyramidal cells, show a late, postnatal remodelling9,24,25,32. With increasing evidence that MFs in the adult rat regulate memory function by remodeling33,34,57–60, it is attractive to hypothesize that improved performance on a number of spatial tasks during development3,13,16,20,41,65,66,71,73 may be related to connectivity- or functional-based changes in the network properties of the DG-CA3 region.\n\nThe main purpose of the present study was to investigate the relationship between the emergence of spatial-behavioral function and the emergence of the neural substrates – both anatomical and functional – that would support optimal spatial-behavioral function. In this respect, one question related to MF connectivity-based developmental changes is whether these changes occur prior to the emergence of spatial behavior or as a result of spatial behavior. This is an issue as adult MF remodelling can occur within a 24-hour period following spatial water maze training33. Another related question is how the functional aspects of hippocampal neurons change during development and whether this too occurs in response to behavioral-based activity. Similar to the MF question is whether hippocampal network activity precedes or results from the emergence of spatial behavior. To investigate this, analysis of c-Fos positive cells in the hippocampal DG, CA3 and CA1 regions was carried out to examine training-associated hippocampal activity and activity associated with neural development.\n\nA final issue explored in the present study was whether improvements in spatial-behavioral performance reflect enhanced hippocampal structural and network activity or emergence of sensori-motor demands required for task performance11,17,67,75. During development, sensori-motor demands of a task could confound firm conclusions concerning the development of spatial-behavioral function (e.g., 30). To clarify this, rats were trained on the hidden-platform water-maze task (aka, Morris water maze task) or a dry maze version of the novelty-preference paradigm where rats are required to recognize that an object is in a place where there had not previously been an object26,54 for one day on PND16, PND18 or PND20. In this case, spatial representational requirements were similar for both tasks but behavioral topology requirements were different – swimming versus exploration. Comparing performance on these two tasks was hypothesized to determine whether the emergence of spatial-behavioral function was based on the ability to generate appropriate behavioral output or based on the hippocampal spatial navigation neural substrates.\n\n\nMaterials and Methods\n\nA total of 51 male Long-Evans rats (LER; Charles River, St. Constant, Quebec, Canada) were used. The day the pups were born was marked as postnatal day 0 (PND0). Pups remained with their dams throughout the duration of the study. Rats were housed in a temperature-controlled vivarium in polycarbonite cages with a 12 hour light-dark cycle. Food and water were provided ad libitum. All experiments were conducted in accordance with the Canadian Council on Animal Care (CCAC) guidelines and specific protocols approved by the Carleton University Animal Care Committee and the Concordia University Animal Care and Use Committee (Animal Use Protocol Number P13-10).\n\nWater maze. The water maze was a white, circular, polypropylene pool measuring 124 cm diameter, 31 cm height and filled with water to a depth of 25 cm. Water temperature averaged 23°C and was rendered opaque by the addition of non-toxic white paint. The platform was made of clear Plexiglas measuring 11 cm diameter and was submerged 2.0 cm below the water surface. Distal visual cues were present on the walls of the room surrounding the maze. Rat movement in the pool was tracked using the HVS Image 2100 Tracking System (version 1/09; HVS Image, Buckingham, UK). The water was skimmed and stirred after each trial and the pool was drained and refilled every day.\n\nDry maze. The dry maze was 122 cm in diameter with Lexan walls and floor measuring 10.5 cm high, 2 mm thick. The floor of the maze was opaque white and the walls were transparent. The two stimulus objects were plastic turquoise water bottles 20 cm high with the widest circumference measuring approximately 23 cm. A small glass jar (7 cm high, 20 cm circumference) was fastened to the bottom of each object with epoxy. The jar lids were inverted and fastened to the floor to allow the objects to be secured in place by screwing the jars into the lids. Holes were drilled at the center of each quadrant to accommodate the lids and to allow counterbalancing of object positions. When not in use, the holes were covered with a small piece of circular white tape. After each trial, the objects and maze were wiped down with water. At the beginning of each day, the floor of the maze and objects were washed with a 70% alcohol solution and the Lexan walls were cleaned with standard window cleaner.\n\nMorris Water Maze (MWM). For the Morris water maze (MWM) portion, there were three groups with five rats per group (n = 15). Rats underwent fixed hidden-platform water-maze training consisting of 8 trials with a 1-minute inter-trial interval to locate the hidden platform. Training was carried out on one day only at PND16, PND18 or PND20. Each trial began from a different start point on the perimeter of the pool. If a rat did not reach the platform within 60 seconds, they were guided to it. Rats remained on the platform for 30 seconds before being removed, dried, and placed into a holding cage for an additional 30 seconds. Following the final trial, each rat was dried with a towel and placed in another holding cage on a heating pad in the housing room for 15 to 20 minutes before being returned to its home cage. Rats were euthanized with an overdose of 60 mg/kg sodium pentobarbital 1 hour later followed by decapitation.\n\nObject in a Novel Location (ONL). For the Object in a Novel Location (ONL) portion, there were three groups of seven rats (n = 21). Each group was trained and tested on a single day: PND16, PND18 or PND20. For training, rats were individually placed into the arena and allowed to explore. Two identical objects were placed in adjacent arena quadrants and remained in the same location for the familiarization phase. There were three familiarization periods each lasting 7 minutes carried out once per hour over 3 hours. Between familiarization periods, rats were placed back into their home cage with their cage mates.\n\nOne and a half hours after the last familiarization period, rats were tested for 5 minutes. For the test, one of the objects was moved to a different quadrant of the arena. This procedure reflects the conventional object-in-novel-place preference test which takes advantage of a rats spontaneous tendency to explore objects that have changed location within an otherwise stable environment19,26,27. When rats display such a preference, it is inferred that they have detected a change in location of the object within the environment. To determine preference during the testing phase, an investigation ratio was calculated during the 5 minute testing phase; the proportion of total object-investigation that was spent investigating the displaced object to the total time investigating both objects (tdisplaced/[tdisplaced + tnot displaced]). A rat was considered to be investigating an object when its head was within 4 cm and its nose was within a 45° angle from being perpendicular to the object. Investigation was also considered when a rat was rearing with at least one forepaw making contact with the object with the head oriented upwards but not when climbing on top of the object. Each rat was tested once to ensure that the exploration ratio was based solely on the change in the position of the object from the familiarization to the testing phase. Rats were euthanized with an overdose of 60 mg/kg sodium pentobarbital 1 hour later followed by decapitation.\n\nHome Cage Controls (HCC). Three additional, separate groups of rats were euthanized with an overdose of 60 mg/kg sodium pentobarbital on PND16, PND18, or PND20 (n = 5/time point; 15 total) with no behavioural manipulations (home cage controls; HCC). All brains were processed immunohistochemically as described below.\n\nTissue preparation. Brains were immersion fixed in 4%-paraformaldehyde (Sigma; USA) in 0.1M phosphate-buffered saline (PBS) overnight at 4°C. This solution was replaced with 30% sucrose in 0.1M PBS the following day and brains were stored at 4°C until sectioning. Brains were sectioned through the dorsal hippocampus at 60µm on a Leica CM1900 cryostat (Weztler, Germany). Because learning-associated changes in mossy fiber axonal input to the CA3 region are restricted to the most rostral levels of the dorsal hippocampus33,59, we focussed our immunohistochemical analyses to this region. Sections were stored in a 0.1% sodium azide solution in 0.1M phosphate buffer (PB) at 4°C.\n\nSynaptophysin staining. Sections were washed for 15 min in a 0.2% Triton-X/0.01 M phosphate-buffered saline (T-PBS) then blocked in a 1x animal free blocker (Vector)/T-PBS solution for 1 h at room temperature. Incubation in the primary antibody (rabbit anti-synaptophysin polyclonal antibody from Chemicon/Millipore (Cat#: AB9272), 1:2500) occurred overnight at room temperature. The following day, sections were washed in T-PBS for 15 minutes followed by a 2 hour incubation in the secondary antibody (1:500 goat anti-rabbit Igg (H+L) polyclonal secondary Alexa Fluor 594 from Molecular Probes; Cat# R37117). Sections were given a final rinse in 0.01 M PBS (pH 7.4) for 15 min then mounted on glass slides and coverslipped with glass coverslips adhered with Fluormount (Sigma).\n\nc-Fos staining. Sections adjacent to those stained for synaptophysin were placed in phosphate-buffered saline with Triton X (T-PBS) for three 5 minute washes. They were then incubated in 0.3% hydrogen peroxide (H2O2) in T-PBS for 15 minutes followed by three, 5 minute washes in T-PBS. Sections were transferred to 1x animal free blocker (AFB; Vector) in T-PBS for 30 minutes at room temperature. Incubation in the primary antibody (rabbit polyclonal anti-c-Fos from Abcam (Cat# ab53036), 1:5000) occurred overnight at room temperature. The following day, tissue was washed in T-PBS for three, 10 minute washes followed by a 2 hour incubation in the secondary antibody (Biotinylated Goat Anti-Rabbit IgG (H+L) Antibody from Vector Laboratories (Cat# BA-1000), 1:1000). Tissue was washed for three 10 minute washes in T-PBS before being placed into an ABC solution (Vector) for 1 hour. The tissue was rinsed in three 5 minute washes using PBS before being placed into a nickel-enhanced 3,3′-Diaminobenzidine (DAB) solution. Sections were then mounted on glass slides, dehydrated (1 minute in distilled H2O; 1 minute in 25% ethanol; 2 minutes in 50% ethanol; 5 minutes in 90% ethanol; 10 minutes in 100% ethanol; 20 minutes in Clearene) and cover slipped.\n\nSynatophysin quantification. Fluorescent images of the CA3 region were captured at 20× magnification using a Retiga-2000R camera (QImaging, BC) and an Olympus BX61 research microscope (Olympus-Canada). Three coronal sections were sampled from the dorsal hippocampus. The anterior boundary was defined as the first section where the upper and lower blades of the granule cell layers were equal in length. The posterior limit was 420 μm from this initial anterior starting location. Three coronal sections were sampled from this anterior-posterior boundary with 120 μm between sections. Synaptophysin staining measurements for each region of interest were taken as averages of the number of puncta from the three coronal levels. The areas of the stratum lucidum (SL) and the stratum oriens (SO) were estimated by outlining the synaptophysin-positive region. Synaptophysin-positive puncta were defined as having an intensity value twice that of the background as measured on the stratum radiatum (SR) of the CA3 region and a size restricted to not greater than 5 µm and not less than 1 µm. Comparisons between groups were made using 1) total puncta in the SL; 2) total puncta in the SO; 3) the ratio of puncta in SO to SL to account for size variations between individual animals. An experimenter who was blind to group assignment carried out all analyses.\n\nc-Fos quantification. Utilizing the Optical Fractionator probe, an estimate of the number of c-Fos positive cells in the dentate gyrus (DG), CA3 and CA1 regions of the dorsal hippocampus was undertaken. Stained sections were visualized using an Olympus BX51 brightfield microscope with a motorized stage (Olympus Canada, Markham, ON) and images captured with an Olympus U-CMAD3 camera. Stereo Investigator (version 11.06.1; MBF Bioscience, Williston, VT) software was used for quantification. The region of interest (DG, CA1 or CA3 of the dorsal hippocampus) for each section was traced digitally at 4× magnification. The anterior boundary was defined as the first section where the upper and lower blades of the granule cell layers were equal in length. The posterior limit was 420 μm from this initial anterior starting location. Based on these limits, the volume of the DG measured 17,892 µm3; CA3 volume was 31,637 µm3; CA1 volume was 12,487 µm3. Because of these volume differences across hippocampal subregions, c-Fos data were expressed as c-Fos+ cells per 100 µm3 to ease visual comparisons across subregions. Cell density (as measured on cresyl violet-stained sections from 3 rats from each age group – PND16, PND18 and PND20) in the DG, CA3 and CA1 regions was not different across the 3 age groups (one-way ANOVA; DG (F(2,6) = 3.29, p = 0.089; CA3 (F(2,6) = 3.60, p = 0.094) and CA1 (F(2,6) = 3.56, p = 0.096). Three coronal sections were sampled from this anterior-posterior limit with 120 μm between sections. Counting parameters were set to a counting frame of 30×30 µm2 and a dissector height of 10 µm between the top and bottom guard zones with average mounted section thickness of 35 μm. c-Fos positive cells were quantified using a 60X magnification lens (oil immersion, NA 1.35) when the uppermost tip of c-Fos positive nuclei were in focus within the counting frame and the dissector height. Stereo Investigator software used planar and depth information for each counted nuclei to calculate the volume for the digitally traced region of interest and provide an estimate of the labeled cells per region of interest (DG, CA1 or CA3).\n\n\nResults\n\nEscape latency (Figure 1A), speed (Figure 1B), pathlength (Figure 1C) and percent pathlength spent swimming along the edge of the pool (thigmotaxis; Figure 1D) data for the water maze task (see Dataset 1) were analyzed using separate two-way, repeated measures ANOVAs (age (PND16, PND18, PND20) as the between factor and trial as the repeated measure). Analysis of latency data (Figure 1A) revealed main effects of age (F(2,12) = 8.49, p < 0.01) and trial (F(7,14) = 2.44, p < 0.05) but no interaction (F(14,84) < 1.0). Tukey HSD post-hoc tests on the main effect of group revealed that the PND20 group showed shorter latencies to locate the platform than the PND16 group (t(8) = 3.37, p < 0.01) and the PND18 group (t(8) = 3.18, p < 0.05).\n\nGroups of rats were trained for one day (PND16, PND18, PND20) to locate the hidden platform on the water maze task. Swimming behaviors were captured with HVS Image software (version 1/09) and included: (A) Latency in seconds (mean ± SEM) to reach the platform (60 s cutoff); (B) speed in meters/second (mean ± SEM); (C) pathlength in meters (mean ± SEM); (D) thigmotaxis defined as percent of total pathlength spent within 10 centimeter of the pool wall (mean ± SEM). Significant group differences were found on latency to reach the platform with PND20 group showing shorter latencies than the PND16 ( p < 0.01) and PND18 (+, p < 0.05). Analysis of speed data revealed faster swim speeds in the PND20 group than both other age groups (, p < 0.01). Measures of thigmotaxis revealed less thigmotaxis in the PND20 group than both other groups (, p < 0.05). A significant interaction (##, p < 0.01) was also found and specific comparisons showed that the PND20 group spent less time along the edge of the pool on Trials 5 and 8 than PND16 and PND18 (A and C; p < 0.05) and on Trial 6 than PND18 (B; p < 0.01). @ and @@ in A and D, respectively, refer to a main effect of Trial.\n\nAnalysis of speed data (meters/sec; Figure 1B) revealed a main effect of age (F(2,12) = 86.18, p < 0.001) but no effect of trial (F(7,14) < 1.0) and no interaction (F(14,84) = 1.43). Tukey HSD post-hoc tests on the main effect of group revealed a significantly faster swim speed in the PND20 group than the PND16 group (t(8) = 13.11, p < 0.001) and PND18 group (t(8) = 9.28, p < 0.001).\n\nAnalysis of pathlength data (meters; Figure 1C) revealed no main effect of age (F(2,12) = 2.11, p = 0.163), no effect of trial (F(7,14) = 2.11, p = 0.051) and no interaction (F(14,84) = 1.01, p = 0.45).\n\nAnalysis of thigmotactic data (percent of the swimming distance spent within 10 cm of the wall of the pool; Figure 1D) revealed a main effect of age (F(2,12) = 7.07, p < 0.01), a main effect of trial (F(7,14) = 4.92, p < 0.001) and a significant interaction (F(14,84) = 2.97, p < 0.01). Tukey HSD post-hoc tests on the main effect of group revealed significantly less thigmotaxis in the PND20 group than the PND16 group (t(8) = 2.64, p < 0.05) and the PND18 group (t(8) = 3.32, p < 0.05). At Trials 5 and 8, the PND20 group showed less thigmotaxis than the PND16 and PND18 groups (p < 0.05) and at Trial 6, the PND20 group showed less thigmotaxis than the PND18 group (p < 0.01).\n\nThe mean investigation ratios (see Dataset 2) for the ONL task are shown in Figure 2A. One-way ANOVA on the investigation ratios between ages revealed a main effect of age (F(2,18) = 4.64, p < 0.05) and Tukey HSD post-hoc tests revealed that the PND20 group showed a larger investigation ratio than both other ages (p < 0.05). The investigation ratios for the PND16 (t(6) = 1.35) and PND18 (t(6) = 0.63) groups were not significantly different from chance (0.5) while the investigation ratio for the PND20 group (t(6) = 2.52, p < 0.05) was significantly greater than chance. An analysis of the times spent investigating the displaced and nondisplaced objects (in seconds; Dataset 2) during the test (Figure 2B) with a two-way ANOVA (age by investigation time of the displaced and nondisplaced objects) revealed no main effects of age (F(2,18) = 1.60), no main effect of investigation time (F(1,2) = 1.38) and no interaction (F(2,18) = 1.67). One-way ANOVA examining the distance traveled (see Dataset 2) during the test session (in meters; Figure 2C) revealed a main effect age group (F(2,18) = 4.02, p < 0.05). Tukey post-hoc test showed that the PND18 group showed a longer distance traveled during the test than the PND20 group (t(12) = 2.50, p < 0.05). No other age group differences were detected (PND16 vs PND18, t(12) = 1.31).\n\nGroups of rats were trained and tested on a single day: PND16, PND18 or PND20. (A) Mean investigation ratios (± SEM) from the entire 5 minute test. Dotted line indicates chance investigation ratios meaning similar times spent investigating the displaced (D) and non-displaced (ND) objects. The PND20 group showed an investigation ratio significantly above chance (* p < 0.01) indicating a preference for the D object over the ND object. The PND20 group also showed a larger investigation ratio than the PND16 and PND18 groups (* p < 0.05). (B) Average total time (seconds ± SEM) spent investigating the D and ND objects. There were no overall age effects or object effects on investigation time. (C) Total distance traveled during the ONL 5 minute test. Data are expressed as average (± SEM) distance in meters. A main effect of age (, p < 0.05) was found with the PND18 group showing a longer distance traveled than the PND20 group (+, p < 0.05).\n\nStratum Lucidum (SL). The average number of synaptophysin-positive puncta (Dataset 3) in the SL (Figure 3A and Supplemental Figure S1) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed a main effect of age only (F(2,2) = 8.53, p < 0.01) (training history F(2,4) < 1.0; interaction F(4,42) < 1.0). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND16 group showed more synaptophysin-positive puncta in the SL than both the PND18 and PND20 groups (p < 0.05).\n\nBrains from PND16, PND18 and PND20 from home cage control (HCC), object in a novel location (ONL) or Morris water maze (MWM) groups were removed and processed immunohistochemically for synaptophysin staining in the CA3 region as a marker for MF connectivity. Left panels show representative staining from the three age groups (images from MWM condition). Abbreviations: SO: stratum oriens; SP: stratum pyramisale; SL: stratum lucidum. Arrows point to synaptophysin-positive puncta in the SO region. Scale bar = 100 µm. (A) Synaptophysin-positive puncta quantified in the SL region revealed a main effect of age with group PND16 showing more staining than PND18 and PND20 (, p < 0.05). (B) Synaptophysin-positive puncta quantified in the SO region revealed a main effect of age with group PND18 showing more staining than PND16 (++, p < 0.01) and group PND20 showing more staining than groups PND16 and PND18 (, p < 0.01). (C) Ratio of synaptophysin-positive puncta quantified in the SO: SL region (to control for potential size variation) revealed a main effect of age with group PND18 showing more staining than PND16 (++, p < 0.01) and group PND20 showing more staining then groups PND16 and PND18 (, p < 0.01).\n\nStratum Oriens (SO). The average number of synaptophysin-positive puncta (Dataset 3) in the SO (Figure 3B and Supplemental Figure S1) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed a main effect of age only (F(2,2) = 91.78, p < 0.001) (training history F(2,4) < 1.0; interaction F(4,42) < 1.0). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND18 group showed more synaptophysin-positive puncta in the SO than the PND16 group (p < 0.01) and the PND20 group showed more synaptophysin-positive puncta in the SO than both the PND16 and PND18 groups (p < 0.001; Figure 3B).\n\nSO:SL Ratio. The ratio (see Dataset 3) of synaptophysin-positive puncta in the SO to puncta in the SL (Figure 3C and Supplemental Figure S1) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed a main effect of age only (F(2,2) = 91.52, p < 0.001) (training history F(2,4) < 1.0; interaction F(4,42) < 1.0). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND18 group showed more synaptophysin-positive puncta in the SO than the PND16 group (p < 0.001) and the PND20 group showed more synaptophysin-positive puncta in the SO than both the PND16 and PND18 groups (p < 0.001; Figure 3C).\n\nDentate Gyrus (DG). The estimated number of c-Fos-positive cells per 100 µm3 (see Dataset 4) in the DG (quantified in Figure 4A; representative images from rats trained on the MWM shown in Figure 5 and for groups HCC and ONL in Supplemental Figure S2 and Supplemental Figure S3) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed main effects of age (F(2,2) = 96.99, p < 0.001) and training history (F(2,4) = 3.45, p < 0.05) but no significant interaction (F(4,42) = 1.23). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND18 group showed more c-Fos positive cells in the DG than the PND16 group (p < 0.01) and the PND20 group showed more c-Fos positive cells in the DG than both other age groups (p < 0.001; Figure 4A). Post-hoc analysis of the main effect of training history revealed fewer c-Fos positive cells associated with MWM task than ONL task (p < 0.05).\n\nThe estimated number of c-Fos-positive cells per 100 µm3 were quantified in the (A) dentate gyrus (DG); (B) CA3 and (C) CA1 hippocampal regions from PND16, PND18 and PND20 groups from home cage control (HCC), object in a novel location (ONL) or Morris water maze (MWM) conditions. (A) DG c-Fos staining revealed more c-Fos-positive cells in the PND18 than the PND16 group (++, p < 0.01) and more positive cells in the PND20 group than the PND16 and PND18 groups (* p < 0.01). (B) CA3 staining revealed more c-Fos positive cells in the PND18 than the PND16 group (++, p < 0.01) and in the pND20 group than the PND16 group (##, p < 0.01). There was also more c-Fos labelling associated with the MWM condition than the HCC and ONL conditions (@@, p < 0.01). (C) CA1 staining revealed more c-Fos positive cells in the PND20 than the PND16 and PND18 groups (, p < 0.01). There was also more c-Fos labelling associated with the MWM condition than the HCC and ONL conditions (@@, p < 0.01).\n\nRepresentative images for immunohistochemical localization of c-Fos in the dentate gyrus (DG), CA3 and CA1 regions. All images are from rats included in the MWM condition. Top row shows sections from PND16; middle row from PND18; bottom row from PND20. Images were taken at 10× with scale bar = 200 μm. Abbreviations: ML: molecular layer of the dentate gyrus; GC: granule cells; S: stratum lucidum; SP: stratum pyramidale; SO: stratum oriens; SR: stratum radiatum; SLM: stratum lacunosum moleculare.\n\nCA3. The estimated number of c-Fos-positive cells per 100 µm3 (see Dataset 4) in the CA3 (quantified in Figure 4B; representative images from rats trained on the MWM shown in Figure 5 and for groups HCC and ONL in Supplemental Figure S2 and Supplemental Figure S3) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed main effects of age (F(2,2) = 28.43, p < 0.001) and training history (F(2,4) = 7.08, p < 0.01) but no significant interaction (F(4,42) = 1.95). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND20 and PND18 groups showed more c-Fos positive cells in the CA3 region than the PND16 group (p < 0.001). Post-hoc analysis of the main effect of training history revealed more c-Fos staining associated with the MWM task than the ONL task and the HCC condition (p < 0.01 for both comparisons).\n\nCA1. The estimated number of c-Fos-positive cells per 100 µm3 (see Dataset 4) in the CA1 (quantified in Figure 4C; representative images from rats trained on the MWM shown in Figure 5 and for groups HCC and ONL in Supplemental Figure S2 and Supplemental Figure S3) was analyzed using a 3×3 univariate ANOVA (age (PND16, PND18, PND20) and training history (HCC, ONL, MWM) as the fixed factors). This analysis revealed main effects of age (F(2,2) = 14.21, p < 0.001) and training history (F(2,4) = 12.49, p < 0.001) but no significant interaction (F(4,42) = 1.26). Tukey HSD post-hoc comparisons on the main effect of age revealed that the PND20 group showed more c-Fos positive cells in the CA1 region than both other age groups (p < 0.001). Post-hoc analysis of the main effect of training history revealed more c-Fos staining associated with the MWM task than the ONL task and the HCC condition (p < 0.01 for both comparisons).\n\n\nDiscussion\n\nThe purpose of the current work was three-fold: 1) to determine whether the emergence of spatial-behavioral function was based on the ability to generate appropriate behavioral output; 2) to assess whether mossy fiber connectivity patterns preceded optimal spatial-behavioral function; 3) to explore functional changes in the hippocampus to determine whether activity in hippocampal networks occurred in a training-dependent or developmentally-dependent fashion. At PND20, optimal spatial behavioral performance emerged on both the water maze task and the object exploration task whereas at PND16 and PND18, optimal spatial behavior was not present on either task. Analysis of synaptophysin staining in the CA3 SL and SO subregions revealed more synaptophysin staining in the SO in the PND18 than the PND16 group and more synaptophysin staining in the SO in the PND20 group than both the PND16 and PND18 groups. Estimates of c-Fos-positive cells in the DG, CA3 and CA1 regions revealed more staining in the DG and CA3 regions in the PND18 and PND20 groups than the PND16 group (and more in the PND20 versus PND18 group) as well as more c-Fos staining in the CA3 and CA1 areas associated with water maze training. The combined behavioral and immunohistochemical results support the hypothesis that hippocampal anatomical and functional neural substrates mature prior to the emergence of optimal spatial-behavioral function. The maturation of these networks may be necessary to support the emergence of optimal spatial behavioral performance.\n\nTraining on two tasks (water and dry maze) that required distinct behavioral topology (swimming versus walking) to complete were used. In this case, the PND20 group showed evidence of optimal spatial proficiency on both tasks when compared to the PND16 and PND18 groups. Behavioral results revealed that the PND20 group given 1 day of water maze training had significantly lower latencies than the groups trained at PND16 or PND18 suggesting the emergence of spatial function by PND20. The PND20 group also showed significantly faster swim speeds. This raises the possibility that the older rats were better able to perform the task than PND16 and PND18 rats because they were stronger and possibly, better swimmers. Rather than spatial function being the reason for the superior water maze performance in the PND20 group, it could be argued that physicality contributed to superior water maze task performance. Because swimming requires a certain level of physical aptitude, an older rat may show better performance because they have more muscle mass to perform the task.\n\nTo examine whether the improvements seen at PND20 were due to spatial or motor development, an experiment was run on a dry maze. This task was less physically demanding than the water maze task (ambulation rather than swimming) and was hypothesized to tease out spatial from motor development. Other reports have shown that by PND15/PND16, rats are able to perform tests of locomotor activity and engage in exploratory behaviours such as rearing that are similar to adults8,74 and what was required for performance of the ONL task used in the present study. In contrast, optimal (i.e., adult) swimming behavior does not fully emerge until PND2268 though PND15 rats are able to keep their face and nose out of water. Thus, performance on the ONL task was utilized as a less physically demanding task yet still able to tap into spatial functions. For the ONL task, the PND20 group performed significantly better than chance but the younger groups (PND16 and PND18) showed chance performance. Because the ONL task takes advantage of the ability to detect changes in the spatial relationship between the two objects but is less physically demanding than the water maze, a preliminary conclusion is that optimal spatial behavioral function emerges around PND20, independent from behavioral topology.\n\nConverging lines of evidence from other behavioral studies point to the postnatal period from PND16 to PND21 as a sensitive developmental timeframe for spatial behavioral function to emerge. Improved performance on a number of spatial tasks during development2–5,13,16,20,21,29,32,38,41,46,64–66,70,71,73,76 point to a sensitive developmental period for spatial behavior to emerge between PND16 and PND21. In addition, optimal performance on a delayed alternation task has been shown to emerge between PND19 and PND27 but in pups with PND10 hippocampal lesions, this behavior fails to fully develop23 suggesting that hippocampal integrity is important for the normal development of spatial behavioral function and other brain structures do not compensate when the hippocampus is damaged during postnatal development6.\n\nIn adult Long Evans rats (LER), the mossy fiber terminal field (as labeled with zinc or synaptophysin) shows a dense projection to the stratum oriens (SO) corresponding to the basal dendrites of the CA3 pyramidal cells33. In an examination of the development of CA3 hippocampal mossy fiber (MF) distribution in LER32, MF innervation of the stratum lucidum (SL) was widespread by PND12 and beginning on PND15, MF staining was evident in the stratum pyramidale (SP) and by PND18 and PND21, widespread MF staining was observed in the SO. By PND24, the SO projection in LER was complete and remained stable into adulthood33.\n\nThe present study carried out an explicit comparison between the development of these MF connectivity patterns and the emergence of spatial behavior. While in a previous study41, using LER rats, dramatic improvement in spatial behavioral function was observed from PND18 to PND21, there was no examination of the MF distribution during this course of behavioral improvement. Utilizing a one-day training procedure on two different tasks and comparing the change in MF connectivity patterns to a home cage control condition allowed us to determine how plasticity in the MF terminal field was associated with the emergence of spatial behavior. In this respect, clear developmental-dependent changes occurred in MF terminal field distribution (inferred from synaptophysin staining) with a gradual increase in staining observed in the SO from PND16 to PND18 to PND20 with SO staining being greatest at PND20. This suggests that developmentally-dependent changes in the distribution of MF terminals, particularly those that terminate on the basal dendrites of CA3 pyramidal neurons, may form an anatomical substrate that supports the emergence of spatial behavior. Such facilitation of spatial behavior could arise as a consequence of shifting input away from inhibitory interneurons10,51,53 thereby reducing the number of synapses on inhibitory interneurons located within the SL72 leading to excitatory inputs on CA3 pyramidal neurons via the basal dendrites in SO.\n\nThe developmentally-dependent changes in presynaptic MF terminals are likely dependent on patterned neural activity within this system that occurs during development39. To examine putative functional changes associated with developmentally-linked improvements in spatial behavior and MF terminal field distribution, c-Fos immunohistochemical staining in the DG, CA3 and CA1 regions of the hippocampus was carried out across the three developmental time points (PND16, PND18 and PND20). Immediate early gene proteins, such as c-Fos, regulate the transcription of additional genes directing a general genomic response to a variety of environmental stimuli69. These regulatory proteins control downstream gene expression and are thought to translate environmental signals into relatively long-term changes in neuronal function28,40,69. As such, c-Fos protein labeling was used in the present work to provide a marker for cells that have recently been activated and are potentially undergoing long-term structural or functional changes.\n\nComparison of c-Fos-positive cells between the different ages revealed more c-Fos labeled neurons in the DG and CA1 regions in the PND20 group than the PND16 and PND18 groups indicating more network activity in these regions at PND20 that may facilitate the emergence of spatial behavior. The PND18 group also showed more c-Fos labeling in the DG and CA3 than the PND16 group, perhaps providing the patterned neural activity required for development changes in the MF connectivity. Finally, there were training-associated elevations in c-Fos staining following the water maze task in the CA3 and CA1 regions.\n\nThe developmental-associated changes in c-Fos labeling in the DG and CA1 regions continued to show increased activity from PND16 to PND18 then more activity at PND20 while the CA3 labeling patterns suggested a plateau of activity at PND20. Activity changes in the DG during development may reflect patterned activity necessary for the stabilization of MF inputs to CA3. Indeed, the wave of increased DG activity at PND18 corresponds to the first occurrence of a significant MF projection to SO; the second wave of DG activity at PND20 corresponds to a further, significant expansion of the MF projection within SO. With the formation and stabilization of this MF-CA3 network, optimal spatial behavior may be possible. A previous report showed that the expression of Homer1a in the DG peaks between PND19 and PND2352 compared to PND9-15 and PND35 to adulthood. This is consistent with the current results showing a continued maturation of network activity in the DG spanning the PND16 – PND20 period. During this maturational period, appropriate networks may be established to support optimal spatial behavioral output.\n\nThe CA1 region also showed continued maturation of network activity spanning PND16 – PND20 (with c-Fos-positive neurons at PND18 > PND16 and PND20 > PND18). Activity patterns in the CA1 region may continue to mature well beyond the PND20 period. As an example, place cells in CA1 with specific spatial firing patterns have been recorded at PND17 with place cell patterns conveying optimal spatial information showing continued development up to PND351,46. Likewise, place cell firing is present from PND16 – PND26 but continues to improve throughout development with stable place cell recordings (i.e., similar to adults) being made at PND2876. Theta-modulated firing is present at PND16 in the hippocampus CA1 region and these responses reach adult proportions by PND2246,76. After PND21, both the magnitude and threshold for post-synaptic induction of long-term potentiation (LTP) are reduced with a corresponding increase in the threshold for presynaptic induction13,22. These data, with the present results, suggest continued maturation of CA1 neural activity beyond the PND20 time point that continues to support the refinement and optimization of spatial information.\n\nThe CA3 and CA1 subregions showed patterned c-Fos labeling that was associated with the water maze task condition. While studies have shown that the dorsal CA3 and CA1 hippocampal subregions contribute to numerous aspects of spatial processing31,36,37,45, there are functional differences between the two regions. The CA3 subregion has been suggested to be important for cognitive functions associated with spatial pattern separation44,63, spatial pattern completion48, novelty detection47 and short-term memory45,49. Likewise, the dorsal CA3 region has been shown to be involved in the early stages of acquiring spatial information but less involved during the retrieval of spatial information after long delays50. The dorsal CA1 region has been shown to be involved in the encoding and retrieval of spatial-shock associations35 and inactivation of the CA1 region with lidocaine prior to water maze training impairs spatial performance14,55,61. As well, CA1 lidocaine injections before a probe retention test impaired the ability of rats to demonstrate spatial function15,56.\n\nIn the present work, both the CA3 and CA1 regions showed c-Fos elevations associated with performance of the water maze task. Because there was no probe retention test, this patterned activity appears to be related to the acquisition of spatial information. For the ONL task, the animals were euthanized following the test session. In this case, there was no significant increase in c-Fos labeling over and above the HCC condition. This suggests the possibility that these regions were not activated during the ONL task or that they were not involved in the retrieval of spatial information. While the present study was not designed to examine differential activation patterns in the CA3 and CA1 regions based on different cognitive demands, it would be of interest to determine if distinct mnemonic functions (e.g., pattern separation mediated by the DG-CA3 network42,62 and pattern completion mediated by the CA3-CA1 network43) would show a different pattern of emergence during development.\n\n\nConclusions\n\nThe combination of neuroanatomical and behavioural results from the present work leads to the hypothesis that this developmental time period (PND16 – PND20) represents a sensitive period for hippocampal anatomical and functional modification leading to the emergence of spatial behavior. The main purpose of the present study was to investigate the relationship between the emergence of spatial-behavior function and the emergence of the neural substrates – both anatomical and functional – that would support optimal spatial-behavioral function. In this respect, MF connectivity-based developmental changes preceded the emergence of spatial behavior. In association with this, functional aspects of hippocampal neurons changed in response to developmental age with functional changes also occurring in response to behavioral-dependent inputs. These data support the hypothesis that network maturation the hippocampus supports the emergence of optimal spatial behavior.\n\n\nData availability\n\nF1000Research: Dataset 1. Watermaze Dataset, 10.5256/f1000research.6822.d9653078\n\nF1000Research: Dataset 2. Drymaze Dataset, 10.5256/f1000research.6822.d9653179\n\nF1000Research: Dataset 3. Synaptophysin Quantification Dataset, 10.5256/f1000research.6822.d9653280\n\nF1000Research: Dataset 4. C-Fos Quantification Dataset, 10.5256/f1000research.6822.d9653381", "appendix": "Author contributions\n\n\n\nRC, NJG, DGM and MRH conceived the study. DGM and MRH designed the experiments. RC and NJG carried out the research. RC prepared the first draft of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nThe authors have no conflict of interest to report.\n\n\nGrant information\n\nThis research was supported by National Science and Engineering Research Council of Canada Discovery grants to DGM (RGPIN 156937) and MRH (RGPIN 341673), a Canadian Foundation for Innovation grant (335892) to MRH and a Fonds de la recherche en santé du Québec to DGM.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to thank Ridhi Nair for quantifying cresyl violet staining.\n\n\nSupplementary Material\n\nBrains from PND16, PND18 and PND20 from home cage control (HCC) and object in a novel location (ONL) groups were removed and processed immunohistochemically for synaptophysin staining in the CA3 region as a marker for MF connectivity. Top panel shows representative staining from the three age groups from the HCC condition and bottom panel show staining from the ONL condition. Abbreviations: SO: stratum oriens; SP: stratum pyramisale; SL: stratum lucidum. Arrows point to synaptophysin-positive puncta in the SO region.\n\nRepresentative images for immunohistochemical localization of c-Fos in the dentate gyrus (DG), CA3 and CA1 regions from rats included in the HCC condition. Top row shows sections from PND16; middle row from PND18; bottom row from PND20. Images were taken at 10×. Abbreviations: ML: molecular layer of the dentate gyrus; GC: granule cells; S: stratum lucidum; SP: stratum pyramidale; SO: stratum oriens; SR: stratum radiatum; SLM: stratum lacunosum moleculare.\n\nRepresentative images for immunohistochemical localization of c-Fos in the dentate gyrus (DG), CA3 and CA1 regions from rats included in the ONL condition. Top row shows sections from PND16; middle row from PND18; bottom row from PND20. Images were taken at 10×. Abbreviations: ML: molecular layer of the dentate gyrus; GC: granule cells; S: stratum lucidum; SP: stratum pyramidale; SO: stratum oriens; SR: stratum radiatum; SLM: stratum lacunosum moleculare.\n\n\nReferences\n\nAinge JA, Langston RF: Ontogeny of neural circuits underlying spatial memory in the rat. Front Neural Circuits. 2012; 6: 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAkers KG, Candelaria-Cook FT, Rice JP, et al.: Delayed development of place navigation compared to directional responding in young rats. Behav Neurosci. 2009; 123(2): 267–275. PubMed Abstract \| Publisher Full Text\n\nAkers KG, Candelaria-Cook FT, Rice JP, et al.: Cued platform training reveals early development of directional responding among preweanling rats in the Morris water task. Dev Psychobiol. 2011; 53(1): 1–12. PubMed Abstract \| Publisher Full Text\n\nAkers KG, Candelaria FT, Hamilton DA: Preweanling rats solve the Morris water task via directional navigation. Behav Neurosci. 2007; 121(6): 1426–1430. PubMed Abstract \| Publisher Full Text\n\nAkers KG, Hamilton DA: Comparison of developmental trajectories for place and cued navigation in the Morris water task. Dev Psychobiol. 2007; 49(6): 553–564. PubMed Abstract \| Publisher Full Text\n\nAltemus KL, Almli CR: Neonatal hippocampal damage in rats: long-term spatial memory deficits and associations with magnitude of hippocampal damage. Hippocampus. 1997; 7(4): 403–415. PubMed Abstract \| Publisher Full Text\n\nAltman J, Das GD: Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. J Comp Neurol. 1965; 124(3): 319–335. PubMed Abstract \| Publisher Full Text\n\nAltman J, Sudarshan K: Postnatal development of locomotion in the laboratory rat. Anim Behav. 1975; 23(4): 896–920. PubMed Abstract \| Publisher Full Text\n\nAmaral DG, Dent JA: Development of the mossy fibers of the dentate gyrus: I. A light and electron microscopic study of the mossy fibers and their expansions. J Comp Neurol. 1981; 195(1): 51–86. PubMed Abstract \| Publisher Full Text\n\nAradi I, Maccaferri G: Cell type-specific synaptic dynamics of synchronized bursting in the juvenile CA3 rat hippocampus. J Neurosci. 2004; 24(43): 9681–9692. PubMed Abstract \| Publisher Full Text\n\nBannerman DM, Good MA, Butcher SP, et al.: Distinct components of spatial learning revealed by prior training and NMDA receptor blockade. Nature. 1995; 378(6553): 182–186. PubMed Abstract \| Publisher Full Text\n\nBayer SA: Development of the hippocampal region in the rat. I. Neurogenesis examined with 3H-thymidine autoradiography. J Comp Neurol. 1980; 190(1): 87–114. PubMed Abstract \| Publisher Full Text\n\nBlair MG, Nguyen NN, Albani SH, et al.: Developmental changes in structural and functional properties of hippocampal AMPARs parallels the emergence of deliberative spatial navigation in juvenile rats. J Neurosci. 2013; 33(30): 12218–12228. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBohbot V, Otáhal P, Liu Z, et al.: Electroconvulsive shock and lidocaine reveal rapid consolidation of spatial working memory in the water maze. Proc Natl Acad Sci U S A. 1996; 93(9): 4016–4019. PubMed Abstract \| Free Full Text\n\nBroadbent NJ, Squire LR, Clark RE: Reversible hippocampal lesions disrupt water maze performance during both recent and remote memory tests. Learn Mem. 2006; 13(2): 187–191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrown RW, Whishaw IQ: Similarities in the development of place and cue navigation by rats in a swimming pool. Dev Psychobiol. 2000; 37(4): 238–245. PubMed Abstract \| Publisher Full Text\n\nCain DP, Saucier D, Boon F: Testing hypotheses of spatial learning: the role of NMDA receptors and NMDA-mediated long-term potentiation. Behav Brain Res. 1997; 84(1–2): 179–193. PubMed Abstract \| Publisher Full Text\n\nCurlik DM 2nd, Difeo G, Shors TJ: Preparing for adulthood: thousands upon thousands of new cells are born in the hippocampus during puberty and most survive with effortful learning. Front Neurosci. 2014; 8: 70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDix SL, Aggleton JP: Extending the spontaneous preference test of recognition: evidence of object-location and object-context recognition. Behav Brain Res. 1999; 99(2): 191–200. PubMed Abstract \| Publisher Full Text\n\nDouglas RJ, Peterson JJ, Douglas DP: The ontogeny of a hippocampus-dependent response in two rodent species. Behav Biol. 1973; 8(1): 27–37. PubMed Abstract \| Publisher Full Text\n\nDumas TC: Early eyelid opening enhances spontaneous alternation and accelerates the development of perforant path synaptic strength in the hippocampus of juvenile rats. Dev Psychobiol. 2004; 45(1): 1–9. PubMed Abstract \| Publisher Full Text\n\nDumas TC: Postnatal alterations in induction threshold and expression magnitude of long-term potentiation and long-term depression at hippocampal synapses. Hippocampus. 2012; 22(2): 188–199. PubMed Abstract \| Publisher Full Text\n\nFreeman JH Jr, Stanton ME: Fimbria-fornix transections disrupt the ontogeny of delayed alternation but not position discrimination in the rat. Behav Neurosci. 1991; 105(3): 386–395. PubMed Abstract \| Publisher Full Text\n\nGaarskjaer FB: The development of the dentate area and the hippocampal mossy fiber projection of the rat. J Comp Neurol. 1985; 241(2): 154–170. PubMed Abstract \| Publisher Full Text\n\nGaarskjaer FB: The organization and development of the hippocampal mossy fiber system. Brain Res. 1986; 396(4): 335–357. PubMed Abstract \| Publisher Full Text\n\nGaskin S, Gamliel A, Tardif M, et al.: Incidental (unreinforced) and reinforced spatial learning in rats with ventral and dorsal lesions of the hippocampus. Behav Brain Res. 2009; 202(1): 64–70. PubMed Abstract \| Publisher Full Text\n\nGaskin S, Tardif M, Cole E, et al.: Object familiarization and novel-object preference in rats. Behav Processes. 2010; 83(1): 61–71. PubMed Abstract \| Publisher Full Text\n\nGoelet P, Castellucci VF, Schacher S, et al.: The long and the short of long-term memory--a molecular framework. Nature. 1986; 322(6078): 419–422. PubMed Abstract \| Publisher Full Text\n\nGreen RJ, Stanton ME: Differential ontogeny of working memory and reference memory in the rat. Behav Neurosci. 1989; 103(1): 98–105. PubMed Abstract \| Publisher Full Text\n\nHenrich-Noack P: Please keep calm: investigating hippocampal function without stress. Front Behav Neurosci. 2014; 8: 356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoang LT, Kesner RP: Dorsal hippocampus, CA3, and CA1 lesions disrupt temporal sequence completion. Behav Neurosci. 2008; 122(1): 9–15. PubMed Abstract \| Publisher Full Text\n\nHolahan MR, Honegger KS, Routtenberg A: Expansion and retraction of hippocampal mossy fibers during postweaning development: strain-specific effects of NMDA receptor blockade. Hippocampus. 2007; 17(1): 58–67. PubMed Abstract \| Publisher Full Text\n\nHolahan MR, Rekart JL, Sandoval J, et al.: Spatial learning induces presynaptic structural remodeling in the hippocampal mossy fiber system of two rat strains. Hippocampus. 2006; 16(6): 560–570. PubMed Abstract \| Publisher Full Text\n\nHolahan MR, Routtenberg A: Lidocaine injections targeting CA3 hippocampus impair long-term spatial memory and prevent learning-induced mossy fiber remodeling. Hippocampus. 2011; 21(5): 532–540. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHunsaker MR, Kesner RP: Dissociations across the dorsal-ventral axis of CA3 and CA1 for encoding and retrieval of contextual and auditory-cued fear. Neurobiol Learn Mem. 2008; 89(1): 61–69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHunsaker MR, Lee B, Kesner RP: Evaluating the temporal context of episodic memory: the role of CA3 and CA1. Behav Brain Res. 2008; 188(2): 310–315. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHunsaker MR, Thorup JA, Welch T, et al.: The role of CA3 and CA1 in the acquisition of an object-trace-place paired-associate task. Behav Neurosci. 2006; 120(6): 1252–1256. PubMed Abstract \| Publisher Full Text\n\nJablonski SA, Schiffino FL, Stanton ME: Role of age, post-training consolidation, and conjunctive associations in the ontogeny of the context preexposure facilitation effect. Dev Psychobiol. 2012; 54(7): 714–722. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJung MW, McNaughton BL: Spatial selectivity of unit activity in the hippocampal granular layer. Hippocampus. 1993; 3(2): 165–182. PubMed Abstract \| Publisher Full Text\n\nKaczmarek L, Chaudhuri A: Sensory regulation of immediate-early gene expression in mammalian visual cortex: implications for functional mapping and neural plasticity. Brain Res Rev. 1997; 23(3): 237–256. PubMed Abstract \| Publisher Full Text\n\nKeeley RJ, Wartman BC, Häusler AN, et al.: Effect of juvenile pretraining on adolescent structural hippocampal attributes as a substrate for enhanced spatial performance. Learn Mem. 2010; 17(7): 344–354. PubMed Abstract \| Publisher Full Text\n\nKesner RP: A behavioral analysis of dentate gyrus function. Prog Brain Res. 2007; 163: 567–576. PubMed Abstract \| Publisher Full Text\n\nKesner RP: Behavioral functions of the CA3 subregion of the hippocampus. Learn Mem. 2007; 14(11): 771–781. PubMed Abstract \| Publisher Full Text\n\nKesner RP: A process analysis of the CA3 subregion of the hippocampus. Front Cell Neurosci. 2013; 7: 78. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKesner RP, Lee I, Gilbert P: A behavioral assessment of hippocampal function based on a subregional analysis. Rev Neurosci. 2004; 15(5): 333–351. PubMed Abstract \| Publisher Full Text\n\nLangston RF, Ainge JA, Couey JJ, et al.: Development of the spatial representation system in the rat. Science. 2010; 328(5985): 1576–1580. PubMed Abstract \| Publisher Full Text\n\nLee I, Hunsaker MR, Kesner RP: The role of hippocampal subregions in detecting spatial novelty. Behav Neurosci. 2005; 119(1): 145–153. PubMed Abstract \| Publisher Full Text\n\nLee I, Jerman TS, Kesner RP: Disruption of delayed memory for a sequence of spatial locations following CA1- or CA3-lesions of the dorsal hippocampus. Neurobiol Learn Mem. 2005; 84(2): 138–147. PubMed Abstract \| Publisher Full Text\n\nLee I, Kesner RP: Differential roles of dorsal hippocampal subregions in spatial working memory with short versus intermediate delay. Behav Neurosci. 2003; 117(5): 1044–1053. PubMed Abstract \| Publisher Full Text\n\nLee I, Rao G, Knierim JJ: A double dissociation between hippocampal subfields: differential time course of CA3 and CA1 place cells for processing changed environments. Neuron. 2004; 42(5): 803–815. PubMed Abstract \| Publisher Full Text\n\nMaccaferri G, Tóth K, McBain CJ: Target-specific expression of presynaptic mossy fiber plasticity. Science. 1998; 279(5355): 1368–1370. PubMed Abstract \| Publisher Full Text\n\nMontes-Rodriguez CJ, Lapointe V, Trivedi V, et al.: Postnatal development of Homer1a in the rat hippocampus. Hippocampus. 2013; 23(10): 890–902. PubMed Abstract \| Publisher Full Text\n\nMori M, Abegg MH, Gähwiler BH, et al.: A frequency-dependent switch from inhibition to excitation in a hippocampal unitary circuit. Nature. 2004; 431(7007): 453–456. PubMed Abstract \| Publisher Full Text\n\nMumby DG, Gaskin S, Glenn MJ, et al.: Hippocampal damage and exploratory preferences in rats: memory for objects, places, and contexts. Learn Mem. 2002; 9(2): 49–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParron C, Poucet B, Save E: Re-evaluation of the spatial memory deficits induced by hippocampal short lasting inactivation reveals the need for cortical co-operation. Behav Brain Res. 2001; 127(1–2): 71–79. PubMed Abstract \| Publisher Full Text\n\nPereira de Vasconcelos A, Klur S, Muller C, et al.: Reversible inactivation of the dorsal hippocampus by tetrodotoxin or lidocaine: a comparative study on cerebral functional activity and motor coordination in the rat. Neuroscience. 2006; 141(4): 1649–1663. PubMed Abstract \| Publisher Full Text\n\nRamirez-Amaya V, Balderas I, Sandoval J, et al.: Spatial long-term memory is related to mossy fiber synaptogenesis. J Neurosci. 2001; 21(18): 7340–7348. PubMed Abstract\n\nRamírez-Amaya V, Escobar ML, Chao V, et al.: Synaptogenesis of mossy fibers induced by spatial water maze overtraining. Hippocampus. 1999; 9(6): 631–636. PubMed Abstract \| Publisher Full Text\n\nRekart JL, Sandoval CJ, Bermudez-Rattoni F, et al.: Remodeling of hippocampal mossy fibers is selectively induced seven days after the acquisition of a spatial but not a cued reference memory task. Learn Mem. 2007; 14(6): 416–421. PubMed Abstract \| Publisher Full Text\n\nRekart JL, Sandoval CJ, Routtenberg A: Learning-induced axonal remodeling: evolutionary divergence and conservation of two components of the mossy fiber system within Rodentia. Neurobiol Learn Mem. 2007; 87(2): 225–235. PubMed Abstract \| Publisher Full Text\n\nRiekkinen P Jr, Ikonen S, Aura J, et al.: Tetrahydroaminoacridine and D-cycloserine fail to alleviate the water maze spatial navigation defect induced by hippocampal inactivation. Eur J Pharmacol. 1999; 366(1): 13–18. PubMed Abstract \| Publisher Full Text\n\nRolls ET: The mechanisms for pattern completion and pattern separation in the hippocampus. Front Syst Neurosci. 2013; 7: 74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRolls ET: A quantitative theory of the functions of the hippocampal CA3 network in memory. Front Cell Neurosci. 2013; 7: 98. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRudy JW: Contextual conditioning and auditory cue conditioning dissociate during development. Behav Neurosci. 1993; 107(5): 887–891. PubMed Abstract \| Publisher Full Text\n\nRudy JW, Morledge P: Ontogeny of contextual fear conditioning in rats: implications for consolidation, infantile amnesia, and hippocampal system function. Behav Neurosci. 1994; 108(2): 227–234. PubMed Abstract \| Publisher Full Text\n\nRudy JW, Stadler-Morris S, Albert P: Ontogeny of spatial navigation behaviors in the rat: dissociation of \"proximal\"- and \"distal\"-cue-based behaviors. Behav Neurosci. 1987; 101(1): 62–73. PubMed Abstract \| Publisher Full Text\n\nSaucier D, Cain DP: Spatial learning without NMDA receptor-dependent long-term potentiation. Nature. 1995; 378(6553): 186–189. PubMed Abstract \| Publisher Full Text\n\nSchapiro S, Salas M, Vukovich K: Hormonal effects on ontogeny of swimming ability in the rat: assessment of central nervous system development. Science. 1970; 168(3927): 147–150. PubMed Abstract \| Publisher Full Text\n\nSheng M, McFadden G, Greenberg ME: Membrane depolarization and calcium induce c-fos transcription via phosphorylation of transcription factor CREB. Neuron. 1990; 4(4): 571–582. PubMed Abstract \| Publisher Full Text\n\nSmith CA, Macdonald A, Holahan MR: Acute postnatal exposure to di(2-ethylhexyl) phthalate adversely impacts hippocampal development in the male rat. Neuroscience. 2011; 193: 100–108. PubMed Abstract \| Publisher Full Text\n\nTonkiss J, Shultz P, Galler JR: Long-Evans and Sprague-Dawley rats differ in their spatial navigation performance during ontogeny and at maturity. Dev Psychobiol. 1992; 25(8): 567–579. PubMed Abstract \| Publisher Full Text\n\nVida I, Frotscher M: A hippocampal interneuron associated with the mossy fiber system. Proc Natl Acad Sci U S A. 2000; 97(3): 1275–1280. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWartman BC, Gervais NJ, Smith C, et al.: Enhanced adolescent learning and hippocampal axonal projections following preadolescent spatial exposure to a water or dry maze. Brain Res. 2012; 1475: 37–48. PubMed Abstract \| Publisher Full Text\n\nWesterga J, Gramsbergen A: The development of locomotion in the rat. Brain Res Dev Brain Res. 1990; 57(2): 163–174. PubMed Abstract \| Publisher Full Text\n\nWhishaw IQ, Cassel JC, Jarrard LE: Rats with fimbria-fornix lesions display a place response in a swimming pool: a dissociation between getting there and knowing where. J Neurosci. 1995; 15(8): 5779–5788. PubMed Abstract\n\nWills TJ, Cacucci F, Burgess N, et al.: Development of the hippocampal cognitive map in preweanling rats. Science. 2010; 328(5985): 1573–1576. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang L, Blomgren K, Kuhn HG, et al.: Effects of postnatal thyroid hormone deficiency on neurogenesis in the juvenile and adult rat. Neurobiol Dis. 2009; 34(2): 366–374. PubMed Abstract \| Publisher Full Text\n\nComba R, Gervais N, Mumby D, et al.: Dataset 1 in: Emergence of spatial behavioral function and associated mossy fiber connectivity and c-Fos labeling patterns in the hippocampus of rats. F1000Research. 2015. Data Source\n\nComba R, Gervais N, Mumby D, et al.: Dataset 2 in: Emergence of spatial behavioral function and associated mossy fiber connectivity and c-Fos labeling patterns in the hippocampus of rats. F1000Research. 2015. Data Source\n\nComba R, Gervais N, Mumby D, et al.: Dataset 3 in: Emergence of spatial behavioral function and associated mossy fiber connectivity and c-Fos labeling patterns in the hippocampus of rats. F1000Research. 2015. Data Source\n\nComba R, Gervais N, Mumby D, et al.: Dataset 4 in: Emergence of spatial behavioral function and associated mossy fiber connectivity and c-Fos labeling patterns in the hippocampus of rats. F1000Research. 2015. Data Source" }	[ { "id": "9642", "date": "31 Jul 2015", "name": "Eric M. Stouffer", "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 conducted behavioral and neuroanatomical experiments to determine if hippocampal neuroanatomical development precedes the development of spatial learning abilities in an early postnatal period in male rats. They examined spatial learning abilities using the Morris water maze (MWM) task and a dry-version of the object in a novel location (ONL) task at postnatal days (PND) 16, 18, and 20 using separate groups of rats. Following performance on these tasks, the rats’ hippocampi were examined using synaptophysin staining and c-Fos staining to demonstrate neuroanatomical development compared to home-cage control groups of rats. Based on the results of the study, the authors suggest that neuroanatomical development of the hippocampus is present at PND 18 compared to PND 16, but that spatial learning abilities are not expressed until PND 20. Therefore, the authors state that hippocampal neuroanatomical development does indeed precede the development of spatial learning abilities in male rats.While I agree with the authors’ main conclusion for this study, I believe that there are some issues with the behavioral data that do not fully support other assertions made by the authors. For example, the authors conclude that the male rats showed spatial learning abilities at PND 20 in both the MWM task and the ONL task. In the MWM task, the authors point to the significant difference in escape latency between the PND 20 rats and the other two age groups. However, this difference can be explained as a result of a difference in swim speed (a physical development issue) rather than as a result of spatial learning. In fact, the authors do point this out in the discussion section as a “possibility”. I would argue that the physical development (swim speed) explanation for the difference in escape latency is not only a possibility, but rather is the most likely explanation of the significant difference in escape latency, especially given the non-difference in pathlength (arguably a more sensitive measure of spatial learning than escape latency) between the three age groups. Therefore, I do not think the authors can claim that the PND 20 rats showed spatial learning abilities in the MWM task. Perhaps if the authors had tested rats at PND 22 or PND 24 they may have seen stronger evidence of spatial learning in the MWM task.Regarding the ONL task, the authors used two main measures of spatial learning: the investigation ratio for the displaced object compared to the non-displaced object and the investigation time for the displaced and non-displaced objects. The investigation ratio data do support the conclusion that PND 20 rats show spatial learning compared to PND 16 and PND 18 rats. The investigation time data, however, did not show significant age differences. However, careful examination of Figure 2B leads me to believe that with a larger sample size that the PND 20 rats would demonstrate a significant increase in investigation time of the displaced compared to the non-displaced objects.Therefore, I do believe that PND 20 rats do show some evidence of spatial learning compared to PND 16 and PND 18 rats, just not as strong as the authors claim. This, however, does not detract from the fact that the authors do clearly demonstrate neuroanatomical development of the hippocampus prior to the development of spatial learning skills. That was the main point of the study, and was the authors’ main conclusion of the study. Overall, I do believe that this study does add to the existing literature.", "responses": [] }, { "id": "9641", "date": "20 Aug 2015", "name": "Bryan D. Devan", "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\nEditorial Note from F1000Research – 26th August 2015: This report has been updated to provide a link to the author's now published Correspondence Article. The phrase \"(forthcoming)\" was removed. No other changes have been made to the text of the report.The study by Comba et al.1 suggests a critical period of prenatal development (PND) in rodents during which neuronal mossy fiber growth in the hippocampus is associated with the emergence of spatial behavior. The researchers emphasize the PND 15-18 month period where this growth connecting the dentate gyrus to the CA3 cellular field is most prominent. It is also noted in the introduction that such grow may be related to the finding that neurogenesis-based processes specific to the hippocampus are also associated with the emergence of spatial learning. The researchers describe a “behavioral typology” control in their research design to dissociate effects based on non-cognitive motor demands from true cognitive information processing, which was supported by their data analyses of place learning in the Morris water maze versus spatial exploration in a dry land task (with swimming being more difficult than common ambulation). Despite the difference in motor demands, there was a common emergence of spatial behavior proficiency on each task at PND20. The researchers also found that enhanced mossy fiber projections, revealed by synaptophysin staining in the CA3 region, preceded the emergence of spatial behavior. Developmentally-dependent functional changes in cFOS positive cells were increased in all hippocampal subregions measured, while training-dependent changes were restricted to the CA3 and CA1 regions for groups trained in the water maze. The researchers conclude that mossy fiber connectivity along with enhanced function of the hippocampus precedes the emergence of spatial behavior at PND20, confirming their hypothesis of a sensitive period for hippocampal growth and the emergence of cognitive-spatial function.This research is very important and exciting when considering past studies of place learning ability in adult rats and research attempts to dissociate motor performance from true cognitive processing. In previous water maze studies of latent learning2,3 using passive placement on the goal in the water maze to view distal cues to form a cognitive map of the environment, mixed results and individual differences seem to obscure matters4-8, leading to one interpretation that movement and cognitive mapping may necessarily occur simultaneously, a finding that has been replicated in humans using a virtual version of the water maze9. Hence, for the water maze at least, movement through the environment seems to be an important constraint on highly proficient spatial learning and navigation. The use of a separate dry land task by Comba et al. with less motoric demands seems to be in agreement with the difference in motor demands between the original rodent version of the task10 requiring swimming and the human virtual version9 using minimal hand/finger movements to navigate. The role of dynamic movement during spatial tasks and the motoric demands have been topics of intense interest with the role of the hippocampus as a substrate for cognitive mapping11,12, path integration13-17, or the conductor of a symphony of dynamic movement and mapping18 as part of a larger neural network of brain systems19 have all been hotly debated theoretically over the years. The separation of cognitive and motor performance associated with developmentally-specific changes in hippocampal circuitry by Comba et al. is an exciting finding that may have important implications for a central role of the hippocampus in cognitive-spatial information processing as it emerges early in development.Consequently, the “behavioral typology” issue in the Comba et al. study, along with other aspects of their research design, is of critical importance in assessing the emerging role of the hippocampus in cognitive-spatial behavior. The following matters should be considered by all interested in this fascinating area of research in general, and in the Comba et al. study in particular.Specific ConsiderationsThe level of spatial proficiency in escaping to a hidden platform for PND20 rats given only 8 trials in the water maze is not comparable to the asymptotic level of escape latency performance (< 10 sec) observed in most water maze studies after considerably more extensive training. The one day water maze training paradigm is likely tapping into ventral hippocampal function in which the animals are just approaching the general location. After more training, dorsal hippocampus forms a more precise representation of the location. The authors should discuss this work20 and an analysis of their data (dorsal versus ventral) would be of interest. On a related matter, escape is not required, and exploration of an object at a novel “place” in the dry land task is very different from the typical water maze procedure, involving the presence of a local cue or familiar beacon (with different motivation). Some might argue that the lack of “true” spatial proficiency in the water maze is a flaw or weakness of the study; however, given the focus on the emergence of spatial behavior, and that well-learned escape responses are dependent on other brain regions21,22 that contribute/correlate with movement parameters19, it seems reasonable to expect less performance-wise using the escape latency measure than what is typically observed in most studies of this type. Consequently, spatial bias on a probe test might be considered as an alternative measure in future studies as it does not depend on a well-learned escape response that may be less hippocampal-dependent and more closely approximates the dwell time that is measured on the spatial exploration task. The object/place task is interesting. Integration with ideas about direct versus indirect measures of memory and the role of the hippocampus in one versus the other, and how these ideas relate to their different measures on this task would be of interest23. There are other obvious differences between the two tasks. Water maze for example is not disrupted by disorientation procedures but a dry-land version of spatial localization is disrupted by disorientation24. This work suggests that the representations are different as well, not just the behavioral topology. This should be discussed. The number of rats in each group for the water maze task is quite low (n = 5/group). A statement on the standardization of immunohistochemical procedures would be reassuring for those not familiar with the specific techniques used. Also, the use of an unbiased stereology technique should be considered. It appears that different behavioral procedures may have been conducted at different institutions. A statement on the time of day of testing and other procedural controls would provide reassurance that there are no threats to internal validity. More information on the recording and quantification of exploratory behavior (e.g., video recording, tracking, and interrater reliability) would be helpful for assessment and replication. It is interesting that the researchers note that PND18 rodents traveled a longer distance in the novel relocation task then the other groups (even though apparently PND20 rats exhibited more exploratory behavior). This finding may warrant further discussion to support the argument that both tasks assess potentially related cognitive functions. Reassurance that no statistical assumptions were violated (e.g., spherecity). Tukey HSDs are specifically based on studentized q related statistics but t-tests were reported. Though this may have been simply an alternative method (e.g. regression-based) to report the Tukey post-hoc results, possibly Bonferroni t-tests with separate mean square error denominators may be optimal for potential corrections to assumption violations.See full Correspondence Article by the Referees", "responses": [] } ]	1	https://f1000research.com/articles/4-396
https://f1000research.com/articles/4-362/v1	24 Jul 15	{ "type": "Correspondence", "title": "Evidence against pain specificity in the dorsal posterior insula", "authors": [ "Karen D. Davis", "M. Catherine Bushnell", "Gian Domenico Iannetti", "Keith St. Lawrence", "Robert Coghill", "M. Catherine Bushnell", "Gian Domenico Iannetti", "Keith St. Lawrence", "Robert Coghill" ], "abstract": "The search for a “pain centre” in the brain has long eluded neuroscientists. Although many regions of the brain have been shown to respond to painful stimuli, all of these regions also respond to other types of salient stimuli. In a recent paper, Segerdahl et al. (Nature Neuroscience, 2015) claims that the dorsal posterior insula (dpIns) is a pain-specific region based on the observation that the magnitude of regional cerebral blood flow (rCBF) fluctuations in the dpIns correlated with the magnitude of evoked pain. However, such a conclusion is, simply, not justified by the experimental evidence provided. Here we discuss three major factors that seriously question this claim.", "keywords": [ "Pain", "insula", "brain imaging", "ASL" ], "content": "\n\nThere are three major factors that we feel negate the claims of the recent study by Segerdahl et al.1 that the dorsal posterior insula (dpIns) is a pain-specific area of the brain.\n\nFirst, the evidence that the dpIns is specific is lacking based on the experimental design and data analysis employed. The methodological approach used by Segerdahl et al.1 was to induce an ongoing pain with capsaicin and then to correlate pain intensity ratings with brain perfusion changes using arterial spin labeling (ASL). ASL is an MRI-based perfusion method that can measure fluctuations in rCBF (akin to PET imaging) without the need for a stimulus, and so its application to study ongoing pain is promising. ASL has been previously used by others2,3 to identify acute and chronic pain-related changes in regional cerebral blood flow (rCBF) but the way Segerdahl et al.1 applied it has several shortcomings. The choice of Segerdahl et al.1 to collect multi-delay ASL data resulted in rCBF images sampled at infrequent intervals of ~45s, which represents a statistically challenging condition because of the small number of data collected. The control experiment using vibrotactile stimuli comprised a very short scan with even fewer data points in only seven subjects – a design that did not match the already low statistical power of the capsaicin experiment. Therefore, the analysis was underpowered and does not constitute a valid control for the pain experiment. This likely contributed to the minimal activation detected anywhere in the brain during the vibrotactile stimulation. The skin is richly innervated by rapidly adapting, low-threshold mechanoreceptors, so this absence of activation is of substantial concern. Even very early PET studies of regional cerebral blood flow (CBF) found robust vibrotactile activation of primary and secondary somatosensory cortex (S1, S2), and the adjacent posterior insula4,5. Most importantly, unlike previous investigations where CBF was directly and statistically compared between pain and innocuous stimulation to evaluate specificity of activation5,6, the Segerdahl et al.1 study performed no such key statistical comparison. Without this direct comparison, and in the absence of a control for vibration intensity, or for stimulus saliency, claims of specificity and pain intensity coding simply cannot be made7. This comparison is crucial given the evidence of a vast predominance of low threshold mechanoreceptive neurons in the posterior insula8 and robust vibrotactile activation of the insula (e.g., see 4).\n\nSecond, the proposition of a very specific “spot” dedicated to pain is critically dependent on the ability of the methodology to localize findings precisely. However, it is challenging to derive an accurate, group-averaged localization of activation within the dpIns given 1) the large intersubject anatomical variability of the insula, in particular the posterior gyri9 and 2) the method of realignment and morphing of brain anatomy into a common space to produce group maps. Inspection of the reported dpIns peak coordinate in the Juelich histologic atlas reveals that this peak activation has a 63% probability of being in the parietal operculum (S2, OP2), and only a 31% probability of being in the insular cortex. These areas are in close approximation, but S2 has a well-documented involvement in both nociceptive and innocuous somatosensory processing (e.g., see 8). No additional procedures were performed to functionally distinguish these two regions.\n\nThird, the interpretation of the findings and proposition of a specific pain center was made without taking into consideration a large body of scientific evidence addressing the brain mechanisms that contribute to pain. Theories of pain have been debated for centuries10, and we still do not know how pain is represented in the brain despite decades of searching for a pain specific brain center. This pursuit for a simple, single pain center however is no longer necessary given the enormity of human neuroimaging data indicating that there is no such dedicated center. Each and every brain area that contains nociceptive neurons also contains non-nociceptive neurons, and neuroimaging has shown that each brain area that responds to noxious stimuli can also respond to non-noxious stimuli11. Rather, multiple, converging lines of evidence strongly indicate that the experience of pain - as any other conscious experience - is constructed from highly distributed cortical processes5,12. For example, many brain regions exhibit activity related to pain intensity (e.g., 12,13). Furthermore, there are several clinical cases of preserved pain perception despite lesions of critical regions including the insula, anterior cingulate, and even the entire contralateral hemisphere14,15. Other studies have shown that interactions among multiple brain regions are critical for distinguishing a state of pain from other highly salient events16.\n\nIt is also useful to place the findings of Segerdahl et al.1 in context given the historical view of insular function. Morphological, physiological and imaging studies throughout the 1980s and 1990s, divided the insula into anterior agranular and posterior granular subregions, with pain-related function attributed to the anterior part, and a variety of other functions, including tactile recognition, attributed to the posterior part (e.g., see 8). Since that time, the anterior insula has been established to be part of a non-specific network related to attention and salience. In addition, there is anatomical and electrophysiological evidence for thermoreceptive processing in the dpIns via a spinal cord lamina 1 pathway17. Although neuroimaging has shown that the dpIns likely has a role in pain and intensity coding, it is critical to reiterate that intensity-coding has also been found for non-pain modalities in this region, including C-fiber mediated pleasant-touch18–20. The last decades have seen several theories of insula function being put forward21. This balanced view of potential dpIns functions is surprisingly absent from the discussion of Segerdahl et al.1. One important theory to consider, put forth by Apkarian’s group13, is that of the “how much” general magnitude-detector function of the insula. Another important theory developed by Craig and colleagues17, proposes the dpIns to be a center for interoceptive integration and awareness. Thus, there are several important issues22 that need to be considered to fully interpret the findings of Segerdahl et al.1. One assumption that drove the approach taken was that of the critical role of intensity-coding as being central to finding a “pain specific” center. We challenge this because although intensity certainly is one classic dimension of pain, there are many other dimensions including location, quality, and unpleasantness that together comprise the experience of pain. Furthermore, none of these dimensions are actually required for a fundamental feeling of pain (see the recent theory put forth by Davis et al.23).\n\nIn conclusion, the extensive evidence about the role of the dpIns is not considered by Segerdahl et al.1 and we note that they do not refute this evidence in their claim to have identified a novel, specific pain center in the dpIns. Such simplistic notions of a specific pain center are incorrect, and therefore dangerous at both an intellectual as well as a clinical level. Here, we suggest an alternate concept of the function of the dpIns based on previous theories and a large body of data that strongly indicate that the dpIns likely is involved in pain but overall is a non-specific perceptual way-station, rather than a specific pain centre. Failure to recognize that many regions activated during nociceptive stimulation are engaging in computational processes related to many things other than pain, lead to interpretations that are fraught with reverse inference11, and they encourage neurosurgeons to pursue lesions for pain control, an approach that has largely been shown to be ineffective since the 1960's24. Their promotion of the concept of a single spot in the brain for pain is even more surprising given the enormous amount of data emerging over the last decade showing the representation of brain function in functional networks, rather than “spots” and the newer view of a “dynamic pain connectome”25. Implications of their concept are far-reaching – from basic theories of pain, to development of “pain-o-meter” type diagnostic tests, to establishing a therapeutic target for clinical pain management26,27.", "appendix": "Author contributions\n\n\n\nKDD and RC prepared the first draft of the manuscript. All authors (KDD, MCB, GI, KSL, and RC) were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting 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\nReferences\n\nSegerdahl AR, Mezue M, Okell TW, et al.: The dorsal posterior insula subserves a fundamental role in human pain. Nat Neurosci. 2015; 18(4): 499–500. PubMed Abstract \| Publisher Full Text\n\nOwen DG, Clarke CF, Ganapathy S, et al.: Using perfusion MRI to measure the dynamic changes in neural activation associated with tonic muscular pain. Pain. 2010; 148(3): 375–86. PubMed Abstract \| Publisher Full Text\n\nLoggia ML, Kim J, Gollub RL, et al.: Default mode network connectivity encodes clinical pain: an arterial spin labeling study. Pain. 2013; 154(1): 24–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBurton H, Videen TO, Raichle ME: Tactile-vibration-activated foci in insular and parietal-opercular cortex studied with positron emission tomography: mapping the second somatosensory area in humans. Somatosens Mot Res. 1993; 10(3): 297–308. PubMed Abstract \| Publisher Full Text\n\nCoghill RC, Talbot JD, Evans AC, et al.: Distributed processing of pain and vibration by the human brain. J Neurosci. 1994; 14(7): 4095–108. PubMed Abstract\n\nIadarola MJ, Berman KF, Zeffiro TA, et al.: Neural activation during acute capsaicin-evoked pain and allodynia assessed with PET. Brain. 1998; 121(Pt 5): 931–47. PubMed Abstract \| Publisher Full Text\n\nPoldrack RA: Inferring mental states from neuroimaging data: from reverse inference to large-scale decoding. Neuron. 2011; 72(5): 692–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson CJ, Burton H: Somatic submodality distribution within the second somatosensory (SII), 7b, retroinsular, postauditory, and granular insular cortical areas of M. fascicularis. J Comp Neurol. 1980; 192(1): 93–108. PubMed Abstract \| Publisher Full Text\n\nRosen A, Chen DQ, Hayes DJ, et al.: A Neuroimaging Strategy for the Three-Dimensional in vivo Anatomical Visualization and Characterization of Insular Gyri. Stereotact Funct Neurosurg. 2015; 93(4): 255–64. PubMed Abstract \| Publisher Full Text\n\nMoayedi M, Davis KD: Theories of pain: from specificity to gate control. J Neurophysiol. 2013; 109(1): 5–12. PubMed Abstract \| Publisher Full Text\n\nIannetti GD, Salomons TV, Moayedi M, et al.: Beyond metaphor: contrasting mechanisms of social and physical pain. Trends Cogn Sci. 2013; 17(8): 371–8. PubMed Abstract \| Publisher Full Text\n\nCoghill RC, Sang CN, Maisog JM, et al.: Pain intensity processing within the human brain: a bilateral, distributed mechanism. J Neurophysiol. 1999; 82(4): 1934–43. PubMed Abstract\n\nBaliki MN, Geha PY, Apkarian AV: Parsing pain perception between nociceptive representation and magnitude estimation. J Neurophysiol. 2009; 101(2): 875–87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStarr CJ, Sawaki L, Wittenberg GF, et al.: Roles of the insular cortex in the modulation of pain: insights from brain lesions. J Neurosci. 2009; 29(9): 2684–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeinstein JS, Khalsa SS, Salomons TV, et al.: Preserved emotional awareness of pain in a patient with extensive bilateral damage to the insula, anterior cingulate, and amygdala. Brain Struct Funct. 2015. PubMed Abstract \| Publisher Full Text\n\nWager TD, Atlas LY, Lindquist MA, et al.: An fMRI-based neurologic signature of physical pain. N Engl J Med. 2013; 368(15): 1388–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCraig AD: Interoception: the sense of the physiological condition of the body. Curr Opin Neurobiol. 2003; 13(4): 500–5. PubMed Abstract \| Publisher Full Text\n\nOlausson H, Lamarre Y, Backlund H, et al.: Unmyelinated tactile afferents signal touch and project to insular cortex. Nat Neurosci. 2002; 5(9): 900–4. PubMed Abstract \| Publisher Full Text\n\nApkarian AV, Bushnell MC, Treede RD, et al.: Human brain mechanisms of pain perception and regulation in health and disease. Eur J Pain. 2005; 9(4): 463–84. PubMed Abstract \| Publisher Full Text\n\nSeminowicz DA, Davis KD: Interactions of pain intensity and cognitive load: the brain stays on task. Cereb Cortex. 2007; 17(6): 1412–22. PubMed Abstract \| Publisher Full Text\n\nMoayedi M, Weissman-Fogel I: Is the insula the “how much” intensity coder? J Neurophysiol. 2009; 102(3): 1345–7. PubMed Abstract \| Publisher Full Text\n\nMoayedi M: All roads lead to the insula. Pain. 2014; 155(10): 1920–1. PubMed Abstract \| Publisher Full Text\n\nDavis KD, Kucyi A, Moayedi M: The Pain Switch: An “ouch” detector. Pain. 2015; in press.\n\nTasker RR: History of lesioning for pain. Stereotact Funct Neurosurg. 2001; 77(1–4): 163–5. PubMed Abstract \| Publisher Full Text\n\nKucyi A, Davis KD: The dynamic pain connectome. Trends Neurosci. 2015; 38(2): 86–95. PubMed Abstract \| Publisher Full Text\n\nReardon S: Neuroscience in court: The painful truth. Nature. 2015; 518(7540): 474–6. PubMed Abstract \| Publisher Full Text\n\nSeminowicz D, Pustilnik A, Rigg S, et al.: Panel 1: Legal and Neuroscientific Perspectives on Chronic Pain. J Health Care Law Policy. 2015; 18: 207–35. Reference Source" }	[ { "id": "9663", "date": "28 Jul 2015", "name": "David Borsook", "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\nWhen Research Reports may be Painful - Open Discourse should PrevailKaren Davis and coauthors, including notable researchers in the pain imaging field, including Robert Coghill and Catherine Bushnell 1, argue strongly against the concept reported in Nature Neuroscience by Segerdahl and colleagues 2; a paper that has quickly proven to be highly controversial for many in the field. The Segerdahl report utilizes an MRI technique known as arterial spin labeling (ASL) to measure quantitative cerebral blood flow as a surrogate marker for neuronal activation 3, which in this case, was combined with a model of capsaicin application to the skin to induce a hypersensitivity to heat stimuli in healthy patents 4,5 as a surrogate model of allodynia in chronic pain.\n\nMany of the arguments raised by Davis et al. relate to the validity of the Segerdahl report, and are of a technical nature. Hopefully, these technical problems can be easily addressed (e.g., in future experiments) or challenged by an understanding of the field and its limitations (as Davis et al so eloquently do - see First and Second Arguments in their paper). What is still unclear is why Segerdahl and colleagues, seem to have overlooked considerable prior work using ASL in experimental pain 6-8, post-surgical pain 9-11, and chronic pain 12-16. There is also a noticeable lack of consideration for the limitations of both the imaging technique and the experimental pain model 5, which itself in healthy volunteers, has some issues relating to reproducibility and its clinical relevance 4. The concept of discovering or defining a pain specific area in pain patients has to be understood in terms of a long history searching for such an area to target with various therapeutic modalities, most notably, neurosurgery has led the charge. The evaluation of putative pain specific areas in acute pain models probably has little if any bearing on the clinical condition of chronic pain.\n\nMore modern concepts of brain-wide integrative processes are now in vogue. Davis authors use the definition of “Pain-Connectome” 17 which while conceptually is not new, adds to the growing literature of Connectomics 18, and should help rid the often used and probably not useful concept or term ‘pain matrix’ from use, given the modern understanding of brain networks. At best, the Segerdahal contribution has raised a vibrant discussion in the field, at is worst it is setting the field back not only because of its purported methodological inaccuracy (as evaluated by Davis et al.), lack of acknowledgement of what has come before, perhaps being too enthusiastic about the results and therefore pushing a notion that is unlikely to be true – finding a single brain area that is a pain specific region. Publications in high impact journals such as Nature Neuroscience (http://www.nature.com/neuro/index.html) carry a great responsibility, since they can (and usually) contribute to a field moving forward, or in a few cases the field becoming ‘stuck’ because of potentially false concepts that then take time for any field to undo. Hopefully this is not one of those issues relating to how high profile papers may occasionally be problematic as previously commented on, for example: “How journals like Nature, Cell and Science are damaging science” (http://www.theguardian.com/commentisfree/2013/dec/09/how-journals-nature-science-cell-damage-science). Having the finding being replicated in the context of chronic pain conditions will be interesting to observe; perhaps Segerdahl et al., have these in the planning stage. This is of particular importance since having reproducible data from chronic pain patients may provide important therapeutic opportunities. What is still to be defined, through a more detailed connectomic understanding, is whether such brain areas may be important through integration of processes such as chronic pain with other brain regions, in remodeling or reconstituting normalization of brain circuits following treatment for chronic pain. Such a notion could perhaps be the real excitement of where the field is headed. A healthy discourse in science can only lead to further evolution in the field and should be open and honest. I believe Davis and colleagues have made such a contribution in their review of the Segerdahl paper.", "responses": [] }, { "id": "9627", "date": "07 Sep 2015", "name": "Apkar Vania Apkarian", "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 correspondence by K. Davis and colleagues regarding the recently published study by Segerdahl et al. raises important and valid concerns regarding the conclusion of the paper. The issue is important as the report appeared in a high visibility journal, and the authors make the strong claim that they have identified a single \"pain center\" in the cortex. The latter has been a quest sought by many pain researchers since the advent of neuroimaging technology. Most importantly a lack of adequate statistical power and a proper control are the most obvious technical weaknesses pinpointed by Davis et al.. Perhaps it would be informative to elaborate on this issue, specifically regarding how an underpowered study can lead into discovering a brain \"specific center\" for pain perception, the validity of which is doubted by senior scientists in the field. Neuroimaging studies, whether based on BOLD or ASL, when attempting to identify brain activity relative to a task commonly first identify in each participant brain activity related to the task, average these patterns across all participants, and then use a set of statistical criteria to determine what brain areas are statistically significantly conveying information about the task. In the present study only one brain region passed the specific criteria used and thus we have a single brain area related to the task. Not surprisingly the area is the posterior insula. A brain region that 10 years ago was described to be most commonly observed activated area to any painful stimuli1 and currently in a PubMed term-based meta-analysis, neurosynth (www.neurosynth.org), it is identified (together with the secondary somatosensory cortex) as the region with highest reverse inference probability (z-score > 13.0) for association with the term “pain”, based on 420 publications. Thus, there is good evidence for this region being involved in pain related studies, and in an underpowered study where high thresholds becomes necessary to identify brain activity it is not surprising that only this one region is identified. Additionally one fully expects that with increased power most of the extended set of brain regions identified in neurosynth, whether called ‘neuromatrix’ or ‘pain connectome’, would also be observed independent of the neuroimaging technology used (tip of the iceberg phenomenon). A simple analogy can be derived from astronomy. Modern telescopes provide us with a picture of the sky full of millions of stars and galaxies. However, even today if we look at the sky by a telescope manufactured by Galileo, or having an equivalent resolution, we will still only observe the handful of stars that Galileo was describing 450 years ago.The other important issue of the paper by Segerdahl et al. regards the conceptual implications, an issue that Davis et al. mention and again is important to elaborate. In the effort of proving that pain is in and of itself a unique sensory system, a large number of pain scientists have espoused the notion of dedicated real estate in the cortex for pain. Yet, isolation of such a single region has the strong implication that the conscious, subjective, and affective perception of pain is all captured in this one brain area. The latter implies the thought experiment of excising the region and placing it in a dish (perhaps also keeping all the tissue that connects it to the periphery), with which act we can claim to recapitulate pain consciousness in a dish, which seems absurd and inconsistent with modern theories relating the brain to perception2.Borsook’s commentary on the problems associated with Segerdahl et al. publication is also very astute3. He points that the competition to publish in high end journals pushes the scientist into making more extravagant conclusions than even the author herself or himself actually does not trust. Yet ultimately responsibility rests on the peer review process, and the latter is not guaranteed to be full proof.", "responses": [] }, { "id": "9718", "date": "12 Oct 2015", "name": "Joel D. Greenspan", "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 discussion initiated by Davis et al. 1, in their critical commentary of Segendahl et al.’s publication in Nature Neuroscience 2 should be recognized as a valuable contribution to our attempt to understand how brain function relates to pain perception. Davis et al., and following commentaries from Apkarian3 and Borsook4, make salient points regarding methodological and interpretive limitations of Segendahl et al.’s published work. One key point is how a weakly powered imaging study can lead to a limited result. In this case, finding that only the dpIns shows a significant relationship to evoked pain intensity is not enough to conclude that the dpIns is necessary and sufficient for pain intensity perception (aka, constituting a “pain center”). Davis et al. and Apkarian specifically decry the concept of a “pain center” in the brain, and I believe our field has rejected this concept as a useful construct. But, to be fair, the term “pain center” does not appear in the Segendahl et al. paper. Two statements in the concluding part of their article are as follows:“Thus, a growing body of literature suggests that a subsection of the posterior insula is both anatomically and functionally well suited to serve a primary and fundamental role in pain processing.”And:\n\n“… we were able to identify the dpIns as subserving a fundamental role in pain and the likely human homolog of the nociceptive region identified from animal studies.”I don’t find these unreasonable statements, although one could argue over what the meaning of “fundamental” is in this context. However, one unjustified claim appears in the abstract: “We exploited arterial spin-labeling quantitative perfusion imaging and a newly developed procedure to identify a specific role for the dorsal posterior insula (dpIns) in pain.” As other commentators noted, Segendahl et al’s work does not provide evidence of specificity of function, either in terms of the dpIns being the only brain region of consequence (note the statistical power issues raised by other referees), nor in terms of specificity of dpIns function (as evidenced by many papers showing non-pain protocol engagement of the region).\n\nSo, if we can agree to abandon the concept of “P1” (primary pain cortex), should we still consider the concept of “N1” – primary nociceptive cortex? We don’t regard V1 or A1 as “centers” for sight or hearing. Rather, they are recognized as essential cerebral areas for early sensory processing leading to their respective perceptual experiences, functioning in concert with other cerebral regions.\n\nIs there a value in looking for an analogous cerebral cortical region for nociception, and if so, would the real estate identified by Segendahl et al. qualify? (I say this in full recognition that the cerebral region identified in their paper, as in nearly all imaging studies targeting this region, is not clearly distinguishable as either dorsal posterior insula vs. parietal operculum/S2.) One very critical difference between the nociceptive vs. visual or auditory systems is that there are multiple thalamo-cortical pathways for nociceptive signals to reach the cortex, as shown most directly in primate studies such as Dum et al.5. This begs the question of whether there is a “primary” nociceptive cortex. In 1999, Howard Fields noted that not everything unpleasant is painful, and that we should have “…a term for the psychophysical property of an unpleasant somatic sensation that allows us to identify it as pain.”6 He offered up the term “algosity”, and suggested that pain could be considered a combination of algosity and unpleasantness. One could also think of algosity as the perceptual consequence of recognizing nociceptor activation. If this concept has value, and I believe that many of us in the field do, should we be able to identify a “N1” cortex that has a fundamental or even an essential role in algosity?", "responses": [] } ]	1	https://f1000research.com/articles/4-362
https://f1000research.com/articles/4-171/v1	29 Jun 15	{ "type": "Case Report", "title": "Case Report: Perioperative management of a pregnant poly trauma patient for spine fixation surgery", "authors": [ "Rashmi Vandse", "Meghan Cook", "Sergio Bergese", "Meghan Cook", "Sergio Bergese" ], "abstract": "Trauma is estimated to complicate approximately one in twelve pregnancies, and is currently a leading non-obstetric cause of maternal death. Pregnant trauma patients requiring non-obstetric surgery pose a number of challenges for anesthesiologists. Here we present the successful perioperative management of a pregnant trauma patient with multiple injuries including occult pneumothorax who underwent T9 to L1 fusion in prone position, and address the pertinent perioperative anesthetic considerations and management.", "keywords": [ "Spine surgery", "pneumothorax", "neuro anesthesia", "pregnant", "poly trauma" ], "content": "Introduction\n\nPerioperative management of a pregnant patients requiring non-obstetric surgery is always challenging for an anesthesiologist. The literature documenting anesthetic, surgical and obstetric management of pregnant poly trauma victims undergoing spine surgery in prone positioning is limited. We present a case of a pregnant polytrauma victim with multiple injuries who subsequently underwent spine fixation surgery in prone position and discuss pertinent anesthetic issues and management.\n\n\nCase presentation\n\nThis is a case of a previously healthy 32 year old female, who presented while 17 weeks pregnant as a level 2 trauma following a motor vehicle collision. She had sustained multiple injuries including Grade II liver laceration, pelvic fracture, bilateral clavicle fractures, C1 transverse process fracture, T11 vertebral body burst fracture, R rib 1–10 fractures, L 1st and 2nd rib fractures, bilateral small pneumothoraces and right pulmonary contusion. On examination, the patient was moderately built and nourished. She was 66 inches tall and weighed 136 pounds. Her vital signs on admission showed: heart rate of 96 beats/minute, respiratory rate of 14–18 breaths/minute, blood pressure of 108/56 mmHg, and O2 saturation of 98% on 2–3 liters of oxygen through nasal cannula. She remained hemodynamically stable throughout and did not show any signs of respiratory distress, although she did have some trouble with coughing and clearing respiratory secretions. A preoperative chest X-ray demonstrated complete collapse of the left lung (Figure 1). The small pneumothorax which was discovered in a computed tomography (CT) of the chest, however, was not apparent in the chest X-ray. After a multidisciplinary discussion, because of the unstable spine fracture, it was decided to perform a posterior T9-L1 fusion under general anesthesia. Her lab values were otherwise normal except for hemoglobin of 9.5 and hematocrit of 27.4.\n\nGeneral anesthesia was induced with propofol, lidocaine, fentanyl and succinylcholine. Following intubation, bronchoscopy was performed and the airway was suctioned given her preoperative chest X-ray. The radial artery was cannulated for hemodynamic monitoring. She was then carefully positioned prone on an open frame Jackson table. Special care was taken to avoid any pressure on the abdomen and all the other pressure points were checked and padded. Anesthesia was maintained with propofol (50 mcg/kg/min) and remifentanil (0.05–0.12 mcg/kg/min) infusions along with 1.0% sevoflurane in 50% oxygen. Phenylephrine was used to support her blood pressure as needed. She remained hemodynamically stable throughout the procedure. She was ventilated with a small tidal volume (300–350 ml) and her peak pressure was closely monitored, which stayed less than 20 cm of H20 throughout. CT-based image guidance was mostly used by the surgeons to limit the intraoperative fluoroscopy. She received 1300 ml of crystalloids and 500 ml of albumin. She produced 400 ml of urine and lost approximately 200 ml of blood. Total duration of anesthesia was approximately 4 hours. She was successfully extubated at the end of the procedure. She remained stable post operatively. However, she did require 22 days of inpatient care due to multiple injuries she had sustained during the trauma. She was successively discharged home. She later came back at term and delivered a healthy baby by elective Caesarean section under general anesthesia.\n\n\nDiscussion\n\nTrauma is estimated to complicate approximately one in twelve pregnancies, and is currently a leading non-obstetric cause of maternal death; moreover, maternal death remains the most common cause of fetal demise1–4. Extensive multidisciplinary planning between the surgeons, intensivists, anesthesiologists and obstetricians is essential to ensure fetal and maternal well-being throughout the perioperative period. The anesthetic considerations of this case were many. We had a pregnant patient requiring spine fixation surgery in prone position. Her management was further convoluted by the associated injuries, most importantly multiple b/l rib fractures and small pneumothoraces.\n\n\nObstetric concerns\n\nOptimum management requires a thorough understanding of normal maternal-fetal physiology, maternal physiologic adaptation to pregnancy and altered drug pharmacodynamics and pharmacokinetics. The increased oxygen requirements, decreased functional residual lung capacity and increased risk for aspiration associated with pregnancy complicates perioperative management by decreasing the time available and the margin of safety. These changes are extensively reviewed in many textbooks and review articles5–7. The gestational age and maturity of the fetus as well as the acute maternal injuries were taken into account when formulating the operative plan. As the patient was in her second trimester, delivery of the fetus was not a feasible option. The obstetric team was consulted who advised fetal heart tones (FHT) monitoring pre and post operatively.\n\nThe deleterious effects of anesthesia on the human fetus have been considered for many years. As such, any drug has the potential to negatively affect the developing human fetus depending on the dose and the time of exposure and there is no “ideal anesthetic agent”5–8.\n\nIt is therefore most prudent to postpone elective surgical procedures until after pregnancy or, if possible, to avoid during the first trimester5–6. There is no convincing evidence that any particular anesthetic drug at clinically used doses is clearly dangerous to the human fetus5–6. Anesthetic goals are to prevent fetal asphyxia by maintaining maternal oxygenation, ventilation and hemodynamic stability and avoid factors that might cause reduction in the uteroplacental perfusion or compromise fetal gas exchange5–7. Large survey studies on women who underwent surgery during pregnancy suggest no increase in congenital anomalies among their offspring but rather an increase in the risk for abortions, growth restriction for reasons mostly attributed to the requirement for surgery but not anesthetic administration5–6. In our patient, we used a combination of intravenous (IV) anesthetics (propofol and remifentanil) with 1% sevoflurane in order to permit Somatosensory Evoked Potential (SSEP) and electromyogram (EMG) monitoring and allowed for rapid awakening at the end of the procedure. In clinical practice both propofol and remifentanil have been used safely in pregnant patients10–13. Caution must be exercised while using propofol infusion for long procedures (>10 hours). Two cases of prolonged IV anesthesia with propofol during pregnancy (14–18 h) resulted in mild metabolic acidosis14. Maintenance of normal maternal blood pressure is imperative because of the relative passive dependence of the uteroplacental circulation and also to avoid spinal cord ischemia. As such a reduction in maternal arterial pressure causes reduced uteroplacental blood flow and fetal ischaemia. We used phenylephrine infusion to maintain MAP above 70 mmHg based on the earlier studies supporting better maternal cardiovascular stability and improved neonatal acid–base status when phenylephrine was used to treat maternal hypotension5–6. Alveolar ventilation increases by 25% to 30% during pregnancy. This increase results in chronic respiratory alkalosis with a Paco2 of 28 to 32 mmHg and a slightly alkaline pH (approximately 7.44), and decreased levels of bicarbonate and buffer base. The Paco2 should be kept in the normal range for pregnancy as both hyper and hypocapnia can compromise the fetus5–6.\n\n\nSurgical positioning\n\nSome of the case reports and small case series have described good fetal outcome among gestational women who had spinal surgery during their pregnancy15–20. Care must be taken while positioning the patient. There are well described and predictable changes in physiology associated with the prone position, including a decrease in cardiac index20. It is attributed to increased intra-thoracic pressures causing a decrease in arterial filling and subsequent increase in sympathetic activity via the baroreceptor reflex. Mean arterial pressure remains the same, owing to an increase in systemic vascular resistance. Aortocaval compression must be avoided as this can lead to significant reductions in maternal cardiac output, systemic blood pressure, and uterine blood flow. This can also lead to epidural venous engorgement and increased surgical bleeding. The study by Nakai et al.21 showed that when pregnant patients were positioned prone by letting the abdomen hang free, there was actually better relief of compression on the large maternal vessels by the gravid uterus when compared to sitting or lateral positions. We used a Jackson frame which helped to avoid any direct compression of the fetus and the great vessels.\n\n\nRib fractures and pneumothorax\n\nThe presence of multiple b/l rib fractures and b/l occult (small) pneumothorax impacted our decision making because of the increased risk for expanding the pneumothorax with positive pressure ventilation. The management of an occult or clinically insignificant pneumothorax in acute trauma patients is controversial. The development of tension pneumothorax intraoperatively requiring emergency chest tube insertion has been reported22. In a prospective randomized study by Enderson et al., 8 out of 21 patients in the observation group demonstrated progression of the occult pneumothorax, 3 of which developed a tension pneumothorax23. They suggested that mechanically ventilated patients with an occult pneumothorax should be managed with a thoracostomy tube. On the contrary, there appears to be a growing recognition that vast majority of cases with an occult pneumothorax can be safely treated without placing a thoracostomy tube in non-ventilated or even mechanically ventilated patients24–27. Hence in the absence of clear cut evidence one must consider the risk versus benefit while making the clinical decision. Thoracostomy is also associated with major complications and has been reported to increase the overall mortality rate26,27. As such, in this case, special attention was paid to the peak airway pressures and plethysmography. Additionally, the general surgery team was made aware of the patient, and a chest tube kit was kept in the room, although it was not needed during this case. Our patient also had decreased aeration on the L side of her lung in the preoperative chest X-ray which was thought to be due to an inability to clear the secretions as a result of splinting. Flexible bronchoscopy and aspiration of secretion was performed after intubation. Postoperatively, meticulous attention was paid to adequate pain control and incentive spirometry, which allowed further improvement in the lung aeration. She was also placed on intermitted BiPAP as needed. Thus, we were able to avoid prolonged intubation as well as chest tube insertion.\n\n\nRadiation exposure\n\nAdditional consideration was also given to radiation exposure. Radiographic studies have shown that radiation exposure poses the greatest teratogenic risk in early pregnancy when organogenesis occurs (2–7 weeks)5,9. Exposure after organogenesis may cause growth restriction, microcephaly, and childhood cancer5,9,28. Fetal risk of malformations is considered to be low with total radiation exposures of less than 50 to 100 mGy (5 to 10 rads)28. In contrasts to the negligible risk of teratogenicity, observational studies suggest that there is a slightly higher risk of childhood cancer at radiation doses greater than or equal to 10 mGy29. Therefore, exposure to radiation should be minimized whenever possible. Computed tomography produces higher levels of radiation exposure than plain radiographs, but even abdominal and pelvic CT scanning usually produces estimated fetal exposures below those typically associated with adverse fetal/neonatal outcomes5,29. In our case, CT-base image guidance was mostly used by the surgeons to limit the intraoperative fluoroscopy.\n\n\nConclusion\n\nSuccessful surgical intervention was accomplished without any major morbidity or mortality due to thorough systematic assessment of individual issues and stratification of management priorities. The ultimate objective is to provide safe anesthesia to the mother while concurrently minimizing the risk of preterm labor or fetal demise. In our case, the patient was successively discharged home and delivered a healthy baby at term without any complications.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and/or clinical images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nRV and MC were involved in preparing the first draft of the manuscript. RV and SB did further literature review and modifications. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting 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\nReferences\n\nMirza FG, Devine PC, Gaddipati S: Trauma in pregnancy: a systematic approach. Am J Perinatol. 2010; 27(7): 579–586. PubMed Abstract \| Publisher Full Text\n\nEl Kady D: Perinatal outcomes of traumatic injuries during pregnancy. Clin Obstet Gynecol. 2007; 50(3): 582–591. PubMed Abstract \| Publisher Full Text\n\nMendez-Figueroa H, Dahlke JD, Vrees RA, et al.: Trauma in pregnancy: an updated systematic review. Am J Obstet Gynecol. 2013; 209(1): 1–10. PubMed Abstract \| Publisher Full Text\n\nDaponte A, Khan N, Smith MD, et al.: Trauma in pregnancy. S Afr J Surg. 2003; 41(2): 51–54; discussion 55. PubMed Abstract\n\nNaughton NN, Cohen SE: Nonobstetric surgery during pregnancy. Chestnut DH ed., Obstetric Anesthesia: principles and practice. Philadelphia: Elsevier Mosby, 2004; 255–72.\n\nReitman E, Flood P: Anaesthetic considerations for non-obstetric surgery during pregnancy. Br J Anaesth. 2011; 107(Suppl 1): i72–i78. PubMed Abstract \| Publisher Full Text\n\nReisner LS, Kuczkowski KM: Anesthetic considerations for complicated pregnancies. In: Creasy, Resnik, Iams (eds). Maternal-fetal medicine. Elsevier Philadelphia. 2003; 1243–1260.\n\nCohen-Kerem R, Railton C, Oren D, et al.: Pregnancy outcome following non-obstetric surgical intervention. Am J Surg. 2005; 190(3): 467–473. PubMed Abstract \| Publisher Full Text\n\nWang LP, Paech MJ: Neuroanesthesia for the pregnant woman. Anesth Analg. 2008; 107(1): 193–200. PubMed Abstract \| Publisher Full Text\n\nGregory MA, Gin T, Yau G, et al.: Propofol infusion anaesthesia for caesarean section. Can J Anaesth. 1990; 37(5): 514–20. PubMed Abstract \| Publisher Full Text\n\nHeesen M, Klöhr S, Hofmann T, et al.: Maternal and foetal effects of remifentanil for general anaesthesia in parturients undergoing caesarean section: a systematic review and meta-analysis. Acta Anaesthesiol Scand. 2013; 57(1): 29–36. PubMed Abstract \| Publisher Full Text\n\nVan de Velde M, Teunkens A, Kuypers M, et al.: General anaesthesia with target controlled infusion of propofol for planned caesarean section: maternal and neonatal effects of a remifentanil-based technique. Int J Obstet Anesth. 2004; 13(3): 153–8. PubMed Abstract \| Publisher Full Text\n\nMertens E, Saldien V, Coppejans H, et al.: Target controlled infusion of remifentanil and propofol for cesarean section in a patient with multivalvular disease and severe pulmonary hypertension. Acta Anaesthesiol Belg. 2001; 52(2): 207–9. PubMed Abstract\n\nHilton G, Andrzejowski JC: Prolonged propofol infusions in pregnant neurosurgical patients. J Neurosurg Anesthesiol. 2007; 19(1): 67–8. PubMed Abstract \| Publisher Full Text\n\nCohen-Kerem R, Railton C, Oren D, et al.: Pregnancy outcome following non-obstetric surgical intervention. Am J Surg. 2005; 190(3): 467–73. PubMed Abstract \| Publisher Full Text\n\nHan IH, Kuh SU, Kim JH, et al.: Clinical approach and surgical strategy for spinal diseases in pregnant women: a report of ten cases. Spine (Phila Pa 1976). 2008; 33(17): E614–619. PubMed Abstract \| Publisher Full Text\n\nBrown MD, Levi AD: Surgery for lumbar disc herniation during pregnancy. Spine (Phila Pa 1976). 2001; 26(4): 440–43. PubMed Abstract \| Publisher Full Text\n\nKim HS, Kim SW, Lee SM, et al.: Endoscopic discectomy for the cauda equina syndrome during third trimester of pregnancy. J Korean Neurosurg Soc. 2007; 42(5): 419–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJea A, Moza K, Levi AD, et al.: Spontaneous spinal epidural haematoma during pregnancy: case report and literature review. Neurosurgery. 2005; 56(5): E1156; discussion E1156. PubMed Abstract\n\nEdgcombe H, Carter K, Yarrow S: Anaesthesia in the prone position. Br J Anaesth. 2008; 100(2); 165–83. PubMed Abstract \| Publisher Full Text\n\nNakai Y, Mine M, Nishio J, et al.: Effects of maternal prone position on the umbilical arterial flow. Acta Obstet Gynecol Scand. 1998; 77(10): 967–69. PubMed Abstract \| Publisher Full Text\n\nRankin D, Mathew P, Kurnutala L, et al.: Tension Pneumothorax during surgery for Thoracic Spine Stabilization in Prone Position: A Case Report and Review of Literature. J Investig Med High Impact Case Rep. 2014; 2(2). Publisher Full Text\n\nEnderson BL, Abdalla R, Frame SB, et al.: Tube thoracostomy for occult pneumothorax: a prospective randomized study of its use. J Trauma. 1993; 35(5): 726–29; discussion 729–30. PubMed Abstract\n\nBrasel KJ, Stafford RE, Weigelt JA, et al.: Treatment of occult pneumothoraces from blunt trauma. J Trauma. 1999; 46(6): 987–990; discussion 990–1. PubMed Abstract\n\nWilson H, Ellsmere J, Tallon J, et al.: Occult pneumothorax in the blunt trauma patient: tube thoracostomy or observation? Injury. 2009; 40(9): 928–931. PubMed Abstract \| Publisher Full Text\n\nOuellet JF, Trottier V, Kmet L, et al.: The OPTICC trial: a multi-institutional study of occult pneumothoraces in critical care. Am J Surg. 2009; 197(5): 581–86. PubMed Abstract \| Publisher Full Text\n\nBall CG, Kirkpatrick AW, Feliciano DV: The occult pneumothorax: what have we learned? Can J Surg. 2009; 52(5): E173–9. PubMed Abstract \| Free Full Text\n\nMcCollough CH, Schueler BA, Atwell TD, et al.: Radiation exposure and pregnancy: when should we be concerned? Radiographics. 2007; 27(4): 909–917; discussion 917–8. PubMed Abstract \| Publisher Full Text\n\nValentin J: Pregnancy and medical radiation, ICRP Publication 84: Approved by the Commission in November 1999. Ann ICRP. 2000; 30(1): 1–43. Publisher Full Text" }	[ { "id": "9420", "date": "10 Jul 2015", "name": "Michael Paech", "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 case report with an extended discussion of perioperative management in this setting. Although the patient was pregnant, she was in the early second trimester, when the relevance of some physiological changes of pregnancy and the absence of a distended abdomen or a viable fetus mean that the management is mainly specific to trauma issues - pregnancy issues are not well illustrated. Consequently the educational value of this report is diminished. My main suggestions are:Delete the sentence in the case presentation “On examination…..” Add a sentence as to whether an attempt made to confirm fetal viability before and after surgery, and by what method. The Discussion section on “obstetric concerns” is too long, a regurgitation of standard teaching and is not specific to this patient. This should be shortened. Likewise much of the first half on “surgical positioning” is not relevant to this patient. Please check for quality of language. Examples of poor use include “of a pregnant patients” ; “the obstetric team was consulted who…” ; “multiple b/l rib……”; “3 of which…”", "responses": [ { "c_id": "1469", "date": "24 Jul 2015", "name": "Rashmi Vandse", "role": "Author Response", "response": "We would like to thank our reviewer for reviewing our case report. We have modified it as per the suggestion. As pointed out by the reviewer, our patient was in her early 2 nd trimester hence lacked the viable fetus or the distended abdomen. Physiologic changes of pregnancy are less pronounced. However as suggested in the literature, surgery during the early pregnancy carry increased risk for miscarriage and it is important to take all precautions to maintain adequate maternal oxygenation and hemodynamic stability which was complicated in our patient due to her polytrauma status. Delete the sentence in the case presentation “On examination…..”- modifiedAdd a sentence as to whether an attempt made to confirm fetal viability before and after surgery, and by what method.- FHT were monitored both pre and post operatively which was reassuring.The Discussion section on “obstetric concerns” is too long, a regurgitation of standard teaching and is not specific to this patient. This should be shortened. Likewise much of the first half on “surgical positioning” is not relevant to this patient.- Has been shortened. However some of the important obstetric issues have been retained for educational purposes. Surgical positioning – Has been shortened. Quality of language has been checked and corrections have been made." } ] }, { "id": "9227", "date": "14 Jul 2015", "name": "Bernard Wittels", "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 report of an exceptionally unusual presentation of a parturient with multiple trauma needing urgent neurosurgery. The authors do a thorough review of important aspects of parturient physiology, as well as the medical and surgical considerations involved. Still, the authors seem to have followed the typical care of any neurosurgical patient for this procedure and were just serendipitous in having a Jackson-type operating table with an open suspended sling that prevented any untoward pressure on the fetus or fetal circulation while in the prone position. If a more unique approach to neurosurgery or neuroanesthesia in the parturient was introduced and proven successful, this report would have much greater value to its readers.The main scientific point is that the positioning and monitoring of a non-pregnant patient for neurosurgery serves well for the polytrauma pregnant patient for many reasons as well described, though not directly. Also missing is the inability to easily monitor the fetal status intraoperatively, and the inability to perform emergent cesarean delivery in this position if fetal distress occurs in a viable fetus. Are there modifications in the anesthetic or surgical approach that could accommodate these factors?", "responses": [ { "c_id": "1468", "date": "24 Jul 2015", "name": "Rashmi Vandse", "role": "Author Response", "response": "We thank our reviewer for their input. We have modified the case report as suggested. As the patient was in her early pregnancy, we did not have significant challenge in positioning her prone. Also since the fetus was non viable we didn’t encounter any need to modify her positioning. As suggested in the literature, surgery in prone position has been carried out safely during the first and early second trimester. As pointed out by the reviewer , one of the drawbacks of prone positioning is inability to monitor fetal status or perform emergent cesarean section for fetal distress in a viable fetus. Spinal surgeries have been performed successfully under epidural anesthesia and lateral positioning has also been utilized safely during the late second and third trimester of pregnancy15. These alternative surgical approaches must be discussed with the surgeon whenever feasible." } ] } ]	1	https://f1000research.com/articles/4-171
https://f1000research.com/articles/4-143/v1	05 Jun 15	{ "type": "Research Note", "title": "Sensorimotor learning deficits observed in children with sports-related concussion", "authors": [ "Jinsung Wang", "Danny Thomas", "Young Ik Cho", "Danny Thomas", "Young Ik Cho" ], "abstract": "Mild traumatic brain injury, commonly known as concussion, occurs frequently in children while playing sports. Whereas the negative impact of sport-related concussions on cognitive function is well known, the effects of concussions on sensorimotor function, especially in children, remain largely unknown. In this observational study, we investigated the association between neurocognitive function, assessed using a trail making test, and sensorimotor function, assessed using a visuomotor adaptation task, in three children who suffered from a concussion while playing sports. For the trail making test, children drew lines to connect numbers and letters alternately on a page. For the visuomotor adaptation task, children performed targeted reaching movements repeatedly under a novel visuomotor condition in which the visual display of reaching movements was rotated 30 degrees about the start position. Their neurocognitive and visuomotor performances were compared to those obtained from three children without a concussion. Results showed that only one of the three concussed children showed a score from the trail making test that was worse than those obtained from control subjects. However, the pattern of visuomotor adaptation observed in the concussed children was different from that observed in the control subjects in several ways. These data indicate that the concussed children may have some deficits in terms of their sensorimotor function as compared to the control subjects. Our findings collectively suggest that children with concussion may have sensorimotor impairments even when they do not seem to have neurocognitive impairments.", "keywords": [ "Mild Traumatic Brain Injury", "Motor Learning", "Visuomotor Adaptation", "Trail Making Test" ], "content": "Introduction\n\nPediatric concussions have been shown to negatively affect neurocognitive function, including poor attention span, impaired memory and learning difficulties1–3. However, the association between pediatric concussion and sensorimotor function has not been investigated. Understanding motor learning impairment following sport-related concussion in children may be critical in guiding safe return to play. In this research, we investigated the pattern of adaptation to a novel visuomotor condition in three children who suffered a concussion while playing sports, and compared their adaptation patterns to those obtained from three children without a concussion. A research paradigm in which individuals adapt to a novel visuomotor condition during reaching movements has been employed extensively in the neuroscience community to understand the neural processes that underlie motor learning in both adults4,5 and children6,7.\n\nWe also examined neurocognitive function of both the concussed children and the control children using an assessment technique called the trail making test (TMT). TMT (part B) is a traditional paper assessment that times subjects as they draw lines to connect numbers and letters alternately on a page. TMT is a validated assessment of concussion, and tests cognitive demands that are also important for sports, including psychomotor processing, visual motor/spatial abilities and mental flexibility8,9.\n\nThe purpose of this observational study was to determine whether concussed children with neurocognitive impairments, as indicated by the TMT scores, would also demonstrate sensorimotor deficits, as indicated by the visuomotor adaptation patterns.\n\n\nMethods\n\nThree children (15 years old), who presented to the Emergency Department at the Children’s Hospital of Wisconsin within 24 hours from the time of injury and who received a diagnosis of concussion (Glasgow Coma Scale ≥ 14), participated in this study. They visited the Neuromechanics Laboratories at the University of Wisconsin-Milwaukee (UWM) 5~8 days following the concussion. Three children (12, 14 and 17 years old), who were recruited from the Milwaukee Metropolitan area, served as controls. Selection criteria for subjects were the same between patients and controls (except their concussion status), which were: subject is 10–17 years of age, regularly participates in an athletic activity, is English-speaking, is right handed, and has no neurological disease or peripheral disorder affecting movement of the right arm. All subjects were recruited and tested in May 2015.\n\nUpon arrival at UWM, the subjects were first administered with the TMT8,9. Following that, they participated in the visuomotor adaptation experiment in which they performed rapid reaching movements from a start circle to a target repeatedly under a normal visuomotor condition (baseline) first, then under a novel visuomotor condition (adaptation). The baseline session (40 trials) was provided for the subjects to become familiarized with the general reaching task with unperturbed visual feedback. In the adaptation session (80 trials), the visual display of reaching movements was rotated 30 degrees counterclockwise about the start circle, such that a hand movement made in the “12 o’clock” direction resulted in a cursor movement made in the “11 o’clock” direction. Continuous visual feedback (in the form of a cursor) was provided throughout the movement in both sessions. A robotic exoskeleton called KINARM (BKIN Technologies Ltd, Kingston, ON, Canada) was used to provide the visuomotor rotation during the experiment and also to collect movement data. The 2-D position of arm segments was sampled at 1,000Hz, low-pass filtered at 15Hz, and differentiated to yield resultant velocity values. Data were processed and analyzed using MATLAB (The Mathworks Inc., Natick, MA) and SPSS.\n\nTo examine performance accuracy, we calculated direction error (DE), which was the angular difference between a vector from the start circle to the target and another vector from the hand position at movement start to that at peak arm velocity. Using the direction error data, we obtained the following measures for each subject: (1) DE at trial 1 in the adaptation session; (2) the first block of DE (i.e., mean of five consecutive trials) in the adaptation session that is not statistically different from the last block of DE in the baseline session; and (3) the rate of performance change during the adaptation session. To obtain the second measure, a priori pairwise comparisons, using t-tests, were made between DE at block 8 from the baseline session and DE at each of the 16 blocks from the adaptation session (starting from block 1). The alpha level was set at .05. To obtain the third measure, a line of approximation was constructed by fitting a logarithmic regression line to the adaptation data, and the slope value was used.\n\nAll subjects and their parents signed the assent/consent forms approved by the Institutional Review Board of UWM (IRB# 15.172).\n\nBecause of the nature of the present study (i.e., observational/case study of concussed children), we only tried to recruit a small number of concussed children. We recruited all our subjects (three concussed children and three controls who met our selection criteria) within a 10-day window.\n\n\nResults\n\nFigure 1c illustrates neurocognitive data from all subjects, indicated by TMT scores. One concussed child (cc1) showed some cognitive slowing on visit 1, as compared with the controls, while the others did not.\n\n(a) Improvements in performance across trials during baseline and adaptation sessions. Upper panel depicts data from 3 controls, lower panel data from 3 concussed children. (b) Improvements in performance across blocks during the baseline (last block only) and adaptation sessions. Numbers under arrows indicate first block of DE that was not statistically different from last block of DE from the baseline phase. (c) TMT scores for every subject. (d) DE at trial 1 during the adaptation session. (d) First block of DE in the adaptation session that was not different from last block of DE in the baseline session. (e) Rate of performance change in the adaptation session. Vertical boxes in red (c–f) indicate variability within each subject group.\n\nThe overall pattern of visuomotor adaptation, illustrated in Figure 1a and b, was somewhat similar between the patients and the controls. In fact, the patients’ baseline performances do not appear to be worse than those of the controls. However, a close examination of the adaptation data revealed the following differences:\n\n1. DE at trial 1 of the adaptation session was somewhat larger for the patients than the controls; and within-group variability for this measure was also larger for the patients (Figure 1d).\n\n2. The first block of DE in the adaptation session that was not statistically different from the last block of DE in the baseline session was observed much later in two of the patients than in the controls; and within-group variability for this measure was much larger for the patients (Figure 1b, e).\n\n3. As indicated by the rate of performance change, it took longer for the patients than the controls to adapt to the visuomotor rotation; and within-group variability for this measure was slightly larger for the patients (Figure 1f).\n\n\nDiscussion\n\nTMT is a validated neurocognitive assessment of pediatric concussion8,9. The average time to complete the TMT part B is 75 seconds in neurologically intact individuals; if the time is longer than 273 seconds, it is considered deficient10,11. One concussed child who participated in our study completed the TMT part B in 98 seconds; and the other two children completed it in less than 50 seconds. According to the TMT scores, thus, one may conclude that all these children who had a concussion 5 to 8 days prior to their participation in this study did not seem to have neurocognitive impairments.\n\nTheir patterns of visuomotor adaptation observed in the patients, however, appear to be somewhat different from those observed in the controls. Specifically, the first block of DE in the adaptation session that is significantly different from the last block of DE of the baseline session occurred later in the patients than in the controls; and the rate of performance change was higher (i.e., slower adaptation) in the patients as well. These data indicate that the patients had more difficulty than the controls while adapting to the novel visuomotor rotation, which points to the possibility of sensorimotor learning deficits in the concussed children.\n\nThe present study has several limitations. First, all the children tested in this study were recruited in the Milwaukee Metropolitan area, which raises a possibility that they may not be the best representative of all children of the same age group. Also, we did not collect from our subjects any information regarding their demographic and social status, which could be considered as potential confounding characteristics. Finally, we did not conduct statistical analyses to quantify the differences between the children with concussion and those without concussion, because the statistical results would not be very meaningful given the small sample size. Nevertheless, our data exhibited qualitative differences between the two groups of children that suggest sensorimotor deficits in concussed children. Further investigation is warranted to demonstrate quantitative differences between children with and children without concussion by utilizing a larger sample size.\n\nIn conclusion, the results from this study indicate that children with concussion may have sensorimotor impairments even when they do not seem to have neurocognitive impairments. Given the importance of children’s ability to rapidly adapt to the dynamic environments and task requirements throughout a sport, our findings suggest that examination of sensorimotor function, in addition to that of neurocognitive function, may provide valuable information when determining whether children are ready to return to play sports or not.\n\n\nData availability\n\nF1000Research: Dataset 1. Kinematic data from visuomotor adaptation task, 10.5256/f1000research.6575.d4898412", "appendix": "Author contributions\n\n\n\nJW, DGT and YIC conceived the study and designed the experiment. JW carried out the research and prepared the first draft of the manuscript. DGT and YIC edited the manuscript. All authors have agreed to the final content of the manuscript.\n\n\nCompeting 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\nReferences\n\nBrowne GJ, Lam LT: Concussive head injury in children and adolescents related to sports and other leisure physical activities. Br J Sports Med. 2006; 40(2): 163–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThomas DG, Collins MW, Saladino RA, et al.: Identifying neurocognitive deficits in adolescents following concussion. Acad Emerg Med. 2011; 18(3): 246–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSandel NK, Lovell MR, Kegel NE, et al.: The relationship of symptoms and neurocognitive performance to perceived recovery from sports-related concussion among adolescent athletes. Appl Neuropsychol Child. 2013; 2(1): 64–9. PubMed Abstract \| Publisher Full Text\n\nGoodbody SJ, Wolpert DM: Temporal and amplitude generalization in motor learning. J Neurophysiol. 1998; 79(4): 1825–38. PubMed Abstract\n\nWang J, Sainburg RL: Adaptation to visuomotor rotations remaps movement vectors, not final positions. J Neurosci. 2005; 25(16): 4024–30. PubMed Abstract \| Publisher Full Text\n\nKagerer FA, Bo J, Contreras-Vidal JL, et al.: Visuomotor adaptation in children with developmental coordination disorder. Motor Control. 2004; 8(4): 450–60. PubMed Abstract\n\nContreras-Vidal JL, Bo J, Boudreau JP, et al.: Development of visuomotor representations for hand movement in young children. Exp Brain Res. 2005; 162(2): 155–64. PubMed Abstract \| Publisher Full Text\n\nTombaugh TN: Trail making test A and B: normative data stratified by age and education. Arch Clin Neuropsychol. 2004; 19(2): 203–14. PubMed Abstract \| Publisher Full Text\n\nBowie CR, Harvey PD: Administration and interpretation of the Trail making Test. Nat Protoc. 2006; 1(5): 2277–81. PubMed Abstract \| Publisher Full Text\n\nGaudino EA, Geisler MW, Squires NK: Construct validity in the Trail Making Test: what makes Part B harder? J Clin Exp Neuropsychol. 1995; 17(4): 529–535. PubMed Abstract \| Publisher Full Text\n\nLezak MD, Howieson DB, Loring DW: Neuropsychological Assessment. 4th ed. New York: Oxford University Press. 2004. Reference Source\n\nWang J, Thomas D, Cho YI: Dataset 1 in: Sensorimotor learning deficits observed in children with sports-related concussion. F1000Research. 2015. Data Source" }	[ { "id": "8919", "date": "24 Jun 2015", "name": "Lauren Sergio", "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\nThis study describes a psychophysical experiment in which three children who had recently experienced a concussion were asked to perform a visuomotor adaptation. Their performance was compared to three children with no concussion history. While the motivation for the study was sound, I have a number of concerns with this manuscript:Introduction: “However, the association between pediatric concussion and sensorimotor function has not been investigated.” An inaccurate statement, de Beaumont has done some work in this area, as has Danckert (review references listed at end). Methods: Won't Measure 2 always co-vary with Measure 3? Perhaps I am misunderstanding something. The rate of adaptation should always be related to the point at which they reach baseline performance, no? Major concern: Given the variety in symptoms and behavioural response to concussion, a study using only 3 children should be considered along the lines of a case study, or extremely preliminary data at least, more suitable for an abstract. Minimally, it should be made clear in the intro that this is a case study (only mentioned at end of methods). In general, this type of study is not appropriate as a case study since adolescent concussion is not, unfortunately, a rare or unusual situation. It would not be overly cumbersome to have an appropriate sample size (10-12 each group), fully analyse the data, and only then report the findings. It should be noted if this was the participants' first concussion or not, since this is often an important distinguishing variable in post-concussion recovery rate. It should also be noted if the experimental group members were still symptomatic (via something like a SCAT3). Major concern: Given the significant neurocognitive and neurological development between 12 and 15, three age-matched controls would have been better. Also, because of sex-related differences in visuomotor control (see Gorbet et al. 2007, 2011) and sex-related developmental differences in this age range, the sex of the children should be stated and matched with controls. Typically this paradigm includes a 'washout' phase whereby participants do a number of trials in a null field. The level of aftereffect serves as an indication of the extent to which the adaptation was incorporated. This might be useful for the present design. TMT result: Again, there is the issue of comparing a 12 year old to 15 year olds to a 17 year old, so it is difficult to judge whether this is an age effect or a concussion effect. Age is typically a strong covariate for coordination measures taken pre/post-concussion (see Dalecki et al. 2015). It appears that the worse the concussed participant did on the TMT, the later the point of adaption-baseline match (cc1 and cc3 vs cc2). This is something which cannot be explored with n=3, but should be noted and examined with the larger data set.Conclusion:The concept and approach are interesting and potentially useful, and indeed the use of sensorimotor impairment in the absence of current-standard symptoms post-concussion is an important area of study. However, the inappropriate presentation as a case study and the lack of age (and sex) controls make it difficult to consider this a full and rigorous study. In its present form it is suitable as a presentation of preliminary data at a conference rather than a cite-able publication. I look forward to seeing the complete study. Minor comments:Methods, paragraph 1: grammar issue (tense change) – “Selection criteria for subjects were the same between patients and controls (except their concussion status), which were: subject is 10–17 years of age,regularly participates in an athletic activity, is English-speaking, is right handed, and has no neurological disease or peripheral disorder affecting movement of the right arm.” Figure 1a: Is the axis mislabeled for figure 1a (left)? The methods state that there are 40 baseline trials. Figure 1b: consider using a different symbol shape for controls, difficult to distinguish open versus filled circles. Figure caption error (two ‘d’ references, no ‘ f’ reference) References:De Beaumont L, Tremblay S, Henry LC, et al. Motor system alterations in retired former athletes: the role of aging and concussion history. BMC Neurol. 2013;13:109.Locklin J1, Bunn L, Roy E, Danckert J. Measuring deficits in visually guided action post-concussion. Sports Med. 2010 Mar 1;40(3):183-7.Gorbet DJ, Sergio LE. Preliminary sex differences in human cortical BOLD fMRI activity during the preparation of increasingly complex visually guided movements. Eur J Neurosci. 2007;25:1228-1239.Gorbet DJ, Staines WR. Inhibition of contralateral premotor cortex delays visually guided reaching movements in men but not in women. Exp Brain Res. 2011 Jul;212(2):315-25.Dalecki MS, Sergio LE. Prolonged cognitive-motor impairments in children with a history of concussion. #2-D-107, Canadian Association for Neuroscience, 2015, Vancouver, BC. http://can-acn.org/documents/2015/CAN2015_Abstract_Booklet.pdf", "responses": [ { "c_id": "1466", "date": "23 Jul 2015", "name": "Jinsung Wang", "role": "Author Response", "response": "Thank you for your review. In our responses below, we have attempted to address each concern raised by you.Introduction: “However, the association between pediatric concussion and sensorimotor function has not been investigated.” An inaccurate statement, de Beaumont has done some work in this area, as has Danckert (review references listed at end):We believe this statement by the reviewer is incorrect, in that neither study investigated association between concussion and sensorimotor function in children. de Beaumont et al. investigated a motor learning issue in retired athletes aged between 51 and 75 years; and Danckert and colleagues investigated a motor control issue in young adults aged between 17 and 27 years. Methods: Won't Measure 2 always co-vary with Measure 3? Perhaps I am misunderstanding something. The rate of adaptation should always be related to the point at which they reach baseline performance, no?:Our measures 2 and 3 do not necessarily co-vary with each other, especially given that measure 3 is influenced by the size of direction error at the beginning of adaptation whereas measure 2 is not.Major concern: Given the variety in symptoms and behavioural response to concussion, a study using only 3 children should be considered along the lines of a case study, or extremely preliminary data at least, more suitable for an abstract. Minimally, it should be made clear in the intro that this is a case study (only mentioned at end of methods):As the reviewer noted, this research note reports movement data obtained from a very small number of children. In fact, we originally intended to publish this short manuscript as a clinical practice article. According to the journal’s guidelines, clinical practice articles include series of case reports, which do not need to describe especially novel or unusual cases. During the in-house check, however, the editor decided that our manuscript would be more suitable as a research note. We have now modified the introduction section to emphasize the nature of this study (i.e., observational/case study).It should be noted if this was the participants' first concussion or not, since this is often an important distinguishing variable in post-concussion recovery rate. It should also be noted if the experimental group members were still symptomatic (via something like a SCAT3):Our patients never had a concussion prior to this one. We did not assess their symptoms on the day they participated in the experiment, although one of the patients (cc1) was not able to speak or walk normally at that time. We have included this information in the methods section now.Major concern: Given the significant neurocognitive and neurological development between 12 and 15, three age-matched controls would have been better. Also, because of sex-related differences in visuomotor control (see Gorbet et al. 2007, 2011) and sex-related developmental differences in this age range, the sex of the children should be stated and matched with controls:We agree that it would have been better if both age and sex were matched perfectly between the patients and the controls. However, we are not aware of any findings that indicate significant differences between boys and girls within the age range of 12-17 years with regard to their visuomotor adaptation capabilities. There is at least some evidence that visuomotor representations used by children older than 11 years are similar to those used by adults. For example, Ferrel et al. (2001, Exp Brain Res) suggested that during targeted reaching movements under a visuomotor rotation condition, visuomotor representations of children who were 6 or 8 years old differed from those of adults, although those of 11 year-old children did not. In addition, our data do not seem to indicate sex-related differences clearly. For example, the adaptation pattern of cc2 (15 year-old female) appears to be very similar to that of ctls 1 and 3 (12 and 14 year-old males, respectively) in terms of our measures 2 and 3. Thus, it seems unlikely that our results were substantially influenced by age- or sex-related differences among the participants. Nonetheless, we agree that we cannot completely exclude the possibility that age and sex of the participants played some role. Therefore, we have now included additional statements in the discussion section to address this point. We have also provided sex information of our participants in the methods section.Typically this paradigm includes a 'washout' phase whereby participants do a number of trials in a null field. The level of aftereffect serves as an indication of the extent to which the adaptation was incorporated. This might be useful for the present design:We did not include a washout phase in our present design, but we plan to do so in the future when we can conduct a more rigorous study with a larger sample size.TMT result: Again, there is the issue of comparing a 12 year old to 15 year olds to a 17 year old, so it is difficult to judge whether this is an age effect or a concussion effect. Age is typically a strong covariate for coordination measures taken pre/post-concussion (see Dalecki et al. 2015):The main result of the present study in terms of TMT is that the TMT scores of all pediatric patients tested in our study (including cc1 whose score was the largest) were below 100 seconds, which is substantially better than the score that is considered deficient (i.e., 273 seconds). Also, our data indicate that the TMT scores of all our subjects except cc1 (15 year-old female) range only from 27 to 53 seconds, which do not appear to vary depending on the subject’s age or sex. Thus, it seems unlikely that our TMT result was influenced substantially by the subject’s age.It appears that the worse the concussed participant did on the TMT, the later the point of adaption-baseline match (cc1 and cc3 vs cc2). This is something which cannot be explored with n=3, but should be noted and examined with the larger data set:We agree that this is something that should be examined (when one has a larger data set). However, we would like to point out that the TMT score of cc3 fell within the range of the TMT scores obtained from the controls and that his score was substantially better than the score considered deficient, which indicates that this patient did not suffer from any cognitive slowing. Given that, our data do not suggest a strong association between TMT performance and visuomotor adaptation at least in the concussed children who were tested in our study. Minor comments:Methods, paragraph 1: grammar issue (tense change) – “Selection criteria for subjects were the same between patients and controls (except their concussion status), which were: subject is 10–17 years of age, regularly participates in an athletic activity, is English-speaking, is right handed, and has no neurological disease or peripheral disorder affecting movement of the right arm.”:We have changed the tense now.Figure 1a: Is the axis mislabeled for figure 1a (left)? The methods state that there are 40 baseline trials:The reviewer is correct that there were 40 baseline trials. We have corrected this error now.Figure 1b: consider using a different symbol shape for controls, difficult to distinguish open versus filled circles:We have now included the lines that are either thick or thin to help distinguish between the patients and controls.Figure caption error (two ‘d’ references, no ‘ f’ reference):We have corrected this error now." } ] } ]	1	https://f1000research.com/articles/4-143
https://f1000research.com/articles/4-22/v1	23 Jan 15	{ "type": "Research Article", "title": "Ligand uptake in Mycobacterium tuberculosis truncated hemoglobins is controlled by both internal tunnels and active site water molecules", "authors": [ "Ignacio Boron", "Juan Pablo Bustamante", "Kelly S Davidge", "Sandip Singh", "Lesley AH Bowman", "Mariana Tinajero-Trejo", "Sebastián Carballal", "Rafael Radi", "Robert K Poole", "Kanak Dikshit", "Dario A Estrin", "Marcelo A Marti", "Leonardo Boechi", "Ignacio Boron", "Juan Pablo Bustamante", "Kelly S Davidge", "Sandip Singh", "Lesley AH Bowman", "Mariana Tinajero-Trejo", "Sebastián Carballal", "Rafael Radi", "Robert K Poole", "Kanak Dikshit", "Dario A Estrin", "Marcelo A Marti" ], "abstract": "Mycobacterium tuberculosis, the causative agent of human tuberculosis, has two proteins belonging to the truncated hemoglobin (trHb) family. Mt-trHbN presents well-defined internal hydrophobic tunnels that allow O2 and •NO to migrate easily from the solvent to the active site, whereas Mt-trHbO possesses tunnels that are partially blocked by a few bulky residues, particularly a tryptophan at position G8. Differential ligand migration rates allow Mt-trHbN to detoxify •NO, a crucial step for pathogen survival once under attack by the immune system, much more efficiently than Mt-trHbO. In order to investigate the differences between these proteins, we performed experimental kinetic measurements, •NO decomposition, as well as molecular dynamics simulations of the wild type Mt-trHbN and two mutants, VG8F and VG8W. These mutations introduce modifications in both tunnel topologies and affect the incoming ligand capacity to displace retained water molecules at the active site. We found that a single mutation allows Mt-trHbN to acquire ligand migration rates comparable to those observed for Mt-trHbO, confirming that ligand migration is regulated by the internal tunnel architecture as well as by water molecules stabilized in the active site.", "keywords": [ "Mycobacterium tuberculosis", "hemoglobin", "water molecules", "ligand interaction" ], "content": "Introduction\n\nMycobacterium tuberculosis, the causative agent of human tuberculosis, affects approximately two billion people world-wide, causing over three millions deaths each year1. The genome of this pathogenic organism includes two genes, glbN and glbO, which encode for two proteins, termed here truncated hemoglobin N (Mt-trHbN) and truncated hemoglobin O (Mt-trHbO), belonging to the truncated hemoglobin (trHb) family of heme proteins, widely distributed in eubacteria, cyanobacteria, microbial eukaryotes and plants2,3.\n\nThe truncated hemoglobin family exhibits a three-dimensional structure similar to the common globin fold of myoglobin, but significantly smaller. The secondary structure of trHbs consists of four α-helices arranged in a two-over-two antiparallel sandwich instead of the common three-over-three helix globin fold. Phylogenetic analysis has distinguished three different groups of truncated hemoglobins, classified as groups I, II and III, also called N, O and P, respectively.\n\nIt has been shown that group I Mt-trHbN catalyzes the detoxification of •NO in the presence of O24,5. The first step of this mechanism involves O2 migration and binding. Subsequently, •NO migrates to the active site and reacts with the heme-bound O2 to yield an unstable peroxynitrite adduct, which isomerizes to generate the relatively innocuous nitrate anion.\n\nSeveral studies have examined the role of internal tunnels in ligand migration in trHbs2,5–8. Three internal tunnels were found in the truncated hemoglobin family: a long tunnel (LT) topologically positioned between helices B and E, and two short tunnels, known as the E7 Gate (E7 gate) and the short tunnel G8 (STG8), which are roughly normal to the LT, as depicted in Figure 1. The E7 tunnel corresponds to the highly conserved E7 pathway widely studied in both myoglobin and hemoglobin9–11. The STG8 tunnel is analogous to that found in Mt-trHbN, next to the key residues VG8 and IH11. Previous results indicate that WG8, an absolutely conserved residue in groups II and III truncated hemoglobins, is involved in hindering ligand migration in Mt-trHbO by blocking both STG8 and LT (Figure 1)12–14. In addition, the presence of a smaller residue at the G8 position in the Mt-trHbO mutant (WG8F) was observed to increase the small ligand association constant, although the molecular details of this process were not investigated12–14. It has also been noted that internal water molecules can block the heme accessibility, thus delaying ligand binding15–17.\n\nThe three pathways, Long Tunnel (LT), E7 Gate (E7 gate) and Short Tunnel G8 (STG8) for ligand migration through the tertiary structure of a typical trHb are shown.\n\nBy performing CO association kinetic constant measurements (kon CO), •NO decomposition, and molecular dynamics (MD) simulations, we addressed molecular mechanisms that control ligand association in M. tuberculosis truncated hemoglobins.\n\n\nMaterials and methods\n\nThe trHbN G8 mutants (VG8W and VG8F) were prepared using the Stratagene QuikChange mutagenesis kit. The following primers were designed using Primer3 http://biotools.umassmed.edu/bioapps/primer3_www.cgi18 to generate single amino acid substitutions (underlined): i) WG8: forward primer 5’–CACTTCAGCCTGTGGGCCGGACACTTGG–3’ and reverse primer 5’–CAAGTGTCCGGCCCACAGGCTGAAGTG–3’; ii) FG8: forward primer 5’–ACCACTTCAGCCTGTTCGCCGGACACTTG–3’ and reverse primer 5’–CAAGTGTCCGGCGAACAGGCTGAAGTGGT–3’. Polymerase chain reaction (PCR) amplification of pET9b carrying the glbN gene with the aforementioned primers was conducted following manufacturer’s instructions. The PCR mix consisted of 5 µl 10x reaction buffer, 5–50 ng double stranded DNA template, 125 ng oligonucleotide primer 1, 125 ng oligonucleotide primer 2, 1 µl dNTP mix, 1 µl PfuUltra HF DNA polymerase and double distilled H2O to a final volume of 50 µl. The PCR reaction was 95°C for 30 s, followed by 16 cycles of: 95°C for 30 s, 55°C for 1 min and 68°C for 4 min. The reaction mix was then digested with DpnI to remove parental methylated DNA. Plasmid containing the mutated gene was then purified and used to transform Escherichia coli XL-1 Blue electrocompetent cells. Cells were provided by Invitrogen. Constructs were checked by sequencing.\n\nAll chemicals and reagents were obtained from Sigma Aldrich, unless indicated otherwise. The trHbN protein was purified using standard techniques reported for other bacterial globins19. Briefly, mutated constructs were used to transform E. coli BL21 DE3 pLysS. Starter cultures grown overnight in LB supplemented with kanamycin (50 μg ml-1) and chloramphenicol (35 μg ml-1) were used to inoculate 6 batches of 1 L LB medium at 1% (v/v), supplemented with kanamycin and 3 μM FeCl3. Once cultures reached an OD600 of around 0.4, expression of trHbN was initiated by the addition of 1 mM IPTG and grown for a further 4 h. Cells were harvested by centrifugation at 5500 rpm for 20 min at 4°C and stored overnight at -20°C. After thawing, cells were resuspended in 40 ml buffer (10 mM TRIS-HCl (pH 7.0) with 1 mM EDTA, 10 mM DTT, 45 μg ml-1 phenylmethylsulphonyl fluoride, 500 μg ml-1 RNase and 50 μl DNase), homogenized using a Douce homogeniser and ultracentrifuged at 44,000 rpm for 1 h at 4°C. The supernatant, red in color, was loaded onto a 30 ml DEAE Sepharose Fast Flow column (Pharmacia Biotech) equilibrated with 10 mM TRIS-HCl (pH 7.0), washed with the same buffer until the UV trace returned to baseline, and eluted via a gradient from 0 to 1 M NaCl in 10 mM TRIS-HCl using an Akta purifier (GE Healthcare Bio-Sciences, Amersham Biosciences, U.K. Ltd.). Fractions that were most red in color were concentrated using a Vivaspin 20 concentrator (Sartorius Stedim Biotech) to around 5 ml and loaded onto a gel filtration Superdex 75 column, equilibrated with 0.15 M NaCl in 10 mM TRIS-HCl (pH 7.5); again, fractions with the most color were collected, combined and stored at -80°C. Purity was checked using gel electrophoresis and analysis of the heme-to-protein ratio (410 nm and 280 nm in the UV-visible absorption spectrum).\n\nRapid mixing experiments were conducted with a thermostated stopped flow apparatus (BioLogic SFM-300). Kinetics of carbon monoxide (CO) binding to determine the kon CO were measured on the deoxy state of mutant and wild type globins at 20ºC. Solutions containing 5 μM protein in a 100 mM sodium phosphate at pH 7.0 were degassed in a nitrogen atmosphere and reduced with an equimolar concentration of sodium dithionite and mixed with increasing CO concentrations. The observed pseudo first-order rate constant (kobs) was determined by fitting the absorbance decay resulting from association of the protein with CO, to a single exponential function. Kinetic rate constants (kon CO) were obtained from the slope of the plots of kobs as a function of CO concentration.\n\nTo determine rates of nitric oxide (•NO) decomposition by wild type and mutant Mt-trHbN proteins, •NO was added, as ProliNONOate, to a solution of 50 mM KPi buffer with 50 μM EDTA (pH 7.5), 100 μM NADPH and 100 nM E. coli ferredoxin reductase inside a thermoregulated, magnetically stirred reaction vessel. Mt-trHbN (2 μM) was added at the apex of the signal response to 2 μM ProliNONOate and •NO decay was followed until depleted using an •NO electrode (World Precision Instruments). Rates of •NO decay were calculated for each protein by determining the time taken for peak •NO concentrations to decay by 0.5 μM and were expressed per μM heme, determined spectrally by the peak in the Soret region at 410 nm.\n\nThe starting structure corresponds to Mt-trHbN crystal structure (PDB entry 1IDR http://www.rcsb.org/pdb/explore/explore.do?structureId=1IDR), at 1.9 Å of resolution20. Amino acids protonation states were assumed based on environment of the residue in the crystal structure. All the solvent-exposed His were protonated at the N-δ delta atom, as well as HisF8, because of its coordination to the heme iron. An octahedral box of 10 Å of radius, which corresponds to 5234 explicit water molecules was added to the system. TIP3P water molecules were used by tLEaP module of the AMBER12 package21. The param99 Amber force field was used for all the aminoacid parameters22 except heme parameters which were developed in our group23 and strongly validated for being used in several studies of heme proteins24–30. Periodic boundary conditions were used for all the simulations performed with a 9 Å cutoff. Particle mesh Ewald (PME) summation method for treating the electrostatic interactions. The SHAKE algorithm was used to keep constant the non-polar hydrogen equilibrium distance. Temperature and pressure were kept constant with Langevin thermostat and barostat, respectively, as implemented in the AMBER12 program21. The equilibration simulation protocol was performed as follow: (i) slowly heating the system from 0 to 300K for 20 ps at constant volume, by using harmonic restraints of 80 kcal/mol A2 for all Cα atoms and (ii) pressure equilibration of the whole system during 1 ns at 300K with restrained atoms in (i). (iii) Unconstrained 100 ns molecular dynamics simulation at constant temperature (300K) was performed.\n\nIn silico mutant proteins were built by using tLEaP module of AMBER12 package21, and underwent the same protocol used for wild type protein. Root Mean Square Deviation (RMSD) was used as structure stability controls. All structures were observed to be stables during the time scale of the simulation (Figure S1).\n\nThe free energy profile for the CO migration process inside the protein tunnel/cavity system was computed by the Implicit Ligand Sampling (ILS) approach that post-processes, using a probe molecule, an MD simulation performed in the absence of the ligand. This method was thoroughly tested for heme proteins32. ILS calculations were performed on a rectangular grid (0.5 Å resolution) that includes the whole simulation box (i.e. protein and the solvent) and the probe used was a CO molecule. Calculations were performed on 5000 frames taken from the last 90 ns of simulation time. The values for grid size, resolution and frame numbers were tested in a previous study32. Analysis of the ILS data was performed using an ad-hoc fortran-90 program available upon request32. Moreover, ILS has been shown to yield quantitative results for ligand migration processes when compared with more costly free energy methods that treat the ligand explicitly.\n\n\nResults\n\nAlthough CO is not the natural ligand of the hemeproteins, it is widely used as a probe for ligand association studies due to its ease of use. In order to address the molecular determinants controlling ligand migration we performed CO ligand association constant measurements of wild type Mt-trHbN and two mutants: VG8F and VG8W. Kinetic traces for CO binding were measured through the absorption changes at the CO adduct peak position (λ=423 nm; Figure 2). Association of CO is well described by a single exponential decay, whose rate constant (kobs) depends linearly on CO concentration and the slope can be interpreted as kon CO. A significant kon CO decrease for VG8F (715 ± 27 mM-1s-1), and an even larger decrease for VG8W (48 ± 1 mM-1s-1) was observed in relation to that observed for the wild type protein (4495± 357 mM-1s-1) (Figure 3). Table 1 summarizes the measured kon CO values for wild type and mutant Mt-trHbs O and N, and is presented alongside literature data.\n\nStopped-flow time course for the reaction of reduced 5 μM wild type (A), VG8F (B), and VG8W (C) mutants in 100 mM phosphate buffer at pH 7. The reaction was monitored at 423 nm (grey dots) and the line shows the best first order fit.\n\nCurves for wild type (green), VG8F (orange) and VG8W (violet) mutants as a function of CO concentration in stopped flow measurements are shown. The time courses are measured at different CO concentrations ranging from 10 to 200 µM (after mixing). Continuous line corresponds to linear fit of kobs rates.\n\nmajor and minor rate contributions to a biphasic fitting are indicated between brackets.\n\nSmall ligand association in the trHb family is presumably regulated by two main processes: i) ligand migration from solvent bulk to the protein distal site cavity, ii) displacement of water molecules from the distal site cavity15–17,35. With this in mind, we performed classical MD simulations as they allow us to investigate both processes involved in ligand association. Ligand migration was studied using ILS calculations for the wild type, as well as both VG8F and VG8W mutant proteins. Displacement of retained water molecules in the distal site was considered by performing classical MD simulations and analyzing the solvation structure in each active site.\n\nThe wild type Mt-trHbN presents two tunnels available for ligand migration, the LT and the STG8 (Figure 4A). On the one hand, the LT connects three internal cavities: (trHb : CO)1, (trHb : CO)2 and (trHb : CO)3. The STG8, on the other hand, has only the distal site cavity, (trHb : CO)1, which is directly connected to the solvent. The VG8F mutant conserves both tunnels, although they are constrained compared to those in the wild type. In the VG8W case, however, the energy profiles suggest a completely blocked STG8 and a LT for which the accessibility to the iron heme is partially reduced.\n\n(A) Schematic representations of the residues involved in the heme distal site and tunnels, the two tunnels and cavities estimated with ILS for the wild type form. (B) Free energy profiles over STG8 and (C) LT along the connection between solvent, (trHb : CO)2 and (trHb : CO)1 cavities for wild type (green), VG8F (orange) and VG8W (violet) mutant forms are shown. Circles represent calculated free energy values with the ILS method and lines correspond to a fitting estimation of these calculated values. The x coordinate represents the Fe-CO distance along the pathways.\n\nIn order to quantify the contribution of the single G8 mutation we computed free energy profiles for CO migration through both LT and STG8 tunnels (Figure 4B, 4C). The free energy was set to a value of 0 kcal/mol where CO ligand is fully solvated- at 13 Å and 24 Å from the Fe atom, for STG8 and LT respectively. Wild type Mt-trHbN presents small barriers (~2 kcal mol-1) for CO migration from the solvent to the active site cavity (trHb : CO)1 through both tunnels.\n\nThe active site water molecules occupancy was computed for all three systems by performing 200 ns of MD simulations with explicit water molecules. In each case a water molecule was able to access the active site and was stabilized by the iron and the distal site residues (Figure 5). Specifically, in both wild type and VG8F Mt-trHbN a water molecule was present for approximately 40% of the length of the simulation (Figure 5A, 5B). The VG8W mutant active site, on the other hand, is occupied by water molecules in 80% of the simulation time, probably due to the hydrogen bonding capacity of W (Figure 5C).\n\n(A) wild type, (B) VG8F and (C) VG8W forms showing, on the basis of MD simulations, the hydrogen-bond network (dotted lines) stabilizing a water molecule above the iron heme. The percentages depicted as insets in the figure correspond to active site water occupancy during MD simulation.\n\nMt-trHbN has previously been described as a dioxygenase, capable of O2-dependent •NO consumption36,37. Consequently, •NO decomposition by purified Mt-trHbN and the VG8F, VG8W mutants was determined in a reaction mixture containing buffer, NADPH and E. coli FdR, to enable cyclic restoration of heme iron to the oxyferrous state. Figure 6A shows that in the absence of protein (red trace) decay of the •NO signal was monophasic until •NO was exhausted. The decay of NO in the presence of Mt-trHbN (black trace) was biphasic, with an almost linear initial rapid rate in decay, which was used to compare the various Mt-trHbN derivatives, followed by a slower rate in decay. This suggests that •NO is not being turned over in a cyclic manner, but is simply binding available heme. •NO consumption results show that the VG8F and VG8W mutans have a statistically significant reduced •NO binding capacity compared to HbN (Figure 6B).\n\n(A) •NO decay was monitored amperometrically in the absence (red trace) and the presence (black trace) of Mt-trHbN added at the apex of the signal response to 2 μM ProliNONOate. Data are representative of 3 technical repeats. (B) Mean rates of •NO decay in the presence of wild type Mt-trHbN or site-directed mutants from 3 technical repeats ± S.E.M P < 0.05, unpaired t-test.\n\n\nDiscussion\n\nCO association kinetic constant measurements as well as MD simulations of Mt-trHbN wild type and site-specific mutants were performed to analyze the role of tunnels and water molecules in the ligand association process. ILS calculations showed that the main tunnels of wild type Mt-trHbN, STG8 and LT, were partially blocked in the VG8F mutant and STG8 was nearly completely blocked in the VG8W mutant. The analysis of water molecules showed that VG8W increases the probablity of the presence of a water molecule in the distal site, which may interfere with the association process. Consistently, the association kinetic constants of CO for both Mt-trHbN mutants showed a decrease of slightly less than one order of magnitude when VG8 is replaced with F and two orders of magnitude when VG8 is replaced with W. Moreover, our data also showed that that both mutants have less capacity of •NO binding than wild type Mt-trHbN.\n\nInterestingly, the Mt-trHbN VG8W mutant presents a similar kon CO to the wild type Mt-trHbO, showing that a single residue is responsible for the differential accessibility in these proteins. The results support the idea that STG8 and LT are the main channels for CO migration in the deoxygenated Mt-trHbN, as blocking these tunnels decreases the ability of CO to access the heme pocket. Although in both the mutated Mt-trHbN and wild type Mt-trHbO the STG8 is blocked by WG8, the LT remains open in Mt-trHbN, allowing CO access into the heme cavity, whereas the main tunnel for CO migration in Mt-trHbO is the E7, as was previously described7. This fact shows that although the kon CO from mutant trHbN and wild type trHbO members are very similar, the ligand enters through different pathways, evidencing the complexity of mechanisms that regulate the ligand association process in these proteins.\n\n\nData availability\n\nF1000Research: Dataset 1. Experimental and theoretical calculations raw data, 10.5256/f1000research.5921.d4209139", "appendix": "Author contributions\n\n\n\nIB, LB, SC, RR, KLD, LAHB designed and performed the experiments. JPB, LB, MAM, DAE designed and analyzed the MD simulations. DAE, MAM, KSD and LB provided expertise in the field. KD, RLP, KLD, SS provided the experimental sample. JPB and LB wrote the manuscript. KSD, SS, LAHB, MTT and RP contributed to the experimental design and preparation of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by Framework program 7 NOstress Grant, CONICET, University of Buenos Aires, and Agencia Nacional de Promoción Científica y Tecnológica, National Institutes of Health Grant R01AI095173 and Universidad de la República (CSIC, Uruguay) to R. R. Additional funding to SC and RR was provided by PEDECIBA (Progama de Desarrollo de Ciencias Básicas, Uruguay) and CeBEM (Centro de Biología Estructural del Mercosur). IB and JPB hold CONICET PhD fellowships. LB is a Pew Latin American Fellow. LB, DAE and MAM are members of CONICET.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nFJ Luque is acknowledged for useful discussions and suggestions. Mehrnoosh Arrar is acknowledged for close reading of the manuscript.\n\n\nSupplementary material\n\n\n\n\nReferences\n\nSudre P, ten Dam G, Kochi A: Tuberculosis: a global overview of the situation today. Bull World Health Organ. 1992; 70(2): 149–159. PubMed Abstract \| Free Full Text\n\nMilani M, Pesce A, Nardini M, et al.: Structural bases for heme binding and diatomic ligand recognition in truncated hemoglobins. J Inorg Biochem. 2005; 99(1): 97–109. PubMed Abstract \| Publisher Full Text\n\nWittenberg JB, Bolognesi M, Wittenberg BA, et al.: Truncated hemoglobins: a new family of hemoglobins widely distributed in bacteria, unicellular eukaryotes, and plants. J Biol Chem. 2002; 277(2): 871–874. PubMed Abstract \| Publisher Full Text\n\nOuellet H, Ouellet Y, Richard C, et al.: Truncated hemoglobin HbN protects Mycobacterium bovis from nitric oxide. Proc Natl Acad Sci U S A. 2002; 99(9): 5902–5907. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrespo A, Martí MA, Kalko SG, et al.: Theoretical study of the truncated hemoglobin, HbN: exploring the molecular basis of the NO detoxification mechanism. J Am Chem Soc. 2005; 127(12): 4433–4444. PubMed Abstract \| Publisher Full Text\n\nFabozzi G, Ascenzi P, Renzi SD, et al.: Truncated hemoglobin GlbO from Mycobacterium leprae alleviates nitric oxide toxicity. Microb Pathog. 2006; 40(5): 211–220. PubMed Abstract \| Publisher Full Text\n\nBoechi L, Mañez PA, Javier Luque FJ, et al.: Unraveling the molecular basis for ligand binding in truncated hemoglobins: the trHbO Bacillus subtilis case. Proteins. 2010; 78(4): 962–970. PubMed Abstract \| Publisher Full Text\n\nBidon-Chanal A, Martı MA, Crespo A, et al.: Proteins. 2007; 464: 457–464.\n\nPerutz MF, Mathews FS: An X-ray study of azide methaemoglobin. J Mol Biol. 1966; 21(1): 199–202. PubMed Abstract \| Publisher Full Text\n\nBoechi L, Arrar M, Marti MA, et al.: Hydrophobic effect drives oxygen uptake in myoglobin via histidine E7. J Biol Chem. 2013; 288(9): 6754–61. PubMed Abstract \| Publisher Full Text \| Free Full Text .\n\nElber R: Ligand diffusion in globins: simulations versus experiment. Curr Opin Struct Biol. 2010; 20(2): 162–167. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOuellet H, Milani M, LaBarre M, et al.: The roles of Tyr(CD1) and Trp(G8) in Mycobacterium tuberculosis truncated hemoglobin O in ligand binding and on the heme distal site architecture. Biochemistry. 2007; 46(41): 11440–11450. PubMed Abstract \| Publisher Full Text\n\nBoechi L, Martí MA, Milani M, et al.: Structural determinants of ligand migration in Mycobacterium tuberculosis truncated hemoglobin O. Proteins. 2008; 73(2): 372–379. PubMed Abstract \| Publisher Full Text\n\nGuallar V, Lu C, Borrelli K, et al.: Ligand migration in the truncated hemoglobin-II from Mycobacterium tuberculosis: the role of G8 tryptophan. J Biol Chem. 2009; 284(5): 3106–3116. PubMed Abstract \| Publisher Full Text\n\nOlson JS, Phillips GN Jr: Myoglobin discriminates between O2, NO, and CO by electrostatic interactions with the bound ligand. J Biol Inorg Chem. 1997; 2(4): 544–552. Publisher Full Text\n\nOuellet YH, Daigle R, Lagüe P, et al.: Ligand binding to truncated hemoglobin N from Mycobacterium tuberculosis is strongly modulated by the interplay between the distal heme pocket residues and internal water. J Biol Chem. 2008; 283(40): 27270–27278. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBustamante JP, Abbruzzetti S, Marcelli A, et al.: Ligand uptake modulation by internal water molecules and hydrophobic cavities in hemoglobins. J Phys Chem B. 2014; 118(5): 1234–45. PubMed Abstract \| Publisher Full Text\n\nRozen S, Skaletsky H: Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 2000; 132: 365–86. PubMed Abstract \| Publisher Full Text\n\nWainwright LM, Wang Y, Park SF, et al.: Purification and spectroscopic characterization of Ctb, a group III truncated hemoglobin implicated in oxygen metabolism in the food-borne pathogen Campylobacter jejuni. Biochemistry. 2006; 45(19): 6003–6011. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMilani M, Pesce A, Ouellet Y, et al.: Mycobacterium tuberculosis hemoglobin N displays a protein tunnel suited for O2 diffusion to the heme. EMBO J. 2001; 20(15): 3902–3909. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPearlman DA, Case DA, Caldwell JW, et al.: Comput Phys Commun. 1995; 91: 1–41.\n\nWang J, Cieplak P, Kollman, PA: How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J Comput Chem. 2000; 21(12): 1049–1074. Publisher Full Text\n\nMarti MA, Capece L, Bidon-Chanal A, et al.: Nitric oxide reactivity with globins as investigated through computer simulation. Methods Enzymol. 2008; 437: 477–498. PubMed Abstract \| Publisher Full Text\n\nBikiel DE, Boechi L, Capece L, et al.: Modeling heme proteins using atomistic simulations. Phys Chem Chem Phys. 2006; 8(48): 5611–5628. PubMed Abstract \| Publisher Full Text\n\nCapece L, Lewis-Ballester A, Marti MA, et al.: Molecular basis for the substrate stereoselectivity in tryptophan dioxygenase. Biochemistry. 2011; 50(50): 10910–10918. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForti F, Boechi L, Bikiel D, et al.: Ligand migration in Methanosarcina acetivorans protoglobin: effects of ligand binding and dimeric assembly. J Phys Chem B. 2011; 115(46): 13771–13780. PubMed Abstract \| Publisher Full Text\n\nGiordano D, Boechi L, Vergara A, et al.: The hemoglobins of the sub-Antarctic fish Cottoperca gobio, a phyletically basal species--oxygen-binding equilibria, kinetics and molecular dynamics. FEBS J. 2009; 276(8): 2266–2277. PubMed Abstract \| Publisher Full Text\n\nNicoletti FP, Droghetti E, Howes BD, et al.: H-bonding networks of the distal residues and water molecules in the active site of Thermobifida fusca hemoglobin. Biochim Biophys Acta 2013; 1834(9): 1901–1909. PubMed Abstract \| Publisher Full Text\n\nMarti MA, Crespo A, Capece L, et al.: Dioxygen affinity in heme proteins investigated by computer simulation. J Inorg Biochem. 2006; 100(4): 761–770. PubMed Abstract \| Publisher Full Text\n\nPerissinotti LL, Marti MA, Doctorovich F, et al.: A microscopic study of the deoxyhemoglobin-catalyzed generation of nitric oxide from nitrite anion. Biochemistry. 2008; 47(37): 9793–9802. PubMed Abstract \| Publisher Full Text\n\nCohen J, Olsen KW, Schulten K: Finding gas migration pathways in proteins using implicit ligand sampling. Methods Enzymol. 2008; 437: 439–457. PubMed Abstract \| Publisher Full Text\n\nForti F, Boechi L, Estrin DA, et al.: Comparing and combining implicit ligand sampling with multiple steered molecular dynamics to study ligand migration processes in heme proteins. J Comput Chem. 2011; 32(10): 2219–2231. PubMed Abstract \| Publisher Full Text\n\nOuellet H, Juszczak L, Dantsker D, et al.: Reactions of Mycobacterium tuberculosis truncated hemoglobin O with ligands reveal a novel ligand-inclusive hydrogen bond network. Biochemistry. 2003; 42(19): 5764–5774. PubMed Abstract \| Publisher Full Text\n\nCouture M, Yeh SR, Wittenberg BA, et al.: A cooperative oxygen-binding hemoglobin from Mycobacterium tuberculosis. Proc Natl Acad Sci U S A. 1999; 96(20): 11223–11228. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOuellet Y, Milani M, Couture M, et al.: Ligand interactions in the distal heme pocket of Mycobacterium tuberculosis truncated hemoglobin N: roles of TyrB10 and GlnE11 residues. Biochemistry. 2006; 45(29): 8770–8781. PubMed Abstract \| Publisher Full Text\n\nLama A, Pawaria S, Bidon-Chanal A, et al.: Role of Pre-A motif in nitric oxide scavenging by truncated hemoglobin, HbN, of Mycobacterium tuberculosis. J Biol Chem. 2009; 284(21): 14457–14468. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPathania R, Navani N, Gardner A, et al.: Nitric oxide scavenging and detoxification by the Mycobacterium tuberculosis haemoglobin, HbN in Escherichia coli. Mol Microbiol. 2002; 45(5): 1303–1314. PubMed Abstract \| Publisher Full Text\n\nMartí MA, Bidon-Chanal A, Crespo A, et al.: Mechanism of product release in NO detoxification from Mycobacterium tuberculosis truncated hemoglobin N. J Am Chem Soc. 2008; 130(5): 1688–1693. PubMed Abstract \| Publisher Full Text\n\nBoron I, Bustamante JP, Davidge KS, et al.: Experimental and theoretical calculations raw data. F1000Research. 2015. Data Source" }	[ { "id": "7468", "date": "09 Feb 2015", "name": "Marco Nardini", "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 manuscript describes CO association kinetic constant measurements, ●NO decomposition, and molecular dynamics simulations on the wild type truncated Hb from Mycobacterium tuberculosis (Mt-trHbN) and two mutants (VG8F and VG8W) which introduce modifications in the two-tunnel system of the protein. The data are then compared to those from Mt-trHbO, suggesting that ligand migration is regulated by the internal tunnel architecture as well as by water molecules stabilized in the active site.Although the topic of the structure and the dynamic behavior of protein matrix tunnels in truncated Hb, and in particular in Mt-trHbN, has been “squeezed” a lot during the past years, the data reported in this paper add some new information and might be of interest in the field. The paper is well written (with regards to the requirements of the journal) and describes a technically sound piece of scientific research with data that supports the conclusions. Indexing is recommended, if the (few) minor comments below are addressed. Minor remarks:Abstract: line 4The authors write that “Mt-trHbO” possesses tunnels that are partially blocked ..” In fact normally trHbOs are associated with internal discrete cavities and not with tunnels. The authors should rephrase the sentence. Abstract: line 11Mt-trHbN should not be in Italics. Abstract: line 12The sentence “mutations introduce modifications in both tunnel topologies” is quite cryptic and it is not clear what the author mean with “tunnel topologies”. The authors should rephrase the sentence to clarify it. Introduction: page 3, first column, line 8The authors might want to include a review on trHbs more recent than that indicated in reference (3). There are several of them published in the last few years. Introduction: page 3, first column, line 23The paragraph starting from line 23 is a bit misleading because the authors try to generalize the description of the protein matrix tunnels in trHbs by mixing what happens in trHbNs and trHbOs. This is confusing since it might give the impression that three tunnels co-exist in trHbs, which is not true. In this respect, Figure 1 contributes a lot to make confusion, since it is not clear which trHb protein represents and it seems that it has three co-existing tunnels. It is probably better to keep separate trHbNs and trHbOs, both in the text description and in Figure 1. The authors should describe the tunnel features in trHbN (short and long tunnel) and trHbO (cavities, small E7 residues an possible E7 gating), and show two panels in Figure 1 with depicted the tunnel/cavity systems in Mt-trHbN (panel A) and Mt-trHbO (panel B), possibly using a similar protein orientation and highlighting the role of the G8 residue in the two cases. Introduction: page 3, second column, line 7The sentence regarding the “internal water molecules” is too generic as it is written now, since it is not clear if the authors refer to globins, to trHbs or to Mt-trHbs. The authors should rephrase the sentence to clarify this issue. Introduction: page 3, second column, line 10The authors should say that the experimental measurements and the MD simulations have been performed only on Mt-trHbN and mutants, and not, for instance, on Mt-trHbO. Materials and Methods: page 3, second column, line 40The purification paragraph seems to refers only to trHbN. What about its mutants? The authors should add a sentence to clarify this issue. Materials and Methods: page 4, first column, line 44The following sentence is not written fully correctly:“Amino acids protonation states were assumed based on environment of the residue in the crystal structure. All solvent-exposed His were protonated at the N- δ delta atom, as well as HisF8, because of its coordination to the heme iron”.One possibility is to rephrase it as follows:“The protonation state of the amino acids was assumed based on the environment of the residues in the crystal structure. All solvent-exposed His residues were protonated at the N-δ atom, as well as the proximal HisF8, because of its coordination to the heme iron”. page 5, second column, line 12It is not clear what the authors mean when they write that the STG8 “has only the distal site cavity, (trHb : CO)1, …”, especially if this sentence is coupled with Figure 4A, where (trHb : CO)1 seems to be connected to STG8 through (trHb : CO)2. page 5, Title of Table 1It is probably better to change “..for wild type and mutants Mt-trHbs O and N” to “..for wild type and mutants of Mt-trHbN and Mt-trHbO” page 6, Figure 4 legendIn the legend of Panel C there is no mention of the (trHb : CO)3 site. References: page 9Reference (21) is missing the title", "responses": [ { "c_id": "1459", "date": "22 Jul 2015", "name": "Leonardo Boechi", "role": "Author Response", "response": "Abstract: line 4The authors write that “Mt-trHbO” possesses tunnels that are partially blocked ..” In fact normally trHbOs are associated with internal discrete cavities and not with tunnels. The authors should rephrase the sentence.We rephrase the sentence “whereas Mt-trHbO possesses tunnels that are partially blocked by a few bulky residues, particularly atryptophan at position G8” by writing “whereas Mt-trHbO possesses tunnels interrupted by a few bulky residues, particularly a tryptophan at position G8” Abstract: line 11Mt-trHbN should not be in Italics.Corrected, thanks Abstract: line 12The sentence “mutations introduce modifications in both tunnel topologies” is quite cryptic and it is not clear what the author mean with “tunnel topologies”. The authors should rephrase the sentence to clarify it.We rephrase the sentence “These mutations introduce modifications in both tunnel topologies and affect the incoming ligand capacity to displace retained water molecules at the active site.” By writing “These mutations affect both the tunnels accessibility as well as the affinity of distal site water molecules, thus modifying the ligand access to the iron” Introduction: page 3, first column, line 8The authors might want to include a review on trHbs more recent than that indicated in reference (3). There are several of them published in the last few years.As suggested by the review we modified the references by more recent ones: Davidge & Dikshit (2013). Introduction: page 3, first column, line 23The paragraph starting from line 23 is a bit misleading because the authors try to generalize the description of the protein matrix tunnels in trHbs by mixing what happens in trHbNs and trHbOs. This is confusing since it might give the impression that three tunnels co-exist in trHbs, which is not true. In this respect, Figure 1 contributes a lot to make confusion, since it is not clear which trHb protein represents and it seems that it has three co-existing tunnels. It is probably better to keep separate trHbNs and trHbOs, both in the text description and in Figure 1. The authors should describe the tunnel features in trHbN (short and long tunnel) and trHbO (cavities, small E7 residues an possible E7 gating), and show two panels in Figure 1 with depicted the tunnel/cavity systems in Mt-trHbN (panel A) and Mt-trHbO (panel B), possibly using a similar protein orientation and highlighting the role of the G8 residue in the two cases.As suggested by the review, we modified the sentence to clarify.“Three internal tunnels were found in the truncated hemoglobin family:” by “Three different internal tunnels have been characterize among the trHb members, in general one or two of these tunnels is found in each protein:”We also change Figure 1 and its caption as suggested.Caption Figure 1. Schematic representation of the two pathways for ligand migration presented in M. tuberculosis trHbN. The Long Tunnel (LT) and Short Tunnel G8 (STG8) are shown in orange. Introduction: page 3, second column, line 7The sentence regarding the “internal water molecules” is too generic as it is written now, since it is not clear if the authors refer to globins, to trHbs or to Mt-trHbs. The authors should rephrase the sentence to clarify this issue.As suggested by the review we clarify the sentence: “It has also been noted that in myoglobin, M. Tuberculosis trHbN as well as in T. fusca trHbO, internal water molecules were observed to block the heme accessibility, thus delaying ligand binding” Introduction: page 3, second column, line 10The authors should say that the experimental measurements and the MD simulations have been performed only on Mt-trHbN and mutants, and not, for instance, on Mt-trHbO.As suggested by the review we clarify the sentence by adding explicitly the name of the protein studied “By performing CO association kinetic constant measurements (…) of Mt-trHbN, we addressed molecular mechanisms that control ligand association in M. Tuberculosis truncated hemoglobins” Materials and Methods: page 3, second column, line 40The purification paragraph seems to refers only to trHbN. What about its mutants? The authors should add a sentence to clarify this issue.We clarify this by modifying “The trHbN protein was” by “The trHbN protein variants were” Materials and Methods: page 4, first column, line 44The following sentence is not written fully correctly:“Amino acids protonation states were assumed based on environment of the residue in the crystal structure. All solvent-exposed His were protonated at the N- δ delta atom, as well as HisF8, because of its coordination to the heme iron”.One possibility is to rephrase it as follows:“The protonation state of the amino acids was assumed based on the environment of the residues in the crystal structure. All solvent-exposed His residues were protonated at the N-δ atom, as well as the proximal HisF8, because of its coordination to the heme iron”.We thank the reviewer for the suggestion, the phrase was modified as suggested. page 5, second column, line 12It is not clear what the authors mean when they write that the STG8 “has only the distal site cavity, (trHb : CO)1, …”, especially if this sentence is coupled with Figure 4A, where (trHb : CO)1 seems to be connected to STG8 through (trHb : CO)2.We modified the sentence by adding information:On the one hand, the LT connects three internal cavities: (trHb : CO) 1 , (trHb : CO) 2 and (trHb : CO) 3 . The STG8, on the other hand, connected to both the cavity (trHb : CO) 2 and the solvent. page 5, Title of Table 1It is probably better to change “..for wild type and mutants Mt-trHbs O and N” to “..for wild type and mutants of Mt-trHbN and Mt-trHbO”We thank the reviewer for the suggestion; the phrase was modified as suggested. page 6, Figure 4 legendIn the legend of Panel C there is no mention of the (trHb : CO)3 site.We thank the reviewer for the suggestion, we rephrase as “Free energy profiles over STG8 (B) and LT (C) connecting the solvent with the distal site through the cavities (trHb : CO) 1, (trHb : CO) 2 and (trHb : CO) 3, for wild type (green), VG8F (orange) and VG8W (violet) mutant Mt-trHbN.” References: page 9Reference (21) is missing the titleWe thank the reviewer; the reference was modified" } ] }, { "id": "7848", "date": "24 Apr 2015", "name": "Michael Wilson", "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 succinct and generally well written paper reporting experimental data on ligand binding to Mt-trHbN and to two mutants (VG8F anndVG8W)).These latter have been designed to test hypotheses regarding the possible routes by which access is gained to the heme by small neutral gaseous molecules. The experiments and the molecular modelling that supports them, and which provides mechanistic insights, have been carefully performed. The results are of interest to the field as they add to the body of accumulated evidence that proteins, including those with the function of binding small neutral molecules, provide specific, and often dynamic, channels to permit rapid access to the binding site. Furthermore the kinetics of binding are seen to be strongly influenced by single amino acid substitution in the access channels. Although the results support the general conclusions drawn by the authors some clarification of a number of points would be helpful. These are given below. Why in Figs 2A and B does time appear not to start at t=0? In Fig and discussion is it proposed that the water molecule is bound to the iron (common for ferric but not for ferrous iron) or stabilised in that location only by hydrogen binding. It is presumed that for the modelling the iron is in the ferrous state as the authors are discussing CO binding. Although the authors discuss the decomposition of NO catalysed by Mt-trHbN in the presence of oxygen the assays do not make it clear that this is the reaction under study. No mention of the oxygen concentration is made in the legend to Fig 6. It seems oxygen is present to account for the disappearance of NO. From Fig 6 it is stated that NO binds to the protein but is not degraded (e.g. to nitrate via peroxynitrite). Is this because regeneration of the reduced heme (necessary for oxygen binding) is so slow or is it because as ferric heme is reduced by the NADPH/ferredoxin system NO binds before oxygen and thus no turnover occurs as it is known that the NO-ferrous complex does not react with oxygen? In any case it is not made clear to what chemical step the measured kinetics refer. It would improve the manuscript if the authors clarified these points.", "responses": [ { "c_id": "1460", "date": "22 Jul 2015", "name": "Leonardo Boechi", "role": "Author Response", "response": "Why in Figs 2A and B does time appear not to start at t=0?We thank the reviewer for noticing that specific issue. We decided to discard the first 8ms because they were too noisy. In Fig and discussion is it proposed that the water molecule is bound to the iron (common for ferric but not for ferrous iron) or stabilised in that location only by hydrogen binding. It is presumed that for the modelling the iron is in the ferrous state as the authors are discussing CO binding. We agree with the reviewer that the water bounds in general very weak to the ferrous iron. However, there is evidence showing that in a polar distal site (with polar residues), as in the case of truncated hemoglobins, water molecules remain inside stabilized by the polar residues, and thus slow ligand binding to ferrous iron (Olson and Phillips Jr, 1997; Ouellet et al., 2008). Although the authors discuss the decomposition of NO catalysed by Mt-trHbN in the presence of oxygen the assays do not make it clear that this is the reaction under study. No mention of the oxygen concentration is made in the legend to Fig 6. It seems oxygen is present to account for the disappearance of NO. From Fig 6 it is stated that NO binds to the protein but is not degraded (e.g. to nitrate via peroxynitrite). Is this because regeneration of the reduced heme (necessary for oxygen binding) is so slow or is it because as ferric heme is reduced by the NADPH/ferredoxin system NO binds before oxygen and thus no turnover occurs as it is known that the NO-ferrous complex does not react with oxygen? In any case it is not made clear to what chemical step the measured kinetics refer. It would improve the manuscript if the authors clarified these points.As suggested by the review we modified the caption of Figure 6, and also we add a sentence.The legend of the caption can be changed to:Figure 6. •NO decomposition by Mt-trHbN at ambient oxygen concentrations (approx. 200 µM, not measured). (A) •NO decay was monitored amperometrically in the absence (red trace) and the presence (black trace) of Mt-trHbN added at the apex of the signal response to 2 μM ProliNONOate. Data are representative of 3 technical repeats. (B) Mean rates of •NO decay in the presence of wild type Mt-trHbN or site-directed mutants from 3 technical repeats ± S.E.M *P < 0.05, unpaired t-test.The sentence added is:The chemical step being measured in this assay is the reaction between •NO and the oxyferrous heme; once this reaction has concluded, we assume that the heme is restored from ferric back to ferrous. We are unsure why the reaction is single turnover but it could be due to (a) rapid binding of •NO to the ferrous complex before oxygen can bind, rendering it unable to bind oxygen and initiate the reaction or (b) due to slow reduction of Mt-trHbN by the non-native E. coli FdR." } ] } ]	1	https://f1000research.com/articles/4-22
https://f1000research.com/articles/4-298/v1	22 Jul 15	{ "type": "Study Protocol", "title": "Examining Factors Influencing Colorectal Cancer Screening of Rural Nebraskans Using Data from Clinics Participating in an Accountable Care Organization: A Study Protocol", "authors": [ "Lufei Young", "Jungyoon Kim", "Hongmei Wang", "Li-Wu Chen", "Jungyoon Kim", "Hongmei Wang", "Li-Wu Chen" ], "abstract": "Background: Although mortality rates of colorectal cancer (CRC) can be significantly reduced through increased screening, rural communities are still experiencing lower rates of screening compared to urban counterparts. Understanding and eliminating barriers to cancer screening will decrease cancer burden and lead to substantial gains in quality and quantity of life for rural populations. However, existing studies have shown inconsistent findings and fail to address how contextual and provider-level factors impact CRC screening in addition to individual-level factors. Purpose: The purpose of the study is to examine multi-level factors related to CRC screening, and providers’ perception of barriers and facilitators of CRC screening in rural patients cared for by accountable care organization (ACO) clinics.Methods/Design: This is a convergent mixed method design. For the quantitative component, multiple data sources, such as electronic health records (EHRs), Area Resource File (ARF), and provider survey data, will be used to examine patient-, provider-, clinic-, and county-level factors. About 21,729 rural patients aged between 50 and 75 years who visited the participating ACO clinics in the past 12 months are included in the quantitative analysis. The qualitative methods include semi-structured in-depth interviews with healthcare professionals in selected rural clinics. Both quantitative and qualitative data will be merged for result interpretation. Quantitative data identifies “what” factors influence CRC screening, while qualitative data explores “how” these factors interact with CRC screening. The study setting is 10 ACO clinics located in nine rural Nebraska counties.Discussion: This will be the first study examining multi-level factors related to CRC screening in the new healthcare delivery system (i.e., ACO clinics) in rural communities. The study findings will enhance our understanding of how the ACO model, particularly in rural areas, interacts with provider- and patient-level factors influencing the CRC screening rate of rural patients.", "keywords": [ "Screening", "Colorectal Cancer", "accountable care organisations", "Barriers", "Rural", "Electronic Health Records", "Mixed Methods" ], "content": "Background\n\nCancer is the second most common cause of death in the US1. The colorectal cancer (CRC) incidence rate for Nebraska is higher than for the US as a whole (50 for men, 37.8 for woman per 100,000 in US vs. 54.9 for men, 42.9 for women per 100,000 in Nebraska)1,2. The CRC mortality rate for Nebraska is also higher than for the US in both men (20.4 vs. 19.1 per 100,000) and women (15 vs. 13.5 per 100,000)1,2. Cancer screening plays a vital role in cancer prevention3. The US Preventive Services Task Force recommends that adults aged between 50 and 75 have a CRC screening, including fecal occult blood testing (FOBT) annually, sigmoidoscopy every 5 years, or colonoscopy every 10 years4. The decrease in both cancer incidence and death rates was significantly associated with the uptake of cancer screening and improved early detection1. However, disparities in CRC screenings persist in rural communities5. Compared to urban residents, rural residents had lower CRC screening rates (48% vs. 55%)6. Remote rural residents had the lowest screening rates overall (45%)6.\n\nStudies have reported factors related to CRC screening rate in rural at three levels: patient5, provider5,7 and contextual (e.g., county, rural clinics)8. Patient-level barriers included social economic status7, family history9, access to care10, comorbidity11, health literacy5, cost5, and healthcare utilization patterns (e.g., regular physician visits)9,12. Among all the patient-level factors, receiving providers’ recommendation was one of the most commonly reported factors associated with CRC screening5,7,13–17. Provider-level factors influencing CRC screening were also well documented, including perceived support18,19, available time and workload18,19, attitude and belief20, competing priorities18,19, and patient load18. Other non-modifiable provider-level factors, such as provider’s age, gender and practice experience, also played a part in the patients’ screening behaviors21. Recently, more studies have begun to examine the contextual factors, such as area poverty rate, rural clinic practice capacity, supply of rural providers (e.g., primary care physicians or specialists), which also significantly affect cancer screening behaviors6,22,23. Despite studies conducted to address the contextual factors associated with CRC screening, the findings have been mixed as a result of variations in research design, conceptual frameworks, the use of incomplete data sources, and measurement issues18,22. For instance, Stimpson et al. found that the supply of specialists (e.g., gastroenterologists) is positively associated with CRC screening based on a Texas-based self-reported survey22, while another study highlighted the importance of both generalists and specialists on CRC screening for the white population only, based on a single state’s Medicare claims data18. The data sources used in each study (i.e., issuance claim data and/or self-report surveys) have inherent problems affecting the reliability and validity of study findings6,22. Furthermore, neither of the studies were designed to address rural specific factors related to CRC screening. As a result, these findings were contrary to what was reported in Greiner’s study24 in which CRC screening among rural populations was not significantly related to the supply of physicians performing endoscopic procedures.\n\nThe interventions designed to improve rural cancer screenings have been primarily focused on overcoming patient and provider level barriers, without much consideration of contextual and delivery system level factors13,15,25–27. Consequently, the sustained effects of these interventions on CRC screening are uncertain. A possible explanation could be that these interventions failed to address the barriers at the healthcare system level. For instance, under the current healthcare delivery system, care providers who were paid by volume experienced high pressure to increase volume as the reimbursement rate declined, which resulted in shortened office visit time and reduced opportunities to recommend preventive services during the visit28. The situation can be worse in rural clinics, where a shortage of primary care providers causes patient overload, with a large number of patient pools being covered by few clinic staff members29,30. Furthermore, without reliable data sources, such as a cancer screening registry or a state-wide electronic medical record system, it is difficult to track rural patients’ cancer screening status objectively, which further makes the evaluation of intervention effects challenging.\n\nAccountable care organizations (ACOs) are a group of health care providers joined together to improve quality of care with lower costs31,32, by emphasizing the mechanism of care coordination, strong patient-physician relationships, use of health information technology, and value-based provider incentive systems32. As a new healthcare delivery alternative, ACOs create opportunities but also challenges for rural healthcare providers31. One of the requirements to become an ACO clinic is mandatory performance data tracking and reporting. This could potentially enhance patient care coordination and increase care providers’ motivation and awareness of CRC screening. However, at the same time, this could potentially increase workload for rural clinics and providers who are already stretched thin with heavy patient loads and limited resources. To date, the interaction between the new healthcare system (ACO clinics) and patient-/provider-level factors affecting CRC screening in rural populations has not been reported.\n\nThe purpose of the proposed study is to examine the mechanisms of multi-level factors associated with colorectal cancer screening within the ACO context in rural Nebraska. To achieve this purpose, we have the following specific aims:\n\n1. To identify patient-, provider-, and county-level factors influencing CRC screening of patients in rural Nebraska using data extracted from electronic health records and surveys provided by the ACO clinic providers.\n\n2. To explore healthcare professionals’ views of barriers and facilitators of CRC screening in the ACO context, using the data collected through in-depth interviews.\n\nBased on our literature review and clinical expert input, we developed a conceptual framework derived from Gelberg-Anderson’s healthcare use behavioral model33. The conceptual framework will assist in understanding rural residents’ cancer screening behavior and its correlation with individual, provider, and county level factors (Figure 1). The model posits that cancer screening is a function of predisposing factors, enabling factors and needs at both the patient and provider levels. The model also posits that county-level factors, such as socioeconomic indicators and rural health resources, influence patient- and provider-level factors. The hypothesis illustrated by the conceptual framework will direct us in study design, variable selection, outcome measure, data collection and analysis, as well as in result interpretation.\n\n\nMethods/Design\n\nThe proposed study will use a convergent mixed method design to identify individual-, provider-, and county-level factors that influence CRC screening (Figure 2). To address the specific aims, we will use multiple data sources including EHR, Areas Resource Files (ARF), and data collected from care provider survey and interviews (Table 1). The study was approved by the University of Nebraska Medical Center Institutional Review Board (IRB) and all participating rural ACO clinics, receiving the number of IRB PROTOCOL # 352-15-EP.\n\nThe quantitative analysis will answer “what” determines patients’ CRC screening by linking individual, provider, and system-level factors to screening outcomes, while the qualitative analysis will address “how,” or in what mechanism, these factors facilitate or hinder CRC screening in the ACO context. Both quantitative and qualitative data will be concurrently collected, and data will be merged during data analysis and result interpretation.\n\nThe study setting is a community-based ACO in rural Nebraska, which started as an advance payment ACO in the Medicare Shared Savings Program with ten independent primary care clinics, taking care of more than 14,000 Medicare patients. These clinics are located in rural counties in Nebraska and range in size from four to twelve primary care providers. All of the ACO clinics have adopted an electronic health records system with varying degrees of implementation.\n\nAim 1. To identify patient-, provider-, and county-level factors influencing CRC screening of patients in rural Nebraska using data extracted from electronic health records and surveys provided by the ACO clinic providers.\n\nData source. The retrospective chart review will be conducted to obtain patient- and provider-level data from the ACO clinics using their electronic health records (EHR). De-identified EHRs will be used for data analysis. The IRB has granted a waiver of patient consent for the retrospective chart review. In addition, county-level characteristics for counties where the patients reside will be obtained from the Area Resource File, administered by U.S. Department of Health and Human Service, Health Resources and Services Administration.\n\nStudy sample. The inclusion criteria for the EHRs are: 1) the patient aged between the ages of 50 and 75 years old; 2) the patient has visited an ACO clinic at least once during the past 12 months. A total number of 21,729 patient records achieves 100% power to detect a small effect size (0.10) using a 1 degree of freedom Chi-Square Test with a significance level (alpha) of 0.05. For the provider- and county-level data, we have a total of over 50 providers including physicians, physician assistants (PAs), and nurse practitioners (NPs) from the participating clinics, providing care to over 20 counties in rural Nebraska. The expected total number of providers and counties would be sufficient to support multi-level analysis adjusting correlations at both the provider-level and the county-level. We could not find previous multi-level studies analogous to our study model. Thus, our analysis will be the first to estimate effect sizes of the explanatory variables at different levels and the random effects, which will benefit power and sample size calculation for a similar study on a larger scale in the future.\n\nStudy procedure. The research staff will work with the clinical data specialist to extract all relevant data fields for patients aged 50 to 75 years old. The dataset will be de-identified for the purpose of confidentiality and protection of patient privacy before being transferred to the researchers for analysis.\n\nVariables. The main outcome variable of interest is whether patients are up-to-date in CRC screening, which is defined based on the US Preventive Services Task Force (USPSTF) guideline: a colonoscopy every 10 years, fecal occult blood test (FOBT) every year, or sigmoidoscopy every 5 years for adults aged 50 to 75 years old with no prior CRC and no family history of CRC. (http://www.uspreventiveservicestaskforce.org/Page/Topic/recommendation-summary/colorectal-cancer-screening) To determine whether there are different barriers for the three types of test, we will also create three dummy variables indicating the type of test patients received for sensitive analysis. Table 2 illustrates patient, provider, and county level variables that will be included in our model. Patient-level data will be obtained from EHRs, provider-level data will be obtained second-handedly from ACO clinics through their administration or provider-survey data; county-level data will be obtained from the publicly available ARF.\n\nAnalysis. We will first run descriptive statistics on all patient characteristics: (1) mean and standard deviation are used to report continuous variables; (2) frequency and percentage are used to report categorical variables. Chi-square tests will then be performed to examine if there are statistically significant differences in each of the patient characteristics between those up-to-date on CRC screening and those not up-to-date. To account for the correlation among patients clustered with provider and county level, a generalized linear mixed effects model will be used to examine the simultaneous effects of all patient-, provider- and county-level characteristics on CRC screening after controlling for other characteristics. This study is the first to control for correlations at two cluster levels when examining the factors influencing CRC screening. Fixed Effects model and Random Effects model will both be conducted to examine the mechanisms that link factors to screening outcomes and the interaction between different levels. SAS version 9.2 will be used for data analysis.\n\nAim 2. Identify healthcare professionals’ view of the challenges and opportunities of CRC screening under ACO context.\n\nData source. The research team will use semi-structured surveys and in-depth interviews. The two methods will be used in parallel to triangulate methodological weaknesses of self-administered surveys and in-depth interviews.\n\nStudy sample. The inclusion criteria for participants for Aim 2 are health care professionals working in rural ACO clinics. Healthcare providers are defined as physicians, PAs, NPs, nurses, and care coordinators.\n\nStudy procedure. a) Survey. A paper-and-pencil, self-reported survey will be distributed to healthcare professionals working in ACO clinics, including physicians, PAs, NPs, nurses, and care coordinators. The survey questionnaire will be developed by the research team in collaborating with ACO partners as a part of ACO’s annual continuing medical education program. The survey will assess healthcare professionals’ knowledge, attitude, practice pattern, and perceived barriers of CRC screening, as well as delivery system characteristics (e.g., ACO) that influence CRC screening. A combination of closed-ended and open-ended question will be used. The pilot testing will be conducted to assess the content validity and reliability of the tool. The survey data will be analyzed using SPSS version 2234. Descriptive analysis will be used to illustrate care provider characteristics, provider enabling factors and needs, and ACO characteristics related to CRC screening.\n\nb) In-depth interview. In parallel with the survey, the research team will conduct interviews with 15–20 key informants, including two or three persons from each professional role in the ACO setting: physicians, PAs, NPs, administrators, nurses, and care coordinators. The research team will use the combination of convenient and purposive sampling, as different professional roles (e.g., administrator or physician) will provide unique aspects about provider and delivery system level factors under the ACO context. Interviewees will be asked about their perception of barriers and facilitators of CRC screening under rural ACO contexts, as well as their opinion of how ACO model is interacting with the promotion of CRC screening (Table 3).\n\nc) Recruitment and data collection mode. The research team will attend regular ACO board meetings and care-coordinator meetings to identify and recruit key informants for interviews. Invitation letters and flyers will also be used to raise awareness of the study and to promote participation rate. Face-to-face or telephone interviews will be conducted depending on the preference of interviewees. The interview will be 30 to 35 minutes in length and will be audio-recorded and transcribed by experienced and professionally trained research staff. A cross-validation of the interview transcript will be conducted by the research team. Table 4 lists the interview protocol regarding the IRB, compensation, interview plan, and confidentiality.\n\nIRB approval will be attained from UNMC and a cover letter expressing research goals, procedures, potential benefits, and risks will be developed by the research team and provided to the participants.\n\nEach interview participant will receive incentives for their time and expertise.\n\nThe interview guide will be developed and reviewed by the research team and expert panels prior to dissemination.\n\nAt least two investigators will participate in the interview. One person will ask questions and another person will take notes. Interviewer training will be completed prior to the interview.\n\nAll interview data will be stored in password protected computers and UNMC secure servers and will not be shared with any other person outside of the research team, except for academic publication. Individual identifiers will not be revealed in publication.\n\nAnalysis. Data will be analyzed by inductive (ground-up) and deductive development and organization of thematic codes. Using the notes taken by the researchers and literature review, the research team will develop a coding structure, which includes key conceptual domains and participant perspectives. Minor modifications will be made iteratively until the model is saturated. Data will be coded and analyzed using NVivo qualitative analysis software (QSR NVivo 10)35.\n\nData management protocol has been developed for this study, including guidelines and procedures for data collection, validation, entry, storage, analysis and dissemination. All study data will be stored in the Research Electronic Data Capture (REDCap) database (http://www.project-redcap.org/). REDCap is a reliable and secure web-based application that allows for comprehensive management of the data collection process that is supported at University of Nebraska Medical Center and University of Iowa. Study participants will have access to de-identified data. The results of this study will be disseminated through publications and presentations. The dataset be provided for public and statistical use.\n\n\nDiscussion\n\nTo our knowledge, this will be the first study quantitatively and qualitatively examining multi-level factors influencing CRC screening in the new healthcare delivery system (i.e., ACO clinics) in rural communities. The evidence of how the new rural ACO clinics interact with county-, provider- and provider-level factors and the combined effects on cancer screening is missing in rural settings. Our study will fill these knowledge gaps through a two-step approach using clinic-level data: 1) quantitatively examine multi-level factors influencing CRC screening in rural adults between age 50 and 75 receiving care from ACO clinics; 2) qualitatively explore factors related to CRC screening guided by the findings from quantitative data. Results from the proposed study will provide practice and managerial implications to the field by helping ACO clinicians and administrators to best utilize ACO infrastructures, such as care coordination, health information technology, value-based incentive system, and reporting of performance measures, to promote cancer screening of rural patients.\n\nIn addition, the study will address the problems with current literature in terms of the inconsistent findings, limitations in data sources, and missing evidence related to CRC screening within the context of the new rural healthcare system. Given that effective and sustained interventions require strategies aligning provider- and patient-level factors with care delivery system and community characteristics28, the findings will help identify and develop strategies to target multi-level factors related to CRC screening in rural areas. Furthermore, the study will provide managerial implications for the operations of ACO organizations and impact policy changes in rural settings.\n\nThe project will help develop the practice-based research network between an academic setting and rural ACO clinics in Nebraska to promote cancer screening. If feasible and sustainable, we will continue to build a larger scale cancer research initiative, as well as extend to other practice-driven research programs (e.g., obesity related cancer prevention and control, interventions to manage cancer in patients with competing co-morbidities, and house-call programs for cancer patients living with complex complications, etc.).\n\nThe partnership between academic and rural ACO clinics will help identify the clinic sites and capstone project topics for students; while the clinics can utilize academic resources to conduct mandatory performance improvement projects and measure tracking.\n\n\nConclusion\n\nEliminating barriers to CRC screening could lead to substantial gains in quality and quantity of life and decrease the CRC burden on public health; however, sustained and effective interventions to promote screening remain uncertain. Our study will help determine the mechanism of effective intervention to optimize CRC screening by qualitatively and quantitatively examining the impact of multi-level factors on CRC screening in rural communities. To explore the additional data resources, we will use clinic data and ACO clinic electronic health records to conduct our study. If successful, our findings will add evidence and inform the design of effective interventions tailored to promote cancer screening in rural populations.", "appendix": "Author contributions\n\n\n\nAll investigators, Dr. Young, Dr. Kim, and Dr. Wang contributed to the development of the study protocol from study concept, design, generating aims, planning data collection and analysis, preparing and revising the manuscript. Dr. Chen provided multiple critiques of the drafts. All authors have agreed to the final version of this protocol.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nResearch reported in this publication was supported by the University of Nebraska Medical Center College of Public Health and Fred & Pamela Buffett Cancer Center Support Grant (P30CA036727).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nEthical approvals\n\nThe study protocol was approved by the Institutional Review Board at the University of Nebraska Medical Center, receiving the number of IRB PROTOCOL # 352-15-EP.\n\n\nReferences\n\nCancer Facts. Cancer facts. 2015.\n\nSiegel RL, Miller KD, Jemal A: Cancer statistics, 2015. CA Cancer J Clin. 2015; 65(1): 5–29. PubMed Abstract \| Publisher Full Text\n\nStern C: Flexible sigmoidoscopy versus fecal occult blood testing for colorectal cancer screening in asymptomatic individuals. Clin J Oncol Nurs. 2014; 18(4): 471–472. PubMed Abstract \| Publisher Full Text\n\nSmith RA, Cokkinides V, Brawley OW: Cancer screening in the United States, 2012: A review of current American Cancer Society guidelines and current issues in cancer screening. CA Cancer J Clin. 2012; 62(2): 129–142. PubMed Abstract \| Publisher Full Text\n\nJilcott Pitts SB, Lea CS, May CL, et al.: \"Fault-line of an earthquake\": a qualitative examination of barriers and facilitators to colorectal cancer screening in rural, Eastern North Carolina. J Rural Health. 2013; 29(1): 78–87. PubMed Abstract \| Publisher Full Text\n\nCole AM, Jackson JE, Doescher M: Urban-rural disparities in colorectal cancer screening: Cross-sectional analysis of 1998-2005 data from the centers for disease control's behavioral risk factor surveillance study. Cancer Med. 2012; 1(3): 350–356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAtassi K, Nemeth L, Edlund B, et al.: Adapting the PPRNet TRIP QI model to increase colorectal cancer screening in primary care.2012; 3: 866–873. Publisher Full Text\n\nShariff-Marco S, Breen N, Stinchcomb DG, et al.: Multilevel predictors of colorectal cancer screening use in California. Am J Manag Care. 2013; 19(3): 205–216. PubMed Abstract\n\nYoung WF, McGloin J, Zittleman L, et al.: Predictors of colorectal screening in rural colorado: Testing to prevent colon cancer in the high plains research network. J Rural Health. 2007; 23(3): 238–245. PubMed Abstract \| Publisher Full Text\n\nSeeff LC, Nadel MR, Klabunde CN, et al.: Patterns and predictors of colorectal cancer test use in the adult U.S. population. Cancer. 2004; 100(10): 2093–2103. PubMed Abstract \| Publisher Full Text\n\nLukin DJ, Jandorf LH, Dhulkifl RJ, et al.: Effect of comorbid conditions on adherence to colorectal cancer screening. J Cancer Educ. 2012; 27(2): 269–276. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCenters for Disease Control and Prevention (CDC). Vital signs: colorectal cancer screening test use--United States, 2012. MMWR Morb Mortal Wkly Rep. 2013; 62(44): 881–888. PubMed Abstract\n\nHoneycutt S, Green R, Ballard D, et al.: Evaluation of a patient navigation program to promote colorectal cancer screening in rural Georgia, USA. Cancer. 2013; 119(16): 3059–3066. PubMed Abstract \| Publisher Full Text\n\nWilkins T, Gillies RA, Harbuck S, et al.: Racial disparities and barriers to colorectal cancer screening in rural areas. J Am Board Fam Med. 2012; 25(3): 308–317. PubMed Abstract \| Publisher Full Text\n\nCurry WJ, Lengerich EJ, Kluhsman BC, et al.: Academic detailing to increase colorectal cancer screening by primary care practices in Appalachian Pennsylvania. BMC Health Serv Res. 2011; 11: 112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoughlin SS, Costanza ME, Fernandez ME, et al.: CDC-funded intervention research aimed at promoting colorectal cancer screening in communities. Cancer. 2006; 107(5 Suppl): 1196–1204. PubMed Abstract \| Publisher Full Text\n\nCoughlin SS, Thompson T: Physician recommendation for colorectal cancer screening by race, ethnicity, and health insurance status among men and women in the United States, 2000. Health Promot Pract. 2005; 6(4): 369–378. PubMed Abstract \| Publisher Full Text\n\nRosenwasser LA, McCall-Hosenfeld JS, Weisman CS, et al.: Barriers to colorectal cancer screening among women in rural central pennsylvania: primary care physicians' perspective. Rural Remote Health. 2013; 13(4): 2504. PubMed Abstract \| Free Full Text\n\nHatcher J, Dignan MB, Schoenberg N: How do rural health care providers and patients view barriers to colorectal cancer screening? Insights from appalachian kentucky. Nurs Clin North Am. 2011; 46(2): 181–192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGimeno García AZ: Factors influencing colorectal cancer screening participation. Gastroenterol Res Pract. 2012; 2012: 483417. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeiss JM, Smith MA, Pickhardt PJ, et al.: Predictors of colorectal cancer screening variation among primary-care providers and clinics. Am J Gastroenterol. 2013; 108(7): 1159–1167. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStimpson JP, Pagán JA, Chen LW: Reducing racial and ethnic disparities in colorectal cancer screening is likely to require more than access to care. Health Aff (Millwood). 2012; 31(12): 2747–2754. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZapka JM, Klabunde CN, Arora NK, et al.: Physicians' colorectal cancer screening discussion and recommendation patterns. Cancer Epidemiol Biomarkers Prev. 2011; 20(3): 509–521. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGreiner KA, Engelman KK, Hall MA, et al.: Barriers to colorectal cancer screening in rural primary care. Prev Med. 2004; 38(3): 269–275. PubMed Abstract \| Publisher Full Text\n\nKreuter MW, Garibay LB, Pfeiffer DJ, et al.: Small media and client reminders for colorectal cancer screening: Current use and gap areas in CDC's Colorectal Cancer Control Program. Prev Chronic Dis. 2012; 9: E131. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeltner FJ, Ely GE, Whitler ET, et al.: Effectiveness of community health workers in providing outreach and education for colorectal cancer screening in Appalachian Kentucky. Soc Work Health Care. 2012; 51(5): 430–440. PubMed Abstract \| Publisher Full Text\n\nMoralez EA, Rao SP, Livaudais JC, et al.: Improving knowledge and screening for colorectal cancer among hispanics: Overcoming barriers through a PROMOTORA-led home-based educational intervention. J Cancer Educ. 2012; 27(3): 533–539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSarfaty M, Wender R, Smith R: Promoting cancer screening within the patient centered medical home. CA Cancer J Clin. 2011; 61(6): 397–408. PubMed Abstract \| Publisher Full Text\n\nRabinowitz HK, Diamond JJ, Markham FW, et al.: Increasing the supply of rural family physicians: Recent outcomes from Jefferson Medical College's Physician Shortage Area Program (PSAP). Acad Med. 2011; 86(2): 264–269. PubMed Abstract \| Publisher Full Text\n\nBodenheimer T, Pham HH: Primary care: current problems and proposed solutions. Health Aff (Millwood). 2010; 29(5): 799–805. PubMed Abstract \| Publisher Full Text\n\nMacKinney AC, Mueller KJ, McBride TD: The march to accountable care organizations-how will rural fare? J Rural Health. 2011; 27(1): 131–137. PubMed Abstract \| Publisher Full Text\n\nMcClellan M, McKethan AN, Lewis JL, et al.: A national strategy to put accountable care into practice. Health Aff (Millwood). 2010; 29(5): 982–990. PubMed Abstract \| Publisher Full Text\n\nGelberg L, Andersen RM, Leake BD: The Behavioral Model for Vulnerable Populations: application to medical care use and outcomes for homeless people. Health Serv Res. 2000; 34(6): 1273–1302. PubMed Abstract \| Free Full Text\n\nArbuckle JL: IBM® SPSS® amos™ 22 User’ s guide. Chicago, IL: IBM. 2013. Reference Source\n\nFraser D: QSR NUD* IST vivo: Reference guide. Qualitative Solutions and Research; 2000. Reference Source" }	[ { "id": "9594", "date": "04 Aug 2015", "name": "Young Dae Kwon", "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 research proposal has dealt with interesting topic, and expects to conduct a comprehensive and systematic study that is able to understand the factors associated with CRC screening of rural area residents based on this. Especially, I think it is great meaningful in that it will carry out with quantitative and qualitative approach together. The overall configuration is well described, but there are minor problems to be improved.The authors described that \"Existing studies have shown inconsistent findings and fail to address how contextual and provider-level factors impact CRC\" in the Background of the Abstract. However, Background in the main manuscript could not clearly show and organize that there is no consistency in the existing research results in detail. In Background, It would be better to explain how considering the patient, provider and county level factor together could improve CRC screening rate or why the comprehensive consideration is so important in CRC screening rate improvement. \"No. Clinic in County’ in Figure1 seems better to be modified as ‘No. of Clinics in County’. In Figure 1, the education variable is included only on the provider factor, but I think it is also an important variable as the patient factor. There are only two quantitative variables in socio-economic indicator of county level: 'poverty rate' and '% uninsured'. Are these two variables sufficiently able to reflect unique characteristics of the rural area? It seems to require a detailed explanation for the abbreviations used in the Figure 2. I think the biggest advantage and characteristic of this study is to perform quantitative approach and qualitative approach together. It would be better to describe in detail how you will conduct integration and linking of two research methods. Instead of just a parallel listing, specific plans for organic connection and interpretation of the results from two methods should be presented.", "responses": [] }, { "id": "9595", "date": "04 Aug 2015", "name": "Paul Terry", "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\nIt would have been interesting to learn why the authors believed the ACO patient population might be unique in terms of their barriers to CRC screening. Given the many other studies cited, were there any hypotheses driving the study other than this ACO is a relatively new healthcare organization? The first mention of “sensitive analysis” under “Study Setting” is unclear. Effect sizes and variability will not be generalizable to other populations, will they? It seems that the authors may have implied this, unless they are planning to address the same issues in a future larger study in rural Nebraska.\n\nIf this latter is the case, then this is a pilot study? Given the size and statistical power, I would think this is not a pilot study. Hence, the scope of the present study was not totally clear. It is unlikely that this study alone will effectively address the mixed results from previous studies. At least the authors might try to justify this further and suggest other necessary components of reconciling the disparate findings. It is not entirely clear how the data will be used to increase CRC screening. What has been the experience of previous investigators who have attempted to use data about “barriers to CRC screening” to increase screening in their populations?", "responses": [] } ]	1	https://f1000research.com/articles/4-298
https://f1000research.com/articles/4-297/v1	22 Jul 15	{ "type": "Review", "title": "Voltage-gated sodium channel as a target for metastatic risk reduction with re-purposed drugs", "authors": [ "Tomas Koltai" ], "abstract": "Objective: To determine the exact role of sodium channel proteins in migration, invasion and metastasis and understand the possible anti-invasion and anti-metastatic activity of repurposed drugs with voltage gated sodium channel blocking properties.Material and methods: A review of the published medical literature was performed searching for pharmaceuticals used in daily practice, with inhibitory activity on voltage gated sodium channels. For every drug found, the literature was reviewed in order to define if it may act against cancer cells as an anti-invasion and anti-metastatic agent and if it was tested with this purpose in the experimental and clinical settings.Results: The following pharmaceuticals that fulfill the above mentioned effects, were found: phenytoin, carbamazepine, valproate, lamotrigine, ranolazine, resveratrol, ropivacaine, lidocaine, mexiletine, flunarizine, and riluzole. Each of them are independently described and analyzed.Conclusions: The above mentioned pharmaceuticals have shown anti-metastatic and anti-invasion activity and many of them deserve to be tested in well-planned clinical trials as adjunct therapies for solid tumors and as anti-metastatic agents. Antiepileptic drugs like phenytoin, carbamazepine and valproate and the vasodilator flunarizine emerged as particularly useful for anti-metastatic purposes.", "keywords": [ "Voltage-gated sodium channels", "cancer", "phenytoin", "flunarizine", "repurposed drugs", "metastasis" ], "content": "Introduction\n\nThe capacity to metastasize is one of the hallmarks of cancer1 and usually death due to cancer is not caused by the primary tumor but rather by the metastatic spread2. The lack of an effective therapy in prevention of metastasis results in a high mortality rate in oncology. So it seems reasonable that if the risk of metastasis can be reduced, the outlook of cancer patients may significantly improve survival and quality of life. Solving the metastasis problem is solving the cancer problem3.\n\nMany natural products, like genistein4, resveratrol5 and curcumin6,7 have shown interesting anti-metastasis activity. The same effect has been observed with older pharmaceuticals like aspirin8, not-as-old pharmaceuticals such as celecoxib6,9; new pharmaceuticals like ticagrelor10, as well as with more sophisticated molecules like dasatinib and ponatinib6 or ultrasophisticated drugs, like polymeric plerixafor11.\n\nMany other compounds have also been identified as possessing anti-metastatic effects, including increases in NO12, cimetidine, doxycycline, heparin and low molecular heparins, and metapristone13.\n\nHigh creativity has been employed in the search for anti-metastatic compounds. For example, Ardiani et al. developed a vaccine-based immunotherapy to enhance CD4 and CD8 T lymphocyte activity against Twist14. Twist is a transcription factor involved in invasion and metastasis.\n\nMany known pharmaceuticals that are, or were, in use for other purposes than cancer treatment are demonstrating anti-metastatic activity. This is the case for thiobendazole, which is an antifungal, anti-parasitic drug that has been used in medical practice for over 40 years, but which also shows anti-migratory and apoptosis-inducing activity15. The introduction of Food and Drug Administration (FDA)-approved products which are used for a purpose different for which it was originally approved is called repurposing of a drug.\n\nMany new drugs are being introduced in the area of anti-metastatic activity. One such example, zoledronic acid16 is a biphosphonate that decreases bone metastasis. Denosumab17 is another example. It is a monoclonal antibody directed against the receptor activator of nuclear factor kappa B ligand (RANKL) that diminishes the number of circulating cancer cells and prevents bone metastasis. It is in Phase II clinical trials and has the advantage of subcutaneous administration, while zoledronic acid requires intravenous route (for further information on these compounds, see clinical trials NCT01952054, NCT01951586, NCT02129699)18.\n\nMetastasis is a multi-step development. The different steps in the metastatic cascade can be targeted with a combination of drugs against each step. Migration and invasion are necessary steps for the metastatic cascade. There is no metastasis without prior migration of malignant cells, so that if migration and invasion are blocked, metastasis should not occur.\n\nInvasion is the first step in metastasis, and in a very simplified view, it can be divided into three stages (Shown schematically in Figure 1):\n\n(uPA: urinary plasminogen activator).\n\n1. Translocation of cells across extracellular matrix barriers\n\n2. Degradation of matrix proteins by specific proteases\n\n3. Cell migration\n\n\nVoltage-gated sodium channels\n\nNeurons and muscle cells (and excitable tissues in general) express voltage-gated sodium channel (VGSC) proteins; tumor cells may also express these proteins. VGSCs are important players in migration and invasion as it will be described in this manuscript.\n\nSodium channels were first described by Hodgkin and Huxley in 1952 and knowledge about structure and physiology of VGSCs are mainly the result of seminal investigations developed by William Catterall19.\n\nSodium channels are glycosylated transmembrane proteins that form passages in the cell membrane for the penetration of sodium into the intracellular space according to their electrical gradients. Voltage-gated sodium channels (also known as VGSCs or ‘NaV’ channels) refers to the mechanism that triggers these proteins to allow sodium movement across the membrane.\n\nThere are nine known VGSCs (NaV1.1 to Nav1.9) that are members of the superfamily of VGCSs. NaV1.1, 1.2, 1.3 and 1.6 are found in the central nervous system. NaV1.4 is found in muscle and NaV1.5 in cardiac muscle20.\n\nVGSC is formed by a large subunit (α) and other smaller subunits (β). The α subunit is the core of the channel and is fully functional by itself, even without the presence of β subunits19–21.\n\nWhen a cell expresses VGSC α subunits, this means that it is capable of conducting sodium into the cell. The structure of VGSC can be seen in Figure 2 and Figure 3. VGSCs modulate the exchange of Na+ across the cell membrane and the inflow of this electrolyte spikes the action potential in excitable tissues22.\n\nIt is well known that expression of VGSCs appears in cancer cells where it is not expressed in their normal counterparts, and plays a significant role in disease progression. Table 1 shows examples of the cancer tissues in which dysregulated expression of VGSCs were identified and the role they play.\n\nTargeting these channels may represent a legitimate way of reducing or blocking the metastatic process.\n\nThe role of sodium channel in invasion, metastasis and carcinogenesis is insufficiently known.\n\n\nSodium channel proteins and cancer\n\nIn 1995, Grimes et al.33 investigated the differential electrophysiological characteristics of VGSCs in two different rodent prostate cancer cell lines: the Mat-Ly-Lu cell line, which is a highly metastatic line (more than 90% of metastasis to lung and lymph nodes under experimental conditions) and the AT-2 cell line with a much lower metastatic potential (less than 10% chance of developing metastasis in experimental conditions). They found fundamental differences in electrophysiological features between these two cell lines which displayed a direct relationship with in vitro invasiveness. Sodium inward currents were detected only in the Mat-Ly-Lu cell line and inhibition of VGSC protein with Tetrodotoxin (TTX; a powerful inhibitor of VGSCs) significantly reduced the capacity for invasion (mean reduction 33%). On the other hand, TTX showed no effect on invasion of AT-2 cell lines. The TTX-induced reduction of invasion showed a direct correlation with the amount of cells expressing VGSC in the culture.\n\nNo fundamental differences in the potassium channels were found between the two cell lines, except for a lower density of potassium channels in the Mat-Ly-Lu cell line. The authors concluded that ion channels may be involved in malignant cell behavior and that VGSCs could play a role in the metastatic process.\n\nIn 1997, Laniado et al.34 investigated the presence of VGSC in human prostate cell lines. As in the Grimes research they used two different cell lines: one with a low metastatic potential: the LN-Cap cell line which is androgen dependent and expresses prostate-specific antigen, and the PC-3 line which is more malignant, does not express prostate-specific antigen and exhibits a high rate of metastatic potential.\n\nAs in the work by Grimes et al., they found that PC-3, the more malignant cell line, expressed VGSC protein and that inhibition of this channel protein with TTX reduced invasion in a significant way. LN-Cap cells did not express VGSC.\n\nOne of the conclusions reached by the authors was that cancer cells expressing functional VGSC had a selective advantage regarding migration and distant metastasis. In the case of both humans and rodents, not all cells in the highly malignant cell cultures showed the presence of the VGSC protein. For example, in PC-3 cell culture only 10% of cells expressed a functional VGSC protein. This is the reason why the authors consider these cells as a clonal evolution that gives pro-tumor and pro-invasive advantages.\n\nThe correlation between VGSC protein expression and invasiveness in human and rat prostate cancer cells was confirmed by Smith et al.35 by comparing seven lines of rat prostate carcinoma cells with different metastatic ability, and nine human prostate carcinoma cell lines. In general, invading capacity of the basement membrane and metastatic ability showed a positive correlation with the percentage of cells expressing VGSC. But this positive correlation between percentage of cells expressing VGSCs and the percentage of cells being invasive occurred only up to 27% of the cells being invasive in the rat series and up to 12% of cells being invasive in the human series. Authors suggested that these discrepancies may be due to the necessity of other factors for invasive capability besides VGSC presence; i.e. this protein may represent a prerequisite for the invasive phenotype but other requirements must also be achieved for a full-blown invasive phenotype. Fraser et al.36 determined the key role played by VGSCs in prostate cancer cells in invasion and motility and showed that TTX and phenytoin (PHEN) that are known VGSC blockers, decreased motility and invasiveness while channel openers increased motility. However, the increased invasion capacity in VGSC-expressing cancer cells is not limited to prostate cancers. The same features were found in breast cancer cell lines MCF-7 (estrogen receptor positive), MDA-MB-231 and MDA-MB-468 (both estrogen receptor negative).\n\nBaciotglu et al.38 when experimenting on a rat model of induced breast cancer showed the importance of inhibiting VGSCs in order to inhibit antioxidant response. They observed a survival improvement in rats treated with a VGSC blocker.\n\nAn important location of VGSCs in cancer cells is in a cellular region directly involved in migration and invasion: the invadopodia. Invadopodias are protrusions of the plasma membrane, rich in actin that are strongly related to degradation of the extracellular matrix (ECM). Figure 4 and Figure 5 summarize how invadopodia works and the relation between VGSC and invadopodia.\n\n(Brisson 201339; Gillet 200940). Acidification activates cathepsine degradation of the extracellular matrix.\n\nThe second mechanism of action is through activation of Src which increases MMP-2 and MMP-9 secretion and activity through phosphorylation of Cortactin. It is postulated that there is a feedback loop starting with MMPs products which induces the development of new invadopodia (Red circle around I). (This figure has been constructed based on references Gillet 200940, Brisson 2013133, Clark 200742 and Mader 201141).\n\nAccording to Brisson et al.39, NaV 1.5 Na+ channels regulate the NHE-1 exchanger protein that increases proton extrusion with extracellular matrix acidification that promotes invasion and migration through activity of cystein cathepsines and degradation of extracellular matrix40.\n\nA second mechanism of invasion promotion was described by Mader et al.41 through the EGFR-Scr-cortactin pathway. Src is activated by VGSCs and Src phosphorylates cortactin. Cortactin is involved in MMP-9 and MMP-2 upregulation and secretion as can be seen in Figure 5. These events lead to matrix degradation, an integral step in cancer cell invasion42.\n\nThere are nine different VGSC α subunits and four different β subunits. The expression of these subunits may vary in the different tumor cells21. For example, NaV 1.5 is overexpressed in astrocytoma, breast and colon cancer. NaV 1.7 is found in breast, prostate and non small-cell lung cancer (NSCLC) and NaV 1.6 in cervical and prostate cancer. This suggests that the α subunits seem to be tissue specific.\n\nThe main players in the invadopodia complex, besides the VGSCs are Src kinase, cortactin and Rho-A GTPase. The exact relation between these players is not fully known and needs further research. (For further reading on invadopodia and cortactin, see references 43,44).\n\nOne possible relation between invadopodia-Src-VGSCs is described in Figure 6.\n\nBerdeaux et al.47 reported that the small molecule GTPase Rho A activity is under control of oncogenic Src and localizes in the invadopodia complex and Durlong et al.48 (2013) showed that Rho-A regulates the expression and activity of NaV1.5 and found a positive feedback between NaV1.5 and Rho A in breast cancer cells. According to Timpson et al.49, cooperation between mutant p53 and oncogenic Ras activates Rho-A.\n\nOnganer and Djamgoz45 proposed the hypothesis that VGSC upregulation enhances the metastatic phenotype by enhancing endocytic membrane activity in SCLC.\n\nAndrikopoulus et al.50 have demonstrated that VGSCs have pro-angiogenic functions by significantly increasing vascular endothelial growth factor (VEGF) signaling in endothelial cells. Endothelial cells express NaV1.5 and NaV1.7. TTX blocks, and NaV1.5 RNAi decreases endothelial cell proliferation and tubular differentiation that are essential steps in the angiogenesis process.\n\nThe important implications of VGSCs in cancer progression and invasion led Litan and Langhans51 to express that cancer is a channelopathy (For further reading on structure and functions of VGSC see reference 52).\n\n\nMaterial and methods\n\nA search was performed in the medical literature to find pharmaceuticals already in use for other purposes than cancer, that as an off-target effect could inhibit VSGCs and to determine if these pharmaceuticals can actually decrease migration, invasion and metastatic potential of cancer. A Pubmed advanced search retrieved 50519 articles under the search condition “voltage-gated sodium channel blocker” during the period of 1981–2015.\n\nThe articles that considered drugs that were not in clinical use or FDA-approved were not included in this study, with the exception of resveratrol and natural polyphenols.\n\nThe following drugs fulfilling these criteria were found: phenytoin, carbamazepine, lamotrigine valproate, ranolazine, resveratrol, ropivacaine, lidocaine, mexiletine, flunarizine, and riluzole.\n\nA new search was performed in Pubmed for each of the above listed pharmaceuticals with two search criteria: 1) the drug and 2) the term cancer. The period considered was from 1962 to the current time.\n\nThose VGSC blocking drugs that exhibited anti-cancer activity based mainly by other mechanisms are only briefly mentioned; valproic acid and lamotrigine probably act against cancer by histone deacetylase inhibition and riluzole’s anti-cancer mechanism is probably related to the glutamatergic pathway.\n\nTetrodotoxin is also analyzed in spite of the fact that it is not in clinical use, because it is the traditional model molecule of VGSC blocking, against which other drugs are comparatively tested in the experimental setting.\n\n\nResults\n\nMany biological toxins like those found in scorpions and sea anemones develop their toxicity by introducing modifications to the properties of VGSCs52. This toxicity can be achieved by inactivation of VGSCs (as in the case of TTX) or on the contrary by persistent activation of the channel (in the case of veratridine, acinitine and many others).\n\nTTX is a powerful biological neurotoxin found in fishes of the Tetraodontiformes order and certain symbiotic bacteria. TTX binds to VGSCs and blocks its activity, mainly in the nervous system. It is used as a biotoxin for defensive or predatory purposes. TTX binds to the extracellular portion of VGSC, disabling the function of the ion channel and results in a very poisonous effect producing death through respiratory paralysis22.\n\nDue to its high toxicity it is not used as a therapeutic agent, but TTX has been very useful in the experimental setting for the study of VGSCs physiology.\n\nPHEN is an anticonvulsant that has been identified as a sodium channel blocker53,54 which has been held responsible for inducing lymphoma, pseudolymphoma, hematological malignancies and other cancers in patients under chronic treatment55. This carcinogenic effect of phenytoin was not confirmed in large epidemiological studies56. PHEN diminishes cell mediated immunity57.\n\nVernillo et al. in 199058 found that phenytoin inhibited bone resorption in rat osteosarcoma cells through significant reduction of collagenase and gelatinase activities. But Dyce et al.59 did not find evidence of PHEN’s gelatinase inhibitory activity in B16 melanoma cells in vitro. This may be evidence of tissue-specific activity which has not been investigated any further. Dyce et al. did not find important anti-metastatic activity either in a melanoma tail injection model in mice. But when the data of this publication is examined in detail, it seems that the anti-metastatic activity is not so low as mentioned by the authors: they found that after injection of tumor cells in the mice protected with PHEN, the animal developed mean pulmonary colonies 4.6 +/- 3.1 but when the mice received no PHEN, developed. 10.2 +/- 9.9 colonies. Beyond any statistical analysis the difference seems important.\n\nYang et al.60 found that NaV 1.5 was over-expressed in breast cancer cells with high metastatic potential, and the anticonvulsivant PHEN had the ability to reduce migration and invasion at clinically achievable concentrations in MDA-MB-231 cells (which are strongly metastatic) and showed no effects on MCF-7 cells with low metastatic potential.\n\nPHEN blocks Na+ channels and has a high affinity for VGSCs in the inactivated state of the channel61. Compared with verapamil, lidocaine and carbamacepine, PHEN had an intermediate potency between verapamil and lidocaine, being verapamil the strongest inhibitor and carbamazepine the weakest.\n\nAbdul et al.62 studied the effect of four anticonvulsants (PHEN, carbamazepine, valproate and ethosuxinide) on the secretion of prostate specific antigen and interleukin-6 in different human prostate cancer cell lines. PHEN and carbamazepine inhibited the secretion of both.\n\nFadiel et al.63 found that PHEN is a strong estrogen receptor α antagonist at clinically achievable concentrations and at the same time is a weak agonist.\n\nThere is an undesired side effect of PHEN that may represent a drawback for its use in cancer: immunological depression73–76. This is an issue that deserves further research. Finally it has to be mentioned that PHEN interacts with many other pharmaceuticals, particularly those usually employed in chemotherapy.\n\nIn summary, the main activities developed by PHEN in relation with cancer are:\n\nVGSC blocking, microtubule polymerization blocking, immunosuppression, calcium channel blocking and enhancement of vinblastine cytotoxicity.\n\nCarbamazepine is a sodium channel blocker, pro-autophagy agent and histone deacetylase inhibitor that has been in use since 1962 for the treatment of seizures, neuropathic pain and bipolar disorders and has shown interesting anti-metastatic potential in the experimental setting77. Studies have also insinuated preventative effects in prostate cancer78.\n\nCarbamazepine induces Her2 protein degradation through the proteosome without modifying its production79. This activity seems to be related to histone deacetylase inhibition rather than VGSC blocking. Growth inhibition in estrogen-receptor positive breast cancer cell lines seems probably a histone deacetylase inhibitor effect80.\n\nOxcarbazepine, a molecule related to carbamazepine is also a sodium channel blocker81 and a potassium channel blocker, but it has not been investigated for cancer.\n\nThe anti-cancer mechanisms shown by carbamazepine are in summary:\n\n1) VGSC blocker as anti-metastatic77.\n\n2) Histone deacetylase inhibition82.\n\n3) Her2 degradation by proteasome79.\n\nAn anticonvulsivant drug that exerts multiple actions related to anti-cancer effects: calcium channel blocker, VGSC blocker, inhibition of histone deacetylase, potentiation of inhibitory activity of GABA, decreases angiogenesis, interferes with MAP kinase pathways and the β catenin-Wnt pathway83. Val is being tested in various clinical trials in leukemias and solid tumors84. Most of the anti-tumor activities of VAL seem to be related to the inhibition of histone deacetylase rather than VGSC blocking and further discussion goes beyond the scope of this review.\n\nRanolazine, (Ranexa) is an antiarrhythmic drug indicated for the treatment of chronic angina that was first approved by FDA in 2006. Common side effects are dizziness, constipation, headache and nausea85.\n\nRanolazine inhibits the late inward sodium current in heart muscle, so that it works as a sodium channel inhibitor. Ranolazine is metabolized by the CYP3A enzyme.\n\nDriffort et al.86 demonstrated that ranolazine inhibition of NaV1.5 reduced breast cancer cells invasiveness in vivo and in vitro using the highly invasive MDA-MB-231 breast cancer cell line. This drug also efficiently decreased the activity of the embryonic/neonatal isoform of NaV1.5 (the active isoform usually found in human breast cancer cells). Ranolazine did not change the viability of the cell. It also decreased the pro-invasive morphology of MDA-MB-231 breast cancer cells. They also demonstrated that injection of cancer cells through the tail vein of nude mice at non-toxic doses achieved a significant reduction in metastatic colonization.\n\nCertain biologically active natural phenols like resveratrol and genistein have shown effects on VGSCs, increasing hyperpolarized potentials during steady state inactivation88.\n\nResveratrol’s inhibitory effects on VGSC has consequences for the behaviour of metastatic cells. Fraser et al.89 showed that resveratrol significantly decreased lateral and transversal motility and invasion capacity of rat prostate cancer cells (MAT-Ly-Lu cells), without changes in cellular viability. They also found that resveratrol inhibited VGSC in a dose-dependent manner and using TTX with resveratrol did not increase VGSC inhibition nor metastatic cell behavior. Resveratrol also inhibits epithelial sodium channels90.\n\nResveratrol is not the only polyphenol with VGSC blocking activity: quercetin and catechin showed similar effects91 in ventricular myocytes and genistein in rat cervical ganglia92 and nociceptive neurons93.\n\nGabapentin is used for the treatment of pain and decreases expression of NaV1.7 and ERK-1/ERK-2 in ganglion neurons94,95 and expression of NaV1.296. We found no publications about gabapentin as a possible anti-metastasis or anti-invasion treatment.\n\nRiluzole is a drug used for amyotrophic lateral sclerosis and it is a known sodium channel blocker. This effect was demonstrated in human prostate cancer cell lines97. Most likely, the anticancer activity of riluzole is mainly related to other anticancer characteristics of this drug like downregulation of the glutamatergic pathway98.\n\nFlunarizine is a calcium channel blocker with a long plasma half-life, used in migraine prevention, vertigo and adjuvant treatment of epilepsy, but has shown important activity as a VGSC blocker99–101. It binds calmodulin.\n\nAt low temperatures (22 degrees) flunarizine potentiate the binding of phenytoin to VGSC102.\n\nFlunarizine has shown anti-cancer activities in lymphoma and multiple myeloma103, and leukaemia104, but these anti-cancer activities were apparently related to induction of apoptosis, which is not a consequence of VGSC blockage. On melanoma cells, flunarizine showed decreased motility and invasion in vitro105,106. Flunarizine inhibited migration and phagocytosis in B16 melanoma cells and M5076 macrophage-like cancer cells107.\n\nAccording to data found in medical literature we may consider anti-cancer activities of flunarizine in the following way:\n\na) Activities dependent on VGSC blocking: decreased motility and invasion105–107.\n\nb) Activities dependent on calcium channel blocking: vasodilatation and increased concentration of chemotherapeutic drugs in tumor tissues108–110, and increased radiosensitivity due to better oxygen delivery to anoxic areas of the tumor111.\n\nc) WNT inhibition103.\n\nd) Inhibition of lymphangiogenesis112.\n\ne) Increase of melphalan`s citotoxicity in resistant ovarian cancer cells113 and in rhabdomyosarcoma114.\n\nf) Positive modulation of doxorubicin in multidrug resistant phenotype colon adenocarcinoma cells115.\n\ng) Decreased blood viscosity improving oxygen delivery to the tumor116.\n\nh) Other anti-tumor activities: apoptosis and growth rate inhibition103,104,117.\n\nFlunarizine has not been tested in cancer trials. The fact that it can significantly reduce motility in melanoma cells which is a highly metastasizing tumor and is also an inhibitor of lymphangiogenesis, makes it an interesting adjuvant therapy that deserves further research. The possible synergy with phenytoin is also an issue that should be explored.\n\nHowever, flunarizine has also shown cytoprotective effects in certain tissues (auditory cells) against cisplatin118 and flunarizine may induce Nrf-2 overexpression that confers resistance to chemotherapy in some tumors like Her2 positive breast cancer119.\n\nLocal anaesthetics eliminate pain through VGSC blocking on nociceptive neurones.\n\nLocal anaesthetics like lidocaine have shown interesting anti-cancer effects in various cancer cells. Lidocaine is a VGSC blocker. The mechanisms involved in decreased proliferation seems related to the inhibitory actions of local anaesthetics on EGFR120 rather than VGSC blocking. Inhibition of invasion found in cancer cells treated with lidocaine (HT1080, HOS, and RPMI-7951) by Mammoto et al. was attributed by the authors to shedding of the extracellular domain of heparin binding epidermal growth factor-like growth factor and not to VGSC blocking121.\n\nBaptista-Hon et al. described a decrease in metastatic potential of colon cancer cells (SW620 cells) by ropivacaine and decrease of Nav1.5 function (adult and neonatal isoforms)122.\n\nPiegeler et al., 2012 identified decreased Src activity produced by amide-linked anaesthetics as an independent mechanism of migration and invasion decrease123.\n\nOther drugs that have shown significant VGSC blocking activity and may have activity in the fight against migration, invasion and metastasis are: fluoxetin blocks NaV1.5124, and mexiletine125.\n\nIntravenous propofol has been recognized as an anti-invasion drug in HeLa, HT1080, HOS and RPMI-75 cells by decreasing actin stress-fiber formation and focal adhesion inhibition, but this drug is not a VGSC blocker and the probable mechanism is through Rho-A modulation126.\n\nAll of the drugs we have mentioned are low cost pharmaceuticals, have predictable and well known side effects, and therefore they are adequate candidates for further clinical trials.\n\n\nDiscussion\n\nFunctionally active VGSCs are expressed in many metastatic cancer cells. This functional expression is an integral element of the metastatic process in many different solid tumors.\n\nThe essential role of this protein in invadopodia has been established, so VGSCs became a legitimate target to decrease migration, invasion and metastasis. Repurposed drugs like anticonvulsants, (phenytoin in particular) have shown interesting anti-invasion effects.\n\nCarbamazepine’s ability to induce Her2 protein degradation should be considered an interesting association to trastuzumab.\n\nTargeting VGSCs may act in synergy with anti-angiogenic treatments and with other chemotherapeutic drugs like vinblastine.\n\nOn the other hand in the review by Besson et al.127 there is a very important remark: VGSC is also present in macrophages and cells related with the immunologic system, so that disrupting VGSC`s activity may deteriorate also anti-tumor immunologic mechanisms.\n\nFlunarizine represents a particularly interesting molecule because it may attack cancer from four different angles: invasion and migration through VGSC blocking, WNT pathway down-regulation, decreased lymphangiogenesis and better oxygenation of hypoxic areas which permits a better arrival of chemotherapeutic drugs and increased sensitivity to radiation. It has never been tested in clinical trials for cancer treatment.\n\n\nFuture directions\n\nNew VGSCs blockers are under research. Sikes et al.67 developed new blockers based on the phenytoin binding site to VGSC. They found compounds with enhanced activity in VGSC blocking and antitumor activity against human prostate cancer cells.\n\nThe association of two or more VGSC blockers may show synergistic enhanced anti-metastatic activity. Nerve growth factor (NGF) increases the number of VGSCs129; tanezumab, a new NGF inhibitor diminishes the amount of VGSCs128, so we may assume that tanezumab may develop synergistic activity with VGSC blockers. Tanezumab has not been tested in cancer and we think it deserves more research because NGF is also an anti-apoptotic protein130.\n\nStettner et al.134 found that men over 50 years of age may be benefited with the use of anticonvulsivants, regarding prostate cancer prevention because they observed lower PSA levels compared with control groups. They also described a ranking of this preventive activity: valproic acid>levetiracetam>carbamazepine/oxcarbazepine>lamotrigine. The authors also observed synergy between these drugs.\n\n\nConclusions\n\nRepurposed VGSC blocker drugs, particularly phenytoin, flunarizine and polyphenols, deserve clinical trials as complementary treatment to decrease the metastatic risk.", "appendix": "Competing interests\n\n\n\nThe authors declared no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no funds were involved in supporting this work.\n\n\nReferences\n\nHanahan D, Weinberg RA: Hallmarks of cancer: the next generation. Cell. 2011; 144(5): 646–74. PubMed Abstract \| Publisher Full Text\n\nMehlen P, Puisieux A: Metastasis: a question of life or death. Nat Rev Cancer. 2006; 6(6): 449–458. PubMed Abstract \| Publisher Full Text\n\nTaketo MM: Reflections on the spread of metastasis to cancer prevention. Cancer Prev Res (Phila). 2011; 4(3): 324–328. PubMed Abstract \| Publisher Full Text\n\nPavese JM, Krishna SN, Bergan RC: Genistein inhibits human prostate cancer cell detachment, invasion, and metastasis. Am J Clin Nutr. 2014; 100(Suppl 1): 431S–436S. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang X, Li X, Ren J: From French Paradox to cancer treatment: anti-cancer activities and mechanisms of resveratrol. Anticancer Agents Med Chem. 2014; 14(6): 806–25. PubMed Abstract \| Publisher Full Text\n\nXiong A, Yang Z, Shen Y, et al.: Transcription Factor STAT3 as a Novel Molecular Target for Cancer Prevention. Cancers (Basel). 2014; 6(2): 926–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKronski E, Fiori ME, Barbieri O, et al.: miR181b is induced by the chemopreventive polyphenol curcumin and inhibits breast cancer metastasis via down-regulation of the inflammatory cytokines CXCL1 and -2. Mol Oncol. 2014; 8(3): 581–95. PubMed Abstract \| Publisher Full Text\n\nLangley RE, Rothwell PM: Aspirin in gastrointestinal oncology: new data on an old friend. Curr Opin Oncol. 2014; 26(4): 441–7. PubMed Abstract \| Publisher Full Text\n\nLi WW, Long GX, Liu DB, et al.: Cyclooxygenase-2 inhibitor celecoxib suppresses invasion and migration of nasopharyngeal carcinoma cell lines through a decrease in matrix metalloproteinase-2 and -9 activity. Pharmazie. 2014; 69(2): 132–7. PubMed Abstract \| Publisher Full Text\n\nGebremeskel S, LeVatte T, Liwski RS, et al.: The reversible P2Y12 inhibitor ticagrelor inhibits metastasis and improves survival in mouse models of cancer. Int J Cancer. 2015; 136(1): 234–40. PubMed Abstract \| Publisher Full Text\n\nLi J, Oupický D: Effect of biodegradability on CXCR4 antagonism, transfection efficacy and antimetastatic activity of polymeric Plerixafor. Biomaterials. 2014; 35(21): 5572–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu Y, Yu T, Liang H, et al.: Nitric oxide inhibits hetero-adhesion of cancer cells to endothelial cells: restraining circulating tumor cells from initiating metastatic cascade. Sci Rep. 2014; 4: 4344. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang J, Chen J, Wan L, et al.: Synthesis, spectral characterization, and in vitro cellular activities of metapristone, a potential cancer metastatic chemopreventive agent derived from mifepristone (RU486). AAPS J. 2014; 16(2): 289–98. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArdiani A, Gameiro SR, Palena C, et al.: Vaccine-mediated immunotherapy directed against a transcription factor driving the metastatic process. Cancer Res. 2014; 74(7): 1945–57. PubMed Abstract \| Publisher Full Text\n\nZhang J, Zhao C, Gao Y, et al.: Thiabendazole, a well-known antifungal drug, exhibits anti-metastatic melanoma B16F10 activity via inhibiting VEGF expression and inducing apoptosis. Pharmazie. 2013; 68(12): 962–8. PubMed Abstract\n\nRietkötter E, Menck K, Bleckmann A, et al.: Zoledronic acid inhibits macrophage/microglia-assisted breast cancer cell invasion. Oncotarget. 2013; 4(9): 1449–60. PubMed Abstract \| Free Full Text\n\nRolfo C, Raez LE, Russo A, et al.: Molecular target therapy for bone metastasis: starting a new era with denosumab, a RANKL inhibitor. Expert Opin Biol Ther. 2014; 14(1): 15–26. PubMed Abstract \| Publisher Full Text\n\nNCT01952054, NCT01951586, NCT02129699. Reference Source\n\nCatterall WA: Voltage-gated sodium channels at 60: structure, function and pathophysiology. J Physiol. 2012; 590(Pt 11): 2577–2589. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTermin A, Martinborough E, Wilson D: Chapter 3 – Recent Advances in Voltage-Gated Sodium Channel Blockers: Therapeutic Potential as Drug Targets in the CNS. Annu Rep Med Chem. 2008; 43: 43–60. Publisher Full Text\n\nBrackenbury WJ: Voltage-gated sodium channels and metastatic disease. Channels (Austin). 2012; 6(5): 352–361. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee CH, Ruben PC: Interaction between voltage-gated sodium channels and the neurotoxin, tetrodotoxin. Channels (Austin). 2008; 2(6): 407–412. PubMed Abstract \| Publisher Full Text\n\nFraser SP, Diss JK, Chioni AM, et al.: Voltage-gated sodium channel expression and potentiation of human breast cancer metastasis. Clin Cancer Res. 2005; 11(15): 5381–9. PubMed Abstract \| Publisher Full Text\n\nRoger S, Rollin J, Barascu A: Voltage-gated sodium channels potentiate the invasive capacities of human non-small-cell lung cancer cell lines. Int J Biochem Cell Biol. 2007; 39(4): 774–786. PubMed Abstract \| Publisher Full Text\n\nBlandino JK, Viglione MP, Bradley WA, et al.: Voltage-dependent sodium channels in human small-cell lung cancer cells: role in action potentials and inhibition by Lambert-Eaton syndrome IgG. J Membr Biol. 1995; 143(2): 153–163. PubMed Abstract \| Publisher Full Text\n\nDiaz D, Delgadillo DM, Hernández-Gallegos E, et al.: Functional expression of voltage-gated sodium channels in primary cultures of human cervical cancer. J Cell Physiol. 2007; 210(2): 469–478. PubMed Abstract \| Publisher Full Text\n\nBennett ES, Smith BA, Harper JM: Voltage-gated Na+ channels confer invasive properties on human prostate cancer cells. Pflugers Arch. 2004; 447(6): 908–914. PubMed Abstract \| Publisher Full Text\n\nGao R, Shen Y, Cai J, et al.: Expression of voltage-gated sodium channel alpha subunit in human ovarian cancer. Oncol Rep. 2010; 23(5): 1293–1299. PubMed Abstract \| Publisher Full Text\n\nHouse CD, Vaske CJ, Schwartz AM, et al.: Voltage-gated Na+ channel SCN5A is a key regulator of a gene transcriptional network that controls colon cancer invasion. Cancer Res. 2010; 70(17): 6957–67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFraser SP, Diss JK, Lloyd LJ, et al.: T-lymphocyte invasiveness: control by voltage-gated Na+ channel activity. FEBS Lett. 2004; 569(1-3): 191–194. PubMed Abstract \| Publisher Full Text\n\nSato K, Ishizuka J, Cooper CW, et al.: Inhibitory effect of calcium channel blockers on growth of pancreatic cancer cells. Pancreas. 1994; 9(2): 193–202. PubMed Abstract \| Publisher Full Text\n\nFulgenzi G, Graciotti L, Faronato M, et al.: Human neoplastic mesothelial cells express voltage-gated sodium channels involved in cell motility. Int J Biochem Cell Biol. 2006; 38(7): 1146–1159. PubMed Abstract \| Publisher Full Text\n\nGrimes JA, Fraser SP, Stephens GJ, et al.: Differential expression of voltage-activated Na+ currents in two prostatic tumour cell lines: contribution to invasiveness in vitro. FEBS Lett. 1995; 369(2-3): 290–294. PubMed Abstract \| Publisher Full Text\n\nLaniado ME, Lalani EN, Fraser SP: Expression and functional analysis of Voltage-activated Na+ channels in human prostate cancer cell lines and their contribution to invasion in vitro. Am J Pathol. 1997; 150(4): 1213–1221. PubMed Abstract \| Free Full Text\n\nSmith P, Rhodes NP, Shortland AP, et al.: Sodium channel protein expression enhances the invasiveness of rat and human prostate cancer cells. FEBS Lett. 1998; 423(1): 19–24. PubMed Abstract \| Publisher Full Text\n\nFraser SP, Salvador V, Manning EA, et al.: Contribution of functional voltage-gated Na+ channel expression to cell behaviors involved in the metastatic cascade in rat prostate cancer: I. Lateral motility. J Cell Physiol. 2003; 195(3): 479–87. PubMed Abstract \| Publisher Full Text\n\nRoger S, Besson P, Le Guennec JY: Involvement of a novel fast inward sodium current in the invasion capacity of a breast cancer cell line. Biochim Biophys Acta. 2003; 1616(2): 107–111. PubMed Abstract \| Publisher Full Text\n\nBatcioglu K, Uyumlu AB, Satilmis B, et al.: Oxidative stress in the in vivo DMBA rat model of breast cancer: suppression by a voltage-gated sodium channel inhibitor (RS100642). Basic Clin Pharmacol Toxicol. 2012; 111(2): 137–41. PubMed Abstract \| Publisher Full Text\n\nBrisson L, Driffort V, Benoist L, et al.: Nav1.5 Na+ channels allosterically regulate the NHE-1 exchanger and promote the activity of breast cancer cell invadopodia. J Cell Sci. 2013; 126(Pt 21): 4835–4842. PubMed Abstract \| Publisher Full Text\n\nGillet L, Roger S, Besson P, et al.: Voltage-gated Sodium Channel Activity Promotes Cysteine Cathepsin-dependent Invasiveness and Colony Growth of Human Cancer Cells. J Biol Chem. 2009; 284(13): 8680–8691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMader CC, Oser M, Magalhaes MA, et al.: An EGFR-Src-Arg-Cortactin pathway mediates functional maturation of invadopodia and breast cancer cell invasion. Cancer Res. 2011; 71(5): 1730–1741. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClark ES, Whigham AS, Yarbrough WG, et al.: Cortactin is an essential regulator of matrix metalloproteinase secretion and extracellular matrix degradation in invadopodia. Cancer Res. 2007; 67(9): 4227–35. PubMed Abstract \| Publisher Full Text\n\nMurphy DA, Courtneidge MA: The ‘ins’ and ‘outs’ of podosomes and invadopodia: characteristics, formation and function. Nat Rev Mol Cell Biol. 2011; 12(7): 413–426. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCosen-Binker LI, Kapus A: Cortactin: the gray eminence of the cytoskeleton. Physiology (Bethesda). 2006; 21: 352–361. PubMed Abstract \| Publisher Full Text\n\nOnganer PU, Djamgoz MB: Small-cell lung cancer (human): potentiation of endocytic membrane activity by voltage-gated Na+ channel expression in vitro. J Membr Biol. 2005; 204(2): 67–75. PubMed Abstract \| Publisher Full Text\n\nTarone G, Cirillo D, Giancotti FG, et al.: Rous sarcoma virus-transformed fibroblasts adhere primarily at discrete protrusions of the ventral membrane called podosomes. Exp Cell Res. 1985; 159(1): 141–57. PubMed Abstract \| Publisher Full Text\n\nBerdeaux RL, Diaz B, Kim L, et al.: Active Rho is localized to podosomes induced by oncogenic Src and is required for their assembly and function. J Cell Biol. 2004; 166(3): 317–323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDulong C, Fang YJ, Gest C, et al.: The small GTPase RhoA regulates the expression and function of the sodium channel Nav1.5 in breast cancer cells. Int J Oncol. 2014; 44(2): 539–547. PubMed Abstract \| Publisher Full Text\n\nTimpson P, McGhee EJ, Morton JP, et al.: Spatial regulation of RhoA activity during pancreatic cancer cell invasion driven by mutant p53. Cancer Res. 2011; 71(3): 747–757. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndrikopoulos P, Fraser SP, Patterson L, et al.: Angiogenic functions of voltage-gated Na+ Channels in human endothelial cells: modulation of vascular endothelial growth factor (VEGF) signaling. J Biol Chem. 2011; 286(19): 16846–16860. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLitan A, Langhans SA: Cancer as a channelopaty: ion channels and pumps in tumor development and progression. Front Cell Neurosci. 2015; 9: 86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDenac H, Mevissen M, Scholtysik G, et al.: Structure, function and pharmacology of voltage-gated sodium channels. Naunyn Schmiedebergs Arch Pharmacol. 2000; 362(6): 453–479. PubMed Abstract\n\nMatsuki N, Quandt FN, Ten Eick RE, et al.: Characterization of the block of sodium channels by phenytoin in mouse neuroblastoma cells. J Pharmacol Exp Ther. 1984; 228(2): 523–30. PubMed Abstract\n\nWillow M, Gonoi T, Catterall WA, et al.: Voltage clamp analysis of the inhibitory actions of diphenylhydantoin and carbamazepine on voltage-sensitive sodium channels in neuroblastoma cells. Mol Pharmacol. 1985; 27(5): 549–58. PubMed Abstract\n\nSchwinghammer TL, Howrie DL: Phenytoin-induced lymphadenopathy. Drug Intell Clin Pharm. 1983; 17(6): 460–2. PubMed Abstract\n\nOlsen JH, Boyce JD Jr, Jenssen JP, et al.: Cancer among epileptic patients exposed to anticonvulsant drugs. J Natl Cancer Inst. 1989; 81(10): 803–8. PubMed Abstract \| Publisher Full Text\n\nOkamoto Y, Shimizu K, Tamura K, et al.: Effects of phenytoin on cell-mediated immunity. Cancer Immunol Immunother. 1988; 26(2): 176–9. PubMed Abstract \| Publisher Full Text\n\nVernillo AT, Ramamurthy NS, Lee HM, et al.: The effect of phenytoin on collagenase and gelatinase activities in UMR 106-01 rat osteoblastic osteosarcoma cells. Matrix. 1990; 10(1): 27–32. PubMed Abstract \| Publisher Full Text\n\nDyce M, Sharif SF, Whalen GF, et al.: Search for anti-metastatic therapy: effects of phenytoin on B16 melanoma metastasis. J Surg Oncol. 1992; 49(2): 107–12. PubMed Abstract \| Publisher Full Text\n\nYang M, Kozminski DJ, Wold LA, et al.: Therapeutic potential for phenytoin: targeting Nav 1.5 sodium channels to reduce migration and invasion in metastatic breast cancer. Breast Cancer Res Treat. 2012; 134(2): 603–615. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRagsdale DS, Scheuer T, Catterall WA, et al.: Frequency and voltage dependent inhibition of type IIA Na+ channels, expressed in a mammalian cell line by local anesthetic, antiarrhytmic and anticonvulsivant drugs. Mol Pharmacol. 1991; 40(5): 756–765. PubMed Abstract\n\nAbdul M, Hoosein N: Inhibition by anticonvulsants of prostate-specific antigen and interleukin-6 secretion by human prostate cancer cells. Anticancer Res. 2001; 21(3B): 2045–8. PubMed Abstract\n\nFadiel A, Song J, Tivon D, et al.: Phenytoin is an estrogen receptor α-selective modulator that interacts with helix 12. Reprod Sci. 2015; 22(2): 146–55. PubMed Abstract \| Publisher Full Text\n\nNelson M, Yang M, Dowle AA, et al.: The sodium channel-blocking antiepileptic drug phenytoin inhibits breast tumour growth and metastasis. Mol Cancer. 2015; 14(1): 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee TS, Chen LC, Liu Y, et al.: 5,5-Diphenyl-2-thiohydantoin-N10 (DPTH-N10) suppresses proliferation of cultured colon cancer cell line COLO-205 by inhibiting DNA synthesis and activating apoptosis. Naunyn Schmiedebergs Arch Pharmacol. 2010; 382(1): 43–50. PubMed Abstract \| Publisher Full Text\n\nLiu Y, Wu J, Ho PY, et al.: Anti-angiogenic action of 5,5-diphenyl-2-thiohydantoin-N10 (DPTH-N10). Cancer Lett. 2008; 271(2): 294–305. PubMed Abstract \| Publisher Full Text\n\nSikes RA, Walls AM, Brennen WN, et al.: Therapeutic approaches targeting prostate cancer progression using novel voltage-gated ion channel blockers. Clin Prostate Cancer. 2003; 2(3): 181–187. PubMed Abstract \| Publisher Full Text\n\nAnderson JD, Hansen TP, Lenkowski PW, et al.: Voltage-gated sodium channel blockers as cytostatic inhibitors of the androgen-independent prostate cancer cell line PC-3. Mol Cancer Ther. 2003; 2(11): 1149–54. PubMed Abstract\n\nLobert S, Ingram JW, Correia JJ, et al.: Additivity of dilantin and vinblastine inhibitory effects on microtubule assembly. Cancer Res. 1999; 59(19): 4816–22. PubMed Abstract\n\nKawamura KI, Grabowski D, Weizer K, et al.: Modulation of vinblastine cytotoxicity by dilantin (phenytoin) or the protein phosphatase inhibitor okadaic acid involves the potentiation of anti-mitotic effects and induction of apoptosis in human tumour cells. Br J Cancer. 1996; 73(2): 183–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLang DG, Wang CM, Cooper BR: Lamotrigine, phenytoin and carbamazepine interactions on the sodium current present in N4TG1 mouse neuroblastoma cells. J Pharmacol Exp Ther. 1993; 266(2): 829–35. PubMed Abstract\n\nTittle TV, Schaumann BA, Rainey JE, et al.: Segregation of the growth slowing effects of valproic acid from phenytoin and carbamazepine on lymphoid tumor cells. Life Sci. 1992; 50(14): PL79–83. PubMed Abstract \| Publisher Full Text\n\nDosch HM, Jason J, Gelfand EW: Transient antibody deficiency and abnormal t-suppressor cells induced by phenytoin. N Engl J Med. 1982; 306(7): 406–409. PubMed Abstract \| Publisher Full Text\n\nSorrell TC, Forbes IJ: Depression of immune competence by phenytoin and carbamazepine. Studies in vivo and in vitro. Clin Exp Immunol. 1975; 20(2): 273–85. PubMed Abstract \| Free Full Text\n\nFontana A, Grob PJ, Sauter R, et al.: IgA deficiency, epilepsy, and hydantoin medication. Lancet. 1976; 2(7979): 228–31. PubMed Abstract \| Publisher Full Text\n\nMarcoli M, Gatti G, Ippoliti G, et al.: Effect of chronic anticonvulsant monotherapy on lymphocyte subpopulations in adult epileptic patients. Hum Toxicol. 1985; 4(2): 147–157. PubMed Abstract\n\nTeichmann M, Kretschy N, Kopf S, et al.: Inhibition of tumour spheroid-induced prometastatic intravasation gates in the lymph endothelial cell barrier by carbamazepine: drug testing in a 3D model. Arch Toxicol. 2014; 88(3): 691–9. PubMed Abstract \| Publisher Full Text\n\nStettner M, Krämer G, Strauss A, et al.: Long-term antiepileptic treatment with histone deacetylase inhibitors may reduce the risk of prostate cancer. Eur J Cancer Prev. 2012; 21(1): 55–64. PubMed Abstract \| Publisher Full Text\n\nMeng Q, Chen X, Sun L, et al.: Carbamazepine promotes Her-2 protein degradation in breast cancer cells by modulating HDAC6 activity and acetylation of Hsp90. Mol Cell Biochem. 2011; 348(1–2): 165–71. PubMed Abstract \| Publisher Full Text\n\nMeng QW, Zhao CH, Xi YH, et al.: [Inhibitory effect of carbamazepine on proliferation of estrogen-dependent breast cancer cells]. Ai Zheng. 2006; 25(8): 967–73. PubMed Abstract\n\nHuang CW, Huang CC, Lin MW, et al.: The synergistic inhibitory actions of oxcarbazepine on voltage-gated sodium and potassium currents in differentiated NG108-15 neuronal cells and model neurons. Int J Neuropsychopharmacol. 2008; 11(5): 597–610. PubMed Abstract \| Publisher Full Text\n\nBeutler AS, Li SD, Nicol R, et al.: Carbamazepine is an inhibitor of histone deacetylases. Life Sci. 2005; 76(26): 3107–3115. PubMed Abstract \| Publisher Full Text\n\nKostrouchová M, Kostrouch Z, Kostrouchová M: Valproic acid, a molecular lead to multiple regulatory pathways. Folia Biol (Praha). 2007; 53(2): 37–49. PubMed Abstract\n\nMichaelis M, Doerr HW, Cinatl J Jr: Valproic acid as anti-cancer drug. Curr Pharm Des. 2007; 13(33): 3378–93. PubMed Abstract \| Publisher Full Text\n\nBanon D, Filion KB, Budlovsky T, et al.: The usefulness of ranolazine for the treatment of refractory chronic stable angina pectoris as determined from a systematic review of randomized controlled trials. Am J Cardiol. 2014; 113(6): 1075–1082. PubMed Abstract \| Publisher Full Text\n\nDriffort V, Gillet L, Bon E, et al.: Ranolazine inhibits NaV1.5-mediated breast cancer cell invasiveness and lung colonization. Mol Cancer. 2014; 13: 264. PubMed Abstract \| Publisher Full Text\n\nDjamgoz MB, Onkal R: Persistent current blockers of voltage-gated sodium channels: a clinical opportunity for controlling metastatic disease. Recent Pat Anticancer Drug Discov. 2013; 8(1): 66–84. PubMed Abstract \| Publisher Full Text\n\nIngolfsson HI, Thakur P, Herold KF, et al.: Phytochemicals perturb membranes and promiscuously alter protein function. ACS Chem Biol. 2014; 9(8): 1788–1798. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFraser SP, Peters A, Fleming-Jones S, et al.: Resveratrol: inhibitory effects on metastatic cell behaviors and voltage-gated Na+ channel activity in rat prostate cancer in vitro. Nutr Cancer. 2014; 66(6): 1047–58. PubMed Abstract \| Publisher Full Text\n\nWeixel KM, Marciszyn A, Alzamora R, et al.: Resveratrol inhibits the epithelial sodium channel via phopshoinositides and AMP-activated protein kinase in kidney collecting duct cells. Plos One. 2013; 8(19): e78019. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWallace CH, Baczkó I, Jones L, et al.: Inhibition of cardiac voltage-gated sodium channels by grape polyphenols. Br J Pharmacol. 2006; 149(6): 657–665. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJia Z, Jia Y, Liu B, et al.: Genistein inhibits voltage-gated sodium currents in SCG neurons through protein tyrosine kinase-dependent and kinase-independent mechanisms. Pflugers Arch. 2008; 456(5): 857–66. PubMed Abstract \| Publisher Full Text\n\nLiu L, Yang T, Simon SA: The protein tyrosine kinase inhibitor, genistein, decreases excitability of nociceptive neurons. Pain. 2004; 112(1–2): 131–41. PubMed Abstract \| Publisher Full Text\n\nZhang JL, Yang JP, Zhang JR, et al.: Gabapentin reduces allodynia and hyperalgesia in painful diabetic neuropathy rats by decreasing expression level of Nav1.7 and p-ERK1/2 in DRG neurons. Brain Res. 2013; 1493: 13–8. PubMed Abstract \| Publisher Full Text\n\nYang RH, Wang WT, Chen JY, et al.: Gabapentin selectively reduces persistent sodium current in injured type-A dorsal root ganglion neurons. Pain. 2009; 143(1–2): 48–55. PubMed Abstract \| Publisher Full Text\n\nLiu Y, Qin N, Reitz T, et al.: Inhibition of the rat brain sodium channel Nav1.2 after prolonged exposure to gabapentin. Epilepsy Res. 2006; 70(2–3): 263–8. PubMed Abstract \| Publisher Full Text\n\nAbdul M, Hoosein N: Voltage-gated sodium ion channels in prostate cancer: expression and activity. Anticancer Res. 2002; 22(3): 1727–30. PubMed Abstract\n\nWall BA, Wangari-Talbot J, Shin SS, et al.: Disruption of GRM1-mediated signalling using riluzole results in DNA damage in melanoma cells. Pigment Cell Melanoma Res. 2014; 27(2): 263–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKiskin NI, Chizhmakov IV, Tsyndrenko AYa, et al.: R56865 and flunarizine as Na+-channel blockers in isolated Purkinje neurons of rat cerebellum. Neuroscience. 1993; 54(3): 575–85. PubMed Abstract \| Publisher Full Text\n\nFischer W, Kittner H, Regenthal R, et al.: Anticonvulsant profile of flunarizine and relation to Na+ channel blocking effects. Basic Clin Pharmacol Toxicol. 2004; 94(2): 79–88. PubMed Abstract \| Publisher Full Text\n\nYe Q, Yan LY, Xue LJ, et al.: Flunarizine blocks voltage-gated Na+ and Ca2+ currents in cultured rat cortical neurons: A possible locus of action in the prevention of migraine. Neurosci Lett. 2011; 487(3): 394–9. PubMed Abstract \| Publisher Full Text\n\nFrancis J, Burnham WM: [3H]Phenytoin identifies a novel anticonvulsant-binding domain on voltage-dependent sodium channels. Mol Pharmacol. 1992; 42(6): 1097–103. PubMed Abstract\n\nSchmeel LC, Schmeel FC, Kim Y, et al.: Flunarizine exhibits in vitro efficacy against lymphoma and multiple myeloma cells. Anticancer Res. 2015; 35(3): 1369–76. PubMed Abstract\n\nConrad DM, Furlong SJ, Doucette CD, et al.: The Ca2+ channel blocker flunarizine induces caspase-10-dependent apoptosis in Jurkat T-leukemia cells. Apoptosis. 2010; 15(5): 597–607. PubMed Abstract \| Publisher Full Text\n\nFink-Puches R, Helige C, Kerl H, et al.: Inhibition of melanoma cell directional migration in vitro via different cellular targets. Exp Dermatol. 1993; 2(1): 17–24. PubMed Abstract \| Publisher Full Text\n\nHofmann-Wellenhof R, Fink-Puches R, Smolle J, et al.: Correlation of melanoma cell motility and invasion in vitro. Melanoma Res. 1995; 5(5): 311–9. PubMed Abstract \| Publisher Full Text\n\nSezzi ML, De Luca G, Materazzi M, et al.: Effects of a calcium-antagonist (flunarizine) on cancer cell movement and phagocytosis. Anticancer Res. 1985; 5(3): 265–71. PubMed Abstract\n\nKaelin WG Jr, Shrivastav S, Jirtle RL: Blood flow to primary tumors and lymph node metastases in SMT-2A tumor-bearing rats following intravenous flunarizine. Cancer Res. 1984; 44(3): 896–9. PubMed Abstract\n\nBellelli A, Camboni C, de Luca G, et al.: In vitro and in vivo enhancement of vincristine antitumor activity on B16 melanoma cells by calcium antagonist flunarizine. Oncology. 1987; 44(1): 17–23. PubMed Abstract \| Publisher Full Text\n\nVaupel P, Menke H: Blood flow, vascular resistance and oxygen availability in malignant tumours upon intravenous flunarizine. Adv Exp Med Biol. 1987; 215: 393–8. PubMed Abstract \| Publisher Full Text\n\nWood PJ, Hirst DG: Cinnarizine and flunarizine as radiation sensitisers in two murine tumours. Br J Cancer. 1988; 58(6): 742–745. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAstin JW, Jamieson SM, Eng TC, et al.: An in vivo antilymphatic screen in zebrafish identifies novel inhibitors of mammalian lymphangiogenesis and lymphatic-mediated metastasis. Mol Cancer Ther. 2014; 13(10): 2450–62. PubMed Abstract \| Publisher Full Text\n\nGornati D, Zaffaroni N, Villa R, et al.: Modulation of melphalan and cisplatin cytotoxicity in human ovarian cancer cells resistant to alkylating drugs. Anticancer Drugs. 1997; 8(5): 509–16. PubMed Abstract \| Publisher Full Text\n\nCastellino SM, Friedman HS, Elion GB, et al.: Flunarizine enhancement of melphalan activity against drug-sensitive/resistant rhabdomyosarcoma. Br J Cancer. 1995; 71(6): 1181–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSilvestrini R, Zaffaroni N, Costa A, et al.: Flunarizine as a modulator of doxorubicin resistance in human colon-adenocarcinoma cells. Int J Cancer. 1993; 55(4): 636–9. PubMed Abstract \| Publisher Full Text\n\nKavanagh BD, Coffey BE, Needham D, et al.: The effect of flunarizine on erythrocyte suspension viscosity under conditions of extreme hypoxia, low pH, and lactate treatment. Br J Cancer. 1993; 67(4): 734–741. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSezzi ML, Zupi G, De Luca G, et al.: Effects of a calcium-antagonist (flunarizine) on the in vitro growth of B16 mouse melanoma cells. Anticancer Res. 1984; 4(4–5): 229–34. PubMed Abstract\n\nSo HS, Kim HJ, Lee JH, et al.: Flunarizine induces Nrf2-mediated transcriptional activation of heme oxygenase-1 in protection of auditory cells from cisplatin. Cell Death Differ. 2006; 13(10): 1763–1755. PubMed Abstract \| Publisher Full Text\n\nKang HJ, YI, Hong YB, et al.: HER2 confers drug resistance of human breast cancer cells through activation of NRF2 by direct interaction. Sci Rep. 2014; 4: 7201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSakaguchi M, Kuroda Y, Hirose M: The antiproliferative effect of lidocaine on human tongue cancer cells with inhibition of the activity of epidermal growth factor receptor. Anesth Analg. 2006; 102(4): 1103–1107. PubMed Abstract \| Publisher Full Text\n\nMammoto T, Higashiyama S, Mukai M, et al.: Infiltration anesthetic lidocaine inhibits cancer cell invasion by modulating ectodomain shedding of heparin-binding epidermal growth factor-like growth factor (HB-EGF). J Cell Physiol. 2002; 192(3): 351–8. PubMed Abstract \| Publisher Full Text\n\nBaptista-Hon DT, Robertson FM, Robertson GB, et al.: Potent inhibition by ropivacaine of metastatic colon cancer SW620 cell invasion and NaV1.5 channel function. Br J Anaesth. 2014; 113( Suppl 1): i39–i48. PubMed Abstract \| Publisher Full Text\n\nPiegeler T, Votta-Velis EG, Liu G, et al.: Antimetastatic potential of amide-linked local anesthetics: inhibition of lung adenocarcinoma cell migration and inflammatory Src signaling independent of sodium channel blockade. Anesthesiology. 2012; 117(3): 548–559. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoulin H, Bruhova I, Timour Q, et al.: Fluoxetine blocks Nav1.5 channels via a mechanism similar to that of class 1 antiarrhythmics. Mol Pharmacol. 2014; 86(4): 378–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Y, Mi J, Lu K, et al.: Comparison of Gating Properties and Use-Dependent Block of Nav1.5 and Nav1.7 Channels by Anti-Arrhythmics Mexiletine and Lidocaine. PLoS One. 2015; 10(6): e0128653. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMammoto T, Mukai M, Mammoto A, et al.: Intravenous anesthetic, propofol inhibits invasion of cancer cells. Cancer Lett. 2002; 184(2): 165–70. PubMed Abstract \| Publisher Full Text\n\nBesson P, Driffort V, Bon É, et al.: How do voltage-gated sodium channels enhance migration and invasiveness in cancer cells? Biochim Biophys Acta. 2015; pii: S0005-2736(15)00136-4. PubMed Abstract \| Publisher Full Text\n\nBrown MT, Herrmann DN, Goldstein M, et al.: Nerve safety of tanezumab, a nerve growth factor inhibitor for pain treatment. J Neurol Sci. 2014; 345(1–2): 139–47. PubMed Abstract \| Publisher Full Text\n\nKalman D, Wong B, Horvai AE, et al.: Nerve growth factor acts through cAMP-dependent protein kinase to increase the number of sodium channels in PC12 cells. Neuron. 1990; 4(3): 355–366. PubMed Abstract \| Publisher Full Text\n\nMnich K, Carleton LA, Kavanagh ET, et al.: Nerve growth factor-mediated inhibition of apoptosis post-caspase activation is due to removal of active caspase-3 in a lysosome-dependent manner. Cell Death Dis. 2014; 5: e1202. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChioni AM, Brackenbury WJ, Calhoun JD, et al.: A novel adhesion molecule in human breast cancer cells: voltage-gated Na+ channel beta1 subunit. Int J Biochem Cell Biol. 2009; 41(5): 1216–1227. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelson M, Millican-Slater R, Forrest LC, et al.: The sodium channel β1 subunit mediates outgrowth of neurite-like processes on breast cancer cells and promotes tumour growth and metastasis. Int J Cancer. 2014; 135(10): 2338–2351. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrisson L, Gillet L, Calaghan S, et al.: NaV1.5 enhances breast cancer cell invasiveness by increasing NHE1-dependent H+ efflux in caveolae. Oncogene. 2011; 30(17): 2070–2076. PubMed Abstract \| Publisher Full Text\n\nStettner M, Krämer G, Strauss A, et al.: Long-term antiepileptic treatment with histone deacetylase inhibitors may reduce the risk of prostate cancer. Eur J Cancer Prev. 2012; 21(1): 55–64. PubMed Abstract \| Publisher Full Text" }	[ { "id": "13165", "date": "19 May 2016", "name": "Stephen G Waxman", "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 article focuses attention on sodium channels as potential therapeutic targets for metastatic disease. Much work needs to be done to determine whether agents that act on sodium channels will blunt metastatic activity in a clinically meaningful manner, but a spectrum of sodium channel blockers with only minimal side-effects are already used clinically. This novel approach thus merits careful study.\nWhile at first glance it may seem surprising that voltage-gated sodium channels are proposed as molecular targets within (presumably) non-excitable cells, there are many precedents for a role of these channels in controlling effector actions in multiple types of cells that have traditionally been considered non-excitable1. One example is provided by astrocytes, which express multiple types of sodium channels that are functional within the cell membrane2. These astrocytic sodium channels provide a return pathway for Na ions that facilitates operation of the Na-K/ATPase in these cells3. Notably, expression of sodium channels within astrocytes is highly dynamic, a phenomenon that is strikingly seen in scarring astrocytes in disorders such as multiple sclerosis (MS) and its models where expression of Nav1.5 is up-regulated 4.\n\nRecent evidence indicates that these glial sodium channels participate in the astrocytic response to injury, via a cascade that involves Na influx that activates reverse (Ca-importing) Na/Ca exchange5. Given the large number of traditionally non-excitable cell-types (including microglia, macrophages, and multiple other cell-types) where expression of voltage-gated sodium channels and a functional role for these channels has been documented (reviewed in Black and Waxman, 20131), sodium channels may emerge as therapeutic targets in multiple disorders.", "responses": [] }, { "id": "14301", "date": "22 Jun 2016", "name": "Michihiro Mutoh", "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 manuscript is well summarized with topics. However, I believe some more information needs to be added.\n\nAs small molecules, such as FDA approved drugs, affect the tissue of the whole body, the information of the data from genetic findings, such as NaV1.5 knockout mice data, is desired to be included in the text\n\nThe idea that selecting the voltage gated sodium channel blocker by repurposing of a drug seems attractive to fight invasion and metastasis of cancer. However, a concrete example for the medical use is not shown in the text. Please provide an idea when a patient could use such a drug.\n\nOxidative stress contributes to the invasion process. Figures 3 & 4 show that proton exfusion also plays a role in the invadopodia complex. Do the author has an idea to connect oxidative stress and the voltage gated sodium channel? If not, please ignore this comment.\n\nFigure 7: “B1 subunit” should be replaced by “b(symbol)1 subunit”.\n\nThe author describes in the text that the voltage gated sodium channel is overexpressed in cancer tissue, and its blockers could be used as cancer chemoprevention agents. It might be worth writing a review article regarding carcinogenesis and the voltage gated sodium channel blocker in the future.", "responses": [] } ]	1	https://f1000research.com/articles/4-297
https://f1000research.com/articles/4-6/v1	09 Jan 15	{ "type": "Observation Article", "title": "Open peer review at four STEM journals: an observational overview", "authors": [ "Emily Ford" ], "abstract": "Open peer review, peer review where authors' and reviewers' identities are disclosed to one another, is a growing trend in scholarly publishing. Through observation of four journals in STEM disciplines, PLoS One, Atmospheric Chemistry & Physics, PeerJ, and F1000Research, an observational overview is conducted. The overview relies on defined characteristics of open peer review. Results show that despite differing open peer review implementations, each journal retains editorial involvement in scholarly publishing. Further, the analysis shows that only one of these implementations is fully transparent in its peer review and decision making process. Finally, the overview contends that journals should clearly outline peer review and editorial processes in order to allow for open peer review to be better understood and adopted by authors, reviewers, editors, and readers of science communications.", "keywords": [ "open peer review", "peer review", "scholarly communication", "science communication", "scholarly publishing" ], "content": "Introduction\n\nIn scholarly publishing open peer review (OPR) is an emerging form of peer review that incorporates disclosure of author and referee identities to one another. Although in its infancy, OPR has been adopted and implemented in a number of disciplines and their respective scholarly publications. In this article I provide some background on open peer review, addressing controversies and divergent opinions. Next I describe, examine, and discuss OPR implementations at four different science, technology, engineering, and mathematics (STEM) journals: PLoS One, Atmospheric Chemistry & Physics, PeerJ, and F1000Research. These observations contribute to our understanding of scholarly publication and scientific communication, as we watch the evolution of scientific vetting and validity determination processes.\n\n\nOpen Peer Review: a definition\n\nUnlike double-blind peer review, which is clearly defined, has clear parameters, and an arguably universal understanding by the scholarly community in how to implement it, OPR is approached and implemented in a variety of ways. There is no one universally accepted definition of OPR, which complicates investigations of its practices. As such, I rely on my previous definition, which understands OPR as any scholarly review mechanism providing disclosure of author and referee identities to one another at any point during the peer review or publication process (Ford, 2013).\n\nOther terms used to discuss OPR are peer-to-peer review and open review. Both these phrases insinuate OPR, but some have approached is as supplementary to formal peer review processes. For example, these review implementations rely on a community’s members to post comments on articles at pre-print servers, such as arXiv, or using comment features via journal websites, such as British Medical Journal and BioMed Central. It should be noted that when I mention open peer review, I discuss it as the formal process via which scholarly articles are vetted for publication.\n\nMentions of OPR in scholarly literature date back to Michael McGiffert’s 1988 article, “Is Justice Blind? An Inquiry into Peer Review”, in which McGiffert argues, based on survey results, that editors should protect the identity of authors, but that editors, “...should leave referees free to decide for themselves whether or not to make themselves known [to the author]” (p. 47, McGiffert, 1988). Over time, attitudes toward OPR have evolved, and support of OPR has grown, although it still remains debated. Although OPR is a phenomenon occurring across the academic disciplines, those in STEM are the most prolific. The first implementation of OPR occurred at Atmospheric Chemistry and Physics with its launch in 2001, which is discussed later in this article.\n\n\nWhy Open Peer Review?\n\nFor many, OPR addresses inherent issues in what has been the gold standard of double-blind peer review. Some see blind review processes as faulty in that referee anonymity allows for referee abuse. Others view OPR as a means to hold referees and authors accountable for their communications (Fitzpatrick, 2010; Cope & Kalantzis, 2009; Mulligan, 2008). It has also been argued that OPR allows for easier identification of scientific misconduct (Boldt, 2011), and that over time the quality of submitted articles will improve (Hu et al., 2010; Prug, 2010). OPR affords referees the ability to gain credit for and cite their contributions to science communication (Boldt, 2011; Bornmann & Daniel, 2010; Fitzpatrick, 2010; Prug, 2010; Pöschl, 2009). More broadly speaking, OPR provides the scholarly community an insight into author/referee conversations during the review process. Surfacing these conversations provides readers an expanded contextual discussion of the subject at hand, and enriches science communication for all stakeholders (Lipworth & Kerridge, 2011; Fitzpatrick, 2010; Friedman et al., 2010; Maharg & Duncan, 2007). Finally, perhaps the most convincing pro argument for OPR asserts that OPR processes allow for quicker publication and dissemination of scientific findings (Hu et al., 2010; Cope & Kalantzis, 2009; Pöschl, 2004).\n\nOne of the major arguments against open peer review is the perceived protection afforded both authors and reviewers in a blind process. For junior researchers serving as reviewers, blind review may allow them to feel more able to provide honest constructive criticism to senior researchers. Similarly, as authors, blind review is perceived as protecting junior researchers from public humiliation (Godlee, 2003). It has also been noted that some reviewers refuse to participate in OPR implementations, or still have concerns about them (Janowicz & Hitzler, 2012). These concerns still pervade conversations about OPR. Most recently, a survey of BMC Pharmacology and Toxicology Editorial Board Members surfaced continuing concerns regarding OPR at the journal. Despite these concerns BMC Pharmacology and Toxicology decided to “...continue with open peer review at BMC Pharmacology and Toxicology because of the ethical grounds for doing so and because the potential benefits outweigh the negatives”, (p. 4, Moylan et al., 2014). This is evidence that despite continuing concerns and resistance to it, OPR will continue to be implemented and evolve in STEM publishing. As such, scholars should understand OPR implementations in order to further innovate and evolve scholarly publishing and scientific communication.\n\n\nMethods\n\nFour STEM journals claiming to use open peer review processes were chosen for observation. These four journals were selected because they represent a difference in relative age, their perceived stature or authority in STEM, and for the salience of information regarding OPR on their respective websites. To review and understand the four different peer review implementations, this observation relied on the eight OPR characteristics I identified in 2013. I relied on these characteristics because there is no other documented common vocabulary used to discuss and analyze open peer review. The characteristics are cited in full below:\n\nSigned review refers to submitted reviews signed by the referee that are either published alongside articles at the time of publication or are signed when an author receives them.\n\nDisclosed review refers to a process in which referees and authors know each others’ identities during the peer review process, enabling them to engage in discussion or discourse.\n\nEditor-mediated review is a characteristic found in most open peer review processes. Editor mediation is any work done by a journal editor to facilitate open peer review. This may include editorial preselection of articles and/or final decision-making for acceptance or rejection of articles. The editor-mediated portion of any open peer review process may or may not be publicly disclosed.\n\nTransparent review refers to complete openness to a distinct community or the public. It allows a public community to watch peer review unfold. Authors and the public know referees’ identities, and referees know authors’ identities. Author responses to referee comments are public. In transparent review the public can see manuscripts, reviews, and replies from authors and public reviewers as well as the published articles.\n\nCrowd-sourced review is a public review process in which any community member may contribute to the article review. In crowd-sourced review there is no limit to the number of comments or reviews an article may receive. In some proposed implementations of crowd-sourced review, there is little editorial mediation of article reviews. Rather, authors may simply submit papers to a preprint server or other community for crowd-sourced commentary.\n\nPre-publication review occurs prior to article publication, and typically occurs in a public space such as a pre-print server.\n\nSynchronous review occurs at the same time as publication of the article. In the literature, synchronous review is approached only theoretically, as part of a novel and completely iterative publishing model.\n\nPost-publication review occurs after an article is published, much like commentary on a blog post (pp. 314–315, Ford, 2013).\n\nUsing these characteristics I examined information for authors and about each publication at their respective websites, promotional materials, blogs, and other materials discussing the OPR processes at each journal. Data for these observations was gathered in mid to late 2013. Publisher/journal policies and practices may have since changed.\n\n\nThe Journals\n\nPLoS One is an international publication of Public Library of Science, a not-for-profit publisher and open access advocacy organization. The journal was formed around the philosophy and practice that all research using scientifically sound research methods should be published regardless of its results, novelty, and/or impact. The journal publishes research articles from science and medical disciplines, including those reporting negative results. By publishing research from multiple disciplines, the journal boasts “PLoS One facilitates the discovery of the connections between papers whether within or between disciplines”.\n\nPLoS One launched in December 2006 and has since seen tremendous growth. It is indexed in numerous databases; is frequently cited as a source of research in news and popular media; and has received positive press for its review process. Even John Bohannon, a science journalist who undertook a sting operation of open access journals in an attempt to uncover poor publishing practices, acknowledged the strength of PLoS One’s review process (Bohannon, 2013, ¶ 9).\n\nAll articles published in PLoS One carry Creative Commons attribution licenses. Although most authors publishing in the journal pay article processing charges (APCs), the journal makes exceptions and waives publishing fees for unfunded research. Moreover, the APC fee model at PLoS One takes into account an author’s country of origin, and whether it is a high, lower middle, or low income nation. In this way the publication aims to make more viable open access publication for authors with disparate economic means.\n\nCompared to the other journals I discuss in this article, PLoS One conservatively approaches OPR. The journal’s peer review process only exhibits a few OPR characteristics, and even then these characteristics are not consistently implemented. It does, however, always use a form of editor-mediation for reviewing and publishing its content. Each submitted article is assigned an Academic Editor, who then determines whether submissions should be considered for peer review, and who then facilitates the peer review process. According to the journal’s review guidelines, these Academic Editors may, “...conduct the peer review themselves, based on their own knowledge and experience; they can take further advice through discussion with other members of the editorial board; they can solicit reports from further referees”.\n\nWhile the editorial mediation of journal articles at PLoS One always occurs, other characteristics of OPR do not. Signed reviews are optional, but they are strongly encouraged. According to its peer review guidelines, “If Peer Reviewers are willing, then they are also identified to the author at the time of decision.” PLoS One does not post reviewer comments to the web alongside published articles, so readers do not benefit from reading discussions that occurred about the topic prior to publication. In addition to PLoS One’s version of signed review, the journal enables crowd-sourced review, which occurs post-publication. Readers may comment on articles that have been published. One unique aspect of this crowd-sourced review process is that PLoS One surfaces any media coverage of published articles by linking to them in an article’s comments section. The result is that the journal is able to create a record of the impact and conversation an article elicits outside of the PLoS One platform and community.\n\nPerhaps the oldest of the open peer reviewed publications is Atmospheric Chemistry and Physics (ACP), the European Geosciences Union’s journal. The journal, which launched in September of 2001 (Pöschl, 2004), publishes research articles, review articles, technical notes, peer-reviewed comments, correigenda, and supplementary materials. All published content is published under a creative commons 3.0 attribution license and authors are subject to APCs.\n\nFollowing an article’s submission to ACP and brief editorial review, it is then hosted on the journal’s pre-print server, Atmospheric Chemistry and Physics Discussion Papers (ACPD) for peer review and crowd-sourced discussion. On this platform reviewer comments and crowd-sourced comments are publicly available. After the discussion period for a paper ends, the ability to comment on the paper is turned off, and the journal editor makes final publishing decisions using submitted referee and public reviews. When the editor accepts an article for publication, the article is published at ACP, where it will also link to its pre-print version, and referee and public comments at ACPD. ACP does not host public commentary on published articles. Although ACP views its peer review process as completely transparent (Pöschl, 2004), it is not. Reviewers may choose to disclose their identities, or they may choose to remain anonymous. True transparency of the review process can only occur when reviewers and public commenters have disclosed their identities.\n\nIn addition to editor mediation and crowd-sourced reviewing, ACP’s process employs disclosed review and pre-publication review. It could be argued that synchronous review occurs, yet ACP does not consider papers posted at ACPD to be “published.” Because these papers are not considered published, their review is not synchronous.\n\nPeerJ is an individual membership-based publisher in the biological and medical sciences. All works published by PeerJ are licensed with a Creative Commons Attribution 3.0 Unported license. Its main publication, PeerJ only publishes research articles. All other publication types are referred to PeerJ PrePrints, its pre-print repository and publication. PeerJ is young. The publisher first announced its publication model in June 2012, and published its first article on February 13, 2013. Its model relies on membership, where individuals pay one fee to PeerJ and become lifetime members. Based on an individual’s membership level—basic, enhanced, or investigator—individuals may publish in PeerJ a dedicated number of times per year.\n\nThe publishing model at PeerJ is similar to the pre-print/publication relationship between ACP and ACPD in that it maintains two publishing platforms: PeerJ and PeerJ PrePrints. Arguably, PeerJ PrePrints is not a publication, but a pre-print repository service. However, unlike ACP, PeerJ considers PeerJ PrePrints a publication, so I will treat it as such alongside PeerJ.\n\nOnly paid PeerJ members may publish work in PeerJ. Works may first be submitted to PeerJ PrePrints, or may be directly submitted to PeerJ. PeerJ’s article review, acceptance and publication model mirrors PLoS One’s; PeerJ accepts scientifically sound research and does not consider an article’s “novelty, interest, or impact” as part of its editorial criteria. Prior to review, submissions to PeerJ undergo editorial vetting by assigned Academic Editors. These Academic Editors are responsible for facilitating peer review of assigned articles to be completed by at least one reviewer, and making final publication decisions. Additionally, Academic Editors are attributed alongside each published article. Authors and reviewers alike are “encouraged” to post the full peer review cycle online alongside final versions of articles, but the journal does not mandate it. Based on information from PeerJ’s website, it is unclear whether author and reviewer identities are disclosed to one another during the review process.\n\nUnlike the crowd-sourced review occurring at PLoS One and ACP, PeerJ utilizes a broader model named Q&A. PeerJ’s Q&A incorporates not only questions and answers regarding pre-prints and articles—which can be posed at the paragraph and figure level—it also allows for free-standing questions. In this way, Q&A is intended to be a platform for anyone to participate in scientific conversation. At PeerJ Q&A includes an incentive system for individual participants in the community. Contributing individuals are awarded points for their engagement. For example, one earns 100 points for authoring an article, 35 points for contributing an open review, etc. These points are displayed on members’ profile pages.\n\nPeerJ has future plans to expand PeerJ PrePrints. The publisher hopes to allow authors to share as much or as little of their publications as they wish. They may decide to openly publish a title, title and abstract, or the whole paper. Additionally, PeerJ PrePrints will allow for authors to share papers “privately” with only particular users, only the PeerJ community, or fully open on the web.\n\nFinally, the fourth of the journals I discuss, F1000Research, most consistently exhibits OPR characteristics. It is an open access journal published by Faculty of 1000. The publisher calls it one of “...four unique services that support and inform the work of life scientists and clinicians” provided by the publisher. F1000Research published its first approved article in July 2012, only six months after Faculty of 1000 announced the new publication (Lawrence, 2012a, 2012b). For published data the journal utilizes a Creative Commons No Rights Reserved license; it requests attribution for works, but anyone anywhere is free to use, build upon, and manipulate works. For published articles the journal uses a Creative Commons Attribution license. The journal itself includes case reports, clinical practice articles, commentary, correspondence, data articles, method articles, opinion articles, research articles, reviews, short research articles, study protocols, systematic reviews, thought experiments, and web tools in the Life Sciences. Authors submitting articles to F1000Research pay APCs (F1000Research, 2013a).\n\nAt F1000Research articles undergo a peer review process after they are published. As such, F1000Research is the only publication I discuss that uses a post-publication review process. In this way the journal is able to speed up publication timelines to disseminate scholarly work; the journal publishes articles within one week of submission. As stated in its referee guidelines, the journal publishes submitted articles that pass initial editorial review for “content, quality, tone and format” as well as completeness, plagiarism, ethical standards, and adherence to author guidelines. In addition to reviews provided each work by two or three designated expert referees—which are attributed to the reviewer and are published online with the work—the scientific public (those affiliated with scientific or medical organizations) may comment on any published article. Any author responses to referees are also public.\n\nIn this publication model it is possible for articles to receive unanimous negative reviews. In this case, articles remain “published”, but are removed from the site’s default search. The site’s interface clearly delineates a work’s referee status and comments using icons to distinguish its status: approved with reservations, approved, and not approved. It is important to note that F1000Research does not consider these statuses as equivalent to the accepted, accepted with revisions, and rejected statuses that one sees in closed review processes. Rather, as stated in the FAQs:\n\nThe term Approved means that the referee thinks that the article is good and has either no suggested revisions or only minor revisions. The term Approved with Reservations means that the referee agrees that the article has scientific merit and is fundamentally sound but would like the author to make further changes to the manuscript. This is approximately equivalent to a request for major revisions or several minor revisions in a traditional journal. In every case, even when all referees approve of the article, future versions are welcome.\n\nBecause F1000Research publishes and attributes all referee responses and author comments, it adheres to a fully transparent peer review process. In addition to transparent reviews provided by pre-selected referees, crowd-sourced review occurs when individuals comment on published articles. The journal exhibits most other OPR characteristics; its reviews are editor-mediated, transparent, referee and author identities are disclosed, and reviews are signed. The only OPR characteristics not exhibited by F1000Research’s process are those related to review timing. Since all review at F1000Research occurs post-publication, the journal does not exhibit pre-publication and synchronous review characteristics. Once a work has been vetted by editors and is made public on the journal’s site, the journal considers it published. Because review occurs post publication, authors receiving critical feedback are encouraged to revise and submit updated versions of articles that will, again, be refereed. The journal uses CrossRef’s CrossMark identification service to assist readers in tracking these article versions and relationships. Even if an author publishes an updated article, previous versions remain published. In this way, publication at F1000Research is a good example of the iterative process of publishing and scientific knowledge and conversation.\n\n\nDiscussion\n\nI examined four journals that boast open peer review processes, PLoS One, ACP, PeerJ, and F1000Research. None of these journals have implemented OPR in the same manner, but they do exhibit many of the same OPR characteristics. Each journal exhibits a form of editor-mediation and each journal vets submitted articles prior to publication for basic quality, scope and adherence to author guidelines. While for some journals, such as F1000Research, this is as far as an editor’s work goes before an article is published, others, such as ACP and PeerJ, allow editors to make final publishing decisions.\n\nEach of the journals allow for some form of crowd-sourced review. At PeerJ the Q&A section includes commentary on articles; it also includes other avenues for the public to engage in scientific conversations. At PLoS One, however, commentary on articles occurs only after an article has been published, and also includes links to all media coverage of articles. Unlike PLoS One, PeerJ, and F1000Research’s crowd-sourcing implementations, ACP’s process only allows for crowd-sourced commentary prior to publication articles during the discussion phase of the OPR process.\n\nAnother commonality between these publications is their varying allowance, encouragement, or mandate for reviewers to sign their commentary. Referees may choose to remain anonymous at ACP and PeerJ (even though disclosure is strongly encouraged), whereas F1000Resarch and PLoS One require referees to disclose their identities. Since author/reviewer identity disclosure is a defining factor of OPR, it could be argued that those articles where reviewer commentary is not attributed are not truly open peer-reviewed. The motivation for publishers to encourage rather than require openness most likely stems from their desire to encourage more authors and reviewers to participate in alternative peer review processes. Publishers may also be attempting to be mindful of different discipline’s accepted publishing practices. In my own view, by not mandating public attributed review, publishers are weakening the power of OPR. However, incremental steps in OPR implementation are necessary to encourage participation and to move OPR to a completely transparent standard in the future.\n\nTwo of the journals discussed above include article pre-print mechanisms. Pre-print servers and mechanisms introduce confusion into understanding publishing and open peer review. Just when is something that is open to be read on the web considered “published?” Where ACP does not consider papers posted to its pre-print space, ACPD, as “published”, PeerJ PrePrints is considered by its publisher a publication. Defining a moment of when scholarly work is “published” will continue to evolve as does scholarly publishing and open peer review processes.\n\nOf the four publishers discussed, I maintain that F1000Research exhibits what we should consider the gold standard of transparent and OPR processes. The publication’s process is completely transparent; it publishes all commentary with attribution and makes salient referee decisions. Moreover, the mechanism it uses to track and correlate article versions and updates enhances and opens scholarly conversations. Yet, F1000Research maintains its editorial voice via editor-mediation prior to an article’s publication and by suppressing from search results articles receiving unanimous negative reviews.\n\n\nConclusion\n\nScholarly journals are beginning to challenge traditional peer review practices by implementing OPR, yet each OPR implementation differs. By observing four different implementations of OPR I conclude that few OPR journals implement truly transparent review, yet each implementation values editorial work. Further, I maintain that distinguishing between publicly available preprints and publicly available published articles unnecessarily muddies the waters in understanding OPR. As OPR implementations proliferate, it is pertinent for journals to clearly outline any peer review process so that readers, authors, and reviewers can fully understand peer review implementations, decision making processes, and to provide for editorial transparency.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed. The author has no affiliation with any of the journals discussed in this article.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBohannon J: Who’s afraid of peer review? Science. 2013; 342(6154): 60–65. PubMed Abstract \| Publisher Full Text\n\nBoldt A: Extending ArXiv.org to Achieve Open Peer Review and Publishing. Journal of Scholarly Publishing. 2011; 42(2): 238–242. Publisher Full Text\n\nBornmann L, Daniel HD: Reliability of reviewers’ ratings when using public peer review: a case study. Learn Publ. 2010; 23(2): 124–131. Publisher Full Text\n\nCope W, Kalantzis M: Signs of epistemic disruption: Transformations in the knowledge system of the academic journal. First Monday. 2009; 14(4–6). Publisher Full Text\n\nFitzpatrick K: Peer-to-peer Review and the Future of Scholarly Authority. Social Epistemology. 2010; 24(3): 161–179. Publisher Full Text\n\nFord E: Defining and Characterizing Open Peer Review: A Review of the Literature. Journal of Scholarly Publishing. 2013; 44(4): 311–326. Publisher Full Text\n\nFriedman R, Whitworth B, Brownstein M: Realizing the Power of Extelligence: A New Business Model for Academic Publishing. International Journal of Technology, Knowledge & Society. 2010; 6(2): 105–117. Reference Source\n\nGodlee F: Peer Review in Health Sciences (2nd ed.). London: BMJ Books. 2003. Reference Source\n\nHu C, Zhang Y, Chen G: Exploring a New Model for Preprint Server: A Case Study of CSPO. Journal of Academic Librarianship. 2010; 36(3): 257–262. Publisher Full Text\n\nJanowicz K, Hitzler P: Open and transparent: the review process of the Semantic Web journal. Learn Publ. 2012; 25(1): 48–55. Publisher Full Text\n\nLipworth W, Kerridge IH, Carter SM, et al.: Should Biomedical Publishing Be “Opened Up”? Toward a Values-Based Peer-Review Process. J Bioeth Inq. 2011; 8(3): 267–280. Publisher Full Text\n\nMaharg P, Duncan N: Black Box Pandora’s Box or Virtual Toolbox? An Experiment in a Journal’s Transparent Peer Review on the Web. International Review of Law Computers & Technology. 2007; 21(2): 109–128. Publisher Full Text\n\nMcGiffert M: Is Justice Blind? An Inquiry into Peer Review. Scholarly Publishing. 1988; 20(1): 43–48.\n\nMoylan EC, Harold S, O’Neill C, et al.: Open, single-blind, double-blind: which peer review process do you prefer? BMC Pharmacol Toxicol. 2014; 15(1): 55. Publisher Full Text\n\nMulligan A: Quality, certification and peer review. Information Services & Use. 2008; 28(3–4): pp. 197–214. Reference Source\n\nPöschl U: Interactive journal concept for improved scientific publishing and quality assurance. Learn Publ. 2004; 17(2): 105–113. Publisher Full Text\n\nPöschl U: Interactive Open Access Peer Review: The Atmospheric Chemistry and Physics Model. Against the Grain. 2009; 21(3): 26–32. Reference Source\n\nPrug T: Open-process academic publishing. Ephemera: Theory & Politics in Organization. 2010; 10(1): 40–63. Reference Source" }	[ { "id": "7269", "date": "12 Jan 2015", "name": "Peter Binfield", "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\nThank you for an interesting article which overviews an emerging form of publication / peer reviewGeneral Comments: At an early point, the author defines OPR as simply naming the authors and reviewers (“any scholarly review mechanism providing disclosure of author and referee identities to one another at any point during the peer review or publication process”), but then goes on (e.g. in the 2nd half of para 1 of “Open Peer Review”) to discuss it as if it also includes making those comments publicly available at the time of publication (something which is certainly considered part of OPR by many people). Therefore, the author either needs to expand their definition, or they need to word their commentary differently.For example, when the author makes statements that:\"More broadly speaking, OPR provides the scholarly community an insight into author/referee conversations during the review process.\"They are implying that the community can read the comments upon publication. This is not the case when using the narrow definition which the author has adopted (which is stated as “any scholarly review mechanism providing disclosure of author and referee identities to one another at any point during the peer review or publication process” and thus does not explicitly include the possibility that peer-review comments will be made public upon publication).I suggest that the author clearly defines what they mean as OPR for the purposes of this article Note: It is interesting that all the studied journals are OA – is that worth a discussion? Is OA more naturally favorable to OPR? Are there any subscription journals operating OPR? If not, why not? Edits:“have approached is as supplementary” should read “have approached it as supplementary” PLoS One should be written PLOS ONE throughout\"individuals may publish in PeerJ a dedicated number of times per year\" edit to: “individuals may publish in PeerJ a specified number of times per year””peer review of assigned articles to be completed by at least one reviewer,” change to: “peer review of assigned articles to be completed by at least two reviewers,”Grammar needs fixing in: “In addition to reviews provided each work by two or three designated expert referees“ Notes:\"By publishing research from multiple disciplines, the journal boasts “PLoS One” use of the word ‘boasts’ seems unnecessarily emotional here. “The first implementation of OPR occurred at Atmospheric Chemistry and Physics with its launch in 2001” This seems unlikely. Is there a reference? The article then says: \"Perhaps the oldest of the open peer reviewed publications is Atmospheric Chemistry and Physics”. So which is it? The oldest? Or ‘perhaps’ the oldest? ”In addition to PLoS One’s version of signed review, the journal enables crowd-sourced review, which occurs post-publication.” Although I note the authors definition of “crowd sourced review” (their bullet 5 in Methods), it is worth saying that at the time this article was written, PLOS ONE did not (and still doesn’t, I believe) regard (or promote) their commenting facility as a form of review. Instead they regard it as a way to enable “post publication commentary”. Same goes for PeerJ. \"One unique aspect of this crowd-sourced review process is that PLoS One surfaces any media coverage of published articles by linking to them\" This isn't necessarily unique, PeerJ does this as well via the ability for any user to “add link”. Also, many other journals do this via their article level metrics provision. ”PeerJ is an individual membership-based publisher in the biological and medical sciences.” At the time this article was written this was correct. However, since that time, the journal changed the description of its model to “publication plans” rather than “memberships” ”All works published by PeerJ are licensed with a Creative Commons Attribution 3.0 Unported license” Correct at the time the article was written. However, since that time the journal has upgraded to CC BY 4.0 ”Based on information from PeerJ’s website, it is unclear whether author and reviewer identities are disclosed to one another during the review process.” I can clarify this – reviewers are encouraged to name themselves (~40% choose to do so). If they do so, then the authors are first made aware of their identities in the decision letter that includes the comments of that reviewer. ”I examined four journals that boast open peer review processes, PLoS One,”. Actually, I am not sure that PLOS ONE would claim this fact (and the author themselves observes that very few elements of OPR are provided by PLOS ONE) ”whereas F1000Resarch and PLoS One require referees to disclose their identities”. This is not the case for PLOS ONE (unless something has changed which I am not aware of). The default for PLOS ONE reviewers is to remain anonymous.", "responses": [ { "c_id": "1194", "date": "04 Feb 2015", "name": "Emily Ford", "role": "Author Response", "response": "Thank you for your thoughtful comments. Your working knowledge of both PeerJ and PLOS ONE will benefit this article's revision, and I appreciate you offering this knowledge in your review.Regarding your note about OPR and the overlap with OA journals: yes, OA journals are generally more sympathetic and willing to try OPR. Since you ask it is clear to me that I have not adequately discussed this overlap in the article, which I will be sure to in a revision. Thank you again for your comments." } ] }, { "id": "7270", "date": "12 Jan 2015", "name": "Scott Walter", "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 study applies the tools of content analysis to the stated editorial policies and procedures of a small number of journals in STEM disciplines adhering (to a greater or lesser degree) to the principles of open peer review. To promote replicability of the study and/or comparison between the set of journals included in the current study and other sets of journals, the study might be revised to include a table documenting the degree to which each journal under consideration did or did not meet the stated OPR principles. While the study is limited to review of relatively new publications and publication models in a defined field, it might also be revised to suggest future research opportunities, e.g., the degree to which OPR principles are employed in more established journals, the degree to which OPR experiments are being explored in other fields.", "responses": [ { "c_id": "1193", "date": "04 Feb 2015", "name": "Emily Ford", "role": "Author Response", "response": "Thank you for your comments and feedback. The suggestion of a table, I think, will help to clearly communication these observations, so a revision will include one. One of the things that I have struggled with is how to represent what might be happening in other disciplinary communities. From what I have found, OPR seems to have started in STEM disciplines. I'm finding it difficult to wrap my thoughts and ideas around OPR in the humanities since humanities publishing is also changing with new ways to express scholarship, aka the digital humanities. In this way it is easier (for me) to begin observing and evaluating OPR in STEM publications. Again, thank you for your thoughts and comments, and I look forward to more comments after I re-submit this article." } ] }, { "id": "7273", "date": "03 Mar 2015", "name": "Ulrich Pöschl", "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\nThanks to the author for an interesting perspective article on open peer review. Following an invitation to review this article, I would like to request a couple of specific corrections and add a general comment and suggestion.Correction of Erroneous Statement: \"Although ACP views its peer review process as completely transparent (Pöschl, 2004), it is not.\"Neither the cited reference nor the ACP journal web pages claim that the peer review process of ACP would be \"completely transparent\". There are good reasons to maintain an option for referees to remain anonymous, and this is also very clearly specified on the journal web pages as well as in the referenced article and in a more comprehensive review article (Pöschl, 2012 as referenced and linked below). Thus, it would indeed be wrong to claim “complete transparency” for the open peer review process for ACP, but it is also wrong to insinuate that ACP or the cited article would have raised this claim. Please discard or correct the erroneous statement accordingly. Moreover, please complement or replace the reference to Pöschl 2004 by a reference to Pöschl 2012, because the latter is more comprehensive and up-to.date:Pöschl U (2012) Multi-stage open peer review: scientific evaluation integrating the strengths of traditional peer review with the virtues of transparency and self-regulation. Front. Comput. Neurosci. 6:33. doi: 10.3389/fncom.2012.00033http://journal.frontiersin.org/article/10.3389/fncom.2012.00033/abstract Correction of Erroneous Statement: “It could be argued that synchronous review occurs, yet ACP does not consider papers posted at ACPD to be “published.” Because these papers are not considered published, their review is not synchronous.” Neither the ACP journal web pages nor the articles providing an authoritative description of the ACP open peer review (Pöschl, 2004 & 2012) have ever indicated that the papers posted in the ACP discussion forum, ACPD, would not be “published”. Well on the contrary, all relevant journal web pages emphasize explicitly that the discussion papers posted and reviewed/discussed in ACPD are publications with permanent public accessibility, archiving and citability. Also my articles explaining the concepts of interactive open access publishing and multi-stage open peer review (Pöschl, 2004 and 2014) specify and emphasize that the discussion papers published in ACPD - as well as in the fifteen interactive open access sister journals of the European Geosciences Union (EGU) - are indeed publications. Thus, it is inappropriate to insinuate the opposite. Please discard or correct the erroneous statement accordingly.http://www.egu.eu/about/statements/position-statement-on-the-status-of-discussion-papers-published-in-egu-interactive-open-access-journals/ http://www.atmospheric-chemistry-and-physics.net/general_information/publication_policy.html http://www.atmospheric-chemistry-and-physics.net/review/review_process_and_interactive_public_discussion.html http://www.atmospheric-chemistry-and-physics.net/general_information/faq.html http://www.ingentaconnect.com/content/alpsp/lp/2004/00000017/00000002/art00005 http://journal.frontiersin.org/article/10.3389/fncom.2012.00033/abstract\n\nGeneral Comment and Suggestion:I understand and respect that the author prefers a fully transparent form of open peer review as implemented by F1000Research. However, F1000Research is a relatively recent follow-up on a series of earlier initiatives that have been pursuing open peer review and are reaching back into the last century. For example, see the British Medical Journal (BMJ), the Electronic Transactions on Artificial Intelligence (ETAI), the Journal of Interactive Media in Education (JIME) and other journals with open peer review as referenced in the recent “research topic” (special issue) on open evaluation (including open peer review) in the open access journal Frontiers in Computational Neuroscience: http://journal.frontiersin.org/researchtopic/137 As detailed in my contribution to that collection of articles (Pöschl 2012) and confirmed by independent studies referenced/linked below, ACP and its sister journals are by most standards of scientific publishing more successful and more efficient than comparable journals with traditional or alternative forms of peer review: http://www.atmospheric-chemistry-and-physics.net/pr_acp_is_interactive_open_access_publishing_able_to_identify_high_impact_submissions.pdf http://www.biomedcentral.com/1471-2288/13/74 Overall, I see no well-founded basis for the claim that the particular form open peer review practiced by F1000Research would deserve the attribute “gold standard of open peer review”, and I would suggest to substantiate or drop this postulate.", "responses": [] }, { "id": "7267", "date": "04 Mar 2015", "name": "Pandelis Perakakis", "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\nThank you for this interesting observation article and for the invitation to contribute my open peer review.In the beginning, the article tries to disambiguate different “flavours\" or implementations of open peer review (OPR) by identifying a number of OPR characteristics. Regarding this effort I would like to note the following issues that I think should be somehow addressed in a future version of the article:As already noted by Peter Binfield, the definition offered is limited to only one of the identified OPR characteristics, namely whether the reviews are signed or not. It is not clear what is the difference between reviews “signed when the author receives them” and disclosed reviews. It seems that in both cases the names of authors and reviewers are disclosed from the beginning of the review process. Is disclosure of the reviewer’s identity necessary to initiate a discussion between authors and reviewers? Perhaps what you refer to here is an infrastructure to allow multiple iterations during the review process, as provided for example by Frontiers. This is different from simply revealing the name of the reviewer. There is a clear distinction between an unsolicited comment on an article and a formal review process by expert peers that receive a review invitation either by an experienced editor or the authors themselves. What is the alternative to editor mediation in peer review, if we consider that unsolicited commentaries do not constitute formal peer reviews? Is there another model where formal peer review is not editor-mediated? I know we proposed such model that we called author-guided peer review in Perakakis, P., Taylor, M., Mazza, M., Trachana, V. (2010). Natural selection of academic papers. Scientometrics, 85 (2), pp. 553–559, but I am not aware of any implementations. I repeat that I do not think we should consider crowd-sourced comments as equal to formal peer reviews.And some other comments to consider:PLOS ONE clearly does not qualify as an OPR journal by any standards. Instead, I would be more interested to see a discussion on the implementation of OPR by Frontiers. I suspect there is a misunderstanding around the terms double-blind and single-blind review. The standard in most journals and disciplines is single-blind review where authors are blind to the name of the reviewer, but reviewers know the name of the authors. Double-blind refers to the process where neither authors nor reviewers are aware about the identity of each other. It is not entirely correct that PLOS ONE “waives publishing fees for unfunded research”. What they do is ask for proofs that authors cannot cover the APCs even from their own budget, sometimes even asking for personal financial records, which by the way I find unacceptable. Proving that there is no formal funding to publish the particular research is not enough to waive APCs. I definitely agree that a comparison table is needed. I detected that the references to the articles by Lawrence are missing. Please double check that all references are included.Thank you again for the opportunity to participate in this process that is very close to my ideal model of OPR and I will be looking forward to a revised version of your manuscript.", "responses": [] } ]	1	https://f1000research.com/articles/4-6
https://f1000research.com/articles/4-289/v1	20 Jul 15	{ "type": "Review", "title": "The role of new echocardiographic techniques in athlete’s heart", "authors": [ "Antonello D'Andrea", "Eduardo Bossone", "Juri Radmilovic", "Pio Caso", "Raffaele Calabrò", "Maria Giovanna Russo", "Maurizio Galderisi", "Eduardo Bossone", "Juri Radmilovic", "Pio Caso", "Raffaele Calabrò", "Maria Giovanna Russo", "Maurizio Galderisi" ], "abstract": "‘Athlete’s heart’ is a common term for the various adaptive changes induced by intensive exercise. Exercise causes alterations of the heart in hemodynamic response to the increased systemic and pulmonary demand during exercise. The understanding of these adaptations is of high importance, since they may overlap with those caused by pathological conditions. Cardiac imaging assessment of the athlete’s heart should begin with a complete echocardiographic examination. In recent years classical echocardiographic surveys have been joined by new developments: tissue Doppler imaging, strain rate echocardiography, and real-time 3-dimensional echocardiography. This review paper focuses on the importance of these new echocardiographic techniques in delineating the morphological characteristics and functional properties of the athlete’s heart.", "keywords": [ "athlete’s heart", "left ventricular hypertrophy", "strain", "tissue Doppler", "three-dimensional echocardiography", "sport" ], "content": "The athlete’s left heart\n\nLong term physical training causes structural, functional and electrical changes in the heart that are physiological responses to the hemodynamic demands of increased cardiac output during effort. This adaptive remodelling can be defined as “athlete’s heart”.\n\nThe understanding of these changes is of high importance, since they have to be distinguished from those caused by pathological conditions. Moreover, there is some evolving evidence suggesting that some of the exercise-induced changes may be associated with acute and chronic cardiac damage and that in a small number of athletes this may predispose to atrial and ventricular arrhythmias. Thus, the need for a standardization of cardiovascular pre-participation screening of competitive athletes for sports eligibility has emerged, since athletes with underlying, masked cardiomyopathy may be at risk of lethal consequences during physical exertion1,2.\n\nAccording to the Morganroth’s original hypothesis, two main models of training can be identified, which cause two distinct patterns of cardiac remodelling (myocardial hypertrophy)2. Endurance training characterizes aerobic sports with dynamic-isotonic muscular involvement – such as long-distance swimming and running. These activities cause a gradual decrease in systemic arterial resistance and an increase in venous return, with a predominant volume overload, with higher left ventricular (LV) end diastolic volume (EDV) and stroke volume (eccentric hypertrophy).\n\nOn the other hand, strength training is typical of anaerobic sports characterized by predominant static-isometric muscular exercise, such as body-building, short-distance running and swimming. These sports categories cause mainly an increase in myocardial wall thickness rather than cavity diameters (concentric remodelling and hypertrophy), in response to the predominant pressure overload.\n\nMorganroth’s original hypothesis has been criticized, because cardiac remodelling is also influenced by other factors like ethnicity, age, sex, genetics and body size. Moreover it has to be noted that most sports are actually characterized by a variable combination of both endurance and strength exercise, rather than only one of them.\n\nStandard echocardiography has an essential role in assessing the characteristics of the athlete’s heart and in differentiating physiological and pathological LV hypertrophy (LVH)3. Previous authors4 in a large series of top level athletes reported that 55% had increased LV end-diastolic diameter and only 15% of them had values > 60 mm, even if ejection fraction (EF) was normal. Competitive athletes have LVH, involving all myocardial segments, with a maximal septal thickness < 12 mm. Conversely, patients with hypertrophic cardiomyopathy (HCM) show increased wall thickness (>15 mm), mainly in the basal septum, and in 20% of cases there is systolic anterior motion (SAM) of the mitral valve, or aortic valve mid-systolic closure5. After a deconditioning period of at least three months a reduction in wall thickness can be observed in athletes, but not in HCM.\n\nIdentification of HCM is challenging, when wall thickness is between 13 and 15 mm (the so-called grey-zone of LVH)7. In the last few years, development of new echocardiographic techniques have improved the knowledge of the athlete’s heart and differential diagnosis of physiological and pathological LVH.\n\nIn the athlete’s heart, LV diastolic function is often supranormal, in particular in endurance-trained athletes, when compared with untrained individuals. LV remodelling in athletes is associated with normal or increased myocardial relaxation, as an expression of increased elastic recoil, different from HCM patients, in whom diastolic dysfunction may be the first expression of the disease and may precede the development of LVH8.\n\nIn athletes transmitral E/A ratio is often > 2, with typical low A velocity (late diastole), and this parameter is useful to distinguish this condition from pathological LVH, where E/A ratio is < 1 and E velocity deceleration time is prolonged9.\n\nPulsed tissue Doppler (TDI)-derived early diastolic myocardial velocity (e’) of basal septal and basal lateral wall is increased in athletes. Conversely, HCM is characterised by an e’ reduction in both the hypertrophic septum and the normal thickness of lateral wall10. Lewis et al. suggested that an e’ peak velocity threshold of < 11.5 cm/sec on TDI can be useful to raise suspicion for pathological LVH11 (Figure 1).\n\nAthletes have no regional diastolic dysfunction (e’/a’ < 1), while this is evident in 25% of myocardial segments of HCM patients and in hypertensive patients12.\n\nFinally E/e’ ratio is low in athletes, but increased in HCM patients, it being related with NYHA class and exercise capacity. Reduction of e’ velocity of both septal and lateral annulus is common after ultra-long duration exercises13.\n\nMoreover, pulsed TDI gives additional information regarding myocardial systolic performance at rest, showing normal or supranormal values in athlete’s heart14. In athletes, LVH is combined with normal EF, normal or supranormal stroke volume and systolic peak velocity (s’) > 9 cm/sec, while in pathological LVH (HCM or arterial hypertension) s’ is < 9 cm/sec, with EF normal or high in early stages and reduced in advanced stages15.\n\nThe athlete’s heart can be considered an interesting model of strain variation at different loading conditions, because there is a LV adaptation of at rest and a load dependency of strain measurement. (Figure 2 and Figure 3).\n\nBull’s eye represents in a single image all myocardial regional deformations, from basal, to middle and apical segments.\n\nIn particular, in athletes mild impairment of global longitudinal strain (GLS), lower apical radial strain and lower twisting at rest than in sedentary controls have been observed, together with an increase of basal and middle radial and circumferential strain16,17. Athletes had higher values for transverse, radial, and circumferential strains when compared with HCM18.\n\nWhile conventional echocardiographic parameters often failed to distinguish between endurance (runners) and strength (bodybuilders) athlete’s heart, a speckle tracking echocardiography (STE) analysis showed a different pattern of myocardial deformation in these two groups: while global radial strain (GRS) was similar, GLS was lower in runners and global circumferential strain (GCS) was lower in bodybuilders: correlations were found in runners between GLS and end-diastolic volume (r = 0.46; p < 0.05) and body surface area (r = 0.49; p < 0.05), while in bodybuilders, GCS was closely related to LV mass (r = 0.61; p < 0.01) and systolic blood pressure (r = 0.42; p < 0.05)19.\n\nAnother study used strain rate imaging to distinguish between individuals with hypertensive LVH and those with strength-training athletic LVH, reporting a significant reduction of systolic and diastolic strain and strain rate in hypertensive individuals, but not in athletes: an e′/a′ ratio >1 was found in 100% of a large population of competitive athletes, 90% of subjects had e′ ≥16 cm/s, s′ ≥10 cm/s, and GLS ≤21%20. Moreover, hearts of hypertensive are characterized by reduced GLS, whereas GCS, GRS, and torsion are similar to those of athletes' hearts: the extent of GLS is strongly associated with LV diastolic function, independently of afterload changes and the degree of LVH21.\n\nSantoro et al. stated that LV apical circumferential strain in endurance athletes group was lower than the strength group and control groups (-21.6 ± 4.1% vs. -26.8 ± 7.7%, p < 0.05; vs. -27.8 ± 5.6%, p < 0.01). The endurance group had lower LV twisting (LVT) and untwisting (UTW) than strength group (6.2 ± 0.1° vs. 12.0 ± 0.1°, p < 0.01; -67.3 ± 22.9°/s vs. -122.5 ± 52.8°/s, p < 0.01) and control group (10.0 ± 0.1°, p < 0.01; -103.3 ± 29.3°/s, p < 0.01)22.\n\nFinally, STE showed reduction of longitudinal, circumferential and radial strains and also reduction and delay of peak twisting in triathletes soon after ultralong-duration exercises23.\n\nThree-dimensional (3D) echocardiography offers the ability to improve the diagnostic capability of cardiac ultrasound for evaluating cardiac anatomy, ventricular function, valvular disease and blood flow velocity. This technique is able to quantify LV volume and mass in a fashion which is similar to cardiac magnetic resonance. However, 3D echocardiography is more reproducible, has lower costs and is applicable to a large population of athletes. 3D echocardiography gives more detailed information than two dimensional (2D) echo techniques, providing data on LV remodelling and function; 3D is better in describing morphological features, showing differences in the length and shape of the LV chamber, which are not adequately assessed using 2D technique24.\n\nUsing 3D echocardiography, Caselli et al. showed LV end-diastolic volumes and mass increased in athletes compared to untrained controls; gender and type of sport had the highest impact on LV remodelling. In particular, male gender and endurance disciplines had the highest impacts on LV end-diastolic volume and mass. Body surface area (BSA) was also an important factor on LV remodelling, while age and blood pressure had only minimal effects. Preserved LV systolic function was observed in athletes, with average values similar in athletes and untrained controls25.\n\nDe Castro et al. measured LV remodeling index (LVRI) to describe the pattern of LV remodelling in athletes: athletes' LVRI was similar to that of controls, suggesting that the LV remodeling associated with intensive athletic conditioning does not alter LV geometry. Athlete’s heart has a “symmetric” remodeling pattern, because an increased cavity dimension and volume are accompanied by an increased thickness and mass of the ventricles, in the absence LV systolic dysfunction26.\n\nMoreover, isometric activity in strength sports had the highest effects on LV wall thickness, while isotonic activities as marathons had the highest impact on LV diastolic cavity diameter27. The athlete’s heart is therefore characterized by harmonic LV remodelling, differently from patients with hypertrophic or dilated cardiomyopathy28 (Table 1).\n\nAtrial function may represent an essential part of cardiac function that is sometimes neglected.\n\nD’Andrea et al. investigated whether mechanical dysfunction in the left atrium (LA) is present in patients with either physiological or pathological LVH using two-dimensional strain rate imaging: LA maximum volume was increased but similar between the two groups of patients with LVH. Peak systolic myocardial atrial strain was significantly impaired in patients with pathological LVH compared with controls and athletes. As assessed by multivariate analysis, LV end-diastolic volume/BSA and LV mass in athletes were the only independent factors influencing LA lateral wall peak systolic strain. In contrast, in hypertensive patients, an independent negative association of LA lateral wall peak systolic strain with both LV mass and circumferential end-systolic stress was observed. Moreover, in the overall population of patients with LVH, LA lateral wall systolic strain was an independent predictor of maximum workload during exercise testing29.\n\n\nAthlete’s right heart\n\nIn the recent years, the substantial structural and functional adaptations of the right heart (RH) have been documented, highlighting the complex interplay with the left heart. There is also evolving evidence of acute and chronic cardiac damage, mainly involving the right heart and which may predispose to atrial and ventricular arrhythmias, configuring an exercise-induced cardiomyopathy. Endurance exercise seems to be associated with the greatest extent of cardiac remodelling, involving both LV and right ventricle (RV), while strength training seems to impact minimally on the RV30–33. Moreover, the reversibility of the changes induced by sport after detraining was considered a typical feature of the athlete’s heart, but several studies have showed that structural and functional recovery might be incomplete, in particular for RV changes and this is particularly true in more practiced athletes33.\n\nStandard echocardiography is the first line imaging exam to differentiate athlete’s heart RV remodeling from pathological conditions. The RH clearly participates in the process of enlargement of the athlete’s heart, with an increase in internal diameters and thickness of its free walls. RV shows greater inflow and outflow dimensions in athletes compared with sedentary controls, with no significant difference in the systolic function. D’Andrea et al. documented that RH measures were all significantly greater in highly-trained endurance athletes, compared to age and sex matched strength-trained athletes34,35.\n\nTypical RV characteristics of the athlete’s heart can resemble those found in arrhythmogenic right ventricle cardiomyopathy (ARVC): in ARVC the enlargement of the RV cavity involves both RV inflow and outflow, and may be associated with RV wall segmental morphological and functional abnormalities; in athletes RV enlargement involves only the inflow tract and systolic function is typically normal36. In addition, the inferior vena cava appeared to be dilated in a study involving 58 endurance athletes37.\n\nLV stroke volume and pulmonary artery systolic pressure (PASP) were found to be powerful independent predictors of both RV and right atrial (RA) dimensions.\n\nInteresting changes in the pulmonary vascular haemodynamics of highly trained athletes can be detected at rest. Concerning the PASP values, whose upper limit of normal was 40 mmHg, endurance-trained athletes showed the highest values, compared with strength trained athletes, and LV stroke volume was an independent predictor of PASP38–40.\n\nResting RV global systolic function as measured by fractional area change (FAC) and Tricuspid Annular Plane Systolic Excursion TAPSE seems to be lower in endurance athletes comparing with non-athletic controls. The reduction was more pronounced in the presence of higher RV dilation.\n\n\nNew right ventricular echocardiographic techniques\n\nConcerning the advanced ultrasound technologies, TDI velocity measurements showed that the early-diastolic phase of LV filling was increased, along with a prolonged isometric relaxation time. LV stroke volume was an independent predictor of the early diastolic velocity (Em) and the time of regional isovolumic release (RTm) of RV free walls34.\n\nAs for RV systolic function, both TDI and 2D-strain-derived deformation indexes are reduced at rest in endurance athletes at the RV inlet and mid-free wall level. These changes in RV function at rest are not caused by myocardial damage, in fact there are no increases in NT-proBNP levels among athletes41,42.\n\nGalderisi et al. showed that by combining 3D echo and STE, RV preload exerts its maximal influence on lateral longitudinal fibres (RV lateral longitudinal strain)41.\n\nA recent study by D’Andrea et al. found comparable 2D and 3D RV systolic indexes between endurance athletes and controls. In this setting, a mild reduction in global RV function could be considered a physiological consequence of RV dilation, since an efficient stroke volume will be reached with higher end-diastolic volumes and then at lower ejection fraction. On the other hand, a severe reduction in RV global systolic function should be considered an abnormal finding even among athletes43 (Figure 4).\n\nDuring exercise, increases in both pressures and volumes were greater for the RV, while increases in wall thickness were relatively less than for the LV. As a result, RV wall stress estimates increased 125% during exercise as compared with a modest 14% increase in LV wall stress44.\n\nHowever, echocardiographic estimates of contractility seem to increase proportional to increases in pulmonary artery pressures during intense exercise of short duration45, suggesting that the RV has the contractile reserve to meet exercise demands, at least for a while.\n\nThe RV is more susceptible than the LV to prolonged exercises and is able to induce cardiac fatigue: many studies reported RV dysfunction after long term exercises, as marathons46–50. D’Andrea et al. observed RV dilatation following an ultra-endurance triathlon without changes of LV dimension, by using M-mode, 2D echo and STE (reduction of longitudinal strain about 15% relative to baseline values)51 (Table 2).\n\n\nConclusions\n\nIn the last few years, clinical exercise practice, both for recreational and competitive purposes has been spreading worldwide and an increase in the number of subjects with features of exercise-induced cardiac remodeling can be expected. It is important to distinguish healthy, physiological modifications of the athlete’s heart from pathological conditions such as cardiomyopathies.\n\nCardiac imaging is essential in identifying cardiovascular disease in athletes, but it must be integrated with medical history, symptoms, age, gender, ECG and genetic analyses.\n\nStandard echocardiography has a pivotal role in assessing the athlete’s heart characteristics while the latest developments in ultrasound techniques, such as TDI, 2D strain imaging and 3D echocardiography are important to improve knowledge about physiological and pathological heart remodeling related to sport exercise.", "appendix": "Author contributions\n\n\n\nAntonello D’Andrea and Eduardo Bossone: design of the study and final technical revision.\n\nJuri Radmilovic: writing the final draft of the manuscript and contributed with English language assistance.\n\nRaffaele Calabrò and Maria Giovanna Russo: drafting the section about new technologies and right ventricle.\n\nPio Caso and Maurizio Galderisi: drafting the section about new technologies and left ventricle.\n\nAll authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no funding was involved in supporting this work.\n\n\nReferences\n\nD'Andrea A, La Gerche A, Golia E, et al.: Physiologic and pathophysiologic changes in the right heart in highly trained athletes. Herz. 2015; 40(3): 369–378. PubMed Abstract \| Publisher Full Text\n\nMorganroth J, Maron BJ, Henry WL, et al.: Comparative left ventricular dimensions in trained athletes. Ann Intern Med. 1975; 82(4): 521–524. PubMed Abstract \| Publisher Full Text\n\nGalderisi M, Cardim N, D’Andrea A, et al.: The multi-modality cardiac imaging approach to the Athlete’s heart: an expert consensus of the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging. 2015; 16(4): 353. PubMed Abstract \| Publisher Full Text\n\nPelliccia A, Maron BJ, Spataro A, et al.: The upper limit of physiologic cardiac hypertrophy in highly trained elite athletes. N Engl J Med. 1991; 324(5): 295–301. PubMed Abstract \| Publisher Full Text\n\nMaron BJ: Structural features of the athlete heart as defined by echocardiography. J Am Coll Cardiol. 1986; 7(1): 190–203. PubMed Abstract \| Publisher Full Text\n\nPelliccia A, Culasso F, Di Paolo FM, et al.: Physiologic left ventricular cavity dilatation in elite athletes. Ann Intern Med. 1999; 130(1): 23–31. PubMed Abstract \| Publisher Full Text\n\nCaselli S, Di Paolo FM, Pisicchio C, et al.: Three-dimensional echocardiographic characterization of left ventricular remodeling in Olympic athletes. Am J Cardiol. 2011; 108(1): 141–147. PubMed Abstract \| Publisher Full Text\n\nCaselli S, Maron MS, Urbano-Moral JA, et al.: Differentiating left ventricular hypertrophy in athletes from that in patients with hypertrophic cardiomyopathy. Am J Cardiol. 2014; 114(9): 1383–9. PubMed Abstract \| Publisher Full Text\n\nGeorge KP, Warburton DE, Oxborough D, et al.: Upper limits of physiological cardiac adaptation in ultramarathon runners. J Am Coll Cardiol. 2011; 57(6): 754–5. PubMed Abstract \| Publisher Full Text\n\nSeverino S, Caso P, Galderisi M, et al.: Use of pulsed Doppler tissue imaging to assess regional left ventricular diastolic dysfunction in hypertrophic cardiomyopathy. Am J Cardiol. 1998; 82(11): 1394–8. PubMed Abstract \| Publisher Full Text\n\nLewis JF, Spirito P, Pelliccia A, et al.: Usefulness of Doppler echocardiographic assessment of diastolic filling in distinguishing “athlete’s heart” from hypertrophic cardiomyopathy. Br Heart J. 1992; 68(3): 296–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCardim N, Oliveira AG, Longo S, et al.: Doppler tissue imaging: regional myocardial function in hypertrophic cardiomyopathy and in athlete’s heart. J Am Soc Echocadiogr. 2003; 16(3): 223–32. PubMed Abstract \| Publisher Full Text\n\nCotrim C, Almeida AR, Miranda R, et al.: Stress-induced intraventricular gradients in symptomatic athletes during upright exercise continuous wave Doppler echocardiography. Am J Cardiol. 2010; 106(12): 1808–1812. PubMed Abstract \| Publisher Full Text\n\nZoncu S, Pelliccia A, Mercuro G: Assessment of regional systolic and diastolic wall motion velocities in highly trained athletes by pulsed wave Doppler tissue imaging. J Am Soc Echocardiogr. 2002; 15(9): 900–5. PubMed Abstract \| Publisher Full Text\n\nVinereanu D, Florescu N, Schulthorpe N, et al.: Differentiation between pathologic and physiologic left ventricular hypertrophy by tissue Doppler assessment of long-axis function in patients with hypertrophic cardiomyopathy or systemic hypertension and in athletes. Am J Cardiol. 2001; 88(1): 53–8. PubMed Abstract \| Publisher Full Text\n\nMatsumura Y, Elliott PM, Virdee MS, et al.: Left ventricular diastolic function assessed using Doppler tissue imaging in patients with hypertrophic cardiomyopathy: relation to symptoms and exercise capacity. Heart. 2002; 87(3): 247–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichand V, Lafitte S, Reant P, et al.: An ultrasound speckle tracking (two-dimensional strain) analysis of myocardial deformation in professional soccer players compared with healthy subjects and hypertrophic cardiomyopathy. Am J Cardiol. 2007; 100(1): 128–32. PubMed Abstract \| Publisher Full Text\n\nRichand V, Lafitte S, Reant P, et al.: An ultrasound speckle tracking (two-dimensional strain) analysis of myocardial deformation in professional soccer players compared with healthy subjects and hypertrophic cardiomyopathy. Am J Cardiol. 2007; 100(1): 128–32. PubMed Abstract \| Publisher Full Text\n\nSzauder I, Kovács A, Pavlik G: Comparison of left ventricular mechanics in runners versus bodybuilders using speckle tracking echocardiography. Cardiovasc Ultrasound. 2015; 13(1): 7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaghir M, Areces M, Makan M: Strain rate imaging differentiates hypertensive cardiac hypertrophy from physiologic cardiac hypertrophy (athlete’s heart). J Am Soc Echocardiogr. 2007; 20(2): 151–7. PubMed Abstract \| Publisher Full Text\n\nGalderisi M, Lomoriello VS, Santoro A, et al.: Differences of myocardial systolic deformation and correlates of diastolic function in competitive rowers and young hypertensives: a speckle-tracking echocardiography study. J Am Soc Echocardiogr. 2010; 23(11): 1190–8. PubMed Abstract \| Publisher Full Text\n\nSantoro A, Alvino F, Antonelli G, et al.: Endurance and Strength Athlete's Heart: Analysis of Myocardial Deformation by Speckle Tracking Echocardiography. J Cardiovasc Ultrasound. 2014; 22(4): 196–204. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDouglas PS, O’Toole ML, Hiller WD, et al.: Different effects of prolonged exercise on the right and left ventricles. J Am Coll Cardiol. 1990; 15(1): 64–9. PubMed Abstract \| Publisher Full Text\n\nDe Castro S, Pelliccia A, Caselli S, et al.: Remodelling of the left ventricle in athlete's heart: a three dimensional echocardiographic and magnetic resonance imaging study. Heart. 2006; 92(7): 975–976. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaselli S, Di Paolo FM, Pisicchio C, et al.: Three-dimensional echocardiographic characterization of left ventricular remodeling in Olympic athletes. Am J Cardiol. 2011; 108(1): 141–7. PubMed Abstract \| Publisher Full Text\n\nDe Castro S, Pelliccia A, Caselli S, et al.: Remodelling of the left ventricle in athlete's heart: a three dimensional echocardiographic and magnetic resonance imaging study. Heart. 2006; 92(7): 975–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpirito P, Pelliccia A, Proschan MA, et al.: Morphology of the “athlete's heart” assessed by echocardiography in 947 elite athletes representing 27 sports. Am J Cardiol. 1994; 74(8): 802–806. PubMed Abstract \| Publisher Full Text\n\nDe Castro S, Caselli S, Maron M, et al.: Left ventricular remodelling index (LVRI) in various pathophysiological conditions: a real-time three-dimensional echocardiographic study. Heart. 2007; 93(2): 205–209. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nD'Andrea A, De Corato G, Scarafile R, et al.: Left atrial myocardial function in either physiological or pathological left ventricular hypertrophy: a two-dimensional speckle strain study. Br J Sports Med. 2008; 42(8): 696–702. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Galderisi M, Sciomer S, et al.: Echocardiographic evaluation of the athlete's heart: from morphological adaptations to myocardial function. G Ital Cardiol (Rome). 2009; 10(8): 533–44. PubMed Abstract\n\nLuijkx T, Cramer MJ, Prakken NH, et al.: Sport category is an important determinant of cardiac adaptation: an MRI study. Br J Sports Med. 2012; 46(16): 1119–24. PubMed Abstract \| Publisher Full Text\n\nWernstedt P, Sjöstedt C, Ekman I, et al.: Adaptation of cardiac morphology and function to endurance and strength training. A comparative study using MR imaging and echocardiography in males and females. Scand J Med Sci Sports. 2002; 12(1): 17–25. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Burns AT, Mooney DJ, et al.: Exercise-induced right ventricular dysfunction and structural remodelling in endurance athletes. Eur Heart J. 2012; 33(8): 998–1006. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Caso P, Sarubbi B, et al.: Right ventricular myocardial adaptation to different training protocols in top-level athletes. Echocardiography. 2003; 20(4): 329–36. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Riegler L, Golia E, et al.: Range of right heart measurements in top-level athletes: the training impact. Int J Cardiol. 2013; 164(1): 48–57. PubMed Abstract \| Publisher Full Text\n\nSen-Chowdhry S, Lowe MD, Sporton SC, et al.: Arrhythmogenic right ventricular cardiomyopathy: clinical presentation, diagnosis, and management. Am J Med. 2004; 117(9): 685–95. PubMed Abstract \| Publisher Full Text\n\nGoldhammer E, Mesnick N, Abinader EG, et al.: Dilated inferior vena cava: a common echocardiographic finding in highly trained elite athletes. J Am Soc Echocardiogr. 1999; 12(11): 988–93. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Naeije R, D'Alto M, et al.: Range in pulmonary artery systolic pressure among highly trained athletes. Chest. 2011; 139(4): 788–94. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Naeije R, D'Alto M, et al.: Range in pulmonary artery systolic pressure among highly trained athletes. Chest. 2011; 139(4): 788–94. PubMed Abstract \| Publisher Full Text\n\nBossone E, Rubenfire M, Bach DS, et al.: Range of tricuspid regurgitation velocity at rest and during exercise in normal adult men: implications for the diagnosis of pulmonary hypertension. J Am Coll Cardiol. 1999; 33(6): 1662–6. PubMed Abstract \| Publisher Full Text\n\nEsposito R, Galderisi M, Schiano-Lomoriello V, et al.: Nonsymmetric myocardial contribution to supranormal right ventricular function in the athlete’s heart: combined assessment by Speckle Tracking and real time three-dimensional echocardiography. Echocardiography. 2014; 31(8): 996–1004. PubMed Abstract \| Publisher Full Text\n\nKing G, Almuntaser I, Murphy RT, et al.: Reduced right ventricular myocardial strain in the elite athlete may not be a consequence of myocardial damage. \"Cream masquerades as skimmed milk\". Echocardiography. 2013; 30(8): 929–35. PubMed Abstract \| Publisher Full Text\n\nD'Andrea A, Riegler L, Morra S, et al.: Right ventricular morphology and function in top-level athletes: a three-dimensional echocardiographic study. J Am Soc Echocardiogr. 2012; 25(12): 1268–76. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Heidbuchel H, Burns AT, et al.: Disproportionate exercise load and remodeling of the athlete's right ventricle. Med Sci Sports Exerc. 2011; 43(6): 974–81. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Burns AT, D'Hooge J, et al.: Exercise strain rate imaging demonstrates normal right ventricular contractile reserve and clarifies ambiguous resting measures in endurance athletes. J Am Soc Echocardiogr. 2012; 25(3): 253–262.e1. PubMed Abstract \| Publisher Full Text\n\nTrivax JE, Franklin BA, Goldstein JA, et al.: Acute cardiac effects of marathon running. J Appl Physiol (1985). 2010; 108(5): 1148–53. PubMed Abstract \| Publisher Full Text\n\nDouglas PS, O'Toole ML, Hiller WD, et al.: Different effects of prolonged exercise on the right and left ventricles. J Am Coll Cardiol. 1990; 15(1): 64–9. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Connelly KA, Mooney DJ, et al.: Biochemical and functional abnormalities of left and right ventricular function after ultra-endurance exercise. Heart. 2008; 94(7): 860–6. PubMed Abstract \| Publisher Full Text\n\nOxborough D, Shave R, Warburton D, et al.: Dilatation and dysfunction of the right ventricle immediately after ultraendurance exercise: exploratory insights from conventional two-dimensional and speckle tracking echocardiography. Circ Cardiovasc Imaging. 2011; 4(3): 253–63. PubMed Abstract \| Publisher Full Text\n\nNeilan TG, Januzzi JL, Lee-Lewandrowski E, et al.: Myocardial injury and ventricular dysfunction related to training levels among nonelite participants in the Boston marathon. Circulation. 2006; 114(22): 2325–33. PubMed Abstract \| Publisher Full Text\n\nD’Andrea A, Caso P, Bossone E, et al.: Right ventricular myocardial involvement in either physiological or pathological left ventricular hypertrophy: an ultrasound speckle-tracking two-dimensional strain analysis. Eur J Echocardiogr. 2010; 11(6): 492–500. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9757", "date": "03 Aug 2015", "name": "Jarosław Kasprzak", "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 paper is a current, high quality review of the novel echocardiographic approaches to quantification of myocardial function in the context of training-induced remodeling.The authors provide a fair overview of methods used for discriminating right and left ventricular pathology which is important for sudden death risk stratification in this subset.The review is of interest for the readership and I suggest to accept it on \"as is\" basis.", "responses": [] }, { "id": "10561", "date": "05 Oct 2015", "name": "Pankaj Garg", "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 very well structured contemporary review paper on role of echocardiography in subjects with athletic heart. Overall, this will add value to this journal. Minor suggestions: In the paragraph of LA function, the introduction line needs a reference. I suggest the following: Leischik et al. (2015). Please clearly define that strain will add incremental value to current routine assessment with TDI in finding out individuals who may develop adverse myocardial fibrosis. Add the following references: Hoffman et al.(2014), O'Keefe et al. (2012).", "responses": [] } ]	1	https://f1000research.com/articles/4-289
https://f1000research.com/articles/4-59/v1	05 Mar 15	{ "type": "Opinion Article", "title": "Digital teaching tools and global learning communities", "authors": [ "Mary Williams", "Patti Lockhart", "Cathie Martin", "Patti Lockhart", "Cathie Martin" ], "abstract": "In 2009, we started a project to support the teaching and learning of university-level plant sciences, called Teaching Tools in Plant Biology. Articles in this series are published by the plant science journal, The Plant Cell (published by the American Society of Plant Biologists). Five years on, we investigated how the published materials are being used through an analysis of the Google Analytics pageviews distribution and through a user survey. Our results suggest that this project has had a broad, global impact in supporting higher education, and also that the materials are used differently by individuals in terms of their role (instructor, independent learner, student) and geographical location. We also report on our ongoing efforts to develop a global learning community that encourages discussion and resource sharing.", "keywords": [ "Plant biology", "science", "education online", "education digital", "education global", "education" ], "content": "Introduction\n\nMore than 20 years ago, the first graphical web browser Netscape Navigator was released, and the world changed forever (Mosaic Communications press release, 1994). Widespread access to the repository of information stored on computers across the globe has changed the way we teach, learn, and communicate. The Internet opened the door to global educational tools, from Wikipedia to massively open online courses (aka MOOCs), and changed the way that students access and engage with information. The opportunities afforded by this international connectedness are still being developed and explored. We describe an innovative, ongoing project to create and disseminate university-level educational materials in plant science by way of their publication in a scholarly journal, we report the new ways these resources are being used across the globe, and we propose new opportunities for enhanced engagement.\n\n\nWhy is plant science important? Food production is a global challenge\n\nWe face many challenges in the 21st century as a consequence of rising population, rising affluence, and energy requirements and global climate changes. It is widely acknowledged that food production must increase by 50% or more by 2050 (Ray et al., 2013), and at the same time it is vital to preserve natural ecosystems and employ more sustainable food production methods. Food production is a global problem that must be addressed locally, and plant scientists, horticulturalists, and agronomists with local knowledge will be at the forefront of these efforts.\n\nAs a first step to addressing the need for increased recruitment and training of plant scientists, the premier plant science journal, The Plant Cell (published by the American Society of Plant Biologists) in 2009 began publishing online educational tools for use in higher education. These materials are presented as a series of articles called “Teaching Tools in Plant Biology”. To date we have published 30 Teaching Tools; each includes a set of image-rich PowerPoint slides from which educators can select parts or complete teaching materials, a review-style article written for advanced undergraduates, recommended reading lists that span recent research and review articles as well as historical reports, and teaching guides with questions for discussion and assessment. Each Teaching Tool is peer-reviewed and regularly updated with content and references, features that were impossible before online publishing.\n\nThe overarching goal of this project is to support plant scientists in their teaching and learning. The targeted users of these resources are the readers of The Plant Cell (access to the articles requires a personal or institutional subscription to the journal) and their students in upper-level undergraduate or graduate courses. Besides summarizing current plant science, specific objectives are to support a variety of teaching and learning styles, to highlight the relevance of plant research to global issues, and to help bridge the gap between textbook-based learning and learning from the primary literature that occurs as students transition from undergraduate to graduate education.\n\nAs online resources, these are used widely across the globe and have been regularly accessed from more than 100 countries. In 2014, the collection received more than 250,000 pageviews. The countries with the greatest number of pageviews were the United States, China, India, Germany, Japan, S. Korea, the United Kingdom, Spain, Canada, and France (Figure 1); each of these countries falls within the top 20 countries in terms of number of Internet users (Internet World Stats, 2014). Although we can track download numbers and sources, we lacked data about who is reading these articles and how they are being used. Therefore, four years into the project, we conducted a survey to learn more about how these resources are used in different countries by instructors, students, and independent learners. A link to the survey was embedded on the download page for the Teaching Tools resources. To encourage survey completion, we provided a $100 Amazon voucher to one randomly selected survey participant. We received nearly 300 completed surveys representing input from 50 countries. About half of our responders were instructors in a course. The remainder were students in a formal course or independent learners not enrolled in a formal course.\n\nMore than 100 countries are represented in the more than 250,000 pageviews accrued during 2014. Since 2011, the US, China and India have been the countries with the largest number of pageviews.\n\n\nTeaching Tools are used in many different ways\n\nWhen we initiated this series, we envisioned the articles being used as a supplement to a textbook, and intentionally did not tie them to the content of any of the several textbooks used to teach advanced plant biology or plant physiology. Through our survey, we learned that in some cases both instructors and students have poor access to textbooks (due to financial or language issues) and many are using the materials from The Plant Cell as a substitute for a textbook (Figure 2). A particularly high proportion of survey respondents from least-developed countries who access the materials through Research4Life/AGORA identify as having poor access to textbooks. Now that we recognize that these articles are in some cases being used in lieu of a textbook, we are including more basic as well as advanced coverage of each topic.\n\nThis was particularly a concern for respondents from least-developed countries (shown in purple) from which access Teaching Tools in Plant Biology is available through AGORA (Access to Global Online Research in Agriculture).\n\nOur survey also reinforced the fact that teaching styles and approaches are quite varied. Although in the US there is wide support from learned societies, funding agencies, and many instructors for an emphasis on student-centered, project-based learning, this approach is not universally embraced, particularly outside the US (AAAS, 2011; Freeman et al., 2014; Tanner, 2009). For example, in China higher education styles tend to be more traditional for a variety of social, cultural, and political reasons (Lee, 2004; Thompson & Ku, 2005; Zhang, 2007; Zhang, 2010). In line with this, 80% of the survey respondents in China reported that the materials are used in courses that are primarily lecture based, whereas in Asia (without China) and Latin America, the largest category of course types was described as “primarily student centered” (54% and 55%) (Figure 3). Respondents from the US or Canada mainly identified their courses “equal parts lecture and non-lecture”.\n\nThis question was only asked to instructors, who selected one category out of the four options indicated.\n\nOur survey responses also indicated that the Teaching Tools are being used differently by students, independent learners, and instructors (Figure 4). Course instructors most highly valued the PowerPoint files as sources of images for preparing lectures, students found the Lecture Notes important, and independent learners rated the references as the most important component (independent learner respondents included undergraduates, graduate students, postdocs, lab heads, staff scientists, teaching staff, and other, a category that included librarian, communicator, and entrepreneur).\n\nInstructors particularly value the PowerPoint files as sources of images for teaching, the students rank the Lecture Notes as most important, and the independent learners rank the references as most important.\n\nAlthough English is the lingua franca of science, about half of the people we surveyed teach or learn in a different language, raising the question of whether we should present these materials in multiple languages. As a first step, we have coordinated the production of translations of one of these lessons, called “Why Study Plants?” into 14 languages (http://www.plantcell.org/site/teachingtools/TTPB1.xhtml), and we are exploring the possibility of translating additional lessons in collaboration with the Global Plant Council.\n\n\nHarnessing social media to develop a global learning community\n\nWe have been exploring the use of social media as a way to connect the geographically widespread users of Teaching Tools in Plant Biology. Scientists are generally considered slow adopters of social media, but there is growing awareness that these tools can be used in a professional context and support professional development (Bik & Goldstein, 2013; Osterrieder, 2013). Since the earliest days of online education, educators have stressed the importance of building community and trust (Brown, 2001; Rovai, 2002); trust is based on positive interactions, whether online or in person. Social media provide the opportunities for community building that were lacking in the earliest iterations of distance learning and have proven to be an effective medium to support interactions among globally distributed professional communities (Claussen et al., 2013; Davies & Glasser, 2014; Evans, 2014; Kietzmann et al., 2011). Our efforts so far have focused on three platforms: Twitter, Facebook, and ScoopIt. As described by Van Noorden (2014), Twitter and Facebook are perceived very differently, with far more of the scientists they surveyed indicating that professionally they prefer Twitter to Facebook. Our data indicate that the geographic distribution of people “following” or “liking” the Teaching Tools in Plant Biology feed differs between the two platforms (Figure 5). Notably, more than two thirds of Twitter followers are from the UK, the US, Canada, and Australia, a fact that emphasizes the importance of language for this type of interaction. By contrast, seven out of eight of the largest groups of followers of the Facebook feed are from countries whose primary languages are not English.\n\nThe geographically broad distribution of followers of both the Teaching Tools articles and the social media feeds provides an interesting challenge for the content curators and creators, which is to be vigilant about the tone and implications of the materials shared. Many companies and organizations have developed social media guidelines for their staff to follow (http://socialmediagovernance.com/policies/), and these are particularly relevant when considering diverse audiences. It is possible for a tweet that is meant to be ironic or humorous to be wildly misinterpreted, and there are plenty of well-publicized social media mistakes to serve as reminders of the importance of the rule “think before you tweet”. Trust is one of the most important factors for the success of online learning, so providing a positive online presence has been one of our goals (Wang, 2014).\n\nA successful online resource has to reflect the values and concerns of a broadly multicultural community (Liu et al., 2010). To further support the global community of plant scientists, we endeavour to use examples and case studies highlighting agricultural challenges from geographically diverse regions. Recent Teaching Tools describe the challenges of seawater incursion in Bangladesh, groundwater limitations in northern China, and phosphate-deprived soils of Chile and South Western Australia. Contributors to our social media communities are excellent sources of information about regional concerns and perspectives.\n\nFinally, it is clear that our social media platforms are helping to connect the core materials to our audience. As an example, on December 30th, 2014, we posted a new Teaching Tool. Forty-eight hours after posting a notice of this new tool on Facebook, the notice had been seen by >3300 people, shared 30 times, and been “liked” 56 times, with more than 500 pageviews of the new article.\n\n\nMoving forward – Supporting interaction and engagement\n\nThrough social media, we are creating an environment in which professors, postdocs, and students can share and discuss ideas about current research in plant biology. The platforms we are using now arrange the information chronologically, but later this year we are launching a custom-built platform that will allow the discussions to be archived by theme and topic as well. We are designing the new platform to encourage direct peer-to-peer sharing of teaching tips and strategies. For example, many users of Teaching Tools ask us to suggest videos or animations to accompany the topics, and we will include a simple user interface to support uploading links to such materials. We envision that when an instructor or student is ready to learn “what’s new in photosynthesis”, he or she can find not only the relevant Teaching Tool, but also links to more recent articles, animations, lesson plans, and teaching ideas contributed from the community. A similar model for introductory biology topics was recently unveiled (CourseSource.org).\n\n\nDoes one size fit all? Should it?\n\nOur own experiences and several other studies suggest that there are many different ways that teachers and learners can use online educational materials, both in and out of the formal classroom environment. Social and cultural issues, centralized control of curriculum and exams, access to textbooks, and fluency in English all affect the learning experience. Support has been strong for the approach we have taken, which is to make available a diversity of materials suitable for teachers and learners to use as they will. Our strategy has been to create and curate high-quality, authoritative content that highlights and interprets the cutting-edge of plant science, and we are increasingly providing an opportunity for educators and learners to share in the development of these resources. We are confident that through Teaching Tools in Plant Biology we are making a significant contribution to developing the next generation of educational tools with global outreach.", "appendix": "Author contributions\n\n\n\nCM and MW conceived the project. PL and MW analyzed the pageviews data. MW designed and analyzed the user survey. MW and CM prepared the first draft of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nMary Williams and Patti Lockhart are employees of the American Society of Plant Biologists (publisher of The Plant Cell), and Cathie Martin is former Editor in Chief of The Plant Cell.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe are grateful to Nancy Winchester and Crispin Taylor for comments on the manuscript.\n\n\nReferences\n\nAmerican Association for the Advancement of Science: Vision and Change in Undergraduate Biology Education: A Call to Action. 2011. Reference Source\n\nBik HM, Goldstein MC: An introduction to social media for scientists. PLoS Biol. 2013; 11(4): e1001535. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrown RE: The process of community-building in distance-learning classes. J Asynchron Learning Networks. 2001; 5: 18–35.\n\nClaussen JE, Cooney PB, Defilippi JM, et al.: Science communication in a digital age: Social media and the American Fisheries Society. Fisheries. 2013; 38(8): 359–362. Publisher Full Text\n\nDavies BJ, Glasser NF: Analysis of www.AntarcticGlaciers.org as a tool for online science communication. J Glaciol. 2014; 60(220): 399–406. Publisher Full Text\n\nEvans P: Open online spaces of professional learning: Context, personalisation and facilitation. TechTrends. 2014; 59(1): 31–36. Publisher Full Text\n\nFreeman S, Eddy SL, McDonough M, et al.: Active learning increases student performance in science, engineering, and mathematics. Proc Natl Acad Sci. 2014; 111(23): 8410–8415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInternet World Stats: Internet Users by Country. 2014. Reference Source\n\nKietzmann JH, Hermkens K, McCarthy IP, et al.: Social media? Get serious! Understanding the functional building blocks of social media. Business Horizons. 2011; 54(3): 241–251. Publisher Full Text\n\nLee D: Web-based instruction in China: Cultural and pedagogical implications and challenges. Educ Technol Res Devel. 2004; 52(1): 101–105. Publisher Full Text\n\nLiu X, Liu S, Lee SH, et al.: Cultural differences in online learning: International student perceptions. Educ Technol Soc. 2010; 13(3): 177–188. Reference Source\n\nMosaic Communications Press Release. 1994. Reference Source\n\nOsterrieder A: The value and use of social media as communication tool in the plant sciences. Plant Methods. 2013; 9(1): 26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRay DK, Mueller ND, West PC, et al.: Yield trends are insufficient to double global crop production by 2050. PLoS One. 2013; 8(6): e66428. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRovai AP: Building sense of community at a distance. Int Rev Res Open Distrib Learning. 2002; 3(1): 1–16. Reference Source\n\nTanner KD: Talking to learn: Why biology students should be talking in classrooms and how to make it happen. CBE Life Sci Edu. 2009; 8(2): 89–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThompson L, Ku HY: Chinese graduate students' experiences and attitudes toward online learning. Educ Media Int. 2005; 42(1): 33–47. Publisher Full Text\n\nVan Noorden R: Online collaboration: Scientists and the social network. Nature. 2014; 512(7513): 126–129. PubMed Abstract \| Publisher Full Text\n\nWang YD: Building student trust in online learning environments. Distance Education. 2014; 35(3): 345–359. Publisher Full Text\n\nZhang LJ: A cultural look at information and communication technologies in Eastern education. Educ Technol Res Devel. 2007; 55(3): 301–314. Publisher Full Text\n\nZhang J: Technology-supported learning innovation in cultural contexts. Educ Technol Res Devel. 2010; 58(2): 229–243. Publisher Full Text" }	[ { "id": "7863", "date": "07 Apr 2015", "name": "Claire Hemingway", "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 present findings on users of a suite of peer-reviewed digital educational materials available to subscribers of The Plant Cell and describe the efforts to develop this global community. This review of the Teaching Tools in Plant Biology’s user community is timely and important for several reasons. The importance of this article lies in part on the very fact of the Teaching Tools in Plant Biology’s existence. Launching the delivery of university-level digital teaching and learning materials with this high impact journal was a bold venture. Scientific societies that publish journals are faced with altered publishing economies as the landscape shifts from subscription based to open access models. At the same time, most scientific societies are also seeking to meet their education mission in ways that have significant impact and potential to be sustained. The approach to publish education articles with a society’s flagship journal remains relatively rare today. Thus, The Plant Cell and the American Society of Plant Biologists deserve recognition for initiating this practice in 2009. As an opinion piece, the authors present a compelling argument that the Teaching Tools in Plant Biology are needed resources. The website analytics document the global reach of the teaching and learning materials. But the major contribution of this article is the user survey results, which shows the richer picture of who uses the resources, in what ways, and for what reasons. While the authors report that about half of the 300 survey respondents were instructors in a course and half were students enrolled in a course or free-choice learners, more information about the demographics of the respondents would have been useful. The tantalizing finding about the influence of institutional, political, and cultural practices as well as the instructor preferences on material use deserves additional attention as the project moves towards its next phase of harnessing social media to develop a global learning community. My primary reservation with this opinion piece is the lack of discussion of digital learning resources, and the trend over the years to shift away from digital repositories of resources to communities of practice engaged around the resources. The article frames the introduction in the general context of the radical changes the internet has brought to the ways educational tools are accessed and used. Then in discussing the future directions, the authors briefly mention the open-access, peer-reviewed journal of evidence-based curricula available at CourseSource. Although I would have liked the authors to present a view of how their materials fit within the constellation of digital learning resources available today, this opinion piece and the Teaching Tools in Plant Biology are a valuable contribution.", "responses": [] }, { "id": "8187", "date": "08 Apr 2015", "name": "Mirza Hasanuzzaman", "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 nice piece of work which examines how the produced materials are being used by different communities. The authors presented the distribution of the followers of Teaching Tools in Plant Biology on social networks. But in some countries some networks (e.g. Facebook in China) may not be available. How have the authors considered this?", "responses": [] } ]	1	https://f1000research.com/articles/4-59
https://f1000research.com/articles/4-50/v1	20 Feb 15	{ "type": "Software Tool Article", "title": "ViennaNGS: A toolbox for building efficient next- generation sequencing analysis pipelines", "authors": [ "Michael T. Wolfinger", "Jörg Fallmann", "Florian Eggenhofer", "Fabian Amman", "Jörg Fallmann", "Florian Eggenhofer", "Fabian Amman" ], "abstract": "Recent achievements in next-generation sequencing (NGS) technologies lead to a high demand for reuseable software components to easily compile customized analysis workflows for big genomics data. We present ViennaNGS, an integrated collection of Perl modules focused on building efficient pipelines for NGS data processing. It comes with functionality for extracting and converting features from common NGS file formats, computation and evaluation of read mapping statistics, as well as normalization of RNA abundance. Moreover, ViennaNGS provides software components for identification and characterization of splice junctions from RNA-seq data, parsing and condensing sequence motif data, automated construction of Assembly and Track Hubs for the UCSC genome browser, as well as wrapper routines for a set of commonly used NGS command line tools.", "keywords": [ "Perl", "next generation sequencing", "RNA-seq", "read mapping", "pipelines" ], "content": "Introduction\n\nNext-generation sequencing (NGS) technologies have influenced both our understanding of genomic landscapes as well as our attitude towards handling big biological data. Emerging functional genomics methods based on high-throughput sequencing allow investigation of highly specialized and complex scientific questions, which continuously poses challenges in the design of analysis strategies. Moreover, the demand for efficient data analysis methods has dramatically increased. While a typical NGS analysis workflow is built on a cascade of routine tasks, individual steps are often specific for a certain assay, e.g. depend on a particular sequencing protocol.\n\nA set of NGS analysis pipelines are available for general1,2, and specialized assays such as de-novo motif discovery3. While these tools mostly cover the elementary steps of an analysis workflow, they often represent custom-tailored solutions that lack flexibility. Web-based approaches like Galaxy4 cover a wide portfolio of available applications, however they do not offer enough room for power users who are used to the benefits of the command line.\n\nThe recently published HTSeq framework5 as well as the biotoolbox package provide library modules for processing high-throughput data. While both packages implement NGS analysis functionality in a coherent manner, we encountered use cases that were not covered by these tools.\n\n\nMotivation\n\nThe motivation for this contribution emerged in the course of the research consortium “RNA regulation of the transcriptome” (Austrian Science Fund project F43), which brings together more than a dozen experimental groups with various thematic backgrounds. In the line of this project it turned out that complex tasks in NGS analysis could easily be automated, whereas linking individual steps was very labour-intensive. As such, it became apparent that there is a strong need for modular and reusable software components that can efficiently be assembled into different full-fledged NGS analysis pipelines.\n\nWe present ViennaNGS, a Perl distribution that integrates high-level routines and wrapper functions for common NGS processing tasks. ViennaNGS is not an established pipeline per se, it rather provides tools and functionality for the development of NGS pipelines. It comes with a set of utility scripts that serve as reference implementation for most library functions and can readily be applied for specific tasks or integrated as-is into custom pipelines. Moreover, we provide extensive documentation, including a dedicated tutorial that showcases core features of the software and discusses common application scenarios.\n\nDevelopment of the ViennaNGS suite was triggered by two driving forces. On the one hand we wanted to return to the open source community our own contribution, which itself is heavily based and dependent on open source software. On the other hand, beside “open science” we advocate for the concept of “reproducible science”6. Unfortunately, and to some extent surprising, bioinformatics analyses are often not fully reproducible due to inaccessibility of tools (keyword “in-house script”) or software versions used. It is therefore essential to ensure the entire chain of reproducibility from data generation to interpretation in the analysis of biological data.\n\n\nMethods\n\nThe major design consideration for the ViennaNGS toolbox was to make available modular and reuseable code for NGS processing in a popular scripting language. We therefore implemented thematically related functionality in different Perl modules under the Bio namespace (Figure 1), partly building on BioPerl7 and the Moose object framework. Our focus is on consistent versioning, facilitated through Github hosting. In addition, ViennaNGS releases are available via the Comprehensive Perl Architecture Network (CPAN), thereby enabling users to get back to previous versions at any time in order to reenact conclusions drawn from shared biological data.\n\nCore modules can be combined in a flexible manner to address individual analysis objectives and experimental setup.\n\nViennaNGS has been designed to close gaps in established analysis workflows by covering a wide range of processing steps from raw data to data visualization. In the following we introduce individual ViennaNGS components and describe their main functionality.\n\nOnce mapped to a reference genome, NGS data is typically stored in the widely used SAM/BAM file format. BAM is a binary format, which can easily be converted into text-based SAM format via samtools8 for downstream analysis. However, modern NGS assays produce hundreds of millions of reads per sample, hence SAM files tend to become excessively large and can have a size of several hundred gigabytes. Given that storage resources are always limited, strategies to efficiently retrieve mapping information from BAM format are an asset. To accomodate that, we provide functionality for querying global mapping statistics and extracting specific alignment information from BAM files directly.\n\nViennaNGS::BamStat extracts both qualitative and quantitative information from BAM files, i.e. the amount of total alignments, aligned reads, as well as uniquely and multi mapped reads. Numbers are reported individually for single-end reads, paired-end fragments and pairs missing a mate. Quality-wise ViennaNGS::BamStat collects data on edit distance in the alignments, fraction of clipped bases, fraction of matched bases, and quality scores for entire alignments. Subsequently, ViennaNGS::BamStatSummary compares different samples in BAM format and illustrates results graphically. Summary information is made available in CSV format to facilitate downstream processing.\n\nEfficient filtering of BAM files is among the most common tasks in NGS analysis pipelines. Building on the BioSamTools distribution, ViennaNGS::Bam provides a set of convenience routines for rapid manipulation of BAM files, including filters for unique and multiple alignments as well as functionality for splitting BAM files by strand, thereby creating two strand-specific BAM files. Results can optionally be converted to BedGraph or BigWig formats for visualization purposes.\n\nProper handling of genomic intervals is essential for NGS analysis pipelines. Several feature annotation formats have gained acceptance in the scientific community, including BED, GTF, GFF, etc., each having its particular benefits and drawbacks. While annotation for a certain organism is often only available in a specific format, interconversion among these formats can be regarded a routine task, and a pipeline should be capable of processing as many formats as possible.\n\nWe address this issue at different levels. On the one hand, we provide ViennaNGS::AnnoC, a lightweight annotation converter for non-spliced genomic intervals, which can be regarded a simple yet powerful solution for conversion of bacterial annotation data. On the other hand we have developed an abstract representation of genomic features via generic Moose-based classes, which provide functionality for efficient manipulation of BED4, BED6, BED12 and GTF/GFF elements, respectively, and allow for BED format conversion facilitated by ViennaNGS::Bed. ViennaNGS::MinimalFeature represents an elementary genomic interval, characterized by chromosome, start, end and strand. ViennaNGS::Feature extends ViennaNGS::MinimalFeature by two attributes, name and score, thereby creating a representation of a single BED6 element. ViennaNGS::FeatureChain pools a set of ViennaNGS::Feature objects via an array reference. All intervals of interest can be covered by a ViennaNGS::FeatureLine object, which holds a hash of references to ViennaNGS::FeatureChain objects (Figure 2).\n\nThis framework can handle annotation data by providing abstract data representations of genomic intervals such as exons, introns, splice junctions etc. It allows for efficient description and manipulation of genomic features up to the level of transcripts (Figure 3). Conversely, it is highly generic and can be extended to hierachically higher levels such as genes composed of different transcript isoforms or clusters of paralogous genes.\n\nSimple intervals (“features”) are characterized by Bio::ViennaNGS::Feature objects (bottom box). At the next level, Bio::ViennaNGS::FeatureChain bundles these, thereby maintaining individual annotation chains for e.g. UTRs, exons, introns, splice junctions, etc. (middle box). The topmost level is given by Bio::ViennaNGS::FeatureLine objects, representing individual transcripts.\n\nAnother cornerstone of NGS analysis pipelines is graphical representation of mapped sequencing data. In this context a standard application is visualization of Chip-seq peaks or RNA-seq coverage profiles. The latter are typically encoded in Wiggle format, or its indexed binary variant, BigWig, which can readily be displayed within a genome browser. In the same line, genomic annotation and intervals are often made available in BigBed format, an indexed binary version of BED. ViennaNGS::Util comes with wrapper routines for automated conversion from common formats like BAM to BigWig or BED to BigBed via third-party utilities9. In addition, we have implemented interfaces for a selection of BEDtools10 components as well as a collection of auxiliary routines.\n\nThe UCSC genome browser allows to display potentially large genomic data sets, that are hosted at Web-accessible locations by means of Track Hubs11. On a more general basis this even works for custom organisms that are not supported by default through the UCSC genome browser, via Assembly Hubs. A typical use case is visualization of genomic annotation, RNA-seq coverage profiles and Chip-seq peaks for Arabidopsis thaliana (which is not available through the generic UCSC browser) via a locally hosted Assembly Hub. ViennaNGS::UCSC provides all relevant routines for automatic construction of Assembly and Track Hubs from genomic sequence and/or annotation. Besides automated Assembly and Track Hub generation, we support deployment of custom organism databases in local mirrors of the UCSC genome browser.\n\nRNA-seq has become a standard approach for gene and transcript quantification by means of measuring the relative amount of RNA present in a certain sample or under a specific condition, thus ideally providing a good estimate for the relative molar concentration of RNA species. Simple “count-based” quantification models assume that the total number of reads mapping to a region can be used as a proxy for RNA abundance12. A good measure for transcript abundance is ideally as closely proportional to the relative molar concentration of a RNA species as possible. Various measures have been proposed, one of the most prominent being RPKM (reads per kilobase per million). It accounts for different transcript lengths and sequencing depth by normalizing by the number of reads in a specific sample, divided by 106. It has, however, been shown that RPKM is not appropriate for measuring the relative molar concentration of a RNA species due to normalization by the total number of reads13,14.\n\nAlternative measures that overcome this shortcoming have been suggested, e.g. TPM (transcript per million) (Equation 1). Here, rather than normalizing by the total number of mapped reads, a proxy for the total number of transcript samples considering the sequencing reads per gene rg is used for normalization (Equation 2). The variable rl is the read length and flg the feature length of a gene region g. Consequently, T can be computed by summing over the set of all genes G.\n\n\n\n\n\nWe provide routines for the computation of TPM values for genomic intervals from raw read counts within ViennaNGS::Expression.\n\nViennaNGS::SpliceJunc addresses a more specific problem, namely characterization of splice junctions which is becoming increasingly relevant for understanding alternative splicing. This module provides code for identification and characterization of splice junctions from short read mappers. It can detect novel splice junctions in RNA-seq data and generate visualization files. While we have focused on processing the output of segemehl15,16, the module can easily be extended for other splice-aware split read mappers.\n\nThe ViennaNGS suite comes with extensive documentation based on Perl’s POD system, thereby providing a single documentation base which is available through different channels, e.g. on the command line via the perldoc utility or on the Web via CPAN. Moreover, we provide ViennaNGS::Tutorial to guide prospective users through the development of basic NGS analysis pipelines. The tutorial is split into different chapters, each covering a common use case in NGS analysis and describing a possible solution.\n\nThe ViennaNGS suite comes with a collection of complementary executable Perl scripts for accomplishing routine tasks often required in NGS data processing. These command line utilities serve as reference implementations of the routines implemented in the library and can readily be used for atomic tasks in NGS data processing. Table 1 lists the utilities and gives a short description of their functionality.\n\nWhile some of these scripts are re-implementations of functionality available elsewhere, they have been developed primarily as reference implementation of the library routines to help prospective ViennaNGS users getting started quickly with the development of custom pipelines.\n\n\nDiscussion\n\nViennaNGS is a comprehensive software library for rapid development of custom NGS analysis pipelines. We have successfully applied its components in the course of an ongoing, large scale collaboration project focusing on RNA regulation. It has been used with different genomics assays in a wide range of biological systems, including human, plants and bacteria. While we have primarily applied ViennaNGS in combination with the short read aligner segemehl15,16, it has also been used with Tophat17 output very recently in a large scale transcriptome study of Ebola and Marburg virus infection in human and bat cells (Hölzer et al., unpublished data). Moreover, ViennaNGS will be used for automated UCSC genome browser integration in an upcoming version of TSSAR18, a recently published approach for characterization of transcription start sites from dRNA-seq data.\n\nViennaNGS is actively developed and its functionality is constantly extended. In this line, we encourage the scientific community to contribute patches and novel features.\n\n\nData availability\n\nInput data for the ViennaNGS tutorial is available from http://rna.tbi.univie.ac.at/ViennaNGS\n\n\nSoftware availability\n\nThe ViennaNGS distribution is available through the Comprehensive Perl Architecture Network (CPAN) at and GitHub.\n\nhttp://search.cpan.org/dist/Bio-ViennaNGS\n\nhttps://github.com/mtw/Bio-ViennaNGS\n\nhttp://dx.doi.org/10.5281/zenodo.15088\n\nThe Perl 5 License", "appendix": "Author contributions\n\n\n\nMTW, JF, FE, FA designed and implemented the software. MTW and FA wrote the manuscript. All authors approved the final manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by the Austrian Science Fund (FWF projects F43 to MTW, FA and FE) and the Research Platform “Decoding mRNA decay in inflammation” by the University of Vienna to JF.\n\n\nAcknowledgments\n\nA preprint version of this article can be found on BioRxiv: http://dx.doi.org/10.1101/013011\n\n\nReferences\n\nFörstner KU, Vogel J, Sharma CM: READemption-a tool for the computational analysis of deep-sequencing-based transcriptome data. Bioinformatics. 2014; 30(23): 3421–3. PubMed Abstract \| Publisher Full Text\n\nBreese MR, Liu Y: NGSUtils: a software suite for analyzing and manipulating next-generation sequencing datasets. Bioinformatics. 2013; 29(4): 494–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeinz S, Benner C, Spann N, et al.: Simple Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for Macrophage and B Cell Identities. Mol Cell. 2010; 38(4): 576–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoecks J, Nekrutenko A, Taylor J, et al.: Galaxy: a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences. Genome Biol. 2010; 11(8): R86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnders S, Pyl PT, Huber W: HTSeq-a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015; 31(2): 166–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStodden V, Leisch F, Peng RD: Implementing Reproducible Research. CRC Press, 2014. Reference Source\n\nStajich JE, Block D, Boulez K, et al.: The Bioperl toolkit: Perl modules for the life sciences. Genome Res. 2002; 12(10): 1611–8. 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\nKent WJ, Zweig AS, Barber G, et al.: BigWig and BigBed: enabling browsing of large distributed datasets. Bioinformatics. 2010; 26(17): 2204–7. 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): 841–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRaney BJ, Dreszer TR, Barber GP, et al.: Track data hubs enable visualization of user-defined genome-wide annotations on the UCSC Genome Browser. Bioinformatics. 2014; 30(7): 1003–1005. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPachter L: Models for transcript quantification from RNA-Seq. arXiv preprint arXiv: 1104.3889. 2011. Reference Source\n\nLi B, Ruotti V, Stewart RM, et al.: RNA-Seq gene expression estimation with read mapping uncertainty. Bioinformatics. 2010; 26(4): 493–500. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWagner GP, Kin K, Lynch VJ: Measurement of mRNA abundance using RNA-seq data: RPKM measure is inconsistent among samples. Theory Biosci. 2012; 131(4): 281–285. PubMed Abstract \| Publisher Full Text\n\nHoffmann S, Otto C, Kurtz S, et al.: Fast mapping of short sequences with mismatches, insertions and deletions using index structures. PLoS Comput Biol. 2009; 5(9): e1000502. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoffmann S, Otto C, Doose G, et al.: A multi-split mapping algorithm for circular, RNA splicing, trans-splicing, and fusion detection. Genome Biol. 2014; 15(2): R34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrapnell C, Pachter L, Salzberg SL: TopHat: discovering splice junctions with RNA-Seq. Bioinformatics. 2009; 25(9): 1105–1111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmman F, Wolfinger MT, Lorenz R, et al.: TSSAR: TSS annotation regime for dRNA-seq data. BMC Bioinformatics. 2014; 15(1). PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8365", "date": "17 Apr 2015", "name": "Angelika Merkel", "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 authors present a useful and relevant toolbox for the analysis of NGS data. Its modular design allows for flexibility in the analysis, and the utilization of track hubs for easy exchange of data as well as visualization with popular tools. A nice implementation is the ability to adapt genome annotations of various formats.Still, I feel the description of the software is rather too general and could be improved.Major Comments:The article lacks any benchmarking or presentation of an example analysis, making it difficult to put the software's performance in perspective with any of the other numerous tools already available. Important for NGS data analysis are specifications for the usage of computational resources (RAM, number of CPUs, processing time, space requirements) and how those scale up with the size of the data set (=number and size of data sets) or type of NGS data (genomic, RNAseq, ChIPseq, Bisulfite-Seq) - all of which are not mentioned. Similarly, the authors do not make any statement on the possibility of parallelization or adaption to cluster infrastructures.Minor comments:Although, truly RPKM has been shown to be inappropriate for measuring the relative molar concentration of a RNA species due to normalization by the total number of reads, it is still widely used. Computing RPMK values as well (optionally) as TPM would allow for comparison with other pipelines.", "responses": [ { "c_id": "1450", "date": "20 Jul 2015", "name": "Michael T. Wolfinger", "role": "Reader Comment", "response": "Thank you very much for taking the time to review our manuscript and for your helpful comments. We have addressed raised issues here in a point-to-point manner, adjusted the text accordingly, and added the requested functionality to the library. We hope that these changes are satisfactory.The article lacks any benchmarking or presentation of an example analysis, making it difficult to put the software's performance in perspective with any of the other numerous tools already available. Important for NGS data analysis are specifications for the usage of computational resources (RAM, number of CPUs, processing time, space requirements) and how those scale up with the size of the data set (=number and size of data sets) or type of NGS data (genomic, RNAseq, ChIPseq, Bisulfite-Seq) - all of which are not mentioned. Similarly, the authors do not make any statement on the possibility of parallelization or adaption to cluster infrastructures.We appreciate this comment and have addressed the concerns at different places throughout the manuscript. These include:Benchmarking and presentation of examples: We have added a section 'Applications' where the process of building custom pipelines is exemplified in terms of the ViennaNGS Tutorials and Utilities. We provide coherent benchmarking data of computer resources required to run the Tutorial pipelines in Table 1. The ViennaNGS Tutorial pipelines have been specifically designed as example implementations of custom ViennaNGS-based analysis workflows. Data-intensive tasks, e.g. BED or BAM filtering, are mainly performed by system calls to third-party tools (bedtools and samtools, respectively), that work on BED or BAM files, regardless whether these originate from RNA-seq or other NGS assays. We appreciate the comment on the possibility of parallelization, which has also been raised by another reviewer, and have added a statement regarding parallelization of ViennaNGS-based pipelines into the Discussion. While the code base not been specifically designed for execution in a parallel environment, specific tasks such as spliting of BAM files can be parallelized trivially within custom ViennaNGS pipelines, provided sufficient IO resources are available. For consistency reasons we stuck to the supplementary shipped data, as included in the Tutorials, for benchmarking. Where applicable, we modified the Tutorials and tested their performance with increasing number of input file (e.g, Tutorial 0) or with increasing size of input file (e.g., Tutorial 2). Both tests showed, as expected, a linear relationship between input, memory and time consumption, respectively.Although, truly RPKM has been shown to be inappropriate for measuring the relative molar concentration of a RNA species due to normalization by the total number of reads, it is still widely used. Computing RPMK values as well (optionally) as TPM would allow for comparison with other pipelines.Thank you very much for this comment. We have added the possibility to compute RPKM alongside TPM within the Bio::ViennaNGS::Expression module and updated the normalize_multicov.pl utility accordingly. Modified versions of the mentioned software are available in Bio::ViennaNGS v0.15." } ] }, { "id": "8057", "date": "23 Apr 2015", "name": "Brad Chapman", "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 authors describe ViennaNGS, a set of Perl modules and scripts to provide RNA-seq analysis and visualization via UCSC integration. The code is nicely written, open source and easy to install via CPAN with cpanminus. Additionally, the documentation is excellent and contains both high level material in the form of blog posts as well as detailed source code descriptions. In reading the paper I found a few areas that would help improve reader's understanding of the toolbox:Please include additional information about what is unique about ViennaNGS in the introduction. Currently it reads generally and is more about pointing about flaws in other software without saying what ViennaNGS provides. The motivation provides much of this text but it seems out of order relative to the introductory material. Please provide benchmarks on your BAM manipulation tools relative other common tools. I don't think this needs to be extensive, but providing a summary of how they perform on a 100Gb 30x whole human genome sequence would be helpful. For filtering comparisons, I suggest comparing with samtools or sambamba (https://github.com/lomereiter/sambamba). For quality control, comparisons to QualiMap (http://qualimap.bioinfo.cipf.es/) or bamtools (https://github.com/pezmaster31/bamtools) would be helpful. Similarly, it would be great to have benchamrking on annotation and BED manipulation tools in ViennaNGS. How does the functionality and timing compare with bedtools? You require and use bedtools for visualization, and it would be useful to clarify benefits and tradeoffs to using ViennaNGS versus interfacing directly with bedtools. How do you handle testing and validation of ViennaNGS tools and pipelines? I saw new tests for UCSC integration coming in during review, which is great. It would be nice to understand the process by which you ensure new development improves (or at least doesn't degrade) the biological results.", "responses": [ { "c_id": "1449", "date": "20 Jul 2015", "name": "Michael T. Wolfinger", "role": "Reader Comment", "response": "Thank you very much for taking the time to review our manuscript. We appreciate your comments and have addressed every issue raised here in a point-to-point manner and modified our manuscript accordingly at different places. We hope that the changes are satisfactory.Please include additional information about what is unique about ViennaNGS in the introduction. Currently it reads generally and is more about pointing about flaws in other software without saying what ViennaNGS provides. The motivation provides much of this text but it seems out of order relative to the introductory material.Thank you very much for this comment, which is highly appreciated. We have re-arranged the Introduction and Methods sections and provide additional information on ViennaNGS' unique selling points, specifically its object oriented design based on the Moose framework and consequently Perl 6 compliance. Please provide benchmarks on your BAM manipulation tools relative other common tools. I don't think this needs to be extensive, but providing a summary of how they perform on a 100Gb 30x whole human genome sequence would be helpful. For filtering comparisons, I suggest comparing with samtools or sambamba (https://github.com/lomereiter/sambamba). For quality control, comparisons to QualiMap (http://qualimap.bioinfo.cipf.es/) or bamtools (https://github.com/pezmaster31/bamtools) would be helpful.ViennaNGS has been designed as a toolbox for building NGS pipelines and does not do any SAM/BAM manipulation itself. For the latter we rely on Bio::DB::Sam, which uses the samtools library internally. Comparison against the mentioned tools is difficult since, to our knowledge, Perl bindings for the mentioned tools are not available.We have benchmarked the ViennaNGS tutorials and provide statistics on time and memory consumption in Table 1. For consistency we have applied the benchmarks to files smaller than the suggested 100Gb 30x coverage, since they are part of our tutorial pipeline and can readily be downloaded from our Web server at http://rna.tbi.univie.ac.at/ViennaNGS. Similarly, it would be great to have benchmarking on annotation and BED manipulation tools in ViennaNGS. How does the functionality and timing compare with bedtools? You require and use bedtools for visualization, and it would be useful to clarify benefits and tradeoffs to using ViennaNGS versus interfacing directly with bedtools.Here the same arguments concerning benchmarking given above apply. Wherever possible we use bedtools for BED manipulation rather than interfacing directly with BED files. The major benefit of using ViennaNGS versus interfacing directly with bedtools is to have data stored consistently in Moose objects which can be referenced throughout the toolbox. As for timing, we do not expect any impact since all bedtools utilities are called via Perl system calls, thus conserving the original bedtools functionality. How do you handle testing and validation of ViennaNGS tools and pipelines? I saw new tests for UCSC integration coming in during review, which is great. It would be nice to understand the process by which you ensure new development improves (or at least doesn't degrade) the biological results.We have added a paragraph outlining the ViennaNGS testing strategy. While we have not yet implemented testing on a global scale, the ViennaNGS::SpliceJunc and ViennaNGS::UCSC modules are currently tested automatically and tests for feature annotation classes will be added in the near future." } ] }, { "id": "8396", "date": "24 Apr 2015", "name": "Björn Voß", "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 their manuscript about ViennaNGS the authors describe a set of perl modules and scripts that is useful to build pipelines for NGS data analysis. A key motivation for this is to promote reproducible science, especially with respect to medium-level users, who often create \"in-house scripts\" for data analysis, which are rarely publicly available. This target community distinguishes ViennaNGS from related approaches, such as Galaxy. The contribution is, thus, relevant and has the potential to serve as a basis for future developments in NGS analysis pipelines. I tested the tutorials and some of the utility scripts and they worked fine. Nevertheless, I think the authors need to clarify some issues and can improve the presentation of their work. Major Comments:The authors should point out clearly, what distinguishes ViennaNGS from other suites. In the end, they need to convince people to use ViennaNGS. For that it would be helpful to clearly state what is hard or even impossible to implement in one of the other systems (galaxy, HTSeq, ...) at best with real world examples. As stated in the title the aim of ViennaNGS is to ease the process of building NGS analysis pipelines. Unfortunately, exactly this aspect is more or less not mentioned in the main text. It would be interesting to know, especially for the data analysts with scripting experience, how such a pipeline looks like and why it is easier to build using ViennaNGS. I do not quite understand the explicit discussion of TPM and RPKM. The differences are extensively discussed in Wagner et al. (2012), which the authors can refer to. Similarly, the description of the accompanying utilities in Table 1 is of minor interest. I would suggest to mention them when the corresponding functionality is described in the main text, e.g., assembly_hub_constructor.pl in the paragraph on Visualization. Furthermore, the authors can explain one tool in detail to show how ViennaNGS pipelines are implemented. BioPerl already provides modules to handle Annotation Features (Bio::SeqFeature), which at first glance seem to provide the same functionality as the ViennaNGS feature annotation classes. Why is there a need for an own class?Minor Comments:An aspect that is becoming more and more important is parallelization. The authors should describe the possibilities of ViennaNGS to be used in cluster or massively parallel environments. The authors should make clear that for some/many tasks they use external tools, such as bedtools2, samtools and tools offered by the UCSC and that the user has to install them on its own. Of course, this is the same as for galaxy and others. I was wondering if ViennaNGS or its pipelines may be integrated into Galaxy. In this way the systems would complement and benefit from each other. At the end of the discussion the authors could provide actual functionalities that they are planning to integrate in the near future. This is interesting for potential users who are missing certain functionalities in the current release. On example is quality control of the raw sequencing data. P.5, Software availability: \"at and\" --> \"and at\"", "responses": [ { "c_id": "1447", "date": "20 Jul 2015", "name": "Michael T. Wolfinger", "role": "Reader Comment", "response": "We would like to thank you for taking the time to review this manuscript, as well as for your helpful comments. We have addressed every issue raised here in a point-to-point manner and modified our manuscript accordingly at different places. We hope that the changes are satisfactory.The authors should point out clearly, what distinguishes ViennaNGS from other suites. In the end, they need to convince people to use ViennaNGS. For that it would be helpful to clearly state what is hard or even impossible to implement in one of the other systems (galaxy, HTSeq,...) at best with real world examples.We appreciate this comment and have added a paragraph in the Introduction, highlighting both the Moose-based object oriented design, and the Perl 6 compliance, which can be regarded as unique selling points of the ViennaNGS suite. ViennaNGS has been developed to provide a toolbox that helps users build their analysis pipelines in Perl, thus targeting researchers who are more literate in Perl than Python. However, ViennaNGS is an open platform and it should therefore be straightforward to implement ViennaNGS-based pipelines within e.g. Galaxy for experienced users. As stated in the title the aim of ViennaNGS is to ease the process of building NGS analysis pipelines. Unfortunately, exactly this aspect is more or less not mentioned in the main text. It would be interesting to know, especially for the data analysts with scripting experience, how such a pipeline looks like and why it is easier to build using ViennaNGS.We have added a new section 'Applications' where the process of building custom pipelines is exemplified in terms of the ViennaNGS Tutorials and Utilities. The ViennaNGS Tutorials explain in detail how custom pipelines can be built for a set of real-world NGS applications. I do not quite understand the explicit discussion of TPM and RPKM. The differences are extensively discussed in Wagner et al. (2012), which the authors can refer to.The extensive discussion of TPM and RPKM, including all formulas, have been removed from the manuscript. Similarly, the description of the accompanying utilities in Table 1 is of minor interest. I would suggest to mention them when the corresponding functionality is described in the main text, e.g., assembly_hub_constructor.pl in the paragraph on Visualization. Furthermore, the authors can explain one tool in detail to show how ViennaNGS pipelines are implemented.We respectfully disagree and think that the ViennaNGS utilities should be mentioned in one place, given that they can be regarded, apart from the ViennaNGS Tutorials, as yet another set of example implementations of ViennaNGS library functions. Moreover, we have shifted the paragraph mentioning the Utilities into the Applications section. BioPerl already provides modules to handle Annotation Features (Bio::SeqFeature), which at first glance seem to provide the same functionality as the ViennaNGS feature annotation classes. Why is there a need for an own class?BioPerl and it's associated modules are a fantastic toolbox for everyday bioinformatics work and we use them whenever applicable (e.g. via Bio::DB::Sam). In general the Bio::Seq and especially Bio::SeqFeature classes allow a multitude of operations on common biological features, their annotations and file formats. ViennaNGS was designed with strong focus on NGS analysis and easy portability to Perl 6. Given that the ViennaNGS feature annotation classes play a pivotal role in current and future development of the toolbox, we decided to implement Moose classeswithout introducing too many dependencies on existing BioPerl modules. We went for a design that specifically fits the needs of NGS analysis and stays as minimal as possible. In this sense, we do not see ViennaNGS in competition to BioPerl but as a boutique alternative for NGS data analysts.An aspect that is becoming more and more important is parallelization. The authors should describe the possibilities of ViennaNGS to be used in cluster or massively parallel environments.We appreciate this comment and have added a statement on parallelization of ViennaNGS-based pipelines into the Discussion. Our focus in the initial development phase of ViennaNGS has not been on parallelization, hence the code base has not been specifically designed for parallel processing in a cluster environment. It should, however, be straightforward to implement certain tasks in multithreaded pipelines.The authors should make clear that for some/many tasks they use external tools, such as bedtools2, samtools and tools offered by the UCSC and that the user has to install them on its own. Of course, this is the same as for galaxy and others.A section listing all third party dependencies has been added to the main text. I was wondering if ViennaNGS or its pipelines may be integrated into Galaxy. In this way the systems would complement and benefit from each other. As mentioned earlier, since ViennaNGS is implemented purely in Perl, it should be straightforward for experienced users to integrate its functionalities into Galaxy, e.g., via the the Galaxy Tool Factory. At the end of the discussion the authors could provide actual functionalities that they are planning to integrate in the near future. This is interesting for potential users who are missing certain functionalities in the current release. On example is quality control of the raw sequencing data.The Discussion has been updated accordingly.P.5, Software availability: \"at and\" --> \"and at\"Done." } ] } ]	1	https://f1000research.com/articles/4-50
https://f1000research.com/articles/4-273/v1	17 Jul 15	{ "type": "Review", "title": "Seeing and believing: recent advances in imaging cell-cell interactions", "authors": [ "Alpha S. Yap", "Magdalene Michael", "Robert G. Parton", "Magdalene Michael" ], "abstract": "Advances in cell and developmental biology have often been closely linked to advances in our ability to visualize structure and function at many length and time scales. In this review, we discuss how new imaging technologies and new reagents have provided novel insights into the biology of cadherin-based cell-cell junctions. We focus on three developments: the application of super-resolution optical technologies to characterize the nanoscale organization of cadherins at cell-cell contacts, new approaches to interrogate the mechanical forces that act upon junctions, and advances in electron microscopy which have the potential to transform our understanding of cell-cell junctions.", "keywords": [ "cell-cell interactions", "cadherin", "adherin junctions", "cell-cell junctions", "imaging", "electron microscopy" ], "content": "Introduction\n\nCell biologists are often resolutely visual people: we believe most what we can see best. This is a heritage of the history of our discipline, which found its roots in work such as Palade’s application of electron microscopy to characterize cellular and subcellular structure. Later, the introduction of antibody technologies allowed morphology to be complemented by molecular specificity. Advances in our understanding of cell biology thus have been driven by the combination of new technologies in microscopy and new reagents that allow us to probe cellular constitution and function.\n\nIn this article, we aim to review how this combination of new technologies and reagents has advanced our understanding of the biology of cadherin-based adherens junctions. We focus on three of these advances. First, we have come to appreciate that adherens junctions are not homogenous collections of cadherin receptors but rather have patterns of organization that are apparent at the nanoscale (smaller than a micron) and mesoscopic scale (tens of microns). Second, we now know that cadherin-based adhesions are active mechanical agents where cells generate force to test their environment and sense forces that are applied upon them. Third, although many of these insights have come from developments in light microscopy, the last 5 to 10 years have also seen the development of dramatic new tools in electron microscopy; these have yet to be widely applied to study cell-cell interactions, but their potential is enormous.\n\nOptical microscopy has been revolutionized by techniques that have overcome the limits that the diffraction of light imposes on spatial resolution1. These include approaches such as structured illumination (Figure 1) and fluorescence photoactivated localization microscopy/stochastic optical reconstruction microscopy (F-PALM/STORM), which are now being applied to the characterization of cell-cell junctions2–4. Already, they have provided valuable insights into how cadherins are organized into clusters at the nanoscale.\n\nCaco-2 cells were stained for E-cadherin, F-actin, and myosin IIA. Details in the region marked by the box are shown on the right side. Bars = 5 μm (on the left side) and 1 μm (for the magnified images on the right).\n\nThe capacity for cadherins to organize into lateral clusters was observed nearly 20 years ago when it was identified as a mechanism that could strengthen cadherin-based adhesion5,6, probably by increasing the avidity of adhesive binding between cadherins and their ligands6. However, those experiments were performed by using reductionist models, such as fibroblasts engineered to express E-cadherin5 or cells adherent to substrata coated with C-cadherin ligands (analogous to the two-dimensional substrata that students of integrin biology have long used to study focal adhesions and focal contacts)6. It was more difficult to determine the extent to which lateral clustering might occur at the native cell-cell contacts formed between cells that express endogenous cadherins, such as simple polarized epithelia. High-resolution confocal imaging had identified clustering in Drosophila embryos7 and cultured mammalian cells8 but did not readily permit quantitative analysis of the extent or nature of this clustering. More commonly, cadherins appeared to distribute extensively at contacts between cells, as if junctions represented carpets of homoligated cadherin complexes.\n\nTwo recent articles applied PALM/STORM to characterize nanoscale E-cadherin distribution in Drosophila embryonic epithelia3 and cultured mammalian cells4. Both clearly demonstrated that E-cadherin was distributed in polydisperse clusters throughout the junctions of these epithelial systems. They confirm that lateral clustering is a fundamental feature of the supramolecular organization of cadherins at junctions. Furthermore, mammalian junctions displayed clusters with a preferred size of approximately 50–60 nm, which then could organize into larger-scale groups4.\n\nMore detailed quantitative analysis also provided provocative insights into the cellular control of clustering. Earlier studies based on analysis of the crystal structure of cadherin ectodomains proposed a model in which trans-interactions between the ectodomains presented on the surfaces of neighboring cells, combined with cis-interactions between ectodomains on the same cell surface, could cause packing into clusters9,10. However, the cytoplasmic tail also supports clustering in cells4,11. Wu et al.4 (2015) found that the molecular density of cadherins could vary even within the same cluster. Some regions within clusters showed high packing density, comparable to that predicted from the crystal structures; this required the ability of cadherins to undergo both cis- and trans-interactions. However, even when the ability to make cis- and trans-interactions was ablated, cells could still make clusters with a size (50–60 nm) similar to those of wild-type cadherins. This implied that adhesive ligation might not be necessary for clustering to occur. Indeed, clusters were observed at the free surfaces of cells, where cadherins could not engage in adhesion, and even with cadherin mutants that lacked the whole adhesive ectodomain4. Instead, clustering required an intact actin cytoskeleton, and detailed inspection suggested that cadherin clusters might be delimited by “corrals” of cortical actin. Consistent with this, Troung Quang et al.3 (2013) demonstrated that F-actin integrity was necessary to stabilize cadherin clusters. Overall, this implies that multiple mechanisms can influence clustering. In one model, cortical actin may define a minimal cadherin cluster, which does not require adhesive ligation; however, the packing of cadherin molecules within clusters is increased upon ligation.\n\nComparison of the two studies also highlights how the operational definition of “clusters” can fundamentally condition the detailed quantitative analysis and its interpretation. For example, although both groups used the same algorithm to analyze their data, they differed in their definition of clusters and hence in the metrics that they used to describe the clusters. Troung Quang et al.3 used a kinetics-based model which defined “size” in stoichiometric terms, as the number of cadherin molecules present within clusters. In contrast, Wu et al. took a more empirical approach that focused on the spatial size of the clusters. What emerged with the first approach, as confirmed by Wu et al., was that the distribution of “sizes” followed a power law, implying that the mechanisms that governed how many cadherin molecules accumulate in a cluster did not have a preferred number. However, a power law relationship was not evident when “size” was defined spatially, as the diameter of the cluster, the data being better fit to a Gaussian distribution. Taken together, these findings suggest that there may be a preferred spatial dimension to a cadherin cluster (approximately 50 nm), but within this physical limit the number of cadherin molecules that can be accumulated varies over a wide range. This emphasizes that how the apparently straightforward notion of “size” is explicitly implemented in the computational analysis will deeply influence data interpretation with these approaches.\n\nMore generally, these studies suggest that the notion of a “cluster” may need to be conceptually defined with greater precision than we have sometimes done in the past. The work of Wu et al. suggests that there may be elemental units that may reflect the spatial organization of the cortical actin cytoskeleton. However, these appear to be able to organize into larger-scale conglomerations and accumulate a variable number of cadherin molecules. It should be remembered that cadherins exist as macromolecular complexes with a range of associated proteins10. So the clusters of cadherins will more likely represent nanoassemblies of many different proteins. What mechanisms define these larger-scale patterns of organization have yet to be established. However, insofar as the phenomenon of receptor clustering has been implicated in regulating cellular processes as fundamental as cell signaling12,13 and receptor sensitivity14, it will be important for us to clearly specify what aspect of “clustering” we are talking about when we come to further analyze the role that clustering plays in cadherin biology.\n\nA fundamental advance in our understanding of cadherin biology has come from the realization that cadherin adhesion serves to couple the contractile cortices of cells together15,16. Indeed, cadherins may promote the biogenesis of the junctional contractile apparatus itself8,17. An important part of this advance has come from the application of tools and theory from the physical sciences to biology, combined with the development of new reagents that allow us to measure molecular-scale tension.\n\nFor example, one of the most popular approaches to assessing tension is to cut regions (cortices, junctions, and whole cells) with a laser and measure the instantaneous velocity of recoil as an index of the tension that had been present beforehand18. This has been used in embryonic tissues19,20 as well as in cell culture models8. Similar nanoablation techniques have been combined with physical theory to characterize patterns of cortical tension in Caenorhabditis elegans embryos21. It should be noted that, though intuitively attractive, the velocity of recoil is not itself a direct measure of tension. Instead, recoil velocity reflects the ratio of tension over frictional forces. When used to infer tension, this assay assumes that the frictional elements (which would reflect the viscoelastic properties of the junctions) do not change between experimental maneuvers18. Ultimately, precise interpretation of recoil velocity needs to be informed by measurements of junctional viscoelasticity22. Other indirect assays have measured junctional movements to infer tension when combined with explicit mechanical models23,24.\n\nThese essentially mesoscopic measurements can be productively complemented by the use of molecular-level tension-sensitive biosensors, such as the Förster resonance energy transfer (FRET)-based system developed by Grashoff et al.25. This sensor reports tension based on the displacement of FRET pairs that are separated by an elastic linker derived from spider silk. The tension sensor (TS) module has been inserted into a range of proteins, where it reported tension over both cadherins (E-cadherin and VE-cadherin26,27) and vinculin at cell-cell junctions28. Of note, the TS module was calibrated in vitro, where it displayed greatest sensitivity over a range of 1–6 pN25. Therefore, its efficacy as a reporter will depend on whether the molecular-level forces that are present fall within its range of sensitivity. Nonetheless, the mesoscopic and molecular-scale approaches to measuring tension are complementary and it is informative to compare both assays, where possible. For example, in mature focal adhesions, which are thought to be sites where contractile force is exerted upon integrin complexes29, vinculin itself can become uncoupled from tension25, despite the integrity of the focal adhesion being unchanged. Thus, molecular-level tension may not always correlate with mesoscopic-level tension.\n\nAn important issue for the future is to better characterize the material properties of cell-cell junctions. Until now, we have lacked the tools to measure those properties. But things have begun to change. He et al.30 (2014) followed the patterns of flow of microbeads injected into Drosophila embryonic epithelia to assess the patterns of mechanical connectivity between cells. They concluded that lateral cell-cell junctions did not present substantive barriers to hydrodynamic flow between cells. Furthermore, Bambardekar et al.22 (2015) demonstrated that it was possible to manipulate cell-cell junctions in Drosophila embryonic epithelium by using optical tweezers and thereby assess the mechanical properties of the junctions. Whether such approaches will be more broadly applicable in other cellular systems remains to be tested.\n\nThe suite of light microscopic techniques available to researchers is impressive, but we are also witnessing a revolution in electron microscopy, from high-resolution structural analysis to ultrastructural analysis of whole tissues in three dimensions (3D). Many of these methods are becoming routine in laboratories throughout the world but have not been extensively applied to the study of cell-cell interactions. Here, we will briefly summarise relevant techniques and their possible applications in this area.\n\nUltrastructural methods can potentially answer how molecular interactions and spatial interactions contribute to the formation and function of junctional assemblies. The ideal method would allow visualization of both the cytoskeleton and membranous elements which together generate the active junctional complex; it should also have the resolution to identify the location of individual protein components in the context of a 3D volume of the cell-cell contact sites. This should include actin and other cytoskeletal networks, cadherin, and actin-binding proteins and should be correlated with real-time observations of junctional dynamics. Although some elements can be recognized by morphology alone (cytoskeleton and junctions), new labeling methods are now facilitating visualization of otherwise undetectable components and can be combined with 3D methods.\n\nConventional electron microscopy, involving chemical fixation and embedding in resin, is still an excellent method for visualization of the membrane and cytoskeletal elements of cell-cell contacts (Figure 2). However, note that the complexity of the junctional cytoskeleton makes detailed analyses of its organization difficult. This can be resolved by electron tomography, which involves tilting a relatively thick (for example, 300 nm) section and obtaining images at different angles relative to the specimen. This provides not only a 3D view through the depth of the specimen but also far greater resolution, allowing identification and tracing of individual elements. This has been used to great effect in recent studies of the actin organization in cultured cells with actin filaments running parallel to the adherens junction31.\n\nMicrotubules are highlighted in green, and putative actin filaments in red. Bar = 500 nm.\n\nNew methods are now providing far greater sample depths and, for the first time, the ability to examine entire cells, large tissue areas, and even entire organisms (albeit the smaller specimens of the animal kingdom). This method, serial blockface scanning electron microscopy (SEM), relies on the imaging of an exposed blockface by SEM in the back-scattered mode32. Material is removed from the blockface, slice by slice using either a knife or a focused ion beam, within the electron microscope, and the exposed blockface is imaged after each slice is removed to generate literally thousands of serial images. Improvements in back-scattered electron detectors now mean that image quality is approaching that of a conventional transmission electron microscope (and the image is contrast-inverted to give a similar appearance). This technique has the potential to provide large-scale information on the way that cells interact in the culture dish but also in a tissue environment, with the capacity to contain numerous cells in a single 3D data set.\n\nThe above methods rely on an initial fixation step, usually using chemical fixatives. The latter can be slow and introduce artefacts, and so there has been a move to cryofixation, usually high-pressure freezing. These methods provide excellent preservation of cellular structures and are becoming routine in many laboratories. However, avoiding chemical fixation by cryofixation introduces another problem: how to go from a frozen sample in liquid nitrogen to an embedded specimen that can be sectioned (note that thin samples can avoid this problem, but this is unlikely to be the case for the study of most cell-cell junctions). Cryosectioning of frozen material provides the optimal method to preserve structure, avoiding both fixatives and any staining process. But this is technically demanding, and the retention of cytoplasmic material can actually hinder visualization of cytoskeletal elements. Freeze substitution, the removal of water at low temperature before embedding, offers a simple and, now, very rapid alternative for embedding in resin after freezing33. Freezing of specimens to sectioning can now be completed in one day. Of particular note for studies aiming to correlate real-time light microscopy with electron microscopy is that methods now exist to maintain the fluorescence of green fluorescent protein and related proteins in resin-processed material34–38. Thus, the behavior of proteins can be followed in real-time, and the cells then fast-frozen to capture a rapid transient event and then processed for embedding in resin. The same material then can be analyzed by light microscopy and by electron microscopy to allow precise correlation of the two sets of observations. Recent modifications of these methods have described fluorescent proteins that are resistant to harsh fixation conditions39, opening the possibility for correlative microscopy to combine super-resolution imaging of fluorescent proteins with electron microscopy to better characterize their local cellular nano-environment.\n\nUltimately, researchers would like to see and recognize all the components involved in cell-cell interactions and understand their precise molecular arrangement. We can already see and recognize some of those components, such as F-actin and junctions, and, as described above, we can see them in 3D and increasingly even in the context of whole tissues. But what about the recognition of other components? Can we imagine visualizing individual cadherin molecules or the key regulators of the junctional actin network in a quantitative fashion? Immunogold labeling has long been used to label on sections, and this method has been the gold standard for ultrastructural localization studies40. However, immunogold labeling is relatively inefficient and labeling is generally restricted to the surface of the section (and therefore is hardly useful for the 3D methods such as electron tomography and serial blockface SEM). The most efficient method, using thawed frozen sections, provides excellent visualization of membranes41 but is not routinely useful for visualizing cytoskeletal structures. But new labeling methods are offering possibilities for genetic tagging of proteins for electron microscopy. Of these, the most promising appears to be a peroxidase construct which can be fused to any protein of interest42. The expressed fusion protein can be visualized by using a simple peroxidase reaction on the fixed material to deposit an electron-dense precipitate at the site of the fusion protein. This method may appear to lack the precision of a particulate marker, but the enzyme is directly fused to the protein of interest rather than being detected with antibodies. Importantly, the reaction product can also be detected within the depth of a thick section (for tomography) or in a whole cell or tissue sample, facilitating detection of a protein of interest by serial blockface SEM. This has immense potential for 3D studies of protein localization.\n\n\nFuture directions\n\nWe are living in a Golden Age for biological imaging, where new microscopy techniques and reagents are allowing us to identify biological structures with unparalleled detail and to interrogate the chemical and physical properties of cells and tissues. Nor is it likely that we have exhausted the possibilities. Already light sheet microscopy in its developing forms provides the opportunity to analyze whole organisms in a comprehensive, dynamic manner43. One consequence of these advances has been the generation of quantitative data, and this has entailed the need for mathematical and statistical tools to analyze often very large data sets. These large data sets carry challenges for how we present and “consume” such data. It seems likely that this will promote an even greater nexus between theory and experiment in biology. Just as seeing can be believing, so can our pre-existing ideas and beliefs influence what we see. The application of new physical theory provides the opportunity to develop predictive models, which are informed by the new dynamic and quantitative data that microscopy provides and which yield predictions for further experimentation. These new advances in microscopy and theory provide the chance for us to interrogate complex biological phenomena at cell-cell junctions across vastly different length and time scales, from molecular events to organismal development.", "appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nAlpha S. Yap and Robert G. Parton are supported by research fellowships from the National Health and Medical Research Council of Australia (NHMRC) (1044041 and 569452). Our work is funded by the NHMRC Program in Membrane Interface Biology (grant 1037320) and grants from the Australian Research Council.\n\n\nReferences\n\nGalbraith CG, Galbraith JA: Super-resolution microscopy at a glance. J Cell Sci. 2011; 124(Pt 10): 1607–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGomez GA, McLachlan RW, Wu SK, et al.: An RPTPα/Src family kinase/Rap1 signaling module recruits myosin IIB to support contractile tension at apical E-cadherin junctions. Mol Biol Cell. 2015; 26(7): 1249–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTruong Quang BA, Mani M, Markova O, et al.: Principles of E-cadherin supramolecular organization in vivo. Curr Biol. 2013; 23(22): 2197–207. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nWu Y, Kanchanawong P, Zaidel-Bar R: Actin-delimited adhesion-independent clustering of E-cadherin forms the nanoscale building blocks of adherens junctions. Dev Cell. 2015; 32(2): 139–54. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nAngres B, Barth A, Nelson WJ: Mechanism for transition from initial to stable cell-cell adhesion: kinetic analysis of E-cadherin-mediated adhesion using a quantitative adhesion assay. J Cell Biol. 1996; 134(2): 549–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYap AS, Brieher WM, Pruschy M, et al.: Lateral clustering of the adhesive ectodomain: a fundamental determinant of cadherin function. Curr Biol. 1997; 7(5): 308–15. PubMed Abstract \| Publisher Full Text\n\nCavey M, Rauzi M, Lenne PF, et al.: A two-tiered mechanism for stabilization and immobilization of E-cadherin. Nature. 2008; 453(7196): 751–6. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nWu SK, Gomez GA, Michael M, et al.: Cortical F-actin stabilization generates apical-lateral patterns of junctional contractility that integrate cells into epithelia. Nat Cell Biol. 2014; 16(2): 167–78. PubMed Abstract \| Publisher Full Text\n\nHarrison OJ, Jin X, Hong S, et al.: The extracellular architecture of adherens junctions revealed by crystal structures of type I cadherins. Structure. 2011; 19(2): 244–56. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBrasch J, Harrison OJ, Honig B, et al.: Thinking outside the cell: how cadherins drive adhesion. Trends Cell Biol. 2012; 22(6): 299–310. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYap AS, Niessen CM, Gumbiner BM: The juxtamembrane region of the cadherin cytoplasmic tail supports lateral clustering, adhesive strengthening, and interaction with p120ctn. J Cell Biol. 1998; 141(3): 779–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIversen L, Tu HL, Lin WC, et al.: Molecular kinetics. Ras activation by SOS: allosteric regulation by altered fluctuation dynamics. Science. 2014; 345(6192): 50–4. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nSaunders TE, Pan KZ, Angel A, et al.: Noise reduction in the intracellular pom1p gradient by a dynamic clustering mechanism. Dev Cell. 2012; 22(3): 558–72. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBray D, Levin MD, Morton-Firth CJ: Receptor clustering as a cellular mechanism to control sensitivity. Nature. 1998; 393(6680): 85–8. PubMed Abstract \| Publisher Full Text\n\nMaître JL, Berthoumieux H, Krens SF, et al.: Adhesion functions in cell sorting by mechanically coupling the cortices of adhering cells. Science. 2012; 338(6104): 253–6. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nLevayer R, Lecuit T: Biomechanical regulation of contractility: spatial control and dynamics. Trends Cell Biol. 2012; 22(2): 61–81. PubMed Abstract \| Publisher Full Text\n\nMichael M, Yap AS: The regulation and functional impact of actin assembly at cadherin cell-cell adhesions. Semin Cell Dev Biol. 2013; 24(4): 298–307. PubMed Abstract \| Publisher Full Text\n\nHutson MS, Tokutake Y, Chang MS, et al.: Forces for morphogenesis investigated with laser microsurgery and quantitative modeling. Science. 2003; 300(5616): 145–9. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nFernandez-Gonzalez R, Simoes Sde M, Röper JC, et al.: Myosin II dynamics are regulated by tension in intercalating cells. Dev Cell. 2009; 17(5): 736–43. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMartin AC, Gelbart M, Fernandez-Gonzalez R, et al.: Integration of contractile forces during tissue invagination. J Cell Biol. 2010; 188(5): 735–49. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMayer M, Depken M, Bois JS, et al.: Anisotropies in cortical tension reveal the physical basis of polarizing cortical flows. Nature. 2010; 467(7315): 617–21. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBambardekar K, Clément R, Blanc O, et al.: Direct laser manipulation reveals the mechanics of cell contacts in vivo. Proc Natl Acad Sci U S A. 2015; 112(5): 1416–21. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nIshihara S, Sugimura K: Bayesian inference of force dynamics during morphogenesis. J Theor Biol. 2012; 313: 201–11. PubMed Abstract \| Publisher Full Text\n\nBrodland GW, Conte V, Cranston PG, et al.: Video force microscopy reveals the mechanics of ventral furrow invagination in Drosophila. Proc Natl Acad Sci U S A. 2010; 107(51): 22111–6. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGrashoff C, Hoffman BD, Brenner MD, et al.: Measuring mechanical tension across vinculin reveals regulation of focal adhesion dynamics. Nature. 2010; 466(7303): 263–6. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nConway DE, Breckenridge MT, Hinde E, et al.: Fluid shear stress on endothelial cells modulates mechanical tension across VE-cadherin and PECAM-1. Curr Biol. 2013; 23(11): 1024–30. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBorghi N, Sorokina M, Shcherbakova OG, et al.: E-cadherin is under constitutive actomyosin-generated tension that is increased at cell-cell contacts upon externally applied stretch. Proc Natl Acad Sci U S A. 2012; 109(31): 12568–73. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLeerberg JM, Gomez GA, Verma S, et al.: Tension-sensitive actin assembly supports contractility at the epithelial zonula adherens. Curr Biol. 2014; 24(15): 1689–99. PubMed Abstract \| Publisher Full Text\n\nBershadsky A, Kozlov M, Geiger B: Adhesion-mediated mechanosensitivity: a time to experiment, and a time to theorize. Curr Opin Cell Biol. 2006; 18(5): 472–81. PubMed Abstract \| Publisher Full Text\n\nHe B, Doubrovinski K, Polyakov O, et al.: Apical constriction drives tissue-scale hydrodynamic flow to mediate cell elongation. Nature. 2014; 508(7496): 392–6. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nBuckley CD, Tan J, Anderson KL, et al.: Cell adhesion. The minimal cadherin-catenin complex binds to actin filaments under force. Science. 2014; 346(6209): 1254211. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nDenk W, Horstmann H: Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol. 2004; 2(11): e329. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMcDonald KL, Webb RI: Freeze substitution in 3 hours or less. J Microsc. 2011; 243(3): 227–33. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKukulski W, Schorb M, Kaksonen M, et al.: Plasma membrane reshaping during endocytosis is revealed by time-resolved electron tomography. Cell. 2012; 150(3): 508–20. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKukulski W, Schorb M, Welsch S, et al.: Correlated fluorescence and 3D electron microscopy with high sensitivity and spatial precision. J Cell Biol. 2011; 192(1): 111–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nKukulski W, Schorb M, Welsch S, et al.: Precise, correlated fluorescence microscopy and electron tomography of lowicryl sections using fluorescent fiducial markers. Methods Cell Biol. 2012; 111: 235–57. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nNixon SJ, Webb RI, Floetenmeyer M, et al.: A single method for cryofixation and correlative light, electron microscopy and tomography of zebrafish embryos. Traffic. 2009; 10(2): 131–6. PubMed Abstract \| Publisher Full Text\n\nSchieber NL, Nixon SJ, Webb RI, et al.: Modern approaches for ultrastructural analysis of the zebrafish embryo. Methods Cell Biol. 2010; 96: 425–42. PubMed Abstract \| Publisher Full Text\n\nPaez-Segala MG, Sun MG, Shtengel G, et al.: Fixation-resistant photoactivatable fluorescent proteins for CLEM. Nat Methods. 2015; 12(3): 215–8, 4 p following 218. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nSlot JW, Geuze HJ: Cryosectioning and immunolabeling. Nat Protoc. 2007; 2(10): 2480–91. PubMed Abstract \| Publisher Full Text\n\nTokuyasu KT: A technique for ultracryotomy of cell suspensions and tissues. J Cell Biol. 1973; 57(2): 551–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartell JD, Deerinck TJ, Sancak Y, et al.: Engineered ascorbate peroxidase as a genetically encoded reporter for electron microscopy. Nat Biotechnol. 2012; 30(11): 1143–8. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nChen BC, Legant WR, Wang K, et al.: Lattice light-sheet microscopy: imaging molecules to embryos at high spatiotemporal resolution. Science. 2014; 346(6208): 1257998. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "9551", "date": "17 Jul 2015", "name": "Pierre-François Lenne", "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", "responses": [] }, { "id": "9552", "date": "17 Jul 2015", "name": "Keith Burridge", "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", "responses": [] }, { "id": "9553", "date": "17 Jul 2015", "name": "Pakorn T. Kanchanawong", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-273
https://f1000research.com/articles/4-269/v1	17 Jul 15	{ "type": "Research Note", "title": "Questioned validity of Gene Expression Dysregulated Domains in Down's Syndrome", "authors": [ "Long H. Do", "William C. Mobley", "Nishant Singhal", "Long H. Do", "William C. Mobley" ], "abstract": "Recently, in studies examining fibroblasts obtained from the tissues of one set of monozygotic twins (i.e. fetuses derived from the same egg) discordant for trisomy 21 (Down syndrome; DS), Letourneau et al., reported the presence of a defined pattern of dysregulation within specific genomic domains they referred to as Gene Expression Dysregulated Domains (GEDDs). GEDDs were described as alternating segments of increased or decreased gene expression affecting all chromosomes. Strikingly, GEDDs in fibroblasts were largely conserved in induced pluripotent cells (iPSCs) generated from the twin’s fibroblasts as well as in fibroblasts from the Ts65Dn mouse model of DS. Our recent analysis failed to find GEDDs. We reexamined the human iPSCs RNAseq data from Letourneau et al., and data from this same research group published earlier examining iPSCs from the same monozygotic twins. An independent analysis of RNAseq data from Ts65Dn fibroblasts also failed to confirm presence of GEDDs. Our analysis questions the validity of GEDDs in DS.", "keywords": [ "Gene expression", "Down syndrome", "iPSC", "RNAseq" ], "content": "Main text\n\nThe surprising findings by Letourneau1 and colleagues prompted us to examine our own, as yet unpublished, Ts65Dn transcriptome data for the developing and mature hippocampus in an attempt to identify GEDDs. Our data provided no evidence for the pattern reported in Letourneau et al.’s work. We first entertained the possibility that GEDDs were not present in post mitotic cells or cells undergoing neural differentiation. However, to ensure that we fully understood the published GEDD data, we examined the entire RNAseq dataset from the Letourneau manuscript, as provided publicly via the Gene Expression Omnibus.\n\nPrincipal component analysis (PCA) of RNAseq replicates from the twin’s fibroblast (T1DS: twin with DS; T2N: disomic twin) revealed a great deal of variability (Figure 1A). When the datasets from Letourneau et al., are compared, in two of four cases, a closer relationship exists between the DS and disomic twin fibroblasts than for replicates from the same individual. [The datasets from Letourneau et al. are denoted by –L]. For example, one of the 2N-hFibro-L replicates clustered more tightly with a DS-hFibro-L replicate than with any of its own replicate (2N) samples.\n\nPCA analysis of global gene expression among RNAseq replicates from the twin’s fibroblasts and iPSCs. Comparing hFibro-L replicates with themselves reveals a high degree of variability along the most significant component, PC1. In addition, there is great variability between hiPSCs-L and hiPSCs-H.\n\nWe next checked the variability of the twin’s iPSCs RNAseq data. We found an additional three RNASeq replicates (hiPSCs-H) performed by the same research group and published earlier using fibroblasts from the same monozygotic twins2. PCA analysis of these data revealed that replicates of hiPSCs-H (H for Hibaoui) clustered well together; however, they did not cluster well with the data for hiPSCs-L (Figure 1A). Altogether, PCA analysis indicated marked variability between datasets, raising the possibility that technical issues in the RNAseq samples or in their analysis compromised the Letourneau study.\n\nTo further explore the additional three hiPSCs-H RNAseq replicates, we searched for GEDDs using methods similar to those utilized by Letourneau and colleagues. Our analysis of the hiPSCs-H did not find conserved patterns indicative of GEDDs. Figure 1B shows the results for two chromosomes, as examples. The authors reported high global gene fold-change correlations between the twin’s fibroblasts and derived iPSCs. Our re-analysis found a similar high correlation value of ρ = 0.82 between the hiPSCs-L and hFibro-L. However, we found the hiPSCs-H poorly correlated with the original datasets; ρ = 0.31 between hiPSCs-L and hiPSCs-H; ρ = 0.07 between hiPSCs-H and hFibro-L (Figure 1C and Supplementary figure 1A).\n\n(1) hiPSCs-L derived from monozygotic twins discordant for DS, Letourneau 2014 (red). (2) hiPSCs-H from the same monozygotic twins discordant for DS from Hibaoui 2014 (blue). (3) Human fibroblasts (hFibro-L) from monozygotic twins discordant for DS from Letourneau 2014 (black). hiPSCs-H lack GEDDs, while original hiPSCs-L and hFibro-L show GEDDs and high Spearman’s correlation (ρ1,3).\n\nWhile iPSCs (hiPSCs-L) from the original study are highly similar to fibroblasts (hFibro-L) with a global correlation of ρ=0.82, hiPSCs-H show poor correlation with those samples (ρ=.07, ρ=0.31).\n\nConservation of GEDDs in Ts65Dn mouse model of DS were quite unexpected given that Ts65Dn mouse is segmentally trisomic (34Mb) for a portion of mouse chromosome 16 (MMU16); the segment contains about 88 mouse homologues to human genes on the long arm of HSA21; it also carries an extra copy of the approximately 10Mb centromeric segment of MMU17 that is not syntenic to any region on human chromosome 213–6. To further explore the possibility of GEDDs, RNAseq data was obtained from three replicates each from Ts65Dn and wild type mouse embryonic fibroblasts (MEFs-D) (D for Do denotes MEFs in the current study). PCA revealed tight clustering between our replicates (Figure 1D), but not those for the MEFs-L samples. While we found expected changes in gene expression in MEFs-D, analysis of MEFs-L and MEFs-D found a poor global correlation (ρ = -0.31) (Figure 1E and Supplementary figure 1B); this was also the case across all mouse chromosomes (for examples see Figure 1F and Supplementary figure 1B). In summary our findings raise serious concerns regarding the validity of GEDDs. We find no evidence for such domains in the studies on DS referenced herein or in cells from the Ts65Dn mouse model of DS.\n\nPCA reveals little variance along the most significant component, PC1, of global gene expression among RNAseq replicates from our repeated experiments, MEFs-D. RNAseq data of mice fibroblasts from the original study, MEFs-L, do not cluster with MEFs-D.\n\nEmbryonic mice fibroblasts examined herein (MEFs-D; blue), from 2N and Ts65DN mice do not show the GEDDs reported for MEFs-L (red).\n\nGlobal gene expression fold-change from MEFs-D and MEFs-L show a poor Spearman’s correlation overall (ρ=-0.31) as well as for each of the chromosomes.\n\n\nMethods\n\nTotal RNA was collected using TRIzol reagent and further purified using RNeasy mini Kit, (Qiagen) from primary mouse embryonic fibroblasts (MEFs) derived from 18.5-day-old Ts65Dn and 2N mouse using manufacturer’s instructions. RNA quality was checked using Tapestation 2200 (Agilent technologies) and quantified using Qubit instrument (Life technologies). TrueSeq stranded mRNA-seq libraries were prepared from 5 μg of total RNA (Illumina mRNA-seq kit, RS-122-2103) and sequenced using Illumina HiSeq 2500 PE-100 (sequences publically available from GEO, accession number: GSE64840). Experiments were performed in triplicate.\n\nRNAseq data from hFibro-L, hiPSCs-L, and hiPSCs-H were downloaded from the Sequence Read Archive (SRP039348, SRP032928) and uploaded to Illumina BaseSpace for mapping (BaseSpace App v1.0, TopHat v2) and differential gene analysis (BaseSpace App v1.1, CuffLinks v2.1.1). PCA was performed using R (v3.1.0) from normalized gene count values (FPKM). Overall Spearman correlation values were calculated from locally weighted scatterplot smoothing (LOWESS) with 30% bandwidth between log2 (FC) gene expression of comparison samples, ordered by genes along each chromosome and plotted using R and custom scripts.\n\n\nSoftware availability\n\nCustom scripts for R used to calculate Spearman correlation values are available at https://github.com/lhdo/GEDDplot\n\nhttps://github.com/F1000Research/GEDDplot/releases/tag/V1\n\nhttp://dx.doi.org/10.5281/zenodo.19232\n\nThe MIT license", "appendix": "Author contributions\n\n\n\nLD contributed in writing the manuscripts and all the scripts used to perform analysis of the data. NS contributed in writing the manuscript and performed all the experiments for sequencing. WM contributed in writing the manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing financial interests.\n\n\nGrant information\n\nNIH grants NS055371 and NS24054, Larry L. Hillblom Foundation, Lumind Foundation (formerly Down Syndrome Research and Treatment Foundation), Research Down Syndrome, Thrasher Research Fund, Adler Foundation, and Alzheimer Association.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nTrueSeq stranded mRNA-seq libraries preparation and sequencing using Illumina HiSeq 2500] was conducted at the IGM Genomics Center, University of California, San Diego, La Jolla, CA.\n\n\nSupplementary materials\n\nSupplemental figure S1. A) Comparison of the gene expression fold-change profiles between T1DS and T2N in human fibroblasts and human iPS cells along all human chromosomes. (1) hiPSCs-L derived from monozygotic twins discordant for DS, Letourneau 2014 (red). (2) hiPSCs-H from the same monozygotic twins discordant for DS from Hibaoui 2014 (blue). (3) Human fibroblasts (hFibro-L) from monozygotic twins discordant for DS from Letourneau 2014 (black). hiPSCs-H lack GEDDs, while original hiPSCs-L and hFibro-L show GEDDs and high Spearman’s correlation (ρ1,3). B) Comparison of the gene expression fold-change profiles between 2N and Ts65Dn fibroblasts along all mouse chromosomes. Embryonic mice fibroblasts MEFs-D (blue), from 2N and Ts65DN mice do not show the GEDDs reported for MEFs-L (red).\n\nClick here to access the data.\n\n\nReferences\n\nLetourneau A, Santoni FA, Bonilla X, et al.: Domains of genome-wide gene expression dysregulation in Down's syndrome. Nature. 2014; 508(7496): 345–350. PubMed Abstract \| Publisher Full Text\n\nHibaoui Y, Grad I, Letourneau A, et al.: Modelling and rescuing neurodevelopmental defect of Down syndrome using induced pluripotent stem cells from monozygotic twins discordant for trisomy 21. EMBO Mol Med. 2014; 6(2): 259–277. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBusciglio J, Capone G, O'Bryan J, et al.: Down syndrome: genes, model systems, and progress towards pharmacotherapies and clinical trials for cognitive deficits. Cytogenet Genome Res. 2013; 141(4): 260–271. PubMed Abstract \| Publisher Full Text\n\nDavisson MT, Schmidt C, Reeves RH, et al.: Segmental trisomy as a mouse model for Down syndrome. Prog Clin Biol Res. 1993; 384: 117–133. PubMed Abstract\n\nReinholdt LG, Ding Y, Gilbert GJ, et al.: Molecular characterization of the translocation breakpoints in the Down syndrome mouse model Ts65Dn. Mamm Genome. 2011; 22(11–12): 685–691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAkeson EC, Lambert JP, Narayanswami S, et al.: Ts65Dn -- localization of the translocation breakpoint and trisomic gene content in a mouse model for Down syndrome. Cytogenet Cell Genet. 2001; 93(3–4): 270–276. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9541", "date": "23 Jul 2015", "name": "Jon Pierce-Shimomura", "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\nDo et al. perform statistical analysis on previously published RNAseq data from Letourneau et al. (2014) and their own new data to test the prominent and intriguing new hypothesis that trisomy 21 may cause up and down regulation of groups of genes associated with specific physical domains on different human chromosomes (Letourneau et al., 2014). These were called gene expression dysregulation domains (GEDDs). Additional evidence suggested that GEDDs may be conserved across human cell types, and surprisingly, may relate to equivalent syntenic domains in the mouse genome after analysis of the Ts65Dn mouse model of Down syndrome (DS) (Letourneau et al., 2014).Here, Do et al., first report that replicates from the Letourneau et al. (2014) dataset are more variable than one would expect as determined by principle component analysis (PCA). This is an important finding because identification of the purported GEDDs and their extrapolated conservation across tissue types and species depends on minimal variation across datasets. To better understand the significance of the apparent variation in replicate datasets (Figure 1A), it would be useful if Do et al. could also plot the percent variance aside each dot accounted for the first principal component. The procedure for data normalization should also be explained more thoroughly in the results section as this may significantly alter the apparent variance by PCA. Do et al. also find that data from the Letourneau et al., 2014 study varies significantly from data from a previous study by the same lab that used human cells derived from the same source in (Hibaoui et al., 2014).Do et al. then analyze how gene expression changes across physical locations on human chromosomes. They replicate the original finding by Letourneau et al. (2014) showing that there are domains of up and down regulated genes from human iPSCS and fibroblasts derived from the same source. However, they also find that these domains fail to correlate with data derived from the same source but published in an earlier study from the same group (Hibaoui et al., 2014). This finding raises significant doubt about the concept of conserved GEDDs if they cannot be replicated from tissue derived from the same individual and collected by the same research group. Do et al., might do well to suggest explanations for the lack of correlation including specific analysis techniques and methodologies.Lastly, Do et al. find that RNAseq datasets from wild-type and the Ts65Dn mouse model published in the (Letourneau et al., 2014) paper show considerable variation from their new set of mouse data as determined by PCA. They also show a lack of correlation between fold-change gene expression for both datasets across chromosomes.Together, this new analysis suggests a re-evaluation of the GEDDs concept related DS. Specific groups of physically-linked genes (domains) may indeed be up and down regulated in DS across individuals and perhaps in mouse models of DS. However, variation within and across RNAseq datasets appears to prevent defining these domains and generalizing them to other individuals and species with current methods of analysis.", "responses": [] }, { "id": "9542", "date": "01 Sep 2015", "name": "Roger Reeves", "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\nWidespread mis-regulation of the expression of disomic genes in trisomic genomes has been established for well over a decade (and seems a self-evident outcome of a trisomy such as that of Hsa21 which includes eight transcription factors, 29 microRNAs and a large number of lncRNAs among other key regulators). Recently, a pattern of chromosome regions showing up- or down- regulation of transcription in the trisomic genome based on RNAseq and substantiated with correlated regional changes in chromosome architecture was reported by Letourneau et al. (Letourneau et al., 2014). They referred to these regions as “gene expression dysregulated domains” or GEDDs. While the original analysis was carried out in fibroblast and iPS lines derived from a co-isogenic pair of monozygotic fetuses, one of which was euploid and the other of which had trisomy 21, Letourneau et al. report that the pattern of GEDDs is conserved in fibroblasts from the outbred Ts65Dn trisomic mouse model of Down syndrome (DS), i.e., the same blocks of disomic genes are up- or down-regulated in a mouse trisomic for orthologs of about half of the genes on Hsa21. These same effects are reported to be undetectable in a pooled comparison of 8 trisomic and 8 euploid cell lines, where they are hypothesized to be masked by normal variability in gene expression expected in individuals who are (obviously) not co-isogenic.In this report, Do et al. were unable to reproduce these GEDD patterns in an analysis of Ts65Dn mice similar to that which formed a part of the Letourneau evidence. Do et al. then proceeded to compare the Letourneau RNAseq results in this paper to RNAseq results generated previously by this group from cell lines of this same twin pair (Hibaoui et al., 2014) and found substantial variability in the results obtained in these independently done experiments. Based on several such comparisons in addition to the absence of GEDDS in their independent experiments, Do et al. conclude that there is not clear evidence for the existence of GEDDs.The comments by Pierce-Shimomura and Nordquist provide an excellent summary of several statistical issues in the analyses. They suggest possible clarifications to the Do study and conclude, with Do, that the existence of GEDDs is not proven. I would note only two additional two points.The PCA analysis in Do et al. comparing datasets from Letourneau and Hibaoui appears to show very strong batch effects, which is in fact the expected outcome for this type of comparison of two studies on different RNA sets prepared and run at different times, probably on different sequencers with different lots of reagents, without reference to placement in sequencing cells, etc. Statistical methods have been developed to clean data for this type of comparison (e.g., Leek JT, Johnson WE, Parker HS, Fertig EJ, Jaffe AE and Storey JD. sva: Surrogate Variable Analysis. R package version 3.16.0.). It is not clear that this has been done in the Do analysis. As in all analyses of large platform datasets, it is impossible to determine from the published and supplemental methods what has actually been done at a level of detail required to reproduce it independently – this applies to both studies. (For the exception proving this rule, consider Gilad Y and Mizrahi-Man O. A reanalysis of mouse ENCODE comparative gene expression data [v1; ref status: indexed, http://f1000r.es/5ez] F1000Research 2015, 4:121 (doi: 10.12688/f1000research.6536.1) ). However, it is immediately obvious that Letourneau et al. have no basis for many of the statistical comparisons used to identify GEDDs – a biological phenomenon - as they have no biological replicates, only technical ones. It was not clear to this reviewer whether Letourneau established four fibroblast clones and/or iPS lines from each twin (and if so, in how many independent transformation experiments) but this would only be a technical replicate for the artifact of that transformation process, not for the biology of GEDDs in trisomy, and therefore it is not the appropriate basis for the statistically-based conclusions that they make about the biology of the effects of trisomy 21 on transcription. For biological conclusions, there is an N of one euploid and one trisomic individual – statistical assessments are not possible with a single comparison. Letourneau et al. argue that genetic variability affecting gene expression levels does not allow detection of GEDDs in any but co-isogenic conditions. However, if GEDDs exist and are so highly conserved in evolution that the same genes are mis-regulated in the same way in outbred* trisomic mice, it is difficult to understand why they are not evident in any comparison of humans with two vs. three copies of Hsa21; if GEDDs only occur in vanishingly rare, coisogenic monozygotic human twin sets discordant for a specific trisomy, the phenomenon would hardly seem worthy of the attention it has received. An explanation for the conservation of this phenomenon in outbred individuals of another species but not between human pairs would strengthen the understanding of this phenomenon. An adequately described experiment comparing multiple individuals with two vs. three copies of Hsa21 including euploid and trisomic sib pairs, in addition to the statistical clarifications suggested by Pierce-Shimomura, would help to establish the existence or not of GEDDs and provide some indication of their relevance to the goal of ameliorating effects of gene dosage in DS. *Ts65Dn mice are maintained as an advanced intercross between any of several C57BL6 and C3H strains. Thus, individual mice and their sibs are not genetically identical and are heterozygous at ~50% of all loci - a different 50 % for each individual. SNPs occur about every 3000 bp between B6 and C3H, while SNPs for a given segment of a human chromosome pair might be on the order of one per 1000 bp. It might be important to consider variability in chr21 alleles in each twin. Assuming the original conceptus was trisomic, if trisomy resulted from a meiosis I error, the trisomic line will carry three sets of Hsa21 alleles while the euploid twin will lack one of the three sets. If the trisomy resulted from a meiosis II error, the euploid twin could be isodisomic for Chr21 and carry only a single set of alleles. One could speculate about the possible impact of isodisomy on genome-wide expression patterns.", "responses": [] } ]	1	https://f1000research.com/articles/4-269
https://f1000research.com/articles/4-264/v1	16 Jul 15	{ "type": "Review", "title": "A developmental biologist’s journey to rediscover the Zen of plant physiology", "authors": [ "José R. Dinneny" ], "abstract": "Physiology, which is often viewed as a field of study distinct from development, is technically defined as the branch of biology that explores the normal function of living organisms and their parts. Because plants normally develop continuously throughout their life, plant physiology actually encompasses all developmental processes. Viewing plant biology from a physiologist’s perspective is an attempt to understand the interconnectedness of development, form, and function in the context of multidimensional complexity in the environment. To meet the needs of an expanding human population and a degrading environment, we must understand the adaptive mechanisms that plants use to acclimate to environmental change, and this will require a more holistic approach than is used by current molecular studies. Grand challenges for studies on plant physiology require a more sophisticated understanding of the environment that plants grow in, which is likely to be at least as complex as the plant itself. Moving the lab to the field and using the field for inspiration in the lab need to be expressly promoted by the community as we work to apply the basic concepts learned through reductionist approaches toward a more integrated and realistic understanding of the plant.", "keywords": [ "Plant physiology", "Plant development", "plant" ], "content": "From development to physiology and back again\n\nI can remember, as an undergraduate, thinking of plant biology as being divided up into compartments just as the courses I was taking were divided: Physiology, Biochemistry, Genetics, Cell Biology, and Development. Biochemistry dealt with proteins, Genetics genes, Cell Biology cells, and Development the origins of form; but what did Physiology deal with? In our plant physiology course, we learned about hormones, hypocotyls, hydraulics, and the like. The experimental sections of the course were a bit more ‘old-fashioned’ than others, implementing cores of potato tubers to examine the process of osmosis. Development, on the other hand, was a particularly exciting field at that time. Mutants causing all sorts of surreal changes to the form of the plant were being identified, especially in the flower1. These pathways fit well into an elementary language of ABCs that is widely conserved across flowering plants2. Flower morphology has been used by systematists to classify the diversity of plant families, in part because of the highly stereotypic nature of these structures. Likewise, I think a lot of progress has been made in the molecular understanding of patterning mechanisms controlling flower development, in part because the process is so resilient to environmental cues. Results can easily be reproduced or confidently refuted by other labs, and standard growth conditions, constructed with simple equipment, can be established while accommodating modest budgets.\n\nIn graduate school, I followed my dream of studying morphogenesis and my studies identified JAGGED and NUBBIN, two genes that promoted growth of lateral organs. I hypothesized that these genes regulated organ shape by locally controlling the rate at which tissues differentiate3,4. But what does ‘differentiation’ actually mean? Developmental biologists often study the outcome of cell-type or organ-type specification pathways by studying the resultant structural features of the cells or tissues. Petals are clearly petals because they develop epidermal cells that are conical and have cuticular ridges, which radiate from the center of the cell1. But what else, besides structure, might be controlled downstream of these developmental pathways? Could the molecular composition of signaling pathways for environmental stimuli also be differentially regulated in each cell? Surely, a root hair cell would need to express different nutrient transporters than a root endodermal cell. Can development be seen as being integrated into a physiological context? Where did development end and physiology begin?\n\nTo address some of these questions, I embarked on a postdoc in the lab of Philip Benfey to determine where physiology and development meet. Recent advances in the use of fluorescence-activated cell sorting enabled cell-type resolution genome-scale transcriptional analyses in plants5. The method was being deployed to characterize the steady-state gene expression programs in the root. We had a general belief, not explicitly communicated, that whatever standard environmental conditions we were using to grow our roots, the transcriptional programs we were characterizing represented some sort of ground state. I wondered, however, whether this concept was incorrect and whether environmental conditions had a much deeper influence on cell type-specific transcriptional programs. I addressed this by profiling each tissue layer after the root was exposed to salt stress as a way of perturbing the system and seeing what changed6. With my colleague Terri Long, who generated a similar data set examining the response to iron deprivation, I performed a meta-analysis of the whole data set. Remarkably, the transcriptional programs that defined the functions of specific cell types in the root changed dramatically between conditions even though the expression of cell identity regulators generally did not change. What defined the epidermis was largely dependent on what environment the epidermis was exposed to. Thus, our concept of cell identity, which was based on a ground state, was too simplistic. Cell identity was a context in which environmental change was interpreted, not a fixed state.\n\nSince my postdoc, my group has been interested in understanding environmental responses by using tools that are often associated with studies in developmental biology. The main hypothesis driving this work is that cell type, organ type, and developmental stage influence how an environmental stimulus is interpreted7. Context informs environmental response.\n\n\nUnderestimating the plant\n\nI must admit that I had originally underestimated the complexity of environmental stress responses. The first underestimation was in time. Although our initial tissue-specific map analyzed the salt-stress response at a single time point, we now know that the root transitions through a series of strikingly different growth and transcriptional programs over the course of the stress treatment8. The second underestimation was in space. Root systems of adult plants are usually composed primarily of lateral root branches. The primary root, though important, essentially provides the foundation for this subsequent development, which is much more extensive and ultimately determines the efficiency with which water and nutrients are absorbed from the soil. I initially assumed that lateral roots would exhibit similar responses as the primary root to salt; however, we were pleasantly surprised that major specialization between root types was apparent in their growth and hormone signaling pathways9. The third and most recent underestimation came with the realization that roots are much more keenly aware of the spatial distribution of environmental cues than expected10,11. In Bao et al.10, we described how roots could sense the distribution of available water around the circumference of a single root tip and use this as a cue to bias the patterning of lateral root branches. Discovery of this phenomenon, termed hydropatterning, indicated that even under presumably homogeneous environmental conditions, such as the agar media in a petri dish, roots sense micro-scale differences in water availability to regulate many aspects of plant development likely to influence water and nutrient uptake.\n\nConsidering these experiences, I can see that I am likely to continue to underestimate the complexity of the plant-environment interaction. Part of this experience is inherent in the standard reductionist approach to science that is often needed to understand anything with a high degree of certainty. I also think, however, that some level of complacency can be involved. Working in highly controlled environments can be challenging, but nowhere near as challenging as studying plant physiology in the field. I have seen this first-hand as we collaborate with Andrew Leakey at the University of Illinois at Urbana-Champaign on water-deficit responses in Setaria viridis. Field experiments require specialized training and organizational skills to work with large teams of undergraduates and field workers who may be non-scientists. It also requires a longer-term perspective on experiments than is supported through many grants. Although such ecophysiological studies are needed, they need not be a requirement for all research questions. However, every plant biologist should have the experience of digging in the dirt if only to gain inspiration for how to design experiments in the lab that mimic aspects of the field environment12.\n\n\nThe new development is the old physiology\n\nDevelopment is the process by which form originates, but it is itself influenced by the a priori physical state of the organism. If environmental stimuli and physiological states of cells are constantly influencing development, then we see that development is one way that the plant responds to a change in environment and just one aspect of a larger physiological continuum. Form influences function, function provides a context for environmental response, environmental response regulates form, and so on in a cycle I like to call the ‘Zen of plant physiology’ (Figure 1).\n\nThe environment is sensed by the plant and leads to changes in the activity of signaling pathways that influence developmental decisions. Development determines the form of the organism and the function of cells that are specified and organized in tissues and organs. Shape, biophysical properties, and sensory pathways are controlled downstream of development and influence how future environmental change is experienced and responded to by the plant. Continuous development throughout the lifecycle of the plant results in a dynamic physiological state that influences, and is influenced by, development and the environment.\n\nThese concepts are actually quite well established in physiology, though perhaps not as well appreciated, given that interest in determining molecular mechanisms has taken over much of modern physiology. Researchers such as Paul Green understood that the physical properties of cells in a multicellular system resulted in complex emergent properties at the systems level that influence future events13. Green and colleagues found that the orientation of cellulose microfibrils changed as morphogenesis occurred in the meristem and they postulated that resultant stress patterns might influence future organogenesis. Recently, there has been a renaissance in appreciation for concepts initially established by Green14. In the pursuit of the origins of phylotaxy, the pattern by which new organ primordia initiate at the shoot apex, a realization that biophysical properties of tissues may play more than a permissive role in the process is becoming apparent15.\n\n\nMeeting future challenges through an integrative understanding of the plant\n\nThe classic physiological concept, of the organism as an integrated system, needs to be resurrected as a guide to modern molecular studies of mechanism. The need for such a perspective becomes obvious when considering how water-deficit responses are often studied. Agar-based media are poorly suited to mimic water deficit in a realistic way. Shoots are typically enclosed in a high-humidity environment that does not allow adaptive responses involving changes in stomatal aperture to conserve water. As soil dries, many physical changes occur that increase osmolarity, air space, and impedance to root growth16. Only osmolarity can be easily modified in gel-based media. Furthermore, heat often co-occurs with water deficit during droughts, and recent work suggests that plants may use temperature as a proxy for monitoring drought17. Currently, we are developing an integrated growth-and-imaging system that allows root structure and gene expression to be simultaneously imaged in soil-grown roots18. Our efforts are aimed at developing a system that preserves many aspects of plant growth in the field while being amenable to studying the mechanisms controlling changes in root growth.\n\nOur imaging system currently uses standard potting mix as the ‘soil-like’ media, and this is certainly another oversimplification of the diverse array of soil types that exist in nature and we will have to address this in the future as well. It is important to realize that the environmental conditions that plants experience have multidimensional complexity that matches or exceeds the complexity of the plant itself. The interface between plant and environment is multifaceted and we need to think of novel ways to characterize this interface. Microbiome studies of root- and shoot-associated microbial populations are an important start19–21. It may be possible to enlist the help of microbes and plant cells themselves in the effort to characterize the environmental inputs that plants experience. Development of chemical and physical biosensors will help in this regard as will determining the exact molecular mechanisms that plants use to sense the various environmental cues that are critical for their survival22,23.\n\nThe next decade in plant physiology looks extremely exciting and challenging. As classic non-model species become easier to explore at the molecular level, the design of growth and phenotyping platforms for these species will become ever more important to place these fine-scale measurements in context. Arabidopsis is small and easy to grow, whereas crops usually require special growth facilities with adequate lighting and nutrients. Larger organs that lack optical transparency prevent many of the standard microscopy methods used in model species.\n\nTwo paradigms are currently being established regarding plant phenotyping and they diverge in their approach to solving the cost-versus-throughput challenge. LemnaTec and others are developing plant growth and imaging platforms that move plants on conveyor systems to deliver hundreds of individuals to stations where watering, treatments, and imaging are performed automatically24,25. Alternative approaches using low-cost equipment such as flat-bed scanners or modified cell-phone cameras deployed en mass allow phenotype-genotype relationships to be explored at a small fraction of the cost26,27. Cost and physiological relevance need to be balanced, however, if the data obtained are aimed at identifying allelic variants that affect adaptive responses. Simulating environmental conditions is an art that needs further attention. How precise must lighting, temperature, and soil conditions be to fool the plant into responding to environmental stimuli as they occur in nature? If we are not mindful of the plant environment we study, we may find that, in the end, we have only been fooling ourselves.", "appendix": "Competing interests\n\n\n\nThe author declares that he has no competing interests.\n\n\nGrant information\n\nWork on plant-environment responses in plants is funded through grants from the National Science Foundation (#MCB-1157895 and IOS-PGRP #420-40-45A) and from the Department of Energy Biological and Environmental Research program (#DE-SC0008769).\n\n\nAcknowledgments\n\nI would like to thank Wei Feng and Neil E. Robbins II for helpful comments during the writing of this manuscript.\n\n\nReferences\n\nBowman JL: Arabidopsis: An Atlas of Morphology and Development. (J. L. Bowman, Ed.). New York: Springer-Verlag New York, Inc. 1994. Publisher Full Text\n\nCoen ES, Meyerowitz EM: The war of the whorls: genetic interactions controlling flower development. Nature. 1991; 353(6339): 31–7. PubMed Abstract \| Publisher Full Text\n\nDinneny JR, Weigel D, Yanofsky MF: NUBBIN and JAGGED define stamen and carpel shape in Arabidopsis. Development. 2006; 133(9): 1645–55. PubMed Abstract \| Publisher Full Text\n\nDinneny JR, Yadegari R, Fischer RL, et al.: The role of JAGGED in shaping lateral organs. Development. 2004; 131(5): 1101–10. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBirnbaum K, Shasha DE, Wang JY, et al.: A gene expression map of the Arabidopsis root. Science. 2003; 302(5652): 1956–60. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nDinneny JR, Long TA, Wang JY, et al.: Cell identity mediates the response of Arabidopsis roots to abiotic stress. Science. 2008; 320(5878): 942–5. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nDinneny JR: Analysis of the salt-stress response at cell-type resolution. Plant Cell Environ. 2010; 33(4): 543–51. PubMed Abstract \| Publisher Full Text\n\nGeng Y, Wu R, Wee CW, et al.: A spatio-temporal understanding of growth regulation during the salt stress response in Arabidopsis. Plant Cell. 2013; 25(6): 2132–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDuan L, Dietrich D, Ng CH, et al.: Endodermal ABA signaling promotes lateral root quiescence during salt stress in Arabidopsis seedlings. Plant Cell. 2013; 25(1): 324–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBao Y, Aggarwal P, Robbins NE 2nd, et al.: Plant roots use a patterning mechanism to position lateral root branches toward available water. Proc Natl Acad Sci U S A. 2014; 111(25): 9319–24. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRobbins NE 2nd, Dinneny JR: The divining root: moisture-driven responses of roots at the micro- and macro-scale. J Exp Bot. 2015; 66(8): 2145–54. PubMed Abstract \| Publisher Full Text\n\nLynch JP, Wojciechowski T: Opportunities and challenges in the subsoil: pathways to deeper rooted crops. J Exp Bot. 2015; 66(8): 2199–210. PubMed Abstract \| Publisher Full Text\n\nGreen PB: Organogenesis-A Biophysical View. Annu Rev Plant Physiol. 1980; 31(1): 51–82. Publisher Full Text\n\nHamant O, Traas J: The mechanics behind plant development. New Phytol. 2010; 185(2): 369–85. PubMed Abstract \| Publisher Full Text\n\nHamant O, Heisler MG, Jönsson H, et al.: Developmental patterning by mechanical signals in Arabidopsis. Science. 2008; 322(5908): 1650–5. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nVerslues PE, Agarwal M, Katiyar-Agarwal S, et al.: Methods and concepts in quantifying resistance to drought, salt and freezing, abiotic stresses that affect plant water status. Plant J. 2006; 45(4): 523–39. PubMed Abstract \| Publisher Full Text\n\nJohansson H, Jones HJ, Foreman J, et al.: Arabidopsis cell expansion is controlled by a photothermal switch. Nat Commun. 2014; 5: 4848. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRellán-Álvarez R, Lobet G, Lindner H, et al.: Multidimensional mapping of root responses to soil environmental cues using a luminescence-based imaging system. bioRxiv. 2015. Publisher Full Text\n\nBulgarelli D, Rott M, Schlaeppi K, et al.: Revealing structure and assembly cues for Arabidopsis root-inhabiting bacterial microbiota. Nature. 2012; 488(7409): 91–5. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKembel SW, O'Connor TK, Arnold HK, et al.: Relationships between phyllosphere bacterial communities and plant functional traits in a neotropical forest. Proc Natl Acad Sci U S A. 2014; 111(38): 13715–20. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLundberg DS, Lebeis SL, Paredes SH, et al.: Defining the core Arabidopsis thaliana root microbiome. Nature. 2012; 488(7409): 86–90. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMonshausen GB, Gilroy S: Feeling green: mechanosensing in plants. Trends Cell Biol. 2009; 19(5): 228–35. PubMed Abstract \| Publisher Full Text\n\nJones AM, Grossmann G, Danielson JÅ, et al.: In vivo biochemistry: applications for small molecule biosensors in plant biology. Curr Opin Plant Biol. 2013; 16(3): 389–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFinkel E: Imaging. With 'phenomics,' plant scientists hope to shift breeding into overdrive. Science. 2009; 325(5939): 380–1. PubMed Abstract \| Publisher Full Text\n\nAraus JL, Cairns JE: Field high-throughput phenotyping: the new crop breeding frontier. Trends Plant Sci. 2014; 19(1): 52–61. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nSlovak R, Göschl C, Su X, et al.: A Scalable Open-Source Pipeline for Large-Scale Root Phenotyping of Arabidopsis. Plant Cell. 2014; 26(6): 2390–403. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMutka AM, Bart RS: Image-based phenotyping of plant disease symptoms. Front Plant Sci. 2014; 5: 734. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "9536", "date": "16 Jul 2015", "name": "Christophe Maurel", "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", "responses": [] }, { "id": "9537", "date": "16 Jul 2015", "name": "David Salt", "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", "responses": [] }, { "id": "9538", "date": "16 Jul 2015", "name": "Jan Traas", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-264
https://f1000research.com/articles/4-144/v1	05 Jun 15	{ "type": "Method Article", "title": "Longitudinal variations of brain functional connectivity: A case report study based on a mouse model of epilepsy", "authors": [ "A. Erramuzpe", "J. M. Encinas", "A. Sierra", "M. Maletic-Savatic", "A.L. Brewster", "Anne E. Anderson", "S. Stramaglia", "Jesus M. Cortes", "A. Erramuzpe", "J. M. Encinas", "A. Sierra", "M. Maletic-Savatic", "A.L. Brewster", "Anne E. Anderson", "S. Stramaglia" ], "abstract": "Brain Functional Connectivity (FC) quantifies statistical dependencies between areas of the brain.FC has been widely used to address altered function of brain circuits in control conditions compared to different pathological states, including epilepsy, a major neurological disorder. However, FC also has the as yet unexplored potential to help us understand the pathological transformation of the brain circuitry.Our hypothesis is that FC can differentiate global brain interactions across a time-scale of days. To this end, we present a case report study based on a mouse model for epilepsy and analyze longitudinal intracranial electroencephalography data of epilepsy to calculate FC across three stages: 1, the initial insult (status epilepticus); 2, the latent period, when epileptogenic networks emerge; and 3, chronic epilepsy, when unprovoked seizures occur as spontaneous events.We found that the overall network FC at low frequency bands decreased immediately after status epilepticus was provoked, and increased monotonously later on during the latent period. Overall, our results demonstrate the capacity of FC to address longitudinal variations of brain connectivity across the establishment of pathological states.", "keywords": [ "Longitudinal study", "brain functional connectivity", "mouse model", "temporal lobe epilepsy", "mouse brain connectivity" ], "content": "Introduction\n\nFunctional Connectivity (FC) quantifies the statistical similarities between brain areas1. FC measures the influences between areas originated by different causes, such as two areas having a shared structural connectivity (wiring connections) or being driven by a common input. As such, studies based on FC are highly valuable for addressing disruptions of brain functioning in neurological disorders such as epilepsy, a major neurological disorder characterized by chronic unprovoked seizures. Indeed, FC studies in epilepsy are abundant2–7, but these studies typically perform group comparisons between health and disease; as an alternative to this approach, we present a longitudinal FC analysis on the same mouse brain across different days.\n\nOur general goal here is to unveil whether FC can account for differences in brain states, across the entire transition from a healthy brain to an epileptic one after an initial episode of status epilepticus. To the best of our knowledge, variations in FC across this transition have not addressed before. Our hypothesis is that FC can indeed differentiate those states.\n\nTo this aim, we introduce a setup based on a classical animal model of mesial temporal lobe epilepsy (MTLE), achieved by intra-hippocampal injection of kainic acid (KA)8,9. It is well-known in this model that after an initial provoked seizure, the latent period emerges and eventually, mouse brain’s resembles the main characteristics of human MTLE (see for instance10 and references therein). Using this validated model of epilepsy, we show herein that FC can indeed differentiate those states when applied to longitudinal data.\n\n\nMethods\n\nAll the experiments were performed employing a C57Bl/6 mouse (The Jackson Laboratory, Sacramento, CA). The animal was housed with ad libitum food and water access, in a 12:12h light cycle. All procedures were approved by the University of the Basque Country EHU/UPV Ethics Committees (Leioa, Spain) and Baylor College of Medicine Institutional Animal Care and Use Committee (Ethical approval number: AN5004; Houston, TX, USA). All animal procedures followed the European directive 2010/63/UE and NIH guidelines.\n\nFor this study, an adult mouse (male, 8 weeks old) was subjected to an intra-hippocampal injection of the glutamate agonist kainic acid (KA, 1nmol of KA in 50 nL, Sigma-Aldrich, St Louis, MO, USA), an experimental model that reliably reproduces the physiopathological features observed in human MTLE10,11.\n\nIn brief, the mouse was anesthetized with ketamine/xylazine (10/1 mg/kg) and received a single dose of the analgesic buprenorphine (1mg/kg) subcutaneously. After positioning in the stereotaxic apparatus, a 0.6mm whole was drilled at coordinates taken from Bregma: anteroposterior (AP) -1.7mm, laterolateral (LL) -1.6mm. A pooled glass microcapillary was inserted at -1.9mm dorsoventral (DV), and 50nL of saline or KA (20mM) were delivered into the right hippocampus using a microinjector (Nanoject II, Drummond Scientific, Broomal, PA, USA). After 2min, the microcapillary was retracted, and the mouse sutured and maintained in a thermal blanket until recovered from anesthesia.\n\nThe mouse was implanted with intracranial electrodes E363/8 platinum/iridium Teflon insulated (PlasticsOne, Roanoke, VA, USA), 0.005mm in diameter and 2mm in length, mounted in a plastic pedestal, which was secured to the skull with dental cement.\n\nAccording to Figure 1, four electrodes were implanted bilaterally in the motor cortex and hippocampus. The four electrodes were positioned at -0.1mm AP, +1.6mm LL, - 1mm DV (left cortex); -0.1mm AP, -1.6mm LL, -1mm DV (right cortex); -1.8mm AP, +1.6mm LL, -2mm DV (left hippocampus); -1.8mm AP, -1.6mm LL, -2mm DV (right hippocampus). The reference electrode was placed at the frontal lobe at +0.1mm AP, +0.1mm LL, -0.5mm DV, and the ground electrode was positioned over the cervical paraspinous area. Hereon, we labeled these electrodes as left cortex (LC), right cortex (RC), left hippocampus (LH) and right hippocampus (RH). The KA injection was applied at the site of the RH electrode (indicated by a red arrow in panel a).\n\na: Experimental setup. The intracranial placement of site recordings consisted on two electrodes placed bilaterally in the cortex (LC and RC, red) and two in the hippocampus (LH and RH, green). b: EEG recording was coupled to videographic recordings for visual confirmation of the seizure events. c1-c3: Examples extracted from the EEG recordings at the day of the injection (c1), the next day (c2) and after 21 days (c3). Overall changes in the electrical potential are shown in the upper row and after filtering for low frequency (1–14 Hz) in the lower row. The red dotted line marks high statistical similarities between electrodes, which provides high values of FC. Notice that RH is the site of the KA injection, and shows a higher epileptogenic activity that can be easily detected by looking at the amplitude of the time series associated to the RH electrode.\n\nRecording sessions had a duration of 4 hours and were performed every day during the first week and every other day for the following weeks with a Nicolet video-electroencephalogram (vEEG) system (NicView 5.71, CareFusion, San Diego, CA, USA). Recordings were first preprocessed with a 60 Hz notch filter and then passed through a (0.5–250) Hz bandwidth filter. Next, data was converted to ASCII using an EEG Converter for further analysis (EegSoft, Inc.). All postprocessing analysis was performed in Matlab (MathWorks Inc., Natick, MA).\n\nChanges in FC patterns were analyzed from these intracranial electroencephalographic data (EEG) across longitudinal sessions, from 0 days post KA injection (0 dpi) to 21 dpi.\n\nEpileptic seizures and artifact-free periods of interictal states were visually classified; seizures were identified according to repetitive-spikes and slow-wave discharges lasting 10 sec or more and synchronized with the behavioral stage 4–5 generalized seizures (monitored by video recording) according to the Racine scale12. Interictal discharges were measured as fast and high amplitude spike events lasting up to 200 msec.\n\nFC was addressed by calculating the correlation (C) and the partial correlations (PC) matrices between the time series electrode data. To calculate both C and PC, let xi be a column vector in which rows represent observations (time points) and i = LH,RH, LC,RC one possible electrode. Here, C and PC were calculated over non-overlapping windows of 1250 time points of interictal activity each (which is equivalent to having time windows of 5 seconds duration, as the sampling frequency was 250 Hz). Then, we build the data set matrix as X ≡ [xLC xRC xLH xRH], a matrix with dimension 1250 times 4.\n\nCalculation of C. Given X, each element matrix Cij is defined as the Pearson’s correlation coefficient between the time-series xi and xj, with i, j = LH,RH, LC,RC. Here, C was calculated using the corr function in Matlab (MathWorks Inc., Natick, MA). In particular, we run C = corr(X), which returns a matrix with dimension 4 times 4. Each element matrix satisfies that -1 ≤ Cij ≤ 1, with high and positive Cij meaning that the two time series xi and xj are correlated, high and negative values means anticorrelated and Cij ≈ 0 that the two time series are statistically uncorrelated, ie., independent.\n\nCalculation of PC. Given X and assuming C to be an invertible matrix, each element PCij is defined as −PijPiiPjj, where P ≡ C-1 is the inverse of the correlation matrix (ie. the so-called precision matrix). Notice that again by construction of PC, we have -1 ≤ PCij ≤ 1. Here, PC was calculated using the partialcorr function incorporated in Matlab (MathWorks Inc., Natick, MA), running the code PC = partialcorr(X), which similar than C has a dimension of 4 times 4.\n\nIt is important to emphasize that when comparing C to PC, high values of Cij are possible due to the presence of common neighbors to both i and j, ie., coming from z ≠ i, j, but PC removes that correlation contribution coming from those other neighbors13.\n\nFC across days. For the following days: 0,1,2,7,14 and 21 dpi, we averaged both C and PC over eight different non-overlapping windows of size 1250 time points each. FC values (mean and standard deviation) were calculated for all days. The raw data corresponding to the eight time windows and all the different days are available below (Dataset 1–Dataset 6).\n\nMotivated by a previous study of synchronization clusters in human temporal lobe epilepsy14, we introduced a network synchronization index that we named the Network Connectivity Index (NCI), which accounted for all electrode interactions, c.f., Figures 2c,d. To calculate NCI, we summed all the absolute values of all matrix elements in either C or PC divided by N(N - 1), a normalization factor equal to the total number of pairs contained in the sum minus the principal diagonal elements; thus, the NCI ignores all diagonal elements Cii and PCii, as they are equal to 1 in both C and PC matrices.\n\na,b: C and PC matrices across different days post KA injection and different frequency bands: low freq (1–14 Hz) and high freq (25–70 Hz). c,d: For the matrices plotted in panels a and b, we calculated the network connectivity index (for definition see methods) and represented across different days and frequency bands. Asterisks mean, for each condition respect to dpi0 (control), statistical significance differences with pvalue smaller than 0.05. C (and to a smaller extent PC) clearly differentiate brain states across days in the lower frequency band (blue line), showing a strong decrement at dpi1 and afterwards, FC started to increase until dpi21.\n\nFor each of the eight non-overlapping windows we calculated one value of NCI. Statistical significance differences between the NCI values at dpi0 (control) in respect to other conditions (1,2,7,14 and 21 dpi) were addressed by performing a paired t-test of the hypothesis that the two data sets (8 values of NCI in each group) have a different mean. Here, the t-test was performed using the ttest function incorporated in Matlab (MathWorks Inc., Natick, MA).\n\nBrain electrophysiological signals are well-known to be a mixture of many different rhythms occurring each at a different time scale15; as a consequence, one electrode data contains activity which results from a superposition of different rhythms. Classical Berger’s criteria (see 15 and references therein) separates brain oscillations occurring within different frequency bands (delta 0.5–4 Hz; theta 4–8 Hz; alpha 8–12 Hz; beta 12–30 Hz; gamma > 30 Hz), the higher the frequency, the fastest the rhythm’s oscillations are.\n\nIn this study, rather than calculating FC measures directly from the raw data, we first band-pass filtered the data and pooled all the frequency classes in two regimes: one occurring at low frequency bands (1–14 Hz, thus merging together delta/theta/alpha bands) and another one focused on high frequency bands (gamma rhythms at 25–70 Hz). To perform band-pass filtering, we applied a zero-phase digital filter to the input data X (ie., the raw data), that depending on the minimum (1 Hz) and maximum (14 Hz) values of frequency to be filtered in, it allows to extract the output data Xlow, which contains the 1–14 Hz contribution of the original signal (see panels in Figures 2c1–c3 for visualization of both X and Xlow). Similarly, the same bandpass filter applied to X but with different minimum (25 Hz) and maximum (70 Hz) frequencies returned Xhigh. The Matlab code of the used function here (named BandPassFilter.m) is available below (Dataset 7). Notice that BandPassFilter.m uses as an input parameter the sampling frequency (here, 250 Hz) and that internally it uses the function filtfilt, incorporated in the default Matlab (MathWorks Inc., Natick, MA).\n\nIn summary, the FC measures (both C and PC and consequently NCI) were calculated over Xlow and Xhigh rather than on X. This selection of frequency bands was performed to show explicit differences in the dynamics between the two highly different regimes.\n\n\nResults\n\nThe setup represented in Figure 1 provided unique data to test our hypothesis that the FC analysis, when applied to longitudinal data, can differentiate between brain states. Taking as an input the recordings obtained from the four electrodes LC, RC, LH and RH, we calculated FC matrices based on C and PC on eight different time-windows (Methods). In all the days, averaging over the eight windows ensured an appropriate sample with regard to the variability in the FC estimation, as the coefficient variation defined as the ratio between the standard deviation and the mean value was around 0.1 or less.\n\nFigures 2a,b correspond to average matrices across the eight segments. The C analysis (Figure 2a) showed a strong non-linear behavior at low frequencies (left column of matrices), as FC values strongly decreased at 1 dpi and after this point the FC values started to increase up to 21 dpi. This tendency, which did not exist in the high-frequency regime (right column of matrices in Figure 2a), confirmed that FC significantly varied across brain states, as data come from a well-validated model of mouse epilepsy.\n\nNext, we quantified connectivity patterns by calculating the NCI (Figures 2c,d) which, by summing the absolute values of all C and PC values (methods), provides information of the overall network connectivity. We analyzed NCI across brain regions and days post KA injection (Figure 2a) and found that hippocampal electrodes LH and RH remained strongly correlated across the time period in comparison to electrodes in the cortex. In particular, LC and RC started highly anticorrelated at 0 dpi, but their mutual correlation was drastically decreased at 1 dpi to eventually start to increase again up to a highly correlated state at 21 dpi (Figure 2a).\n\nThe PC analysis in the low frequency regime was also able to differentiate between hippocampal electrodes LH and RH, as the FC value between these two electrodes was high across the experimental period. Furthermore, the PC analysis showed small variations across days in comparison to the C analysis, and this occurred for both high and low frequency regimes. This has a particular interest, as PC removes interactions in a given pair coming from common neighbors, the so-called indirect effects. Thus, the indirect effects captured by C but not by PC were dominant at low frequencies.\n\n\nDiscussion\n\nCan FC differentiate between brain states when applied to longitudinal data? To answer this question, we have made use of an animal model of MTLE to address the variations in FC across the transition from an initial episode of status epilepticus to seizure chronification. We addressed FC by calculating C and PC. C (but not PC) revealed interactions through common neighbors (i.e., network effects) at low-frequency bands. More precisely, the network index for C showed a strong drop-off in the overall brain connectivity at 1 dpi but it smoothly increased for several days afterwards. This tendency might be correlated with the latency period, namely, the time interval between the original brain insult and the clinical presentation of the first spontaneous seizure. During this latency period, the transition from a healthy brain into an epileptic one, or epileptogenesis, occurs due to changes in the molecular, cellular, and network properties of the brain in response to the initial precipitating event. What we particularly show here is that the NCI for C (and to a smaller extent PC) works as a readout of the changes in brain functioning that take place during the latency period. In the near future, we aim to correlate the FC results with studies at the molecular and cellular level, in an integrative approach to better understand the process of epileptogenesis to eventually open new venues for more efficient therapeutic strategies.\n\nWe are emphasizing the advantages of using animal models for studying epilepsy. Thus, rather than performing group comparison (health vs. epilepsy) as it is normally done when studying disease, our setup allowed us for addressing longitudinal variations in FC on the same animal from the initial episode of status epilepticus to chronic epilepsy. Although the results analyzed here correspond to a very limited sample (n=1), we firmly believe that the same analysis can be applied to larger samples, allowing for studying longitudinal group FC patterns rather than individual results.\n\nIn summary, FC calculated from intracranial electroencephalography works as a readout of brain functioning and provides a straightforward measure for studying the effect of biological alterations occurring at the molecular, cellular and physiological scales during the transition from a healthy brain into an epileptic one. Importantly, the FC analysis presented here is based on longitudinal recordings of the same experimental subject allowing a continuous and precise temporal resolution that makes the calculation of FC meaningful and robust. Finally, it is important to remark that our analysis, bringing together experimental, mouse-based disease, biological science and computational science, might help to pave the road for further, more elaborated, computational algorithms as tools for analysis and validation of other diseases.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw intracranial data for recording day dpi0, 10.5256/f1000research.6570.d4898916\n\nF1000Research: Dataset 2. Raw intracranial data for recording day dpi1, 10.5256/f1000research.6570.d4899017\n\nF1000Research: Dataset 3. Raw intracranial data for recording day dpi2, 10.5256/f1000research.6570.d4899118\n\nF1000Research: Dataset 4. Raw intracranial data for recording day dpi7, 10.5256/f1000research.6570.d4899219\n\nF1000Research: Dataset 5. Raw intracranial data for recording day dpi14, 10.5256/f1000research.6570.d4899320\n\nF1000Research: Dataset 6. Raw intracranial data for recording day dpi21, 10.5256/f1000research.6570.d4899421\n\nF1000Research: Dataset 7. Band Pass Filter, 10.5256/f1000research.6570.d4899522", "appendix": "Author contributions\n\n\n\nConceived and designed the research: JME, AS, SS, JMC; Performed the acquisitions: AS, ALB; Analyzed the data: AE; Wrote the paper: AE, JME, AS, MMS, ALB, AEA, SS and JMC. All authors have read and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors have declared no competing interests.\n\n\nGrant information\n\nWork supported by Ikerbasque: The Basque Foundation for Science, Gobierno Vasco (Saiotek SAIO13-PE13BF001) and Euskampus at UPV/EHU to J.M.C.; Ikerbasque Visiting Professor and Bizkaia Talent (AYD-000-285) to S.S.; a Pre-doctoral contract from the Basque Government (Eusko Jaurlaritza), grant PRE/2014/1/252, to A.E.; the Spanish Ministry of Economy and Competitiveness with FEDER funds to J.M.E. (SAF2012-40085 and RYC-2012-11137) and A.S. (BFU2012-32089 and RYC-2013-12817); to the Basque Government (Saiotek SPC12UN014) and Ikerbasque start-up funds to J.M.E. and A.S; NIH Intellectual and Developmental Disabilities Research Grant (P30HD024064) and Dana Foundation, McKnight Endowment for Science Work grants to M.M-S; grants from NIH R01 NS, 39943 and 49427 to A.E.A; and T32 NS and 43124 to A.L.B., who in addition is a recipient of an Epilepsy Foundation Postdoctoral Fellowship.\n\n\nReferences\n\nFriston KJ: Functional and effective connectivity in neuroimaging: A synthesis. Hum Brain Mapp. 1994; 2(1–2): 56–78. Publisher Full Text\n\nBlumenfeld H, McNally KA, Vanderhill SD, et al.: Positive and negative network correlations in temporal lobe epilepsy. Cereb Cortex. 2004; 14(8): 892–902. PubMed Abstract \| Publisher Full Text\n\nWaites AB, Briellmann RS, Saling MM, et al.: Functional connectivity networks are disrupted in left temporal lobe epilepsy. Ann Neurol. 2006; 59(2): 335–343. PubMed Abstract \| Publisher Full Text\n\nBettus G, Guedj E, Joyeux F, et al.: Decreased basal fMRI functional connectivity in epileptogenic networks and contralateral compensatory mechanisms. Hum Brain Mapp. 2009; 30(5): 1580–1591. PubMed Abstract \| Publisher Full Text\n\nChavez M, Valencia M, Navarro V, et al.: Functional modularity of background activities in normal and epileptic brain networks. Phys Rev Lett. 2010; 104(11): 118701. PubMed Abstract \| Publisher Full Text\n\nWu GR, Chen F, Kang D, et al.: Multiscale causal connectivity analysis by canonical correlation: theory and application to epileptic brain. IEEE Trans Biomed Eng. 2011; 58(11): 3088–3096. PubMed Abstract \| Publisher Full Text\n\nBernhardt BC, Hong S, Bernasconi A, et al.: Imaging structural and functional brain networks in temporal lobe epilepsy. Front Hum Neurosci. 2013; 7: 624. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchwarcz R, Zaczek R, Coyle JT: Microinjection of kainic acid into the rat hippocampus. Eur J Pharmacol. 1978; 50(3): 209–220. PubMed Abstract \| Publisher Full Text\n\nBen-Ari Y: Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience. 1985; 14(2): 375–403. PubMed Abstract \| Publisher Full Text\n\nSierra A, Martin-Suárez S, Valcarcel-Martin R, et al.: Neuronal hyperactivity accelerates depletion of neural stem cells and impairs hippocampal neurogenesis. Cell Stem Cell. 2015; 16(5): 488–503. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBouilleret V, Ridoux V, Depaulis A, et al.: Recurrent seizures and hippocampal sclerosis following intrahippocampal kainate injection in adult mice: electroencephalography, histopathology and synaptic reorganization similar to mesial temporal lobe epilepsy. Neuroscience. 1999; 89(3): 717–729. PubMed Abstract \| Publisher Full Text\n\nRacine RJ: Modification of seizure activity by electrical stimulation. II. Motor seizure. Electroencephalogr Clin Neurophysiol. 1972; 32(3): 281–294. PubMed Abstract \| Publisher Full Text\n\nMäki-Marttunen V, Diez I, Cortes JM, et al.: Disruption of transfer entropy and inter-hemispheric brain functional connectivity in patients with disorder of consciousness. Front Neuroinform. 2013; 7: 24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPalmigiano A, Pastor J, Garcia de Sola R, et al.: Stability of synchronization clusters and seizurability in temporal lobe epilepsy. PLoS One. 2012; 7(7): e41799. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuzsaki G: Rhythms of the Brain. Oxford University Press, 2006. Publisher Full Text\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 1 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 2 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 3 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 4 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 5 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 6 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source\n\nErramuzpe A, Encinas JM, Sierra A, et al.: Dataset 7 in: Longitudinal variations of brain functional connectivity: A case report study based on a mouse model for epilepsy. F1000Research. 2015. Data Source" }	[ { "id": "8921", "date": "15 Jun 2015", "name": "Roma Siugzdaite", "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 investigated functional connectivity (FC) changes during different time points in one mouse brain with the lesion to hippocampus. The study is very interesting and informative in several ways:they tracked longitudinal variations of FC through 3 weeks after the lesion was induced; they investigated three different stages of epilepsy; they introduced a network synchronization measure NCI which takes into account interactions between all electrodes; they divided frequencies into low (1-14Hz) and high (25-70Hz) frequency bands, in which they found interactions revealed through common neighbours; NCI was reflecting Correlation results in low frequency band.Even though the study is the first step to understand dynamics and (probably in the future the mechanism) of epilepsy using functional connectivity tools, I have few concerns.First of all since it is a case study, we are looking only at one animal data. We have to keep it in mind. The three time points: initial insult, latent period and chronic epilepsy are described as the interval of interest. Identification of interictal states are described in methods, but in the results this information was not taken into account. How number of seizures were distributed in different stages and how it could influence changes in FC? In the experimental protocol part when the injection is explained there is a mistake \"saline or KA were delivered\". It was only one mouse, so I suppose that only Kainic Acid was injected (and not saline, was mostly used for control condition). Since T-test is performed on a small sample size, it would be good to calculate also the effect size. In the results part the sentence starts \"We analysed NCI across brain regions...\" and then authors talk about correlations from electrodes in hippocampus, that is in C analysis, but not in NCI. Personally for me the results are very interesting, especially at DPI 0. We would expect to see differences in connections RH to RC, and from there RH with other regions in case of hippocampus damage, but it's not described in the discussion at all. Methodologically paper is very good, though some physiological explanations or hypothesis in discussion would be useful. For example, why correlations remain strong despite the lesion in RH with LH in low freq; and between LH and LC in high frequency?This article is acceptable, although these minor corrections should be made.", "responses": [ { "c_id": "1454", "date": "16 Jul 2015", "name": "Jesus Cortes", "role": "Author Response", "response": "First of all since it is a case study, we are looking only at one animal data. We have to keep it in mind.We fully agree with the reviewer that these results are based on a single animal study. Indeed, this fact has been stated explicitly in several occasions throughout the text. Therefore, we are aware that these results need for a further validation. However, we really would like to emphasize that what we are presenting here is a new methodologically setup to tackle brain connectivity on same brain in the transition from a healthy to pathological brain. The three time points: initial insult, latent period and chronic epilepsy are described as the interval of interest. Identification of interictal states are described in methods, but in the results this information was not taken into account. How number of seizures were distributed in different stages and how it could influence changes in FC?We agree with the reviewer that it would be very interesting to correlate the number of seizures with the FC across the different stages (initial insult, latent period, chronic epilepsy); however, our methodological approach does not allow us to properly estimate the number of seizures per day when seizures are less frequent because of the low sampling (4h recording session per day). To avoid misguiding the readers, we have put less emphasis in these three periods in the abstract, and incorporated the new sentence:“To this end, we present a case report study based on a mouse model for epilepsy and analyze longitudinal intracranial electroencephalography data of epilepsy to calculate FC changes from the initial insult (status epilepticus) and over the latent period, when epileptogenic networks emerge, and at chronic epilepsy, when unprovoked seizures occur as spontaneous events.” In the experimental protocol part when the injection is explained there is a mistake \"saline or KA were delivered\". It was only one mouse, so I suppose that only Kainic Acid was injected (and not saline, was mostly used for control condition).Thanks to the reviewer for this comment. We have corrected this error and removed “saline”. Since T-test is performed on a small sample size, it would be good to calculate also the effect size.We want to thank the reviewer for making us to clarify this. We have calculated the size effects by computing the Cohen’s d parameter (Cohen, 1988). When comparing the NIC value at dpi0 with other days, respectively, (dpi1, dpi2,dpi7, dpi14 and dpi 21), the Cohen’s d parameter based on C results was (2.9797, 2.9716, 1.4273, 0.7480, 0.5870). Similarly, for NCI based on PC, we found (3.3111, 3.2127, 2.9904 , 2.9513 , 2.8848), which indicated high size effects for both C/PC NCI indexes (the higher the Cohen’s distance, the bigger the size effects and in any case the size effects were high).This text has been added in the final paragraph of Results in the revised version.We also added a new reference:Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates. In the results part the sentence starts \"We analysed NCI across brain regions...\" and then authors talk about correlations from electrodes in hippocampus, that is in C analysis, but not in NCI.We thank the reviewer for this comment. We now edited the text and corrected this. Personally for me the results are very interesting, especially at DPI 0. We would expect to see differences in connections RH to RC, and from there RH with other regions in case of hippocampus damage, but it's not described in the discussion at all.We thank the reviewer for this insight. Sample size, n=1, makes difficult to generalize; however, we added the following paragraph in Discussion, as suggested by the reviewer:“It is tempting to speculate that the increasing synchronization of the network observed after the initial status epilepticus is driven by the reorganization of the circuitry which occurs during the latent period. For instance, synchronization is higher between right and left hippocampus, as is expected from their direct connection via the hippocampal commisure, while it is lower between hippocampus and the primary somatosensorial cortex, as their connection is more indirect through a relay in the entorhinal cortex. Furthermore, our results insinuate that this synchronization is exclusive of low frequency bands (1-14Hz, delta/theta/alpha bands combined together), suggesting an underlying specificity of the hippocampal circuitry. Nonetheless, we are cautious to reach biologically relevant conclusions with n=1 and we will test these hypotheses in the future in larger samples of data. Rather, our data serves as a proof of principle that longitudinal variations of brain functional connectivity detect changes in brain connectivity in the epileptic mouse brain during the reorganization of the hippocampal circuitry after the initial status epilepticus.” Methodologically paper is very good, though some physiological explanations or hypothesis in discussion would be useful. For example, why correlations remain strong despite the lesion in RH with LH in low freq; and between LH and LC in high frequency?This is indeed an interesting observation but difficult to be extrapolated any interpretation based on only n=1 data set. It is well known, as it is acknowledged now in the manuscript, that different electrophysiological rhythms have different behavioral and cognitive correlates, and the fact that hippocampus electrodes are much higher correlated at slow rhythms and cortex electrodes do correlate at high rhythms, although very striking, needs more data to have a clearer interpretation.This point also is related to previous concern, number 6." } ] }, { "id": "9189", "date": "24 Jun 2015", "name": "Ruedi Stoop", "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 manuscript provides evidence that for studying brain diseases, the use of animal models for brain connectivity analysis is opening promising avenues. In an animal model case study, the authors analyze here in the same brain the transition from health to disease. Their methodology reveals an important link between molecular and large-scale brain imaging. Despite the convincing results the authors obtain, the reader should be aware that these results, although methodologically consistent, are in need of further validation, as they are based on an isolated one-animal study.", "responses": [] } ]	1	https://f1000research.com/articles/4-144
https://f1000research.com/articles/4-261/v1	16 Jul 15	{ "type": "Research Article", "title": "SwissPalm: Protein Palmitoylation database", "authors": [ "Mathieu Blanc", "Fabrice David", "Laurence Abrami", "Daniel Migliozzi", "Florence Armand", "Jérôme Bürgi", "Françoise Gisou van der Goot", "Mathieu Blanc", "Fabrice David", "Laurence Abrami", "Daniel Migliozzi", "Florence Armand", "Jérôme Bürgi" ], "abstract": "Protein S-palmitoylation is a reversible post-translational modification that regulates many key biological processes, although the full extent and functions of protein S-palmitoylation remain largely unexplored. Recent developments of new chemical methods have allowed the establishment of palmitoyl-proteomes of a variety of cell lines and tissues from different species. As the amount of information generated by these high-throughput studies is increasing, the field requires centralization and comparison of this information. Here we present SwissPalm (http://swisspalm.epfl.ch), our open, comprehensive, manually curated resource to study protein S-palmitoylation. It currently encompasses more than 5000 S-palmitoylated protein hits from seven species, and contains more than 500 specific sites of S-palmitoylation. SwissPalm also provides curated information and filters that increase the confidence in true positive hits, and integrates predictions of S-palmitoylated cysteine scores, orthologs and isoform multiple alignments. Systems analysis of the palmitoyl-proteome screens indicate that 10% or more of the human proteome is susceptible to S-palmitoylation. Moreover, ontology and pathway analyses of the human palmitoyl-proteome reveal that key biological functions involve this reversible lipid modification. Comparative analysis finally shows a strong crosstalk between S-palmitoylation and other post-translational modifications. Through the compilation of data and continuous updates, SwissPalm will provide a powerful tool to unravel the global importance of protein S-palmitoylation.", "keywords": [ "S-palmitoylation", "palmitoyl-proteomes", "database", "proteomics", "Acyl-biotin exchange", "Acyl-RAC" ], "content": "Introduction\n\nS-palmitoylation is defined as the enzymatic attachment of a 16-carbon-chain palmitic acid to cysteine residues via a thioester bond1,2. S-palmitoylation increases the local hydrophobicity of proteins, driving their association with cellular membranes and their segregation to specific membrane domains3 and also affect protein stability and protein-protein interactions4–12. S-palmitoylation appears to occur in all eukaryotes since it was found in yeast, parasites, worm, flies and plants. Importantly, it is the only reversible lipid post-translational modification (PTM) identified to date13 and thus can dynamically regulate the function of proteins. For example, S-palmitoylation/S-depalmitoylation cycles control the shuttling of the small GTPases H-Ras and N-Ras between the Golgi and the plasma membrane thus regulating cell growth14,15. The dynamics of S-palmitoylation are mediated by the opposing activities of two families of enzymes: palmitoyltransferases (PATs), which catalyze the attachment of palmitate to specific cysteine residues while thioesterases detach it5,16,17 and Acyl Protein Thioesterases (APTS) which remove the acyl chain.\n\nPalmitoylated proteins are involved in key biological processes including ion transport, receptor function, membrane trafficking, signaling, cell growth, development, neuronal plasticity and immune response5,18–20. Not surprisingly S-palmitoylation has been linked to a variety of human diseases, including neurological diseases such as Huntington’s disease21, schizophrenia22, Alzheimer’s disease23 and cancer (gastric, bladder, lung, colorectal, carcinoma24–29). In addition, pathogens utilize host S-palmitoylation machinery to promote infection30.\n\nDespite the growing evidences of the importance of S-palmitoylation and the attractive therapeutic perspectives that it provides, the full extent of the S-palmitoylated proteomes is not well defined and our understanding of the mechanisms that regulates S-palmitoylation and its consequences on cellular and organismal levels function remains incomplete. This is due in part to several factors; technical difficulties in studying lipid modifications, the absence of a consensus sequence for S-palmitoylation identifiable with bioinformatics means and the lack of antibodies that would recognize S-palmitoylated cysteines, in analogy to phosphospecific antibodies. The field has therefore long relied almost exclusively on metabolic labeling with radioactive palmitate, immuno-precipitation of specific proteins, and autoradiography - sometimes requiring months of exposure times.\n\nNovel methods now allow the capture of cellular or tissue palmitoyl proteomes followed by their identification by mass spectrometry. So far two major methods have been used. The first is Acyl Biotin Exchange (ABE), and its derivative Acyl Resin Assisted Capture (Acyl-RAC)31–33. It provides a snapshot of the S-palmitoylated proteins in a cell or tissue. It is based on the selective cleavage of thioester bonds, as found between the palmitate moiety and the cysteine residue, using neutral hydroxylamine (HA), and the subsequent capture of the liberated thiol groups with a thiol-specific reagent coupled - directly or indirectly - to beads. The second method is based on the metabolic labeling of S-palmitoylated proteins in live cells with palmitate analogues containing an azide or alkyne group, and thus will reveal the proteins that have undergone S-palmitoylation during the labeling time. After cell lysis, the labeled proteins can be isolated using click chemistry and affinity capture34.\n\nSo far, 19 palmitoyl-proteomes have been published, identifying thousands of new putative S-palmitoylated proteins26,31,33–49, and many more are expected to be reported. Since these palmitoyl-proteomes were obtained from various cell lines and tissues originating from different species, the field is in need of integration of these datasets in order to compare them, identify orthologs, and analyze what conditions lead to changes in S-palmitoylation. Importantly, the above-described capture techniques not only differ in the population of S-palmitoylated proteins that they aim to capture but also in some drawbacks. For example, incomplete alkylation of proteins during ABE/Acyl-RAC isolation will lead to a poor enrichment ratio and thus false negatives, but this step is absent from the in vivo labeling followed by click chemistry protocol. Thus comparing studies performed with different techniques increases the confidence that a protein represents a true positive. Moreover, different studies have used different thresholds and criteria to establish their confidence levels. Finally, low-abundance proteins might be missed in studies in cells or tissues, while being enriched in studies on subcellular fractions.\n\nTo provide the scientific community with a tool to extract information from the comparison of different palmitoyl-proteomics studies, we have created SwissPalm, which combines results from large-scale palmitoyl-proteomics studies with curation from the literature of small S-palmitoylation studies. SwissPalm is a user-friendly web resource that allows users to search for proteins of interest through all published palmitoyl-proteomes, determine the predicted S-palmitoylation sites, identify orthologues, compare palmitoyl-proteomes and more. Combination of the available data raises the confidence that a protein of interest is indeed palmitoylated, as we have herein validated on the chaperone complex CCT. A palmitoylation database also provides the opportunity to compare with other databases, leading for example to the generalized cross-talk between palmitoylation and ubiquitination.\n\n\nMethods\n\nThe database was designed to offer a general solution for storing knowledge on protein S-palmitoylation obtained from different types of studies: from biochemical studies focusing on a specific protein, to large-scale analyses by mass spectrometry-based proteomics.\n\nProtein hits and sites. In SwissPalm, two main objects have been designed to store the information related to S-palmitoylation. First, a Hit represents the knowledge that a given protein (or isoform) has been found as S-palmitoylated in a given study. Second, a Site is defined by a Hit and the position in the related sequence where the S-palmitoylation event was identified. A given Hit can have none, one or several associated sites. Hit and Site records are labeled with a unique identifier SPalmH# and SPalmS#, respectively. The enzyme(s) catalyzing the S-palmitoylation/S-depalmitoylation reactions on a Hit or Site can also be indicated.\n\nProtein sequences and annotations. Protein sequences are the primary and central source of information of the SwissPalm database. The UniProt Knowledge Base (UniProtKB) was chosen because it is a stable and regularly updated resource for protein sequences. One protein entry in UniProtKB can contain one or several sequences corresponding to alternative products (isoforms) of a single gene. Therefore, in the database, proteins and isoforms represent two distinct reference objects, and the known isoform-specific S-palmitoylation information can be reported to one or the other. The SwissPalm database is built for a set of species for which S-palmitoylation events have been reported for at least one protein. For each of these species, UniProtKB/SwissProt - the manually curated part of UniProtKB - sequences are inserted in the database. UniProtKB/TrEMBL entries are taken into account only if a palmitoyl-proteome hit protein is not found in UniProtKB/SwissProt. Moreover, UniProtKB provides curated functional annotations on proteins, e.g. ubiquitination or phosphorylation, which is useful to compare to S-palmitoylation. Information extracted from UniProtKB entries includes subcellular localization and sequence features (topological domains, variants, post-translational modifications, etc.). Post-translational modification information was also retrieved from Phosphosite50.\n\nMappings to other protein databases like RefSeq or genome-specific databases (e.g. The Arabidopsis Information Resource (TAIR), Mouse Genome Informatics (MGI), or Saccharomyces Genome Database (SGD)) were obtained either through the UniProt mapping API or from the databases themselves and inserted in SwissPalm as protein references. Gene Ontology (GO) terms and annotations were downloaded from the GO website and inserted in our database. A full mapping of UniProtKB entries and GO terms was computed, taking into account the lineage of GO terms and stored in the database. Thus, all UniProtKB entries associated with a given GO term can be found.\n\nOrthologs of Hits were extracted from OMA (Orthologs MAtrix) groups51 and OrthoDB52 and included in the database. Orthology relations are indicated in both databases at the level of proteins. For each protein, orthology groups represented in OrthoDB are listed non-redundantly. In OrthoDB, one protein can present several orthologs in the same species but in different orthology groups. To simplify the orthology information from OrthoDB, orthology groups were sorted by their size (number of orthologs in the group) from the smallest to the biggest, and only the first encountered protein for each species was selected. OMA group orthologs were used to complement this resource. These orthologs were used in the application for three purposes: the comparison of palmitoyl-proteomes of different species and the comparison of palmitoyl acyl transferases (PATs) from different organisms, and finally the analysis of conservation of known S-palmitoylation sites across different species.\n\nDue to the increasing number of palmitoyl-proteome studies and of known involved proteins and orthologs, the comparison of palmitoyl-proteomes had to be precomputed to keep the website responding fast. The results of the comparison are stored in a specific table of the database, which is queried to present results in the palmitoyl-proteome comparison tool. For each protein, two types of multiple sequence alignments (MSA) were performed (when relevant) using MAFFT []: a MSA of isoform sequences and a MSA of protein orthologs (taking into account only the main isoform if several exist). For each sequence and each cysteine, a prediction of S-palmitoylation sites has been performed using two tools: PalmPred53 and CSS-Palm 4.054. Results are stored in the database.\n\nSwissPalm is a standard Ruby-on-Rails (RoR) application with a PostgreSQL database backend. The update procedure is fully automated and includes every step of download, parsing and loading in the database of the external and computed data. It is possible to initially update the list of species to load in the system. The application consists of a public website to browse and search the S-palmitoylation data in different contexts and an administration interface to manually add any curated information. The search is powered by a systematic indexation system of all known protein identifiers, all possible key-words from the UniProtKB description of proteins and GO terms. Mass spectrometry data from literature can be loaded through a specific function part of the SwissPalm server used in admin mode. The system uses the database to directly load the lacking sequences and validate the possibility of protein S-palmitoylation. The original file (tab-delimited format) is kept on the server file system, and in the database only internal protein/isoform IDs (corresponding to UniProtKB entries or sequences) are stored. A dedicated interface has been set up to allow the curation of hits and sites extracted from the literature. Several validators help to keep consistency in the database; for example, a systematic check is done at the level of the sequence to verify the presence of a cysteine at the indicated position.\n\nGO term, STRING, DAVID and CORUM analysis. GO terms, coming from RDAVIDWebService55,56, with enrichment scores >1.6 and corrected p-values <0.01 (using a Benjamini-Hochberg correction for multiple testing) were selected. In order to cluster GO terms by group of ontology, GO terms with more than ten proteins but fewer than 500 were selected. GO terms sharing the same biological functions enriched in S-palmitoylation Hit proteins were clustered using the R environment and the results were visualized with RCytoscape57 and Cytoscape version 2.8.3 (Prefused forced direct layout)58. The distance between GO terms corresponds to the inverse numbers of proteins common to the two terms. Uniprot IDs corresponding to the human and mouse palmitoyl proteomes were searched against the STRING database version 9.159 for protein-protein interactions. Only interactions between the proteins belonging to the palmitoyl proteome dataset were selected using a confidence score ≥ 0.9 (high confidence). For CORUM analysis, complexes present in the CORUM database60 with more than six proteins were selected and classified by the % of S-palmitoylation Hit proteins in these complexes.\n\nTwo sample logo analysis. Sequence motif analyses were performed with the Two-Sample Logo software61 using the 535 known sites of S-palmitoylation present in SwissPalm. Sequences collected in this study were compared with a set of 5000 same-length protein sequences containing cysteine residues randomly selected but not annotated or predicted as S-palmitoylated. Two groups of aligned sequences were statistically analyzed by the binomial test (p < 0.01) against a binomial distribution.\n\nIn order to illustrate the potential of the SwissPalm database to increase the confidence level as to whether a given protein is palmitoylated or not, we have validated the palmitoylation of an identified protein complex, CCT. Similarly, the comparison of SwissPalm with ubiquitination databases suggested that proteins that can be palmitoylated also appear to be ubiquitinated. We therefore investigated this experimentally.\n\nCells, Antibodies and Reagents. HeLa cells (human uterine cervical carcinoma) from ATCC were grown at 37°C in complete modified Eagle's medium (MEM) (Sigma), supplemented with 10% fetal bovine serum (FBS) (Brunschwig), L-glutamine, penicillin and streptomycin. HAP1 knockout cell lines were purchased from Horizon Genomics (Vienna, Austria) and were grown in complete Dulbecco’s MEM (DMEM) (Gibco), supplemented with 10% FBS, 2 mM L-glutamine, penicillin and streptomycin. The DHHC2 clone (09818-03) contains a 2bp insertion in exon 1, the DHHC5 clone (30129-12) contains a 10bp insertion in exon 2 and the DHHC6 clone (13474-01) contains a 5bp insertion in exon 2. Human Anti CCT1 (ab109126- Rabbit monoclonal, RRID:AB_10864216), CCT2 (ab92746- Rabbit monoclonal, RRID:AB_10565196), CCT3 (ab167559- Mouse polyclonal), CCT4 (ab49151- Rabbit polyclonal, RRID:AB_2073761), and CCT5 (ab129016- Rabbit monoclonal, RRID:AB_11154964) were obtained from Abcam. Human anti-α Ubiquitin (P4D1) (mouse monoclonal) was from Santacruz (sc-8017, RRID:AB_628423) and anti human α-actin (Mouse monoclonal) from Millipore (MAB 1501, RRID:AB_1675188).\n\nImmunoprecipitation. HeLa cells were lysed for 30 min at 4°C in IP buffer (0.5% NP-40, 500 mM Tris–HCl pH 7.4, 20 mM EDTA, 10 mM NaF, 2 mM benzamidine, and Roche protease inhibitor cocktail) followed by centrifugation for 3 min at 2000 g. The supernatants were pre-cleared with protein G-agarose-conjugated beads (GE Healthcare) and incubated for 16 h at 4°C with antibodies and beads. The beads were washed three times with the immunoprecipitation buffer and resuspended in sample buffer (2 ×) after the final wash. The samples were heated at 95°C for 5 min and migrated on SDS–PAGE. Western blotting was performed using the iBlot (Invitrogen) according to the manufacturer's instructions.\n\nAcyl-RAC. Protein S-palmitoylation was assessed by the Acyl-RAC assay as previously described33, with some modifications. Briefly, total protein extracts (2 mg/ml) from cultured cells were incubated for 4 h at 42°C in buffer containing 1.25% SDS, 0.75% Triton X-100, Hepes 62.5 mM 1 mM EDTA, 20 mM methyl methanethiosulfonate (MMTS-Sigma), 8M Urea and protease inhibitor cocktail (Roche) at pH 7.4. MMTS was removed from the protein extract by chloroform–methanol precipitation followed by five methanol washes. Protein pellets were dried and solubilized in binding buffer (1% SDS, 1 mM EDTA, Hepes 100 mM, Urea 8M and protease inhibitor cocktail at PH 7.4). Supernatants were split into two: one sample was supplemented with 1 M hydroxylamine hydrochloride (pH 7.4) and 16.5 mg thiopropyl beads (Sigma); the other sample contained no hydroxylamine but 1M Tris (pH 7.4) and the same amount of thiopropyl beads. After overnight incubation at RT, beads were washed with washing buffer at least five times, and bound proteins were eluted by incubation of beads with 2 × SDS-PAGE loading buffer at 95°C for 5 min. Finally, samples were submitted to SDS-PAGE and analyzed by immunoblotting.\n\nMetabolic labeling with 3H-palmitic acid. HeLa cells were incubated for 2 h at 37°C in IM (Glasgow minimal essential medium buffered with 10 mM Hepes, pH 7.4) with 200 μCi/ml 3H-palmitic acid (9,10-3H(N)) (American Radiolabeled Chemicals, Inc.). Cells were washed and the cell lysate was subjected to immunoprecipitation of the protein of interest as described above. After the washes, beads were incubated for 5 min at 95°C in reducing sample buffer (2 ×) prior to SDS–PAGE. After the SDS–PAGE, the gel was incubated with a fixative solution (25% isopropanol, 65% H2O, 10% acetic acid), followed by a 30 min incubation with signal enhancer Amplify NAMP100 (Amersham). The dried gels were exposed to a Hyperfilm MP (Amersham). Chemical removal of S-palmitoylation was performed by treating cell extracts for 10 minutes at room temperature with 1 M hydroxylamine hydrochloride (Sigma) pH 7.4.\n\n\nResults\n\nCurated articles. By the time of submission, 303 published studies describing 365 S-palmitoylated proteins were incorporated into SwissPalm (http://swisspalm.epfl.ch/hits). Strict criteria were applied to consider a protein as S-palmitoylated: proteins must have been demonstrated as S-palmitoylated by two independent methods or by one method and mutagenesis of S-palmitoylation sites. As a result, 535 known S-palmitoylation sites were included in SwissPalm. In addition, palmitoyltransferases and thioesterases demonstrated to be involved in the S-palmitoylation and S-depalmitoylation of those proteins were annotated.\n\nPalmitoyl-proteomes. 19 palmitoyl-proteome screens using ABE-, Acyl-RAC- or click chemistry-based studies were selected from published literature (for details and references see: http://swisspalm.epfl.ch/studies?large_scale=1). They cover seven species and ten palmitoyl proteomes performed with ABE, 2 with Acyl-RAC and 7 with click chemistry (Figure 1A).\n\nA: Database content: Primary data on S-palmitoylation of proteins are extracted from MS large scale experiments on different species. Curated data on S-palmitoylation are obtained from the literature and input together with the MS information in the same data structure. In order to perform complex query related to S-palmitoylation, we have integrated in the database various external sources like orthology databases (Ortho_DB and OMA), UniProt features and subcellular localization information and Interpro domains. We ran also programs to have additional data, like multiple alignments of orthologous proteins or protein isoforms, and results from existing S-palmitoylation site predictors (CSS-Palm 4.0 and PalmPred). Web interface: The web interface presents the knowledge on S-palmitoylation in protein-centric pages. These pages are accessible through queries on a search engine. Some tools to analyse S-palmitoylation datasets are available online, like the orthologs comparison tool, aiming to perform cross-species S-palmitoylation comparison. B: 19 palmitoyl-proteomes from 7 species and various cell types and tissues were selected from published literature and integrated to SwissPalm. In total the dataset includes 5199 proteins.\n\nIn all cases, the specificity of the capture is controlled by the addition of hydroxylamine (HA). The enrichment ratio (+HA/-HA) for ABE and Acyl-RAC and (-HA/+HA) for click chemistry samples determines the likelihood of proteins to be S-palmitoylated and the cut off in these high-throughput datasets is defined arbitrarily. Accordingly to the vast majority of the palmitoyl-proteomes studies, a ratio (+HA/-HA) greater than two for either ABE, and Acyl-RAC, or a ratio (-HA/+HA) greater than two for click chemistry was selected to add proteins into SwissPalm. In some studies, authors have classified their dataset of S-palmitoylated proteins into high confidence (HC, (+HA/-HA ratio > 20), Medium confidence (MC, 5 < ratio < 20) and low confidence (LC, 2 < ratio < 5) groups. The same classification was kept and annotated in SwissPalm, according to each study. As a result, 5199 unique proteins were incorporated into SwissPalm, all of them showing a ratio +HA/-HA higher than two in each case. The human dataset contains 1453 annotated hits from six different cell types (DUI145, Platelet, B-cell, Umbilical vein blood, HEK 293, Jurkat) while the mouse dataset comprises 1747 proteins from six different cell types or tissues (RAW 264, 3T3-L1, Dendritic cell, Neuronal stem cell, T cell hybridoma, Brain extract) (Figure 1B).\n\nSwissPalm-mediated improvement of confidence. The information generated by independent palmitoyl-proteome studies was used to build a filter system that increases the hit confidence. First we annotated the proteins that are defined as high confidence by the authors of the studies (468 for human and 347 for mouse) (Figure 2A and Figure 2B, Supplementary table S1A and Supplementary table S1B). Second, we assume that the likelihood of a protein to be a true positive increases with its presence in multiple independent palmitoyl-proteomes. It is however important to keep in mind that even if proteins were isolated using the ABE/Acyl-RAC method of labeling followed by click chemistry, all proteins that contain a thioester bond, not related to S-palmitoylation, will be recovered and thus constitute false positives.\n\nA: Analysis of the 1372 human hits contained in 8 palmitoyl proteomes: 672 S-palmitoylation hits are present in at least 2 human palmitoyl proteomes or are annotated as “high confident hits” (HC). 204 are only found in at least 2 human palmitoyl proteomes, 136 are only classified as HC and 332 S-palmitoylation hits are both found in more than one palmitoyl-proteome and classified as HC. Out of the 672 hits, 345 are identified with 2 independent techniques. 63 out of the 672 S-palmitoylation hits have been validated in targeted studies, while 24 out of 700 hits only found in 1 human palmitoyl proteome have been validated. B: Analysis of the 1747 mouse hits contained in 6 palmitoyl proteomes as described in A. C: Analysis of the 2541 human orthologous hits contained in 19 palmitoyl proteomes as described in A. D: Number of the S-palmitoylation hits by the occurrence of palmitoyl-proteomes in which they have been identified.\n\n536 human proteins and 501 mouse proteins were found in at least two independent palmitoyl-proteomes (Figure 2A and Figure 2B, Supplementary table S1A and Supplementary table S1B). To exploit the information from palmitoyl-proteomes contained in other species, we looked for human orthologs of S-palmitoylated proteins from other species. Out of the orthologs of 2541 human hits in SwissPalm, 1160 were present in at least two independent palmitoyl proteomes (Figure 2C, Supplementary table S2). Well-studied S-palmitoylated proteins such as calnexin (CALX), Thioredoxin-related transmembrane protein 1 (TMX1), Synaptosomal-associated protein 23 (SNAP23), Ras-related protein (RRAS), Phosphatidylinositol 4-kinase IIα (P4K2A), transferrin receptor (TFR1), G-proteins (GNAI2, GNAI3), Ras related proteins were identified in multiple palmitoyl-proteome screens (at least in 10 out of 19) in at least four different species (Table 1, Supplementary table S3)6,50,62–67.\n\nStrikingly, ≈ 90% (1255 proteins) of the proteins present in more than one screen or annotated as high confidence hits have not yet been validated in targeted studies. In addition, more than 40% (524 proteins) of the proteins present in more than one screen were not annotated as high confidence hits in any of the studies (Figure 2D, Supplementary table S2). Finally, we annotated the proteins that were identified by two distinct independent techniques (metabolic labelling and ABE-based methods), the overlap of false positive proteins generated by these two techniques being very limited. This corresponded to 345 proteins for human and 387 proteins for mouse (Figure 2A and Figure 2B; Supplementary table S1A and Supplementary table S1B).\n\nAltogether, using multiple screens, independent methods and extending the Hits by an ortholog search, we have identified a list of high confidence S-palmitoylated protein hits, which could be useful for further study.\n\nSearch page. The web-based search tool is accessible from any page of the SwissPalm website: www.swisspalm.epfl.ch and produces as output a list of proteins with a direct link to detailed information for each protein (Figure 3A). Protein searches are initiated by submitting a protein name, gene name or Uniprot_ID in the query section and protein IDs that match the string search will be returned in the results sections. Protein searches can be restricted to specific criteria. These include: ‘species’, ‘presence in palmitoyl-proteome screens’, ‘predicted to be S-palmitoylated’ or ‘obtained from a third party application/tool’ (e.g. from a complex query on the UniProt website). Search can also be restricted to the “reviewed” annotated UniProtKB/SwissProt proteins. Batch searches are possible by submitting a list of identifiers in a tab/csv file (in the first column). An advanced search tool is available to help the user choosing among identifiers recognized by the search engine. Also, it presents controlled vocabularies that can be used to perform complex queries within the database (Motif Search, GO term and Subcellular localization).\n\nA: SwissPalm search page: Example of query for “calnexin” shows that it has been found in palmitoyl-proteomes from several species: human (7 out of 8 screens), mouse (6 out of 6), rat (1 out of 1) and Arabidopsis thaliana (1 out of 1). For human, mouse and rat calnexin was classified as a high confident hit and for human and mouse identified by two independent techniques (metabolic labeling and chemical capture). Finally, calnexin S-palmitoylation was also subject to targeted studies and 2 cysteine residues (502 and 503 in human calnexin) were identified. B: Results Page from human calnexin display summary boxes containing the main information related to S-palmitoylation: number of occurrences in palmitoyl-proteome screens and targeted studies, sites information, cysteine prediction.\n\nProtein information page. A system of summary boxes gives a quick overview on the main information related to S-palmitoylated proteins. This includes the number of times the protein is cited in proteome or targeted studies, information on experimental sites, and high confidence predictions from CSS-Palm 4.0 and PalmPred (Figure 3B). Other information includes a global alignment of isoform sequences highlighting all cysteine residues (Figure 4A), protein topology, disulfide bond positions and prediction scores (Figure 4B), information on orthologous proteins and an alignment of them (Figure 3D), as well as GO terms and references (cell types, techniques, subcellular localization).\n\nA: (upper) Global alignment of isoform sequences highlighting all cysteine residues. (lower) When available, information on protein topology, disulfide bond involvement and prediction scores from CSS-Palm 4.0 and PalmPred are provided for each cysteine residue in the different isoform sequences. B: Global alignment of orthologs sequences show conserved cysteine residues (502 and 503 in human calnexin) across species.\n\nEstimation of the human palmitome. We made use of the information gathered in SwissPalm to obtain a current estimation of the human palmitoyl-proteome. We found using Uniprot annotation that 6.8% of the human (1453 proteins) and 9.5% (1747) of the mouse proteomes may undergo S-palmitoylation (Figure 5A, Supplementary table S3). Since several identified S-palmitoylation sites are conserved across species, we integrated 891 mouse orthologs that were not present in human palmitoyl proteomes. The joint dataset indicates that 11% (2339 proteins) of the human proteome may undergo S-palmitoylation. Extending the analysis to 2649 human orthologs present across all species, 12.47% of the human proteome may be subject to S-palmitoylation (Figure 5A, Supplementary table S3). Moreover, since more than half of the proteins described in published targeted studies were not identified in any of the palmitoyl proteome screens (Figure 5B), the current figure of 9 to 12% is probably an underestimate.\n\nA: Percentage of S-palmitoylation hits combined from 8 human or 6 palmitoyl proteomes in the human and mouse proteomes. The analysis was extended to human orthologs of mouse and all S-palmitoylation hits present in the dataset. B: Percentage of targeted studies present in palmitoyl-proteomes. C: Topology of human and mouse S-palmitoylation hits. D and E: Distribution of human and mouse S-palmitoylation hits in cellular compartments. F: Enrichment of amino acid nearby validated S-palmitoylated cysteines in all, only cytoplasmic or only membrane proteins.\n\nProtein localization and S-palmitoylation motif. Using the UniProt annotation on protein localization, 45% of both the human and mouse S-palmitoylation hits (679 and 974 for proteins respectively) were annotated as membrane proteins, 30% as cytosolic (445 and 723 proteins respectively) (Figure 5D). This represents a modest enrichment in membrane proteins, more specifically of Type I membrane proteins, when compared to the total human and mouse proteomes (Figure 5C, Supplementary table S4, Supplementary figure S1). S-palmitoylation hits are found in all major compartments of the endomembrane system – endoplasmic reticulum, Golgi apparatus, lysosome, endosome – as well as in mitochondria (Figure 5E). The latter finding is intriguing since none of the DHHC palmitoyl transferases have so far been localized to mitochondria68.\n\nAdditionally, 377/1453 human and 422/1747 mouse proteins identified in palmitoyl proteomes were associated with the nucleus which represents around 25% of the total S-palmitoylated human and mouse proteins (Figure 5D). While these might be cytosolic proteins associated with the outer nuclear membrane, the higher number of nuclear proteins raises the possibility that S-palmitoylation may occur in the nucleus and could be used to reversibly target nuclear proteins to the inner nuclear membrane. Presence of a palmitoyltransferase in the nuclear envelope has been reported for the yeast protein rif169. Palmitoylation of nuclear proteins will clearly require validation.\n\nIn order to investigate whether certain amino acids were preferentially enriched around S-palmitoylation sites, we compared the amino acid environment of the 535 sites present in SwissPalm against a random set of cysteines not predicted as S-palmitoylated (Materials and methods). A two sample logo analysis shows that S-palmitoylated cysteines are preferentially surrounded by other cysteines, have hydrophobic amino acid to the N-terminal side and charged amino acid to C-terminal side of the S-palmitoylated cysteine (Figure 5F). We repeated the analysis, separating soluble from transmembrane proteins. With the exception of neighboring cysteines, no amino acid specificity was found for soluble proteins. For membrane proteins, despite the fact that we did not classify them as type I, II or multispanning, S-palmitoylation sites appear to preferentially locate to the C-termini of transmembrane domains, there appears to be multiple cysteines and these are preferentially followed with charged residues. This residues could contribute to the inside positive rule of membrane proteins and also ensure interaction with the cytosolic domain with negatively charged lipid head groups at the plasma membrane for example, as observed for myristoylated proteins70.\n\nThis analysis does not exclude the presence of specific motifs that would ensure S-palmitoylation by a given palmitoyltransferase. Such an analysis will require extensive knowledge of the specific target of the DHHC enzymes.\n\nAltogether this analysis indicates that S-palmitoylation is not restricted to a specific compartment and that it is slightly more frequent in membrane proteins.\n\nOntology and network analysis. To search for cellular functions enriched in S-palmitoylated proteins, we performed pathway- and network-based analysis on both human and mouse palmitoyl-proteomes. Gene ontology56,71 and protein-protein interaction analysis (through the STRING database) were performed on proteins found by two independent techniques or by targeted studies: 470 proteins for human and 443 for mouse (Supplementary figure S2).\n\nConsistent with the findings reported in some targeted studies, specific biological functions/processes (Figure 6A, Supplementary figure S3 and Supplementary figure S4, Supplementary table S6, methods) are enriched in potentially S-palmitoylated proteins: membrane organisation, cytoskeleton organization, protein localization, cell localization and cell signalling. For example, proteins involved in vesicular trafficking were enriched such as subunits of the coatomer complex, clathrin as well as a significant number of SNARE proteins (Supplementary table S5). We also found 289 hit proteins associated with the GO term Cytoskeleton. These include microfilament- and microtubule-related proteins such as actin subunits, dynein and vimentin but also proteins involved in the formation and regulation of these structures: profilin I, Arp2-Arp3 related proteins. Consistent with a role of S-palmitoylated proteins in cytoskeleton dynamics, tubulin S-palmitoylation was shown to contribute to its association with membrane72 (Supplementary table S5). More recently, S-palmitoylation of LIM Kinase-1 was shown to regulate spine-specific actin polymerization and thereby morphological plasticity73. The full view of the role of S-palmitoylation in cytoskeleton organization will clearly require further studies.\n\nA: GO term analysis of 470 human S-palmitoylation hits found by 2 independent techniques or by targeted studies. GO terms sharing the same biological functions enriched in S-palmitoylation Hit proteins are clustered. The distance between GO terms corresponds to the inverse numbers of proteins common to the two terms and the size of the circle to the number of proteins associated with the GO term. B: Protein-protein interactions networks analysis of 470 human S-palmitoylation hits found by 2 independent techniques or by targeted studies using STRING software. The interactions (high confidence score > 0.9) are shown in evidence view (pink: experimental evidences and blue database evidences).\n\nMore unexpected processes were also found to be enriched: regulation of cell death, metabolism, mRNA metabolic processes, generation of precursor metabolites and energy, ribonucleotides metabolic processes (Figure 6B, Supplementary table S5 and Supplementary table S6). Whether this is due to the abundance of these proteins or a functional requirement for S-palmitoylation remains to be investigated. It is tempting to speculate that S-palmitoylation might regulate the association of large cytosolic protein complexes, such as ribosomes or proteasomes with membranes to ensure an efficient regulation of protein synthesis and degradation (Figure 6B, Supplementary table S5)74.\n\nEnrichment in protein complexes. Intrigued by the presence of a large number of protein complexes enriched in S-palmitoylated proteins, we analysed the human and mouse palmitoyl proteome dataset using CORUM, a database of manually curated and validated mammalian protein complexes60. A clear enrichment of S-palmitoylated proteins in protein complexes was observed: 18% (463/2558) of the human and 23% (217/938) of the mouse S-palmitoylation hits were part of a complex compared to the 6.8% (1455/21267) of human and 9.5% (1747/18402) of S-palmitoylation hits in their corresponding proteomes. 27 human and 20 mouse complexes possess more than 50% of their subunits S-palmitoylated (Figure 7A and Figure 7B). Intriguingly, the majority of the components of the CCT micro complex, also called TCP-1 ring complex (TRiC), appeared to be S-palmitoylated in both human and mouse analyses (Figure 7C). The TRiC complex is a cytosolic molecular chaperone that promotes folding of 10 to 15 percent of cellular proteins75. It is composed of two identical stacked rings, each of which contains eight different subunits. We were able to experimentally confirm the S-palmitoylation of CCT components; CCT1 and CCT2 could be labeled with 3H-palmitate in a hydroxylamine-dependent manner (Figure 7D). These subunits, as well as CCT3, CCT4 and CCT5, also were positive for S-palmitoylation by Acyl-RAC. CCT subunits assemble into the TRiC complex but independent roles as individual subunits proteins have also been proposed76. What the functional consequence of this S-palmitoylation is remains to be established.\n\nA and B: Human and mouse protein complexes enriched in S-palmitoylation hits using the CORUM database. Proteins complexes containing more than 6 proteins and enriched by at least 50% of S-palmitoylation hits were selected. C: Representation of the TRiC complex subunits. Color circles represent the species in which each CCT subunit was identified as S-palmitoylated. The star indicated proteins identified by 2 independent techniques. D: Palmitoylation of the subunits was validated by Acyl-RAC on CCT1, CCT2, CCT3, CCT4 and CCT5 subunits and by 3H-palmitate labelling on CCT1 and CCT2 subunits. (TCE: Total cell extract, NH2OH: Hydroxylamine treatment, 3H-palm: radioactive palmitate signal, WB: Western blot signal).\n\nPTMs and S-palmitoylation. S-palmitoylation is not the sole PTM that can occur on cysteine residues. They can in particular also undergo other post-translational modifications on cysteine residues such as S-nitrosylation and S-glutathionylation77,78. Site competition can thus potentially occur and has been observed77,78. To assess the potential amplitude of this competition, we compared the datasets from two available databases, Db_SNO and Db-GSH, which contain proteomic hits for S-nitrosylation and S-gluthationylation, respectively79,80. Comparison of the three datasets shows that more than 40% (1031 out of 2483) of S-palmitoylated protein hits had at least one S-nitrosylated or one S-glutathionylated site and 19% (470 proteins) had both (Figure 8A). This strong overlap suggests that either the isolation methods share the same drawbacks or there is cross-talk between these PTMs.\n\nA: Venn diagram displaying the strong co-occurrence of S-palmitoylation, S-nitrosylation and S-glutathionylation modifications within the same proteins. For this analysis human orthologs of mouse proteins were included. B and C: Percentage of phosphorylation and ubiquitination sites in human and mouse S-palmitoylation hits compared to the total protein population. C: Ubiquitination pattern in WT and DHHC2, 5 and 6 HAP1 cells. Total protein extracts from HAP1 cells were subject to Western blot and probe using anti-ubiquitin antibody and actin as loading control.\n\nSeveral studies, including those from our laboratory, have reported cross-talk between S-palmitoylation and phosphorylation or ubiquitination7,12,81–88. To investigate whether this might be a general feature, we used the Phosphosite database to analyse the presence of identified phosphorylation or ubiquitination sites among S-palmitoylation hits50. We found using Uniprot annotation that 75% of the human hit proteins contained a phosphorylation site compared to 64% for total proteins. Strikingly 61% of S-palmitoylation hits contained an ubiquitination site, compare to 33% for the human proteome (Figure 8B). Similar ratios were found for the mouse S-palmitoylated proteins (Figure 8C). The high abundance of ubiquination sites in S-palmitoylation hits however does not imply that proteins become simultaneously S-palmitoylated and ubiquitinated. On the contrary we have found that S-palmitoylation tends to prevent ubiquitination, protecting membrane proteins for example from lysosomal or endoplasmic-reticulum associated degradation (ERAD) targeting. Consistent with our observations we found that ubiquitination was markedly increased in HAP1 cell lines knocked-out for the palmitoyltransferases ZDHHC2, ZDHHC5 or ZDHHC6 (using the CRISPR-CAS9 system) when compared to control cells (Figure 8D).\n\n\nConclusion\n\nAs the first database dedicated to protein S-palmitoylation, SwissPalm currently assembles 5199 proteins from 19 palmitoyl-proteomes, 7 species and 17 different tissues/cell lines. This database will be updated as new palmitoyl-proteomes appear in the literature and authors are encouraged to directly contact SwissPalm through the website (webmaster.bbcf@epfl.ch). By curating information from multiples screens, filters could be established to improve the confidence in positive hits from palmitoyl-proteomics. In addition, SwissPalm provides various tools to ease the study of S-palmitoylation (multiple alignments of orthologous proteins and their isoforms, site prediction score from CSS-Palm and PalmPred). The combined dataset reveals that S-palmitoylation is a widespread post-translational modification that affects key biological processes. SwissPalm, by collecting and sorting information from different sources in a single web page, will accelerate research on S-palmitoylation.\n\n\nSoftware availability\n\nThe code for the Ruby-on-Rails application for the SwissPalm database can be accessed online at https://github.com/bbcf/swisspalm\n\nhttps://github.com/F1000Research/swisspalm.git\n\nhttp://dx.doi.org/10.5281/zenodo.19106\n\nThe Ruby-on-Rails application for the SwissPalm database is distributed under the MIT license.", "appendix": "Author contributions\n\n\n\nMB and GVDG conceived the study. MB and LA designed and performed the experiments. FD designed and implements the database. DM curated articles. JB performed the Uniprot topology analysis. FA and MB designed and performed the Gene Ontology analysis. MB, FD and GVDG wrote the manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing interests.\n\n\nGrant 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 n. 340260 - PalmERa. GVDG also received support from the Swiss National Science Foundation grant (http://www.snf.ch), the National Centre of Competence in Research in Chemical Biology and the SystemsX.ch, the Swiss Initiative in systems Biology evaluated by the Swiss National Science Foundation (LipidX). MB was a recipient of Novartis Foundation for Medical Biological Research Fellowship and an EMBO Long term Fellowship.\n\nI confirm that the 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 Alan Bridge, Maria Eugenia Zaballa and Oksana Andrei Sergeeva for helpful advice.\n\n\nSupplementary materials\n\nHuman palmitoyl-proteome\n\nClick here to access the data.\n\nMouse palmitoyl-proteome\n\nClick here to access the data.\n\nOrthologs palmitoyl-proteome\n\nClick here to access the data.\n\nTopology of S-palmitoylated proteins\n\nClick here to access the data.\n\nGene Ontology analysis\n\nClick here to access the data.\n\nString network analysis\n\nClick here to access the data.\n\nGO term analysis of 443 mouse S-palmitoylation hits found by 2 independent techniques or by targeted studies. GO terms sharing the same biological functions enriched in S-palmitoylation Hit proteins are clustered. The distance between GO terms corresponds to the inverse numbers of proteins common to the two terms and the size of the circle to the number of proteins associated with the GO term.\n\nProtein-protein interactions networks analysis of 443 human S-palmitoylation hits found by 2 independent techniques or by targeted studies using STRING software. The interactions (high confidence score > 0.9) are shown in evidence view (pink: experimental evidences and blue database evidences).\n\n\nReferences\n\nSalaun C, Greaves J, Chamberlain LH: The intracellular dynamic of protein palmitoylation. J Cell Biol. 2010; 191(7): 1229–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlaskovic S, Blanc M, van der Goot FG: What does S-palmitoylation do to membrane proteins? FEBS J. 2013; 280(12): 2766–74. PubMed Abstract \| Publisher Full Text\n\nLevental I, Lingwood D, Grzybek M, et al.: Palmitoylation regulates raft affinity for the majority of integral raft proteins. Proc Natl Acad Sci U S A. 2010; 107(51): 22050–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbrami L, Kunz B, Deuquet J, et al.: Functional interactions between anthrax toxin receptors and the WNT signalling protein LRP6. Cell Microbiol. 2008; 10(12): 2509–19. PubMed Abstract \| Publisher Full Text\n\nFukata Y, Fukata M: Protein palmitoylation in neuronal development and synaptic plasticity. Nat Rev Neurosci. 2010; 11(3): 161–75. PubMed Abstract \| Publisher Full Text\n\nLakkaraju AK, Abrami L, Lemmin T, et al.: Palmitoylated calnexin is a key component of the ribosome-translocon complex. EMBO J. 2012; 31(7): 1823–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGauthier-Kemper A, Igaev M, Sündermann F, et al.: Interplay between phosphorylation and palmitoylation mediates plasma membrane targeting and sorting of GAP43. Mol Biol Cell. 2014; 25(21): 3284–99. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRossin A, Durivault J, Chakhtoura-Feghali T, et al.: Fas palmitoylation by the palmitoyl acyltransferase DHHC7 regulates Fas stability. Cell Death Differ. 2015; 22(4): 643–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong C, Lange JJ, Samovski D, et al.: Ubiquitination and degradation of the hominoid-specific oncoprotein TBC1D3 is regulated by protein palmitoylation. Biochem Biophys Res Commun. 2013; 434(2): 388–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhu YC, Li D, Wang L, et al.: Palmitoylation-dependent CDKL5-PSD-95 interaction regulates synaptic targeting of CDKL5 and dendritic spine development. Proc Natl Acad Sci U S A. 2013; 110(22): 9118–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLinder ME, Deschenes RJ: Palmitoylation: policing protein stability and traffic. Nat Rev Mol Cell Biol. 2007; 8(1): 74–84. PubMed Abstract \| Publisher Full Text\n\nAbrami L, Kunz B, Iacovache I, et al.: Palmitoylation and ubiquitination regulate exit of the Wnt signaling protein LRP6 from the endoplasmic reticulum. Proc Natl Acad Sci U S A. 2008; 105(14): 5384–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nResh MD: Covalent lipid modifications of proteins. Curr Biol. 2013; 23(10): R431–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRocks O, Peyker A, Kahms M, et al.: An acylation cycle regulates localization and activity of palmitoylated Ras isoforms. Science. 2005; 307(5716): 1746–52. PubMed Abstract \| Publisher Full Text\n\nLynch SJ, Snitkin H, Gumper I, et al.: The differential palmitoylation states of N-Ras and H-Ras determine their distinct Golgi subcompartment localizations. J Cell Physiol. 2015; 230(3): 610–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGreaves J, Chamberlain LH: DHHC palmitoyl transferases: substrate interactions and (patho)physiology. Trends Biochem Sci. 2011; 36(5): 245–53. PubMed Abstract \| Publisher Full Text\n\nFukata Y, Iwanaga T, Fukata M: Systematic screening for palmitoyl transferase activity of the DHHC protein family in mammalian cells. Methods. 2006; 40(2): 177–82. PubMed Abstract \| Publisher Full Text\n\nEl-Husseini AE, Craven SE, Chetkovich DM, et al.: Dual palmitoylation of PSD-95 mediates its vesiculotubular sorting, postsynaptic targeting, and ion channel clustering. J Cell Biol. 2000; 148(1): 159–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMisra C, Restituito S, Ferreira J, et al.: Regulation of synaptic structure and function by palmitoylated AMPA receptor binding protein. Mol Cell Neurosci. 2010; 43(4): 341–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHayashi T, Thomas GM, Huganir RL: Dual palmitoylation of NR2 subunits regulates NMDA receptor trafficking. Neuron. 2009; 64(2): 213–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingaraja RR, Huang K, Sanders SS, et al.: Altered palmitoylation and neuropathological deficits in mice lacking HIP14. Hum Mol Genet. 2011; 20(20): 3899–909. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu H, Abecasis GR, Heath SC, et al.: Genetic variation in the 22q11 locus and susceptibility to schizophrenia. Proc Natl Acad Sci U S A. 2002; 99(26): 16859–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizumaru C, Saito Y, Ishikawa T, et al.: Suppression of APP-containing vesicle trafficking and production of beta-amyloid by AID/DHHC-12 protein. J Neurochem. 2009; 111(5): 1213–24. PubMed Abstract \| Publisher Full Text\n\nIwabuchi M, Sheng H, Thompson JW, et al.: Characterization of the ubiquitin-modified proteome regulated by transient forebrain ischemia. J Cereb Blood Flow Metab. 2014; 34(3): 425–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYamamoto Y, Chochi Y, Matsuyama H, et al.: Gain of 5p15.33 is associated with progression of bladder cancer. Oncology. 2007; 72(1–2): 132–8. PubMed Abstract \| Publisher Full Text\n\nKang R, Wan J, Arstikaitis P, et al.: Neural palmitoyl-proteomics reveals dynamic synaptic palmitoylation. Nature. 2008; 456(7224): 904–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOyama T, Miyoshi Y, Koyama K, et al.: Isolation of a novel gene on 8p21.3–22 whose expression is reduced significantly in human colorectal cancers with liver metastasis. Genes Chromosomes Cancer. 2000; 29(1): 9–15. PubMed Abstract \| Publisher Full Text\n\nBirkenkamp-Demtroder K, Christensen LL, Olesen SH, et al.: Gene expression in colorectal cancer. Cancer Res. 2002; 62(15): 4352–63. PubMed Abstract\n\nYeste-Velasco M, Mao X, Grose R, et al.: Identification of ZDHHC14 as a novel human tumour suppressor gene. J Pathol. 2014; 232(5): 566–77. PubMed Abstract \| Publisher Full Text\n\nBlanc M, Blaskovic S, van der Goot FG: Palmitoylation, pathogens and their host. Biochem Soc Trans. 2013; 41(1): 84–8. PubMed Abstract \| Publisher Full Text\n\nRoth AF, Wan J, Bailey AO, et al.: Global analysis of protein palmitoylation in yeast. Cell. 2006; 125(5): 1003–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWan J, Roth AF, Bailey AO, et al.: Palmitoylated proteins: purification and identification. Nat Protoc. 2007; 2(7): 1573–84. PubMed Abstract \| Publisher Full Text\n\nForrester MT, Hess DT, Thompson JW, et al.: Site-specific analysis of protein S-acylation by resin-assisted capture. J Lipid Res. 2011; 52(2): 393–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartin BR, Cravatt BF: Large-scale profiling of protein palmitoylation in mammalian cells. Nat Methods. 2009; 6(2): 135–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang W, Di Vizio D, Kirchner M, et al.: Proteome scale characterization of human S-acylated proteins in lipid raft-enriched and non-raft membranes. Mol Cell Proteomics. 2010; 9(1): 54–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowal L, Yang W, Freeman MR, et al.: Proteomic analysis of palmitoylated platelet proteins. Blood. 2011; 118(13): e62–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIvaldi C, Martin BR, Kieffer-Jaquinod S, et al.: Proteomic analysis of S-acylated proteins in human B cells reveals palmitoylation of the immune regulators CD20 and CD23. PLoS One. 2012; 7(5): e37187. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarin EP, Derakhshan B, Lam TT, et al.: Endothelial cell palmitoylproteomic identifies novel lipid-modified targets and potential substrates for protein acyl transferases. Circ Res. 2012; 110(10): 1336–44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilson JP, Raghavan AS, Yang YY, et al.: Proteomic analysis of fatty-acylated proteins in mammalian cells with chemical reporters reveals S-acylation of histone H3 variants. Mol Cell Proteomics. 2011; 10(3): M110.001198. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMerrick BA, Dhungana S, Williams JG, et al.: Proteomic profiling of S-acylated macrophage proteins identifies a role for palmitoylation in mitochondrial targeting of phospholipid scramblase 3. Mol Cell Proteomics. 2011; 10(10): M110.006007. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRen W, Jhala US, Du K: Proteomic analysis of protein palmitoylation in adipocytes. Adipocyte. 2013; 2(1): 17–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYount JS, Moltedo B, Yang YY, et al.: Palmitoylome profiling reveals S-palmitoylation-dependent antiviral activity of IFITM3. Nat Chem Biol. 2010; 6(8): 610–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi Y, Martin BR, Cravatt BF, et al.: DHHC5 protein palmitoylates flotillin-2 and is rapidly degraded on induction of neuronal differentiation in cultured cells. J Biol Chem. 2012; 287(1): 523–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartin BR, Wang C, Adibekian A, et al.: Global profiling of dynamic protein palmitoylation. Nat Methods. 2011; 9(1): 84–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHemsley PA, Weimar T, Lilley KS, et al.: A proteomic approach identifies many novel palmitoylated proteins in Arabidopsis. New Phytol. 2013; 197(3): 805–14. PubMed Abstract \| Publisher Full Text\n\nWan J, Savas JN, Roth AF, et al.: Tracking brain palmitoylation change: predominance of glial change in a mouse model of Huntington's disease. Chem Biol. 2013; 20(11): 1421–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang MM, Wu PY, Kelly FD, et al.: Quantitative control of protein S-palmitoylation regulates meiotic entry in fission yeast. PLoS Biol. 2013; 11(7): e1001597. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJones ML, Collins MO, Goulding D, et al.: Analysis of protein palmitoylation reveals a pervasive role in Plasmodium development and pathogenesis. Cell Host Microbe. 2012; 12(2): 246–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWei X, Song H, Semenkovich CF: Insulin-regulated protein palmitoylation impacts endothelial cell function. Arterioscler Thromb Vasc Biol. 2014; 34(2): 346–54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHornbeck PV, Kornhauser JM, Tkachev S, et al.: PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res. 2012; 40(Database issue): D261–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltenhoff AM, Schneider A, Gonnet GH, et al.: OMA 2011: orthology inference among 1000 complete genomes. Nucleic Acids Res. 2011; 39(Database issue): D289–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKriventseva EV, Tegenfeldt F, Petty TJ, et al.: OrthoDB v8: update of the hierarchical catalog of orthologs and the underlying free software. Nucleic Acids Res. 2015; 43(Database issue): D250–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKumari B, Kumar R, Kumar M: PalmPred: an SVM based palmitoylation prediction method using sequence profile information. PLoS One. 2014; 9(2): e89246. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRen J, Wen L, Gao X, et al.: CSS-Palm 2.0: an updated software for palmitoylation sites prediction. Protein Eng Des Sel. 2008; 21(11): 639–44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFresno C, Fernandez EA: RDAVIDWebService: a versatile R interface to DAVID. Bioinformatics. 2013; 29(21): 2810–1. PubMed Abstract \| Publisher 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\nShannon PT, Grimes M, Kutlu B, et al.: RCytoscape: tools for exploratory network analysis. BMC Bioinformatics. 2013; 14: 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmoot ME, Ono K, Ruscheinski J, et al.: Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011; 27(3): 431–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFranceschini A, Szklarczyk D, Frankild S, et al.: STRING v9.1: protein-protein interaction networks, with increased coverage and integration. Nucleic Acids Res. 2013; 41(Database issue): D808–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRuepp A, Waegele B, Lechner M, et al.: CORUM: the comprehensive resource of mammalian protein complexes--2009. Nucleic Acids Res. 2010; 38(Database issue): D497–501. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVacic V, Iakoucheva LM, Radivojac P: Two Sample Logo: a graphical representation of the differences between two sets of sequence alignments. Bioinformatics. 2006; 22(12): 1536–7. PubMed Abstract \| Publisher Full Text\n\nGreaves J, Gorleku OA, Salaun C, et al.: Palmitoylation of the SNAP25 protein family: specificity and regulation by DHHC palmitoyl transferases. J Biol Chem. 2010; 285(32): 24629–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWurtzel JG, Kumar P, Goldfinger LE: Palmitoylation regulates vesicular trafficking of R-Ras to membrane ruffles and effects on ruffling and cell spreading. Small GTPases. 2012; 3(3): 139–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlvarez E, Girones N, Davis RJ: Inhibition of the receptor-mediated endocytosis of diferric transferrin is associated with the covalent modification of the transferrin receptor with palmitic acid. J Biol Chem. 1990; 265(27): 16644–55. PubMed Abstract\n\nLynes EM, Bui M, Yap MC, et al.: Palmitoylated TMX and calnexin target to the mitochondria-associated membrane. EMBO J. 2012; 31(2): 457–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLinder ME, Middleton P, Hepler JR, et al.: Lipid modifications of G proteins: alpha subunits are palmitoylated. Proc Natl Acad Sci U S A. 1993; 90(8): 3675–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLu D, Sun HQ, Wang H, et al.: Phosphatidylinositol 4-kinase IIalpha is palmitoylated by Golgi-localized palmitoyltransferases in cholesterol-dependent manner. J Biol Chem. 2012; 287(26): 21856–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOhno Y, Kashio A, Ogata R, et al.: Analysis of substrate specificity of human DHHC protein acyltransferases using a yeast expression system. Mol Biol Cell. 2012; 23(23): 4543–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark S, Patterson EE, Cobb J, et al.: Palmitoylation controls the dynamics of budding-yeast heterochromatin via the telomere-binding protein Rif1. Proc Natl Acad Sci U S A. 2011; 108(35): 14572–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nResh MD: Fatty acylation of proteins: new insights into membrane targeting of myristoylated and palmitoylated proteins. Biochim Biophys Acta. 1999; 1451(1): 1–16. PubMed Abstract \| Publisher 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): 25–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaron JM, Vega LR, Fleming J, et al.: Single site alpha-tubulin mutation affects astral microtubules and nuclear positioning during anaphase in Saccharomyces cerevisiae: possible role for palmitoylation of alpha-tubulin. Mol Biol Cell. 2001; 12(9): 2672–87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGeorge J, Soares C, Montersino A, et al.: Palmitoylation of LIM Kinase-1 ensures spine-specific actin polymerization and morphological plasticity. Elife. 2015; 4: e06327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSha Z, Brill LM, Cabrera R, et al.: The eIF3 interactome reveals the translasome, a supercomplex linking protein synthesis and degradation machineries. Mol Cell. 2009; 36(1): 141–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKabir MA, Uddin W, Narayanan A, et al.: Functional Subunits of Eukaryotic Chaperonin CCT/TRiC in Protein Folding. J Amino Acids. 2011; 2011: 843206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoobol A, Carden MJ: Subunits of the eukaryotic cytosolic chaperonin CCT do not always behave as components of a uniform hetero-oligomeric particle. Eur J Cell Biol. 1999; 78(1): 21–32. PubMed Abstract \| Publisher Full Text\n\nHo GP, Selvakumar B, Mukai J, et al.: S-nitrosylation and S-palmitoylation reciprocally regulate synaptic targeting of PSD-95. Neuron. 2011; 71(1): 131–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHess DT, Stamler JS: Regulation by S-nitrosylation of protein post-translational modification. J Biol Chem. 2012; 287(7): 4411–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee TY, Chen YJ, Lu CT, et al.: dbSNO: a database of cysteine S-nitrosylation. Bioinformatics. 2012; 28(17): 2293–5. PubMed Abstract \| Publisher Full Text\n\nChen YJ, Lu CT, Lee TY, et al.: dbGSH: a database of S-glutathionylation. Bioinformatics. 2014; 30(16): 2386–8. PubMed Abstract \| Publisher Full Text\n\nCharych EI, Jiang LX, Lo F, et al.: Interplay of palmitoylation and phosphorylation in the trafficking and localization of phosphodiesterase 10A: implications for the treatment of schizophrenia. J Neurosci. 2010; 30(27): 9027–37. PubMed Abstract \| Publisher Full Text\n\nTian L, Jeffries O, McClafferty H, et al.: Palmitoylation gates phosphorylation-dependent regulation of BK potassium channels. Proc Natl Acad Sci U S A. 2008; 105(52): 21006–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDorfleutner A, Ruf W: Regulation of tissue factor cytoplasmic domain phosphorylation by palmitoylation. Blood. 2003; 102(12): 3998–4005. PubMed Abstract \| Publisher Full Text\n\nSoskic V, Nyakatura E, Roos M, et al.: Correlations in palmitoylation and multiple phosphorylation of rat bradykinin B2 receptor in Chinese hamster ovary cells. J Biol Chem. 1999; 274(13): 8539–45. PubMed Abstract \| Publisher Full Text\n\nYount JS, Karssemeijer RA, Hang HC: S-palmitoylation and ubiquitination differentially regulate interferon-induced transmembrane protein 3 (IFITM3)-mediated resistance to influenza virus. J Biol Chem. 2012; 287(23): 19631–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbrami L, Leppla SH, van der Goot FG: Receptor palmitoylation and ubiquitination regulate anthrax toxin endocytosis. J Cell Biol. 2006; 172(2): 309–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFairbank M, Huang K, El-Husseini A, et al.: RING finger palmitoylation of the endoplasmic reticulum Gp78 E3 ubiquitin ligase. FEBS Lett. 2012; 586(16): 2488–93. PubMed Abstract \| Publisher Full Text\n\nValdez-Taubas J, Pelham H: Swf1-dependent palmitoylation of the SNARE Tlg1 prevents its ubiquitination and degradation. EMBO J. 2005; 24(14): 2524–32. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9524", "date": "27 Jul 2015", "name": "Howard C. Hang", "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 database developed by Blanc and colleagues is an excellent and valuable resource for researchers working on S-palmitoylation. The database allows to visualize and compare 19 proteomics datasets published in the last few years for protein S-palmitoylation. Proteomics data often generate some false positive hits and SwissPalm-mediated improvement of confidence will help scientists interpret the results of these proteomics studies and could also use as a reference database for future proteomics studies. The other tools added to the database (multiple alignments of orthologous proteins and their isoforms, site prediction score from CSS-Palm and PalmPred) are also valuable. The database interface is well described in the paper. As a result, the database is easy to use and the data easily extracted from the website. The comparative analysis of metabolic labeling and hydroxyl amine-sensitive enrichment methods should encourage researchers to employ both methods for high-quality proteomic analysis of S-fatty-acylated proteins.", "responses": [] }, { "id": "9527", "date": "30 Jul 2015", "name": "Luke Chamberlain", "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\nSwissPalm represents an extremely valuable resource for the palmitoylation research community, and the authors have undertaken a large volume of work to produce this palmitoylation database. There are also some useful and interesting comparative analyses presented in the paper. As commented by the authors, up to 10% of the proteome might be subject to palmitoylation, emphasising both the importance of this lipid modification and the importance of this new resource. Information has been curated from a growing number of protein-palmitoylome studies and from papers focused on individual palmitoylated proteins, and the database will be continuously updated.I have only minor comments for improvement:Page 3, first paragraph: “while thioesterases detach it and Acyl Protein Thioesterases (APTS) which remove the acyl chain”. There appears to be some repetition here and “while thioesterases detach it” could be removed. Figure 1A: text states 19 palmitoyl-proteome screens were selected but first panel of Figure 1A suggests 18? Should there be 6 studies from human using chemical capture? Also text states 5199 proteins were incorporated but figure shows 5169? Figure 2A-C: some text is partially obscured. Page 8, section on “Protein information page”: the figure numbers given in text of this section do not match with the figures. Figure 5A: values of 1455, 2341 and 2651 are given in Figure, whereas corresponding text (pages 8 and 9) states 1453, 2339 and 2649. The text corresponding to Figure 7A indicates 27 human and 20 mouse protein complexes - the authors might provide some additional information as to how these values correspond with the bar graph in Figure 7A (which does not show 27/20 complexes). The legend of Figure 8 gives “C” rather than “D” for text corresponding to panel D.", "responses": [] }, { "id": "9523", "date": "13 Aug 2015", "name": "Anthony Magee", "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 very valuable contribution to the field of palmitoylation. It not only provides a wealth of data for researchers but also thoughtful interpretations.I found the location of cysteine residues in proteins particularly helpful (Fig. 4). Also, the analysis of residues surrounding palmitoylated cysteines is informative, and in essence comes to the same conclusion that I and several others have come to (on a less systematic basis!) that there is no real \"consensus\" sequence for palmitoylation. This is in contrast to a few papers in the literature which erroneously purport to identify such a consensus.The enrichment of palmitoylated proteins in complexes is interesting and suggests a possible role of fatty acyl moieties in mediating complex formation. The interplay between palmitoylation, phosphorylation, ubiquitination and other PTMs is also nicely highlighted.The only major limitation of the paper and the website I have found is the omission of the MBOAT family of protein acyltransferases. Although only one of these is strictly probably a S-palmitoyltransferase (Hedgehog acyltransferase, Hhat), two members of the family (Porcupine and Ghrelin O-acyltransferase, GOAT) are acyltransferases that transfer fatty acyl groups onto proteins. I think it would be helpful if they could be incorporated, and would make the website more comprehensive for investigators interested in protein acylation in its broadest sense. Admittedly, Hhat is thought by some to be a N-palmitoyltransferase, but it is highly likely that it is a S-palmitoyltransferase and that the modified Hedgehog substrates then undergo spontaneous S-to-N acyl transfer resulting in amide-linked palmitate. Porcupine transfers the closely related fatty acyl group palmitoleate to a serine hydroxyl of Wnt proteins and GOAT transfers an octanoyl group to a serine of ghrelin, so these are not \"S-palmitoyltransferases\" but they share enough in common as protein fatty acyltransferases to warrant inclusion in the database. A minor point, but Hhat has also recently been identified by Konitsiotis et al. to be palmitoylated (Konitsiotis et al., 2015) so should be included in the database at the next revision.One minor correction I would suggest - the authors generally use the term \"palmitoyl proteome\" throughout, which was agreed to be the appropriate term at the second dedicated meeting on protein palmitoylation in Edinburgh last year. However, the less suitable term \"palmitome\" has slipped through in one place, the heading at the bottom right of page 8.", "responses": [] } ]	1	https://f1000research.com/articles/4-261
https://f1000research.com/articles/4-259/v1	16 Jul 15	{ "type": "Research Article", "title": "PHENOstruct: Prediction of human phenotype ontology terms using heterogeneous data sources", "authors": [ "Indika Kahanda", "Christopher Funk", "Karin Verspoor", "Asa Ben-Hur", "Indika Kahanda", "Christopher Funk", "Karin Verspoor" ], "abstract": "The human phenotype ontology (HPO) was recently developed as a standardized vocabulary for describing the phenotype abnormalities associated with human diseases. At present, only a small fraction of human protein coding genes have HPO annotations. But, researchers believe that a large portion of currently unannotated genes are related to disease phenotypes. Therefore, it is important to predict gene-HPO term associations using accurate computational methods. In this work we demonstrate the performance advantage of the structured SVM approach which was shown to be highly effective for Gene Ontology term prediction in comparison to several baseline methods. Furthermore, we highlight a collection of informative data sources suitable for the problem of predicting gene-HPO associations, including large scale literature mining data.", "keywords": [ "human phenotype ontology", "structured SVM" ], "content": "Introduction\n\nIn the medical context a phenotype is defined as a deviation from normal morphology, physiology, or behavior1. The human phenotype ontology (HPO) is a standardized vocabulary that describes the phenotype abnormalities encountered in human diseases2. It was initially populated using databases of human genes and genetic disorders such as OMIM3, Orphanet4 and DECIPHER5, and was later expanded using literature curation. The hierarchical structure of the HPO is very similar to that of the Gene Ontology (GO)6, and it too has the structure of a directed acyclic graph (DAG); like GO, more general terms are found at the top, and term specificity increases from the root to the leaves. This implies the “true-path rule”: whenever a gene is annotated with a given term, that implies all its ancestor terms.\n\nHPO is composed of three subontologies: organ abnormality, mode of inheritance, and onset and clinical course. Organ abnormality is the main subontology which describes clinical abnormalities (Figure 1). The mode of inheritance subontology describes the inheritance patterns of the phenotypes. The onset and clinical course subontology describes the typical time of onset of clinical symptoms and their speed of progression. The organ abnormality, mode of inheritance and onset and clinical course subontologies are composed of ~10000, 25 and 30 terms respectively. Throughout this paper, the organ abnormality, the mode of inheritance, and the onset and clinical course subontologies will be referred to as the Organ subontology, Inheritance subontology and Onset subontology, respectively.\n\nAll HPO parent-child relationships represent “is-a” relationships.\n\nThe HPO web site (http://www.human-phenotype-ontology.org) provides gene-disease-HPO annotations that can be used for research involving human diseases. Over 50,000 annotations of hereditary diseases are available at the moment. Specifically, the genes are annotated with a set of phenotype terms based on their known relationships with diseases (Figure 2).\n\na) general format of annotations: genes are annotated with a set of phenotype terms based on their known relationships with diseases b) an example annotation: the amyloid precursor protein (APP) gene is associated with Alzheimer’s disease and cerebroarterial amyloidosis. Therefore, the APP gene is annotated with the set of HPO terms (Organ in orange, Inheritance in green) associated with these diseases.\n\nCurrently, only a small fraction (~3000) of human protein coding genes are known to be associated with hereditary diseases, and only those genes have HPO annotations at the moment. But researchers believe that there are many other disease-causing genes in the human genome and estimate that another 5000 genes can be associated with phenotypes (Peter Robinson, personal communication, 2014). However, experimentally finding disease-causing genes is a highly resource consuming and difficult task7. Therefore, it is important to explore the feasibility of developing computational methods for predicting gene-HPO associations. While there is a plethora of computational approaches for the related task of prediction of gene-disease associations8, no computational method that directly predicts gene-HPO term associations exists at this time.\n\n\nApproach\n\nWe define the HPO prediction problem as directly predicting the complete set of HPO terms for a given gene. This problem is a hierarchical multilabel classification (HMC) problem9, as a given gene can be annotated with multiple labels, and the set of labels have a hierarchy associated with them.\n\nThe traditional approach for solving HMC problems is to decompose the problem into multiple single label problems and apply independent binary classifiers for each label separately10; however, this approach has several disadvantages. First, independent classifiers are not able to learn from the inter-relationships between the labels. Second, the leaf terms typically have a low number of annotated examples making it difficult to learn an effective classifier. Furthermore, the predicted labels are typically hierarchically inconsistent, i.e. a child term (e.g. Hearing abnormality) is predicted while its parent term (e.g. Abnormality of ear) is not—making it difficult to interpret the predictions. To remedy this problem, an additional reconciliation step of combining independent predictions to obtain a set of predictions that are consistent with the topology of the ontology is required (see e.g. 11 for a discussion of several reconciliation methods that are effective for GO term prediction).\n\nAn alternative approach is to use a single classifier that learns a direct mapping from inputs to the space of hierarchically consistent labels; this can be achieved using structured prediction, which is a framework for learning a mapping from inputs to label spaces that have a structure associated with them12. This framework can capture information from the inter-relationships between labels and allows the prediction of a set of labels that are hierarchically consistent, eliminating the need for multiple classifiers, and the need for establishing hierarchical consistency between the predictions. Previously we have shown the effectiveness of modeling the GO term prediction problem using a structured prediction framework in a method called GOstruct13,14. In this work we demonstrate the effectiveness of this strategy for HPO term prediction using the same methodology, and explore a variety of data sources that are useful for this task, including large scale data extracted from the biomedical literature.\n\n\nMethods\n\nOur models are provided with feature vectors and HPO annotations. Each gene/protein was characterized by several sets of features generated using four data sources: Network, GO, literature and variants, which are described below. We used the UniProt ID mapping service (http://www.uniprot.org/mapping/) for mapping genes to proteins.\n\nGene-HPO annotations were downloaded from the HPO website (http://www.human-phenotype-ontology.org). We ignored the global root term (“ALL”) and root terms of the three subontologies. We also removed terms that were not annotated to 10 or more genes. Then we mapped the genes to proteins and generated corresponding protein-HPO annotations (see Table 1).\n\nThe “unique terms” column provides both the number of terms and the number of leaf terms; the “annotations” column provides the number of annotations, as well as their number when expanded using the true-path rule.\n\nWe extracted protein-protein interactions and other functional association network data (i.e. co-expression, co-occurrence, etc.) from BioGRID 3.2.10615, STRING 9.116 and GeneMANIA 3.1.2 (http://pages.genemania.org/data/) databases.\n\nThe BioGRID database provides protein-protein interaction networks acquired from physical and genetic interaction experiments. STRING provides networks based on several different evidence channels (co-expression, co-occurrence, fusion, neighborhood, genetic interactions, physical interactions, etc.). We combined edges from the two databases by taking the union of interactions from BioGRID and STRING and represented each gene by a vector of variables, where component i indicates if the corresponding protein interacts with protein i in the combined network.\n\nThe GeneMANIA website (http://pages.genemania.org/data/) provides a large number of protein-protein interaction/association networks generated using several types of evidence: co-expression, co-localization, genetic interactions, physical interactions and predicted interactions. A gene is represented by a vector of variables for each network, where component i indicates if the corresponding protein interacts with protein i with respect to that particular network.\n\nWe extracted GO6 annotations from the GO web site (http://www.geneontology.org/) and Uniprot-goa (http://www.ebi.ac.uk/GOA). We excluded all annotations that were obtained by computational methods. A gene is represented as a vector of indicator variables in which variable i is 1 if it is annotated with GO term i.\n\nWe used two different sources for generating literature features: abstracts extracted from Medline on 10-23-13 and full-text articles extracted from PubMed Open Access Collection (PMCOA) on 11-06-13. A natural language processing pipeline was utilized to characterize genes/proteins by same-sentence word occurrences extracted from these sources, forming a bag-of-words (BoW) representation for each gene17. First, all words were lower-cased and stop words were removed. Then they were further filtered to keep only the low frequency words (i.e. words that are present only in less than 1% of the proteins in the data). A gene is represented by a vector in which the element i gives the number of times the word i occurred in the same sentence with that gene/protein.\n\nWe extracted all the disease variants in the human genome and their associated diseases from UniProt (http://www.uniprot.org/docs/humsavar). This data provides variants that have been found in patients and the disease-association is reported in literature. We also extracted gene-disease associations from the HPO website. This data associates a protein with diseases that are known to occur when the associated gene is mutated. To generate features from this data, we first extracted for each protein pi its set of associated diseases (Di) from the protein-disease associations. Then we retrieved the set of disease variants (Vi) associated with all diseases in Di from the UniProt disease variants data. Finally, each gene was represented by a vector in which element j indicates if the variant j is in Vi.\n\nIn this work we compare a structured support vector machine approach against several baseline methods: a) binary support vector machines (SVMs) and b) a state-of-the-art HMC method based on decision tree ensembles (Clus-HMC-Ens). In this section we describe PHENOstruct and the two baseline methods. In addition, we assessed the performance of: c) an indirect method that first predicts disease terms for a gene using a structured model and then maps them to HPO terms and d) using OMIM disease terms predicted by PhenoPPIOrth18 followed by mapping the OMIM terms to HPO terms. We describe these two additional methods in the Supplementary material (see section “Additional methods”). All methods except PhenoPPIOrth were provided the same data.\n\nIn earlier work we developed the GOstruct method which uses structured SVMs (SSVM) for GO term prediction13. In this work we apply the same methodology to HPO term prediction and refer to it as PHENOstruct to emphasize the different problem domain. Unlike collections of binary classifiers applied independently at each node of the hierarchy, PHENOstruct predicts a set of hierarchically consistent HPO terms for a given gene (Figure 3). More specifically, PHENOstruct learns a compatibility function that models the association between a given input and a structured output12, in this case the collection of all hierarchically consistent sets of HPO terms. Let 𝒳 be the input space where genes are represented and let 𝒴 be the space of labels. The set of HPO terms associated with a given gene is collectively referred to as its (structured) label. 𝒴 represents each HPO subontology in a vector space where component i represents term i. Given a training set {(xi, yi)}i=1n where xi∈𝒳 and yi∈𝒴, the compatibility function f : 𝒳 × 𝒴 → ℛ maps input-output pairs to a score that indicates how likely is a gene x to be associated with a collection of terms represented by y. The predicted label ŷ for an unseen input x can then be obtained by using the argmax operator as ŷ = argmaxy∈𝒴c f(x, y) where 𝒴c ⊂ 𝒴 is the set of all candidate labels. In this work we use the combinations of all terms in the training set as the set of candidate labels 𝒴c.\n\nPHENOstruct takes the set of feature vectors and HPO annotations associated with each gene as input for training. Once trained, it can predict a set of hierarchically consistent HPO terms for a given test gene. PHENOstruct is trained on and makes predictions for a single subontology at a time (DAGs belonging to Organ, Inheritance and Onset subontologies are shown in orange, green and blue, respectively).\n\nIn order to obtain correct classification, the compatibility value of the true label (correct set of HPO annotations) of an input needs to be higher than that of any other candidate label (Figure 4). PHENOstruct uses structured SVM (SSVM) training where this is used as a (soft) constraint; it tries to maximize the margin, or the difference between the compatibility value for the actual label and the compatibility for the next best candidate12. In the structured-output setting, kernels correspond to dot products in the joint input-output feature space, and they are functions of both inputs and outputs. PHENOstruct uses a joint kernel that is the product of the input-space and the output-space kernels:\n\nThe compatibility function, which is the key component of the structured prediction framework, measures compatibility between a given input and a structured output. The compatibility function of the true label (correct set of HPO annotations) is required to be higher than that of any other label. and the difference between these two scores (margin) is maximized.\n\nK((x1, y1), (x2, y2)) = K𝒳(x1, x2)K𝒴(y1, y2).\n\nThe motivation for this form is that two input/output pairs are considered similar if they are similar in both their input space features and their labels; the output space kernel, for which we use a linear kernel between label vectors, captures similarity of the annotations associated with two genes; the input space kernel combines several sources of data by the addition of multiple input-space kernels, one for each data source. Each kernel is normalized according to\n\nKnorm(z1, z2) = K(z1, z2)/K(z1, z1)K(z2, z2)\n\nbefore being used with the joint input-output kernel. The Strut library (http://sourceforge.net/projects/strut/) with default parameter settings was used for the implementation of PHENOstruct.\n\nAs a baseline method we trained a collection of binary SVMs, each trained on a single HPO term. Binary SVMs were trained using the PyML (http://pyml.sourceforge.net) machine learning library with default parameter settings. We used linear kernels for each set of input space features.\n\nClus-HMC-Ens is a state-of-the-art HMC method based on decision tree ensembles which has been shown to be very effective for GO term prediction19. In our study, we provide exactly the same set of features used with PHENOstruct as input to Clus-HMC-Ens and use parameter settings that provided the best performance for GO term prediction (https://dtai.cs.kuleuven.be/clus/hmc-ens/). The number of bags used was 50 for the Inheritance and Onset subontologies; 10 bags were used for the Organ subontology because of the large running times for this subontology.\n\nClassifier performance was estimated using five-fold cross-validation. Since typically scientists/biologists are interested in knowing the set of genes/proteins associated with a certain HPO term, we primarily use a term-centric measure for presenting results. Term-centric measures average performance across terms as opposed to protein-centric measures which average performance across proteins as described elsewhere20. More specifically, we use the macro AUC (area under the receiver operating curve), which is computed by averaging the AUCs across HPO terms. For comparing performance across classifiers, p-values were computed using paired t-tests. Additionally, we report performance in terms of several protein-centric measures (precision, recall, F-max) in the Supplementary material (Table S3 and Table S4). Definitions of all performance measures are given in the Supplementary material. PHENOstruct assigns a confidence score to each predicted HPO term, which is computed using the compatibility function as described elsewhere14. The onset and clinical course subontology includes terms such as pace of progression, age of onset and onset which are only used for grouping terms. We ignore these grouping terms when computing performance.\n\n\nResults and discussion\n\nAs illustrated in Table 2, PHENOstruct significantly outperforms Clus-HMC-Ens and the binary SVMs in the Organ and Onset subontologies. This suggests that modeling the HPO prediction problem as a structured prediction problem is highly effective. It is interesting to note that the biggest improvement of PHENOstruct over binary SVMs is seen in the Organ subontology. Given its very large number of terms, as well as the deep hierarchy, this further confirms the value of the structured approach. PHENOstruct outperforms binary SVMs in the Inheritance and Onset subontologies but to a lesser extent than in the Organ subontology because they are far less complex than the Organ subontology. We note that the two methods that first predict OMIM terms, which are then mapped to HPO terms performed poorly (see details in the Supplementary material). It is also interesting to see that Clus-HMC-Ens performs worse than binary SVMs with respect to macro AUC (Table 2) but performs slightly better than binary SVMs according to protein-centric F-max (Table S3).\n\nPerformance across the three HPO subontologies for PHENOstruct, binary SVMs and Clus-HMC-Ens measured using the macro AUC. P-values provide the significance level for the difference between the corresponding method and PHENOstruct.\n\nThe average macro AUC for the Inheritance subontology is 0.74. Terms are displayed in ascending order of frequency.\n\nPHENOstruct’s average AUC for the Organ and Inheritance subontologies are 0.73 and 0.74, respectively. Even though the Organ subontology is a far more complex subontology than the Inheritance subontology (with thousands of terms and 13 levels as opposed to tens of terms and only 3 levels) they show similar performance. The Onset subontology is the hardest to predict accurately, with an average AUC of 0.64. Only six Onset subontology terms have individual AUCs above 0.7 (Table 4).\n\nThe average macro AUC for the Onset subontology is 0.64. Terms are displayed in ascending order of frequency.\n\nEven though PHENOstruct outperforms the baseline methods, there is much room for improvement, especially in the Onset subontology. The small number of annotated genes in this subontology (Table 1) makes it difficult to train an effective model while the incomplete nature of the current gold standard used for evaluation tends to underestimate performance of classifiers21. See section for a detailed analysis of false positives.\n\nIn general, Organ subontology terms with few annotations show a mix of both high and low performance as illustrated in Figure 5. This suggests that PHENOstruct is not necessarily affected by the frequency of the terms. But, terms with more annotations tend to show moderate performance. See Figure 6 for an example of experimental and predicted annotations (Organ subontology) for a protein. It is interesting to note that “polygenic inheritance” and its parent term “mulifactorial inheritance” have the lowest number of annotations as well as the lowest individual AUCs in the Inheritance subontology (see Table 3). These are the two terms with the lowest AUC with binary SVMs as well (see Table S6). It is not surprising that these two terms have lower accuracy because each describes inheritance patterns that depend on a mixture of determinants. Moreover, the diseases inherited in this manner – termed complex diseases – are not as well characterized and annotated compared to Mendelian/single gene diseases. On the other hand, the mitochondrial inheritance term has an exceptional AUC of 0.98. It is also the term with the highest AUC with the binary SVMs as well (see Table S6). The human mitochondrial DNA was the first significant part of the human genome to be fully sequenced, two decades before the completion of the human genome project22. Due to this, and the relative ease of sequencing the mitochondrial genome23, diseases caused by mutations in human mitochondrial DNA have been reported very early24,25. It is likely that this well-studied nature of mitochondrial DNA leads to the high performance of the mitochondrial inheritance term.\n\nPerformance for each term is displayed using AUC against its frequency. The average AUC for the Organ subontology is 0.73.\n\na) experimental annotation of protein P43681 b) PHENOstruct’s prediction for P43681 (protein-centric precision and recall for this individual protein is 1.0 and 0.62, respectively).\n\nAs a potential improvement to PHENOstruct we explored an approximate inference algorithm that replaces computation of the most compatible label by looping overall combinations of labels that occur in the training data with a dynamic programming algorithm that performs approximate evaluation of all possible combinations of hierarchically consistent labels. However, this led to a slight decrease in performance, showing the advantage of considering only the biologically relevant combinations. Further research should consider other alternatives.\n\nAll experiments were performed on Linux running machines with 8 cores (64-bit, 3.3GHz) and 8GB memory. Combined running times for performing five-fold cross-validation for all three subontologies are: binary SVMs: 55 hours, Clus-HMC-Ens: 825 hours and PHENOstruct: 90 hours.\n\nWe performed the following set of experiments in order to identify the most effective data sources for HPO prediction using PHENOstruct. First, to identify the individual effectiveness of each source, we performed a series of experiments in which we provided features generated from a single source of data at a time as input to PHENOstruct. Then to understand how much each data source is contributing to the overall performance we conducted leave-one-source-out experiments.\n\nIn all three subontologies, network data is the most informative individual data source as illustrated in Figure 7. Moreover, it is by far the main contributor to the overall performance both in the Organ and Inheritance subontologies (Figure 8). This is intuitive because if two genes/proteins are known to be interacting and/or active in the same pathways it leads to association with the same/similar diseases/phenotypes.\n\nResults are shown for each source of data: network (functional association data); Gene Ontology annotations; literature mining data; genetic variants; and the model that combines all features together.\n\nAlthough the genetic variant features provide the lowest performance in the Organ and Onset subontologies, leaving out variant data hurts the overall performance noticeably in all three subontologies as can be seen in Figure 8. This suggests that variant data are very useful especially as a complementary dataset to the others. Moreover, we found that variant data are very effective for predicting cancer-related terms in the Organ subontology (see Table S1).\n\nIt is very encouraging to see that the literature data with a simple BoW representation by itself is very informative (Figure 7) and leaving out literature features shows considerable performance drop in the other two subontologies (Figure 8). In an analysis of the SSVM weight vector, we found that the majority of the most important tokens extracted from literature consist of names of proteins, genes and diseases (see Table S2).\n\nWe also considered an alternative representation of the literature data where a gene is represented by a vector in which the element i gives the number of times the word i occurred in the same sentence with that particular gene/protein divided by the total number of unique genes/proteins that word co-occurred with. This representation is analogous to the TFIDF (term frequency ∗ inverse document frequency) representation typically used in information retrieval and text mining26. However, these features led to slight deterioration of performance in all three subontologies (macro AUCs 0.60, 0.58 and 0.56 for Organ, Inheritance and Onset subontologies, respectively).\n\nAlthough GO features provide the second best individual performance both in the Organ and Onset subontologies (Figure 7), their contribution to the overall performance is very minimal (Figure 8). In fact leaving out GO features increases the overall performance in the Inheritance and Onset subontologies. The incompleteness of GO annotations may have contributed towards this.\n\nFinally, the combination of all the features provides higher performance than individual feature sets in all three subontologies as can be seen in Figure 7. However, leaving out GO features in the Inheritance and Onset subontologies, led to improved performance, suggesting that not all sources contribute to the overall performance. This shows that the selection of data sources must be performed carefully in order to find the optimal combination of sources for each subontology.\n\nLike other biological ontologies, the HPO is incomplete due to various factors such as slowness of the curation process27. In other words, the set of HPO annotations we considered as the gold standard does not fully represent all the phenotypes that should be associated with the currently annotated genes; this leads to performance estimates that underestimate the true performance of a classifier21. To explore this issue, we selected 25 predictions made by PHENOstruct which were considered false positives according to the current gold standard and looked for evidence in the current biomedical literature that can be used as evidence for those predictions. For 14 of those predictions we were able to find supporting evidence. The details of the complete validation process are given in the Supplementary material.\n\n\nConclusions and future work\n\nThis is the first study of directly predicting gene-HPO term associations. We modeled this problem as a hierarchical multi-label problem and used the SSVM framework for developing PHENOstruct. Our results demonstrate that using the SSVM is more effective than the traditional approach of decomposing the problem into a collection of binary classification problems. In our experiments we evaluated several types of data which were found to be informative for HPO term prediction: networks of functional association, large scale data mined from the biomedical literature and genetic variant data.\n\nThere are several ways in which this work can be extended. For the literature data we used a simple BoW representation. An alternative is to try and extract gene-HPO term co-mentions directly; in the context of GO term prediction we have found that both approaches lead to similar overall performance17. However, co-mentions have the added value that they are easy to verify by a human curator. Another source of information that can be utilized is semantic similarity of HPO terms to other phenotypic ontologies such the mammalian phenotype ontology, which is currently used for annotating the rat genome28. Finally, exploring the effectiveness of combining all three subontologies, as opposed to treating them as three independent subontologies as we have done here, is also worth exploring.\n\nAlthough PHENOstruct outperformed the baseline methods, there is considerable room for improvement in all three subontologies. While some improvement can likely be obtained as described above, its performance will also improve as the number of HPO annotations increases. HPO is a relatively new ontology that will likely see substantial growth in the coming years, which will help in improving the accuracy of computational methods that contribute to its expansion.\n\n\nData and software availability\n\nZenodo: Data and software associated with PHENOstruct:Prediction of human phenotype ontology terms using heterogeneous data sources, 10.5281/zenodo.1876429", "appendix": "Author contributions\n\n\n\nIK and AB conceived and designed the method and experiments. CF and KV developed a NLP pipeline and generated literature features. IK performed all experiments with PHENOstruct. All authors read and approved the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work was supported by the NSF Advances in Biological Informatics program through grants number 0965768 (awarded to Dr. Ben-Hur) and 0965616 (originally awarded to Dr. Verspoor).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nIn this section we analyze the performance of the features generated from genetic variant data in detail. The macro AUC for the variants data is only 0.56 in the Organ subontology. However, 37 terms have an AUC equal to or above 0.9. As listed in the Table S1, 21 out of those 37 terms well-predicted by the variant data are terms related to cancer. But interestingly, only 53 out of 1796 of all the Organ subontology terms are related to cancer. This shows strong evidence that the genetic variant data are highly effective for predicting cancer related phenotype terms. Terms predicted with high accuracy by the literature features do not show a similar tendency (data not shown).\n\nTerms are listed in the ascending order of their individual AUCs. 21 out of the 37 (57%) terms well-predicted Organ subontoloy terms by the variant data are terms related to cancer.\n\nIn the Inheritance subontology it was noticeable that the variant data are more effective for the categories with fewer annotations compared to the literature features (data not shown). The average number of annotations of the Inheritance subontology HPO categories with relatively higher AUC by variant features (compared to literature features) is only 46. This trend is also visible, albeit to a lesser extent, in the Organ and Onset subontologies as well; the corresponding numbers for the Organ and the Onset subontologies are 26 and 91, respectively. Furthermore, only mitochondrial inheritance term achieves an AUC above 0.9 with all data sources. However, with variant data alone, both mitochondrial inheritance and somatic mutation terms achieve AUCs above 0.9.\n\nIn order to identify the most important literature features, we looked at the weight vectors of the structured SVM model underlying PHENOstruct that was trained only on the literature features. Typically, the input space features with higher weight in the weight vector correspond to the features that are considered most important by the model for the given predictive task.\n\nIn the dual formulation of the Structured SVM, αij values are defined for each pair of example i and structured output j. In order to calculate the weight vector for a specific HPO term j (Wj), we first identified the subset of input examples (i.e. proteins) that are annotated with the given term (referred to as Sj). Then Wj is the summation of αij × xi where xi is the feature vector of example i and xi ∈ Sj. Features with higher weights in the weight vector Wj correspond to the features that were considered most informative by the model for the task of predicting the term j.\n\nWe trained PHENOstruct on literature features and computed the weight vectors as described above. Then we ranked the literature features by their weights and examined the top-100 literature features. In the Organ subontology we analysed the top-100 literature features with respect to the 8 HPO terms that have individual AUCs above 0.9. For those 8 terms the union set of top-100 features is composed of 107 unique tokens. By far, the majority (>70%) of these tokens are genes/ proteins/ protein complexes/ pathways names. Another 12 tokens are disease/phenotype names (Table S2).\n\nThe union set of the top-100 literature features with respect to the 8 HPO terms that have individual AUCs equal to or above 0.9. It is composed of 107 unique tokens. The token “-308” and “t308” in the “proteins/protein complexes” category are due to mis-tokenization of “miR-308”. Similarly, “-238” in the same category is due to mis-tokenization of “BQ-23”. Also “=-galcer” in the same category originated from α-galcer and β-galcer due to mis-handling of UTF characters α and β.\n\nWe describe here the results of experiments that we conducted with two additional methods.\n\nSSVM → disease → HPO method This is an indirect method that first predicts gene-disease associations and then maps them to HPO terms using associations available on the HPO website. This method uses the same input space data as PHENOstruct and learns a structured SVM using the same methodology. Using this model it predicts diseases along with confidence scores for unseen genes. Subsequently, the predicted scores for disease terms are directly transferred to all the HPO terms associated with those diseases (Figure S1). When multiple diseases are associated with a single HPO term, scores are accumulated. It is surprising that this method shows mediocre performance (Table S3). One of the main reasons for this is the low performance of the underlying SSVM for predicting disease terms (average AUC of 0.64), which consequently affects the accuracy of predicted HPO terms.\n\nThis method takes feature vectors and disease annotations associated with each gene as the input for training a SSVM model. Then, it predicts diseases for unseen genes using the learned model. Subsequently, the predicted scores for disease terms are directly transferred to all the HPO terms associated with those diseases.\n\nPhenoPPIOrth We also evaluated the performance of PhenoPPIOrth (Wang et al., 2013). PhenoPPIOrth is a computational tool that can predict a set of diseases for a given human gene. Specifically, it predicts OMIM disease terms for human genes using protein-protein interaction and orthology data. Then it also maps the predicted OMIM terms to HPO terms using the disease-HPO mapping available in the HPO website11. We downloaded the pre-computed preditions from the PhenoPPIOrth website. Compared to PHENOStruct, PhenoPPIOrths performance was quite low (see Table S3). It is important to note that PhenoPPIOrth makes predictions for only a subset of proteins with respect to all three ontologies (1487, 175 and 155 in Organ, Inheritance and Onset subontologies, respectively). One of the main reasons is that HPO annotations are generated using three sources: OMIM, Orphanet and DECHIPER but PhenoPPIOrth predicts only OMIM terms.\n\nWe use term-centric AUC or macro AUC as our primary evaluation measure for reporting results. In addition, we use several protein-centric measures. Protein-centric precision and recall at a given threshold t are defined as\n\nPrpc(t) = 1N∑i=0NTP(t)iTP(t)i + FP(t)i,\n\nRcpc(t) = 1N∑i=0NTP(t)iTP(t)i + FN(t)i,\n\nwhere TP(t)i, FP(t)i and FN(t)i are the number of true positives, number of false positives and number of false negatives w.r.t. protein i at threshold t. Now we can define protein-centric F-max as\n\nFpc-max = maxt2Prpc(t)Rcpc(t)Prpc(t) + Rcpc(t).\n\nIn this section we present performance of all five methods using several performance measures.\n\nThe performance is evaluated using macro AUC and protein-centric F-max, Precision and Recall (as defined above) on the complete HPO graph (i.e. true-path rule is applied to annotations and predictions).\n\nThe performance is evaluated by using the exact annotations (i.e. only leaf terms) as ground truth. In other words, true-path rule is not applied. Performance is presented using macro AUC and protein-centric F-max, Precision and Recall as defined above.\n\nFirst we ranked the test proteins in descending order of the protein-centric precision of their Organ subontology predictions made by PHENOstruct. Then we retrieved the 25 false positive predictions for the top 17 proteins in that list. Next, we performed online searches using the pair of protein name and phenotype name as the query for the search engine. This resulted in a list of publications for each false positive prediction. Then we manually extracted the excepts from those papers that contained supporting evidence that suggests the particular false positive is in fact correct. Using this manual process we found evidence for 14 of the 25 false predictions considered for this study (see Table S5). For two of the cases the evidence comes from studies involving mice (indicated within parentheses with the PubMed ID). Overall success of this study strongly suggests that the performance of PHENOstruct is under-estimated due to the incompleteness of the current gold standard.\n\nThe columns “HPO term”, “PubMed ID” and “Evidence” provides the false positive prediction made by PHENOStruct for the given gene, PubMed ID of the literature that contains evidence which actually suggests that the prediction should be considered true and the excerpt from that literature which contains the evidence, respectively. We used the 25 false positive predictions for the 17 proteins that had the highest individual protein-centric precision and found evidence for 14 predictions. Two of the evidence comes from studies involving mice (indicated within parentheses with the PubMed ID)\n\nThe macro AUC for the Inheritance subontology is 0.72. Terms are displayed in ascending order of frequency.\n\n\nReferences\n\nRobinson PN: Deep phenotyping for precision medicine. Hum Mutat. 2012; 33(5): 777–780. PubMed Abstract \| Publisher Full Text\n\nKhler S, Doelken SC, Mungall CJ, et al.: The human phenotype ontology project: linking molecular biology and disease through phenotype data. Nucleic Acids Res. 2014; 42(Database issue): D966–D974. 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\nAymé S, Schmidtke J: Networking for rare diseases: a necessity for Europe. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz. 2007; 50(12): 1477–1483. PubMed Abstract \| Publisher Full Text\n\nBragin E, Chatzimichali EA, Wright CF, et al.: DECIPHER: database for the interpretation of phenotype-linked plausibly pathogenic sequence and copy-number variation. Nucleic Acids Res. 2014; 42(Database issue): D993–D1000. 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\nRobinson PN, Köhler S, Oellrich A, et al.: Improved exome prioritization of disease genes through cross-species phenotype comparison. Genome Res. 2014; 24(2): 340–348. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoreau T, Tranchevent LC: Computational tools for prioritizing candidate genes: boosting disease gene discovery. Nat Rev Genet. 2012; 13(8): 523–536. PubMed Abstract \| Publisher Full Text\n\nBi W, Kwok JT: Multi-label classification on tree- and dag-structured hierarchies. In Lise Getoor and Tobias Scheffer, editors, New York, NY, USA, ACM. Proceedings of the 28th International Conference on Machine Learning (ICML-11). 2011; 17–24. Reference Source\n\nSilla CN Jr, Freitas AA: A survey of hierarchical classification across different application domains. Data Min Knowl Discov. 2011; 22(1–2): 31–72. Publisher Full Text\n\nObozinski G, Lanckriet G, Grant C, et al.: Consistent probabilistic outputs for protein function prediction. Genome Biol. 2008; 9(Suppl 1): S6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsochantaridis I, Joachims T, Hofmann T, et al.: Large margin methods for structured and interdependent output variables. J Mach Learn Res. 2005; 6: 1453–1484. Reference Source\n\nSokolov A, Ben-Hur A: Hierarchical classification of gene ontology terms using the GOstruct method. J Bioinform Comput Biol. 2010; 8(2): 357–376. PubMed Abstract \| Publisher Full Text\n\nSokolov A, Funk C, Graim K, et al.: Combining heterogeneous data sources for accurate functional annotation of proteins. BMC Bioinformatics. 2013; 14(Suppl 3): S10. PubMed Abstract \| Free Full Text\n\nChatr-aryamontri A, Breitkreutz BJ, Heinicke S, et al.: The BioGRID interaction database: 2013 update. Nucleic Acids Res. 2013; 41(Database issue): D816–D823. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSzklarczyk D, Franceschini A, Kuhn M, et al.: The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res. 2011; 39(Database issue): D561–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFunk C, Kahanda I, Ben-Hur A, et al.: Evaluating a variety of text-mined features for automatic protein function prediction with GOstruct. J Biomed Semantics. 2015; 6: 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang P, Lai WF, Li MJ, et al.: Inference of gene-phenotype associations via protein-protein interaction and orthology. PLoS One. 2013; 8(10): e77478. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchietgat L, Vens C, Struyf J, et al.: Predicting gene function using hierarchical multi-label decision tree ensembles. BMC Bioinformatics. 2010; 11: 2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRadivojac P, Clark WT, Oron TR, et al.: A large-scale evaluation of computational protein function prediction. Nat Methods. 2013; 10(3): 221–227. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuttenhower C, Hibbs MA, Myers CL, et al.: The impact of incomplete knowledge on evaluation: an experimental benchmark for protein function prediction. Bioinformatics. 2009; 25(18): 2404–2410. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnderson S, Bankier AT, Barrell BG, et al.: Sequence and organization of the human mitochondrial genome. Nature. 1981; 290(5806): 457–465. PubMed Abstract \| Publisher Full Text\n\nTaylor RW, Turnbull DM: Mitochondrial DNA mutations in human disease. Nat Rev Genet. 2005; 6(5): 389–402. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWallace DC, Singh G, Lott MT, et al.: Mitochondrial DNA mutation associated with Leber’s hereditary optic neuropathy. Science. 1988; 242(4884): 1427–1430. PubMed Abstract \| Publisher Full Text\n\nHolt IJ, Harding AE, Morgan-Hughes JA: Deletions of muscle mitochondrial DNA in patients with mitochondrial myopathies. Nature. 1988; 331(6158): 717–719. PubMed Abstract \| Publisher Full Text\n\nJones KS: A statistical interpretation of term specificity and its application in retrieval. J Doc. 1972; 28(1): 11–21. Publisher Full Text\n\nBaumgartner WA Jr, Cohen KB, Fox LM, et al.: Manual curation is not sufficient for annotation of genomic databases. Bioinformatics. 2007; 23(13): i41–i48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith CL, Eppig JT: The mammalian phenotype ontology: enabling robust annotation and comparative analysis. Wiley Interdiscip Rev Syst Biol Med. 2009; 1(3): 390–399. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKahanda I, Funk C, Verspoor K, et al.: Data and software associated with PHENOstruct: Prediction of human phenotype ontology terms using heterogeneous data sources. Zenodo. 2005. Data Source" }	[ { "id": "9567", "date": "14 Aug 2015", "name": "Peter N. Robinson", "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 present a clever strategy for using a machine learning approach to predict associations between genes and human phenotype ontology (HPO) terms. The HPO, as many other ontologies like the Gene Ontology, has a hierarchical structure such that annotations to HPO terms are inherited up the structure of the ontology. For instance, if we say that a patient has \"ventricular septal defect\", we implicitly annotate the patient to all of the ancestor terms of \"ventricular septal defect\" such as \"abnormality of the ventricular septum\". This creates a problem for naive machine learning approaches that make HPO term/Gene association predictions one at a time. If a prediction is YES for \"ventricular septal defect\" but NO for \"abnormality of the ventricular septum\", then the result is mutually inconsistent. A number of machine learning algorithms have emerged to tackle this problem, including one that the authors previously learned for an analogous project with Gene Ontology predictions. In general, the paper is very well done and it is likely to be accessible to a wide audience because it is well written. The topic of HPO annotation prediction is very new, and there is no other published work on the topic that I am aware of at present, although a number of groups, including the authors, participated in a CAFA competition at the 2014 ISMB.Suggestions:The authors should cite GeneMania (The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function Warde-Farley et al, 2010, NAR). The authors should state the version of the HPO and the HPO annotation data they used. In the meantime, the number of annotations has increased substantially, and a number of improvements to the HPO structure have been made (for instance, the Organ abnormality term has been renamed to Phenotypic abnormality). It would be nice to have a little more self-contained explanation about some of the methods employed, such as Clu-HMC-Ens. The authors observed an excellent AUROC score for mitochondrial inheritance, and speculate that the reason is that the mitochondrial genome is well studied. I suspect that the true reason might be that mitochondrial genes have a very specialized functional profile (energy etc) that is much more homogeneous than say \"autosomal recessive\". The Mammalian phenotype ontology is not only used to annotate the rat genome, but is the major tool used to annotate the mouse genome, and is an extremely useful resource used now by the International Mouse Phenotyping Consortium the Mouse Genome Informatics group, and many others. This should be added to the text.", "responses": [] }, { "id": "9995", "date": "28 Aug 2015", "name": "Shaillay Dogra", "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\nA well written article with detailed methodology towards mapping genes to diseases. The method proposes to overcome the limitations of traditional approaches, which take single-label at a time. Author's approach uses structured prediction that takes into account related set of labels.Some general points for authors consideration:What is a possible, direct application of this method? How can this be integrated with other software tools or called as a service, for example? Different data sources have been used of which protein-protein interactions, for example, is a noisy one with false-interactions reported via experimental methods like co-precipitation and yeast-two-hybrid systems. Can the author's comment on how the quality of data affects their approach? How to deal with noisy data sources or reduce the weight of contribution from that particular data source? How easy is to update this, for example, integrating latest PubMed abstracts? What's the pipeline or process to do so?Some specific points for authors to consider:Why do we expect more genes to be disease-causing? Is it just a general line of reasoning that functional genes should cause an aberrant phenotype if they do not work properly? Perhaps the authors can expand a bit more on the problem formulation - \"therefore, it is important to explore the feasibility...\" Under \"Approach\", perhaps the authors can explain a bit on HMC for the benefit of the readers. Also, a bit more on \"structured prediction/learning\" in layman terms or illustrated with an example can help the reader grasp the concept. Authors can consider simplifying this into shorter sentences for easy grasp - \"An alternate approach is to use a single classifier...\" Under \"HPO annotations\" - for general understanding, could the authors tell more about why they 'removed terms that were not annotated to 10 or more genes.' Under \"Literature\" - abstracts extracted are from 2013 and not up to date with 2015. Under \"Literature\" - for general understanding, could the authors tell more about why they 'filtered to keep only the low frequency words'. Under \"variants\" - does this data from Uniprot covers data sources like clinvar, dbGaP, GWAS studies etc? Under \"Models\", for the general understanding of the reader could the authors expand on what is meant by a structured model? Figure 3 - last part of panel is barely legible. Under \"Evaluation\" could the authors expand on what's implied by term-centric and protein-centric, F-max? \"The human mitochondrial DNA... Due to this, and the relative...\" - I am not sure if this is the reason. Authors can consider simplifying this into shorter sentences for easy grasp - \"As a potential improvement to PHENOstruct...\" I really liked the way authors dealt with \"validating false positives\" (text and table S5). Figure S1 - last part of panel is almost illegible. Supplementary material, \"Performance Measures\" - for the benefit of the readers, what does F-max mean in a literal, intuitive sense? Table S4 -- for the benefit of the readers, what is 'true-path rule', 'ground truth', 'macro AUC'...?", "responses": [] } ]	1	https://f1000research.com/articles/4-259
https://f1000research.com/articles/4-257/v1	15 Jul 15	{ "type": "Case Report", "title": "Case Report: melanoma and melanocytic nevus differentiation with reflectance confocal microscopy.", "authors": [ "Joanna Łudzik", "Alexander M Witkowski", "Giovanni Pellacani", "Joanna Łudzik", "Giovanni Pellacani" ], "abstract": "Historically, melanoma has been typically diagnosed by naked-eye examination and confirmed with invasive biopsy. However, recently the use of reflectance confocal microscopy enables non-invasive bedside diagnosis of clinically equivocal lesions. We present a case in which reflectance confocal microscopy was used to evaluate two skin lesions in the same patient confirming the diagnosis of a melanoma and potentially avoiding invasive biopsy in the second benign melanocytic lesion. Clinicians should be aware of the availability of new non-invasive technologies that can aid in early diagnosis of malignant skin tumors and potentially reduce the number of benign lesion excisions.", "keywords": [ "Reflectance confocal microscopy", "Melanoma", "Dysplasia", "Nevi", "Dermatology" ], "content": "Introduction\n\nWe report a case of a patient with multiple atypical melanocytic nevi evaluated with dermoscopy and reflectance confocal microscopy during a referral skin cancer control visit.\n\n\nBackground\n\nSkin tumor diagnosis can be difficult due to the variable clinical presentation of skin lesions. In order to correctly identify melanoma at its earliest stage, the use of dermoscopy has been shown to significantly increase the sensitivity and specificity of diagnosis when compared to traditional naked-eye examination1,2. In equivocal cases benign lesions may be excised when further cytological information is required to rule out malignancy. Recently reflectance confocal microscopy (RCM) use in clinical practice has been shown to further improve early melanoma diagnosis non-invasively by providing an in-vivo optical biopsy at histologic resolution down to a depth of 200 µm of skin tissue3–7. Moreover RCM has been shown to significantly reduce the number of unnecessary excisions in different settings8–10.\n\nIn this article we review the clinical, dermoscopic, and RCM presentation of two lesions in the same patient controlled with the gold standard of histopathology diagnosis.\n\n\nCase report\n\nA 65 year old Caucasian female (Fitzpatrick skin type III) presented to the dermatology department at the University of Modena and Reggio Emilia (UNIMORE) with referral from a general practitioner for two skin lesions. Her past medical history and family history were negative for melanoma. No other significant medical history was noted. The patient reported having several invasive biopsies of dysplastic nevi in the past, the last reported biopsy in 2012. In 2013 the patient had her last naked-eye skin cancer screening by a private dermatologist with no worrisome skin lesions identified or recommended for biopsy.\n\n\nClinical naked-eye examination findings\n\nThe patient presented with a high numerosity of multiple irregularly shaped nevi located mainly on the back and lower legs. Lesion number 1, located on the upper right back, presented with ABCD (asymmetry, irregular borders, multiple colors, diameter >6mm) positive criteria and was of highest concern as it was the largest solitary macule on the back. Lesion number 2, located on the upper left shoulder, also presented with ABCD (asymmetry, irregular borders, multiple colors, diameter >6mm) positive criteria (Figure 1A, B, D).\n\nA. Clinical overview of the patient. B. Lesion 1: Melanoma - naked-eye clinical close-up. C. Lesion 1: Melanoma - digital dermoscopy view. D. Lesion 2: Dysplastic nevus - naked-eye clinical close-up. E. Lesion 2: Dysplastic nevus - digital dermoscopy view.\n\n\nDigital dermoscopy findings\n\nDermoscopy evaluation was performed with both a handheld dermatoscope and sequential digital dermoscopy (videodermoscopy). Lesion 1 presented with dermoscopic findings including: asymmetry, irregular reticular network with areas of eccentric hyperpigmentation, blue-white areas, and peppering representing early regression (Figure 1C). Lesion 2 presented with dermoscopic findings including: asymmetry and eccentric hyperpigmented network (Figure 1E).\n\n\nReflectance confocal microscopy findings\n\nAfter dermoscopic evaluation and storage of both lesions in the UNIMORE digital dermoscopy database system the patient was referred for further evaluation with reflectance confocal microscopy. RCM images were obtained with a reflectance confocal microscope (Vivascope1500; MAVIG GmBH, Munich, Germany) using a 830 nm laser at a maximum power of 20 mW. RCM images of 0.5 × 0.5 mm were acquired with a lateral resolution of 1 μm and an axial resolution of 3–5 μm and stitched into composite images that covered between 4 to 8 square mm mosaics (VivaCube; MAVIG GmBH, Munich, Germany). A minimum of three mosaics were obtained at different depths, corresponding to the stratum granulosum/spinosum, the dermo-epidermal junction, and the papillary dermis.\n\nLesion number 1 presented with the following findings at the dermo-epidermal junction: predominant meshwork architecture composed of enlarged interpapillary spaces with junctional nests containing atypical melanocytes. Additionally, there were areas of loss of architectural structure and replacement by non-specific architecture with bundles of atypical dendritic-type melanocytes. The epidermis presented with complete disarrangement with an atypical honeycombed pattern and presence of a high numerosity of heterogeneously shaped pagetoid cells (Figure 2). RCM examination was therefore confirming the diagnosis of melanoma, later confirmed by histopathology report.\n\nReflectance confocal microscopy (RCM) imaging. A. Mosaic-map overview. B. Presence of non-specific pattern (), atypical meshwork pattern (↑), aggregates of dendritic-type atypical melanocytes in bundles (^), location: dermo-epidermal junction. C. Disarrangement of the epidermis with an atypical honeycombed pattern and presence of a high numerosity of heterogeneously shaped pagetoid cells (), location: epidermis.\n\nLesion number 2 presented with the following findings at the dermo-epidermal junction (DEJ): predominant ringed and clod architecture, representing junctional lentiginous proliferation of melanocytes and dermal nests respectively, with absence of atypical cells. Additionally at the DEJ there were few areas of meshwork architecture. The epidermis presented with a regular honeycombed pattern with few inflammatory cells (Figure 3). RCM examination was therefore suggestive of a dysplasic nevus, later confirmed by histopathology report.\n\nReflectance confocal microscopy (RCM) imaging. A. Mosaic-map overview. B. Presence of ringed and clod (*) architecture, representing junctional lentiginous proliferation of melanocytes and dermal nests respectively, location dermo-epidermal junction. C. Regular honeycombed pattern with few inflammatory cells, location: epidermis.\n\n\nDiscussion\n\nThe purpose of our case-report was to present a typical scenario encountered by clinicians in daily practice where multiple lesions are referred for skin cancer examination. The methodology of full body dermoscopy evaluation to identify potentially high risk skin lesions and further evaluation with RCM imaging provides trained experts with cellular information about skin lesions non-invasively at the bedside. Ultimately this information can aid in early diagnosis of malignant skin tumors and moreover potentially reduce removal of benign lesions, saving patients from unnecessary scaring and healthcare costs. In the case of our patient after RCM evaluation was performed it was recommended to the patient to remove the melanoma (lesion 1) and to follow-up the melanocytic nevus (lesion 2) with annual sequential digital dermoscopy (videodermoscopy) evaluation. Due to the patient’s request and concern both lesions were removed and sent for histopathology evaluation where lesion 1 was confirmed to be a melanoma (0.62 mm depth) and lesion 2 a dysplastic nevus. In conclusion, this case is a classic example where implementation of non-invasive screening methods can help confirm tumor diagnosis immediately at the bedside and help to reduce the waiting time for necessary removal of a melanoma and potentially reduce the unnecessary excision of a dysplastic nevus.\n\n\nConsent\n\nWritten informed consent for publication of patient clinical details and/or clinical/digital dermoscopy/RCM images was obtained from the patient.", "appendix": "Author contributions\n\n\n\nJL and AW conceived the case-report. AW carried out patient examination and image collection. JL and GP contributed to the design and preparation of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting 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\nReferences\n\nRosado B, Menzies S, Harbauer A, et al.: Accuracy of computer diagnosis of melanoma: a quantitative meta-analysis. Arch Dermatol. 2003; 139(3): 361–7; discussion 366. PubMed Abstract \| Publisher Full Text\n\nLallas A, Argenziano G: Dermatoscope--the dermatologist’s stethoscope. Indian J Dermatol Venereol Leprol. 2014; 80(6): 493–4. PubMed Abstract \| Publisher Full Text\n\nArgenziano G, Cerroni L, Zalaudek I, et al.: Accuracy in melanoma detection: a 10-year multicenter survey. J Am Acad Dermatol. 2012; 67(1): 54–9. PubMed Abstract \| Publisher Full Text\n\nArgenziano G, Moscarella E, Annetta A, et al.: Melanoma detection in Italian pigmented lesion clinics. G Ital Dermatol Venereol. 2014; 149(2): 161–6. PubMed Abstract\n\nPellacani G, Pepe P, Casari A, et al.: Reflectance confocal microscopy as a second-level examination in skin oncology improves diagnostic accuracy and saves unnecessary excisions: a longitudinal prospective study. Br J Dermatol. 2014; 171(5): 1044–51. PubMed Abstract \| Publisher Full Text\n\nStanganelli I, Longo C, Mazzoni L, et al.: Integration of reflectance confocal microscopy in sequential dermoscopy follow-up improves melanoma detection accuracy. Br J Dermatol. 2015; 172(2): 365–71. PubMed Abstract \| Publisher Full Text\n\nBraga JC, Macedo MP, Pinto C, et al.: Learning reflectance confocal microscopy of melanocytic skin lesions through histopathologic transversal sections. PLoS One. 2013; 8(12): e81205. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRajadhyaksha M, Grossman M, Esterowitz D, et al.: In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J Invest Dermatol. 1995; 104(6): 946–52. PubMed Abstract \| Publisher Full Text\n\nScope A, Benvenuto-Andrade C, Agero AL, et al.: In vivo reflectance confocal microscopy imaging of melanocytic skin lesions: consensus terminology glossary and illustrative images. J Am Acad Dermatol. 2007; 57(4): 644–58. PubMed Abstract \| Publisher Full Text\n\nŁudzik J, Witkowski AM, Pellacani G: Pseudomelanoma follow-up of a recurrent naevus with dermoscopy and reflectance confocal microscopy. J Eur Acad Dermatol Venereol. 2015. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9591", "date": "22 Jul 2015", "name": "Marco Ardigò", "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 described case report on the use of RCM for melanoma and dysplastic nevi evaluation. Clinic, dermoscopic and confocal figures are well presented and clearly described the case.", "responses": [] }, { "id": "9603", "date": "23 Jul 2015", "name": "Joyce Lee", "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 nice concise case report to increase our awareness of the availability of the new non-invasive tool (RCM), which can aid in early diagnosis of melanoma and hence, reducing the number of benign lesions excision especially for patients with atypical melanocytic nevi syndrome.The knowledge added is limited but useful for other practitioners.The title, abstract, article content and overall discussions are appropriate with good confocal and dermoscopy images attached. However, it would be nice to have histological images for correlation as well. The conclusions are sensible and balanced as well. However, it is important to indicate in the case report that the RCM is a useful complementary tool but should not replace clinical judgement on the necessity to closely monitor nevi and still offer excision if in doubt.", "responses": [] }, { "id": "10277", "date": "11 Sep 2015", "name": "Gerardo Ferrara", "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\nIn my eye, the manuscript is very good. I wonder if the Authors could provide the histological images of the cases in order to achieve a greater completeness in their report.In the background section, 3rd sentence, 'cytologcal infomation' should be changed into 'histopathological information'.I have no additional comment other than congratulating with the authors for their work.", "responses": [] } ]	1	https://f1000research.com/articles/4-257
https://f1000research.com/articles/3-289/v1	24 Nov 14	{ "type": "Research Article", "title": "Platforms for publishing and archiving computer-aided research", "authors": [ "Konrad Hinsen" ], "abstract": "Computational models and methods take an ever more important place in modern scientific research. At the same time, they are becoming ever more complex, to the point that many such models and methods can no longer be adequately described in the narrative of a traditional journal article. Often they exist only as part of scientific software tools, which causes two important problems: (1) software tools are much more complex than the models and methods they embed, making the latter unnecessarily difficult to understand, (2) software tools depend on minute details of the computing environment they were written for, making them difficult to deploy and often completely unusable after a few years. This article addresses the second problem, based on the experience gained from the development and use of a platform specifically designed to facilitate the integration of computational methods into the scientific record.", "keywords": [ "Computational models", "publishing", "software tools", "scientific record" ], "content": "Introduction\n\nIn the course of a few decades, computers have become essential tools in scientific research and have profoundly changed the way scientists work with data and with theoretical models. Until now, these changes have had little impact on the scientific record, which still consists mainly of narratives published in articles that are limited in size and type of contents, and linked to each other through citations. Some particularly data-intensive fields of research also have their own digital repositories. An early example is the Protein Data Bank1, which publishes and archives structures of biological macromolecules. However, for most domains of research, no such repositories exist, and datasets are mostly neither published nor archived.\n\nWhile the technology used for publishing and archiving the scientific record has shifted from the printing press and libraries to PDF files and Web servers, the kind of information that is being stored has hardly changed. Some scientific journals offer the possibility of submitting “supplementary material” with articles, as a way to circumvent the habitual length restrictions, and for providing unprintable information such as videos. In principle, the data underlying an article can be published as supplementary material as well but this remains an exception and is in fact of little practical interest. The reasons are the various restrictions on file formats and file sizes imposed arbitrarily by different scientific journals, but also the often difficult access to these electronic resources, which usually requires a careful study of each journal’s Web site. Only the recent advent of Web repositories2–4 and peer-to-peer networks5 for scientific data has finally made the publication of scientific data accessible to any scientist willing to do so.\n\nThe increasing number of mistakes found in published scientific findings based on non-trivial computations6,7 has made evident the necessity of making computational science more transparent by publishing software and datasets along with any descriptions of the results obtained from them. While this is now technically possible, and initiatives have been started to create incentives for scientists to invest the additional effort required for making such material available8, much more work remains to be done to ensure that published software and datasets can actually be understood, verified, and reused by other scientists. This is particularly important because computational methods are becoming an essential aspect of all scientific research, including experimental and theoretical work in which computations are not the main focus of activity. It is therefore more appropriate to discuss these issues in the context of “computer-aided research” rather than the more narrow specialty called computational science.\n\nThe most current efforts in this direction (see e.g. Ref. 9–11) start from the status quo of computation in science and propose small-step improvements in order to facilitate adoption by the scientific community. The work presented in this article takes the opposite approach of starting from the requirements of the scientific record and exploring how software and electronic datasets need to be prepared in order to become useful parts of this record. Both approaches are complementary: while ease of adoption is important for rapid improvement, it is also important to have a clear idea of the goal that should ultimately be reached, in order to avoid getting stuck in technological dead-ends.\n\nThe main contribution made by this work are the following insights:\n\nThe traditional distinction between “software” and “data” is not adapted to the needs of scientific communication. It should be replaced by a distinction between “computational tools” and “scientific contents”. Scientific contents are the information that is conserved permanently in the scientific record. It includes experimental data, theoretical models, and computational protocols.\n\nTheoretical models and computational protocols include algorithms, whose permanent conservation requires a precise and stable representation with well-defined semantics.\n\nScientific contents consist of distinct information items linked by dependency and provenance information, which must be stored in the scientific record as well in order to ensure reproducibility. The same information can be used for attributing credit to everyone involved in producing the information.\n\nA proof-of-concept implementation shows that these goals are attainable with the existing technology.\n\nThese insights are the result of developing a new computational framework, called ActivePapers12, and using it for several research projects in the field of biomolecular simulation. The ActivePapers framework is not the principal result of this work, and it will be described only in as much as its technical characteristics matter for the conclusions. In fact, one of the conclusions of this study is that the current ActivePapers framework does not satisfy all the requirements for integrating software and datasets into the scientific record. Nevertheless, ActivePapers is a useful tool in spite of its imperfections, and researchers interested in exploring the future of computational science are invited to use it and build on it.\n\n\nThe state of the art\n\nThe main consequence of the computerization of scientific research has been an enormous increase in the volume and complexity of scientific information. In the pre-computing era, experimental data rarely exceeded a couple of printed tables, and the description of experimental protocols and theoretical models rarely exceeded a few pages. This information could easily be recorded in text form, replicated on printed paper, and stored in libraries. Moreover, each such piece of information could be read, understood, and verified by an individual scientist with sufficient experience in the underlying domain of research.\n\nToday’s computer-aided research is characterized by large datasets which can only be stored electronically, and processed by software that embodies theoretical models and methods. While there is no fundamental difference between a printed table and a computer file, or between an experimental protocol and a piece of software, there is a very important practical difference: the size and complexity of the electronic versions often puts them beyond the limit of what an individual scientist can understand or verify. Moreover, neither datasets nor software have traditionally been published along with the articles describing a scientific study, making verification strictly impossible even in cases of moderate size and complexity.\n\nThis situation has lead to numerous mistakes in published results based on computation, of which the identified and publicized cases6,7 are only the tip of the iceberg. In fact, these cases plus our daily experience with buggy computer software in other aspects of life should make us consider any computational result in science suspect, unless clear evidence for verification and quality assurance is provided by the authors. This evidence includes software testing, formal proof of correctness, comparison of the outcomes of independent computational studies, and proof of rigor in the non-automatic aspects of the application of computational protocols. However, the most basic requirement for building confidence in computational results is total transparency: the publication of all datasets and software used during a scientific study. This is the main goal of the Reproducible Research movement, which has been gaining traction over the last years13–15. The transition to more trustworthy computer-aided research requires changes of both technological and social nature. The computational tools that scientists use must make replication not only possible, but straightforward. The maintainers of the scientific record must integrate electronic datasets and software into their archiving and publication process and reject submissions based on computational work that is not made fully transparent. But most of all, scientists must adopt a much more critical attitude towards computational results. The current tacit convention in computational science is that published results are assumed correct unless there is clear evidence suggesting a mistake. As C.A.R. Hoare famously said16, “There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies”. This observation applies equally to computational science, where the vast majority of published results have no obvious deficiencies, but are obtained using software that is much too complicated for anyone to be certain about its correctness.\n\nThe term “scientific record” refers to the totality of published scientific findings in history. It started to become organized in 1665 with the creation of the first scientific journal, the Philosophical Transactions of the Royal Society. A scientific journal publishes articles, which are narratives explaining the motivation for a specific study, the methods being applied, the observations made, and the conclusions drawn. The exact observations are provided in the form of tables and figures. To this day, scientific journals are the most visible part of the scientific record, although digital repositories have become an important second pillar in several domains of research. These repositories store datasets that are too big to be included as tables or figures. In addition, repositories facilitate the reuse of data in other studies, including the application of data mining techniques in meta-analysis studies.\n\nOne of the main characteristics of the scientific record is its permanence. Once an article is published in a journal, or a dataset in a database, a permanent reference is attached to it, and the publisher or database accepts the moral obligation to make the information accessible through this reference for as long as possible. Traditionally, this reference is a citation taking a more or less standardized format. With the transition to electronic publishing, the role of the permanent reference is fulfilled by a Digital Object Identifier (DOI)17, which is defined by the international standard ISO 26324:2012. For published articles, permanence also applies to the contents: once published, they cannot change. Mistakes detected after publishing can be corrected only by publishing a separate short article called “erratum”. The situation is less clear for repositories, some of which, for example the Protein Data Bank (PDB)1, do correct formal mistakes in their electronic records, but do not allow changes that would modify their scientific interpretation. More recently, the idea of versioning has been applied by some publishers, e.g. F1000Research: a DOI refers permanently to a specific article, but multiple versions of this article remain accessible permanently in order to document the evolution of the publication.\n\nThe permanence of the scientific record applies only to the preservation of the original expression of each information item, but not to its semantics. A published article can become unintelligible because of changes in terminology and in the scientists’ education. A modern physicist would not recognize the theory of classical mechanics in Newton’s “Philosophiæ Naturalis Principia Mathematica”18 without prior training in Latin and in the history of science. The longevity of electronic datasets relies on a careful documentation of the data models and data formats being used, and on proper curation of submissions by the database managers to ensure adherence to these formats. Scientific journals generally use the PDF/A format for their articles, which is a variant of the popular PDF format designed specifically for long-term archiving, defined by the international standard ISO 19005:2005. The latest generation of uncurated Web repositories2–4 allows any computer file to enter the scientific record without even requiring a definition of the data format. It is to be expected that much of the information in these repositories will quickly become unusable.\n\nSoftware presents a particular challenge because formal data models for executable computer code are rare and in particular non-existent for the programming languages commonly used in computational science today. This is one of the key problems addressed in this work. Many practitioners consider the idea of preserving scientific software for many years unrealistic, and some even argue that it is unnecessary because computational methods change so rapidly that their long-term conservation is of no interest. The latter argument is manifestly not valid in general. As an example, the DSSP method for defining secondary-structure elements in proteins19 was published more than thirty years ago and is still widely used today. The example of DSSP is also interesting in that the most widely used software implementing DSSP today20 does not in fact implement the exact method published in the original paper. The differences are not documented anywhere at this time, and scientific papers using the modern software systematically cite the original paper without further comment. It can thus be assumed that most DSSP users are not aware of the fact that they are using a modified method. If the original method had been published in executable form and preserved until today, such discrepancies could have been avoided.\n\nOne of the cornerstones of scientific research is the reproducibility of scientific findings: in principle, anyone applying a published research protocol to a sufficiently similar object of study should obtain very similar results. The notion of reproducibility is necessarily imprecise. In performing experiments, the samples and environments are never exactly the same. Moreover, the description of an experimental protocol is never fully complete, because the experimenters cannot know with certainty which parameters need to be recorded. The level of similarity required in the experimental setup and in the results for the latter to be considered a successful reproduction thus varies considerably across domains of research. A reproduction attempt, whether successful or not, always yields new scientific knowledge, because it explores the impact of variations in the protocols, environments, and samples.\n\nIn computational science, the protocols, input data, and computational environment can in principle be recorded exactly, being digital information. A sufficiently complete recording of this information makes a computational study replicable: another scientist can re-run the exact same computation and obtain exactly identical results. Reproducibility, like in experimental science, refers to the less precise idea of re-doing computations based on the description of the methods in the published narrative, which is supposed to describe the “important” aspects (e.g. a numerical algorithm) but not details generally considered to be irrelevant (e.g. compiler and library version numbers).\n\nReplicability and reproducibility in computational science play different roles in the scientific process. Replicability is part of quality assurance. If scientist B can replicate scientist A’s computation, this shows that A has provided a sufficiently precise and complete record of the original work. This is far from trivial because a precise and complete description of a computational study is a dataset that is both large and complex. Moreover, it is often difficult to obtain in today’s computing environments. Reproducibility plays the same role as in other branches of science: it establishes which aspects of a computational protocol are important for reaching specific conclusions.\n\nHowever, replicability and reproducibility are not completely independent. For all but the simplest computational methods, reproducibility requires replicability. If a reproduction attempt leads to significantly different results, the cause of the differences must be explored. This is in practice only possible if the original results can at least be replicated. Otherwise, the most probable explanation for the difference is that the original study was insufficiently documented, and is thus of very limited value. But exploring the differences requires more than replicability: the original computational protocol must also be understandable by the scientist who sets out to explore the failure of a reproduction attempt. The ultimate problem is that computational methods have become very complex. A summary in the narrative is not sufficiently detailed to allow reproduction. On the other hand, the actual executable computer code is precise and complete, but even more complex than the method because it also needs to take into account complex technical issues such as performance and resource management. I have outlined solutions to this problem in Ref. 21.\n\nDifferent parts of the scientific process impose different criteria for replicability and reproducibility. Conducting a study requires only short-time local replicability for quality assurance, i.e. the authors must be able to re-run their own computation in their own computational environment. Collaboration requires short-time non-local replicability, because co-workers usually have somewhat different computational environments. Pre- or post-publication peer review requires more stringent short-time non-local replicability, because reviewers are likely to have significantly different computing environments at their disposal. Peer review also requires a minimal level of reproducibility in that reviewers must at least be convinced that they could reproduce the findings if they tried, even though for lack of time they rarely do. Publication and archiving as part of the scientific record add the requirements of long-term reproducibility and thus long-term replicability. All archived data must remain usable and understandable for as long as the study remains of scientific interest, which is usually a few decades.\n\nSince replicability is a purely technical aspect of computer-aided research, it can in principle be guaranteed by software tools. Many ongoing research and development projects aim to create such tools, although for pragmatic reasons the goal is rarely complete replicability, but rather some weaker requirement that is easier to achieve with existing technology. Reproducibility implies human understanding and thus cannot be ensured or even verified automatically. Nevertheless, software tools can help in the process of documenting computational methods in a way that is both replicable and understandable. In the following, I will summarize the currently most popular approaches.\n\nAn approach that has already been widely adopted in domains of research that make heavy use of computation is the conservation of datasets in digital repositories issuing permanent references. A scientific narrative is separately published in a journal, and uses the permanent reference to establish a link with the data. This link between narrative and datasets can be tightened to the point where articles no longer contain any data in the form of tables, but only permanent references to repository entries. A good example for this strategy in chemistry is described in Ref. 22. In that particular work, the software does not enter the scientific record at all, and appears only in metadata stored along with the computational results, where it identifies the software packages, version numbers, and other details of the computational environment. Other initiatives (see e.g. 23, 24) advocate storing a snapshot of the program source code in a digital repository as a dataset, in order to provide at least an archive of the exact version of the software that was used, even if it is difficult or impossible to re-run that software later. This ambivalent attitude towards software stems from the recognition of its fundamental importance on one hand and from the practical impossibility to fully integrate today’s scientific software into the scientific record on the other hand.\n\nThe computational notebook approach11, pioneered by Mathematica25 and recently popularized by the Jupyter project (formerly known as the IPython notebook)26, builds on earlier developments in literate programming27, which have also been applied to computational science directly28,29. It aims to integrate computational methods expressed as working code with input/output data and the scientific narrative. It permits a seamless transition from interactive exploratory work to a documented computational method that can be shared and published. Compared to traditional scripts, computational notebooks represent an important advance in improving reproducibility through improving human understanding. However, none of the existing notebook implementations improve on scripts in replicability, which remains local and short-term. Like a script, a notebook depends on the computational environment in which it was generated. This environment is neither conserved nor even documented in the notebook. A few years later, a notebook still provides a human-readable and rather detailed description of the method, but re-running it is likely to be difficult or impossible. Moreover, the notebook approach does not take into account datasets, unless they are small enough to be included as literal data into the notebook itself. All other data is accessed by usually nonpermanent references such as filenames or Universal Resource Locators (URLs).\n\nSimilar remarks apply to workflow management systems such as Kepler30 or VisTrails31. In fact, workflows, scripts, and notebooks all refer to the same basic concept: the outer algorithmic layer that defines a specific computational study in terms of more generic components. The differences lie in the user interface and in the kind of components that can be used (libraries, executables, Web services, etc.). Some workflow managers can archive these components partially, and also some kinds of datasets, but such support is neither complete nor exhaustive, because the technology on which today’s scientific software is built does not allow this.\n\nThe most comprehensive approach to archiving scientific software in an executable form is based on virtual machine technology32–34. The authors of a computational study produce a virtual machine image that contains their complete computational environment, starting with the operating system, in addition to the problem-specific data and workflows. The resulting archives are in general too big to be archived in today’s general-purpose digital repositories. Moreover, it is not possible to refer to or re-use individual pieces of software or data inside a virtual machine image, nor is it straightforward in general to analyze the software or data except by the tools explicitly provided by the authors. But most importantly, the longevity of archived virtual machine images is uncertain. Executing such an image requires complex and sophisticated software, which for the moment is produced and maintained by non-scientific organizations for reasons completely unrelated to science. Once technological progress makes these efforts obsolete, it must be expected that computations archived as virtual machine images will become unusable.\n\nA conclusion that can be drawn from the approaches summarized above is that today’s computational scientists cannot publish their work in a form that is at the same time executable, understandable by human readers, and reusable. Of the three basic kinds of information in computational science, software, data, and narrative, it is clearly the software that is at the root of the difficulties.\n\nFor understanding the difficulties caused by software, and thus to identify the possible solutions, it is useful to introduce the concept of a platform for scientific computing. A platform defines the interface between a computational infrastructure and computational contents. In particular, the platform defines the exact data formats that the contents must respect, and specifies how each data item will be interpreted. For example, the MP3 standard defines a platform for handling music in computers. It defines a file format for storing sound samples, and defines how an MP3 player interprets the data in such files. Any MP3 player is an implementation of the MP3 platform. Any MP3 file is a piece of contents for the MP3 platform. In general, a platform can be more complex and define formats for many different kinds of data. There are also customizable platforms that define some basic features of their contents but also a mechanism for adding more specific features. The best-known example is the XML platform, which allows working with generic structured text data, including the definition of more specific subformats through DTDs or schemas.\n\nFor a piece of software, the platform required to run it varies considerably as a function of how the software is presented. A compiled executable for the Microsoft Windows platform has very specific requirements concerning the instruction set of the real or virtual processor used to run it, but also concerning how the operating system services are accessed. In addition, its correct function may depend on specific versions of specific dynamically loadable libraries. Although most aspects of the Microsoft Windows platform are documented to some degree, there is no comprehensive documentation for the total, and its complexity makes it unlikely that such a documentation will ever be produced. In practice, only a test run can establish whether a given program works on a given machine. For software published in source code form, the requirements are very different but equally complex. Typical dependencies include a compiler or interpreter for a specific programming language, specific versions of specific libraries, and sometimes even specific files being accessible in specific locations. None of these details are documented comprehensively, which is why installation and deployment of software is very difficult. Even standardized programming languages are not defined with precise semantics, a situation that has already caused many serious problems (see e.g. Ref. 35 for the languages C and C++). This lack of a precisely defined and stable platform for executable code is also the root cause of non-replicability in computer-aided research.\n\nOf the various attempts to remedy this situation, the best known and most successful one is the Java Virtual Machine (JVM)36, originally defined as a support for running software written in the Java language. Today the JVM hosts a variety of languages and ensures a high level of interoperability between them. The goal of the JVM developers was to enable the distribution of executable code via the Web, which users could run in their browsers without any prior installation or configuration. This goal has overall been reached successfully, and with remarkable stability: Java code written in 1995 can still be run without modification. There are only two aspects in which the JVM platform failed to attain universal portability: (1) interfacing to certain operating-system services, such as user interface layers or concurrency management, and (2) floating-point computations. The latter failure is due to a deliberate decision to give up the precise initial specification of the JVM in favor of a less rigid one that leaves more room for performance optimizations. It is still possible to use the original precise floating-point semantics, but in practice this feature is hardly used because most computer users give a higher priority to performance than to replicability.\n\nThe reasons for the JVMs success in establishing a stable software platform are various, and to a significant degree due to the interplay of the commercial strategies of the major companies in the computing market. Among the technical reasons, the main one is the choice of a data model for executable code that is situated at a higher level of abstraction than machine code, but at a lower level than typical programming languages. Machine code evolves rapidly because of progress in processor design, and programming languages evolve, somewhat less rapidly, because of advances in software engineering. Stability can only be found in between these two extremes. Other stable software platforms have adopted the same fundamental approach. In particular, the ECMA standard CLI37 can be considered a more modern implementation of the basic JVM idea. Google’s much more recent Portable Native Client (PNaCL) platform38 chose a more low-level code representation defined by the LLVM project39 and a less precisely defined computational environment, in order to facilitate the adaptation of software written in traditional programming languages. It is too early to say if this approach will turn out to be successful.\n\nIn summary, the JVM experience proves that a significant aspect of the software portability problem can be solved: it is possible to define a stable platform for executable code. The difficulties encountered with the portability of JVM code can be avoided by limiting oneself to the important subset of pure computations, i.e. software that transforms input data into output data but does not interact with its environment in any other way. This observation is important because the scientific aspects of software are always pure computations.\n\n\nActivePapers\n\nThe goal of the ActivePapers research and development project is to define a platform for publishing and archiving computer-aided research. Such a platform should ideally meet all of the following requirements:\n\nA published electronic dataset, in the following called an ActivePaper, should contain all the data, code, and narrative related to a research project, with internal links among all the pieces of information that indicate dependencies and provenance.\n\nAn ActivePaper should be able to refer to data items in previously published ActivePapers. Such references should allow both re-use and attribution of scientific credit.\n\nAn ActivePaper should support large datasets by ensuring compact storage and high-performance data access.\n\nThe representation of executable code inside an ActivePaper should be well-defined, stable, and sufficiently simple to allow implementation on future computing systems with minimal effort. The execution of any piece of code from an ActivePaper should always produce exactly the same results at the bit level.\n\nAny code stored in an ActivePaper should be safe to execute, i.e. it should not be able to cause any harm to the computing environment it is executed on.\n\nAn ActivePaper should contain metadata for provenance tracking and reproducibility.\n\nIt is important to note that it is not required that all software used for a computational study be stored in ActivePapers. On the contrary, it is to be expected that important software tools remain forever outside of the ActivePaper universe and work on ActivePapers as data. This includes everything requiring user interaction, from authoring tools to data visualization programs, and also highly machine-specific software such as batch execution managers. It is also possible to write external code accelerators that take code from an ActivePaper and execute it after optimization and/or parallelization, guaranteeing identical results. While the current state of the art does not provide techniques for making such code accelerators both general and efficient, it is possible and even straightforward to write problem-specific code accelerators, which are simply efficient reimplementations (in a language like Fortran or C) of algorithms stored in an ActivePaper, with the equivalence of the results verified by extensive tests.\n\nThe original ActivePapers architecture40, which was subsequently implemented in the “ActivePapers JVM edition”, was a proof-of-concept design intended to show that it is possible with existing technology to meet all these requirements. The key design and implementation choices were\n\nAn ActivePaper is a file in HDF5 format41. The HDF5 format ensures flexibility, compactness, and high-performance data access.\n\nHDF5 dataset attributes are used to store metadata, including a dataflow graph that records provenance.\n\nAny data item inside a published ActivePaper can be referenced by the combination of the ActivePaper’s DOI and the HDF5 path to the dataset.\n\nExecutable code is stored as JVM bytecode. Any other code representation, in particular human-readable source code in any language, is admissible if a compiler or interpreter exists in the form of JVM bytecode.\n\nThe JVM security model is used to prevent executable code in an ActivePaper from accessing any data outside of the ActivePapers platform. This ensures both security (the user’s computing environment is protected) and the absence of unrecorded dependencies.\n\nIndividual programs inside an ActivePaper can be declared as data importers, in which case they have unrestricted read access to anything, including local files and network resources. They share the write restrictions of all other code, meaning that they cannot modify anything outside of the ActivePapers platform. Moreover, they are never run automatically, but only on explicit user request.\n\nAn implementation of the original ActivePapers platform is available from the ActivePapers Web site12. Its only dependencies are (1) a Java Virtual Machine implementation, (2) the HDF5 library, and (3) JHDF542, a Java interface to the HDF5 library. The ActivePapers software provides a command-line interface for creating ActivePapers, inspecting their contents and metadata, and for running the embedded executable code. This is clearly a minimal working environment. For production use by a wide community of computational scientists, many convenience functions would have to be added: a code and data editor, data visualization, data management, etc.\n\nAn important design decision is related to the management of the metadata that tracks dependencies and provenance. The ActivePapers platform creates and updates this metadata automatically during program execution. From the user’s point of view, an ActivePaper is a collection of datasets and programs, of which the latter can be run individually just like traditional executables or scripts. The ActivePapers platform tracks all data accesses from programs and generates the dependency graph from them. When a program is re-run, typically after modification, all the datasets it generated earlier are deleted automatically. Moreover, when a program reads data generated by another program which has been modified since it was last run, the modified program is re-run automatically to ensure coherence of all data. This automatic dependency handling has worked well in practice. It is the exact opposite of the approach taken by automation tools such as make43, which execute programs according to a manually prepared definition of the dependencies between their results.\n\nThe main difficulty with the original ActivePapers platform is the lack of scientific software compatible with its constraints. All code running inside of the ActivePapers platform must exist as JVM bytecode. All code storing data in an ActivePaper must use the HDF5 library. All code that falls into both categories, which includes in particular the workflow of a specific scientific project, must exist as JVM bytecode accessing the HDF5 library. There is almost no publicly available code that meets these requirements, mostly due to the lack of popularity of the JVM in scientific computing.\n\nIn order to gain experience with the ActivePapers approach in practice, a second implementation was developed for the Scientific Python ecosystem44. Its dependencies are the Python language45, the HDF5 library, the h5py library60 for interfacing HDF5 to Python, and the NumPy library46 which is a dependency of h5py. For the Python edition of ActivePapers, all executable code must exist in the form of Python scripts, which access the datasets through the h5py library. Libraries that contain compiled code, which are very common, cannot be placed inside an ActivePaper, but can be declared as an external dependency. This effectively means that the platform required for using an ActivePaper with such a dependency includes that library in addition to the packages listed above. Adding external dependencies is clearly not desirable from a replicability point of view, but it provides a short-term workaround to the fundamental problem that most scientific software is not ready for long-term replicability.\n\nThe Scientific Python ecosystem provides a large choice of libraries that can be used within these constraints, and the Python language is already very popular for scientific computing, making the ActivePapers Python edition a good vehicle for testing the ActivePapers approach on real research projects. On the other hand, the Python edition cannot fulfill all the requirements listed above. In particular, the Python language lacks sufficiently strong security mechanisms to implement a useful level of user protection. A more subtle problem is the stability of the platform itself. The Python language has no formal specification and in fact evolves together with its principal implementation. The scientific libraries, in particular NumPy, also evolve rather rapidly, with only moderate efforts to maintain compatibility with older versions. The ActivePapers platform records the version of all libraries that were used in the preparation of an ActivePaper, but the long-time usability of these versions is questionable, as in general only the current versions can be expected to work in current computing environments.\n\nThe Python edition of the ActivePapers platform has been used for several research projects in the field of biomolecular simulation, some of which have already been published47–49. Each publication has one or more ActivePaper files deposited as supplementary material, but all the files are also available in digital repositories with DOIs. Among the published ActivePapers, there are software libraries50,51, a database of protein structures52, and combinations of datasets and code that document computational studies53–55. Additional published ActivePapers contain obsolete versions of the pyMosaic library56–58. These files remain permanently available because other ActivePapers depend on them. They also remain usable for as long as the underlying platform remains compatible.\n\nOne problem encountered in the course of these research projects is the relatively low size limit that today’s digital repositories impose on archived files. Zenodo4 provides the most generous limit of 2 GB per file. However, the input data for one study55 contains ten Molecular Dynamics (MD) trajectories for lysozyme in solvent, and requires 10 GB of storage even in compressed form. Since these data were not essential for the subsequent analysis step, which requires only the rigid-body motion of the protein, they were removed from the published files. The alternative would have been to publish each MD trajectory separately as an ActivePaper, and use DOI-based references in the analysis step to refer to this data. Because such references are nearly transparent to the user (the dependencies are downloaded automatically when needed), file size limits apply in practice only to individual HDF5 datasets.\n\nActivePapers proposes another mechanism to reduce file sizes: the deletion of recomputable datasets. Any dataset that was generated by a program stored in an ActivePaper can be replaced by a dummy dataset that retains only the dependency metadata. The full dataset can be re-computed on demand, or automatically when another program tries to read it. Recomputation consists in rerunning the program that generated the dataset initially. This mechanism makes sense only if the replicability of a dataset is guaranteed. In practice, this applies to any program that does not use floating-point operations. The latter are insufficiently specified in most of today’s programming languages, including Python, and therefore floating-point computations can produce different results when the same program is run on two different computers.\n\nExperience with the two current implementations of the ActivePapers idea has shown that all of the requirements defined at the outset can be fulfilled and that the approach works well in practice. In particular, the ActivePapers project has shown that installation-free software deployment and long-time software conservation are possible, contrary to a common belief in the scientific computing community. As mentioned earlier, ActivePapers can achieve these goals because the science part of software takes the form of pure computations, which are possible identically in all computational environments.\n\nThe existence of two distinct ActivePapers platforms is an historical accident that is clearly not desirable. The envisaged solution is a split of the ActivePapers platform into two parts: a data publishing system, which defines the HDF5 conventions for ActivePapers, in particular the metadata, and a code execution system that defines how specific datasets in ActivePapers are interpreted as executable code. Only the second part would differ between the current two implementations, and its separation also opens the way for additional execution systems for other code representations. There is in fact no fully satisfying code representation for scientific computations at this time, which is a strong argument for flexibility in the platform definition.\n\nOne line of future development is an integration of a narrative into the computational methods stored in an ActivePaper. Work on integrating the ActivePapers Python edition with the Jupyter project26 (formerly the IPython notebook) is underway. Unfortunately, non-fundamental technical issues make this a non-trivial project: the various components (HDF5, Python, Jupyter) have different and conflicting requirements and restrictions concerning concurrency. Aside from these software engineering issues, the main question to be solved is how to reconcile the interactivity of the notebook approach with the permanence requirements of the scientific record. The coherence of code and results in a notebook is guaranteed only if it has been executed linearly from start to end. Any interactive manipulation results in general in a non-replicable state. Two solutions are currently explored. The first solution marks notebooks as non-replicable except when executed linearly. No ActivePaper containing such non-replicable notebooks should be accepted by a digital repository. The second solution is to record all interactive code execution in a log, which can then be replayed. After a complete linear execution of the notebook code, the log of interactive executions is then deleted.\n\nAnother direction for future developments explores how to provide a realistic transition from today’s scientific computing environments to future ones that take into account the needs for publishing and archiving computations. One important advantage of the ActivePapers approach in this context is that the minimal requirements for adopting it are modest: any software tool that can work with the ActivePapers file format, which is HDF5 plus a small set of conventions, can read and write publishable datasets. With a very small additional effort, software tools can be adapted to handle ActivePapers metadata and thus ensure dependency and provenance tracking. None of this requires that the software live inside the ActivePapers platform. The challenge for future ActivePapers developments is to facilitate the transition of computational methods from subroutines hidden inside software tools to precise specifications that become part of the scientific record.\n\n\nConclusion\n\nAs I have pointed out earlier21, today’s scientific software fulfills two distinct roles: it is a tool that permits doing computations, but also the only precise and complete description of the models and methods applied in these computations. This situation is the result of the growing complexity of computational methods in science, which make the documentation of these methods in the traditional narrative of a journal article impossible. Understanding and evaluating computational science requires both the possibility to read the source code of all software and to run it on suitable input data. A useful documentation of computational science in the scientific record thus requires archiving all software parts that have an influence on computational results, in a form that can be both inspected and executed, for as long as the study remains relevant for science, which is typically several decades. The ActivePapers project has shown that these goals are achievable in principle using existing technology. It has defined two variants of a platform that gives computational methods the status of publishable content with well-defined data formats that guarantee long-term replicability. However, it has also shown that the vast majority of today’s scientific software is not easily integrated into such a platform. The main reason is that most of the computing technology used by scientists was developed outside of scientific research, for domains of application where replicability is not important.\n\nA key ingredient in the transition from the current state of the art, in which scientific software cannot be fully archived in the scientific record, is a clear distinction between scientific models and methods on one hand, and software tools on the other hand. It is only the models and methods that need to be archived, but not the tools. The long-term usability of the models and methods is guaranteed by a complete and precise specification of their data formats, rather than by a preservation of the tools that work on them. Computational tools must in fact evolve with the progress of technology in order to remain useful to the communities that develop and apply them59. This distinction is completely analogous to how other digital content is handled. We archive articles in PDF format, movies in MPEG3 format, or protein structures in mmCIF format because these formats are well documented and allow anyone, at any point in time, to interpret the archived contents, even if today’s software tools are no longer usable because of the inherent instability of computational environments.\n\nWhat distinguishes computational models and methods from articles, movies, and protein structures is their algorithmic nature, which makes them look like “software” rather than “data”. However, this distinction between software and data, although deeply ingrained in the habits of computational scientists, is not fundamental: software is just a specific kind of data, defined by the existence of some mechanism to execute it. It is very common to use software that treats other software as data, e.g. compilers, interpreters, workflow managers, debuggers, etc. In the context of scientific communication, we should treat software exactly like other kinds of data. The fundamental distinction is not “software” vs. “data”, but “computational tool” vs. “scientific content”.\n\nPublishing and archiving scientific results has always involved an additional effort compared to keeping personal records. Experimentalists do not publish their raw lab notebooks with the user manual of their scientific equipment as an appendix. Theoreticians do not submit scans of their hand-written notes for publication. Publication always implies presenting the work that has been done and its results in a form that is understandable to and usable by other scientists. The same principle applies to computational science, whose practitioners need to be prepared to invest additional effort to make their libraries, programs, and scripts suitable for publishing.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe development of ActivePapers and the first research project in which it as applied were supported by the French “Agence Nationale de la Recherche” (Contract No. ANR-2010-COSI-001-01).\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\nBerman H, Henrick K, Nakamura H: Announcing the worldwide Protein Data Bank. Nature Struct Biol. 2003; 10(12): 980. PubMed Abstract \| Publisher Full Text\n\nDryad. 2014. Reference Source\n\nfigshare. 2014. Reference Source\n\nZenodo. 2014. Reference Source\n\nAcademic torrents. 2014. Reference Source\n\nMerali Z: Computational science: ...Error. Nature. 2010; 467(7317): 775–777. PubMed Abstract \| Publisher Full Text\n\nJoppa LN, McInerny G, Harper R, et al.: Computational science. Troubling trends in scientific software use. Science. 2013; 340(6134): 814–815. PubMed Abstract \| Publisher Full Text\n\nPriem J, Taraborelli D, Groth P, et al.: altmetrics: a manifesto. 2014. Reference Source\n\nStodden V, Miguez S: Best Practices for Computational Science: Software Infrastructure and Environments for Reproducible and Extensible Research. Soc Sci Res Network. 2013; 2322276. Reference Source\n\nStodden V, Leisch F, Peng RD: Implementing Reproducible Research. Chapman and Hall/CRC. 2014. Reference Source\n\nShen H: Interactive notebooks: Sharing the code. Nature. 2014; 515(7525): 151–152. PubMed Abstract \| Publisher Full Text\n\nHinsen K: ActivePapers - computational science made reproducible and publishable. 2014. Reference Source\n\nPeng RD: Reproducible research in computational science. Science. 2011; 334(6060): 1226–1227. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDonoho DL: An invitation to reproducible computational research. Biostatistics. 2010; 11(3): 385–388. PubMed Abstract \| Publisher Full Text\n\nStodden V: Reproducible research: Tools and strategies for scientific computing. Comput Sci Eng. 2012; 14(4): 11–12. Publisher Full Text\n\nHoare CAR: The emperor’s old clothes. Communications of the ACM. 1981; 24(2): 75–83. Publisher Full Text\n\nThe DOI system. 2014. Reference Source\n\nNewton I: Philosophiae Naturalis Principia Mathematica. R Soc. 1686. Reference Source\n\nKabsch W, Sander C: Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers. 1983; 22(12): 2577–2637. PubMed Abstract \| Publisher Full Text\n\nHekkelman M: DSSP 2.1.0. Reference Source\n\nHinsen K: Computational science: shifting the focus from tools to models. [v2; ref status: indexed, http://f1000r.es/3p2]. F1000Res. 2014; 3: 101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarvey MJ, Mason NJ, Rzepa HS: Digital data repositories in chemistry and their integration with journals and electronic notebooks. J Chem Inf Model. 2014; 54(10): 2627–2635. PubMed Abstract \| Publisher Full Text\n\nexec&share. 2014. Reference Source\n\nMozilla Science Lab, Github, and figshare. Code as a research object. 2014. Reference Source\n\nWolfram Research, Inc. Mathematica 2.0. 1991.\n\nProject Jupyter. 2014. Reference Source\n\nKnuth DE: Literate programming. The Computer journal. 1984; 27(2): 97–111. Reference Source\n\nSchulte E, Davison D: Active documents with orgmode. Comput Sci Eng. 2011; 13(3): 66–73. Publisher Full Text \| Free Full Text\n\nXie Y: Dynamic Documents with R and knitr. Chapman & Hall. 2013.\n\nThe Kepler Project. Reference Source\n\nVisTrails. Reference Source\n\nVan Gorp P, Grefen P: Supporting the internet-based evaluation of research software with cloud infrastructure. Softw Syst Model. 2012; 11(1): 11–28. Publisher Full Text\n\nGent I, Kotthoff L: recomputation.org home page. 2014. Reference Source\n\nBoettiger C: An introduction to Docker for reproducible research, with examples from the R environment. arXiv.org. 2014. Reference Source\n\nRegehr J: A guide to undefined behavior in C and C++. Reference Source\n\nLindholm T, Yellin F: The Java Virtual Machine Specification. Prentice Hall. 1999. Reference Source\n\nECMA Standard 335: Common Language Infrastructure CLI. Reference Source\n\nPortable Native client: The “pinnacle” of speed, security, and portability. 2014. Reference Source\n\nLattner C, Adve V: LLVM: A compilation framework for lifelong program analysis and transformation. In Proceedings of the 2004 International Symposium on Code Generation and Optimization, San Jose, CA, USA, 2004; 75–88. Reference Source\n\nHinsen K: A data and code model for reproducible research and executable papers. Pro Comput Sci. 2011; 4: 579–588. Publisher Full Text\n\nThe HDF Group. Hierarchical data format version 5. Reference Source\n\nScientific IT Services of ETH Zürich. JHDF5, HDF5 for Java. Reference Source\n\nWikipedia. Make (software) — Wikipedia, the free encyclopedia. 2014. Reference Source\n\nMillman KJ, Aivazis M: Python for scientists and engineers. Computing in Science Engineering. 2011; 13(2): 9–12. Publisher Full Text\n\nPython Software Foundation. The Python language. 2014. Reference Source\n\nNumPy development team. NumPy. 2014. Reference Source\n\nHinsen K, Hu S, Kneller GR, et al.: A comparison of reduced coordinate sets for describing protein structure. J Chem Phys. 2013; 139(12): 124115. PubMed Abstract \| Publisher Full Text\n\nChevrot G, Hinsen K, Kneller GR: Model-free simulation approach to molecular diffusion tensors. J Chem Phys. 2013; 139(15): 154110. PubMed Abstract \| Publisher Full Text\n\nHinsen K: MOSAIC: a data model and file formats for molecular simulations. J Chem Inf Model. 2014; 54(1): 131–137. PubMed Abstract \| Publisher Full Text\n\nHinsen K: ImmutablePy 0.1 in ActivePapers format. figshare. 2013. Reference Source\n\nHinsen K: pyMosaic 0.3.1. Zenodo. 2014. Reference Source\n\nHinsen K: ASTRAL-SCOPe subset 2.04 in ActivePapers format. Zenodo. 2014. Reference Source\n\nHinsen K, Shuangwei Hu, Kneller GR, et al.: A comparison of reduced coordinate sets for describing protein structure. figshare. 2013. Reference Source\n\nChevrot G, Hinsen K, Kneller GR: Model-free simulation approach to molecular diffusion tensors: Water. figshare. 2013. Reference Source\n\nChevrot G, Hinsen K, Kneller GR: Model-free simulation approach to molecular diffusion tensors: Lysozyme. figshare. 2013. Reference Source\n\nHinsen K: pyMosaic 0.1.1 in ActivePapers format. figshare. 2013. Reference Source\n\nHinsen K: pyMosaic 0.2.0. Zenodo. 2014. Reference Source\n\nHinsen K: pyMosaic 0.3.0. Zenodo. 2014. Reference Source\n\nKatz DS, Allen G, Chue Hong N, et al.: First workshop on sustainable software for science: Practice and experiences (WSSSPE): submission and peer-review process, and results. arXiv.org, abs/1311.3523, 2013. Reference Source\n\nDrummond C: Replicability is not reproducibility: nor is it good science. 2009. Reference Source" }	[ { "id": "6774", "date": "09 Dec 2014", "name": "Neil Chue Hong", "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\nI, Neil Chue Hong, have reviewed this research article following the principles set out in the Open Science Peer Review Oath v1 (DOI: 10.12688/f1000research.5686.1). This article by Konrad Hinsen discusses the very important issue of how we capture the detail of the dependencies and environment surrounding the software tools that we use, such that we can guarantee that the tools may be used in the future to replicate research and reuse the tools. Overall, it is a comprehensive summary of most of the area.However as it stands, I believe that the article could be significantly improved by deciding whether the paper is to be a fully comprehensive \"state of the art\" summary, looking at the specific advantages and disadvantages of each approach; or one which sets out a shorter summary of the state of the art (as is present in the article in this version), and then goes on to describe the lessons learned from the ActivePapers work, in which case I would suggest a change in title to reflect the emphasis on ActivePapers as a primary example of a platform for publishing and archiving computer aided research. Therefore I am marking this as \"Approved with reservations\" as it requires a structural change, rather than because I believe the work contained in it is not technically sound.In both scenarios for how the article could be rewritten, I believe the article would benefit from the use of more devices to highlight key points and comparisons, for instance by use of tables to compare the effect of different types of platforms on the ability to define environmental dependencies, or linkage with input data.There is one important area of research that should be covered in the discussion of the state of the art - that around \"significant properties\" of software. In particular, the work of Brian Matthews at STFC in this area has previously considered the issue of capturing and prioritising details of both the environment, and how a computational tool is expected to function, thus forming a theoretical basis for describing the way that many current implementations ranging from dependency managers like Maven, through configuration management tools like Docker, Vagrant and Conda, to virtualisation. See:http://www.jisc.ac.uk/media/documents/programmes/preservation/spsoftware_report_redacted.pdfhttp://ijdc.net/index.php/ijdc/article/view/148In terms of other areas where I felt that additional discussion would have provoked more debate, these would be around the trade-offs surround floating point operations, a discussion of other bytecode platforms, and around the long lifetime of successful pieces of software, in particular around trust and how it is mechanically/technically checked.This last point is illustrated in this example from random sampling: https://cryptogenomicon.wordpress.com/2014/10/13/cryptic-genetic-variation-in-software-hunting-a-buffered-41-year-old-bug/As minor points that I believe would improve the papers I would suggest the following:The term \"Web repositories\" for platforms like FigShare and Zenodo is not commonly used - indeed, it is more commonly used to refer to repositories of web pages. I would suggest the more commonly used \"digital repositories\" term, or perhaps \"web accessible third party digital repositories\"? On page 3, it would be useful to describe what makes \"Web repositories\" better. I would suggest it is cost (most are free for openly licensed deposits) and the ability to generate citable DOI s On page 3, when talking about the versioning used by F1000Research, it should be clarified that \"a DOI refers permanently to a specific version of an article\" On page 3, the author's example of the DSSP method still being widely used today could be seen to contradict the earlier arguments surrounding the inability of code to preserved effectively. Whilst I agree that as stated on Page 5 \"This lack of a precisely defined and stable platform for executable code is also the root cause of non-replicability in computer-aided research\" I feel that the author could discuss the tradeoffs (mostly optimisation and performance based) in more detail and put across their opinion of which are most important. On page 6, I had a little difficulty with the statement that \"the scientific aspects of software are always pure computation\". I think that I understand what the author means, but as written it makes me want to identify a counter example. In fairness, I haven't been able to find one yet. On page 7, I note that all the examples of ActivePapers I could find on FigShare have the author as a co-author on them. This means it is unclear whether ActivePapers Python edition is indeed suitable for the wide variety of scientific research areas, as it is unclear whether they represent a representative selection of the use cases for ActivePapers. I believe that the author could do more to support their statement that the fundamental distinction is between \"computational tool\" and \"scientific content\". This could probably be done by making it clearer how current platforms do or don't support this conceptual split, and whether those that do support the split lead to a more accurate ability to replicate research at a later date.As a final note, I would say that it is the tacit convention in all science that published results are assumed correct unless there is clear evidence to suggest otherwise.I believe that with some structural changes to give it a clear narrative emphasis, and better figures to present the information that this research article would provide significant information to the community in this area.", "responses": [ { "c_id": "1234", "date": "24 Feb 2015", "name": "Konrad Hinsen", "role": "Author Response", "response": "My goal was not to write a review article on the topic of computational reproducibility, but to present the conclusions from developing ActivePapers and applying it in real-life research projects. The article has been restructured accordingly. The review of prior work has been shortened and reframed in the context of the ActivePapers project. The \"lessons learned\" from ActivePapers are presented in more detail.I was not aware of the work of Brian Matthews on software preservation in the context of scientific computing. This is indeed very relevant and is mentioned in the revision. I have also expanded the discussion of bytecode platforms and floating-point operations. The discussion of bytecode platform also addresses the issue of trust and security.The minor points have all been addressed in the revised manuscript.Concerning the example of DSSP, it is the method but not the original code that is still widely used. Few users of today's code know that it differs from the originally published method. I do not know if the original code actually implemented the method described in the paper. It is no longer available and I have never seen it.I have expanded my statement that \"the scientific aspects of software are always pure computation\" into a more detailed paragraph, and referred to it in several places in the article. I believe that this point is very important but often overlooked. The structure of an ActivePaper, combining software and the data it works one, makes it clear that only pure computations can be replicable. I then found out that this is common knowledge in other fields, such as programming language theory.I have also expanded the discussion of the factors that have until now prevented the establishment of a stable platform for scientific computing. While performance is often quoted as an important factor, as the reviewer remarks, I do not think this can be backed up by much evidence. To the best of my knowledge, no attempt has been made to design and implement a high-performance yet rigorously defined platform, so it cannot be claimed that this is impossible. The problem is rather that for economical reasons, progress in computing happens as a sequence of small, localized changes: a revision of a language, then a new processor generation, a library update, etc. Each local change must work correctly and efficiently with the existing ecosystem of computing technology. A stable platform definition requires coordinated changes in several technological layers, which is difficult to achieve.It is well possible that the only currently published ActivePapers are those that I cite, and for which I am a co-author. I know about a few other groups experimenting with ActivePapers, but they have not published any results in this form yet. However, I do not make the claim that ActivePapers is in its current form a good solution for all branches of computational science. Like any other tool, it was written with specific techniques and workflows in mind. Like for any other tool, only long-time experience will show how universallly applicable it is. Moreover, the difficulty of integrating legacy software is a real problem for adoption. This article doesn't pretend to do more that present the lessons learned from the applications that are cited.I cannot support my opinion about the distinction between tools and scientific content by any empirical evidence, because to the best of my knowledge no existing framework or tool chain for scientific computing supports the implementation of such an approach.I agree that the tacit assumption of correctness of published results is applied everywhere in science. The particularity of computational science is that this is usually impossible to verify, both for the authors of a scientific study and for their readers." } ] }, { "id": "6988", "date": "26 Jan 2015", "name": "Mercè Crosas", "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 of the Paper This article summarizes arguments in support of reproducibility for scientific research, specifically with respect to computational science. It raises important issues about scholarly communication and reproducibility of previous research work, and it presents ActivePapers as a solution to many problems associated with reproducibility.Reaction Upon reading the article, our reaction is twofold. First, we do not understand the author's primary purpose. Is this a review article or a research article? Is this an article summarizing arguments in support of reproducibility? Is it making the case for ActivePapers? Second, regardless of the primary purpose, we believe the work misses key existing technologies and practices.Structural IssuesIs this a review article or a research article? If this is a review article, replicability and reproducibility should be discussed in more detail, and ActivePapers in less. If this is a research article, we suggest less review of the ActivePapers platform, and more on its contribution and the ways it addresses problems that other systems do not.The introduction states, \"the work presented in this article takes the opposite approach of starting from the requirements of the scientific record and exploring how software and electronic datasets need to be prepared in order to become useful parts of this record.\" What are the requirements of the scientific record? They seem scattered throughout the next few sections, such as \"the most basic requirement for building confidence in computational results is total transparency\" (p. 3). The philosophy of science literature might be helpful in summarizing the basic requirements.Based on our reading, we do not think the main contributions of this paper are the bulleted items in the introduction, but rather the ideal platform requirements on page six. These logically flow from a discussion of the requirements of the scientific record followed by a discussion of tools for reproducible and replicable research.On Data ReuseThere are preservation practices and tools, which follow the approach widely used in Libraries, that help to make a dataset reusable in the longer term:First, researchers are encouraged to use and share datasets in formats commonly used by their discipline, and when possible, formats that do not depend on proprietary software. When the data file format depends on a specific software (or is not considered a preservation format), there exist software tools that convert the proprietary format to a preservation format. For example, for tabular data in SPSS and STATA format which depend on a specific statistical package, the files can be converted to a plain text (tab-delimited or CSV file) plus a metadata file (in XML or JSON format) that contains information about the columns in the original tabular data file. Information is not lost and it can be re-combine to generate an SPSS or STATA file.This preservation feature can be found in public data repositories such as the Dataverse repository software (dataverse.org). There are software projects that focus on other similar automated preservation tools to re-format data files into preservation formats and provide additional preservation metadata. One example is Archivematica. These preservation tools are important for data repositories if they want to make their data accessible and reusable in the future, when the original software might be obsolete.On Methods/Code/SoftwareOne approach on sharing the code used to model/analyze a research work is by using an open-source language like R, where the models and packages are shared and disseminated through a common R package repository (CRAN). Some R packages have explored solutions for reproducibility by tracking the detailed information on what model/code ran and the computing configuration (see http://arxiv.org/pdf/1501.02284.pdf).Also, there are on-going efforts in reproducibility that support hosting code and provide executable functionality to run in the cloud (http://researchcompendia.org/).Minor IssuesThe word conservation should be replaced with preservation. Conservation generally implies the careful maintenance of a finite resource, while preservation implies the protection of a thing in order to keep that thing as is.We recommend reviewing the work and guidelines provided by the Data Citation Principles: https://www.force11.org/datacitation which can apply to software as well, and should help towards giving due importance to software and providing formal long term access and reuse, if the principles are followed.In the conclusion: \"the main reason is that most of the computing technology used by scientists was developed outside of scientific research, for domains of application where replicability is not important.\" Is this statement supported?Indexability This article requires considerable work to be indexed. First, clarify the type of article that is being written, and restructure accordingly. Second, the author misses some key existing technologies and practices that should be discussed.", "responses": [ { "c_id": "1235", "date": "24 Feb 2015", "name": "Konrad Hinsen", "role": "Author Response", "response": "The article has been completely restructured in order to make its primary goal stand out: report the lessons learned from developing and using a platform for publishing and archiving computational science. The revised article starts with an analysis of the needs of the scientific record, deduces technical requirements for a platform designed to meet these needs, and then describes the concrete technical choices made in the two implementations of ActivePapers, followed by a report of the lessons learned from its use. The review of existing technology and the description of the ActivePapers platform are still quite long, but inevitable because I am not aware of any other work I could refer the reader to for this essential background information. Even for fundamental issues such as the requirements of the scientific record, I did not find any publication discussing the specific problems of software and datasets. Software in science is almost always discussed exclusively from the point of view of its utility, ignoring its role as an encoding of scientific knowledge.I fully agree with the reviewers' point of view on data reuse and how to improve it, but I don't see much adoption of these techniques in my scientific environment. Scientists prefer the simplicity of just publishing the files they have on their Web site or on a no-questions-asked digital repository. It's encouraging to see that other domains have succeeded in establishing better habits.I was not aware of the work by Becker et al. on the tools switchr and GRANbase, which are based on very similar ideas as ActivePapers. A short comparison has been added to the revised article. I was aware of Research Compendia, which however does not address software preservation or even execution in its current version. I do mention the two cloud-hosting sites with on-line execution that I know of: Exec&Share (http://www.execandshare.org) and Elsevier's Collage system (http://collage.elsevier.com).The term \"conservation\" has been replaced by \"preservation\" in the revision, and the related issues of software preservation and software deployment are now the common thread around which the presentation is organized. Existing work on the preservation of electronic artifacts has been integrated into the discussion." } ] } ]	1	https://f1000research.com/articles/3-289
https://f1000research.com/articles/4-69/v1	16 Mar 15	{ "type": "Opinion Article", "title": "Tourette Syndrome research highlights 2014", "authors": [ "Cheryl A Richards", "Kevin J Black", "Kevin J Black" ], "abstract": "About 200 journal articles reported research on Tourette syndrome and other tic disorders in 2014. Here we briefly summarize a few of the reports that seemed most important or interesting, ranging from animal models to human studies. Readers can comment on our choices or provide their own favorites using the tools on the online article.", "keywords": [ "review", "histamine", "animal models", "premonitory urge", "MRI", "treatment", "remission", "inheritance" ], "content": "Introduction\n\nThe available information on Tourette syndrome (TS) is steadily increasing (Figure 1), and keeping up with the published literature is therefore an increasing challenge. This article introduces a Highlights article to the F1000Research: Tics portal, to showcase some of the most noteworthy publications from the previous calendar year.\n\nThe number of new publications on Tourette syndrome or other tic disorders each year was estimated from PubMed. The colored line represents locally weighted scatterplot smoothing (LOWESS) of the primary data.\n\nPubMed was searched on 16 Feb 2015 using the search string “(“Tic Disorders”[MeSH] OR Tourette NOT Tourette[AU]) AND 1800[PDAT] : yyyy[PDAT]” for each year yyyy from 1950 through 2014. Publications per year were computed as the difference of each year’s cumulative publications from the previous year’s. (This strategy addresses PubMed’s double-counting electronic and paper publication dates for about 250 publications since 2005). The graph was generated by matplotlib in python (see supplementary material).\n\n\nMethods\n\nThis article is not a systematic review but summarizes the authors’ personal views. We used the following approach to identify pertinent publications: personal reading, asking colleagues for suggestions, F1000Prime, and a PubMed search (Figure 1, legend). Of course this approach will miss some TS-related publications that do not appear in PubMed, or that will be indexed in PubMed in coming months. We chose to focus on articles with final publication dates in 2014.\n\n\nResults\n\nHere we present, in no particular order, examples of TS research published in 2014 that we felt were of special interest.\n\nCastellan Baldan et al.1,2 report on characterization of a possible animal model for Tourette syndrome, chosen because of a family with Tourette syndrome linked to a loss-of-function mutation in the histidine decarboxylase gene3,4. Histidine decarboxylase knock-out mice5 exhibited tic-like stereotypies after a D-amphetamine challenge (see Figure 1, panel E in ref. 2) and increased striatal dopamine during the nocturnal (awake) period. The amphetamine-induced stereotypies were decreased by administration of the dopamine D2 receptor antagonist haloperidol, and histamine infusion reduced striatal dopamine levels and amphetamine-induced stereotypies. Prepulse inhibition and dopamine D2/D3 receptor binding were altered in both the knock-out mice and the small available sample of human carriers of the histidine decarboxylase mutation.\n\nAppropriate animal models of TS could provide pathophysiological insights or speed identification and development of novel treatments. Two informative reviews of mouse models for tic disorders were published in 20146,7. Both reviews agreed that animal models need appropriate face, constructive, and predictive validity to be useful, but Godar et al. argue that a focus on intermediate phenotypes, which involve more elemental neuroanatomic and functional deficits, results in animal models with specific measurable parameters6. They then describe several lines of knockout mutant mice using candidate genes for TS, animal models examining the link between early neuroinflammation and TS pathogenesis, and pharmacological models. Pappas et al. include mouse models for Rett’s syndrome and primary dystonia in addition to TS/OCD7. They developed a test battery to characterize variations across mouse strains in terms of putative tic-like symptoms (i.e., head twitches and body jerks induced by administration of a selective 5-HT2 receptor agonist), amphetamine-induced stereotypies (i.e., wall-rearing and head-down sniffing), perseverative responding on an attentional set-shifting task involving binary choice, and spontaneous locomotion8. Although DOI administration produced head twitches and body jerks in all subjects, the SJL and C57 mice spent a longer time performing body jerks than the ABH and CD1 mouse strains, with the C57 mice also showing increased elevated levels of spontaneous locomotion and increased perseveration on the set-shifting task. The authors argue that using specific behavioral parameters to investigate differences among mouse strains will allow identification of strains that may represent “pure” TS while other strains may represent TS along with behaviors that reflect common co-morbidities (e.g., hyperactivity, compulsions).\n\nAt the November, 2014, annual meeting of the Society for Neuroscience, Xu et al. presented evidence that removing about half of the cholinergic interneurons in the dorsolateral striatum reproduced some features of TS in a rodent model9; see also10. This model was inspired by the fascinating autopsy studies that found decreased numbers of striatal interneurons in Tourette syndrome11. At the same meeting, McCairn and colleagues presented a non-human primate model that included a variety of movements and vocalizations, including tics with reasonable face validity12. As the validity of animal models increases, so does the expectation that they may provide additional insights concerning TS and its associated comorbidities.\n\nNeuner et al.13 examined tic-related neural activity in ten adults with TS using fMRI, estimating the timing of brain activity (as reflected by BOLD signal) to a resolution shorter than that of the individual image acquisitions by taking advantage of the essentially random temporal distribution of tics with respect to the timing of image acquisition. This strategy is reminiscent of the event-related analysis of positron emission tomography regional brain blood flow images developed by Silbersweig and colleagues14. Two seconds before tic occurrence, BOLD activity increased in the supplementary motor area (SMA), ventral primary motor cortex, primary sensorimotor cortex and parietal operculum. One second before tics, activation was seen in the anterior cingulate, putamen, insula, amygdala, cerebellum and extrastriatal visual cortex, and at tic onset activation was seen in the primary motor and somatosensory cortices, the thalamus, and the central operculum. Cortical structure BOLD signal clearly preceded signal in subcortical structures. In addition, resting state data demonstrated that network strength in the same SMA regions correlated with ratings of recent tic severity; the authors suggest that abnormal baseline activity in the SMA may contribute to tic generation. Diffusion tensor imaging identified lower connectivity values (CI), consistent with altered white matter structure, in almost two-thirds of the tracts examined in 15 adults with “pure” TS compared to healthy controls15. After correction for multiple comparisons, 10 tracts were identified as having significantly lower CI in the patient group. Two of these involved connections between the SMA or preSMA and pallidum, and were excluded from further analysis. Significant correlations between YGTSS scores and individual tract CI values were found for the M1–OFC and preSMA–putamen, although none of these remained significant after correcting for multiple comparisons.\n\nIn another fMRI study, adults with “pure” TS performed similarly to control subjects on a stop signal reaction time task, consistent with the conclusion that tics occur without broadly insufficient action inhibition16. However, TS subjects exhibited greater dorsal premotor activation during the Go condition compared to the StopSuccess condition. Increased right pre-SMA activation was associated with successful stop trials in healthy controls. For TS subjects, activation in the SMA proper during StopSuccess compared to Go trials was positively correlated with motor tic frequency. The involvement of SMA in both proactive and reactive control was discussed and the authors suggest that greater SMA activation in patients with higher tic frequency may reflect a stronger need for tic inhibition.\n\nAn overview of SMA syndromes and related research explores how the SMA may be involved in both initiation and inhibition of movements along with providing a tonic interhemispheric balance17. This concept of interhemispheric balance may explain why both SMA activation and SMA inhibition can reduce tics, and may also explain why SMA activation can produce echophenomena in healthy controls.\n\nA derived measure called regional homogeneity (“ReHo”) in the left inferior frontal gyrus increased in 14 subjects with pure TS during tic suppression compared to free ticcing18. ReHo increases were positively correlated with participants’ ability to inhibit their tics both inside and outside the scanner. In another report from the same group, grey matter volumes in the right inferior frontal gyrus and left frontal pole were reduced in adults with “pure” TS compared to healthy controls but these reductions were not correlated with Yale Global Tic Severity Scale (YGTSS) scores or the ability to inhibit tics19.\n\nGABA concentrations in the SMA, measured using magnetic resonance spectroscopy, were significantly higher in 15 adolescents with TS compared to 14 age- and gender-matched controls; there were no group differences for GABA concentrations within M1 or primary visual cortex20. The fMRI BOLD signal change within SMA was negatively correlated with SMA GABA levels supporting the idea that Magnetic Resonance Spectroscopy (MRS)-GABA concentrations are associated with localized increases in tonic inhibition. In a small subset of TS subjects, single-pulse transcranial magnetic stimulation (TMS) delivered to the hand area of the left M1 region preceding movement of the right hand revealed a significant negative correlation between MRS-GABA in the SMA and cortical-spinal excitability within the left M1. Fractional anisotropy values within the corpus callosum for TS subjects were positively correlated with the SMA GABA values and with motor tic severity. These authors suggest that enhanced control over volitional movements and tic suppression may be the result of increased tonic inhibition due to the localized increases in extracellular GABA within SMA.\n\nEight adult subjects who completed Comprehensive Behavioral Intervention for Tics (CBIT) treatment were compared with matched controls on a visual priming task that was used to measure response inhibition21. No significant between-group differences were found in task-related BOLD signal in regions of interest (putamen, caudate, and prefrontal cortex regions BA 11, 44 and 47) before or after CBIT training (with retesting over a comparable amount of time for controls). However, there was a significant group by time interaction because putamen activation decreased in the TS subjects from time 1 to time 2 while it increased in the control subjects. A significant negative correlation between change in IFG activation and change in YGTSS Total Tic Scores was also found. The authors point out that this result is somewhat difficult to interpret given that prior research has indicated that frontal regions are involved in tic suppression.\n\nSeveral groups have recently studied the substantial effects small head motions can have on BOLD fMRI. It is becoming apparent that many of the established techniques to control for movement effects are frequently not sufficient. Functional connectivity analyses have been especially affected, since small head movements during scanning can produce artifactual connectivity signal (i.e., bias not just noise). Fortunately, robust methods exist for preventing such artifact, at the cost of potentially longer acquisition times22. However, movement also interferes with task fMRI analysis. In one task fMRI study of 73 TS subjects ranging in age from 9–15 years of age, only 38 subjects remained after excluding subjects with less than 70% accuracy on a rule-switching task and with at least 3 runs out of 6 with root-mean-square head movement estimates below 1.5 mm23. This is not all explained by tics, since 33 of 53 healthy, tic-free children aged 7–9 were excluded for <60% task accuracy or RMS > 1.5 mm. Even with these relatively stringent requirements, frame-by-frame motion censoring excluded an additional 15–20% of the data. However, this approach bought cleaner signal; motion censoring performed better than all forms of motion regression23.\n\nMinor head motion has been shown to affect structural brain imaging as well. Diffusion MRI is especially sensitive to motion24. A recent study showed that small head movements that do not cause visible artifact in structural brain images can also produce spurious reductions in estimated gray matter volume or cortical thickness25.\n\nMany studies provide minimal information about the specific methods used to control for movement in a patient population that by definition is going to exhibit more movement than the average subject. Therefore inadequate control of subject movement may have contributed to some of the inconsistent results in past neuroimaging studies.\n\nBehavior therapy has been studied as a treatment for tics for many years26. However, its adoption in clinical practice has lagged for a number of reasons27. A meta-analysis that appeared in 2014 may help convince skeptics of its efficacy. The meta-analysis included 8 randomized control trials of TS behavior therapy with a total of 438 TS subjects28. There was no evidence of publication bias. Treatment effects were in the medium to large range, with a number needed to treat (NNT) of only 3, comparable to the most effective class of medications (antipsychotics). Participants who were more likely to respond to behavior therapy were older, had more therapeutic contact and less likely to have comorbid ADHD. At this point, the evidence base for behavior therapy’s efficacy in treating tics is stronger than for any other class of treatments except antipsychotics.\n\nAlthough research continues to demonstrate the value of behavioral treatments such as CBIT, a limited number of therapists have been trained to administer CBIT. Given the distance that many patients live from potential therapists there is a need for alternative forms of treatment. Blount et al. reported on treatment administered using an intensive outpatient procedure (i.e., several hours daily over a four day period) to two pediatric outpatients. This treatment resulted in significant tic reductions that were maintained at follow-up 6 to 7 months later29. This form of treatment may be more convenient for patients and their families who need to travel a significant distance for treatment, but a randomized control trial will be needed to replicate these promising results.\n\nA small group of 18 participants, ranging in age from 10 to 20 years old, performed an Xbox® kickboxing exercise routine with a 5-minute easy exercise session followed by a 2-minute break and then a 5-minute exercise session that was more physically demanding30. Tic counts based on video recordings were lower during both exercise sessions compared to during a baseline interview (i.e., completion of the Physical Activity Questionnaire for Adolescents, discussion about hobbies and other leisure activities). Interestingly, tic frequency was higher during the more demanding exercise session (which also occurred after the subjects had been exercising for a longer amount of time) than during the easier exercise session. Although tic frequency increased significantly during a post-interview completed about 30 minutes after the end of the exercise sessions, the frequency was still below that seen during the pre-exercise baseline. Exercise also resulted in significantly reduced self-reported anxiety which was maintained during the post-exercise interview. It was suggested that exercise might have been effective in reducing tic frequency because it improved executive control functions, or because it served as a distraction task taking attention away from the tics, or because exercise functioned as a competing response which made it more difficult to perform the tics. Behavioral treatments, such as CBIT, tend to involve multiple components and, consequently, it is difficult to determine whether all components are necessary for all patients. Using a simple intervention such as that used by Nixon et al. may make it easier to identify the underlying mechanism that makes the intervention effective and may help identify whether certain patient subgroups are more likely to respond to a particular treatment component.\n\nA preliminary randomized control trial examined electrodermal biofeedback during three 30-minutes sessions each week for 4 weeks in 21 adults with TS31. Both sham and actual biofeedback produced similar decreases in tic frequency and similar improvements in well-being. The authors noted that tics occurring during the biofeedback sessions resulted in competing phasic electrodermal arousal responses making it difficult for patients to sustain a reduction in sympathetic tone, suggesting that modifications in the treatment protocol might increase effectiveness. The sham procedure, which involved providing feedback to subjects so that they thought that they were successfully altering their electrodermal activity, also resulted in a significant decrease in tics. This is surprising since placebo effects are minimal in pharmacological trials32.\n\nPremonitory urges have been considered to have an important role in tic generation, and CBIT includes using a competing response to prevent a tic from occurring until the urge decreases sufficiently so that the tic will not occur. A number of articles in 2014 addressed how premonitory sensations and urges relate to tics and tic suppression.\n\nCapriotti et al.33 examined the effects of negative reinforcement on premonitory urges in 13 children and adolescents with TS or chronic tic disorder (CTD). Subjects rated their urges to tic during three conditions: baseline during which they freely ticced, reinforced tic suppression and reinforced tic suppression with escape. During the escape condition, subjects could initiate a 10 second break during which they could freely tic. When the break was over, the reinforced tic suppression began again. Tic rates were significantly lower during reinforced suppression conditions compared to baseline free-to-tic conditions, although tic rates were significantly higher during the breaks in the escape condition compared to non-break periods. Urge ratings were significantly higher during the reinforced tic suppression conditions compared to the baseline periods and in the escape condition, urge intensity went down from break onset to the end of a break. These results support the hypothesis that premonitory urges are maintained through a process of negative reinforcement.\n\nMany people with tics say that they perform tics to decrease the intensity of premonitory urges because the urges are so bothersome. The relationship between feelings of discomfort and habituation was studied in 90 healthy undergraduate humans with no tic diagnosis34. A 2×2 experimental design was used with subjects either receiving an air puff to the eye or hearing a sound, and either receiving an instruction to blink or no instruction to blink. When subjects received the air puff and instructions to blink, the air puff was less annoying but the EMG response of the orbicularis oculi muscle continued and the length of the EMG response actually increased. When subjects received the air puff without any instructions about blinking, habituation to the air puff occurred. These results indicated that blinking was reinforced by the decrease in annoyance and yet this process also prevented habituation from occurring. A similar process may establish the association between premonitory urges and tic behaviors; if so, this may provide an interesting “animal” model of tics for certain studies.\n\nTreatment-naive children and teenagers with chronic tic disorders were compared while being allowed to tic freely and while receiving reinforcement for suppressing their tics35. Attentional difficulties and age did not affect ability to suppress tics. Interestingly, subjects were able to suppress tics associated with more intense urges just as much as tics associated with less intense urges.\n\nThe Premonitory Urge for Tics Scale (PUTS)36 has been the primary instrument for evaluating premonitory urges in children and adolescents. A 9 item version is frequently used because one item (“I am able to stop my tics, even if only for a short period of time”) did not correlate well with other test items. Interest in determining the reliability and validity in older adolescents and adults produced several studies that were published in 2014. The 10-item PUTS was completed by 102 adults at two specialist clinics. Again item 10 demonstrated relatively low item-total correlation37. The PUTS total score correlated only slightly with scores on the Motor tic, Obsessions and Compulsions, Vocal tic Evaluation Survey (MOVES) (total 0.34, motor 0.28, vocal 0.27), supporting the view that tics and premonitory urges may involve different processes. In general, however, the PUTS was considered to have acceptable reliability and validity when used with adults. Another study examined PUTS scores in 122 older adolescents and adults with TS or CTD38. A third study examined the use of the PUTS in 100 adults with TS39. PUTS scores were related to obsessive-compulsive symptoms, anxiety, attentional problems and quality of life. Half of the total sample had “pure” TS while the other half had comorbid conditions (including 23 with OCD, 15 with ADHD, and 6 with anxiety). For patients with “pure” TS, premonitory urges were negatively related to quality of life scores while a weaker relationship was seen between these two variables for patients with comorbid conditions. When stepwise multiple linear regression analyses were performed, PUTS scores for the “pure” subgroup were only predicted by MOVES obsessive-compulsive subscale scores, while for the subgroup with comorbidities only anxiety scores were predictive of premonitory urges.\n\nAt this time the PUTS is the only empirically validated measure of premonitory urge severity. However, Capriotti et al. pointed out that the PUTS is relatively insensitive to change and is of limited validity in children under the age of 1033. They suggested the number of breaks taken during the tic suppression reinforcement + escape trials as an alternative way of measuring premonitory urge intensity. New approaches to quantifying urge intensity would be welcome.\n\nTwo stress-induction tasks (i.e., public speech, discussion of family conflict) were used to study 8 TS children with comorbid anxiety40. Tic frequency did not increase during periods of increased heart rate, and during the public speech task tic frequencies were actually lower during periods of increased heart rate. The authors point out the only psychophysiological measure of stress used in this study was heart rate and that future studies may benefit from simultaneously assessing a variety of measures of stress (e.g., respiratory rate, ECG, eye tracking) and examining effects on premonitory urge intensity in addition to tic occurrence.\n\nShprecher et al. reported a retrospective follow-up study of tic remission41. A brief survey was used to assess current symptoms of 53 TS patients who were 13–31 years old and had been seen previously in a TS clinic. At the time of the follow-up subjects were seen in person or contacted by telephone. The survey results were consistent with past research about TS and comorbid ADHD and OCD. Mean symptom onset was age 7.9 for both tics and ADHD and 9.2 for OCD. Peak symptom severity was reported to be around age 11–13 for all three conditions with a decline in symptom severity beginning around age 14–15. Symptom remission was reported in 32%, 23%, and 21% of subjects for tics, ADHD, and OCD respectively.\n\nLimited longitudinal follow-up data are available for tic disorders other than TS. Bisker and colleagues reviewed 43 children with no prior diagnosis of Tourette syndrome who had been diagnosed with ocular tics by a pediatric neuro-ophthalmologist42. An average of 6 years after their initial consultation, 32 of the children were located for follow-up. Of these, 44% had persistent ocular tics, 9% had developed nonocular motor tics, and 16% had developed both nonocular motor tics and vocal tics. In other words, the tic disorder remitted in less than a third of the patients available for follow-up.\n\nTo this point, the highly heritable nature of TS has remained a tantalizing clue rather than the key to understanding pathophysiology. However, recently an international collaboration reported an intriguing result. A recent genome-wide association study had identified a number of single nucleotide polymorphisms (SNPs) as possibly associated with TS. The group genotyped 42 of these SNPs in over 1200 individuals, half from unrelated TS cases and half from controls matched for ancestry69. A risk score based on each individual’s alleles at the 42 SNPs was able to predict diagnosis significantly better than chance; this result supports the conclusion that at least some of these SNPs are true risk alleles for TS. One of the SNPs remained significant after correction for multiple comparisons, and the authors discuss nearby genes that could produce relevant changes in brain structure or function.\n\nThe most recent International Scientific Symposium on Tourette Syndrome (New York, 2009) led to a set of review articles on TS updated for publication in 201427,43–50. Also in 2014 the Tourette Syndrome Association announced that it had joined with two European TS groups to sponsor the “First World Congress on Tourette Syndrome and Tic Disorders,” to be held in London in June, 2015 (http://www.tsa-usa.org/tscongress2015.html). Finally, it is difficult to resist pointing out that 2014 also marked the introduction of a publication channel devoted entirely to tics, F1000Research: Tics51. New submissions are warmly invited!\n\n\nDiscussion\n\nWe have provided summaries of some of the articles published in 2014 that we think will contribute to further advances in the field. They cover a broad spectrum: animal models, neuroimaging, and pharmacological and nonpharmacological treatment. The choice of articles was admittedly subjective and most likely incomplete; in fact, we have listed a few more papers in Box 1. However, one of the beauties of this publication venue is that readers who feel we have misjudged are welcome to add their own recommendations to the comments section of this article online.\n\nWe look forward to reprising this “highlights” page at the end of 2015, and would be grateful for article nominations or other suggestions from readers. Box 2 starts off this process by listing some meeting presentations and preprints that caught our interest but had not appeared in final form by the end of 2014. We hope that 2015 brings important breakthroughs in our understanding of the causes, mechanisms and treatment of tic disorders.\n\n“Altered synaptic plasticity in Tourette’s Syndrome and its relationship to motor skill learning”52\n\n“Environmental circumstances influencing tic expression in children”53\n\n“The modulating role of stress in the onset and course of Tourette’s Syndrome: A review”54\n\n“Tic-related obsessive-compulsive disorder (OCD): Phenomenology and treatment outcome in the Pediatric OCD Treatment Study II”55\n\n“Set-shifting deficits: A possible neurocognitive endophenotype for Tourette Syndrome without ADHD”56\n\n“Variables associated with tic exacerbations in children with chronic tic disorders”57\n\nPrenatal and perinatal risk factors for TS58\n\n“Tics are caused by alterations in prefrontal areas, thalamus and putamen, while changes in the cingulate gyrus reflect secondary compensatory mechanisms”59\n\n“Meta-cognitions in Tourette syndrome, tic disorders, and body-focused repetitive disorder”60\n\n“Dysregulated intracellular signaling in the striatum in a pathophysiologically grounded model of Tourette syndrome”61\n\n“Don’t look”: seeing your own tics makes them more frequent62\n\nAstrocyte metabolism in TS70\n\nFunctional connectivity and machine learning in TS63\n\nReward enhances tic suppression very early in the course of tic disorders64\n\nTranscriptome analysis of the human striatum in Tourette syndrome65\n\nInfluence of gender on Tourette syndrome beyond adolescence66\n\nAttention and tic suppression in TS67\n\nMindfulness-based stress reduction68\n\nAblation of striatal cholinergic interneurons10\n\n\nData availability\n\nF1000Research: Dataset 1. Publications on Tourette syndrome: 1950–2014, 10.5256/f1000research.6209.d4418971", "appendix": "Author contributions\n\n\n\nBoth authors contributed to all phases of this work, were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed. Dr. Black is an (unpaid) member of the F1000Research Advisory Board.\n\n\nGrant information\n\nThis work was supported in part by the U.S. National Institutes of Health (NIH), grant R21 NS091635, and by the McDonnell Center for Systems Neuroscience at Washington University.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary materials\n\nIPython notebook. This is an IPython notebook file for anyone who wants to recreate or update Figure 1. Click here to access the data. http://dx.doi.org/10.5256/f1000research.6209.s44187\n\nSame notebook in HTML. The same file for viewing in a web browser, for readers who do not use IPython. Click here to access the data.\n\n\nReferences\n\nCastellan Baldan L, Williams KA, Gallezot JD, et al.: Histidine decarboxylase deficiency causes Tourette syndrome: parallel findings in humans and mice. Neuron. 2014; 81(1): 77–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCastellan Baldan L, Williams K, Gallezot JD, et al.: Erratum. Histidine decarboxylase deficiency causes Tourette syndrome: parallel findings in humans and mice. Neuron. 2014; 82(5): 1186–1187. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nErcan-Sencicek AG, Stillman AA, Ghosh AK, et al.: L-histidine decarboxylase and Tourette’s syndrome. N Engl J Med. 2010; 362(20): 1901–1908. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaragiannidis I, Dehning S, Sandor P, et al.: Support of the histaminergic hypothesis in Tourette syndrome: association of the histamine decarboxylase gene in a large sample of families. J Med Genet. 2013; 50(11): 760–764. PubMed Abstract \| Publisher Full Text\n\nParmentier R, Ohtsu H, Djebbara-Hannas Z, et al.: Anatomical, physiological, and pharmacological characteristics of histidine decarboxylase knock-out mice: evidence for the role of brain histamine in behavioral and sleep-wake control. J Neurosci. 2002; 22(17): 7695–7711. PubMed Abstract\n\nGodar SC, Mosher LJ, Di Giovanni G, et al.: Animal models of tic disorders: a translational perspective. J Neurosci Methods. 2014; 238: 54–69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPappas SS, Leventhal DK, Albin RL, et al.: Mouse models of neurodevelopmental disease of the basal ganglia and associated circuits. Curr Top Dev Biol. 2014; 109: 97–169. PubMed Abstract \| Publisher Full Text\n\nOnori MP, Ceci C, Laviola G, et al.: A behavioural test battery to investigate tic-like symptoms, stereotypies, attentional capabilities, and spontaneous locomotion in different mouse strains. Behav Brain Res. 2014; 267: 95–105. PubMed Abstract \| Publisher Full Text\n\nXu M, Kobets A, Du JC, et al.: Targeted ablation of cholinergic interneurons in the dorsolateral striatum produces behavioral manifestations of Tourette syndrome. Program No. 518.03. 2014 Neuroscience Meeting Planner. [Internet]. Targeted ablation of cholinergic interneurons in the dorsolateral striatum produces behavioral manifestations of Tourette syndrome. 2014 [cited 2015 Jan 16]. Reference Source\n\nXu M, Kobets A, Du JC, et al.: Targeted ablation of cholinergic interneurons in the dorsolateral striatum produces behavioral manifestations of Tourette syndrome. Proc Natl Acad Sci U S A. 2015; 112(3): 893–898. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKataoka Y, Kalanithi PS, Grantz H, et al.: Decreased number of parvalbumin and cholinergic interneurons in the striatum of individuals with Tourette syndrome. J Comp Neurol. 2010; 518(3): 277–291. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCairn K, Nagai Y, Hori Y, et al.: Program No. 517.06. 2014 neuroscience meeting planner. [Internet]. Distinct cortical and subcortical networks drive myoclonic and vocal tics in the nonhuman primate model of Tourette syndrome: A PET and electrophysiological study. 2014 [cited 2015 Jan 15]. Reference Source\n\nNeuner I, Werner CJ, Arrubla J, et al.: Imaging the where and when of tic generation and resting state networks in adult Tourette patients. Front Hum Neurosci. 2014; 8: 362. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSilbersweig DA, Stern E, Schnorr L, et al.: Imaging transient, randomly occurring neuropsychological events in single subjects with positron emission tomography: an event-related count rate correlational analysis. J Cereb Blood Flow Metab. 1994; 14(5): 771–782. PubMed Abstract \| Publisher Full Text\n\nCheng B, Braass H, Ganos C, et al.: Altered intrahemispheric structural connectivity in gilles de la Tourette syndrome. Neuroimage Clin. 2014; 4: 174–181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGanos C, Kühn S, Kahl U, et al.: Action inhibition in Tourette syndrome. Mov Disord. 2014; 29(12): 1532–1538. PubMed Abstract \| Publisher Full Text\n\nPotgieser AR, de Jong BM, Wagemakers M, et al.: Insights from the supplementary motor area syndrome in balancing movement initiation and inhibition. Front Hum Neurosci. 2014; 8: 960. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGanos C, Kahl U, Brandt V, et al.: The neural correlates of tic inhibition in gilles de la Tourette syndrome. Neuropsychologia. 2014; 65: 297–301. PubMed Abstract \| Publisher Full Text\n\nGanos C, Kühn S, Kahl U, et al.: Prefrontal cortex volume reductions and tic inhibition are unrelated in uncomplicated GTS adults. J Psychosom Res. 2014; 76(1): 84–87. PubMed Abstract \| Publisher Full Text\n\nDraper A, Stephenson MC, Jackson GM, et al.: Increased GABA contributes to enhanced control over motor excitability in Tourette syndrome. Curr Biol. 2014; 24(19): 2343–2347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeckersbach T, Chou T, Britton JC, et al.: Neural correlates of behavior therapy for Tourette’s disorder. Psychiatry Res. 2014; 224(3): 269–274. PubMed Abstract \| Publisher Full Text\n\nPower JD, Mitra A, Laumann TO, et al.: Methods to detect, characterize, and remove motion artifact in resting state fMRI. Neuroimage. 2014; 84: 320–341. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSiegel JS, Power JD, Dubis JW, et al.: Statistical improvements in functional magnetic resonance imaging analyses produced by censoring high-motion data points. Hum Brain Mapp. 2014; 35(5): 1981–1996. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLe Bihan D, Poupon C, Amadon A, et al.: Artifacts and pitfalls in diffusion MRI. J Magn Reson Imaging. 2006; 24(3): 478–488. PubMed Abstract \| Publisher Full Text\n\nReuter M, Tisdall MD, Qureshi A, et al.: Head motion during MRI acquisition reduces gray matter volume and thickness estimates. Neuroimage. 2015; 107: 107–115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlack KJ: Behavior therapy for Tourette syndrome.2014. Reference Source\n\nCapriotti MR, Himle MB, Woods DW: Behavioral treatments for Tourette syndrome. J Obsessive Compuls Relat Disord. 2014; 3(4): 415–420. Publisher Full Text\n\nMcGuire JF, Piacentini J, Brennan EA, et al.: A meta-analysis of behavior therapy for Tourette syndrome. J Psychiatr Res. 2014; 50: 106–112. PubMed Abstract \| Publisher Full Text\n\nBlount TH, Lockhart ALT, Garcia RV, et al.: Intensive outpatient comprehensive behavioral intervention for tics: A case series. World J Clin Cases. 2014; 2(10): 569–577. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNixon E, Glazebrook C, Hollis C, et al.: Reduced Tic Symptomatology in Tourette Syndrome After an Acute Bout of Exercise: An Observational Study. Behav Modif. 2014; 38(2): 235–263. PubMed Abstract \| Publisher Full Text\n\nNagai Y, Cavanna AE, Critchley HD, et al.: Biofeedback treatment for Tourette syndrome: a preliminary randomized controlled trial. Cogn Behav Neurol. 2014; 27(1): 17–24. PubMed Abstract \| Publisher Full Text\n\nCubo E, González M, Singer H, et al.: Impact of placebo assignment in clinical trials of tic disorders. Mov Disord. 2013; 28(9): 1288–1292. PubMed Abstract \| Publisher Full Text\n\nCapriotti MR, Brandt BC, Turkel JE, et al.: Negative Reinforcement and Premonitory Urges in Youth With Tourette Syndrome: An Experimental Evaluation. Behav Modif. 2014; 38(2): 276–296. PubMed Abstract \| Publisher Full Text\n\nBeetsma DJV, van den Hout MA, Engelhard IM, et al.: Does repeated ticking maintain tic behavior? An experimental study of eye blinking in healthy individuals. Behav Neurol. 2014; 2014: 753020. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpecht MW, Nicotra CM, Kelly LM, et al.: A Comparison of Urge Intensity and the Probability of Tic Completion During Tic Freely and Tic Suppression Conditions. Behav Modif. 2014; 38(2): 297–318. PubMed Abstract \| Publisher Full Text\n\nWoods DW, Piacentini J, Himle MB, et al.: Premonitory Urge for Tics Scale (PUTS): initial psychometric results and examination of the premonitory urge phenomenon in youths with Tic disorders. J Dev Behav Pediatr. 2005; 26(6): 397–403. PubMed Abstract \| Publisher Full Text\n\nCrossley E, Seri S, Stern JS, et al.: Premonitory urges for tics in adult patients with Tourette syndrome. Brain Dev. 2014; 36(1): 45–50. PubMed Abstract \| Publisher Full Text\n\nReese HE, Scahill L, Peterson AL, et al.: The premonitory urge to tic: measurement, characteristics, and correlates in older adolescents and adults. Behav Ther. 2014; 45(2): 177–186. PubMed Abstract \| Publisher Full Text\n\nEddy CM, Cavanna AE: Premonitory Urges in Adults With Complicated and Uncomplicated Tourette Syndrome. Behav Modif. 2013; 38(2): 264–275. PubMed Abstract \| Publisher Full Text\n\nConelea CA, Ramanujam K, Walther MR, et al.: Is There a Relationship Between Tic Frequency and Physiological Arousal? Examination in a Sample of Children With Co-Occurring Tic and Anxiety Disorders. Behav Modif. 2014; 38(2): 217–234. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShprecher DR, Gannon K, Agarwal N, et al.: Elucidating the nature and mechanism of tic improvement in Tourette syndrome: a pilot study. Tremor Other Hyperkinet Mov (N Y). 2014; 4: 217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBisker ER, McClelland CM, Brown LW, et al.: The long-term outcomes of ocular tics in a pediatric neuro-ophthalmology practice. J AAPOS. 2014; 18(1): 31–35. PubMed Abstract \| Publisher Full Text\n\nBlack KJ, Jankovic J, Hershey T, et al.: Progress in research on Tourette syndrome. J Obsessive Compuls Relat Disord. 2014; 3(4): 359–362. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChurch JA, Schlaggar BL: Pediatric Tourette syndrome: insights from recent neuroimaging studies. J Obsessive Compuls Relat Disord. 2014; 3(4): 386–393. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEddy CM, Cavanna AE: Tourette syndrome and obsessive compulsive disorder: Compulsivity along the continuum. J Obsessive Compuls Relat Disord. 2014; 3(4): 363–371. Publisher Full Text\n\nGilbert DL, Jankovic J: Pharmacological treatment of Tourette syndrome. J Obsessive Compuls Relat Disord. 2014; 3(4): 407–414. Publisher Full Text\n\nLeckman JF, King RA, Bloch MH: Clinical features of Tourette syndrome and tic disorders. J Obsessive Compuls Relat Disord. 2014; 3(4): 372–379. Publisher Full Text\n\nPauls DL, Fernandez TV, Mathews CA, et al.: The Inheritance of Tourette Disorder: A review. J Obsessive Compuls Relat Disord. 2014; 3(4): 380–385. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScahill L, Specht M, Page C: The Prevalence of Tic Disorders and Clinical Characteristics in Children. J Obsessive Compuls Relat Disord. 2014; 3(4): 394–400. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVisser-Vandewalle V, Huys D, Neuner I, et al.: Deep Brain Stimulation for Tourette syndrome: The Current State of the Field. J Obsessive Compuls Relat Disord. 2014; 3(4): 401–406. Publisher Full Text\n\nBlack KJ: F1000Research: Tics welcomes you to 21st century biomedical publishing. [v1; ref status: not peer reviewed, http://f1000r.es/4o1]. F1000Res. 2014; 3: 272. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrandt VC, Niessen E, Ganos C, et al.: Altered synaptic plasticity in Tourette’s syndrome and its relationship to motor skill learning. PLoS One. 2014; 9(5): e98417. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaurín B, Serrano M, Fernández-Alvarez E, et al.: Environmental circumstances influencing tic expression in children. Eur J Paediatr Neurol. 2014; 18(2): 157–162. PubMed Abstract \| Publisher Full Text\n\nBuse J, Kirschbaum C, Leckman JF, et al.: The Modulating Role of Stress in the Onset and Course of Tourette’s Syndrome: A Review. Behav Modif. 2014; 38(2): 184–216. PubMed Abstract \| Publisher Full Text\n\nConelea CA, Walther MR, Freeman JB, et al.: Tic-related obsessive-compulsive disorder (OCD): phenomenology and treatment outcome in the Pediatric OCD Treatment Study II. J Am Acad Child Adolesc Psychiatry. 2014; 53(12): 1308–1316. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEddy CM, Cavanna AE: Set-Shifting Deficits: A Possible Neurocognitive Endophenotype for Tourette Syndrome Without ADHD. J Atten Disord. 2014. PubMed Abstract \| Publisher Full Text\n\nHimle MB, Capriotti MR, Hayes LP, et al.: Variables Associated With Tic Exacerbation in Children With Chronic Tic Disorders. Behav Modif. 2014; 38(2): 163–183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMathews CA, Scharf JM, Miller LL, et al.: Association between pre- and perinatal exposures and Tourette syndrome or chronic tic disorder in the ALSPAC cohort. Br J Psychiatry. 2014; 204(1): 40–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMüller-Vahl KR, Grosskreutz J, Prell T, et al.: Tics are caused by alterations in prefrontal areas, thalamus and putamen, while changes in the cingulate gyrus reflect secondary compensatory mechanisms. BMC Neurosci. 2014; 15: 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nO’Connor K, St-Pierre-Delorme M-È, Leclerc J, et al.: Meta-cognitions in Tourette syndrome, tic disorders, and body-focused repetitive disorder. Can J Psychiatry. 2014; 59(8): 417–425. PubMed Abstract \| Free Full Text\n\nRapanelli M, Frick LR, Pogorelov V, et al.: Dysregulated intracellular signaling in the striatum in a pathophysiologically grounded model of Tourette syndrome. Eur Neuropsychopharmacol. 2014; 24(12): 1896–1906. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrandt VC, Lynn MT, Obst M, et al.: Visual feedback of own tics increases tic frequency in patients with Tourette’s syndrome. Cogn Neurosci. 2015; 6(1): 1–7. PubMed Abstract \| Publisher Full Text\n\nGreene DJ, Church JA, Adeyemo B, et al.: Support vector machine classification of pediatric Tourette syndrome using resting state functional connectivity. Program No. 517.03. 2014 Neuroscience Meeting Planner [Internet]. Support vector machine classification of pediatric Tourette syndrome using resting state functional connectivity. 2014; [cited 2015 Jan 14]. Reference Source\n\nGreene DJ, Koller JM, Robichaux-Viehoever A, et al.: Reward enhances tic suppression in children within months of tic disorder onset. Dev Cogn Neurosci. 2015; 11: 65–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLennington JB, Coppola G, Kataoka-Sasaki Y, et al.: Transcriptome Analysis of the Human Striatum in Tourette Syndrome. Biol Psychiatry. 2014. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLichter DG, Finnegan SG: Influence of gender on Tourette syndrome beyond adolescence. Eur Psychiatry. 2015; 30(2): 334–40. PubMed Abstract \| Publisher Full Text\n\nMisirlisoy E, Brandt V, Ganos C, et al.: The Relation Between Attention and Tic Generation in Tourette Syndrome. Neuropsychology. 2014. PubMed Abstract \| Publisher Full Text\n\nReese HE, Vallejo Z, Rasmussen J, et al.: Mindfulness-based stress reduction for Tourette syndrome and chronic tic disorder: A pilot study. J Psychosom Res. 2015; 78(3): 293–8. PubMed Abstract \| Publisher Full Text\n\nPaschou P, Yu D, Gerber G, et al.: Genetic association signal near NTN4 in Tourette syndrome. Ann Neurol. 2014; 76(2): 310–315. PubMed Abstract \| Publisher Full Text\n\nde Leeuw C, Goudriaan A, Smit AB, et al.: Involvement of astrocyte metabolic coupling in Tourette syndrome pathogenesis. Eur J Hum Genet. 2015; [in press]. PubMed Abstract \| Publisher Full Text\n\nRichards CA, Black KJ: Dataset 1, in: Tourette Syndrome research highlights 2014. F1000Research. 2015. Data Source" }	[ { "id": "8212", "date": "21 Apr 2015", "name": "Davide Martino", "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\nI enjoyed reading this useful and well written contribution.I have a few comments:In the subsection of Results entitled “Other animal models”, the reader might benefit from a brief explanation of the model presented by McCairn et al: did this use a GABA antagonist like bicuculline? In the subsection entitled “A look inside: neuroimaging studies”, I personally found interesting a paper which is not cited and evaluates the potential of deep repetitive transcranial magnetic stimulation (deep rTMS) to the supplementary motor area (Bloch et al.,2014) In respect to neurobiofeedback, as commented on in page 4, another useful paper on ADHD and tic disorders is Gevensleben et al. (2009). Moreover, a 2015 citation focusing on neurofeedback is this useful review: Farkas et al. (2015). My main criticism is that there is no citation at all about deep brain stimulation, which is an area currently avidly investigated in Tourette syndrome. It is possible that the authors felt that none of the articles published in 2014 on DBS in Tourette syndrome is sufficiently interesting, but I would at least discuss this in the article and add the following citation in Box 2: Schrock et al., (2015). Additional 2014 publications of interest:Chao TK, Hu J and Pringsheim T: Prenatal risk factors for Tourette syndrome: a systematic review. BMC Pregnancy and Childbirth. 2014; 14(53).Gilbert DL, Budman CL, Singer HS, et al:, A D1 receptor antagonist, ecopipam, for treatment in Tourette syndrome. Clin Neuropharmacol. 2014; 37(1): 26-30.Wijemanne S, Wu LJ and Jankovic J: Long-term efficacy and safety of fluphenazine in patients with Tourette syndrome. Mov Disord. 2014; 29(1):126-130.", "responses": [ { "c_id": "1433", "date": "14 Jul 2015", "name": "Kevin J Black", "role": "Author Response F1000Research Advisory Board Member", "response": "We appreciate Dr. Martino's thoughtful comments and have used them to improve the manuscript for version 2.Brief explanation of the model presented by McCairn et al.Yes, this used bicuculline injected into the putamen or nucleus accumbens; we have added this to the manuscript. Recommends Bloch et al.This is a very interesting article, but the final version was not yet published as of May, 2015. We have added it to Box 2. \"In respect to neurobiofeedback, as commented on in page 4, another useful paper on ADHD and tic disorders is Gevensleben et al. (2009). Moreover, a 2015 citation focusing on neurofeedback is this useful review: Farkas et al. (2015).\"Good articles. Readers can find the citation to Gevensleben et al. 2009 in your comments. We have added the Farkas et al. (2015) reference to Box 2. No citation about DBS.We agree this is an important area of work in TS, but relatively few DBS papers on TS were published in 2014. We have added a paragraph on DBS, and have added the important 2015 paper you suggested to Box 2. Recommends Chao et al., prenatal risk factors for Tourette syndromeWe agree and have added this to Box 1.Recommends Gilbert et al., ecopipam studyThis paper was overlooked by accident. We agree this was an important pilot study on a potentially new avenue of treatment for TS and have added a short paragraph.Recommends Wijemanne et al., fluphenazine report.Thank you. We have added this to the section on medication treatment." } ] }, { "id": "7989", "date": "29 Apr 2015", "name": "Barbara Coffey", "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 authors report on highlights of the growing medical and scientific literature on Tourette Syndrome in 2014. The aim was not to conduct a systematic review, but instead to discuss noteworthy articles of interest. Topics ranged from animal models and neuroimaging to behavioral treatment. Strengths of this manuscript were 1) cogent and thoughtful summary/discussion of the articles, 2) an informal and readable style, and 3) interesting content. There were several weaknesses, which if addressed, would strengthen the manuscript. Methods: it would be helpful to know the approximate denominator of articles reviewed, and what proportion were selected. Results: it would be helpful to know how/why topic areas were chosen. Weight seemed to lean strongly toward neuroimaging, which may reflect the authors’ primary interests and expertise. Organization of topic areas could be improved; as it stands, the reader is moved from neuroscience (animal models, imaging) to behavior therapy, premonitory urges, and longitudinal course, and then back to neuroscience (genetics). It might read more easily to start with basics (neuroscience, genetics, neuroimaging) through phenomenology (course, urges, role of stress) to treatment. The manuscript could be improved with the inclusion of pharmacotherapy updates; although the Discussion summarized that the “broad spectrum of articles” covered “animal models, neuroimaging, and pharmacological and non-pharmacological treatment,” the only treatments discussed were behavior therapy, exercise and biofeedback. Suggested pharmacotherapy additions include:Gilbert, D. et al.: “A D1 receptor antagonist, ecopipam, for treatment of tics in Tourette Syndrome” Clin Neuropharmacol 2014; 37 (1) 26-30Malaty, I.A. and Akbar U: Updates in Medical and Surgical Therapies for Tourette Syndrome; 2014; Curr Neurol Neurosci Rep 14: 458; this is a review with an annotated bibliography. Work to look for in 2015: Bachmann, CJ et al.: Trends in Psychopharmacological Treatment of Tic Disorders in Children and Adolescents in Germany; Eur Child Adolesc Psychiatry 2015; 24 (2); 199-207Lastly, there appeared to be a relative lack of articles on aspects of psychiatric comorbidity. Of particular interest is the recently published study from the Tourette Syndrome Association International Consortium for Genetics: Hirschtritt, M. et al.: Lifetime Prevalence, Age of Risk, and Genetic Relationships of Comorbid Psychiatric Disorders in Tourette Syndrome; JAMA Psychiatry doi:10.1001/jamapsychiatry.2014.2650", "responses": [ { "c_id": "1434", "date": "14 Jul 2015", "name": "Kevin J Black", "role": "Author Response F1000Research Advisory Board Member", "response": "We thank Dr. Coffey for her very thoughtful comments, and have revised the manuscript accordingly. Replies to specific comments follow.Requests the “approximate denominator of articles reviewed.”We read at least the abstract from the 200 papers identified in the PubMed searches. We have added this number to the introduction. We cite well over 50 of them. Originally we intended to discuss only about 5 or 10 papers, but it was harder than expected to omit anything. “It would be helpful to know how/why topic areas were chosen … may reflect the authors’ primary interests”Doubtless the selection of articles reflects some influence of our own interests. As you noted, we at least warned the reader that our selection was arbitrary. We attempted to respond to the theme of what reports were most likely to be of greatest interest in the future, and we now explain this choice in the first sentence of Results. “Organization of topic areas could be improved”Thank you for the useful suggestion. We have rearranged the topics in a more structured fashion: Etiology, pathophysiology, phenomenology and natural history, and treatment. We left out pharmacologyWe agree. Please see our comments in the response to Dr. Martino's review about Gilbert et al. We have added the Malaty and Akbar citation to the treatment section, and the Bachmann et al. review to Box 2. “Relative lack of articles on … psychiatric comorbidity.” Recommends Hirschtritt et al. 2015.We are also very interested in psychiatric comorbidity, and we agree that the Hirschtritt et al article is very important. As it was published in April, 2015, we added it to Box 2." } ] } ]	1	https://f1000research.com/articles/4-69
https://f1000research.com/articles/4-235/v1	13 Jul 15	{ "type": "Review", "title": "A ChIP on the shoulder? Chromatin immunoprecipitation and validation strategies for ChIP antibodies", "authors": [ "Fiona C. Wardle", "Haihan Tan", "Haihan Tan" ], "abstract": "Chromatin immunoprecipitation (ChIP) is a technique widely used in the study of epigenetics and transcriptional regulation of gene expression. However, its antibody-centric nature exposes it to similar challenges faced by other antibody-based procedures, of which the most prominent are issues of specificity and affinity in antigen recognition. As with other techniques that make use of antibodies, recent studies have shown the need for validation of ChIP antibodies in order to be sure they recognize the advertised protein or epitope. We summarize here the issues surrounding ChIP antibody usage, and highlight the toolkit of validation methods that can be employed by investigators looking to appraise these reagents.", "keywords": [ "Chromatin immunoprecipitation", "antibody", "validation", "transcription factor", "modified histone" ], "content": "Introduction\n\nChromatin immunoprecipitation (ChIP) is a technique that has revolutionized our ability to identify regulatory sequences and epigenetic marks in the genome, and in doing so decipher networks of gene expression regulation that drive cell identity during development, disease, regeneration and evolution. It is used to determine if a protein of interest binds to, or is localized at, a specific DNA sequence. For example, it can be used to show where transcription factors and modified histones bind to a particular region of DNA in particular cells, allowing the identification of functional genomic sequences (e.g. 1).\n\nOriginally developed in bacterial cells2 but soon applied to other cell and tissue types, particularly Drosophila cells and embryos (e.g. 3–6), ChIP methods initially identified DNA bound by a protein of interest via a candidate approach for genomic regions suspected to interact with the protein, or by cloning and sequencing of immunoprecipitated DNA3–6. Over more than a decade, these methods allowed many regulatory regions to be identified, but it has been with the invention of genome-scale methods, such as microarrays (ChIP-chip) and more recently next generation sequencing (ChIP-seq;7–13), that the use of this technique has greatly expanded and allowed whole genome regulatory landscapes to be uncovered.\n\nChIP is now routinely used to map the genomic distribution of transcription factors, chromatin remodelling factors, and histone modifications in numerous cell types and organisms, and it is a technique central to the efforts of large-scale genomics consortia to map the regulatory genome. Indeed, the ENCODE and modENCODE communities used ChIP-seq in more than 100 cell types in mouse, human, Drosophila and C. elegans to map binding of over 140 DNA-interacting factors1,14–17.\n\nHowever, this technique depends on the use of antibodies to recognize the target protein of interest, and as with all techniques that rely on antibodies, issues of specificity and affinity arise18. There is a growing realization among the scientific community that not all antibodies work as advertised, with problems of antibody specificity and variability causing projects to stall and published results to be irreproducible (discussed in 18,19). Due to these problems there have been recent calls to standardize antibody manufacture, validation and reporting in publications (e.g. 19–23). In this review, we will highlight some of the concerns and challenges that arise in selecting and validating antibodies for ChIP; we also discuss the need for validation standards, and highlight the validation guidelines used by ENCODE, modENCODE and other animal genome annotation consortia as a minimum standard for ChIP assays24.\n\n\nChIP method overview\n\nChIP usually involves lightly fixing cells of interest, usually with formaldehyde, to cross-link proteins and DNA. An alternative is native ChIP where a cross-linking reagent is not used. Chromatin is isolated from these cells and fragmented into pieces, usually in the range of 200–500 base pairs. This fragmentation may be enzymatic, e.g. with micrococcal nuclease, or mechanical, e.g. by sonication. An antibody recognizing the protein of interest, coupled to beads or other solid support, is then used to purify the protein, with its attached DNA, away from the rest of the sample. The cross-links, if present, are reversed and proteins in the sample are then degraded, leaving purified DNA that was associated with the protein of interest. Typically nowadays, this DNA is analysed either by high throughput sequencing to identify the genomic regions associated with the protein of interest, or by PCR with specific primers if binding sites are already known.\n\nThis procedure is a fairly involved process and the outcome is critically dependent on the quality of the antibody used – both its specificity and affinity for the protein of interest. Therefore like other procedures involving antibodies, it is crucial to validate antibodies used in ChIP to be confident of the results obtained. Many of the challenges faced in validating antibodies for other procedures are similar for ChIP, although there are issues unique to this protocol (Box 1).\n\n\n\n- Commercially available ChIP-grade antibodies may not be validated between lots. They may also non-specifically cross-react with similar proteins, or not be validated in the organism or cell type of interest.\n\n- Validation of non-ChIP-grade antibodies may be challenging, costly, and time-consuming.\n\n- The choice of monoclonal antibodies recognizing only one epitope, or of polyclonal antibodies raised only to a part of the target protein, may reduce pull down of the target protein.\n\n- The choice of polyclonal antibodies results in production of limited quantities of serum/antibody.\n\n- Tagging of target proteins for tag-based pull downs may interfere with endogenous protein function.\n\n- The straight overexpression of tagged proteins may result in spurious DNA binding.\n\n- Antibodies validated as specific for the target protein may not bind to the target with high affinity in ChIP.\n\n\nSelecting an antibody\n\nThe choice of antibody for ChIP will depend on the target protein of interest and the antibodies that are already available. It may be that the protein of interest is well-studied, and that well-characterized antibodies are commercially available, which have been used and previously validated in ChIP in the cells of interest. In this case, little or no additional validation may be required. Indeed with the surge in the use of ChIP methodologies, many companies now sell ChIP-validated antibodies. This situation is most common for antibodies against histone modifications and dozens of companies offer such ChIP-validated antibodies (e.g. 25).\n\nHowever, even with these ‘validated’ antibodies, it is critical to confirm what validation assays have been performed, and whether cross-reactivity has been reported. As an example, Egelhofer and colleagues tested over 200 commercially available antibodies raised to different histone modifications and found that more than 25% were not strictly specific to the modification advertised26. This may in part be due to different lots of the same antibody differing considerably in their specificity26, so it is crucial to know which lot(s) any validations have been performed on, and whether the current lot in hand has been tested.\n\nUnfortunately, not all companies repeat validation tests on new lots27, and so the user may need to validate the antibody themselves (more of which below). This issue also highlights the need for authors to include the lot number, company, and catalog number in materials and methods sections, when publishing papers in which antibodies are used. In addition, if the antibody has not been validated in the cells of interest, it should be validated in those specific cells before experimental use, since it may not behave in the same way as in tested cells24.\n\nIt may, on the other hand, be the case that the ChIP will be against a protein for which no validated antibody has been produced. This is especially pertinent in the case of non-mammalian tissues, such as zebrafish or Xenopus, or if investigating a protein that has not received much attention in the past. In these situations, commercially available antibodies raised against homologous proteins from different species, or custom-manufactured antibodies, will have to be tested. Unfortunately, there are no hard and fast rules defining an archetypal antibody best for ChIP; however, we set out below some of the advantages and disadvantages of different types of antibodies.\n\n\nMonoclonal vs polyclonal\n\nChIP can be performed using either monoclonal or polyclonal antibodies, each of which have their own advantages and disadvantages. Monoclonal antibodies bear the obvious advantage of being a continuous, generally consistent, and potentially unlimited resource. The major disadvantage of a monoclonal antibody is its recognition of only one epitope, thus if the protein of interest can form a complex with other proteins (a highly probable scenario), it is possible that the epitope will be masked, decreasing the chances of pulling down all instances of the protein bound to DNA.\n\nPolyclonal antibodies raised against the whole target protein get around this problem since the antibodies will recognize multiple epitopes, increasing the likelihood that there will be free epitopes, even in a protein complex, to interact with the antibody. However, polyclonal antibodies that are raised only to a subdomain or a peptide of the protein will recognize a reduced number of epitopes, bringing this type of antibody closer to the monoclonal condition. Another drawback of polyclonal antibodies is that only a limited quantity of serum, and therefore antibody, can be produced per batch (the amount is dependent on the species in which the antibody is raised). Since each new batch of serum may differ in its characteristics, each fresh batch of antibody must be tested as a new entity.\n\n\nEndogenous protein vs tagged protein\n\nThe majority of ChIP experiments currently rely on antibodies that recognize an endogenous protein expressed in cells of interest. However, a workable, validated antibody to a protein of interest cannot always be identified, despite extensive testing. To circumvent this problem, tagged versions of target proteins, expressed in relevant cells of interest, are sometimes used. This brings with it the advantage that well-characterized antibodies to epitope tags (e.g. V5, HA and His) are commercially available and therefore provide a consistent source of reagents.\n\nNonetheless we note that there are also disadvantages to this approach. One possible issue is that the tag may interfere with endogenous protein function, and potentially its interaction with DNA. It is hence advisable to compare the function of tagged and untagged proteins in suitable assays to confirm that they are functionally equivalent, before moving onto ChIP experiments. Additionally, in order to rule out non-specific binding of the anti-epitope tag antibody in the sample, a control ChIP reaction from identical cells, bar the expression of the tagged protein, should also be included (discussed in more detail below;24).\n\nIn many situations, it is preferable to express the tagged protein at levels comparable to the endogenous protein, for instance by harnessing endogenous regulatory sequences to drive its expression (e.g. 28). This is because the overexpression of some, though not all, transcription factors may result in spurious binding29,30. Another, perhaps better, approach is to express the protein in cells mutant for the endogenous target, at levels that rescue the mutant phenotype (e.g. 31,32). However, we note that in certain situations, over- or ectopic expression of DNA-binding proteins may produce the phenotype being studied, such as in lineage reprogramming or induced differentiation, in which case the non-endogenous binding of the factor gives useful information on the function and regulatory circuits controlled by that protein in such circumstances (e.g. 30,33).\n\nRecent technological advances in genome editing, such as TALEN and CRISPR/Cas9 technologies, now allow tags to be knocked into a specific locus to produce a fusion protein in many more different cell types and organisms than had previously been possible (e.g. 34,35). Creating a tagged fusion protein obviates the issue of the protein being expressed exogenously at higher-than-endogenous levels, although not the disadvantage that the tag may interfere with protein function. Nevertheless, given that this approach potentially allows any protein to be tagged and ChIPed in any organism or cell type, its use seems set to increase massively in the future.\n\n\nStandard assays for ChIP antibody validation\n\nIn order for investigators to have confidence in an antibody’s specificity for the protein of interest, it is critical to have working standards and reporting guidelines23, and this is as true for ChIP as for any antibody-based technique. It is prudent to take as much care as possible to ensure ChIP reagent fidelity in order to maximize the accuracy of research output. This has been particularly recognized by the ENCODE and modENCODE consortia, which have published guidelines for the standards required of ChIP-seq experiments for inclusion in their data pipeline24. Other genome annotation consortia such as FAANG (36; http://www.faang.org/) and IHEC (37; http://ihec-epigenomes.net/) have also adopted such standards in order that data can be compatible and comparable, and we suggest that these validation assays are a useful toolkit for all researchers performing ChIP experiments. The guidelines typically suggest a two-step validation procedure: initially the antibody is tested in an immunoblot or immunofluorescence assay, followed by at least one secondary validation assay; these are described in brief below (see also Box 2). However, it should be noted that others have suggested these guidelines are not sufficiently stringent, and additional controls/validations may be required in certain experiments38.\n\n\n\n- Recombinant target proteins, or cell/nuclear lysates of relevant tissues may be immunoblotted with the antibody, and a strong immunoreactive band should be observed around the expected molecular weight of the target protein.\n\n- A modification to the immunoblot assay is the ChIP-immunoblot. ChIP eluates containing antibody-target protein complexes are blotted and probed with antibody.\n\n- The antibody may be tested in an immunofluorescent assay, where staining should be observed in the nuclei of target protein-expressing cells.\n\n- Further immunoblot or immunofluorescence-based validation of antibody specificity may be carried out on cells/tissues in which the target protein has been knocked out or knocked down. The immunoreactive signal should be absent or greatly reduced.\n\n- Proteins may also be translated in vitro or expressed in cells, and these samples may be tested in immunoblots or immunofluorescence, as an alternative method of evaluating reactivity.\n\n- Immunoprecipitation with the antibody, followed by mass spectrometry-based sequencing, should identify a majority of the target protein in the pull down fraction.\n\n- A second antibody to the target protein, or a tag-based pull down of tagged target proteins, may be used as an independent test. There should be good overlap of results between the different pull downs.\n\n- A search for DNA motifs beneath ChIP peaks may be undertaken, and should enrich for the known binding consensus of the target protein.\n\n- Peptide binding/competition assays may be performed to evaluate antibody specificity between the target epitope/protein and related proteins. ChIP-peptide methods may be used to quantitatively measure antibody specificity and affinity.\n\n- Stable isotope labelling of amino acids in cell culture (SILAC) may also be used to quantitatively measure antibody specificity and affinity.\n\n- To test for cross-reactivity against proteins related to the target, immunoblots or immunofluorescent experiments may be performed in cells/tissues in which the related protein(s) have been depleted. The antibody signal should not experience a reduction in this case.\n\n- It may be cost-effective to first appraise antibodies in a medium/high-throughput pilot ChIP assay. If the antibody exhibits a reasonable ChIP signal, other validation steps may subsequently be undertaken.\n\n\nPrimary validation\n\nThe primary validation most often employed is the immunoblot or western blot. This assay can be performed on cell or nuclear lysates, with the expectation that an immunoreactive band will be seen at the expected (or known) molecular weight for the protein of interest. In practice, it is likely the blot will reveal (many) other immunoreactive bands, which can suggest that the antibody recognizes other proteins in the sample. This may not be a problem if these other proteins are non-nuclear and hence not present in the chromatin sample being ChIPed; one way to test this is to perform the immunoblot with separate cytoplasmic and nuclear extracts. As a guideline, ENCODE accepts the immunoblot validation if the primary immunoreactive band makes up more than 50% of the blot signal. The immunoreactive band should be of the size expected for the protein, although if not, this will not necessarily rule out a specific signal since many factors, such as post-translation modifications, can affect the electrophoretic mobility of a protein. Especially in these situations, a secondary validation (see below) using cells with reduced or absent levels of the target protein will aid in determining whether the band represents the protein of interest.\n\nInstead of a straight immunoblot assay, it can be advisable to perform a ChIP-immunoblot. In this version of the technique, the eluted protein-antibody complex from the ChIP is saved and run on a gel, then blotted and probed using the antibody against the protein of interest. This assay can be informative of whether the protein alone is pulled down in the ChIP reaction, or if other proteins are also in the eluate, suggesting cross-reactivity or non-specific binding. It can also be a useful guide as to whether the antibody will be successful in the full ChIP assay.\n\nImmunoblots may however be challenging in some systems and with some antibodies, especially where transcription factors are be expressed at low levels. In this case, other methods may need to be considered to show that the antibody recognizes the protein of interest. For instance, the candidate protein can be overexpressed in the cells, or translated in vitro, and immunoblots performed on these protein samples39,40. Alternatively, an immunofluorescence assay may be used as a primary validation of the antibody24, with the expectation that staining should be seen in the nuclei of cells in which the target is known to be expressed.\n\n\nSecondary validation assays\n\nGiven the caveats of the above primary assays, additional assays should be used to add support that an antibody is specific. These secondary assays address slightly different issues, and the more of these validation steps that can be taken, the better.\n\nIn order to further validate the specificity of the antibody immunoblots, immunofluorescence or ChIP assays can be carried out on samples from cells in which the target protein is knocked out or knocked down. In these experiments, the signal for the target protein should be absent or reduced in the mutant/knockdown cells compared to the control. As a guide, data is accepted into the ENCODE pipeline if the immunoreactive signal is reduced by at least 70% in immunoblot or immunofluorescence, or if the ChIP-seq (or ChIP-chip) signal is reduced by at least 50%, in the mutant or knocked down cells24.\n\nImmunoprecipitation followed by mass spectrometry-based sequencing can also be performed, with the expectation that the protein of interest will be identified in the sample24. The presence of other proteins in the sample may not be problematic if these do not bind DNA. However, if other DNA-binding proteins are present these may represent non-specific binding of the antibody; conversely, they may merely represent other proteins that normally occur in a complex with the protein of interest on DNA41. For the ENCODE project, the presence of other DNA-binding proteins was accepted, and samples entered the analysis pipeline, if the other DNA-binding proteins were present at a lower level than the target protein24.\n\nAnother validation approach is to use multiple antibodies to the same protein of interest that target different parts of the protein (or protein complex) in a ChIP-seq assay. With this approach, a sizable overlap of protein-bound peaks in each ChIP should be seen; for instance, ENCODE has historically accepted an overlap of 75% of shared targets, although more recently another quality measure based on the irreproducible discovery rate has been employed24. If other antibodies are not available then using an epitope-tagged version of the protein, and ChIPing with an anti-epitope tag antibody, is an alternative approach and should also give substantial overlap with the endogenous antibody (although caveats apply; see section on tagged proteins above). Indeed, even if an antibody has satisfied other secondary validations, it is still good practice to perform a ChIP with two different antibodies, when available.\n\nFinally, a validation which applies specifically to ChIP-chip or ChIP-seq assays is motif enrichment. Since transcription factors recognize and bind specific DNA sequences, that sequence should be found under bound peaks by a motif search24. ENCODE guidelines suggest searching for a known motif in a defined set of high-quality peaks, with data accepted if the motif is more than 4-fold enriched over all other accessible regions and is present in more than 10% of peaks24. Alternatively if a de novo search for motifs reveals the known binding site, this can also corroborate that the antibody pulls down the protein of interest. However, it is worth noting that target proteins may not interact directly with DNA, in which case the lack of an enriched motif does not preclude the antibody being specific.\n\n\nValidation of histone modification antibodies\n\nFor anti-histone antibodies, additional validations to test the specificity and affinity of the antibodies are recommended. For instance in the ENCODE project, anti-histone antibodies were initially tested in immunoblots against a dilution series of whole-cell or nuclear extracts, and recombinant unmodified histones24,26. Although histones are highly conserved, antibody reactivity may vary between different species and so the antibodies were tested against lysates from each species used in the ChIP assays24. In the case of histones, the guideline is that the specific histone band should make up at least 50% of the immunoblot signal and show at least 10-fold enrichment over any other individual band and the recombinant unmodified histone band24.\n\nPeptide binding or peptide competition assays, using histone tail peptides with particular modifications, are another class of methods to evaluate specificity of anti-histone antibodies24,26. In these tests, an enrichment in binding signal for the modification compared to other modifications should be seen; for ENCODE this enrichment was set at 10-fold24. However, it should be noted that the lack of a signal does not rule the antibody out, since the antibody may not recognize a short peptide in an in vitro environment, but may still be able to bind to the modification in a ChIP assay. That being the case, an improvement to the assay may be the use of peptide ChIP, which allows a quantitative measure of specificity and affinity42.\n\nOther possible validations include mass spectrometry of immunoprecipitated samples as described above, with the target histone species accounting for at least 80% of the immunoprecipitated sample24. An alternative, where resources allow, is mass spectrometry of immunoprecipitated samples after stable isotope labelling of amino acids in cell culture (SILAC). This method compares two samples that incorporate different isotopes of carbon or nitrogen (such as 12C,13C,14N or 15N), allowing the relative abundance of immunoprecipitated proteins to be determined in different samples, hence giving a quantitative measure of antibody specificity and affinity as described by Peach and colleagues43. In addition, immunoprecipitation from cells depleted of or mutant for particular histone modifying enzymes, if available, may be used to validate that an antibody is specific to a particular histone modification24,44. Finally, once ChIP-seq data for the antibody is generated, binding profiles should be inspected for recognized patterns, such as for well-characterized modifications like H3K4me3 at transcription start sites; if binding is as previously established, this can also be taken as further corroboration that the antibody is specific and behaves as expected24.\n\n\nCross-reactivity with family members\n\nAnother issue to be aware of is potential cross-reactivity of the antibody with other proteins related to the protein of interest. For antibodies directed against members of a multi-gene family, it is best to use an antibody that recognizes regions unique to that particular family member of interest. Of course this is not always possible, either due to the nature of the protein, or the lack of suitable antibodies. Moreover, it may be that even if this precaution is taken, the antibody may still non-specifically bind to other family members of the target. Thus validation of the antibody should take related proteins into account.\n\nSequencing an immunoprecipitation reaction by mass spectrometry will give information on whether other family members are present in the sample (see secondary validation above). However, it may not always be possible to sequence the proteins by mass spectrometry, although it should be possible to take a candidate approach and test whether an antibody cross-reacts with related proteins that are expressed at the same time and in same place as the protein of interest. For instance, this can be tested in samples that are knocked down for the related family member, with the expectation that if the antibody spuriously recognizes this related protein, then the immunoreactive signal will be reduced or absent compared to the control. In vitro or in vivo translated proteins for related family members can also be used in immunoblots to ascertain whether the antibody cross-reacts with these related proteins (e.g. 39).\n\n\nWill your antibody work in a ChIP assay?\n\nUnfortunately, even after all these assays have been performed for specificity, they are no guarantee that the antibody will have a high affinity for your protein of interest in a ChIP assay and give a good signal. For instance, Egelhofer and colleagues found that out of 80 anti-histone antibodies that had passed two validation assays (dot blot and immunoblot), 16 (20%) failed to produce a reliable ChIP-seq signal, despite 13 of those being advertised as ChIP-grade (see Supplementary Table 1 of 26). Similarly, Landt and colleagues reported that of 227 transcription factor antibodies that passed two ENCODE validation assays, only 44 (19%) also functioned in ChIP-seq assays24.\n\nAs a considerable amount of effort can be expended in validating an antibody that subsequently fails to give a reliable signal in a ChIP assay, it may be best from a practical point of view to re-order the procedural sequence. For instance, it may be more efficient and cost-effective to initially test candidate antibodies in a medium throughput ChIP assay such as ChIP-string45 or a pilot ChIP-seq assay46. If a good ChIP signal is seen with a particular antibody (and especially if the known motif is identified underneath peaks of binding as a validation for specificity), then further validation steps, as suggested above, can be performed before continuing to use that antibody.\n\n\nConclusions\n\nChIP is now a standard assay used to identify and study protein-DNA interactions, with its use greatly enhancing our understanding of how the genome is regulated in development and disease, and how it has evolved over time. However, its widespread use should not breed complacency in researchers, since the data generated can only be as good as the antibody used. We reiterate that it is critically important that ChIP antibodies be properly validated; tied to this, it is essential that the validations are properly reported on companies' antibody information sheets, and in research publications. We have highlighted a toolkit of possible measures that may be harnessed in validation studies, based on ENCODE guidelines, although we note that this list is not exhaustive and investigators should apply due consideration to the uniqueness of every experimental system and how validation may best be performed in each.", "appendix": "Author contributions\n\n\n\nF.C.W conceived the manuscript and prepared the manuscript. H.T. contributed to preparation of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nENCODE Project Consortium: An integrated encyclopedia of DNA elements in the human genome. Nature. 2012; 489(7414): 57–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGilmour DS, Lis JT: Detecting protein-DNA interactions in vivo: distribution of RNA polymerase on specific bacterial genes. Proc Natl Acad Sci U S A. 1984; 81(14): 4275–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGilmour DS, Lis JT: In vivo interactions of RNA polymerase II with genes of Drosophila melanogaster. Mol Cell Biol. 1985; 5(8): 2009–18. PubMed Abstract \| Free Full Text\n\nGilmour DS, Pflugfelder G, Wang JC, et al.: Topoisomerase I interacts with transcribed regions in Drosophila cells. Cell. 1986; 44(3): 401–7. PubMed Abstract \| Publisher Full Text\n\nGould AP, Brookman JJ, Strutt DI, et al.: Targets of homeotic gene control in Drosophila. Nature. 1990; 348(6299): 308–12. PubMed Abstract \| Publisher Full Text\n\nSolomon MJ, Larsen PL, Varshavsky A: Mapping protein-DNA interactions in vivo with formaldehyde: evidence that histone H4 is retained on a highly transcribed gene. Cell. 1988; 53(6): 937–47. PubMed Abstract \| Publisher Full Text\n\nBarski A, Cuddapah S, Cui K, et al.: High-resolution profiling of histone methylations in the human genome. Cell. 2007; 129(4): 823–37. PubMed Abstract \| Publisher Full Text\n\nIyer VR, Horak CE, Scafe CS, et al.: Genomic binding sites of the yeast cell-cycle transcription factors SBF and MBF. Nature. 2001; 409(6819): 533–8. PubMed Abstract \| Publisher Full Text\n\nJohnson DS, Mortazavi A, Myers RM, et al.: Genome-wide mapping of in vivo protein-DNA interactions. Science. 2007; 316(5830): 1497–502. PubMed Abstract \| Publisher Full Text\n\nMikkelsen TS, Ku M, Jaffe DB, et al.: Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature. 2007; 448(7153): 553–60. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNg P, Tan JJ, Ooi HS, et al.: Multiplex sequencing of paired-end ditags (MS-PET): a strategy for the ultra-high-throughput analysis of transcriptomes and genomes. Nucleic Acids Res. 2006; 34(12): e84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRen B, Robert F, Wyrick JJ, et al.: Genome-wide location and function of DNA binding proteins. Science. 2000; 290(5500): 2306–9. PubMed Abstract \| Publisher Full Text\n\nRobertson G, Hirst M, Bainbridge M, et al.: Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing. Nat Methods. 2007; 4(8): 651–7. PubMed Abstract \| Publisher Full Text\n\nGerstein MB, Lu ZJ, Van Nostrand EL, et al.: Integrative analysis of the Caenorhabditis elegans genome by the modENCODE project. Science. 2010; 330(6012): 1775–87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nmodENCODE Consortium, Roy S, Ernst J, et al.: Identification of functional elements and regulatory circuits by Drosophila modENCODE. Science. 2010; 330(6012): 1787–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMouse ENCODE Consortium, Stamatoyannopoulos JA, Snyder M, et al.: An encyclopedia of mouse DNA elements (Mouse ENCODE). Genome Biol. 2012; 13(8): 418. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang J, Zhuang J, Iyer S, et al.: Sequence features and chromatin structure around the genomic regions bound by 119 human transcription factors. Genome Res. 2012; 22(9): 1798–812. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaker M: Reproducibility crisis: Blame it on the antibodies. Nature. 2015; 521(7552): 274–6. PubMed Abstract \| Publisher Full Text\n\nSchonbrunn A: Editorial: Antibody can get it right: confronting problems of antibody specificity and irreproducibility. Mol Endocrinol. 2014; 28(9): 1403–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBordeaux J, Welsh A, Agarwal S, et al.: Antibody validation. Biotechniques. 2010; 48(3): 197–209. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBradbury A, Pluckthun A: Reproducibility: Standardize antibodies used in research. Nature. 2015; 518(7537): 27–9. PubMed Abstract \| Publisher Full Text\n\nPolakiewicz RD: Antibodies: The solution is validation. Nature. 2015; 518(7540): 483. PubMed Abstract \| Publisher Full Text\n\nReiss PD, Min D, Leung MY: Working towards a consensus for antibody validation [v1; ref status: not peer reviewed, http://f1000r.es/4o9]. F1000Res. 2014; 3: 266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLandt SG, Marinov GK, Kundaje A, et al.: ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Res. 2012; 22(9): 1813–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoens G, Rusu D, Bultot L, et al.: Characterization and quality control of antibodies used in ChIP assays. Methods Mol Biol. 2009; 567: 27–43. PubMed Abstract \| Publisher Full Text\n\nEgelhofer TA, Minoda A, Klugman S, et al.: An assessment of histone-modification antibody quality. Nat Struct Mol Biol. 2011; 18(1): 91–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoskuil J: Commercial antibodies and their validation [v2; ref status: indexed, http://f1000r.es/4jp]. F1000Res. 2014; 3: 232. PubMed Abstract \| Publisher Full Text\n\nPoser I, Sarov M, Hutchins JR, et al.: BAC TransgeneOmics: a high-throughput method for exploration of protein function in mammals. Nat Methods. 2008; 5(5): 409–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFernandez PC, Frank SR, Wang L, et al.: Genomic targets of the human c-Myc protein. Genes Dev. 2003; 17(9): 1115–29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYao Z, Fong AP, Cao Y, et al.: Comparison of endogenous and overexpressed MyoD shows enhanced binding of physiologically bound sites. Skelet Muscle. 2013; 3(1): 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWindner SE, Doris RA, Ferguson CM, et al.: Tbx6, Mesp-b and Ripply1 regulate the onset of skeletal myogenesis in zebrafish. Development. 2015; 142(6): 1159–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeichsenring M, Maes J, Mössner R, et al.: Pou5f1 transcription factor controls zygotic gene activation in vertebrates. Science. 2013; 341(6149): 1005–9. PubMed Abstract \| Publisher Full Text\n\nMazzoni EO, Mahony S, Iacovino M, et al.: Embryonic stem cell-based mapping of developmental transcriptional programs. Nat Methods. 2011; 8(12): 1056–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMali P, Yang L, Esvelt KM, et al.: RNA-guided human genome engineering via Cas9. Science. 2013; 339(6121): 823–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZu Y, Tong X, Wang Z, et al.: TALEN-mediated precise genome modification by homologous recombination in zebrafish. Nat Methods. 2013; 10(4): 329–31. PubMed Abstract \| Publisher Full Text\n\nAndersson L, Archibald AL, Bottema CD, et al.: Coordinated international action to accelerate genome-to-phenome with FAANG, the Functional Annotation of Animal Genomes project. Genome Biol. 2015; 16: 57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBae JB: Perspectives of international human epigenome consortium. Genomics Inform. 2013; 11(1): 7–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParseghian MH: Hitchhiker antigens: inconsistent ChIP results, questionable immunohistology data, and poor antibody performance may have a common factor. Biochem Cell Biol. 2013; 91(6): 378–94. PubMed Abstract \| Publisher Full Text\n\nMorley RH, Lachani K, Keefe D, et al.: A gene regulatory network directed by zebrafish No tail accounts for its roles in mesoderm formation. Proc Natl Acad Sci U S A. 2009; 106(10): 3829–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelson AC, Cutty SJ, Niini M, et al.: Global identification of Smad2 and Eomesodermin targets in zebrafish identifies a conserved transcriptional network in mesendoderm and a novel role for Eomesodermin in repression of ectodermal gene expression. BMC Biol. 2014; 12: 81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMohammed H, D'Santos C, Serandour AA, et al.: Endogenous purification reveals GREB1 as a key estrogen receptor regulatory factor. Cell Rep. 2013; 3(2): 342–9. PubMed Abstract \| Publisher Full Text\n\nNishikori S, Hattori T, Fuchs SM, et al.: Broad ranges of affinity and specificity of anti-histone antibodies revealed by a quantitative peptide immunoprecipitation assay. J Mol Biol. 2012; 424(5): 391–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeach SE, Rudomin EL, Udeshi ND, et al.: Quantitative assessment of chromatin immunoprecipitation grade antibodies directed against histone modifications reveals patterns of co-occurring marks on histone protein molecules. Mol Cell Proteomics. 2012; 11(5): 128–37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHo JW, Jung YL, Liu T, et al.: Comparative analysis of metazoan chromatin organization. Nature. 2014; 512(7515): 449–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRam O, Goren A, Amit I, et al.: Combinatorial patterning of chromatin regulators uncovered by genome-wide location analysis in human cells. Cell. 2011; 147(7): 1628–39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGasper WC, Marinov GK, Pauli-Behn F, et al.: Fully automated high-throughput chromatin immunoprecipitation for ChIP-seq: identifying ChIP-quality p300 monoclonal antibodies. Sci Rep. 2014; 4: 5152. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "10037", "date": "15 Sep 2015", "name": "Niall Dillon", "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 review by Wardle and Tan provides a clear and accessible treatment of the issue of antibody validation for chromatin immunoprecipitation (ChIP). The technical aspects of validating antibodes are covered effectively in the review. The authors also discuss the problems that can arise even with supposedly validated commercial antibodies – for example the fact that companies often change batches of polyclonal antibodies without revalidating them. This is an important problem affecting molecular biology and epigenetic studies and the review is a timely contribution to the discussion about how to address it.", "responses": [] }, { "id": "10036", "date": "12 Oct 2015", "name": "Ana Pombo", "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 manuscript by Wardle and Tan provide an overview of technical concerns and strategies for quality control of antibody specificity in chromatin immunoprecipitation (ChIP) assays. The manuscript summarizes quality standards described previously by others, in particular those used in ENCODE and others. I found the manuscript well written and very useful in general, although felt it could have considered more broadly the use of ChIP to map chromatin bound proteins or DNA modifications, beyond mapping transcription factors or histone modifications. ChIP is also used to map enzymatic activities that work on chromatin, such as DNA and RNA polymerases, chromatin remodelers and RNA processing machinery, DNA repair complexes, which can also be modified by post-translation modifications (PTMs). These other applications of ChIP may be beyond the intended focus of the manuscript, making the content feel incomplete on occasions. The authors may consider stating the focus of their review on ChIP for transcription factors and histone modifications, or alternatively expand the content to a broader coverage of the topic.Major comments:One main aspects that I felt could be improved is the fact that detection of post-translational modifications (PTMs) is not specific to histone modifications, and could work better in a separate section not associated with any kind of protein.The second aspect was on specific remarks about whether ChIP is expected or not to only immunoprecipitate a single protein. Specific points:a) \"Monoclonal vs polyclonal\"It is argued that monoclonal antibodies have a drawback of only detecting one epitope. It could be pointed out that this can be an advantage in situations where researchers only wish to map a single epitope, as is the case for PTMs. In this case, a polyclonal seems to have only disadvantages, in particular the batch to batch unreliability.b) \"Primary validation\" part 1I was expecting to find the use of ELISA with specific peptides or recombinant proteins containing or not the epitope (e.g. peptides with/without PTM, or recombinant protein with/without modified or detected aminoacid(s)) discussed in this section, but it only appears later when discussing histones. c) \"Primary validation\" part 2\"This assay [western blot of ChIP chromatin] can be informative of whether the protein alone is pulled down in the ChIP reaction, or if other proteins are also in the eluate, suggesting crossreactivity or non-specific binding.\" Relatively few proteins will bind to chromatin alone, as the author acknowledges later in the manuscript, namely the subunit of a complex will immunoprecipitate with the other proteins of the same complex. I would not recommend a western blot of ChIPed chromatin as an assay of choice to judge the specificity of an antibody in ChIP. Alternative simpler methods can help judge the enrichment of chromatin bound by a given protein.Subsequent section \"The presence of other proteins in the sample may not be problematic if these do not bind to DNA.\" In line with the above comment, I would suggest either removing the comments that argue that only one protein can be immunoprecipitated for the ChIP to be specific, as this is in most cases an incorrect assumption.d) \"Validation of histone modification antibodies\"Several aspects of this section apply not only to histone modification antibodies, but to any PTMs to chromatin binding factors and enzymatic activities. The arguments for peptide binding etc should be put forward in a broader sense, not only for PTMs, but also to prove specificity of an antibody (mono- or polyclonal) to a given peptide.Edits:- Page 3. Use of word \"ChIPed\".- Page 4. \"immunofluorescent assay\", I would suggest \"immunofluorescence assay\"", "responses": [] } ]	1	https://f1000research.com/articles/4-235
https://f1000research.com/articles/4-180/v2	13 Jul 15	{ "type": "Research Article", "title": "Re-analysis of metagenomic sequences from acute flaccid myelitis patients reveals alternatives to enterovirus D68 infection", "authors": [ "Florian P. Breitwieser", "Carlos A. Pardo", "Steven L. Salzberg", "Florian P. Breitwieser", "Carlos A. Pardo" ], "abstract": "Metagenomic sequence data can be used to detect the presence of infectious viruses and bacteria, but normal microbial flora make this process challenging. We re-analyzed metagenomic RNA sequence data collected during a recent outbreak of acute flaccid myelitis (AFM), caused in some cases by infection with enterovirus D68. We found that among the patients whose symptoms were previously attributed to enterovirus D68, one patient had clear evidence of infection with Haemophilus influenzae, and a second patient had a severe Staphylococcus aureus infection caused by a methicillin-resistant strain. Neither of these bacteria were identified in the original study. These observations may have relevance in cases that present with flaccid paralysis because bacterial infections, co-infections or post-infection immune responses may trigger pathogenic processes that may present as poliomyelitis-like syndromes and may mimic AFM. A separate finding was that large numbers of human sequences were present in each of the publicly released samples, although the original study reported that human sequences had been removed before deposition.", "keywords": [ "microbiome", "metagenomics", "neurological infections", "computational biology", "next-generation sequencing", "sequence alignment" ], "content": "Background\n\nMetagenomic shotgun sequencing, in which DNA or RNA is extracted from a tissue sample and then sequenced, has the potential to detect a wide range of infections. Deep whole-genome shotgun (WGS) sequencing can detect bacteria, viruses, and eukaryotic pathogens with equal effectiveness, as long as the infectious agent is similar to a species that has been previously sequenced. Sequencing databases already contain thousands of known species, and as this number grows, the sensitivity of WGS will grow as well.\n\nIn 2014, a large outbreak of infection with enterovirus D68 was associated with both severe respiratory illness and acute paralysis, which the U.S. Centers for Disease Control and Prevention (CDC) named acute flaccid myelitis (AFM)1. Samples collected from 48 patients were sequenced and shown to form a novel strain, Clade B1, based on phylogenetic analysis of 180 complete enterovirus D68 sequences2. The same study conducted metagenomic sequencing of cerebrospinal fluid (CSF) and/or nasopharyngeal (NP) swabs from 22 of these patients and found enterovirus D68 in some NP samples that were positive based on PCR testing.\n\nThe identification of species from a WGS sample is a challenging problem that has spurred the development of multiple new computational methods3–5. Because of the large size of next-generation sequencing data sets, these methods need to be very fast, but in the context of clinical diagnosis, they also need to be accurate. We downloaded the 31 next-generation sequencing (NGS) samples from the Greninger et al.2 study (NCBI accession SRP055445) and re-analyzed them using a computational pipeline based on the recently developed Kraken metagenomic analysis software4, a very fast and sensitive system that can be customized to use a database containing any species whose sequences are available.\n\n\nAlternative infectious diagnoses in two subjects\n\nAmong the 22 subjects for which NGS data were available, we found at least two that had far greater numbers of sequences (reads) from a bacterial pathogen than from enterovirus D68. Neither subject had been reported in 2 as having a bacterial infection.\n\nIn one subject, US/CA/09-871, reported by Greninger et al.2 as positive for enterovirus D68 through PCR and metagenomic NGS, we found in the NP swab sample an overwhelming presence of bacterial sequences from Haemophilus influenzae, a known cause of meningitis and neurological complications that was a common infection prior to the development of an effective vaccine.\n\nSpecifically, we identified 2,389,621 reads from H. influenzae in this subject, with the closest similarity to strain R2846. These reads comprise 93% of all microbial reads identified at the species level in the sample. Greninger et al.2 reported 2,742 reads (in their Supplementary Table 4) matching enterovirus D682 but did not report finding any H. influenzae reads from this sample. Our analysis found 1,330 reads matching enterovirus D68.\n\nTo confirm the identity of these reads, we aligned them separately to the complete genome of H. influenzae R2846, and we found that the reads completely covered the genome. Dividing the genome into 100 kilobase windows, depth of coverage varied from 266–828 reads/100Kbp, with far deeper coverage as expected at the 16S ribosomal RNA genes.\n\nThe enterovirus D68 isolated from patient US/CA/09-871 differed from the others in that it appeared in 2009, well before the 2014 outbreak, and that it grouped with Clade C, phylogenetically distinct from Clade B1 that was associated with AFM. This patient was reported2 as having respiratory illness but not AFM. The sequence evidence here suggests that the patient might have had complications from H. influenzae-associated infection, although no clinical or CSF data was available for our re-analysis.\n\nIn a second subject, US/CA/12-5837, we found a strikingly large number of reads from Staphylococcus aureus in the NP swabs. The two separate NGS files associated with this subject contained 6,858,453 and 1,343,806 reads, comprising 70% and 84% (respectively) of all non-human reads identified at the species level in each sample. The closest match was S. aureus subsp. aureus MRSA252, a methicillin-resistant strain. The coverage was deep enough, approximately 40X, that it would be possible to assemble this genome separately from the reads here (Figure 1). Greninger et al.2 reported 2,790 reads from enterovirus D68 in this subject (our analysis found 1,641) but did not report any from S. aureus.\n\nHigh peaks correspond to 16S rRNA genes. Red line: median coverage; blue line: mean coverage.\n\nPatient US/CA/12-5837 was sampled in 2012, two years before the outbreak of AFM, although this patient was described in Greninger et al.2 as positive for enterovirus D68 based on clinical PCR testing and metagenomic sequencing. This patient is reported to be one of the first patients with enterovirus-D68-positive AFM2, but the sequence evidence indicates a severe S. aureus infection that might explain at least some of the patient’s symptoms. S. aureus has been implicated in neurological complications such as myelitis6 and meningitis7 by mechanisms that involve not only direct invasion into the central nervous system (CNS), but also immunopathogenic responses triggered by superantigens that can target the CNS8. At a minimum, S. aureus infection was overlooked by the previous analysis. Although the potential role of bacterial infection in the neurological disease that affected these two subjects is difficult to assess because of the lack of clinical and CSF information, its involvement as a pathogenic co-factor should be evaluated.\n\n\nHuman reads included in database submission\n\nThe metagenomics data (NCBI accession SRP055445) released by Greninger et al.2 comprise 43 files which cover 22 of the 48 subjects from their study (in their Supplementary Table 1); the study did not conduct NGS for all subjects. Our metagenomics pipeline identifies human reads at the same time that it searches for pathogens; therefore we scanned the data for human as well as microbial content. Greninger et al.2 reported that all human sequences had been removed from these files. We found, however, that all samples contained large numbers of human reads, ranging from a low of 18,215 to a high of 6,159,868. These comprised as few as 0.5% to as many as 95.6% of the reads in each sample, as shown in Table 1.\n\nShown are the number of reads in each sample that clearly match the human genome and do not match any microbial species. AFM: acute flaccid myelits; NP: nasopharyngeal swap; CSF: cerebrospinal fluid.\n\nThe inclusion of human sequence data in the files deposited at NCBI was likely a result of a computational method (SURPI5) that was insufficiently sensitive. Although the exact cause cannot be determined here, it is well known that sequence alignment algorithms often trade speed for sensitivity; e.g., by allowing fewer mismatches, an aligner can process reads at a much higher rate, at the cost of missing some alignments. It is less clear why the very large numbers of matches to two bacteria were missed; for both these bacteria, complete genomes from multiple strains are available in GenBank. We used both the Kraken system4 and the Bowtie2 aligner9 to ensure both sensitivity and speed in our analysis.\n\nRelease of sequence data is highly valuable, if not essential, for reproducibility and validation of sequencing-based studies. Failure to filter human reads from a sample is not uncommon; a recent study10 found that Human Microbiome Project samples, from which human DNA was supposed to have been removed, contain up to 95% human sequence. This suggests that future efforts to deposit microbiome data need to employ more sensitive computational screens in order to avoid the unintentional release of human sequence data.\n\n\nMethods\n\nSequences were extracted from SRP055445 and each file was separately run through the Kraken program version 0.10.6-beta (https://github.com/DerrickWood/kraken)4, which identifies species by comparison with a database of all 31-bp sequences in all species. The database included the human genome (version GRCh38.p2), all complete bacterial and viral genomes, selected fungal pathogens, and known laboratory vector sequences from the NCBI UniVec database (http://www.ncbi.nlm.nih.gov/tools/vecscreen/univec). Percentages of bacterial and viral reads in each sample were re-computed after excluding human and vector sequences. Reads matching more than one species were classified at the genus level or above. Reads from H. influenzae and S. aureus were re-aligned using Bowtie2 version 2.2.59, a very fast and sensitive program for alignment of NGS reads to a reference genome, with the --local option. Bowtie2 was also used to re-align all reads from US/CA/12-5837 and US/CA/09-871 to the sequence of multiple enterovirus D68 strains (GenBank accessions JX101846.1, AY426531.1, KM851231.1, KM892500.1, KM892501.1, KM881710.2, KP745751.1, KP745755.1, KP745757.1, KP745760.1, KP745764.1, KP745766.1, and KP745767.1). We report the highest number of reads matching any one of these strains.", "appendix": "Author contributions\n\n\n\nSLS conceived the study. FBP ran the computational analyses. SLS, CAP, and FBP jointly analyzed the computational results and wrote the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing interests.\n\n\nGrant information\n\nThis work was supported in part by the National Institutes of Health under grant R01-HG007196 and by the U. S. Army Research Office under grant number W911NF-14-1-0490.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nCenters for Disease Control and Prevention. Notes from the field: acute flaccid myelitis among persons aged ≤21 years - United States, August 1-November 13, 2014. MMWR Morb Mortal Wkly Rep. 2015; 63(53): 1243–1244. PubMed Abstract\n\nGreninger AL, Naccache SN, Messacar K, et al.: A novel outbreak enterovirus D68 strain associated with acute flaccid myelitis cases in the USA (2012–14): a retrospective cohort study. Lancet Infect Dis. 2015; 15(6): 671–82. PubMed Abstract \| Publisher Full Text\n\nBrady A, Salzberg SL: Phymm and PhymmBL: metagenomic phylogenetic classification with interpolated Markov models. Nat Methods. 2009; 6(9): 673–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWood DE, Salzberg SL: Kraken: ultrafast metagenomic sequence classification using exact alignments. Genome Biol. 2014; 15(3): R46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNaccache SN, Federman S, Veeraraghavan N, et al.: A cloud-compatible bioinformatics pipeline for ultrarapid pathogen identification from next-generation sequencing of clinical samples. Genome Res. 2014; 24(7): 1180–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaini M, Prasad K, Ling LM, et al.: Transverse myelitis due to Staphylococcus aureus may occur without contiguous spread. Spinal Cord. 2014; 52(Suppl 2): S1–2. PubMed Abstract \| Publisher Full Text\n\nAguilar J, Urday-Cornejo V, Donabedian S, et al.: Staphylococcus aureus meningitis: case series and literature review. Medicine (Baltimore). 2010; 89(2): 117–25. PubMed Abstract \| Publisher Full Text\n\nStach CS, Herrera A, Schlievert PM: Staphylococcal superantigens interact with multiple host receptors to cause serious diseases. Immunol Res. 2014; 59(1–3): 177–81. 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): 357–359. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmes SK, Gardner SN, Marti JM, et al.: Using populations of human and microbial genomes for organism detection in metagenomes. Genome Res. 2015. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9479", "date": "15 Jul 2015", "name": "David J. Lipman", "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 paper provides a straightforward reanalysis of the metagenomic data presented in Greninger et. al. and does not contradict the basic interpretation of the results in that paper. As noted in the exchange of comments on this paper by Greninger et al. and Breitwieser et. al., while Greninger et al. noted the staph and h.flu sequences in the supplementary data, they did not note these in the paper itself or in the table in the supplementary data that provided additional details - despite the fact that these were present in very high proportions. They also didn't report the presence of human sequence. If all these data were presented explicitly and in the main body of Greninger et al., a reader would be more aware of the challenges of these metagenomic approaches in infectious disease. I hope this paper will encourage this awareness and stimulate discussion so I note the comments on this paper from both sets of authors and I commend Breitwieser et al. for taking the time to respond as they have.", "responses": [] }, { "id": "9476", "date": "12 Aug 2015", "name": "Yoav Gilad", "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 analysis is described in detail and the findings are clear. I found the comments of the authors of the original paper of interest, and the responses of Breitwieser et al. appropriate. I do think that it would be fair to mention in the abstract that this analysis does not challenge the main findings of the original paper. As currently written, the abstract alone does not make this point.", "responses": [] } ]	2	https://f1000research.com/articles/4-180
https://f1000research.com/articles/4-232/v1	13 Jul 15	{ "type": "Systematic Review", "title": "Chemotherapy for Late-Stage Cancer Patients: Meta-Analysis of Complete Response Rates", "authors": [ "Martin L. Ashdown", "Andrew P. Robinson", "Steven L. Yatomi-Clarke", "M. Luisa Ashdown", "Andrew Allison", "Derek Abbott", "Svetomir N. Markovic", "Brendon J. Coventry", "Martin L. Ashdown", "Andrew P. Robinson", "Steven L. Yatomi-Clarke", "M. Luisa Ashdown", "Andrew Allison", "Svetomir N. Markovic", "Brendon J. Coventry" ], "abstract": "Complete response (CR) rates reported for cytotoxic chemotherapy for late-stage cancer patients are generally low, with few exceptions, regardless of the solid cancer type or drug regimen. We investigated CR rates reported in the literature for clinical trials using chemotherapy alone, across a wide range of tumour types and chemotherapeutic regimens, to determine an overall CR rate for late-stage cancers. A total of 141 reports were located using the PubMed database. A meta-analysis was performed of reported CR from 68 chemotherapy trials (total 2732 patients) using standard agents across late-stage solid cancers—a binomial model with random effects was adopted. Mean CR rates were compared for different cancer types, and for chemotherapeutic agents with different mechanisms of action, using a logistic regression. Our results showed that the CR rates for chemotherapy treatment of late-stage cancer were generally low at 7.4%, regardless of the cancer type or drug regimen used. We found no evidence that CR rates differed between different chemotherapy drug types, but amongst different cancer types small CR differences were evident, although none exceeded a mean CR rate of 11%. This remarkable concordance of CR rates regardless of cancer or therapy type remains currently unexplained, and motivates further investigation.", "keywords": [ "Advanced Cancer", "Metastatic Cancer", "Chemotherapy", "Complete Response Rates", "Meta-analysis" ], "content": "Introduction\n\nDespite intense efforts to improve treatment of advanced cancer over many years with numerous cytotoxic agents and dose regimens, some observers have reported that there has been little substantial improvement in treatment outcomes over the last several decades for most cancer types1–5. Several notable exceptions exist, where more successful clinical remission and even cure rates have been shown using chemotherapeutic approaches, such as for testicular cancer using platinum-based agents, and acute childhood leukaemia using vinca alkaloids6–9. Unfortunately, the same has not been true for most other advanced solid malignancies that cause mortality in an estimated 160,000 cancer patients per week internationally, with over 11,000 cancer deaths per week in the USA10.\n\nIn 2006, Kiberstis and Travis1 commented that “An examination of the annual statistical data compiled by the American Cancer Society quickly reveals that the rate of mortality from cancer has changed very little over the past 50 years,” showing little departure from a point made by Bailar2 from a mortality evaluation of the national cancer program between 1950 and 1990, stating “In the end, any claim of major success against cancer must be reconciled with this [increasing U.S. mortality] figure. I do not think such reconciliation is possible and again conclude, as I did seven years ago, that our decades of war against cancer have been a qualified failure.” Again in 1997, Bailar and Gornik commented, “Observed changes in mortality due to cancer primarily reflect changing incidence or early detection. The effect of new treatments for cancer on mortality has been largely disappointing”2.\n\nThis lack of progress persists despite efforts to improve fundamental understanding of cancer growth models11–14, 1.56 million published papers, and around US $200 billion expenditure on cancer research up until 2006 in the US alone, since the National Cancer Act was passed in 19711. This suggests a problem might exist with the current paradigm and the assumption that cytotoxic chemotherapies are acting against cancer cells per se, rather than by some other mechanism. In 2010, Lawrence Baker, Professor of Internal Medicine and Michigan Medical School and Chairman of the Southwest Oncology Group stated, “I am trying to get people to stop saying how successful the cancer research enterprise is, it is not true” and “Cure is clearly the expectation of society”4. Cure, or long-term survivals, are associated with the relatively rare event of complete response (CR), where all cancer disappears as a result of chemotherapy.\n\nThe above statements are significant when considering standard chemotherapy where CR rates, in late-stage disease, are particularly static and therefore disappointing. Using breast cancer as an example, Frasci et al.15 recently reported from the Milan NCI experience, a 7% pathological complete response (pCR) rate using neo-adjuvant combined doxorubicin-paclitaxel and 6% CR rate for advanced breast cancer using an anthracycline-based regimen16. In 1581 patients treated between 1973 and 1982 with consecutive first-line standard-dose doxorubicin and alkylating agent combinations, 263 (16.6%) patients achieved a CR and 49 (3.1%) remained disease free for more than 5 years, and 26 patients (1.5%) remained in first CR at 15 years median follow-up17. A recent study of 2100 patients in 42 phase II trials (70 trial arms) using cytotoxic chemotherapy for metastatic melanoma that completed accrual in the years from 1975 to 2006, conducted by the Southwest Oncology Group, Eastern Cooperative Oncology Group, Cancer and Leukemia Group B, North Central Cancer Treatment Group, and the Clinical Trials Group of the National Cancer Institute of Canada, showed no statistically significant change in progression free survival, or in long-term overall survival over this time period18.\n\nChemo-resistance metabolic pathways have been widely assumed to be the reason for the development of reduced cytotoxicity against many tumor types19, and may be part of the answer. Clonal genetic diversity and clonal outgrowth of less chemo-susceptible tumor clones is another explanation that has been advanced, related to the existence of tumor stem-cells with the capacity to better adapt and grow in response to environmental selection pressure20,21. These theories have led to the development of increased dosage regimens, high-intensity dosing, more frequent-dosing regimens, high-dose myelo-ablative chemotherapy with cellular re-infusion methods, and the use of multiple agent chemotherapy regimens. All have become popular in standard medical oncology practice, but there remains little evidence that more chemotherapy is better in terms of clinical outcomes—undesirable toxicity to normal tissues is often a significant problem for the patient, causing treatment limitations and considerable cost in economic and human terms22. Moreover, CR rates and overall survival have not appreciably improved for most individual cancer types, and to our knowledge, wider analysis of CR rates across many different cancer types, or chemotherapy drug types, has not been performed.\n\nFew studies have addressed the reported CR rates over time for systemic cytotoxic chemotherapeutic treatment across a broad range of advanced solid tumours in a systematic manner. Systematic meta-analysis of CR rates across the spectrum of solid tumours and cytotoxic drug types appeared to be lacking in the literature.\n\nThe objective was to investigate CR rates reported in the literature in clinical trials for advanced cancer treatment across a wide range of cancer types and chemotherapeutic regimens used to date, by conducting a meta-analysis to compare the CR rates and to determine an overall CR rate.\n\n\nMaterials and methods\n\nWe performed this meta-analysis in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA)23.\n\nStudies were considered initially eligible for evaluation if they were published on the recognized PubMed database, in the English language, and accessible in abstract form. Only full text studies with reported valid randomized series of cases of advanced cancer, treated by described methods of chemotherapy administration alone without concomitant surgery or radiotherapy—both of which may confound data interpretation—and clearly reported CR rates were included, published between 2000 and 2006 inclusive. Study follow-up had to be of sufficient length to permit adequate assessment of CR rates.\n\nThe available literature was searched using the PubMed database (http://www.ncbi.nlm.nih.gov/pubmed/), hosted by the National Center for Biotechnology Information (NCBI), U.S. National Library of Medicine. The search date range was 1st January 2000 to 31st December 2006. This time period “snapshot” was chosen commencing in 2000 because of improved standardization of clinical trial response rate reporting after 1999 with the publication of RECIST criteria24 and before the introduction of newer pathway blocking ‘targeted’ agents from 2007 onwards, which progressively were added to chemotherapy agents, and might have confounded the analysis.\n\nThe following search terms “phase 2/3”, “chemotherapy”, “cancer”, “late stage”, and “complete response” were used. To increase the specificity of the query, specific chemotherapy agents (eg. vinblastine, Taxol, cisplatin, 5FU) were included in the search criteria. The search was restricted to clinical trials, reported in English. Using these criteria, 141 candidate trials were identified by abstract. A spreadsheet containing the gathered data is included in the Supplemental material S1.\n\nThree analysts (MLA, SLY-C and BJC) independently examined those studies that were reported in journals that were accessible by the authors. The CR rates had to be recorded in sufficient detail to enable assessment. Trials involving chemotherapy in combination with surgery or radiotherapy were excluded in order to observe clinical responses to chemotherapy alone. Trials were also excluded if their reported response rates in the text were inconsistent with presented data. Disagreements between the analysts about exclusions were resolved by discussion. After exclusions, sixty-eight clinical trials with a total of 2732 assessable patients remained25–92.\n\nData was extracted as already outlined, using a preliminary screen of two analysts identifying the content validity of each study: advanced cancer, numbers of patients >10 and use of a chemotherapeutic agent or agents alone—with no potentially confounding surgery/radiotherapy or other treatments—for suitable follow-up to clearly report any CR rate. Full texts were obtained and a third analyst scrutinized the papers for the details to verify that the abstract CR rates and reported cases were accurately reported. Any paper that could not meet the above criteria or was otherwise unable to be validated was excluded by agreement. Any discrepancy was solved by repeated review, discussion and agreement. Data was collected in a spreadsheet, de-identified and used for statistical analysis.\n\nTo investigate the risk of possible bias in individual clinical trial studies, we relied on standardized reporting methods of clinical trial results as outlined by WHO and RECIST response rate criteria introduced in 200024. The assumption being, all trials would experience similar occurrence of overstatement or understatement of efficacy. As a preliminary guide, to assess feasibility and whether our exclusion criteria might represent bias, we carried out an analysis of CR and Partial Response (PR) data from the published abstracts on the full 130 clinical trials initially identified as eligible. This resulted in an average CR rate of 8.3%, which suggested that more detailed analysis was both feasible and likely to not represent appreciable bias.\n\nThe summary measures used in these studies are the mean CR rate separately determined from meta-analysis of (i) included studies across all cancer types, and (ii) across different primary drug group types specified according to the overall mode of action of each drug type used in the relevant study.\n\nClinical trial data were entered into a spreadsheet—this spreadsheet can be openly accessed in the Supplemental material S1. Fields recorded for each of the 68 trials include: inclusion/exclusion, %CR, %PR, cancer type, drug type, journal citation etc.\n\nThe data were analyzed using a random-effects binomial distribution meta-analysis, for which the response variable was the proportion of patients that were reported to achieve a CR. Separate random-effects binomial distribution meta-analyses were also performed by tumor type and by primary drug type.\n\nWe also applied a generalized linear model to the observed number of complete responders for each study, assuming that the distribution of the response variable was conditionally binomial, and using the logit link function. We tested the effects of tumor type and primary drug type using marginal likelihood ratio tests, that is, each effect was tested with the other included in the model. The analyses were all performed in R version 2.11.1 (http://www.r-project.org), an open-source statistical environment93.\n\nTo investigate the risk of possible bias by individual clinical trial studies, we undertook a sensitivity analysis. This revealed a surprising consistency of ~7% average CR rate. This suggests, that although the population study with respects to cancer type, age, sex of patient and drugs used, were heterogeneous, the CR rate was fixed or stable at ~7%. To illustrate the possible effects of publication bias we have generated a standard funnel plot94,95, as shown in Figure 1. Here we see, the cohort size N for each chemotherapy trial on the vertical axis and the fraction of the cohort with CR on the horizontal axis. The red vertical line represents the average of the CR fraction = 0.0706, which corresponds to 7.06%. By visual inspection overall skew in the plot is small, and thus the effects of publication bias appear to be minimal for this study.\n\nFor each of the 68 chemotherapy trials, the size N of the cohort is plotted on the vertical axis. On the horizontal axis the fraction of the cohort that displayed complete response (CR) is plotted. The red vertical line represents the average of the CR fraction = 0.0706, which corresponds to 7.06%. A standard deviation each side of the average is represented by the curved red lines, calculated by assuming a binomial model. The points appear to be clustered in bands, due to the affect of data quantisation—this is due to the fact that integer numbers of patients have a CR. Thus the first band of points, toward the bottom left of the graph, represent all the trials with a total CR = 1 person, the next band represents all the trials with total CR = 2 people, and so on. For higher total CR numbers, the bands do not appear due to the sparsity of data for large trials. On the funnel plot we see a large spread in CR fraction for 6 trials that are optimistically over a standard deviation, however, the overall skew is nevertheless small. Thus the effects of publication bias appear to be minimal for this study.\n\nIn addition, as a preliminary measure, to see if our exclusion criteria were feasible and whether any bias might exist, we carried out exactly the same analysis on the full 141 clinical trials from CR and PR data in the initial identified published abstracts.\n\n\nResults\n\nReferring to the flow chart in Figure 2, we see that from 141 records resulting from the PubMed database search, 11 were not accessible as full text. Of the remaining 130, only 68 of these, without either missing or incomplete data, were evaluable. The 62 exclusions were due to small numbers of patients in the study, treatment of patients by surgery, difficulty substantiating the conclusions from the data provided, or absence of information on precise CR or PR rates. Exclusion were due to: (i) errors or inconsistencies appeared present, (ii) ambiguity in the patient numbers was detected, (iii) unclear staging, (iv) non-advanced cancer was included, or (v) surgery was used—sometimes post-chemotherapy. The final 68 papers were checked for duplicate studies. None were found, and thus no further exclusions were made on this basis.\n\nThe final 68 studies represented a total of 2732 patients.\n\nOur analyses involved 68 trials and included studies across more than 10 different cancer types and 7 different cytotoxic drug types. These included by (a) cancer type; squamous cell carcinoma (9 trials), renal cell carcinoma (1) prostate (3), pancreatic carcinoma (6), ovarian carcinoma (11), other carcinoma (9), non-small cell lung carcinoma (4), mesothelioma (1), melanoma (4), gastric carcinoma (6), colorectal carcinoma (6), breast carcinoma (6) and brain malignancies (2); and by (b) primary drug type; topoisomerase inhibitor (TPI), T-Cell cytokine inhibitor (cyclosporine) (TCCi), spindle poison (SP), nucleoside analogue (NA), DNA/RNA synthesis inhibitor (DRsi), antimetabolite (AM), alkylating agent (ALK), antibiotic (AB).\n\nThe total number of assessable patients was 2732, of which a total of 193 were classified as complete responders (7.41%). A total of 768 patients were reported as partial responders (28.1%).\n\nIndividual estimates of mean CR rates within the cancer type groupings were: SCC (10.97%); RCC (8.12%); Prostate (10.9%); Pancreatic (3.2%); Ovarian (6.08%); Other (10.57%); NSCLC (5.02%); Mesothelioma (6.33%); Melanoma (8.48%); Gastric (6.79%); Colorectal (6.71%); Breast (10.14%); Brain (6.45%). These are shown collectively in Figure 3.\n\nThese are for (a) Cancer type (SCCa = Squamous cell carcinoma; RCC = renal cell carcinoma; NSCLC = non-small cell lung carcinoma) and by (b) Primary drug type, (abbreviation key; TPI = Topoisomerase inhibitor, TCCi = T-Cell Cytokine inhibitor (Cyclosporine), SP = Spindle poison, NA = Nucleoside analogue, DRsi = DNA/RNA synthesis inhibitor, AM = Antimetabolite, ALK = Alkylating Agent, AB = Antibiotic). The thin vertical blue line denotes a 7.4% CR rate, which is the overall estimate from the meta-analysis.\n\nThe overall meta-analysis estimate of the CR rate from random-effects was 7.41%, [95% confidence intervals (6.27%, 8.64%)].\n\nA generalized linear model that was fitted to the CR rates found a statistically significant relationship between cancer type and CR rate (χ2 = 23.0 on 12 d.f., p = 0.028) and no evidence of a statistically significant relationship between drug type/regimen and CR rate (χ2 = 7.87 on 7 d.f., p = 0.343).\n\nDespite the statistically significant difference between cancer types, the mean CR rates for each individual cancer type only ranged from 3.2% (pancreatic cancer) to 10.9% (prostate cancer)—see Figure 3.\n\nThe ‘other carcinoma’ group comprised cancers of the biliary tract, cervix, uterus, salivary gland, urothelium, and small-cell lung carcinoma. The number of chemotherapy trials using a single agent alone was 17, while the remainder used more than one agent. The number of patients in each trial ranged from 10 to 134, with a median of 37.5.\n\n\nDiscussion\n\nThe complete response (CR) rate has unfortunately remained relatively static for most advanced solid cancers for many decades despite major improvements in understanding considerable new information concerning the molecular genetics, intracellular pathways, adhesion mechanisms, stromal characteristics, angiogenesis, metastatic processes, and immunology, relating to many cancer cell types and modalities of treatment. Similarly, advances in organic chemistry, molecular structural crystallography, synthesis, and pharmaceutical production have not led to expected rapid advances. Numerous different, often quite ingenious, approaches have failed to significantly improve CR rates for survival in patients with advanced cancers20. With this in mind we performed a meta-analysis of existing trials and treatment modalities across most common cancer types over the seven years between 2000 and 2006 inclusive.\n\nIn this paper, we report the results of a meta-analysis of 68 chemotherapy trials for cancer treatment, in which we sought to evaluate the CR rate for late-stage cancer patients receiving chemotherapy, embracing a wide variety of solid tumors and drug types. Overall, the CR rate for patients with most types of late-stage cancers receiving chemotherapy are between 5% and 10%. Our meta-analysis suggests that the CR rate for patients with most types of late-stage cancers receiving chemotherapy are between 5% and 10% across many cancer types, with an actual mean CR rate of 7.41% approximately bisecting this estimated range.\n\nThis study did not set out to include cancers that represent rare notable exceptions, such as testicular carcinoma and childhood acute lymphoblastic leukemia, which are known to be highly chemo-responsive with CR rates of around 80–90%8,9. However, most advanced solid tumors in adults (eg. colon, breast, prostate, melanoma, lung) are unfortunately typically lethal with CR rates that are almost reciprocal to those mentioned. The common advanced solid cancers, therefore, formed the focus of the present meta-analysis.\n\nThe CR rate was surprisingly concordant across the trials despite wide differences in tumor type, chemotherapy combination and mechanism of drug action. This suggests that the range of agents being used is low in direct anti-tumor activity. That finding has been explained by possible development of ‘chemo-resistance’, however, it appears to occur in about 93% of patients for the majority of tumour types and agents. An alternative explanation is interference with the host immune system in an adverse manner, which would abrogate generation or perhaps diminish the effectiveness of an existing anti-tumor immune response96. Many systemically administered chemotherapeutic agents exert their effects non-specifically on many rapidly dividing tissues other than the cancer itself. Anti-angiogenic activity or direct injury to intra-tumoral blood vessels has been proposed as the mechanism of action of some agents, while others induce DNA or mitochondrial damage.\n\nUnfortunately, selectivity for the malignant cells in vivo is often poor and it has been widely appreciated that injury to normal tissues occurs together with tumor cell damage, causing typical side effects such as nausea, diarrhea, marrow suppression, stomatitis, mucositis, and hair loss22. Some of these may be dose limiting and often significantly reduce the quality of life of the patient. In addition to marrow suppressive effects, recent data indicates that the immune system is often injured during chemotherapy treatment, with reduction in the white blood cell count being commonly noted. Leucopaenia may be severe, but is sometimes more subtle, and in particular T-cells can be ablated20. Subsets of rapidly dividing T-cells may be particularly vulnerable to injury, and paradoxically, may be even more susceptible than the cancer cells. If the ablated cells are effector T-cells, then any effective immune response may be eliminated or reduced. However, if regulatory T-cells are selectively ablated or depleted, then the T-effector response may be enhanced96–98. Global depletion of both groups of T-cells may also result in the reduced ability to mount an effective immune response, and this may possibly ‘re-set’ the immune response.\n\nA total of 768 patients were reported as partial responders, with an estimated partial response rate from the meta-analysis of 27.9%—and when combined with complete responders (CR + PR) providing an overall response rate of 35.3%. Effectively, 64.7% of patients were ‘non-responders’ to therapy by not achieving a measurable clinical response. The heterogeneity of clinical responses might reasonably be explained by the random manipulation of the immune system being determined by mathematical probability when chemotherapeutic regimens are randomly applied without consideration of which particular subsets of susceptible immunological cells are actively proliferating at the time of each dosing20,96–100.\n\nIn order to be confident that our systematic exclusions did not skew the results unreasonably, we repeated the meta-analysis using the abstracts from the full complement of 130 papers for which suitable CR results were provided within the summaries. The estimated CR rate was 8.43%, with 95% confidence interval (7.18%, 9.78%), which differed only slightly from the rate reported for the studies that were within the scope of the full-article 68 trial meta-analysis.\n\nThe following caveats should be considered when interpreting our results. Inevitably, the definition of CR varied slightly amongst some studies. The CR was usually defined as complete regression of all detectable tumour clinically or radiologically, while PR was defined more variably. The studies were not necessarily a random sample because they were selected from the PubMed database between 2000 and 2006 and required full-text availability with adequate data for the purposes of the current studies, introducing the potential for inadvertent selection and publication bias. However, every effort was made to avoid this problem. Finally, the analyses included here were the reported clinical response statistics rather than the reported rate estimates and confidence intervals. Hence the analysis might conflict with those reported in the actual paper. A larger study may overcome/diminish such biases.\n\nAnother point is that this study is limited to papers from the PubMed database. The usual recommendation for meta-analyses is to use multiple databases101. However, this is with the intention of mitigating against the omission of negative results that may bias association studies (eg. establishing a positive association between a cancer and a genetic marker). However, this paper is not an association study; we are merely demonstrating the observation that complete response rates for chemotherapy are low across a reasonably broad range of literature. The funnel plot in Figure 1 demonstrates lack of bias as there are roughly equal numbers of points higher and lower than the two standard deviation lines.\n\n\nConclusions\n\nThe knowledge gained from this meta-analysis offers a broad view of the effects of chemotherapy on CR rates for cancers of specific types; collectively for many types of cancers; and for chemotherapy agents with different mechanisms of action. It is particularly noteworthy that the confidence intervals lie progressively closer to the respective mean CR rate for the agents that have historically been in use for the longest time and for which the most clinical trial data exists. This indicates that the agents that have been in use the longest time have a ’real’ CR rate that approximates 7%—an interesting result despite years of use of these agents, even in multiple combination regimens. The potential for variability amongst the individual clinical studies included in the meta-analysis is recognized, however, selection of higher quality valid clinical trials using chemotherapy alone as the sole treatment modality aimed to reduce the natural heterogeneity amongst the studies. In this meta-analysis we further sought to include many cancer types to purposely examine CR rates within a spectrum of cancers. The data offer what we contend is a relatively unbiased and ‘clean’ view of representative ‘real world’ clinical data, from a wide range of sources—indicative of true international clinical experience. The funnel plot indicates that this objective satisfies optimal minimisation of publication bias.\n\nThe significance of the results of our meta-analysis are that across different tumour types, and regardless of different chemotherapy agents/approaches, the CR rates to cancer have remained essentially static and locked at about 7%, despite over 7 years of diligent clinical effort during which these trials were conducted. This might suggest that probabilistic effects are operating to mathematically restrict the ability to manipulate the clinical CR rate—if this can be further understood or overcome, CR rates may potentially be capable of being significantly increased.\n\nThe data might offer an alternative approach to thinking about the possible mechanisms of action of chemotherapy, irrespective of their direct effects on the tumour cells. In considering the consistently low mean CR rates of 7.4% across many tumor types, and using different chemotherapy agents, the findings are suggestive that the paradigm of chemotherapy directly acting for tumour cell killing per se, might be incorrect. Rather, the effects of chemotherapy on immunological cells at any point in time may be of considerable significance, depending on whether rapidly proliferating T-effector cells or T-regulatory cell subgroups are selectively ablated by cytotoxicity, as has been reported in some mouse studies99,100. The balance of the anti-tumor immune response may be pivotally controlled by the relative ablation of either effector or regulatory cell populations, and thereby determine the growth or destruction of the tumor. Indeed, higher doses of cytotoxics might ablate effector T-cells, thereby blunting the immune response and the ability of the patient’s own immune system to be effective. The recent findings concerning the efficacy of metronomic low-dose chemotherapies, administered more frequently, regularly and chronically, would also suggest that the immune system may be pivotal in determining the anti-tumor effect and clinical outcome of the patient102.\n\nWe are currently taking this approach clinically with some success, and are actively investigating the possible timing of chemotherapy and immunotherapy103–105. Optimally timed doses are capable of manipulating the immune response by selective immuno-stimulation or immuno-ablation. The intention is to maximize the in-vivo T-effector response, while minimizing the T-regulatory responses, within the patient, to provide a potential means of synchronizing the ongoing immune response in the patient for improving clinical outcome20,96.\n\nThe main finding of this study is the remarkable concordance of CR rates amongst studies of patients with different cancer types, and also amongst a range of cytotoxic chemotherapy types. This notable similarity in CR rates regardless of cancer or therapy type remains currently unexplained, and requires further intensive investigation.", "appendix": "Author contributions\n\n\n\nConceived and designed the meta-analysis: MLA1. Performed the meta-analysis: MLA1, APR. Screened the data: MLA1, SLY-C, BJC. Analyzed the data: MLA1, APR, SLY-C, MLA2, SNM, BJC, AA, DA. Contributed analysis tools: MLA1, APR, SLY-C, MLA2, SNM, BJC, AA, DA. Wrote the manuscript: MLA1, BJC, DA. Proofed the manuscript: MLA1, APR, SLY-C, MLA2, SNM, BJC, AA, DA.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing financial interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nAR thanks Cecilia Li of the University of Melbourne, for discussions and assistance on statistical aspects. We would also like to thank Tom Sullivan, Data Management Centre, University of Adelaide for helpful advice and additional checking some of the data analysis. Thanks are due to Michael Quinn, University of Melbourne, for useful discussions. We would also like to thank Barbara La Scala and Peter Grossman, University of Melbourne, for statistical assistance and discussion.\n\n\nAdditional information\n\nSupplementary information accompanies this paper as follows:\n\nS1 Spreadsheet of collated information. The spreadsheet contains the raw collated information from the 68 papers included in this study. The columns are described as follows: (A) paper/trial number in order of search download, (B) paper citation details, (C) year published, (D) N = number of patients in trial, (E) CR, number of complete responses post therapy reported in trial, (F) PR, partial responses in trial, (G) cancer type, (H)-(L) chemotherapeutic agent/s used in trial, (M) drug type, (N) drug/s used as mono or multi agent combination.\n\nClick here to access the data.\n\nS2 PRISMA checklist. Standard checklist for compliance to PRISMA standards for the meta-analysis (available at http://www.prismastatement.org/statement.htm).\n\nClick here to access the data.\n\n\nReferences\n\nKiberstis PA, Travis J: Celebrating a glass half-full. Science. 2006; 312(5777): 1157. Publisher Full Text\n\nBailar JC 3rd, Gornik HL: Cancer undefeated. N Eng J Med. 1997; 336(22): 1569–1574. PubMed Abstract \| Publisher Full Text\n\nMorgan G, Ward R, Barton M: The contribution of cytotoxic chemotherapy to 5-year survival in adult malignancies. Clin Oncol (R Coll Radiol). 2004; 16(8): 549–560. PubMed Abstract \| Publisher Full Text\n\nTuma RS: Large trials, small gains: is change on the way? J Natl Cancer Inst. 2010; 102(16): 1216–17. PubMed Abstract \| Publisher Full Text\n\nBeardsley T: A war not won. Sci Am. 1994; 270(1): 130–138. PubMed Abstract \| Publisher Full Text\n\nJeetle SS, Berney DM: Testing testes: problems and recent advances in the diagnosis of testicular tumours and implications for treatment. Clin Oncol (R Coll Radiol). 2012; 24(1): 30–38. PubMed Abstract \| Publisher Full Text\n\nNapper AD, Watson VG: Targeted drug discovery for pediatric leukemia. Front Oncol. 2013; 3: 170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVoutsadakis IA: The chemosensitivity of testicular germ cell tumors. Cell Oncol (Dordr). 2014; 37(2): 79–94. PubMed Abstract \| Publisher Full Text\n\nAsselin BL, Gaynon P, Whitlock JA: Recent advances in acute lymphoblastic leukemia in children and adolescents: an expert panel discussion. Curr Opin Oncol. 2013; 25(Suppl 3): S1–S13; quiz S14-6. PubMed Abstract \| Publisher Full Text\n\nSiegel R, Naishadham D, Jemal A: Cancer statistics, 2013. CA Cancer J Clin. 2013; 63(1): 11–30. PubMed Abstract \| Publisher Full Text\n\nBellomo N, Li NK, Maini PK: On the foundations of cancer modelling: selected topics, speculations, and perspectives. Math Models and Methods in Appl Sci. 2008; 18(4): 593–646. Reference Source\n\nSpencer SL, Berryman MJ, García JA, et al.: An ordinary differential equation model for the multistep transformation to cancer. J Theor Biol. 2004; 231(4): 515–524. PubMed Abstract \| Publisher Full Text\n\nRyan M, Berryman MJ, Abbott D: Modeling the effect of p53 on tumor heterogeneity and the mutator phenotype. Proc. SPIE 5651, Biomedical Applications of Micro- and Nanoengineering II. (Ed. Dan V. Nicolau) Sydney, Australia. 2005; 5651: 144–152. Publisher Full Text\n\nBerryman MJ, Spencer SL, Allison AG, et al.: Fluctuations and noise in cancer development. Proc. SPIE 5471, Noise in Complex Systems and Stochastic Dynamics II, Maspalomas, Gran Canaria, Spain. 2005; 5471(1): 322–332. Publisher Full Text\n\nFrasci G, D'Aiuto G, Comella P, et al.: Southern Italy Cooperative Oncology Group (SICOG). Weekly cisplatin, epirubicin, and paclitaxel with granulocyte colony-stimulating factor support vs triweekly epirubicin and paclitaxel in locally advanced breast cancer: final analysis of a SICOG phase III study. Br J Cancer. 2006; 95(8): 1005–1012. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShehata S, Saad E, Goda Y, et al.: A phase II study of gemcitabine combined with vinorelbine as first-line chemotherapy for metastatic breast cancer. Hematol Oncol Stem Cell Ther. 2010; 3(1): 1–6. PubMed Abstract \| Publisher Full Text\n\nPagani O, Senkus E, Wood W, et al.: International guidelines for management of metastatic breast cancer: can metastatic breast cancer be cured? J Natl Cancer Inst. 2010; 102(7): 456–463. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKorn EL, Liu PY, Lee SJ, et al.: Meta-analysis of phase II cooperative group trials in metastatic stage IV melanoma to determine progression-free and overall survival benchmarks for future phase II trials. J Clin Oncol. 2008; 26(4): 527–534. PubMed Abstract \| Publisher Full Text\n\nLu HP, Chao CC: Cancer cells acquire resistance to anticancer drugs: an update. Biomed J. 2012; 35(6): 464–472. PubMed Abstract\n\nCoventry BJ, Ashdown ML: Complete clinical responses to cancer therapy caused by multiple divergent approaches: a repeating theme lost in translation. Cancer Manag Res. 2012; 4: 137–149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavies P: What’s really going on in those cancer cells? New Scientist. 2013; 217(2898): 24–25. Reference Source\n\nLivshits Z, Rao RB, Smith SW: An approach to chemotherapy-associated toxicity. Emerg Med Clin North Am. 2014; 32(1): 167–203. PubMed Abstract \| Publisher Full Text\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151(4): 264–269, W64. PubMed Abstract \| Publisher Full Text\n\nTherasse P, Arbuck SG, Eisenhauer EA, et al.: New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Nat Cancer Insti. 2000; 92(3): 205–216. PubMed Abstract \| Publisher Full Text\n\nLluch A, Ojeda B, Colomer R, et al.: Doxorubicin and paclitaxel in advanced breast carcinoma: importance of prior adjuvant anthracycline therapy. Cancer. 2000; 89(11): 2169–2175. PubMed Abstract \| Publisher Full Text\n\nCocconi G, Mambrini A, Quarta M, et al.: Vinorelbine combined with paclitaxel infused over 96 hours (VI-TA-96) for patients with metastatic breast carcinoma. Cancer. 2000; 88(12): 2731–2738. PubMed Abstract \| Publisher Full Text\n\nTrudeau M, Seymour R, Stanimir G, et al.: A phase II trial of mitomycin, ifosfamide and cisplatin in recurrent carcinoma of the cervix. Int J Gynecol Cancer. 2000; 10(3): 207–211. PubMed Abstract\n\nKaufman D, Raghavan D, Carducci M, et al.: Phase II trial of gemcitabine plus cisplatin in patients with metastatic urothelial Cancer. J Clin Oncol. 2000; 18(9): 1921–1927. PubMed Abstract\n\nHeinemann V, Wilke H, Mergenthaler HG, et al.: Gemcitabine and cisplatin in the treatment of advanced or metastatic pancreatic cancer. Ann Oncol. 2000; 11(11): 1399–1403. PubMed Abstract\n\nGebbia V, Testa A, Borsellino N, et al.: Cisplatin and vinorelbine in advanced and/or metastatic adenocarcinoma of the endometrium: a new highly active chemotherapeutic regimen. Ann Oncol. 2001; 12(6): 767–772. PubMed Abstract\n\nDouillard JY, Lerouge D, Monnier A, et al.: Combined paclitaxel and gemcitabine as first-line treatment in metastatic non-small cell lung cancer: a multicentre phase II study. Br J Cancer. 2001; 84(9): 1179–1184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShin DM, Khuri FR, Glisson BS, et al.: Phase II study of paclitaxel, ifosfamide, and carboplatin in patients with recurrent or metastatic head and neck squamous cell carcinoma. Cancer. 2001; 91(7): 1316–1323. PubMed Abstract \| Publisher Full Text\n\nMeropol NJ, Niedzwiecki D, Hollis D, et al.: Phase II study of oral eniluracil, 5-fluorouracil, and leucovorin in patients with advanced colorectal carcinoma. Cancer. 2001; 91(7): 1256–1263. PubMed Abstract \| Publisher Full Text\n\nLiu JH, Yang MH, Fan FS, et al.: Tamoxifen and colchicine-modulated vinblastine followed by 5-fluorouracil in advanced renal cell carcinoma: a phase II study. Urology. 2001; 57(4): 650–654. PubMed Abstract \| Publisher Full Text\n\nAiroldi M, Pedani F, Succo G, et al.: Phase II randomized trial comparing vinorelbine versus vinorelbine plus cisplatin in patients with recurrent salivary gland malignancies Cancer. 2001; 91(3): 541–547. PubMed Abstract\n\nBaekelandt M, Lehne G, Tropé CG, et al.: Phase I/II trial of the multidrug-resistance modulator valspodar combined with cisplatin and doxorubicin in refractory ovarian cancer. J Clin Oncol. 2001; 19(12): 2983–2993. PubMed Abstract\n\nMa BB, Tannock IF, Pond GR, et al.: Chemotherapy with gemcitabine-containing regimens for locally recurrent or metastatic nasopharyngeal carcinoma. Cancer. 2002; 95(12): 2516–2523. PubMed Abstract \| Publisher Full Text\n\nKim TW, Kang YK, Ahn JH, et al.: Phase II study of capecitabine plus cisplatin as first-line chemotherapy in advanced gastric cancer. Ann Oncol. 2002; 13(12): 1893–1898. PubMed Abstract \| Publisher Full Text\n\nEspinosa E, Zamora P, Millá A, et al.: A phase II trial of cisplatin and vinorelbine in patients with recurrent or metastatic squamous cell carcinoma of the head and neck. Head Neck. 2002; 24(12): 1054–1059. PubMed Abstract \| Publisher Full Text\n\nSoule SE, Miller KD, Porcu P, et al.: Combined anti-microtubule therapy: a phase II study of weekly docetaxel plus estramustine in patients with metastatic breast cancer. Ann Oncol. 2002; 13(10): 1612–1615. PubMed Abstract \| Publisher Full Text\n\nNeri B, Cini G, Doni L, et al.: Weekly gemcitabine plus Epirubicin as effective chemotherapy for advanced pancreatic cancer: a multicenter phase II study. Br J Cancer. 2002; 87(5): 497–501. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaïeb J, Mitry E, Boige V, et al.: Optimization of 5-fluorouracil (5-FU)/cisplatin combination chemotherapy with a new schedule of leucovorin, 5-FU and cisplatin (LV5FU2-P regimen) in patients with biliary tract carcinoma. Ann Oncol. 2002; 13(8): 1192–1196. PubMed Abstract \| Publisher Full Text\n\nCho EK, Lee WK, Lim DY, et al.: Epirubicin, cisplatin, and protracted venous infusion of 5-fluorouracil for advanced gastric carcinoma. J Korean Med Sci. 2002; 17(3): 348–352. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeo JH, Whang YM, Kim BS, et al.: A phase II study of paclitaxel by 24-hour infusion and ifosfamide in anthracycline-resistant metastatic breast carcinoma. Cancer. 2002; 94(7): 1925–1930. PubMed Abstract \| Publisher Full Text\n\nPlaxe SC, Blessing JA, Bookman MA, et al.: Phase II trial of pyrazoloacridine in recurrent platinum-sensitive ovarian cancer: A gynecologic oncology group study. Gynecol Oncol. 2002; 84(1): 32–35. PubMed Abstract \| Publisher Full Text\n\nEdmonson JH, Blessing JA, Cosin JA, et al.: Phase II study of mitomycin, doxorubicin, and cisplatin in the treatment of advanced uterine leiomyosarcoma: A gynecologic oncology group study. Gynecol Oncol. 2002; 85(3): 507–510. PubMed Abstract \| Publisher Full Text\n\nGonzález-Martin A, Crespo C, García-López JL, et al.: Ifosfamide and vinorelbine in advanced platinum-resistant ovarian cancer: excessive toxicity with a potentially active regimen. Gynecol Oncol. 2002; 84(3): 368–373. PubMed Abstract \| Publisher Full Text\n\nBafaloukos D, Aravantinos G, Fountzilas G, et al.: Docetaxel in combination with dacarbazine in patients with advanced melanoma. Oncology. 2002; 63(4): 333–337. PubMed Abstract \| Publisher Full Text\n\nComella P, Casaretti R, Crucitta E, et al.: Oxaliplatin plus raltitrexed and leucovorin-modulated 5-fluorouracil i.v. bolus: a salvage regimen for colorectal cancer patients. Br J Cancer. 2002; 86(12): 1871–1875. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHan JY, Lee DH, Kim HY, et al.: A phase II study of weekly irinotecan and capecitabine in patients with previously treated non-small cell lung cancer. Clin Cancer Res. 2003; 9(16 Pt 1): 5909–5914. PubMed Abstract\n\nPetty RD, Nicolson MC, Skaria S, et al.: A phase II study of mitomycin C, cisplatin and protracted infusional 5-fluorouracil in advanced pancreatic carcinoma: efficacy and low toxicity. Ann Oncol. 2003; 14(7): 1100–1105. PubMed Abstract \| Publisher Full Text\n\nChua DT, Sham JS, Au GK: A phase II study of capecitabine in patients with recurrent and metastatic nasopharyngeal carcinoma pretreated with platinum-based chemotherapy. Oral Oncol. 2003; 39(4): 361–366. PubMed Abstract \| Publisher Full Text\n\nHirsh V, Whittom R, Ofiara L, et al.: Weekly paclitaxel and gemcitabine chemotherapy for metastatic non-small cell lung carcinoma (NSCLC): a dose-optimizing phase II trial. Cancer. 2003; 97(9): 2242–2247. PubMed Abstract \| Publisher Full Text\n\nBodurka DC, Levenback C, Wolf JK, et al.: Phase II trial of irinotecan in patients with metastatic epithelial ovarian cancer or peritoneal cancer. J Clin Oncol. 2003; 21(2): 291–297. PubMed Abstract \| Publisher Full Text\n\nOhtsu A, Boku N, Yoshioka T, et al.: A Phase II study of irinotecan in combination with 120-h infusion of 5-fluorouracil in patients with metastatic colorectal carcinoma: Japan Clinical Oncology Group Study (JCOG9703). Jpn J Clin Oncol. 2003; 33(1): 28–32. PubMed Abstract \| Publisher Full Text\n\nFracasso PM, Blessing JA, Morgan MA, et al.: Phase II study of oxaliplatin in platinum-resistant and refractory ovarian cancer: a gynecologic group study. J Clin Oncol. 2003; 21(15): 2856–2859. PubMed Abstract \| Publisher Full Text\n\nAravantinos G, Bafaloukos D, Fountzilas G, et al.: Phase II study of docetaxel-vinorelbine in platinum-resistant, paclitaxel-pretreated ovarian cancer. Ann Oncol. 2003; 14(7): 1094–1099. PubMed Abstract \| Publisher Full Text\n\nRose PG, Blessing JA, Ball HG, et al.: A phase II study of docetaxel in paclitaxel-resistant ovarian and peritoneal carcinoma: a Gynecologic Oncology Group Study. Gynecol Oncol. 2003; 88(2): 130–135. PubMed Abstract \| Publisher Full Text\n\nOh WK, Halabi S, Kelly WK, et al.: A phase II study of estramustine, docetaxel, and carboplatin with granulocyte-colony-stimulating factor support in patients with hormone-refractory prostate carcinoma: Cancer and Leukemia Group B 99813. Cancer. 2003; 98(12): 2592–2598. PubMed Abstract \| Publisher Full Text\n\nKoletsky AJ, Guerra ML, Kronish L: Phase II study of vinorelbine and low-dose docetaxel in chemotherapy-naive patients with hormone-refractory prostate cancer. Cancer. 2003; 9(4): 286–292. PubMed Abstract \| Publisher Full Text\n\nReina JJ, Aparicio J, Salvador J, et al.: A multicenter phase II study of irinotecan (CPT-11) alternated with 5-fluorouracil and leucovorin as first-line treatment of patients with metastatic colorectal cancer. Cancer Chemother Pharmacol. 2003; 52(4): 339–345. PubMed Abstract \| Publisher Full Text\n\nJaeckle KA, Hess KR, Alfred Yung WK, et al.: Phase II evaluation of temozolomide and 13-cis-retinoic acid for the treatment of recurrent and progressive malignant glioma: a North American Brain Tumor Consortium study. J Clin Oncol. 2003; 21(12): 2305–2311. PubMed Abstract \| Publisher Full Text\n\nFreyer G, Delozier T, Lichinister M, et al.: Phase II study of oral vinorelbine in first-line advanced breast cancer chemotherapy. J Clin Oncol. 2003; 21(1): 35–40. PubMed Abstract\n\nChitapanarux I, Tonusin A, Sukthomya V, et al.: Phase II clinical study of irinotecan and cisplatin as first-line chemotherapy in metastatic or recurrent cervical cancer. Gynecol Oncol. 2003; 89(3): 402–407. PubMed Abstract \| Publisher Full Text\n\nBajetta E, Di Bartolomeo M, Mariani L, et al.: Randomized multicenter Phase II trial of two different schedules of irinotecan combined with capecitabine as first-line treatment in metastatic colorectal carcinoma. Cancer. 2004; 100(2): 279–287. PubMed Abstract \| Publisher Full Text\n\nFriedland DM, Dakhil S, Hollen C, et al.: A phase II evaluation of weekly paclitaxel plus carboplatin in advanced urothelial cancer. Cancer Invest. 2004; 22(3): 374–382. PubMed Abstract \| Publisher Full Text\n\nLee J, Park JO, Kim WS, et al.: Phase II study of gemcitabine combined with uracil-tegafur in metastatic pancreatic cancer. Oncology. 2004; 66(1): 32–37. PubMed Abstract \| Publisher Full Text\n\nLorusso D, Naldini A, Testa A, et al.: Phase II study of pegylated liposomal doxorubicin in heavily pretreated epithelial ovarian cancer patients. May a new treatment schedule improve toxicity profile? Oncology. 2004; 67(3–4): 243–249. PubMed Abstract \| Publisher Full Text\n\nPectasides D, Pectasides M, Farmakis D, et al.: Oxaliplatin plus high-dose leucovorin and 5-fluorouracil (FOLFOX 4) in platinum-resistant and taxane-pretreated ovarian cancer: a phase II study. Gynecol Oncol. 2004; 95(1): 165–172. PubMed Abstract \| Publisher Full Text\n\nBorrega P, Velasco A, Bolaños M, et al.: Phase II trial of vinorelbine and estramustine in the treatment of metastatic hormone-resistant prostate cancer. Urol Oncol. 2004; 22(1): 32–35. PubMed Abstract \| Publisher Full Text\n\nPerez EA, Hillman DW, Mailliard JA, et al.: Randomized phase II study of two irinotecan schedules for patients with metastatic breast cancer refractory to an anthracycline, a taxane, or both. J Clin Oncol. 2004; 22(14): 2849–2855. PubMed Abstract \| Publisher Full Text\n\nPerez EA, Geoffroy FJ, Hillman S, et al.: Phase II study of oral etoposide and intravenous paclitaxel in extensive-stage small cell lung cancer. Lung Cancer. 2004; 44(3): 347–353. PubMed Abstract \| Publisher Full Text\n\nShepard RC, Levy DE, Berlin JD, et al.: Phase II study of gemcitabine in combination with docetaxel in patients with advanced pancreatic carcinoma (E1298). A trial of the eastern cooperative oncology group. Oncology. 2004; 66(4): 303–309. PubMed Abstract \| Publisher Full Text\n\nMuggia FM, Blessing JA, McGehee R, et al.: Cisplatin and irinotecan in squamous cell carcinoma of the cervix: a phase II study of the Gynecologic Oncology Group. Gynecol Oncol. 2004; 94(2): 483–484. PubMed Abstract \| Publisher Full Text\n\nTsuda H, Hashiguchi Y, Nishimura S, et al.: Phase I–II study of irinotecan (CPT-11) plus nedaplatin (254-S) with recombinant human granulocyte colony-stimulating factor support in patients with advanced or recurrent cervical cancer. Br J Cancer. 2004; 91(6): 1032–1037. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller DS, Blessing JA, Waggoner S, et al.: Phase II evaluation of 9-aminocamptothecin (9-AC, NSC #603071) in platinum-resistant ovarian and primary peritoneal carcinoma: a Gynecologic Oncology Group Study. Gynecol Oncol. 2005; 96(1): 67–71. PubMed Abstract \| Publisher Full Text\n\nLevy T, Inbar M, Menczer J, et al.: Phase II study of weekly topotecan in patients with recurrent or persistent epithelial ovarian cancer. Gynecol Oncol. 2004; 95(3): 686–690. PubMed Abstract \| Publisher Full Text\n\nLong HJ 3rd, Blessing JA, Sorosky J: Phase II trial of dacarbazine, mitomycin, doxorubicin, and cisplatin with sargramostim in uterine leiomyosarcoma: a Gynecologic Oncology Group study. Gynecol Oncol. 2005; 99(2): 339–342. PubMed Abstract \| Publisher Full Text\n\nTsavaris N, Kosmas C, Skopelitis E, et al.: A phase II study of the docetaxel-carboplatin chemotherapy regimen in advanced non-small-cell lung cancer. Lung. 2005; 183(6): 405–416. PubMed Abstract \| Publisher Full Text\n\nÖman M, Lundqvist S, Gustavsson B, et al.: Phase I/II trial of intraperitoneal 5-Fluorouracil with and without intravenous vasopressin in non-resectable pancreas cancer. Cancer Chemother Pharmacol. 2005; 56(6): 603–609. PubMed Abstract \| Publisher Full Text\n\nFeun L, Marini A, Moffat F, et al.: Cyclosporine A, alpha-lnterferon and interleukin-2 following chemotherapy with BCNU, DTIC, cisplatin, and tamoxifen: a phase II study in advanced melanoma. Cancer Invest. 2005; 23(1): 3–8. PubMed Abstract \| Publisher Full Text\n\nMillar J, Scullin P, Morrison A, et al.: Phase II study of gemcitabine and cisplatin in locally advanced/metastatic oesophageal cancer. Br J Cancer. 2005; 93(10): 1112–1116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang SH, Kim MK, Lee TK, et al.: Temozolomide chemotherapy in patients with recurrent malignant gliomas. J Korean Med Sci. 2006; 21(4): 739–744. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaber DA, Okeke RI, Arce-Lara C, et al.: A phase II study of extended dose temozolomide and thalidomide in previously treated patients with metastatic melanoma. J Cancer Res Clin Oncol. 2006; 132(9): 611–616. PubMed Abstract \| Publisher Full Text\n\nHwu WJ, Panageas KS, Menell JH, et al.: Phase II study of temozolomide plus pegylated interferon-alpha-2b for metastatic melanoma. Cancer. 2006; 106(11): 2445–2451. PubMed Abstract \| Publisher Full Text\n\nPinto C, Marino A, De Pangher Manzini V, et al.: Sequential chemotherapy with cisplatin/gemcitabine (CG) followed by mitoxantrone/methotrexate/mitomycin (MMM) in patients with malignant pleural mesothelioma. A multicenter Italian Phase II Study (SITMP1). Lung Cancer. 2006; 52(2): 199–206. PubMed Abstract \| Publisher Full Text\n\nGoto A, Yamada Y, Yasui H, et al.: Phase II study of combination therapy with S-1 and irinotecan in patients with advanced colorectal cancer. Ann Oncol. 2006; 17(6): 968–973. PubMed Abstract \| Publisher Full Text\n\nFeliu J, Salud A, Escudero P, et al.: XELOX (capecitabine plus oxaliplatin) as first-line treatment for elderly patients over 70 years of age with advanced colorectal cancer. Br J Cancer. 2006; 94(7): 969–975. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPapakostas P, Tsavdaridis D, Kosmidis P, et al.: Combination docetaxel (Taxotere), fluorouracil, and leucovorin (TFL), as first-line chemotherapy in advanced gastric cancer: a Hellenic Cooperative Oncology Group phase II study. Gastric Cancer. 2006; 9(1): 26–31. PubMed Abstract \| Publisher Full Text\n\nBaek JH, Kim JG, Jeon SB, et al.: Phase II study of capecitabine and irinotecan combination chemotherapy in patients with advanced gastric cancer. Br J Cancer. 2006; 94(10): 1407–1411. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSantini D, Graziano F, Catalano V, et al.: Weekly oxaliplatin, 5-fluorouracil and folinic acid (OXALF) as first-line chemotherapy for elderly patients with advanced gastric cancer: results of a phase II trial. BMC Cancer. 2006; 6: 125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith HO, Jiang CS, Weiss GR, et al.: Tirapazamine plus cisplatin in advanced or recurrent carcinoma of the uterine cervix: a Southwest Oncology Group study. Int J Gynecol Cancer. 2006; 16(1): 298–305. PubMed Abstract \| Publisher Full Text\n\nR Development Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. 2009. Reference Source\n\nBegg CB, Mazumdar M: Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994; 50(4): 1088–1101. 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): 629–634. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoventry BJ, Ashdown ML, Quinn MA, et al.: CRP identifies homeostatic immune oscillations in cancer patients: a potential treatment targeting tool? J Transl Med. 2009; 7: 102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAshdown ML, Coventry BJ: A matter of time. Australasian Science. 2010; (5): 18–20. Reference Source\n\nAshdown ML, Coventry BJ: Window of opportunity. Australasian Science. 2014; (6): 16–19. Reference Source\n\nAwwad M, North RJ: Cyclophosphamide-induced immunologically mediated regression of a cyclophosphamide-resistant murine tumor: a consequence of eliminating precursor L3T4+ suppressor T-cells. Cancer Res. 1989; 49(7): 1649–1654. PubMed Abstract\n\nDarrasse-Jèze G, Bergot AS, Durgeau A, et al.: Tumor emergence is sensed by self-specific CD44hi memory Tregs that create a dominant tolerogenic environment for tumors in mice. J Clin Invest. 2009; 119(9): 2648–2662. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSagoo GS, Little J, Higgins JP: Systematic reviews of genetic association studies. Human Genome Epidemiology Network. PLOS Med. 2009; 6(3): e1000028. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndré N, Carré M, Pasquier E: Metronomics: towards personalized chemotherapy? Nat Rev Clin Oncol. 2014; 11(7): 413–431. PubMed Abstract \| Publisher Full Text\n\nCoventry BJ, Ashdown ML: The 20th anniversary of interleukin-2 therapy: bimodal role explaining longstanding random induction of complete clinical responses. Cancer Manag Res. 2012; 4: 215–221. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoventry BJ, Lilly CA, Hersey P, et al.: Prolonged repeated vaccine immuno-chemotherapy induces long-term clinical responses and survival for advanced metastatic melanoma. J Immunother Cancer. 2014; 2: 9. Publisher Full Text\n\nDutcher JP, Wiernik PH: Deconstructing and reinventing the IL-2 paradigm: can alternate dosing schedules enhance tumor effect? Kidney Cancer J. 2014; 11: 22–28." }	[ { "id": "9467", "date": "12 Aug 2015", "name": "Joseph J. Drabick", "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 fascinating review by Ashdown and colleagues that examines the CR rate to various cytotoxic regimens for various cancers as part of meta-analysis focusing on the attainment of CR. The remarkable concordance of CR rate across cancer and therapy types is very intriguing. The authors explore possible explanations for this phenomenon and surmise it is most likely immune in nature. This is a remarkable finding that begs for some solution and to my knowledge is a novel observation.", "responses": [] }, { "id": "11333", "date": "25 Nov 2015", "name": "Gábor Balázsi", "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 statistical analysis is appropriate and rigorous. The results are interesting – similar rates of CR for a variety of cancers and drug types. I would like to recommend indexation after a few minor comments have been addressed:It would be interesting and important to include a statement on the rate of Spontaneous Regression (tumour regressing without any treatment). This could simply be extracted from previous publications (no additional analysis needed). This would answer the question: does chemotherapy provide any benefit compared to no treatment at all? It would be interesting to add a sentence on comparing the rates of CR and SR. In the Introduction, a sentence on the molecular mechanisms (such as gene regulation) leading to metastasis would be useful. A relevant citation could be Lee et al. (2014). It would also be interesting to add a sentence on any trends observable over time. If the data are insufficient to make a statement, then the authors should write a sentence stating that.", "responses": [ { "c_id": "1713", "date": "02 Dec 2015", "name": "Derek Abbott", "role": "Author Response", "response": "The authors respond to Professor Balázsi's suggestions as follows:In clinical terms, spontaneous regression is an exceedingly rare event that an individual clinician might perhaps be lucky enough to observe in an entire career of clinical practice - many do not. As such, it is therefore far less than the CR rates that we have identified and reported here in this meta-analysis, and the CR rates are most likely to be the effect(s) of treatment.We agree that molecular mechanisms and gene regulation are important aspects in neoplasia and oncogenesis, but these aspects are not precisely central to the subject of our meta-analysis, and so we did not stray into those areas in this paperThis is an interesting point, however, since we necessarily confined our studies to the time period from 2000 - 2007, we did not see a trend over that short time. However, the suggestion is a particularly interesting one and potentially relevant over a longer period. This is an open question for future study." } ] } ]	1	https://f1000research.com/articles/4-232
https://f1000research.com/articles/3-212/v1	05 Sep 14	{ "type": "Research Article", "title": "Anterior and posterior subareas of the dorsolateral frontal cortex in socially relevant decisions based on masked affect expressions", "authors": [ "Denise Prochnow", "Sascha Brunheim", "Hannes Kossack", "Simon B. Eickhoff", "Hans J. Markowitsch", "Rüdiger J. Seitz", "Sascha Brunheim", "Hannes Kossack", "Simon B. Eickhoff", "Hans J. Markowitsch", "Rüdiger J. Seitz" ], "abstract": "Socially-relevant decisions are based on clearly recognizable but also not consciously accessible affective stimuli. We studied the role of the dorsolateral frontal cortex (DLFC) in decision-making on masked affect expressions using functional magnetic resonance imaging. Our paradigm permitted us to capture brain activity during a pre-decision phase when the subjects viewed emotional expressions below the threshold of subjective awareness, and during the decision phase, which was based on verbal descriptions as the choice criterion. Using meta-analytic connectivity modeling, we found that the preparatory phase of the decision was associated with activity in a right-posterior portion of the DLFC featuring co-activations in the left-inferior frontal cortex. During the subsequent decision a right-anterior and more dorsal portion of the DLFC became activated, exhibiting a different co-activation pattern. These results provide evidence for partially independent sub-regions within the DLFC, supporting the notion of dual associative processes in intuitive judgments.", "keywords": [ "functional connectivity", "dorsolateral frontal cortex", "masked affect expressions", "functional magnetic resonance imaging", "decision-making" ], "content": "Introduction\n\nReading of, and reacting to the numerous dynamic and variable nonverbal signals that are sent out voluntarily or unintentionally in an everyday social situation is challenging and requires the interaction of many brain systems (Frith & Frith, 2003; Xi et al., 2011). Particularly in social situations, people tend to evaluate their surroundings, including their interaction partner (Ellsworth & Scherer, 2003). The human face is the most important object for such an evaluation, since it acts as a key component in conveying socially relevant messages in rapid succession (Ekman & Friesen, 1969). Owing to the complexity of social encounters and the many communicative signals produced by rapidly changing facial expressions, it appears likely that some facial expressions might be too subtle to be perceived fully consciously by the addressee. However, even these transient signals might be of high relevance in “gut-feeling”-based social decisions. For example, inferring even a slightly aggressive emotional state from another’s behavior or facial expression might be crucial for the decision between appeasement in order to avoid confrontation or provocation. Thereby, understanding the mental state of others can be self-profitable for the individual.\n\nThe affective primacy hypothesis (Murphy & Zajonc, 1993) highlights the effects of not consciously perceived affective information, stating that affect can be elicited prior to cognitive processing even when its origin is not consciously accessible. In line with this assumption, studies have shown that subliminal stimuli are processed similarly to consciously accessible stimuli (Henson et al., 2008; Nomura et al., 2004; Prochnow et al., 2013b). Hence they are able to affect attitudes and judgments which are potent determinants of decision-making in complex situations (Dimberg et al., 2000; Li et al., 2008; Moskowitz et al., 2012; Ruys & Aarts, 2012; Sweeny et al., 2009; Winkielman et al., 2005).\n\nDecision-making as a term subsumes multiple aspects such as different phases as well as the circumstances of decision-making, such as risky decisions and ambiguous decisions (Bechara et al., 2005). Typically, gambling paradigms are used to study decision-making (Bechara et al., 1994; Bechara et al., 2005; Brand et al., 2005; Brand et al., 2006). However, there exist also standardized paradigms with more emphasis on social aspects like the Ultimatum Game or the Prisoner’s Dilemma Game (Baumgartner et al., 2011; Güth et al., 1982; Sanfey, 2007; van ’t Wout et al., 2005). Due to the omnipresence of decisions in everyday life, many different experimental settings are suited to assess socially relevant decisions and decision-making often appears to be implicitly studied in mental state reasoning or theory of mind (ToM) paradigms (Hall et al., 2010; Hooker et al., 2008; Mériau et al., 2006; Prochnow et al., 2013a; Reniers et al., 2012; Walter et al., 2004). Recent evidence, however, suggests that gaming and ToM scenarios are based at least partly on different neural circuits (Xi et al., 2011).\n\nSvenson’s “Differentiation and Consolidation Theory” (1996) considers decision-making as the result of a number of different sub-processes. These comprise a pre-decision phase during which different choice alternatives are compared, the decision itself and a post-decision consolidation phase. Following the theory, a number of studies investigated the preparatory processes of different kinds of real-life and gaming decisions and found that the ventromedial frontal cortex (VMFC) and dorsolateral frontal cortex (DLFC) are related to the computation of decision values (Camus et al., 2009; Hall et al., 2010; Jocham et al., 2012; Litt et al., 2010; Reniers et al., 2012; Sokol-Hessner et al., 2012; van ’t Wout et al., 2005). Further evidence suggests that both regions continuously share information during this process (Baumgartner et al., 2011; Sokol-Hessner et al., 2012), along with other interconnected areas within the prefrontal cortex (Miller & Cohen, 2001). The DLFC has also been identified as crucially involved in decisions involving ambiguity or uncertainty, paradigms which are considered being predominantly cognitive in nature (Hosseini et al., 2010; Krain et al., 2006). Accordingly, the DLFC has traditionally been linked to cognitive control and monitoring processes (Cole & Schneider, 2007; Durston et al., 2003; Milham et al., 2003; Wagner et al., 2001).\n\nHowever, increasing evidence shows, that DLFC engagement is not limited to decision and judgment tasks in a predominantly cognitive environment but is found in social and affective contexts as well (Bzdok et al., 2012a; Hall et al., 2010; Lawrence et al., 2006; Opialla et al., 2014; Prochnow et al., 2013a; Prochnow et al., 2013b; Prochnow et al., 2014b; Silvers et al., 2014; Thirioux et al., 2014; Walter et al., 2004). Anatomically, the DLFC has close connections to the parietal and premotor cortices, via the thalamus to the cerebellum (Hoshi, 2006) and also to regions that have been critically implicated in mentalizing, such as the temporo-parietal junction (Bzdok et al., 2012b; Kucyi et al., 2012), the anterior cingulate cortex (ACC), and right-inferior frontal gyrus (IFG) (Cieslik et al., 2013). Notably, in line with previous research highlighting the important role of the DLFC in the preparatory stages of a decision, we found DLFC activity when subjects were presented with either subtle or prominent emotional expressions on which a subsequent decision should be based (Prochnow et al., 2013b; Prochnow et al., 2014b). Conversely, the DLFC became also engaged late during the actual discrimination and categorization of evolving emotional facial expressions, even when the executive load was partly controlled (Prochnow et al., 2013a). While in our studies the activation tended to be located in posterior parts of the DLFC during preparation of the decision, it was located more anterior when the decision itself took place.\n\nIn the current functional magnetic resonance imaging (fMRI) study we extended the earlier study (Prochnow et al. 2013b) to investigate the role of the dorsolateral frontal cortex (DLFC) in socially relevant decisions based on subtle emotional information. In the light of our previous results implicating the DLFC both in the preparatory stage of decision-making as well as in the actual decision, our novel paradigm permitted differentiating between both sub-processes within the same decision process. In particular, we presented facial expressions showing very short (40 ms) happy, angry or sad expressions, which were immediately superimposed by a neutral expression of the same actor, which masked the subtle emotional expression the participants had to evaluate. In this preparatory stage of the decision process, the subjects were already aware that a decision had to be made on the basis of the ambiguous facial expression but necessary information to actually make the decision was still lacking. The actual decision could not been made until pairs of emotional adjectives serving as the decision criterion were presented along with the instruction to decide which adjective matched best the previously seen facial expression. This approach permitted us to explore the role of the DLFC in relation to different aspects of socially-relevant decisions.\n\nWe hypothesized that the DLFC becomes active when socially relevant decisions based on subtle emotional information which is not accessible to fully conscious perception are made. Specifically, based on our own previous data, as well as evidence from primate studies and network analyses (cf. Hoshi, 2006 for a review; Cieslik et al., 2013), we predicted that the pre-decision phase and subsequent decision engage different subareas within the DLFC, and that this at least partly functional specialization is reflected by different co-activation patterns.\n\n\nMaterials and methods\n\nThe screening of the participants comprised of assessments of handedness (Edinburgh inventory, Oldfield, 1971), alexithymia (TAS-20, Bagby et al., 1994), depressiveness (BDI, Hautzinger et al., 1994), empathy (SPF, German adaptation of the Interpersonal Reactivity Index, http://psydok.sulb.uni-saarland.de/volltexte/2009/2363/pdf/SPF_Artikel.pdf) and affect (PANAS, Watson et al., 1988) in order to only enroll participants with an intact ability to understand emotions and infer emotional states. Exclusion criteria were: left handedness, signs of alexithymia (TAS-20 > 52) or depressiveness (BDI > 9), low self-reported empathy (SPF scale fantasy < 10, SPF scale perspective-taking < 13, SPF scale empathic concern < 12), critical life events during the last year (assessed by means of a short self-developed questionnaire asking whether the participants recently experienced the loss of a beloved one or other traumata), a predominantly negative mood on the day of testing (PANAS negative affect > positive affect), intake of psychotropic drugs or a contraindication of fMRI scanning. Contraindications could be pregnancy, fMRI incompatible or irremovable metals like pacemakers or implants, claustrophobia, and fraction anomalies of sight that could not be corrected by MRI suitable glasses or contact lenses. Participants were recruited using flyers on the university campus. From the 18 participants fulfilling the inclusion criteria for the fMRI study, six were later excluded from data analysis due to movement artifacts or reports of being aware of the subtle emotional expressions indicating a too low threshold of subjective awareness which would have been a confounding factor (see the next section for more information on the debriefing procedure). All participants had normal or corrected-to-normal vision and gave informed written consent to participate in the fMRI study and for publication of the study results. Experiments were approved by the ethics committee of the Heinrich-Heine University Düsseldorf (project # 3614) and conducted according to the Declaration of Helsinki. Statistical data analysis was performed on the data from the remaining 12 healthy volunteers (5 men/7 women) who had a mean age of 23.8 (SD = 3.0) and a median of 16.5 (9–18) years of education.\n\nDuring fMRI scanning, participants lay supine in the scanner and viewed the experimental stimuli through a mirror attached to the head coil. The images were presented using presentation software (Version 14.9, Neurobehavioral Systems Inc., Albany CA). During stimulation, participants were presented with male and female facial expressions of emotion depicting happiness, anger or sadness via projection on a semitransparent screen installed in the scanner room using an LCD-projector positioned outside the scanner room (Ekman & Friesen Picture Set, Ekman & Friesen, 1976). They were followed by pairs of emotional adjectives presented as text on screen for 3000 ms (e.g. sorrowful (betrübt) – annoyed (verärgert)) after a jittered (400–4800 ms) time interval. They were instructed to imagine being confronted with someone showing the particular facial expression and to press one of two response buttons (left, right) to decide which adjective corresponded best to the affect of the person depicted. If they felt that none of the adjectives would match, they were requested to choose the best fit (forced choice paradigm).\n\nIn 96 experimental trials which were scanned consecutively in one scanning session, the facial expressions of emotion were shown for only 40 ms and then superimposed by a masking neutral expression of the same person for 360 ms. Each emotion (happy, angry, sad) was repeated 32 times in a pseudorandomized order. In addition, there were another 96 trials in which no masking technique was applied and the emotional expression lasted for 400 ms (for a comparison of the masked emotional and unmasked emotional conditions, see Prochnow et al., 2013b), as well as scrambled images of the facial expressions to measure baseline. Masking is a common technique validated by many studies suited to prevent a short stimulus from being consciously perceived (e.g. Dimberg et al., 2000; Suslow et al., 2013). In order to ensure that despite of the masking technique, our subjects were not aware of the masked emotional expression, they were subjected to a post scanning debriefing similar to the one described in Chartrand & Bargh (1996). The debriefing consisted of increasingly precise questions about the assumed goal of the study, the perception of the stimuli and the procedure. Most participants thought the study was about decision-making or subjective judgments of different facial expressions. However, eight participants (26%) had a suspicion that there were emotional faces presented very shortly before the neutral faces. These were excluded from further data analysis. Furthermore, 78% reported to have noticed a flickering in some of the trials, but did not attribute any meaning to this phenomenon. In fact, the flickering appeared during the switch between the emotional expression and the neutral masking expression.\n\nThe “pictures of facial affect” dataset is one of the most intensively studied facial expression datasets of all times (e.g. Adolphs, 2002; Seitz et al., 2008). It contains expressions of six basic emotions, as well as a neutral reference expression of male and female actors. All neutral faces used as masks in the current study were previously rated neutral in a pre-study with 30 volunteers. In the pre-study, the participants were required to rate whether a presented facial expression represented one of the six basic emotions (anger, sadness, fear, disgust, happiness, surprise) or a neutral expression and to which degree (measured in percent) the expression represented each of the emotions or neutrality. In addition, the emotional adjectives used as the response criteria were matched for word frequency, perceived arousal and dominance (SAM, Bradley & Lang, 1994) based on data from another pre-study in 44 volunteers.\n\nScanning was performed on a 3 T Siemens Trio TIM MRI scanner (Erlangen, Germany) using an EPI-GE sequence (TR = 2000 ms, TE = 30 ms, flip-angle = 90°). The whole brain was covered by 28 transversal slices oriented parallel to the bi-commissural plane (in-plane resolution = 1.5 mm × 1.5 mm, slice thickness = 4.0 mm, interslice gap = 0 mm). In each run, 1200 volumes were acquired. The first three volumes of each session did not enter the analysis. A 3D-T1-weighted image (gradient echo sequence) with high-resolution consisting of 192 sagittal slices and 1 mm × 1 mm resolution was also acquired in each subject (TR = 2300 ms, TE = 2.98 ms, flip angle = 90°).\n\nFMRI scanning was followed by approximately 6 min of anatomical scanning. Post-scanning, participants rated all stimuli on the dimensions arousal, valence and dominance (SAM, Bradley & Lang, 1994) and were debriefed about the experiment.\n\n\nData processing and analysis\n\nBehavioral data were analyzed using SPSS software PASW, Predictive Analysis Software, version 20). Prior to analysis, all statistical data were tested for normal distribution using Kolmogorov-Smirnov test. For comparison of means, single factor analyses of variance (ANOVA) were used.\n\nThe Brainvoyager QX software package (Brain Innovation, Maastricht, The Netherlands) was used for the analysis of imaging data. Functional data were pre-processed including Gaussian spatial smoothing (FWHM = 8), temporal filtering, removal of linear trends and movement correction. In each subject, the 2-D slice time-course image data were co-registered with the volumetric 3-D Gradient Echo data sets from the same session.\n\nWe analyzed the blood oxygenation level dependent (BOLD) changes in a mixed rapid event-related model and entered the planned contrasts in a random effects group analysis. The whole-brain analysis was based on a general linear model (GLM) and a deconvolution approach which allowed the capturing of event-related brain activity at different time steps after event onset, estimating the hemodynamic response function (HRF). The third volume (4000 ms after event onset) was chosen in order to map activation patterns when the blood oxygen dependent (BOLD) increase was close to peak. In this exploratory study, clusters of activations were considered significant when they surpassed a p < 0.005 and had a minimal cluster size of 405 voxels in 3D space (equivalent to 15 cohesive voxels). This procedure corrects for the limited spatial resolution and the autocorrelation of adjacent voxels in the fMRI images and for multiple comparisons (Knorr et al., 1993; Worsley et al., 1992). The following regressors were included: baseline, pre-decision phase, decision phase, and motor control. Scrambled faces (generated by a self-programmed software) served as the baseline condition, and motor control reflected a simple motor response task (reacting towards an unrelated target word out of two words) in order to subtract motor and reading related activity.\n\nIn addition to the whole brain analysis, the activated clusters in the DLFC during the preparatory decision phase as well as the decision itself were defined as regions of interest (ROI) in order to extract their parameter estimates (β) for statistical comparison of the degree of activation between conditions. To ensure comparability, we defined all activated regions within the DLFC as ROIs with a maximum cluster spread range of 10 mm around the peak of activation. All coordinates are given as peak coordinates in Talairach space (Talairach & Tournoux, 1988).\n\nWe used meta-analytic connectivity modeling (MACM) to explore the task-based functional connectivity of the two ROIs identified in this study in the DLFC. After identification of all experiments in the BrainMap database (www.brainmap.org; Laird et al., 2011; Laird et al., 2009) which report activation of the seed regions, quantitative meta-analysis permitted testing for convergence across the clusters of activation reflecting co-activation with the seed regions (Eickhoff et al., 2010). Our analysis was based on approximately 7500 experiments from the BrainMap database reflecting functional mapping studies involving group analyses on healthy participants. Importantly, in order to ensure a completely data-driven approach, all experiments fulfilling the above-mentioned criteria were included regardless of behavioral classification. In a first step, all experiments reporting foci within a 5 mm radius of the seed regions were identified (Cieslik et al., 2011; Eickhoff et al., 2011a), followed by activation likelihood estimation (ALE) to discover co-activations across experiments (Eickhoff et al., 2010; Eickhoff et al., 2009). Importantly, ALE is based on the assumption that the reported foci are not single points but function as centers for 3D Gaussian probability distributions considering the focus-related spatial uncertainty using an empirical model of between-subject and between-template variance (Eickhoff et al., 2009). Voxel-wise combination of the probabilities related to all foci then permitted creating modelled activation (MA) maps (Turkeltaub et al., 2012). These were subsequently merged in order to get voxel-wise and noise-corrected ALE-scores representing the concordance of results at a family-wise error (FWE) corrected p-threshold of p < 0.05 (Eickhoff et al., 2012).\n\nIn a further step, difference maps contrasting functional connectivity maps of the two defined DLFC ROIs were obtained based on their voxel-wise differences as extracted from their MACM-maps. Subsequently, two groups of experiments were formed by pooling and randomly assigning them to same-size groups (Eickhoff et al., 2011b). A repeated (10,000 times) subtraction of the group’s voxel-wise ALE-scores resulted in an empirical null distribution of ALE-score differences between the two conditions. This was followed by thresholding the map of true differences at a probability of p > 0.95 for a true difference between both. To avoid false positive voxels, the resulting maps were masked with the respective main effect of the minuend connectivity map and the minimal cluster size was 20 cohesive voxels.\n\n\nResults\n\nThe fMRI study was preceded by a behavioral study in 32 healthy subjects (mean age 23.9 years, SD = 2.3) testing whether the experimental manipulation was successful (cf. Prochnow et al., 2013b). We found that the subtle masked facial expressions of emotion affected the adjective choice and were thus suitable for a study on decision-making (for a detailed description of the statistical results, please refer to Prochnow et al., 2013b).\n\nWe first present the activation patterns obtained by whole-brain analysis with emphasis on the masked facial expressions of emotion at the pre-decision phase and the subsequent actual decision. Second, we report the comparisons based on the regionally extracted parameter estimates (β) for the two activated areas in DLFC. And finally, we describe the functional connectivity of these seed regions in DLFC.\n\nPre-decision phase: masked facial expressions vs. baseline. In the pre-decision phase, comparing masked emotional facial expressions with scrambled images of faces (baseline) resulted in a bilateral activation of the occipital cortex extending to the fusiform gyrus, of the caudal intraparietal sulcus, as well as of the right superior temporal sulcus, left premotor cortex and most importantly of a right posterior portion of the DLFC (x = 44, y = 16, z = 27, Figure 1).\n\nThese activation clusters were used to define regions of interest based on their activation peaks plus a cluster spread range of 10 mm. The diagrams show their degrees in percent signal change at both events of interest.\n\nDecision phase: decisions based on masked affect expressions vs. motor control. At the moment of the actual decision as indicated by the subjects choice of one of two emotional adjectives following a masked emotional face, we found activation of the left cuneus, left putamen, left paracingulate gyrus, right inferior frontal gyrus and, most importantly, of an anterior portion of the right DLFC (x = 50, y = 28, z = 36, Figure 1).\n\nRegion of interest (ROI) analysis. The activation peak of the ROI related to pre-decisional masked face presentation was located posterior within the DLFC, while the activation peak of the ROI related to the decision phase was located more anterior with a Euclidean distance of 16.16 mm to the ROI related to pre-decisional masked face presentation. This distance exceeded the spatial resolution of the fMRI images (8 mm full width and half maximum (FWHM)).\n\nWe conducted pairwise t-tests to compare parameter estimates between the two DLFC ROIs (for their definition see the Materials and methods section) at α = 0.05, and additionally calculated effect sizes (Cohen’s d) due to the small sample size. The parameter estimates related to pre-decisional masked face presentation did not differ significantly from those during the decision phase (T = -1.02, df = 11, p = 0.329; Cohen’s d = 0.2).\n\nCorrelation analyses revealed that no correlation was found between parameter estimates related to pre-decisional masked face presentation and the decision phase. Notably, the parameter estimates of the decision phase of the masked emotional faces correlated significantly with the accuracy of related decisions following sad expressions. However, parameter estimates in none of the defined DLFC ROIs correlated with self-reported empathy (SPF questionnaire), mood (BDI, Hautzinger et al., 1994) or emotional competence (TAS-20, Bagby et al., 1994).\n\nFunctional connectivity analyses. For the computation of co-activation maps using ALE-based meta-analysis, the posterior ROI related to pre-decisional masked facial expressions and the anterior ROI related to the actual decision phase in the DLFC were used as seed regions. Both were associated with bilateral co-activations in the DLFC and the adjacent premotor cortex. Also, there was task-dependent co-activation in the dorsomedial frontal cortex and around the intraparietal sulcus which was found bilaterally in relation to the seed region associated with pre-decisional masked facial expressions and exclusively right-sided regarding the seed region representing the subsequent decision phase. In addition, the seed region in the DLFC related to pre-decisional masked facial expressions featured co-activations in the inferior frontal gyrus bilaterally and in the left fusiform gyrus.\n\nThe conjunction between co-activations related to both DLFC seed regions comprised two clusters of co-activations in the DLFC, one located more anterior and the other more posterior, a cluster in the left intraparietal sulcus and a cluster in the dorsomedial frontal cortex which included parts of the pre-supplementary motor area (pre-SMA) (Figure 2).\n\nContrasting the co-activation patterns between the two seed regions yielded a more distributed pattern of co-activated clusters in relation to the DLFC seed region associated with the decision phase. This seed region featured stronger co-activations in the left and right DLFC, the adjacent premotor cortex, the dorsomedial frontal cortex, the left pre-SMA and around the left intraparietal sulcus (Figure 2). Interestingly, the seed region in relation to the decision phase was associated with stronger co-activations in two distinct DLFC clusters bilaterally, an anterior and a posterior one, whereas the seed region of pre-decisional masked facial expressions featured a stronger co-activation in a right DLFC region located between these two clusters. Also, it was associated with stronger co-activations in the right inferior frontal gyrus (Figure 2).\n\n\nDiscussion\n\nThis study aimed at identifying the brain areas related to different aspects of decision-making based on masked emotional information that presented a model of daily interpersonal interactions. Specifically, we used a paradigm capable of distinguishing the activation patterns during a preparatory decision phase when not all decision-relevant information was present, from activation patterns related to the decision itself. We found the right DLFC to be involved in both decision stages at clearly different positions: a posterior portion became activated when the actual decision was made as indicated by the subject’s button press (decision). The pre-decision phase during which the subjects were presented with masked emotional facial expressions, which they had to evaluate later, was associated with an activation increase in the right anterior DLFC. No significant differences were found in the degree of activation between both sub-regions, as indicated by the extracted parameter estimates.\n\nThere is a large body of evidence implicating the DLFC in decision-making tasks (Basten et al., 2010; Domenech & Dreher, 2010; Gilbert et al., 2010; Hall et al., 2010; Hayama & Rugg, 2009; Hosseini et al., 2010; Huettel & Misiurek, 2004; Plassmann et al., 2007), especially when the decisions are characterized by some degree of ambiguity (Christakou et al., 2009; Kahnt et al., 2011; Krain et al., 2006). Moreover, DLFC activity has been found in various higher-order cognitive tasks such as working memory and monitoring tasks (Durston et al., 2003; Kellermann et al., 2012; Opitz et al., 2000; Wagner et al., 2001) and cognitive control tasks (Cieslik et al., 2010; Cole & Schneider, 2007; Eickhoff & Grefkes, 2011; Milham et al., 2003; Jakobs et al., 2009). These are considered pre-dominantly “cold” cognitively-driven tasks (Zelazo & Muller, 2002) and may act as key players in self-related control tasks such as decision-making and choice (reviewed by Banfield et al., 2004).\n\nHowever, even though affect-based decisions have been traditionally linked to the recruitment of the ventromedial and orbitofrontal prefrontal cortex, which we failed to observe in the current study (Chib et al., 2009; Grabenhorst & Rolls, 2011; Krain et al., 2006; Smith et al., 2010; Zelazo & Muller, 2002), we consistently found DLFC activation in affective judgment tasks (Prochnow et al., 2013a; Prochnow et al., 2013b; Prochnow et al., 2014b). Our observations are supported by studies using affective tasks which implicitly studied decisions in an affective context (Bzdok et al., 2012a; Lawrence et al., 2006; Opialla et al., 2014; Silvers et al., 2014; Thirioux et al., 2014; Walter et al., 2004) In order to model daily interpersonal interactions we intentionally created a decision-making paradigm in which the subjects had to base their decisions on subtle and thus ambiguous facial expressions. Following the affective primacy hypothesis (Murphy & Zajonc, 1993), the emotional expressions were considered to elicit an affective response in the observer even though the subjects were not aware of having seen them, similarly as to what Ekman has described as micro expressions (Ekman, 1992; Shen et al., 2012). The short emotional expression was thus expected to add an emotional flavor onto the masking neutral expression which loaded an ambiguous stimulus with a specific emotional state (Rohr et al., 2012; Prochnow et al., 2013b).\n\nIn the current study, as well as in previous studies (Prochnow et al., 2013b; Prochnow et al., 2014b), we show that already during the presentation of pre-decisional masked facial expressions a posterior and more ventral portion of the DLFC became activated. According to anatomical coordinates, this activation cluster corresponded to dorsolateral frontal regions found in normative decision-making (Baumgartner et al., 2011) and ill-structured problem-solving (Gilbert et al., 2010), indicating its importance in the decision-making process. During this preparatory stage of decision-making, when not all necessary information to make a goal-directed decision is present, Svenson’s theory assumes that calculation of decision values takes place (Svenson, 1996). Evidence for the involvement of the DLFC in the calculation of decision values comes from a growing number of studies (Camus et al., 2009; Litt et al., 2010; Plassmann et al., 2007; Sokol-Hessner et al., 2012). Notably, a more anterior and dorsal portion of the DLFC became activated when the adjectives offered as the decision criteria were presented and the subjects had to make a decision (forced choice paradigm). This result is in line with our previous study showing anterior DLFC engagement during online emotion discrimination and categorization (Prochnow et al., 2013a) and suggests that the anterior portion of the DLFC is associated with uncertain decisions (Hosseini et al., 2010).\n\nDLFC activations reported in the literature are heterogeneous in their locations and also regarding their related tasks. Most clusters are situated in close proximity to the anterior cluster found here or even more anterior. Functionally, they are referred to working memory and monitoring (Rottschy et al., 2012; Wagner et al., 2001), self-reflection (Herwig et al., 2012), cognitive control or cognitive conflict (Cieslik et al., 2010; Eickhoff & Grefkes, 2011; Jakobs et al., 2009; Milham et al., 2003) and different aspects of decision-making (Krain et al., 2006; Plassmann et al., 2007; Prochnow et al., 2013a). Especially, there seems to be a conceptual overlap of studies examining cognitive control, cognitive conflict and decision-making depending on the focus of the study. Whereas studies focusing on decision-making, including the current study, implicitly study aspects of cognitive control, studies on cognitive control appear to imply aspects of decision-making. In order to get further insights into the functional connectivity of the DLFC, this study also focused on the identification of co-activations of the two subareas within the DLFC obtained in the whole brain analysis.\n\nThe analyses of functional connectivity showed that the posterior DLFC cluster activated during the pre-decision phase featured stronger co-activations in the right inferior frontal gyrus (IFG) and in a DLFC area located between the precentral and inferior frontal sulcus. By contrast, the anterior portion of the DLFC that became activated during the actual decision was associated with stronger co-activations in two DLFC areas framing the DLFC region co-activated in relation to the posterior DLFC seed region. In addition, it featured co-activations of the premotor cortex, a dorsomedial frontal region, the left pre-SMA and the left intraparietal sulcus. Activation of the IFG has been found repeatedly in tasks involving low-level empathy (Carr et al., 2003; Lamm et al., 2007; Lindenberg et al., 2012; Schulte-Rüther et al., 2007; Seitz et al., 2008; Shamay-Tsoory et al., 2009; Prochnow et al., 2013a), most likely because it is considered an important node of the putative human mirror neuron system (Rizzolatti & Craighero, 2004). Moreover, the left IFG is well known to accommodate Broca’s speech area (Lindenberg et al., 2007) and its activation might therefore also reflect covert speech. Accordingly, in our paradigm one would expect left IFG activity to co-occur during the actual decision since at this stage, the subjects were confronted with verbal descriptions in form of two emotional adjectives they were required to choose in order to respond. Instead, the whole brain analysis showed an activation increase in the right inferior frontal gyrus during the actual decision, and neither the pre-decision phase, nor the actual decision was associated with an activation increase in the left IFG in this sample. However, although the pre-decision phase does not involve any explicit speech component, it remains impossible to control for covert speech in fMRI tasks like ours.\n\nInterestingly, in the current study activity in the anterior portion of the DLFC associated with the actual decision was also accompanied by an activation increase in the left paracingulate gyrus. This dorsomedial prefrontal region has been found relevant for rapid interpersonal evaluations (Cooper et al., 2012) and theory of mind (Hooker et al., 2008; Schulte-Rüther et al., 2007). Moreover, the adjacent pre-SMA has been shown to be crucial in the context of the generation of the so-called Bereitschaftspotential to perform a movement (Shibasaki & Hallett, 2006), as well as for movement selection (Deiber et al., 1991; Hoffstaedter et al., 2013). Interestingly, it was not only found active during the recognition of emotions in static emotional facial expressions (Seitz et al., 2008) but also when dynamically evolving emotional facial expressions had to be discriminated (Prochnow et al., 2013a). These observations suggest that the dorsomedial portion of the prefrontal cortex including the adjacent pre-SMA becomes involved when an external mental state needs to be transferred into an internal frame of reference (Seitz et al., 2006; Seitz et al., 2009).\n\nIn addition to the identification of different patterns of functional connectivity between the posterior DLFC region related to the pre-decision phase and the anterior region related to the decision phase, we were interested in the co-activations shared by both DLFC regions. These were bilateral anterior and posterior areas in the DLFC, the dorsomedial frontal cortex including the pre-SMA and the left intraparietal sulcus, suggesting a common network allowing for visuo-spatial and time-related attention (Culham & Kanwisher, 2001; Davranche et al., 2011; Grefkes & Fink, 2005) and self-referential valuation (Seitz et al., 2006; Seitz et al., 2009).\n\nIn the current study, activations of the two subregions in the DLFC were clearly lateralized to the right cerebral hemisphere featuring co-activations distributed over both hemispheres. This result corresponds to behavioral evidence showing that not consciously accessible faces affected choices regardless of the visual hemifield to which they were presented while, in contrast, subliminally presented words affected choices only when they were presented to the left cerebral hemisphere (Henke et al., 1994).\n\nPossible limitations of the current study should not go unmentioned. We considered the moment when our subjects viewed the emotional masked facial expressions the preparatory stage of the actual decision since not all relevant information was present to make a goal-directed choice. It cannot, however, be ruled out that instead of measuring a pre-decision phase and the actual decision, there were two different decisions following one-another. A first partial decision based on only the visual information and the outside of subjective awareness elicited affective response and a subsequent decision when the emotional adjectives as the decision criterion were available. For example, Wunderlich et al. (2010) provided evidence that people are able to partially make a choice in stimulus space before knowing the motor mapping associated with the final decision. Independent of these theoretical considerations, our fMRI and functional connectivity data showed that both time points were associated with the involvement of different parts of the DLFC indicating functional specialization in the DLFC. Instead of representing a pre-decision phase and the decision itself, the anterior-posterior subdivision could also reflect different degrees to which the decision was goal-directed.\n\n\nConclusions\n\nIn conclusion, our data suggest that the DLFC is crucial for decisions involving masked, and thus, ambiguous affective information. Moreover, by use of categorical and functional connectivity image analysis approaches we provide evidence for partially independent sub-regions within the right DLFC. Whereas the posterior portion of the right DLFC was relevant for the preparatory phase within the decision process when not all the necessary information for a goal-directed choice were available, the anterior sub-region appeared to be related to later goal-directed decision stages involving sustained attention for time, space and valuation. These results may be related to the notion of dual associative processes in intuitive judgments (Morewedge & Kahneman, 2010).\n\n\nParticipant consent\n\nAll participants gave informed written consent to participate in the fMRI study. Experiments were approved by the local ethics committee and conducted according to the Declaration of Helsinki.\n\n\nData availability\n\nfigshare: Statistical data of subareas of the dorsolateral frontal cortex in socially relevant decisions based on masked affect expressions. Doi: 10.6084/m9.figshare.1153792 (Prochnow et al., 2014a).", "appendix": "Author contributions\n\n\n\nDP and RS conceived the study. DP, RS and SB designed the experiments. DP, HK and SB carried out the research. SB provided technical support during data collection. SE contributed to the design of the experiments and provided expertise in MACM. DP, HK and SB analyzed the fMRI data, SE carried out the MACM. DP prepared the first draft of the manuscript under supervision of RS and HM. All authors have agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe study was financially supported by 7FP of the European commission (RGS).\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\nThe authors thank Erika Rädisch for her assistance and support with the fMRI measurements.\n\n\nReferences\n\nAdolphs R: Neural systems for recognizing emotion. Curr Opin Neurobiol. 2002; 12(2): 169–177. PubMed Abstract \| Publisher Full Text\n\nBagby RM, Parker JD, Taylor GJ: The twenty-item Toronto Alexithymia Scale--I. Item selection and cross-validation of the factor structure. J Psychosom Res. 1994; 38(1): 23–32. PubMed Abstract \| Publisher Full Text\n\nBanfield JF, Wyland CL, Macrae CN, et al.: The cognitive neuroscience of self-regulation. In Baumeister, R.F. & Vohs, K.D. (Eds.) Handbook of self-regulation: Research, theory, and applications, New York: Guilford Press, 2004; (pp. 62–83). Reference Source\n\nBasten U, Biele G, Heekeren HR, et al.: How the brain integrates costs and benefits during decision making. Proc Natl Acad Sci U S A. 2010; 107(50): 21767–21772. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaumgartner T, Knoch D, Hotz P, et al.: Dorsolateral and ventromedial prefrontal cortex orchestrate normative choice. Nat Neurosci. 2011; 14(11): 1468–1474. PubMed Abstract \| Publisher Full Text\n\nBechara A, Damasio AR, Damasio H, et al.: Insensitivity to future consequences following damage to human prefrontal cortex. Cognition. 1994; 50(1–3): 7–15. PubMed Abstract \| Publisher Full Text\n\nBechara A, Damasio H, Tranel D, et al.: The Iowa Gambling Task and the somatic marker hypothesis: some questions and answers. Trends Cogn Sci. 2005; 9(4): 159–162. PubMed Abstract \| Publisher Full Text\n\nBradley MM, Lang PJ: Measuring emotion: the Self-Assessment Manikin and the Semantic Differential. J Behav Ther Exp Psychiatry. 1994; 25(1): 49–59. PubMed Abstract \| Publisher Full Text\n\nBrand M, Fujiwara E, Borsutzky S, et al.: Decision-making deficits of korsakoff patients in a new gambling task with explicit rules: associations with executive functions. Neuropsychology. 2005; 19(3): 267–277. PubMed Abstract \| Publisher Full Text\n\nBrand M, Labudda K, Markowitsch HJ: Neuropsychological correlates of decision-making in ambiguous and risky situations. Neural Netw. 2006; 19(8): 1266–1276. PubMed Abstract \| Publisher Full Text\n\nBzdok D, Langner R, Hoffstaedter F, et al.: The modular neuroarchitecture of social judgments on faces. Cereb Cortex. 2012a; 22(4): 951–961. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBzdok D, Schilbach L, Vogeley K, et al.: Parsing the neural correlates of moral cognition: ALE meta-analysis on morality, theory of mind, and empathy. Brain Struct Funct. 2012b; 217(4): 783–796. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCamus M, Halelamien N, Plassmann H, et al.: Repetitive transcranial magnetic stimulation over the right dorsolateral prefrontal cortex decreases valuations during food choices. Eur J Neurosci. 2009; 30(10): 1980–1988. PubMed Abstract \| Publisher Full Text\n\nCarr L, Iacoboni M, Dubeau MC, et al.: Neural mechanisms of empathy in humans: a relay from neural systems for imitation to limbic areas. Proc Natl Acad Sci U S A. 2003; 100(9): 5497–5502. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChartrand TL, Bargh JA: Automatic activation of impression formation and memorization goals: Nonconscious goal priming reproduces effects of explicit task instructions. J Pers Soc Psychol. 1996; 71(3): 464–478. Publisher Full Text\n\nChib VS, Rangel A, Shimojo S, et al.: Evidence for a common representation of decision values for dissimilar goods in human ventromedial prefrontal cortex. J Neurosci. 2009; 29(39): 12315–12320. PubMed Abstract \| Publisher Full Text\n\nChristakou A, Brammer M, Giampietro V, et al.: Right ventromedial and dorsolateral prefrontal cortices mediate adaptive decisions under ambiguity by integrating choice utility and outcome evaluation. J Neurosci. 2009; 29(35): 11020–11028. PubMed Abstract \| Publisher Full Text\n\nCieslik EC, Zilles K, Caspers S, et al.: Is there “one” DLPFC in cognitive action control? Evidence for heterogeneity from co-activation-based parcellation. Cereb Cortex. 2013; 23(11): 2677–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCieslik EC, Zilles K, Grefkes C, et al.: Dynamic interactions in the fronto-parietal network during a manual stimulus-response compatibility task. Neuroimage. 2011; 58(3): 860–869. PubMed Abstract \| Publisher Full Text\n\nCieslik EC, Zilles K, Kurth F, et al.: Dissociating bottom-up and top-down processes in a manual stimulus -response compatibility task. J Neurophysiol. 2010; 104(3): 1472–1483. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCole MW, Schneider W: The cognitive control network: Integrated cortical regions with dissociable functions. Neuroimage. 2007; 37(1): 343–360. PubMed Abstract \| Publisher Full Text\n\nCooper JC, Dunne S, Furey T, et al.: Dorsomedial prefrontal cortex mediates rapid evaluations predicting the outcome of romantic interactions. J Neurosci. 2012; 32(45): 15647–15656. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCulham JC, Kanwisher NG: Neuroimaging of cognitive functions in human parietal cortex. Curr Opin Neurobiol. 2001; 11(2): 157–163. PubMed Abstract \| Publisher Full Text\n\nDavranche K, Nazarian B, Vidal F, et al.: Orienting attention in time activates left intraparietal sulcus for both perceptual and motor task goals. J Cogn Neurosci. 2011; 23(11): 3318–3330. PubMed Abstract \| Publisher Full Text\n\nDeiber MP, Passingham RE, Colebatch JG, et al.: Cortical areas and the selection of movement: a study with positron emission tomography. Exp Brain Res. 1991; 84(2): 393–402. PubMed Abstract \| Publisher Full Text\n\nDimberg U, Thunberg M, Elmehed K: Unconscious facial reactions to emotional facial expressions. Psychol Sci. 2000; 11(1): 86–89. PubMed Abstract \| Publisher Full Text\n\nDomenech P, Dreher JC: Decision threshold modulation in the human brain. J Neurosci. 2010; 30(43): 14305–14317. PubMed Abstract \| Publisher Full Text\n\nDurston S, Davidson MC, Thomas KM, et al.: Parametric manipulation of conflict and response competition using rapid mixed-trial event-related fMRI. Neuroimage. 2003; 20(4): 2135–2141. PubMed Abstract \| Publisher Full Text\n\nEickhoff SB, Bzdok D, Laird AR, et al.: Activation likelihood estimation meta-analysis revisited. Neuroimage. 2012; 59(3): 2349–2361. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEickhoff SB, Bzdok D, Laird AR, et al.: Co-activation patterns distinguish cortical modules, their connectivity and functional differentiation. Neuroimage. 2011b; 57(3): 938–949. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEickhoff SB, Grefkes C: Approaches for the integrated analysis of structure, function and connectivity of the human brain. Clin EEG Neurosci. 2011; 42(2): 107–121. PubMed Abstract \| Publisher Full Text\n\nEickhoff SB, Jbabdi S, Caspers S, et al.: Anatomical and functional connectivity of cytoarchitectonic areas within the human parietal operculum. J Neurosci. 2010; 30(18): 6409–6421. PubMed Abstract \| Publisher Full Text\n\nEickhoff SB, Laird AR, Grefkes C, et al.: Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: a random-effects approach based on empirical estimates of spatial uncertainty. Hum Brain Mapp. 2009; 30(9): 2907–2926. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEickhoff SB, Pomjansi W, Jakobs O, et al.: Neural correlates of developing and adapting behavioral biases in speeded choice reactions—An fMRI study on predictive motor coding. Cereb Cortex. 2011a; 21(5): 1178–1191. PubMed Abstract \| Publisher Full Text\n\nEkman P: Facial expressions of emotion: an old controversy and new findings. Philos Trans R Soc Lond B Biol Sci. 1992; 335(1273): 63–69. PubMed Abstract \| Publisher Full Text\n\nEkman P, Friesen W: Nonverbal leakage and clues to deception. Psychiatry. 1969; 32(1): 88–106. PubMed Abstract\n\nEkman P, Friesen W: Pictures of facial affect. Palo Alto, CA Consulting Psychologists Press. 1976. Reference Source\n\nEllsworth PC, Scherer K: Appraisal processes in emotion. In R. J. Davidson, K. R. Scherer, & H. Goldsmith (Eds.), Handbook of Affective Sciences. Oxford, UK: Oxford University Press. 2003; pp. 572–595. Reference Source\n\nFrith U, Frith CD: Development and neurophysiology of mentalizing. Philos Trans R Soc Lond B Biol Sci. 2003; 358(1431): 459–473. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGilbert SJ, Zamenopoulos T, Alexiou K, et al.: Involvement of right dorsolateral prefrontal cortex in ill-structured design cognition: an fMRI study. Brain Res. 2010; 1312: 79–88. PubMed Abstract \| Publisher Full Text\n\nGrabenhorst F, Rolls ET: Value, pleasure and choice in the ventral prefrontal cortex. Trends Cogn Sci. 2011; 15(2): 56–67. PubMed Abstract \| Publisher Full Text\n\nGrefkes C, Fink GR: The functional organization of the intraparietal sulcus in humans and monkeys. J Anat. 2005; 207(1): 3–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGüth W, Schmittberger R, Schwarze B: An experimental analysis of ultimatum bargaining. J Econ Behav Organ. 1982; 3(4): 367–388. Publisher Full Text\n\nHall J, Whalley HC, McKirdy JW, et al.: A common neural system mediating two different forms of social judgement. Psychol Med. 2010; 40(7): 1183–1192. PubMed Abstract \| Publisher Full Text\n\nHautzinger M, Bailer M, Worall H, et al.: Beck Depressions-Inventar (BDI) Testhandbuch. Huber, Bern. 1995. Reference Source\n\nHayama HR, Rugg MD: Right dorsolateral prefrontal cortex is engaged during post-retrieval processing of both episodic and semantic information. Neuropsychologia. 2009; 47(12): 2409–2416. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHenke K, Landis T, Markowitsch HJ: Subliminal perception of words and faces. Int J Neurosci. 1994; 75(3–4): 181–187. PubMed Abstract \| Publisher Full Text\n\nHenson RN, Mouchlianitis E, Matthews EJ, et al.: Electrophysiological correlates of masked face priming. Neuroimage. 2008; 40(2): 884–895. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerwig U, Kaffenberger T, Schell C, et al.: Neural activity associated with self-reflection. BMC Neurosci. 2012; 13: 52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoffstaedter F, Grefkes C, Zilles K, et al.: The “what” and “when” of self-Initiated movements. Cereb Cortex. 2013; 23(3): 520–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHooker CI, Verosky SC, Germine LT, et al.: Mentalizing about emotion and its relationship to empathy. Soc Cogn Affect Neurosci. 2008; 3(3): 204–217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoshi E: Functional specialization within the dorsolateral prefrontal cortex: a review of anatomical and physiological studies of non-human primates. Neurosci Res. 2006; 54(2): 73–84. PubMed Abstract \| Publisher Full Text\n\nHosseini SM, Rostami M, Yomogida Y, et al.: Aging and decision making under uncertainty: behavioral and neural evidence for the preservation of decision making in the absence of learning in old age. Neuroimage. 2010; 52(4): 1514–1520. PubMed Abstract \| Publisher Full Text\n\nHuettel SA, Misiurek J: Modulation of prefrontal cortex activity by information toward a decision rule. Neuroreport. 2004; 15(12): 18883–1886. PubMed Abstract\n\nJakobs O, Wang LE, Dafotakis M, et al.: Effects of timing and movement uncertainty implicate the temporo-parietal junction in the prediction of forthcoming motor actions. Neuroimage. 2009; 47(2): 667–677. PubMed Abstract \| Publisher Full Text\n\nJocham G, Hunt LT, Near J, et al.: A mechanism for value-guided choice based on the excitation-inhibition balance in prefrontal cortex. Nat Neurosci. 2012; 15(7): 960–961. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKahnt T, Heinzle J, Park SO, et al.: Decoding different roles for vmPFC and dlPFC in multi-attribute decision making. Neuroimage. 2011; 56(2): 709–715. PubMed Abstract \| Publisher Full Text\n\nKellermann TS, Sternkopf MA, Schneider F, et al.: Modulating the processing of emotional stimuli by cognitive demand. Soc Cogn Affect Neurosci. 2012; 7(3): 263–273. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKillgore WD, Yurgelun-Todd D: Activation of the amygdala and anterior cingulate during nonconscious processing of sad versus happy faces. Neuroimage. 2004; 21(4): 1215–1223. PubMed Abstract \| Publisher Full Text\n\nKnorr U, Weder B, Kleinschmidt A, et al.: Identification of task-specific rCBF changes in individual subjects: validation and application for PET. J Comput Assist Tomogr. 1993; 17(4): 517–528. PubMed Abstract\n\nKrain AL, Wilson AM, Arbuckle R, et al.: Distinct neural mechanisms of risk and ambiguity: a meta-analysis of decision-making. Neuroimage. 2006; 32(1): 477–484. PubMed Abstract \| Publisher Full Text\n\nKucyi A, Hodaie M, Davis KD: Lateralization in intrinsic functional connectivity of the temporoparietal junction with salience- and attention-related brain networks. J Neurophysiol. 2012; 108(12): 3382–3392. PubMed Abstract \| Publisher Full Text\n\nLaird AR, Eickhoff SB, Fox PM, et al.: The BrainMap strategy for standardization, sharing, and meta-analysis of neuroimaging data. BMC Res Notes. 2011; 4: 349. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaird AR, Eickhoff SB, Kurth F, et al.: ALE Meta-Analysis Workflows Via the Brainmap Database: Progress Towards A Probabilistic Functional Brain Atlas. Front Neuroinform. 2009; 3: 23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLamm C, Batson CD, Decety J: The neural substrate of human empathy: effects of perspective-taking and cognitive appraisal. J Cogn Neurosci. 2007; 19(1): 42–58. PubMed Abstract \| Publisher Full Text\n\nLawrence EJ, Shaw P, Giampietro VP, et al.: The role of ‘shared representations’ in social perception and empathy: an fMRI study. Neuroimage. 2006; 29(4): 1173–1184. PubMed Abstract \| Publisher Full Text\n\nLi W, Zinbarg RE, Boehm SG, et al.: Neural and behavioral evidence for affective priming from unconsciously perceived emotional facial expressions and the influence of trait anxiety. J Cogn Neurosci. 2008; 20(1): 95–107. PubMed Abstract \| Publisher Full Text\n\nLindenberg R, Fangereau H, Seitz RJ: “Broca’s area” as a collective term? Brain Lang. 2007; 102(1): 22–29. PubMed Abstract \| Publisher Full Text\n\nLindenberg R, Uhlig M, Scherfeld D, et al.: Communication with emblematic gestures: shared and distinct neural correlates of expression and reception. Hum Brain Mapp. 2012; 33(4): 812–823. PubMed Abstract \| Publisher Full Text\n\nLitt A, Plassmann A, Shav B, et al.: Dissociating valuation and saliency signals during decision-making. Cereb Cortex. 2010; 21(1): 95–102. PubMed Abstract \| Publisher Full Text\n\nMériau K, Wartenburger I, Kazzer P, et al.: A neural network reflecting individual differences in cognitive processing of emotions during perceptual decision making. Neuroimage. 2006; 33(3): 1016–1027. PubMed Abstract \| Publisher Full Text\n\nMilham MP, Banich MT, et al.: Competition for priority in processing increases prefrontal cortex’s involvement in top-down control: an event-related fMRI study of the stroop task. Brain Res Cogn Brain Res. 2003; 17(2): 212–222. PubMed Abstract \| Publisher Full Text\n\nMiller EK, Cohen JD: An integrative theory of prefrontal cortex function. Annu Rev Neurosci. 2001; 24: 167–202. PubMed Abstract \| Publisher Full Text\n\nMorewedge CK, Kahneman D: Associative processes in intuitive judgment. Trends Cogn Sci. 2010; 14(10): 435–440. PubMed Abstract \| Publisher Full Text\n\nMoskowitz GB, Stone J, Childs A: Implicit stereotyping and medical decisions: unconscious stereotype activation in practitioners’ thoughts about African Americans. Am J Public Health. 2012; 10(5): 996–1001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMurphy ST, Zajonc RB: Affect, cognition, and awareness: affective priming with optimal and suboptimal stimulus exposures. J Pers Soc Psychol. 1993; 64(5): 723–739. PubMed Abstract \| Publisher Full Text\n\nNomura M, Ohira H, Haneda K, et al.: Functional association of the amygdala and ventral prefrontal cortex during cognitive evaluation of facial expressions primed by masked angry: an event-related fMRI study. Neuroimage. 2004; 21(1): 352–363. PubMed Abstract \| Publisher Full Text\n\nOldfield RC: The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971; 9(1): 97–113. PubMed Abstract \| Publisher Full Text\n\nOpialla S, Lutz J, Scherpiet S, et al.: Neural circuits of emotion regulation: a comparison of mindfulness-based and cognitive reappraisal strategies. Eur Arch Psychiatry Clin Neurosci. 2014. PubMed Abstract \| Publisher Full Text\n\nOpitz B, Mecklinger A, Friederici AD: Functional asymmetry of human prefrontal cortex: encoding and retrieval of verbally and nonverbally coded information. Learn Mem. 2000; 7(2): 85–96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPlassmann H, O’Doherty J, Rangel A: Orbitofrontal cortex encodes willingness to pay in everyday economic transactions. J Neurosci. 2007; 27(37): 9984–9988. PubMed Abstract \| Publisher Full Text\n\nProchnow D, Brunheim S, Kossack H, et al.: Statistical data of subareas of the dorsolateral frontal cortex in socially relevant decisions based on masked affect expressions. figshare. 2014a. Data Source\n\nProchnow D, Höing B, Kleiser R, et al.: The neural correlates of affect reading: an fMRI study on faces and gestures. Behav Brain Res. 2013a; 237: 270–277. PubMed Abstract \| Publisher Full Text\n\nProchnow D, Kossack H, Brunheim S, et al.: Processing of subliminal facial expressions of emotion: a behavioral and fMRI study. Soc Neurosci. 2013b; 8(5): 448–461. PubMed Abstract \| Publisher Full Text\n\nProchnow D, Steinhäuser L, Brunheim S, et al.: Differential emotional state reasoning in young and older adults: Evidence from behavioral and neuroimaging data. J Neurol Psychol. 2014b; 2(1): 1–8. Reference Source\n\nReniers R, Corcoran R, Völlm BA, et al.: Moral decision-making, ToM empathy and the default mode network. Biol Psychol. 2012; 90(3): 202–210. PubMed Abstract \| Publisher Full Text\n\nRizzolatti G, Craighero L: The mirror neuron system. Annu Rev Neurosci. 2004; 27: 169–192. PubMed Abstract \| Publisher Full Text\n\nRohr M, Degner J, Wentura D: Masked emotional priming beyond global valence activations. Cogn Emot. 2012; 26(2): 224–244. PubMed Abstract \| Publisher Full Text\n\nRottschy C, Langner R, Dogan I, et al.: Modelling neural correlates of working memory: a coordinate-based meta-analysis. Neuroimage. 2012; 60(1): 830–846. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRuys KI, Aarts H: I didn’t mean to hurt you! Unconscious origins of experienced self-agency over others’ emotions. Emotion. 2012; 12(1): 132–141. PubMed Abstract \| Publisher Full Text\n\nSanfey AG: Social decision-making: insights from game theory and neuroscience. Science. 2007; 318(5850): 598–602. PubMed Abstract \| Publisher Full Text\n\nSchulte-Rüther M, Markowitsch HJ, Fink GR, et al.: Mirror neuron and theory of mind mechanisms involved in face-to-face interactions: a functional magnetic resonance imaging approach to empathy. J Cogn Neurosci. 2007; 19(8): 1354–1372. PubMed Abstract \| Publisher Full Text\n\nSeitz RJ, Franz M, Azari NP: Value judgments and self-control of action: the role of the medial frontal cortex. Brain Res Rev. 2009; 60(2): 368–378. PubMed Abstract \| Publisher Full Text\n\nSeitz RJ, Nickel J, Azari NP: Functional modularity of the medial prefrontal cortex: involvement in human empathy. Neuropsychology. 2006; 20(6): 743–751. PubMed Abstract \| Publisher Full Text\n\nSeitz RJ, Scherfeld D, Friederichs S, et al.: Valuating other people’s emotional face expression: a combined functional magnetic resonance imaging and electroencephalography study. Neuroscience. 2008; 152(3): 713–722. PubMed Abstract \| Publisher Full Text\n\nShamay-Tsoory SC, Aharon-Peretz J, Perry D: Two systems for empathy: a double dissociation between emotional and cognitive empathy in inferior frontal gyrus versus ventromedial prefrontal lesions. Brain. 2009; 132(Pt 3): 617–627. PubMed Abstract \| Publisher Full Text\n\nShen XB, Wu Q, Fu XL: Effects of the duration of expressions on the recognition of microexpressions. J Zhejiang Univ Sci B. 2012; 13(3): 221–230. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShibasaki H, Hallett M: What is the Bereitschaftspotential? Clin Neurophysiol. 2006; 117(11): 2341–2356. PubMed Abstract \| Publisher Full Text\n\nSilvers JA, Weber J, Wager TD, et al.: Bad and worse: neural systems underlying reappraisal of high- and low-intensity negative emotions. Soc Cogn Affect Neurosci. 2014. PubMed Abstract \| Publisher Full Text\n\nSmith DV, Hayden BY, Truong TK, et al.: Distinct value signals in anterior and posterior ventromedial prefrontal cortex. J Neurosci. 2010; 30(7): 2490–2495. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSokol-Hessner P, Hutcherson C, Hare T, et al.: Decision value computation in DLPFC and VMPFC adjusts to the available decision time. Eur J Neurosci. 2012; 35(7): 1065–1074. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSuslow T, Kugel H, Ohrmann P, et al.: Neural correlates of affective priming effects based on masked facial emotion: an fMRI study. Psychiatry Res. 2013; 211(3): 239–245. PubMed Abstract \| Publisher Full Text\n\nSvenson O: Decision Making and the Search for Fundamental Psychological Regularities: What Can Be Learned from a Process Perspective? Organ Behav Hum Decis Process. 1996; 65(3): 252–267. Publisher Full Text\n\nSweeny TD, Grabowecky M, Suzuki S, et al.: Long-lasting effects of subliminal affective priming from facial expressions. Conscious Cogn. 2009; 18(4): 929–938. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTalairach J, Tournoux P: Co-planar stereotaxic atlas of the human brain. Stuttgart: Thieme. 1988. Reference Source\n\nThirioux B, Mercier MR, Blanke O, et al.: The cognitive and neural time course of empathy and sympathy: an electrical neuroimaging study on self-other interaction. Neuroscience. 2014; 267: 286–306. PubMed Abstract \| Publisher Full Text\n\nTurkeltaub PE, Eickhoff SB, Laird AR, et al.: Minimizing within-experiment and within-group effects in Activation Likelihood Estimation meta-analyses. Hum Brain Mapp. 2012; 33(1): 1–13. PubMed Abstract \| Publisher Full Text\n\nvan ’t Wout M, Kahn RS, Sanfey AG, et al.: Repetitive transcranial magnetic stimulation over the right dorsolateral prefrontal cortex affects strategic decision-making. Neuroreport. 2005; 16(16): 1849–1852. PubMed Abstract\n\nWagner AD, Maril A, Bjork RA, et al.: Prefrontal contributions to executive control: fMRI evidence for functional distinctions within lateral Prefrontal cortex. Neuroimage. 2001; 14(6): 1337–1347. PubMed Abstract \| Publisher Full Text\n\nWalter H, Adenzato M, Ciaramidaro A, et al.: Understanding intentions in social interaction: the role of the anterior paracingulate cortex. J Cogn Neurosci. 2004; 16(10): 1854–1863. PubMed Abstract \| Publisher Full Text\n\nWatson D, Clark LA, Tellegren A: Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988; 54(6): 1063–1070. PubMed Abstract \| Publisher Full Text\n\nWinkielman P, Berridge KC, Wilbarger JL: Unconscious affective reactions to masked happy versus angry faces influence consumption behavior and judgements of value. Pers Soc Psychol Bull. 2005; 31(1): 121–135. PubMed Abstract \| Publisher Full Text\n\nWorsley KJ, Evans AC, Marrett S, et al.: A three-dimensional statistical analysis for CBF activation studies in human brain. J Cereb Blood Flow Metab. 1992; 12(6): 900–918. PubMed Abstract \| Publisher Full Text\n\nWunderlich K, Rangel A, O’Doherty JP: Economic choices can be made using only stimulus values. Proc Natl Acad Sci U S A. 2010; 107(34): 15005–15010. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXi C, Zhu Y, Niu C, et al.: Contributions of subregions of the prefrontal cortex to the theory of mind and decision making. Behav Brain Res. 2011; 221(2): 587–593. PubMed Abstract \| Publisher Full Text\n\nZelazo PD, Muller U: Executive function in typical and atypical development. In Goswami, U. Handbook of childhood cognitive development. Oxford: Blackwell. 2002; 445–489. Publisher Full Text" }	[ { "id": "6218", "date": "25 Sep 2014", "name": "Motoaki Sugiura", "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 authors tried to demonstrate that anterior and posterior subregions of the DLFC are involved in different phases (i.e., pre-decision and decision per se, respectively) of socially relevant decisions based on subliminal emotional information. The addressed issue is important in related fields and the approach is unique. However, I consider that the following methodological issues require clarification before the evaluation of the work. Major points:Is it reasonable to use the contrasts “masked facial expressions vs. baseline” and “decisions based on masked affect expressions vs. motor control” to isolate the DLPF subregions for the socially relevant decisions based on subliminal emotional information? Activation identified in these contrasts may be affected by variety of perceptual and cognitive processes. How can the authors’ assumption that activation in these contrasts reflected only the decision making process be justified?Also, in the MACM approach, authors did not make the selection of the data based on the task or contrast. How, then, can the authors’ argue the identified connectivity specifically reflected the decision making process?The authors used 1) “p < 0.005 and minimal cluster size of 405 voxels” for the fMRI analysis, 2) “a family-wise error (FWE) corrected p-threshold of p < 0.05” for the MACM/ALE analysis, and 3) some Monte-Crlo-like simulation described after the explanation of 2. I don’t understand the reason for using both 2 and 3 for the MACM/ALE analysis, and also see how adaptation of these three different thresholds/criteria across three cases was justified.In the task, how can activation for the pre-decision and decision phases be separated despite the fixed close interval of the two phases?Region of interest (ROI) analysis, paragraph 2 - “We conducted pairwise t-tests to…”: Why was this uncommon analysis necessary? Details:Introduction, paragraph 3, last sentence - “Recent evidence, however, suggests that gaming and ToM scenarios are based at least partly on different neural circuits”: I don’t get the nuance; was there an assumption that the gaming and ToM perfectly share the neural substrate?FMRT data analysis, paragraph 3 - “To ensure comparability…”: I don’t understand what is meant by “comparability” (i.e., with what?) or what has exactly been done here. In addition, authors should be clear about the “template” for the normalization; “Talairach space” sometimes means just a 3D brain space.Functional connectivity analyses, paragraph 1 - “above-mentioned criteria”: which criteria do you mean?", "responses": [ { "c_id": "1291", "date": "20 Apr 2015", "name": "Denise Prochnow", "role": "Reader Comment", "response": "We thank you for reviewing our manuscript. Based on his thoughtful and constructive criticism, we revised our manuscript carefully and made the following changes. Please find below our point by point answers to the important comments you made.You wrote “Is it reasonable to use the contrasts “masked facial expressions vs. baseline” and “decisions based on masked affect expressions vs. motor control” to isolate the DLPF subregions for the socially relevant decisions based on subliminal emotional information? Activation identified in these contrasts may be affected by variety of perceptual and cognitive processes. How can the authors’ assumption that activation in these contrasts reflected only the decision making process be justified?”We are thankful for this comment and now explain in more detail in the methods section the rationale behind our paradigm (p.5). In the first part of the visual stimulation, corresponding to the pre-decision phase, masked faces or scrambled faces were presented. They had the identical visual input. However, when the subjects were exposed to the faces they were tuned to do a subsequent decision, since they were instructed to choose one out of two emotional adjectives presented to them after the faces. Thus, beyond brain areas related to emotional face perception also brain areas related to decision making should become engaged. In the decision phase the subjects had to indicate the appropriate word describing the emotion seen previously in the masked face by a button press. In the motor control condition, the subject had to press the right or left button according to the visually presented instruction. Thus, visual processing of the word list, the selection of the active finger and the button press were identical in both conditions. However, the decision which word described the previously face was only present in the decision condition. You wrote “Also, in the MACM approach, authors did not make the selection of the data based on the task or contrast. How, then, can the authors’ argue the identified connectivity specifically reflected the decision making process?”We thank you for this point and added some information to the results section (p.9). In the MACM approach all experiments in the BrainMap database reporting activation of the seed regions were identified irrespective of their behavioral classification. From these data a quantitative meta-analysis was performed that tested the presence of activation clusters in the imaging data of the database resulting in a statistical map of co-activations with the seed regions. We agree with the reviewer that this analysis reflects the functional connectivity of the DLPF but does not the decision making process. You wrote “The authors used 1) “p < 0.005 and minimal cluster size of 405 voxels” for the fMRI analysis, 2) “a family-wise error (FWE) corrected p-threshold of p < 0.05” for the MACM/ALE analysis, and 3) some Monte-Crlo-like simulation described after the explanation of 2. I don’t understand the reason for using both 2 and 3 for the MACM/ALE analysis, and also see how adaptation of these three different thresholds/criteria across three cases was justified.”We also thank you for your comment on our choice of statistical thresholds. The fMRI experiment was based on 18 healthy subjects. This statistical thresholding we used corresponds to established criteria as we outline in our paper (please refer also to Prochnow et al., 2013b). The MACM/ALE analysis was based on data of about 7500 imaging studies. Therefore, far more rigid threshold criteria were appropriate and used as done also in the previous work referred to in our manuscript. This was also the case for the calculation of the difference maps. You wrote “In the task, how can activation for the pre-decision and decision phases be separated despite the fixed close interval of the two phases?”We thank you for mentioning this important aspect and modified the description of the fMRI data analysis (p.6) based on your comment. The separation of the two phases was possible by the event-related character of the scanning procedure which allowed performing two time-locked events with a temporal separation of 2.600 ms on average (jittered time interval) while the scanning repetition time was 2000 ms. You wrote “Region of interest (ROI) analysis, paragraph 2 - “We conducted pairwise t-tests to…”: Why was this uncommon analysis necessary?”As we used the extracted activation estimates to explore the spatial separation and correlation with the behavioral data of these two regions of interest, we tested also if the areas would reveal different degrees of activation in the pre-decision phase and the decision phase. As we state in the paper this small volume analysis excluded such a difference. You wrote “Introduction, paragraph 3, last sentence - “Recent evidence, however, suggests that gaming and ToM scenarios are based at least partly on different neural circuits”: I don’t get the nuance; was there an assumption that the gaming and ToM perfectly share the neural substrate?”We thank you for drawing our attention to this orthographical error. We referred to gambling, not gaming scenarios. You wrote “FMRT data analysis, paragraph 3 - “To ensure comparability…”: I don’t understand what is meant by “comparability” (i.e., with what?) or what has exactly been done here. In addition, authors should be clear about the “template” for the normalization; “Talairach space” sometimes means just a 3D brain space.”“Functional connectivity analyses, paragraph 1 - “above-mentioned criteria”: which criteria do you mean?”We changed the sentences mentioned in order to be more precise (p.7)." } ] }, { "id": "7367", "date": "03 Mar 2015", "name": "Gonzalo G de Polavieja", "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 authors use fMRI experiments to argue that the dorsolateral frontal cortex (DLFC) responds in decision-making to masked affect human faces.The reference images for the pre-decision phase are scrambled images. Scrambled images have a number of obvious features as they are not (a), faces, (b) smooth and (c) they have local contrast different to faces. So it is not obvious to me why the experimental condition, respect to this baseline, corresponds to affect faces, and not more generally to (a) some structure of faces, (b) smooth figures or (c) particular contrast structures. Why wouldn't you need to use additional controls to eliminate these options? (i.e. normal faces or objects)Related to this, you mention that you eliminate the 26% of subjects that suspect the presence of emotional faces and that 78% reported flickering. Although I didn't understand whether this flickering came from real flickering or not, I could not find in the text whether some subjects could report seeing the transition from scramble image to face. Some of the reported data (mean age, % subjects that suspect emotional faces, etc) are identical to Prochnow et al 2013b, but I cannot find mention in the present ms that the data collected is the same than in Prochnow et al 2013b or that the subjects are the same. Is this the case? Please, clarify. Data presented are comparison of means, but the data obtained is distributions. Did you find whether other parameters of the distributions for other brain regions were significant (median, mode, variance). Is there individual variability? As you made many preliminary tests on subjects (for example, empathy), it would be very interesting to correlate these results with brain activations.", "responses": [ { "c_id": "1290", "date": "20 Apr 2015", "name": "Denise Prochnow", "role": "Reader Comment", "response": "We thank you for reviewing our manuscript and for the constructive comments he made. Based on these, we revised our manuscript carefully in order to improve it. Please find below our point by point answers to your constructive criticism.You wrote “The reference images for the pre-decision phase are scrambled images. Scrambled images have a number of obvious features as they are not (a), faces, (b) smooth and (c) they have local contrast different to faces. So it is not obvious to me why the experimental condition, respect to this baseline, corresponds to affect faces, and not more generally to (a) some structure of faces, (b) smooth figures or (c) particular contrast structures. Why wouldn't you need to use additional controls to eliminate these options? (i.e. normal faces or objects)“ We thank you for this important comment and would like to point out that scrambled images were used to subtract activation related to visual processing. As you correctly mention, scrambled images differ from facial images in smoothness and local contrast.They are identical in overall contrast and color but do not convey any face information or meaning. Thus, we consider they are suited to maximize the cognitive comparison to faces. You wrote “Related to this, you mention that you eliminate the 26% of subjects that suspect the presence of emotional faces and that 78% reported flickering. Although I didn't understand whether this flickering came from real flickering or not, I could not find in the text whether some subjects could report seeing the transition from scramble image to face.” We thank you for this important point as well. Indeed, the majority of subjects reported having noticed a flickering throughout the experiment. As explained now in detail on page 5, this perception was due to the rapid succession of the emotional expressions presented below the threshold of subjective awareness and the clearly visible neutral masking expression. Please note that the subjects did not report a break but just a flicker which was due to the fact that there were only slight details changing in the face as the position of the eyebrows and / or mouth. For comparison, the transition between the facial stimuli and the scrambled image was clearly visible and was not perceived as a flickering. You wrote “Some of the reported data (mean age, % subjects that suspect emotional faces, etc) are identical to Prochnow et al 2013b, but I cannot find mention in the present ms that the data collected is the same than in Prochnow et al 2013b or that the subjects are the same. Is this the case? Please, clarify.” We are also thankful for this important information. We now provide more detailed information on the relationship between both articles and explain that both are based on the same data set albeit studying entirely different aspects. Prochnow et al., 2013b compared brain activation patterns between facial expressions of emotion which were either clearly visible or presented below the threshold of subjective awareness and superimposed by a neutral masking facial expression. On the contrary, concerning the experimental paradigm the current article focuses on the decision aspect of the overall paradigm since the subjects were instructed to decide which of two subsequently presented emotional adjectives best described the mood observed in the previously seen face. Furthermore, this paper addresses the novel issue of functional connectivity of the activated lateral prefrontal cortex. You wrote “Data presented are comparison of means, but the data obtained is distributions. Did you find whether other parameters of the distributions for other brain regions were significant (median, mode, variance).” We have analyzed our fMRI data with established image analysis tools of BRAIN VOYAGER which computes t-tests on a voxel by voxel basis. A similar approach is utilized by MCAM ALE. Thus, we adhered to standards that have been developed for functional imaging data which allow to compared studies of different laboratories, since they are generally used. You wrote “Is there individual variability? As you made many preliminary tests on subjects (for example, empathy), it would be very interesting to correlate these results with brain activations.” We thank you for mentioning this interesting aspect. We fully agree that this is an important issue. For that reason we extracted the local values in two areas of interest in the dorsolateral prefrontal cortex, our area of interest, to address this issue. Unfortunately, a study on the individual variability in an omnibus approach would be a study in its own right and go far beyond the present communication." } ] } ]	1	https://f1000research.com/articles/3-212
https://f1000research.com/articles/4-196/v1	07 Jul 15	{ "type": "Research Article", "title": "Journal ratings as predictors of articles quality in Arts, Humanities and Social Sciences: an analysis based on the Italian Research Evaluation Exercise", "authors": [ "Andrea Bonaccorsi", "Tindaro Cicero", "Antonio Ferrara", "Marco Malgarini", "Andrea Bonaccorsi", "Tindaro Cicero", "Antonio Ferrara" ], "abstract": "The aim of this paper is to understand whether the probability of receiving positive peer reviews is influenced by having published in an independently assessed, high-ranking journal: we eventually interpret a positive relationship among peer evaluation and journal ranking as evidence that journal ratings are good predictors of article quality. The analysis is based on a large dataset of over 11,500 research articles published in Italy in the period 2004-2010 in the areas of architecture, arts and humanities, history and philosophy, law, sociology and political sciences. These articles received a score by a large number of externally appointed referees in the context of the Italian research assessment exercise (VQR); similarly, journal scores were assigned in a panel-based independent assessment, which involved all academic journals in which Italian scholars have published, carried out under a different procedure. The score of an article is compared with that of the journal it is published in: more specifically, we first estimate an ordered probit model, assessing the probability for a paper of receiving a higher score, the higher the score of the journal; in a second step, we concentrate on the top papers, evaluating the probability of a paper receiving an excellent score having been published in a top-rated journal. In doing so, we control for a number of characteristics of the paper and its author, including the language of publication, the scientific field and its size, the age of the author and the academic status. We add to the literature on journal classification by providing for the first time a large scale test of the robustness of expert-based classification.", "keywords": [ "Journal rankings", "ANVUR", "Arts", "Humanities and Social Sciences", "Journal ratings", "Article Quality" ], "content": "Introduction\n\nThere is a large degree of agreement on the notion that research assessment in humanities and social sciences (HSS) is made more complex by a variety of factors. First, in these fields the structure of academic publication is largely different, with a large weight assigned to books and monographs, and to production in national language (Finkenstaedt, 1990). Consequently, the bibliometric approach is considered to be of limited usefulness (Nederhof et al., 1989), not only because journals are a small fraction of total production and indexed journals are a tiny fraction of the population of journals in HSS, but also because the meaning of citations is different (Frost, 1979). Third, and even more challenging, there is evidence that the number of research quality criteria is larger in HSS than in other fields, and also that there is less agreement on these criteria (Hemlin, 1996; Hemlin & Gustafsson, 1996; Hug et al., 2013; Hug et al., 2014; Ochsner et al., 2012; Ochsner et al., 2013).\n\nFaced with these challenges, the state of the art of the assessment of research in HSS at international level has followed several directions. On the one hand, it is agreed that peer review is still the most important evaluation methodology, so large efforts are made in making it more sophisticated, methodologically controlled, based on sound principles of evaluation methodology in social sciences, and free from unwanted biases, distortions and unexpected side effects. Under this agenda, issues such as the notion of originality, unorthodox science, or interdisciplinarity are under examination (Guetzkow et al., 2004; Hammarfelt, 2011). On the other hand, there are many efforts to classify and evaluate non-indexed journals (mainly in national languages), as one of the main vehicles for academic communication. An additional line of work refers to the classification of books and publishers.\n\nThis paper reports on a large experiment in the classification of journals in HSS carried out in Italy in the 2012–2014 period for the National Scientific Habilitation (Abilitazione Scientifica Nazionale; ASN). The exercise was based upon a mandatory provision in the law to rate all journals, in order to calculate the overall academic production of all candidates to the national procedure to become associate professor or full professor. This exercise asked the National Agency for the Evaluation of Universities and Research Institutes (ANVUR) to evaluate all journals in which at least one Italian scholar published at least one paper in the 2002–2012 period, for a total of more than 60,000 titles.\n\nWhile the rating of journals has been followed in several national contexts, it is only in the Italian exercise that there is the opportunity to carry out a controlled experiment in order to test the robustness of journal classification. In fact, we have two independent evaluations carried out on the same set of journals. On the one hand, a panel of experts classified all journals as academic and non-academic (i.e. popular, professional, technical, cultural and political etc.), and rated the subset of academic journals in A-rated and non A-rated. The rating exercise was done on the basis of the reputation, esteem, diffusion and impact of journals, that is, on a qualitative, expert-based, reputational basis. On the other hand, we also have the rating of individual articles published in those journals, which have been done by a large number of individual referees (not panels) and summarized with a consensus agreement approach by expert panels, who however acted independently from the other panels, and without exchange of information. This peculiarity of the Italian context and the time sequence of events creates a favorable condition for carrying out a controlled experiment.\n\nThis paper extends to all HSS, with the exception of economics and business, the analysis initiated by Ferrara & Bonaccorsi (2015) on journals in the area of philosophy and history. In the following, we first introduce the database used for the analysis and hence we test for the influence of the journal class on the article score. Some consideration on the results obtained will conclude the paper.\n\n\nMethods\n\nThe paper is based on a dataset including data on all the journal articles submitted for evaluation by Italian scholars in the disciplinary areas of architecture, arts and humanities, history and philosophy, law and sociology and political science. Submissions for evaluation took place within the framework of VQR 2004–10, Italy’s national research assessment exercise involving all professors and researchers affiliated to the Italian universities and Public Research Organizations (PROs) as of November 2011. According to adopted rules, research evaluation in HSS was entirely based on peer review; research quality was assessed against the criteria of relevance, meaning contribution to the advancement of the state of art in the field, also in terms of adequacy, efficacy, timeliness and duration of impacts; originality and innovation, meaning contribution to the creation of new knowledge in the field; internationalization, meaning position in the international research landscape. Evaluation has been conducted by five Groups of Evaluation Experts (GEV in the Italian acronym), one for each area in HSS (Architecture; Arts and Humanities; History and Philosophy; Law; Sociology and political sciences); reviewers were instructed by GEV to evaluate articles only on the basis or their merit, regardless of the journal in which they are published in and of the language of publication. Each article had a possible rating of Excellent (A), Good (B), Fair (C) or Limited (D); to each class corresponded a score ranging from 1 (for articles A-rated) to zero (for articles deemed as limited). Negative scores were also assigned in case the article was deemed as non-academic (-1) or for plagiarism or fraud (-2, see Ancaiani et al., 2015 for details). Limited to the human and social sciences, a substantial fraction of articles – namely, 6,701 out of 11,660 (Table 1) – appeared on journals deemed as ‘A-class’ according to the procedure of ASN, intended to select the best researchers for the ranks of associate and full professors. Those journals, according to the relevant Ministerial Decree (No. 76/2012), were those ‘internationally recognized as excellent because of the rigor of their procedures of peer review and because of their diffusion among, esteem by, and impact on, the scholarly community of a field, as indicated also by their presence in the major national and international databases’ (our translation). Most of the remaining articles appeared on journals deemed as ‘academic’ for the purposes of the ASN, while a minority were published in journals that remained ‘uncategorized’. The main feature of the dataset, thus, is that it allows the comparison between the evaluations of journals and individual articles.\n\nA preliminary analysis shows that there is a relationship between the evaluation of individual articles and that of journals where the article is published (Table 2). The non-parametric test for categorical data (Pearson χ2) is statistically significant at 1% (All the statistical analyses have been performed using the software STATA ver. 13 (http://www.stata.com/stata13/)), showing that the two distributions are not independent and hence the two ratings are mutually related. In the following, we will analyze more thoroughly this relationship, also controlling for a number of author-level and article-level variables.\n\nPearson χ2=630.9; p-value=0.000\n\n\nThe influence of journal classification on the article score\n\nWe assume that the probability for an article i, published in the journal j, of receiving a score equal to x ∈ {-2; 1} is influenced by the class assigned to the journal, once controlling for a number of characteristics of the article:\n\nP (Scorei,j = x) = F(Journal classi,j, Paper characteristicsi,j) (1)\n\nAmong the controls, we consider the language of publication (Italian or not) and the age (distinguishing among 3 age classes, less than 40 years, between 41 and 55 years and more than 55 years), scientific sector of activity (Scientific Areas 8, 10, 11, 12, 14), academic status (full professor; associate professor; researcher; other) and gender of the researcher. We also add the consideration of two binary variables controlling for the existence of international co-author(s) and for the nationality of the referees (allowing for the possibility of international referees). We finally add a variable taking into account the size of the scientific area of the author. The model is estimated as an ordered probit, an extension of the standard binary probit model, used when the dependent variable takes the form of a ranked and multiple discrete variable, considering alternatively the whole sample or each scientific area; in the first case, we also control for possible area-specific effects. In order to avoid the “dummy trap”, we normalize with respect to articles written in Italian with no international co-author, evaluated by an Italian reviewer, presented by a female researcher in sociology and political science, aged less than 40: i.e. the statistical significance, sign and magnitude of estimated parameters are to be interpreted as differentials with respect to this control group. The total number of available observations amounts to 11,660 varying from a minimum of 918 in architecture to a maximum of 3,838 in law (Table 3).\n\n* p<0.01, p<0.05, * p<0.1\n\nThe main result is that both at the aggregate level and in each scientific area the article score is higher as the journal ranking gets better: in other words, the probability of receiving a high score grows if the article is published in a high-ranking journal according to the evaluation of the ASN’s experts. As for the control variables, we confirm most of the results already emerged in a previous paper on the same data (Cicero et al., 2014), namely, that article scores are higher for papers not written in Italian, with international co-authors, published by an under-40, male full or associate professor. Moreover, we also find that at aggregate level and in most areas an international reviewer and a lower number of professors in the specific scientific sector (SSD) are associated with an higher article score: a possible interpretation of the first result is that the expert groups responsible for the evaluation (GEV) mostly assign to international reviewers more internationalized papers, that are considered to have an higher probability of receiving a high score, given also that the level of internationalization was one of the evaluation criteria according to VQR rules (see again Ancaiani et al., 2015). As for a negative relationship among area size and article score, this result emerged already in Ferrara & Bonaccorsi (2015) for the scientific fields in history and philosophy and is now extended to all HSS: a possible interpretation is that small fields may be favored by a “proximity bias” among authors and reviewers, thus resulting, ceteris paribus, in higher article scores.\n\nAs a final check, we concentrate on the probability of receiving an excellent score and relate it to the fact that the article is published in a top, A-Class journal, once controlling for the same variables considered in model 1:\n\nP (Scorei,j = “E”) = F(Journal classi,j = “A”, Paper characteristicsi,j) (2)\n\nIn (2), F is the logistic function and the model is estimated as a logit, a class of models allowing to predict the binary response based on the specified predictors. A desirable feature of the logit model is that the regression coefficients may easily be transformed in odds ratio, expressing the change in the odds of the occurrence under scrutiny (in our case, the odds for a paper of receiving an ‘Excellent’ evaluation) due to a small change of a given predictor: in our case, we are particularly interested in the odds associated with the classification of a journal as a top, Class A journal. Estimation results for both the aggregate sample and each scientific area are presented in Table 4.\n\n* p<0.01, p<0.05, * p<0.1\n\nAccording to logit estimations, the probability of receiving an excellent evaluation is positively affected by the journal in which the paper is published in: more specifically, publishing in a class A journal almost doubles the probability of receiving an excellent evaluation. Looking at the results in each scientific area, the odds of receiving an excellent evaluation are more than doubled by the publication in a Class A journal in architecture and history and philosophy; the effect is somewhat lower, but still highly significant, in law, and arts and humanities, while disappearing in sociology and political sciences. Logit estimation also broadly confirms the results already emerging from the ordered probit model: the odds of receiving an excellent evaluation are increased by publishing in a foreign language, with an international co-author (albeit only in law and architecture) and when the submitting author is 40 years old or younger, an associate or full professor and a male. Gender effect is in fact significant at the aggregate level and in architecture and humanities, but not in the remaining areas. Also in this case, having an international reviewer and publishing in a SSD characterized by a lower number of full professors helps in obtaining an excellent evaluation.\n\n\nConclusions\n\nUsing a very large dataset of journal articles published in HSS, the paper proves that independent classifications of journals may be considered as good predictors of the score assigned to individual articles. More specifically, we find that, after controlling for a number of articles’ characteristics, the probability of receiving a better score grows with the quality profile of the journal the article is published in; moreover, the probability of receiving an excellent score almost doubles when the paper is published in a top, A-Class journal. The findings hold both at the aggregate level and for each specific sub-area considered in the analysis. While peer review has to remain the main evaluation methodology, our results indicate that journal classifications may be considered as a useful supporting tool in large evaluation exercise, since it may provide reviewers with valuable information apt to support expert evaluation.\n\n\nData availability\n\nThe authors hold the view that it is important to allow the free access to data used in the article in order to enable others to replicate the study. However, information used in the article were gathered by the national agency responsible for evaluation of the University and research system in Italy (ANVUR), in the framework of this VQR exercise. In this context, ANVUR asked Italian professors to provide access to their publications, assuming the commitment not to disclose to the public, unless in an aggregate form, any data concerning the publications submitted for the evaluation and, most importantly, the results of the evaluation itself. This is deemed as necessary in order to guarantee the full anonymity of evaluations performed on each individual publications and on each Italian professor. For this reason, as the public agency in charge of evaluating research of Italian universities, ANVUR does not allow to make information about individual evaluations available to the general public.\n\nThe information used to generate data in this article concerning journal classification is available to the public at the following URL: http://www.anvur.it/index.php?option=com_content&view=article&id=254&Itemid=315&lang=it.", "appendix": "Author contributions\n\n\n\nThe paper is the result of a common effort of the authors. However, Andrea Bonaccorsi can be credited for the “Introduction” and the “Conclusions”, while Antonio Ferrara took care of the “Methods” section and Tindaro Cicero and Marco Malgarini were jointly responsible for the estimates contained in the “The influence of journal classification on the article score” section. All authors have read and agreed to the final content of the manuscript.\n\n\nCompeting 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\nReferences\n\nAncaiani A, Anfossi A, Barbara A, et al.: Evaluating Scientific Research in Italy: the 2004–10 Research Evaluation Exercise. Res Eval. 2015. Publisher Full Text\n\nCicero T, Malgarini M, Benedetto S: Research quality, characteristics of publications and socio-demographic features of Universities and Researchers: evidence from the Italian VQR 2004–2010 evaluation exercise. Proceedings of the science and technology indicators Conference 2014 Leiden “Context Counts: Pathways to Master Big and Little Data”. 2014. Reference Source\n\nFerrara A, Bonaccorsi A: How robust is journal rating in Humanities and Social Sciences? Evidence from a large scale multi-method exercise. Submitted for publication. 2015.\n\nFinkenstaedt T: Measuring research performance in the humanities. Scientometrics. 1990; 19(5–6): 409–417. Publisher Full Text\n\nFrost CO: The use of citations in literary research: a preliminary classification of citation functions. The Library Quarterly: Information, Community, Policy. 1979; 49(4): 399–414. Reference Source\n\nGuetzkow J, Lamont M, Mallard G: What is originality in the humanities and the social sciences? Am Sociol Rev. 2004; 69(2): 190–212. Publisher Full Text\n\nHammarfelt B: Interdisciplinarity and the intellectual base of literature studies: Citation analysis of highly cited monographs. Scientometrics. 2011; 86(3): 705–725. Publisher Full Text\n\nHemlin S: Social studies of the humanities: a case study of research conditions and performance in ancient history and classical archaeology, and English. Res Eval. 1996; 6(1): 53–61. Publisher Full Text\n\nHemlin S, Gustafsson M: Research production in the arts and humanities. A questionnaire study of factors influencing research performance. Scientometrics. 1996; 37(3): 417–432. Publisher Full Text\n\nHug SE, Ochsner M, Daniel HD: Criteria for assessing research quality in the humanities: a Delphi study among scholars of English literature, German literature and art history. Res Eval. 2013; 22(5): 369–383. Publisher Full Text\n\nHug SE, Ochsner M, Daniel HD: A framework to Explore and Develop Criteria for Assessing Research Quality in the Humanities. International Journal for Education Law and Policy. Forthcoming. 2014; 10(1): 1–14. Reference Source\n\nNederhof AJ, Zwaan RA, De Bruin RE, et al.: Assessing the usefulness of bibliometric indicators for the humanities and the social and behavioural sciences – a comparative study. Scientometrics. 1989; 15(5–6): 423–435. Publisher Full Text\n\nOchsner M, Hug SE, Daniel HD: Indicators for Research Quality for Evaluation of Humanities Research: Opportunities and Limitations. Bibliometrie - Praxis und Forschung. 2012; 1: 4. Reference Source\n\nOchsner M, Hug SE, Daniel HD: Four types of research in the humanities: setting the stage for research quality criteria in the humanities. Res Eval. 2013; 22(2): 79–92. Publisher Full Text" }	[ { "id": "9402", "date": "20 Jul 2015", "name": "Geoffrey Williams", "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\nOverview This papers sets out to exploit some of the massive amount of data collected during the Italian 2012-2014 National Scientific Habilitation (Abilitazione Scientifica Nazionale; ASN). The database created for this evaluation exercise listed the academic output of all Italian researchers applying to become associate or full professors within the Italian university of research institutes. Data was classified by output type – academic, non-academic, journal articles, books, monographs… - and also by the disciplinary section of the candidate. Interestingly, the collected data for article publications went through two independent review processes with a panel ranking journals as academic and non-academic, and giving different weighting to the academic series, and another panel carrying out peer review to look at the quality of the articles submitted. This aim of the study described in this paper was to see if there is agreement between the rankings produced by the two panels that would allow journal rank to be a reasonable proxy for quality. It looks at a selected number of very different fields from the HSS, namely architecture, arts and humanities, history and philosophy, law and sociology and political science. The study proceeds by a statistical analysis of journals ranking and the classification of individual articles given by reviewers on a scale from A (excellent) to D (limited). It also takes into account a series of variables such as the language of production and field size as well as the discipline and age of the writer. The results are clearly of great interest in developing new bibliometric tools for handling data in large evaluation exercises in that they show a clear correlation between journal ranking and the outcome of peer review appraisal. The outcomes indicate that those articles with the best evaluation appeared in class A journals and that thus, although peer review remains important, a journal ranking may be a good proxy for quality. Assessment As this is a short article, much background knowledge would be needed to fully apprehend the criteria behind the original exercise. We have little information as to either the peer review panels and the reality of the criteria they applied, or how the panel classifying journals work, and the extent that it simply followed existing classifications as that of ERIH. These are both key elements in evaluating the output as they represent sources of potential bias and the influence of normative approaches to quality. The following comments thus concern more potential underlying biases in data collection and evaluation policy that the methodology applied here. The first question applies to the ranking of the journals. Who are the experts and what is the danger of field bias? ERIH has come in for enormous criticism as the degree to which the experts represent a field is far from clear. In the case of this assessment, we do not know how the panels were constituted and to what extent the inclusion of a journal in an A list was free of domain bias. In France, the AERES agency had to abandon lists for the HSS as being too hotly disputed, and the field of law had simply ranked its own journals as A anyway. It is relatively easy to highlight a group of high profile, high impact journals in any field, but much more difficult to obtain clear criteria for ranking other journals as reputation measures can be fearfully biased. The same problems arise with peer review, namely representativity of the panels and relevance of criteria to individual fields. We do not know how many reviewers read any individual paper and the extent they actually read the whole paper and were more competent to judge than the field specific peer reviewers who initially reviewed the article. The outcomes from the UK REF showed that reviewers only skimmed publications due to lack of time, given the volume of data to be treated, did these reviewers read more closely. Another issue arises from the allocation of experts, notably international experts who were chosen because the article was already deemed as having a quality potential. Another important factor is the variety within the broad field of SSH, and even variety within disciplines. Architecture might be expected to have a more engineering dissemination profile, whereas political science and sociology can be at the end of the HSS spectrum that is closer to the sciences. The area 11, History, philosophy, pedagogy and psychology, is particularly wide as psychology can have a very different dissemination pattern to history, and is often not included or treated as borderline, within the HSS sphere. Within language, there will be great variation between areas. The oft-cited maxim that humanities researchers write books has been shown to be far too simplistic and is just an example of how broad brush strokes can hide the diversity within fields. The article itself does point out that there can be perverse effects of an evaluation process, especially if it is normative. This may be happening here with internationalisation. Internationalisation is obviously an added value in research, but it should not be seen as a necessary prerequisite of quality. A relative lack of internationalisation is inherent in many humanities disciplines, and is particularly common in law. This is not a lack of quality, but simply due to the national orientation in the field of study. Penalising by too rigid evaluation criteria would be a bad thing as it is for evaluators to understand a field and adapt, not to try and change the field to suit their criteria. These factors do not change the interest of the methodology adopted, but do need to be considered before making policy decisions based on outcomes. The methodology itself is through and opens vistas for analysing what is happening in assessment exercises and how it affects dissemination practice. I have only one minor gripe in what is otherwise a very clear and stimulating paper, and that is with data presentation. For the two datasets shown in tables 1 and 2, it would have been nice to have had averages shown so that we have an instantly visible means of comparison. This would have revealed that law beats all disciplines with a very high level of A class journals, 68% as opposed to only 39% in architecture. These differences merit comment. The article uses sound methodology and opens perspectives for more detailed work, and hopefully a close grain analysis within disciplines that will take into account researcher motivations. This is vital as correlation at this level does not necessarily justify the use of bibliometric criteria over peer review as there are inevitably built in biases in both quantitative and qualitative approaches. Until in-depth studies of how and why researchers disseminate are carried out, the picture will always be falsified. Thus, this research opens interesting channels, and always calls for much more sociological analysis before confirming outcomes.", "responses": [] }, { "id": "9401", "date": "04 Aug 2015", "name": "Alesia Zuccala", "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\nBrief Overview:This study is situated within a national research evaluation context in Italy (i.e., designed for the selection of the best candidates for the ranks of associate and full professor) and focuses on two distinct rating exercises, one for journals and one for individual research articles published in five fields of the Social Sciences and Humanities. The authors take advantage of a fortuitous data set, and use an ordered probit-model to compare the score given by expert peer reviewers to 11,500 research articles (i.e., Excellent-A; Good-B; Fair-C; Limited-D) and the rating of the journal (classified as A or non-A) in which the individual articles were published. For all papers, a series of additional variables are taken into account, including a) language of the article, b) the scientific area of the author, c) the author’s age, academic status and gender, and d) the inclusion of an international co-author. In terms of referees, an allowance is made for the possibility that he/she was international. The purpose of this controlled experiment was to test for the robustness of expert-based journal ratings by determining the probability of a paper receiving a high independent review score, where the journal in which it was published also received a high independent score.Assessment:I tend to agree with the first reviewer in that more background information is needed regarding the criteria behind the original two rating exercises, although I basically find that for a very short article, the statistical aspects of the methodology are quite thorough.I have only one comment about the rating exercises. The article indicates that different panels or groups of individual experts were chosen: some were assigned to provide the journal classifications (but we do not know how many), and a separate group of others (i.e., one non-panel, plus a consensus panel) were asked to rate individual articles from the specific fields (also how many?). Neither group were said to have exchanged information; thus acted independently of one another. The peer reviewers of the individual articles were “instructed to evaluate articles only on the basis of their merit regardless of journal and of language of publication” (p. 3). Here, I am curious specifically about how much information pertaining to the journal (e.g., header; footer; abstract; citation style; volume number) had been removed prior to the article evaluation/rating procedure? Were the reviewers in this experiment truly blind to the journal’s influence? In certain areas of the Humanities, where there are notably fewer A-class journals, perhaps the format of the article instantly gave away the type of journal in which it was published. This could mean that when an article had been rated as being “excellent”, the reviewer was actually making a simultaneous judgement on the journal as well. I would therefore like to assume that all of the articles peer-reviewed in this experiment were non-formatted pre-prints (?); otherwise, this could be a factor which is contributing to the mutually related ratings.", "responses": [] }, { "id": "9400", "date": "10 Aug 2015", "name": "Chiara Faggiolani", "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\nConsistency of title and abstract and potential interest of the subject for readersThe paper deals with a particularly relevant and controversial topic: journal ratings as predictors of articles quality in no-bibliometrics areas (architecture, arts and humanities, history and philosophy, law and sociology and political science).The aim of this paper is to understand whether the probability of receiving positive peer reviews is influenced by having published in an independently assessed, high-ranking journal.The title is appropriate for the content of the article and the abstract represent a suitable summary of the work. The paper is interesting and provides a useful exposure to the objectives proposed. It is well structured and clear in every step, although the article is brief and therefore omits some very important details. Article content: Have the design, methods and analysis of the results from the study been explained and are they appropriate for the topic being studied?The paper analyses a large dataset composed of over 11,500 research articles published in Italy in the period 2004-2010 in the areas of architecture, arts and humanities, history and philosophy, law, sociology and political sciences (non-bibliometric areas).The authors used an ordered profit model to understand whether the probability of receiving positive peer reviews is influenced by having published in an independently assessed, high-ranking journal. The results show that independent classifications of journals may be considered as good predictors of the score assigned to individual articles.In the manuscript, the authors use a big dataset that represents a very important first evaluation of the Italian university research. The paper is remarkable and underlines how a journal of high quality can ensure an excellent rating for a researcher. The method is suitable to achieve the goals. However, it should be highlighted that:1) in the same non-bibliometric area, but in different scientific fields, monographs or articles of books rather than articles of the journal could be important, then, also kind of publication could be affect the evaluation.2) the method adopted is appropriated, but it could be relevant applied clustering techniques to detect the profiles of micro-sectors (scientific fields), and find clusters of scientific fields, which although belonging to different areas have similar;3) the paper deserves a greater depth to allow you to specify in detail the analysis. In fact, the authors present the results at macro level, and not enter into the details of different scientific fields. It would be interesting to have a longer version of the paper. Conclusions: Are the conclusions sensible, balanced and justified on the basis of the results of the study?Using a very large dataset of journal articles published in HSS, the paper proves that journal ranking may be considered as a good predictors of scores assigned to individual articles.The probability of receiving an excellent score almost doubles when the paper is published in a top journal.The results indicate that journal classifications may be a useful supporting tool in large evaluation excercise since it may provide reviewers with valuable information apt to support expert evaluation.It is very important to emphasize, in agreement with The Leiden Manifesto for research metrics that «the best decisions are taken by combining robust statistics with sensitivity to the aim and nature of the research that is evaluated. Both quantitative and qualitative evidence are needed; each is objective in its own way. Decision-making about science must be based on high-quality processes that are informed by the highest quality data» (Hicks, et al., 2015).", "responses": [] } ]	1	https://f1000research.com/articles/4-196
https://f1000research.com/articles/4-58/v1	27 Feb 15	{ "type": "Opinion Article", "title": "Finding small molecules for the ‘next Ebola’", "authors": [ "Sean Ekins", "Christopher Southan", "Megan Coffee", "Christopher Southan", "Megan Coffee" ], "abstract": "The current Ebola virus epidemic may provide some suggestions of how we can better prepare for the next pathogen outbreak. We propose several cost effective steps that could be taken that would impact the discovery and use of small molecule therapeutics including: 1. text mine the literature, 2. patent assignees and/or inventors should openly declare their relevant filings, 3. reagents and assays could be commoditized, 4. using manual curation to enhance database links, 5. engage database and curation teams, 6. consider open science approaches, 7. adapt the “box” model for shareable reference compounds, and 8. involve the physician’s perspective.", "keywords": [ "ebola", "text mining", "open science", "outbreak", "patents", "databases", "box model" ], "content": "Introduction\n\nThe current Ebola virus (EBOV) epidemic points to opportunities for preparing for the next pathogen outbreak or newly identified infectious disease. While control measures and therapeutic strategies have certainly been learned from past outbreaks1, they may be insufficient to control a new one. Given the rapid evolution of viruses and the inexorable global increase in human mobility, this quote from a recent Nature editorial seems prescient “because one thing is clear: whether it is Ebola virus, another filovirus or something completely different, there will be a next time”2. For comparison we can also look at infectious diseases we do have treatments for but still need to improve and/or circumvent drug resistance. For example in our experience in tuberculosis and malaria the patchiness of explicit chemistry connectivity between papers, patents and database entries impedes progress. In the case of EBOV with less than 1500 papers in PubMed and only 25 crystal structures in the Protein Data Bank (at the time of writing), mechanistic aspects that could open the way for therapeutic developments are still not elucidated. Like others3–5 our focus is on small molecule interventions6–8 and we have therefore considered various steps that might help prepare for future pathogens as follows:\n\n\nText mine the literature\n\nThe highest quality and density of information about pathogens resides in peer reviewed publications, patents and databases. In recent years, text mining in general and natural language processing in particular, has become the method of choice for the extraction and collation of facts from document corpora9. This could thus have a rapid payoff in mining for the similarities and differences between emergent versus known pathogens. In the case of EBOV we immediately found antiviral medicinal chemistry basic recall searches (i.e. not authentic text mining) had specificity challenges, even just associated with synonyms for EBOV, related isolates and phylogenetic neighbors (e.g. Marburg virus)10. A corollary of this is that full text of at least EBOV papers could be released for text mining outside pay-walls, by agreement with publishers.\n\n\nPatent assignees and/or inventors should openly declare their relevant filings\n\nPatents contain more published medicinal chemistry data than papers11. For example nearly 200 WO patents for HIV protease inhibitors can be retrieved by a simple word search. This information source also presents a paradox in being, on the one hand, difficult to extract structured data from because of varying degrees of obfuscation, but on the other, full-text is easier to access than papers. In addition, not only does PubChem contain over 18 million structures from patents but also SureChEMBL now automatically extracts the chemistry from newly published filings within days. Preliminary queries indicate the patent corpus covering direct EBOV entry or replication inhibitor chemistry (or for host processing proteases as targets) is small but would nonetheless be very important to access. The only way to make retrieval rapid and complete is for assignees to openly declare their relevant published patent titles and numbers that are inevitably missed by keyword searching. In addition extraction would be much more effective if they re-surfaced the data to make it more accessible. This could be as simple as just uploading an Excel sheet to Figshare (or other open repository) with a few hundred rows of structures (linked to PubChem CIDs where these are already out there), activity values and short assay descriptions, rather than leaving the community to grapple with a hundred page PDF. We realize this is unprecedented but as a type of emergency response has to be considered. Assignee organizations should also encourage their inventors to do exactly this. Logically, another ‘precedent breaker’ can be considered, namely that applicants publish or surface their anti-pathogen patent results effectively the day after filing. This may sound scary to some, but IP rights are conserved while, in community terms, 18 months are cut off the “information shadow” phase. Another move in the right direction has been shown by the World Intellectual Property Organization Re:Search Consortium. Their initiative to open up patents for neglected tropical disease research could also be extended to cover filoviruses and other viruses (http://www.wipo.int/research/en/about/index.html).\n\n\nReagents and assays could be commoditized\n\nResearch can be accelerated by the collaborative exchange of assay reagents and protocols between teams and this has other positive consequences beyond just speed. Crucially, it contributes to inter-lab reproducibility if assays are made robust enough to be transferred. In addition, structure activity relationship (SAR) results will have reduced variance and will thus be more comparable between laboratories. This reciprocity becomes particularly valuable if a pharmaceutical company or other organization (or Molecular Libraries Screening Centre, or Euro Screen) engages to run a high throughput screen. Consequently, multiple collaborators can pick up the baton of analog expansion of confirmed hits via the same standardized assay. A good example in the EBOV case is a recent publication of PDB structures for small molecules that bind the filovirus VP35 protein and inhibit its polymerase cofactor activity12. Supplies of assay-ready VP35 (even from a reagent vendor) would thus be valuable to expand take-up by more screening centers.\n\n\nUsing manual curation to enhance database links\n\nOnce a search of various publications and databases is complete one should be able to navigate reciprocally from a molecule identifier to a structure or from a target to modulating chemistry. This is not always the case, for example where pharmaceutical company lead structures are obfuscated13. A relevant example of useful linkages is exemplified by “Pyridinyl imidazole inhibitors of p38 MAP kinase impair viral entry and reduce cytokine induction by Zaire ebolavirus in human dendritic cells”14. It certainly helped that SB202190 is PubChem positive as CID 5353940 (NJNKPVPFGLGHPA-UHFFFAOYSA-N). This is well-linked as a kinase inhibitor but not directly to these recent EBOV results. The essential role of facilitating such bioactive chemistry linkage is taken by curated databases15–17. These not only curate literature-extracted activity results into structured records but also merge this connectivity into PubChem. The overall data findability/linkability could be further enhanced if the MeSH system could fast-track new EBOV-specific indexing and shorten the lag time.\n\n\nEngage database and curation teams\n\nData mining is expedited by activity results and associated chemical structures being captured in databases. But the time to publish a paper and index the contents into structured records may take years. However, major chemistry resources such as PubChem18, ChEMBL19 and ChemSpider20 are all willing (by prior arrangement) to take direct submissions (e.g. from EBOV screening teams) possibly even pulled directly from electronic lab notebooks (ELNs) along with the crucial metadata. The same approach of data generators actively engaging with speeding up and improving the transfer of their own results into databases applies equally to the sequence side of things. For EBOV the bioinformatics and genomics communities appear demonstrably ahead of the game compared to the medicinal chemistry community. For example the ViralZone in Europe and the US Virus Pathogen Resource were quickly established as integrated knowledge portals21 (http://www.viprbrc.org/brc/home.spg?decorator=vipr). Significantly, the related problems of terminology mapping for curators and retrieval specificity for users is already being addressed22. What might be less well known is that authors submitting sequences can apply this rule by engaging directly with database staff (e.g. including feedback for MeSH indexing) to ensure the rapid and precise annotation of new virus entries.\n\n\nConsider Open Science approaches\n\nThe advantages of this to pathogen drug discovery (including the abrogation of intellectual property generation) primary data sharing can then become instantaneous and global23. The consequent shortening of drug research stages can be dramatic. For example an InChIkey surfaced from an ELN (or other open instantiations such as Wikis or Figshare (http://figshare.com/)) means that chemistry becomes findable within hours of Google indexing24. It also frees teams from the ‘tyranny of novelty’ where new leads can be rationally optimized from pre-existing ones. In contrast the conventional IP-centric research model not only includes the years of delay to prepare a paper (i.e. 18 months after a patent application) but, even then, not all relevant antiviral chemistry flows from papers into public databases. This can also expedite the free exchange of reagents and protocols. The Open Science model can also leverage the “wisdom of the crowd” such that a global volunteer cadre of experienced chemists (industry and academic) can immediately participate both in SAR interpretation and in the design cycle. We would also suggest open sharing not only of small molecules or data sets from relevant assays, but also of a range of predictive (and sharable) models or hypotheses that can be used for virtual screening. Any SAR data from the literature can be mined to understand physicochemical properties or molecular features important for antiviral activity. Such ligand-based computational models in turn could then be used for searching additional libraries of compounds (e.g. pharma companies might even implement this on their complete proprietary screening collections and share the results). Curated data sets could be used to construct “whole cell” virus specific machine learning models, similar to those for Mycobacterium tuberculosis25. Computational algorithms like Connectivity Map26, SEA27 and others could be implemented to enable fast querying of the data so that the most similar virus to an unknown could be found, and from there the most active compounds. Software like Euretos BRAIN, could be used to mine relationships between different biological terms and molecules that can then be used for target inference28.\n\n\nAdapt the “box” model for shareable reference compounds\n\nThe idea here is to create an openly available diverse set of compounds that are not likely to yield false positives, aggregators or other undesirable structural types commonly termed PAINS29. These “box” compounds would possess known antiviral and anti-pathogen activity plated out for wide availability. The notable precedents here are the MMV Malaria Box (www.ebi.ac.uk/chemblntd) (upcoming), MMV Pathogen Box (http://www.mmv.org/research-development/project-portfolio/pathogen-box) and the NIH Clinical Collection (http://www.nihclinicalcollection.com/). These could be hosted by a third party on behalf of NIAID, CDC etc. This model is flexible in terms of multiple “boxes” being possible. For example sets of ~1400 screening-ready FDA approved drugs (that physicians would have ready access to), would be the first logical pass for repurposing investigations. The next “box” could include the ~8000 structures in PubChem that include an International Nonproprietary Name (INN) designation and therefore in most cases have clinical testing. The advantages of what we can call ‘virtuous circularity’ of connectivity, apply exactly in this case. Specifically the major chemical databases can ensure a) they tag availability (i.e. a retrievable flag “this compound is in free box Y”, b) the publications, patents and historical assay results are linked to the same entries and crucially c) new results (with appropriate provenance) from users of box compounds are promptly added back into the database records. By logical extension the computational modelling efforts can then loop through more rounds of improvement, new testing, leading to better hits that are put back into the ‘box’.\n\n\nInvolve the physician’s perspective\n\nAs we are seeing with EBOV, clinicians and healthcare providers are the first line of defense for the rest of the world. They are also at the greatest risk from the pathogen themselves. They are also clearly in the best position to decide how to treat their patients; the steps above should result in treatments that can actually be obtained8 and tolerated by the patient. Physicians with experience of treating infectious diseases could be engaged to group treatments as 1. drugs which they would use in patients who are very ill; 2. those drugs which would be of concern as they may do more harm than good and 3. those drugs which might be used regardless (to explore if effective). In the case of a virulent pathogen with only palliative treatment options, from a physician’s perspective, anything that reduces mortality (even slightly) is crucial or which may even have other clinical endpoints (such as reduced hospitalization or reduced symptoms) even if mortality isn’t affected. Another way to think of this is to increase the number of patients that can be treated.\n\nIn conclusion, what we are seeing now has a precedent in other viruses we were not “expecting” (e.g. HIV). Even decades on we have combination therapies to control the disease but no cure or vaccine. For EBOV we have had nearly 40 years to prepare. The cost effective suggestions above could be implemented to prepare for when the next new pathogen arrives, otherwise we will be in the same situation again. We propose that as new pathogens are identified we should be able to rapidly identify new antiviral drugs as well as establish where approved drugs that physicians have experience with, can be effective. This approach could be applicable to other infectious diseases beyond those which we currently know.", "appendix": "Author contributions\n\n\n\nAll authors contributed to the writing of the final manuscript.\n\n\nCompeting interests\n\n\n\nSE is a consultant for CDD.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nDhillon RS, Srikrishna D, Sachs J: Controlling Ebola: next steps. Lancet. 2014; 384(9952): 1409–11. PubMed Abstract \| Publisher Full Text\n\nAnon: Call to action. Nature. 2014; 514(7524): 535–536. PubMed Abstract \| Publisher Full Text\n\nPicazo E, Giordanetto F: Small molecule inhibitors of ebola virus infection. Drug Discov Today. 2014; 20(2): 277–286. PubMed Abstract \| Publisher Full Text\n\nDe Clercq E: Ebola virus (EBOV) infection: Therapeutic strategies. Biochem Pharmacol. 2015; 93(1): 1–10. PubMed Abstract \| Publisher Full Text\n\nSouthan C: Anti-Ebola medicinal chemistry - time for crowd sourcing?2014. Reference Source\n\nEkins S, Freundlich JS, Coffee M: A common feature pharmacophore for FDA-approved drugs inhibiting the Ebola virus [v2; ref status: indexed, http://f1000r.es/4wt]. F1000Res. 2014; 3: 277. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLitterman NK, Lipinski CA, Ekins S: Small molecules with antiviral activity against the Ebola virus [v1; ref status: indexed, http://f1000r.es/523]. F1000Res. 2015; 4: 38. Publisher Full Text\n\nEkins S, Coffee M: FDA approved drugs as potential Ebola treatments [v1; ref status: approved 1, http://f1000r.es/53k]. F1000Res. 2015; 4: 48. Publisher Full Text\n\nRebholz-Schuhmann D, Oellrich A, Hoehndorf R: Text-mining solutions for biomedical research: enabling integrative biology. Nat Rev Genet. 2012; 13(12): 829–39. PubMed Abstract \| Publisher Full Text\n\nSouthan CD: Anti-Ebola medicinal chemistry - time for crowd sourcing?2014. Reference Source\n\nSouthan CD: Expanding opportunities for mining bioactive chemistry from patents. Drug Discov Today Technol. 2014. Publisher Full Text\n\nBrown CS, Lee MS, Leung DW, et al.: In silico derived small molecules bind the filovirus VP35 protein and inhibit its polymerase cofactor activity. J Mol Biol. 2014; 426(10): 2045–58. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSouthan C, Williams AJ, Ekins S: Challenges and recommendations for obtaining chemical structures of industry-provided repurposing candidates. Drug Discov Today. 2013; 18(1–2): 58–70. PubMed Abstract \| Publisher Full Text\n\nJohnson JC, Martinez O, Honko AN, et al.: Pyridinyl imidazole inhibitors of p38 MAP kinase impair viral entry and reduce cytokine induction by Zaire ebolavirus in human dendritic cells. Antiviral Res. 2014; 107: 102–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu T, Lin Y, Wen X, et al.: BindingDB: a web-accessible database of experimentally determined protein-ligand binding affinities. Nucleic Acids Res. 2007; 35(Database issue): D198–201. 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): D1083–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPawson AJ, Sharman JL, Benson HE, et al.: The IUPHAR/BPS Guide to PHARMACOLOGY: an expert-driven knowledgebase of drug targets and their ligands. Nucleic Acids Res. 2014; 42(Database issue): D1098–106. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Y, Bolton E, Dracheva S, et al.: An overview of the PubChem BioAssay resource. Nucleic Acids Res. 2010; 38(Database issue): D255–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaulton A, Bellis LJ, Bento AP, et al.: ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2012; 40(Database issue): D1100–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPence HE, Williams AJ: ChemSpider: An Online Chemical Information Resource. J Chem Educ. 2010; 87(11): 1123–1124. Publisher Full Text\n\nHulo C, de Castro E, Masson P, et al.: ViralZone: a knowledge resource to understand virus diversity. Nucleic Acids Res. 2011; 39(Database issue): D576–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuhn JH, Andersen KG, Bào Y, et al.: Filovirus RefSeq entries: evaluation and selection of filovirus type variants, type sequences, and names. Viruses. 2014; 6(9): 3663–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobertson MN, Ylioja PM, Williamson AE, et al.: Open source drug discovery - a limited tutorial. Parasitology. 2014; 141(1): 148–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSouthan C: InChI in the wild: an assessment of InChIKey searching in Google. J Cheminform. 2013; 5(1): 10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEkins S, Freundlich JS, Reynolds RC: Are bigger data sets better for machine learning? Fusing single-point and dual-event dose response data for Mycobacterium tuberculosis. J Chem Inf Model. 2014; 54(7): 2157–65. PubMed Abstract \| Publisher Full Text\n\nLamb J, Crawford ED, Peck D, et al.: The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science. 2006; 313(5795): 1929–35. PubMed Abstract \| Publisher Full Text\n\nBesnard J, Ruda GF, Setola V, et al.: Automated design of ligands to polypharmacological profiles. Nature. 2012; 492(7428): 215–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Haagen HH, Hoen PA, Mons B, et al.: Generic information can retrieve known biological associations: implications for biomedical knowledge discovery. PLoS One. 2013; 8(11): e78665. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBaell JB, Holloway GA: New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J Med Chem. 2010; 53(7): 2719–2740. PubMed Abstract \| Publisher Full Text" }	[ { "id": "8258", "date": "13 Apr 2015", "name": "Qiaoying Zeng", "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 article proposed an integrative strategy in anticipation of the next outbreak of an emerging or reemerging infectious disease like Ebola. The authors suggested seven cost effective steps that help to establish a synergistic mechanism for a fast discovery of small molecule therapeutics. Under this mechanism, the information and data from publications, patents, and database, even some reagents and assays in different labs, could be shared instantaneously and efficiently among scientific and clinical communities around the globe. The direct experience of physicians is also a great plus. These cost effective approaches could be implemented to prepare in advance for the next pathogen outbreak or newly identified infectious disease that is definite no matter when or where it starts. The article is interesting and will be beneficial for an efficient battle against the future new outbreaks of infectious diseases. Minors: Some sentences are confused and needed to be more concise or clarified:In the case of EBOV we immediately found antiviral medicinal chemistry basic recall searches (i.e. not authentic text mining) had specificity challenges, even just associated with synonyms for EBOV, related isolates and phylogenetic neighbors (e.g. Marburg virus). The advantages of this to pathogen drug discovery (including the abrogation of intellectual property generation) primary data sharing can then become instantaneous and global.", "responses": [ { "c_id": "1437", "date": "03 Jul 2015", "name": "Sean Ekins", "role": "Author Response", "response": "Thank you for these suggestions which have now been addressed in the latest version." } ] }, { "id": "8318", "date": "14 Apr 2015", "name": "Martin Zacharias", "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 recent Ebola virus outbreak came as a surprise and the authors suggest possible steps to improve the situation in case of another Ebola epidemic outbreak. The main focus is on cost effective easy implementable steps that could be taken to accelerate drug development or other therapeutic approaches. I think the suggested strategies and cost effective approaches are relevant not only in case of Ebola but could also be useful in case of other infectious diseases. The paper is well and clearly written and of interest for a broad readership in the area of pharmaceutical research, medicine and other biomedical disciplines.", "responses": [ { "c_id": "1436", "date": "03 Jul 2015", "name": "Sean Ekins", "role": "Author Response", "response": "Thank you for this review." } ] } ]	1	https://f1000research.com/articles/4-58
https://f1000research.com/articles/3-139/v1	01 Jul 14	{ "type": "Software Tool Article", "title": "Biological network analysis with CentiScaPe: centralities and experimental dataset integration", "authors": [ "Giovanni Scardoni", "Gabriele Tosadori", "Mohammed Faizan", "Fausto Spoto", "Franco Fabbri", "Carlo Laudanna", "Gabriele Tosadori", "Mohammed Faizan", "Fausto Spoto", "Franco Fabbri", "Carlo Laudanna" ], "abstract": "The growing dimension and complexity of available experimental data generating biological networks has increased the need for tools allowing to categorize nodes by their topological relevance in biological networks. Here we present CentiScaPe, a Cytoscape app specifically designed to calculate centrality indexes for the identification of the most important nodes of a network. CentiScaPe is a comprehensive suite of algorithms dedicated to network nodes centrality analysis, computing several centralities for undirected, directed and weighted networks. The results of the topological analysis can be also integrated with data sets from lab experiments, such as expression or phosphorylation levels of the proteins represented in the network, using the graphical features of the tool. This opens a new perspective in the analysis of biological networks, since integration of topological analysis with lab experimental data can increase the predictive power of a bioinformatical analysis.", "keywords": [ "Biological processes can be displayed as networks where the nodes represent biological entities", "and edges represent interactions between these entities. Several kinds of biological networks have been introduced", "including metabolic networks", "gene networks", "signal transduction networks", "and protein-protein interaction networks1. Such networks are a static representation of the dynamics of biological processes", "where molecular interactions give rise to cascades of reactions", "called pathways", "determining the life processes of living organisms. Even if the scientific community is far from being able to simulate the dynamic behavior of such pathways", "important information can be extracted from the topological analysis of biological networks2", "3 since the structure of a network can affect its function4. In this context", "several global parameters are commonly used to describe the properties of the whole networks", "while centralities5 are parameters allowing the identification of the most important nodes which could be key regulators in the biological process being studied." ], "content": "Introduction\n\nBiological processes can be displayed as networks where the nodes represent biological entities, and edges represent interactions between these entities. Several kinds of biological networks have been introduced, including metabolic networks, gene networks, signal transduction networks, and protein-protein interaction networks1. Such networks are a static representation of the dynamics of biological processes, where molecular interactions give rise to cascades of reactions, called pathways, determining the life processes of living organisms. Even if the scientific community is far from being able to simulate the dynamic behavior of such pathways, important information can be extracted from the topological analysis of biological networks2,3 since the structure of a network can affect its function4. In this context, several global parameters are commonly used to describe the properties of the whole networks, while centralities5 are parameters allowing the identification of the most important nodes which could be key regulators in the biological process being studied.\n\nIn this paper we present CentiScaPe 2.1, a Cytoscape app6,7 for network centralities analysis. While the built in Analyze Network tool is oriented to characterize the global behavior of the network. Provided with several global network statistics, CentiScaPe is designed to identify the most relevant nodes, and it provides a more complete set of centralities. This new version introduces the computation of centralities for directed and weighted networks not available in any other Cytoscape app. As in the previous version, it computes Average Distance, Diameter, Degree, Stress, Betweenness, Radiality, Closeness, Centroid Value, and Eccentricity. Other parameters as Eigenvector, Bridging centrality and Edge Betweenness have been added, making it the most complete app for network centrality analysis.\n\nA web version of CentiScaPe, FastCentiScaPe, is also available (http://www.cbmc.it/fastcent/), which performs very fast computations for large networks sending the network to a multiprocessor server. The centrality analysis results are sent to the user by e-mail in xgmml format.\n\nCentiScaPe’s main goal is to produce results that should drive further lab experiments, as the high score nodes identified by the computation can be considered as potential targets for drug development and new experiments.\n\n\nMethods and implementation\n\nTo calculate all the centralities, the computation of the shortest path between each pair of nodes in the graph is needed. The algorithm for the shortest path is the Dijkstra algorithm8, that has been adjusted in order to compute all the shortest paths pairs (needed for Stress and Betweenness computation)9 and to use also edge direction and edge weight in the shortest path identification. A good description of this and other centrality algorithms can be found in Koschützki et al.5. CentiScaPe 2.1 introduces the computation of centralities for directed networks, networks where the edges are considered to have a direction. Consequently, some nodes cannot be reached by others: given two nodes s and t, it is possible that there is no path from s to t, or from t to s or both. Several centrality parameters are based on the computation of the shortest path between each couple of nodes and are not defined when there are two nodes not connected by a path. In this case CentiScaPe consider the distance from s to t equal to infinity. The centrality definitions have been modified to consider this case (see supplementary materials: CentralitiesTutorial), so the directed centralities can also be used to compute centralities for disconnected networks, i.e. networks where some nodes cannot be reached by others. This gives the use of centrality analysis great flexibility.\n\nIn the case of weighted networks the edges are supposed to have a numerical integer attribute depending on experimental data or on some feature of the network. This numerical value is treated as a distance in the computation of the shortest path between two nodes. In an unweighted network the distance between two node connected by an edge is equals to 1, and the distance of two generic nodes is the number of edge of the shortest path connecting them. In a weighted network the distance is the sum of the attribute values of the edges connecting the nodes, so the shortest path is not necessary the one with the lowest number of edges, but the one with the least distance.\n\nCentiScaPe is written in Java as a Cytoscape app, in order to exploit all the excellent features of Cytoscape and to reach the largest number of users. It has a multi-thread core that can exploit multiprocessor architecture. The Java library JFreechart10 has been used for some of the graphic features.\n\n\nResults and discussion\n\nThe Main use of CentiScaPe is to rank the nodes of a network depending on their topological and experimental relevance. The numerical results are saved as node, edge or network attributes in the Cytoscape attributes browser, depending on the kind of parameters, so all the Cytoscape features for managing attributes are supported. After the computation the centralities are treated as normal Cytoscape attributes. CentiScaPe can be used in undirected networks11, in directed networks and in weighted networks.\n\nCentralities for directed networks (see Supplementary Files: CentralitiesTutorial) are useful in the case of metabolic networks where the direction is from substrates and reactants to the products of the chemical reactions and in signal transduction networks, where the direction depends on the flux of information. Considering direction in the computation of centralities can lead to different and more precise results than the undirected version.\n\nAs example, in Figure 1 the computation of the directed and undirected Stress applied to a network of Oncogenes is shown (see Supplementary Files: Oncogenes.txt and Oncogenes_edge_directions.txt). Results of both the computations are shown. The image, obtained using Cytoscape, represents the different Stress values using the color and size of nodes. The node’s size represents the value obtained using the directed Stress, so the bigger the node the higher the value; the color represents the values obtained using the undirected Stress: red is used for the highest values, green for the lowest values. For example a large green node is interesting because it means that a node with a high value of the new algorithm has a low value using CentiScaPe. While analyzing the oncogenes network we saw that the large red node, AKT1, shows how its stress values are high using both algorithms. However the green medium-sized node, FANCE or RAF1, shows how, using undirected Stress, we obtain a low value for stress, but using the new algorithm we obtain a high stress value. The opposite situation is found in the third highlighted node, the small yellow node, RB1, in the right bottom corner. Here, the value computed with undirected Stress is not very high, similar to the red node, but the value computed with the directed Stress is very low. This can be explained by suggesting that AKT1 is a very important node, that is essential for maintaining connections within this network. For RB1 and CREB1 the situation is not very clear because we have opposite situations. If we use CentiScape, for FANCE, we assume that this node is not essential, but using the directed values it appears to be very central.\n\nSize represents directed Stress, color undirected Stress (green=low, red=high).\n\nSecond important features of the new version of CentiScaPe is the possibility of computing centralities for weighted networks, networks where the edges are provided with an attribute that can be interpreted as a distance between the two connected nodes.\n\nIn the network depicted in Figure 2 we have a distance (dist) attribute for each edge. We have dist(A,B)=2, dist(B,C)=3 and dist(A,C)=7. Since A and C are connected by a single edge, in an unweighted computation, the distance from A to C is equal to 1. But if the attributes of the edges are considered as distances, the shortest path between A and C is the one passing through B (=2+3=5) since it is shorter than the one connecting A directly to C (=7). The computation of weighted shortest paths will result in completely different values compared to a situation where the weight is not considered. The user should take care that the weight is used in cases where close nodes are more important than distant nodes. So, depending on the meaning of the attributes, one can use the value or its reciprocal. For example if the attribute represents the speed of a reaction instead of a distance, the reciprocal should be used.\n\nIf the weight of the edge is not considered the shortest path from A to C is the directed edge, and the distance is 1. If the weights are considered the shortest path from A to C is the one passing through B, and the distance is equals to 5.\n\nAn example where weighted networks centrality analysis is used, is provided by Currie et al.12 where an euclidean distance is given to each edge depending on the difference between the phosphorylation level of the proteins connected by that edge.\n\nAll the graphical features of the previous version of CentiScaPe, such as the plot by node, the plot by centrality and the boolean-based result panel have been maintained in the new version. A complete guide can be found in Scardoni et al.11 or in the CentiScaPe userguide available from the website.\n\n\nConclusions\n\nCentiScaPe 2.1 has been enriched with new centrality parameters such as Eigenvector, Bridging centrality and Edge Betweenness centrality, and with the possibility to analyze directed and weighted networks. It allows integrating centrality-based network analysis with experimental data. The results of the computation can be used and exported as Cytoscape attributes, allowing the user to exploit all the other features of Cytoscape and its apps. Compared to the built-in Analyze Network tool of Cytoscape, CentiScaPe is an excellent integrative tool allowing the identification of potential target nodes from both the topological and the experimental point of view, and can be considered as an essential instrument for the characterizations of nodes in order to drive further experiments.\n\n\nSoftware availability\n\nSoftware available from the Cytoscape App Store: http://apps.cytoscape.org/apps/centiscape\n\nLatest source code: https://bitbucket.org/giovanniscardoni/centiscapepublic/\n\nSource code as at the time of publication: https://bitbucket.org/F1000Research/centiscapepublic-archive\n\nArchived source code as at the time of publication: http://dx.doi.org/10.5281/zenodo.1065213\n\nLicense: Lesser GNU Public License 3.0: https://www.gnu.org/licenses/lgpl.html", "appendix": "Author contributions\n\n\n\nGS is the main designer and developer of CentiScaPe. CL contributed to the design of CentiScaPe and performed the experiments. GT contributed to the definition of the directed centralities and performed the examples of usage in the Supplementary files. MF developed the directed centralities. FS contributed to the development of the last version of CentiScaPe. FF is the main developer of the web version of CentiScaPe.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by: Italian Association for Cancer Research (AIRC, IG 8690) (C.L.); Fondazione Cariverona; Nanomedicine project University of Verona and Fondazione Cariverona (C.L.). Directed centralities have been developed thanks to the GSoC2013 program.\n\n\nSupplementary files\n\nSupplementary files available from: https://f1000researchdata.s3.amazonaws.com/supplementary/4477/dd4541d4-bdfe-461a-be92-f2e2cf9f2d57.zip\n\n\nReferences\n\nGuido Caldarelli. Scale-Free Networks: Complex Webs in Nature and Technology (Oxford Finance). Oxford University Press, USA. 2007. Reference Source\n\nJeong H, Tombor B, Albert R, et al.: The large-scale organization of metabolic networks. Nature. 2000; 407(6804): 651–654. PubMed Abstract \| Publisher Full Text\n\nBarabási AL, Oltvai ZN: Network biology: understanding the cell’s functional organization. Nat Rev Genet. 2004; 5(2): 101–113. PubMed Abstract \| Publisher Full Text\n\nStrogatz SH: Exploring complex networks. Nature. 2001; 410(6825): 268–276. PubMed Abstract \| Publisher Full Text\n\nDirk K, Lehmann KA, Peeters L, et al.: Centrality indices. In Ulrik Brandes and Thomas Erlebach, editors, Network Analysis: Methodological Foundations, Springer. 2005; 16–61. Publisher Full Text\n\nCline MS, Smoot M, Cerami E, et al.: Integration of biological networks and gene expression data using Cytoscape. Nat Protoc. 2007; 2(10): 2366–2382. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaito R, Smoot ME, Ono K, et al.: A travel guide to Cytoscape plugins. Nat Methods. 2012; 9(11): 1069–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDijkstra EW: A note on two problems in connexion with graphs. Numerische Mathematik. 1959; 1(1): 269–271. Publisher Full Text\n\nScardoni G, Laudanna C: Centralities based analysis of complex networks. In Yagang Zhang, editor, New Frontiers in Graph Theory. InTech. 2012. Publisher Full Text\n\nGilbert D: JFreeChart. Reference Source\n\nScardoni G, Petterlini M, Laudanna C: Analyzing biological network parameters with CentiScaPe. Bioinformatics. 2009; 25(21): 2857–2859. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCurrie HN, Vrana JA, Han AA, et al.: An approach to investigate intracellular protein network responses. Chem Res Toxicol. 2014; 27(1): 17–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScardoni G, Tosadori G, Faizan M, et al.: F1000Research-centiscapepublic-archive. ZENODO. 2014. Data Source" }	[ { "id": "5358", "date": "07 Jul 2014", "name": "Ferenc Jordan", "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\nAnalysing several centrality measures in parallel is important and there is a need for this, however, it would be even nicer to compare them (see http://www.sciencedirect.com/science/article/pii/S0304380007001184 ) and discuss better their biological relevance (i.e. what is Eigenvector centrality good for?). Central nodes can be of key regulator function but this is not a result, this is still a hypothesis. It should be mentioned briefly that either simulations or lab experiments should reinforce these findings. For infinite distances, using the reciprocal distance matrix has already been suggested as an elegant solution (see http://link.springer.com/article/10.1007/BF01164642#page-1). It should be clarified a bit more that a large weight can be considered as a short or as a long path, depending on its biological meaning. In Figure 1, the grid layout algorithm could be replaced by some better one, I think. Finally, a quick English check would be welcome.", "responses": [] }, { "id": "5341", "date": "11 Jul 2014", "name": "Ankush Sharma", "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 research article, the authors describe the CentiScaPe app for Cytoscape (and its web version FastCentiScaPe) for computing the several centrality parameters in networks, analyzing the elemental level importance of each node in the network, based on network topology. The addition of centralities parameters, especially Bridging Centrality and Eigenvector centrality, appears to be useful to improve metrics, to infer the informational flow across overlapping modules, and as a measure of the influence of a node in a network. The article is well written with an appropriate title and an abstract which provides sufficient details. The description of the purpose, implementation and usage of CentiScaPe app are informative and detailed for the audience. Minor comments:The example of calculating stress in the network of oncogenes in the manuscript should comply with experimental functional relevance, e.g. a decrease in bridging centrality is observed for key intermodular nodes upon heat shock response or oxidative stress, partially disintegrating overlapping modules into local communities which alters the centrality statistics of the nodes 1,2. It would be good to see such implementation in the example showing applicability of CentiScaPe. I suggest the authors have a small example of the analysis comparing connected and disconnected networks, to support computation of centralities for networks with a disconnected component. It would be better to implement a user-friendly exploration of results and user registration feature instead of email authorization, in order to avoid delays in using the web version FastCentiScaPe. Correction in the name of built-in tool i.e. NetworkAnalyzer , instead of \"Analyze Network\" tool. Usage of appropriate words such as “more comprehensive” in place of “more complete” in introduction section.", "responses": [] } ]	1	https://f1000research.com/articles/3-139
https://f1000research.com/articles/4-195/v1	07 Jul 15	{ "type": "Software Tool Article", "title": "Connecting the pieces: Using ORCIDs to improve research impact and repositories", "authors": [ "Mohamed Baessa", "Thibaut Lery", "Daryl Grenz", "J. K. Vijayakumar", "Thibaut Lery", "Daryl Grenz", "J. K. Vijayakumar" ], "abstract": "Quantitative data are crucial in the assessment of research impact in the academic world. However, as a young university created in 2009, King Abdullah University of Science and Technology (KAUST) needs to aggregate bibliometrics from researchers coming from diverse origins, not necessarily with the proper affiliations. In this context, the University has launched an institutional repository in September 2012 with the objectives of creating a home for the intellectual outputs of KAUST researchers. Later, the university adopted the first mandated institutional open access policy in the Arab region, effective June 31, 2014. Several projects were then initiated in order to accurately identify the research being done by KAUST authors and bring it into the repository in accordance with the open access policy.\nIntegration with ORCID has been a key element in this process and the best way to ensure data quality for researcher’s scientific contributions. It included the systematic inclusion and creation, if necessary, of ORCID identifiers in the existing repository system, an institutional membership in ORCID, and the creation of dedicated integration tools. In addition and in cooperation with the Office of Research Evaluation, the Library worked at implementing a Current Research Information System (CRIS) as a standardized common resource to monitor KAUST research outputs. We will present our findings about the CRIS implementation, the ORCID API, the repository statistics as well as our approach in conducting the assessment of research impact in terms of usage by the global research community.", "keywords": [ "ORCID", "Research Impact", "Research repositories", "KAUST", "CRIS" ], "content": "Introduction\n\nThe quality of a research institute can be defined in many ways; quality of its teaching, graduates, research outputs, social, environmental and economic impacts. Indeed, many of these attributes are incorporated into various international ranking formulae which are beginning to garner more and more media and political attention year on year. In this context, it is important for a young university, such as King Abdullah University of Science and Technology (KAUST), created 5 years ago, to gather, curate, monitor and analyze all the publications produced by its researchers. Recognizing the fundamental importance of preserving and sharing such knowledge, the University Library launched its digital repository services in September 2012 (available at http://repository.kaust.edu.sa/kaust/). As correctly mentioned by Richard K. Johnson (The former library reference and collection manager 2009 – 2013) “University rankings typically look at the volume and influence of an institution’s research output, when works are openly accessible on the Internet, they get more citations. And citations are the whole game”. We present here our approach to the objectives and ensuing policies of an academic repository, its integration with other academic systems, its usage as a research evaluation tool and its importance for the university.\n\nThe repository (http://repository.kaust.edu.sa/kaust/) services aim to bring together all of the university’s intellectual output, published and unpublished works alike. After several months of evaluation and feasibility study, the library and IT teams evaluated a number of issues (e.g., technology selection, hosting strategy, depositor authentication) and recruited and trained staff for the program. The following three objectives were established for the repository [http://www.dpconline.org/newsroom/whats-new/933-whats-new-issue-50-nov-2012#oneworld]:\n\n➢ Provide persistent access to university intellectual assets, including grey literature (e.g., technical reports, conference papers, theses etc.) and research data, in order to preserve and share scientific knowledge created at KAUST.\n\n➢ Showcase the intellectual output of KAUST research, the development of international research networks and collaborations, and support graduate student and post-doc recruitment.\n\n➢ Expand the impact of KAUST research, which contributes to increased awareness of and growing prestige for our new and ambitious university.\n\nThe real challenge for any repository initiative lies in trying to serve the needs of a variety of stakeholders. Dealing with this challenge requires identifying different groups of stakeholders, understanding their motivations and needs, and then striving to meet their aspirations. For example, researchers’ concerns lie with the impact of their research within the scholarly community; in order to have an impact, their research outputs in the repository need to be discoverable in ways that allow them to be disseminated widely among the relevant community of researchers. [http://www.sparc.arl.org/sites/default/files/ir_final_release_102.pdf]\n\nOn the other hand, the research evaluation office seeks to locate an accurate and comprehensive list of different university research outputs, to be able to analyze it and better evaluate the effectiveness and significance of research activities.\n\nThe library has taken different approaches to building the collection and to meeting the needs and concerns of varied stakeholders. This paper will discuss policy and technology developments and their role in building the repository collections, how ORCID implementation is helping in locating and linking research, how these efforts relate to research evaluation at KAUST, and finally will discuss the role of the library in engaging the community to learn about related issues.\n\n\nDevelopment of repository policies\n\nPolicy plays an important role in terms of the repository collections development by granting authority and shaping the framework of assigned roles and responsibilities. The primary relevant policies so far have been for management of theses and dissertations, and for open access to published research.\n\n\nElectronic theses and dissertations policy\n\nTheses and dissertations represent the scholarly achievement of KAUST students, and from the beginning the university has made a decision to capture, preserve, and provide access only to the electronic versions of this work, with no printed versions required. Theses and dissertations were managed within SharePoint (Microsoft) for the first two years while decisions were made regarding a long-term policy and platform. With the decision in 2012 to move theses and dissertations to the repository and provide public access, the theses and dissertations policy addressed issues like copyright while also setting clear roles and responsibilities [https://www2.le.ac.uk/library/downloads/etd2014/etds-in-a-new-university], the following figure, Figure 1 demonstrates some unique repository materials that have been cited.\n\nThe electronic theses and dissertations policy retains a non-exclusive license for the university to make dissertations or theses publicly available, with an option for the student to embargo public access for one year. The submission workflow has also been standardized, with students submitting the final version of their work to their graduate program coordinators, who then deposit the files in the repository, where the deposit is checked by the repository team before being made publicly available. The repository now has 338 theses and 80 dissertations made available through this policy, with approximately 30 citations.\n\n\nOpen access policy\n\nThe rise of the concept of open access to research or scholarly output has led many funding bodies, institutions, and governments to adopt open access policies [http://legacy.earlham.edu/~peters/fos/newsletter/02-02-09.htm#choicepoints]. When KAUST adopted an open access policy covering published research in June 2014, it became the first proponent of such a policy in the Middle East [http://hdl.handle.net/10754/337608].\n\nUnder the policy each KAUST author grants a non-exclusive license to the university to exercise all rights under copyright relating to each of the authors’ scholarly research articles for the purpose of open dissemination. A committee consisting of faculty, library and legal office representatives, conducted over 15 meetings to address all issues and concerns prior to adoption of the policy. The result was a carefully balanced policy and supporting procedures and tools, including an author addendum to notify a publisher of the policy’s existence, an automated waiver form for a researcher to waive the open access (but not the deposit) requirement of the policy, and a cover sheet on repository items to support version awareness and accuracy in citation. The repository now has full-text deposits of over 1100 research publications, approximately 400 of which are items published after adoption of the open access policy. As illustrated on Antelman case study Open access publishing may reach more readers, the following Figure 2 shows the effect of policy implementation on usage of the repository content.\n\nKAUST repository usage captured from Google Analytics.\n\n\nRepository integration with external systems\n\nThe adoption of the KAUST open access policy was just part of a broader expansion of the KAUST research repository towards being a reliably comprehensive resource for all of the university’s research outputs. This expansion has necessitated efforts to track the appearance of KAUST research in external systems and to bring information from external systems into the repository in order to properly contextualize the archived research.\n\n\nUse of ORCID IDs\n\nOne significant effort has been the integration of ORCID IDs into the KAUST repository. This began in August 2014 with the addition of features to our DSpace repository software allowing for the entry and display of ORCID IDs. This development work was undertaken by our hosting provider Biomed Central and consisted of the following elements:\n\nAddition of ORCID IDs along with the names of various contributor types (authors, thesis/dissertation advisors, and thesis/dissertation committee members) either through the item submission form or through the metadata editing functions.\n\nOptions for retrieval of ORCID IDs during item submission, either via Crossref (if ORCID IDs are available in the metadata associated with a DOI) or via the ORCID public API (based on name search and selection).\n\nDisplay of ORCID IDs on the individual item pages and in the browse authors listing along with links to search the repository by the ORCID ID and to the matching ORCID profile.\n\nAt this time we also instituted a policy of requiring ORCID IDs to be added for the student authors of newly deposited theses and dissertations in the repository. User expectations that use of ORCID IDs would have additional benefits (such as the automatic addition of thesis information to ORCID profile pages) contributed to our decision at this time to pursue ORCID membership. Our goals for institutional membership in ORCID included being able to assist all KAUST-affiliated authors with the registration of ORCID IDs and also help them derive benefits from the use of ORCID by making it easier to maintain up-to-date publication lists in their public profiles by pushing work information from the KAUST repository into their ORCID profiles.\n\nWe then explored the relative benefits of creating and hosting our own integration with the ORCID member API or pursuing integration through our repository provider (Biomed Central’s Open Repository) or through our CRIS provider (Elsevier’s Pure). After investigation we decided that developing our own integration would allow us to more rapidly make effective use of our ORCID membership while preserving our options for moving to as-yet-to-be-developed integrations from commercial providers at a later date.\n\n\nUse of the ORCID member API\n\nOur internally developed integration with the ORCID member API provides a web interface built with PHP on a MySQL backend through which users can interact with the ORCID system to create an ORCID ID and grant permissions to KAUST to add work and affiliation information to their ORCID profiles. We launched this service in January 2015 with emails to faculty explaining the benefits of ORCID and providing a custom link to our intranet tool for creation of a new ORCID ID or connection of an existing ORCID ID. We then followed up with additional email communication to those who did not use the tool in response to the first email. We have since followed the same process with postdoctoral researchers and research scientists at KAUST. Our current uptake rates are 82% for faculty, 52% for postdocs and 13% for research scientists (who were first contacted at the end of March). Our plan is to implement a similar process for current students while also working internally to include ORCID creation in the procedures for incoming faculty, staff and students in the future.\n\nThe functionality of adding work information from the KAUST repository to ORCID profiles by using the KAUST/ORCID integration tool has been well received and has prompted some individuals to deposit works in the repository. However, users have also indicated a desire for additional functionality, allowing them to easily reuse their ORCID work lists on their personal web pages or in other online profiling systems.\n\n\nInclusion of sources of research usage metrics\n\nAs we explore ways to derive benefits from ORCID use, we have been looking specifically at how external systems can be leveraged to answer questions about how, and how much, KAUST research outputs are being used. The repository statistics themselves are derived from Google Analytics and provide indicators of the most used items within the repository. On our request our repository provider has also added Scopus citation counts and the Altmetric.com donut badge to the display page for individual items in the repository. In order to have a more comprehensive view of research usage we are also procuring a subscription to Ebsco’s PlumX service.\n\nWe made the decision to use PlumX after evaluation of it in comparison with a similar product from Altmetric.com called Altmetric for Institutions. While there is some overlap in the metrics types and sources used by the two services (primarily those from social media sites), we also found significant differences. Altmetric.com has emphasized finding meaningful mentions of published articles and conference papers in the news media, on blogs, and in government documents. However, the broader net that PlumX throws in other areas struck us as likely to give us a more comprehensive view of the varied types of research output produced by KAUST authors (such as computer code, datasets, videos and presentations), and also the varied metrics associated with their usage (downloads, views, citations, etc.). PlumX’s use of the ORCID public API to remain regularly updated with new works added to a researcher’s profile also presents an attractive method of showing researchers a benefit of using the KAUST repository, ORCID and the KAUST/ORCID integration.\n\n\nMoving towards use of a CRIS system\n\nA key feature of our future university research tracking and evaluation system will be a current research information system that may become the central locus for information exchange between these varied systems (the repository, ORCID, PlumX, etc.). The Research Evaluation Office has taken the lead in this area by purchasing the Pure system from Elsevier. The initial work has focused on integrating with KAUST’s internal information systems, after completion of which the groundwork will be laid for university-wide use and implementation of connections to external systems to better assess the quality of research done by our faculty. Indeed, we have more than 130 professors coming from all around the world, bringing with them their personal academic and bibliographical history. As a consequence, accurately collecting all the present and past information is a real challenge. Additionally, most of our researchers are involved in collaborations with top scientists worldwide. This brings another degree of complexity when dealing with the measurement of outcomes and their impact.\n\nThe integration of ORCID in our system will help us to reduce the errors of affiliations of researchers by giving them a unique and centrally managed identifier. We are starting to use the ORCID identifier connected to our internal reference as the entry point for our users in the Pure system. To evaluate researcher publication records we collect a full list of publications from the researchers themselves and check their validity in Web of Science, Scopus and other online databases. We include the publications prior to their start at KAUST in order to assess the relative productivity of researchers since their arrival compared to their previous career. Unfortunately, the common online tools such as Scopus, Scival, Web of Science or InCites do not have the flexibility to refine the analysis to include start and end dates of affiliations and other research outputs, such as grey literature.\n\nAs an example, we present here an analysis of the number of citations per publication for institutional, national, international and corporate collaborative outputs for 86 Institutions that had significant collaborations (generating 10 publications or more) in place with KAUST researchers over the period of 2011–2013. The sample set of institutions was diverse in both geographical location and international ranking and as such provided a representative world average.\n\nOn average the quality/impact of a publication (as measured in citations per article) was higher when the publication included a national (4.8) or international (7.4) collaborator (Figure 3). Of particular note, those publications that included a corporate collaborator (10.3) were on average the most impactful, having the highest average citations per publication, while institutional collaborations (3.6) had the lowest number of citations. These results demonstrate that international scientific collaborations with KAUST can be seen as drivers to international impact and prestige, and that such an impact increases with the number of partners and the industry involvement within those collaborations.\n\nIn terms of research assessment, the main key indicators that we are using are the Field Weighted Citation Impact, the average number of citations per publication, the percentage of publications in top percentile journals and the number of collaborative publications. In this context, we have compiled results for several universities to benchmark KAUST against other similar institutions, as presented in Figure 4.\n\nThanks to the library repository, the central CRIS system and our internal analysts, those results have allowed the senior management to understand the quality of the research performed at KAUST as well as its impact at the international level. As a consequence, our researchers are also taking a more active part in the collection of data about their publications and other research outputs.\n\n\nCommunity engagement: role of the university library\n\nThe roles of the university library are changing in relation to scholarly communications and academic research practices [Malenfant, 2010]. At KAUST University Library we are actively exploring new roles that we can play in relation to storage and dissemination of all forms of scholarly research with an overall goal of increasing the impact that KAUST research has in the global research community. We also support university stakeholders as they gather information about the impact of KAUST research and select tools to evaluate this information. A key element of this work is interacting with students, researchers and faculty to introduce the available tools for measuring and monitoring the impact of research, while also answering questions about how, when and why to use a given tool.\n\nAlong with helping individuals with queries the library has also developed a program of scheduled workshops to be held every semester. Below is a summary of the relevant areas covered in these sessions:\n\nValue of using citation databases in the literature search\n\nUnderstanding citation metrics and tools (h-Index, Impact Factor, Altmetrics etc.)\n\nRole of publications in effecting institutional rankings\n\nUnderstanding researcher profiling (ORCID, Google Scholar etc.)\n\nBenefits of open access and explanation of how to use the institutional repository\n\nAcademic honesty, plagiarism, and the use of similarity checking tools\n\nFeedback received from participants shows that they found the trainings beneficial to their understanding of how to communicate their research. Based on the positive feedback received, the library is planning to offer additional training to support researchers in navigating confusing and complex aspects of the changing scholarly communications landscape while connecting the pieces of the research life cycle, and thus enhancing the profile of individual researchers, increasing the impact of their research and improving the reputation of our institution.\n\nThe KAUST Library also conducts regional outreach in the Middle East to promote open access and research service initiatives. We shared our experiences at the Special Library Association Arab Gulf Chapter conferences in two panels (2013 and 2014) and at the American Library Association’s Sharjah Book Fair conference in 2014. We wish to direct the attention of professional librarians and the academic community in the region towards better sharing and management of information regarding their own institutions’ research activities.\n\n\nConclusion\n\nKAUST also believes that good practices in research data management (RDM) are going to be a key part of responsible research moving forward. We expect to have an increasing role in facilitating the management and sharing of research data in ways aligned with the KAUST mission and beneficial to the global research community. There remain numerous questions to address regarding data storage, sharing, reuse and citation, along with issues of varied institutional and funder requirements in terms of data governance. Services and initiatives around data infrastructure, stewardship, and management support will be important next steps for KAUST in connecting the pieces of scholarly communication.\n\n\nData availability\n\nF1000Research: Dataset 1. Unique repository materials citations, 10.5256/f1000research.6502.d50203\n\nF1000Research: Dataset 2. Bibliometric benchmarking of KAUST University compared to other similar institutions, 10.5256/f1000research.6502.d50204", "appendix": "Author contributions\n\n\n\nMB and DG collected information about KAUST repository establishment, policy integration and tools; TL collected information and present information about research evaluation; JK collected information about community engagement. All authors wrote the manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare no competing interests.\n\n\nGrant information\n\nThe authors declared that no funds were involved in supporting this work.\n\n\nAcknowledgements\n\nResearch and results reported in this publication were supported by the King Abdullah University of Science and Technology (KAUST). In developing the ideas presented here, Thibaut Lery has received helpful input from Dr. Manus Ward, Fadzai Chikwava, Lina Ekere and Eirini Mastoraki.\n\n\nReferences\n\nAntelman K: “Do open-access articles have a greater research impact?” College & Research Libraries. 2004; 65(5): 372–382. Publisher Full Text\n\nBaessa M, Lery T, Grenz D, et al.: Dataset 1 in: Connecting the pieces: Using ORCIDs to improve research impact and repositories. F1000Research. 2015. Data Source\n\nBaessa M, Lery T, Grenz D, et al.: Dataset 2 in: Connecting the pieces: Using ORCIDs to improve research impact and repositories. F1000Research. 2015. Data Source\n\nMalenfant KJ: Leading change in the system of scholarly communication: A case study of engaging liaison librarians for outreach to faculty. College & Research Libraries. 2010; 71(1): 63–76. Publisher Full Text" }	[ { "id": "9385", "date": "13 Jul 2015", "name": "Sarah L. Shreeves", "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 article outlines the infrastructure in place at KAUST to support both research outputs, particularly those that are not formally published, and research metrics. While the tools themselves are not new or unique, the implementation at a very young university (established in 2009) and the integration of ORCID into multiple systems makes this article acceptable for the Software Tool section.The abstract is appropriate. The datasets for figures were provided.The article itself is descriptive rather than research, but provides the appropriate context and information for understanding how the systems and tools interact. I would suggest a small change to the title which concentrates on ORCID - the article itself is a bit more broad ranging than that and it might be more appropriate to have a broader title.", "responses": [] }, { "id": "10068", "date": "01 Sep 2015", "name": "Antonella De Robbio", "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 will suggest a slight modification of the title from “Connecting the pieces: Using ORCIDs to improve research impact and repositories” to “Connecting the pieces to improve research impact and repositories: Moving towards use of a CRIS system”The abstract provides an adequate summary of the article.The paper is coherent and good structured, comprehensive explanation of argument treated. Level of writing is appropriate for journal’s audience, also regarding the style and terminology, it is clear it gives a good overview on management and organizational aspects. The paper offers a good level on overviews on IR inside Open Access context: principles, strategy, channels and tools are good defined. Information on policy and infrastructures are very useful for other institutions. Data and figures presented updated and well embedded inside the context. Results are presented accurately and the conclusions are justified and supported by the data.", "responses": [] } ]	1	https://f1000research.com/articles/4-195
https://f1000research.com/articles/4-194/v1	07 Jul 15	{ "type": "Systematic Review", "title": "Systematic Review and Meta-analysis of Prevalence of Coronary Artery Disease in Adult Patients with Cardiac Myxomas", "authors": [ "Matheus Silva", "Matheus Carneiro", "Júlio Nunes", "Antônio da Silva", "Marcos de Sousa", "Matheus Carneiro", "Júlio Nunes", "Antônio da Silva", "Marcos de Sousa" ], "abstract": "Background: Studies have reported varied prevalence estimates of coronary artery disease (CAD) in cardiac myxoma patients. We performed a systematic review and meta-analysis of observational studies to summarize the point prevalence of CAD in adults with cardiac myxomas. Methods and Results: Two independent investigators searched MEDLINE and LILACS databases using the terms \"Myxoma”, \"Coronary Angiography\" and \"Coronary Disease\" from inception through December 2014 for all relevant studies. We included 6 observational studies. Publication bias was evaluated through Egger's test and Trim and Fill method. A pooled estimate of CAD prevalence with corresponding 95% confidence interval (CI) was calculated based on a random-effects model. The pooled CAD prevalence in adult cardiac myxoma patients was 20.7% with low heterogeneity (I2 = 14.86%). Conclusions: It is a matter of debate if preoperative coronary angiography must be done as a routine procedure. Although coronary disease and angiographically detectable neovascularity can alter surgical management, more studies are needed to evaluate this question.", "keywords": [ "Systematic review", "meta-analysis", "coronary artery disease", "coronary angiography", "myxoma" ], "content": "Introduction\n\nMyxomas are the most common primary cardiac tumors, although extremely rare. As an example, in one series of over 12,000 autopsies, only two were identified, for an incidence of less than 0.02 percent1. Histologically, these tumors are composed of scattered cells within a mucopolysaccharide stroma. The cells originate from a multipotent mesenchyme that is capable of endothelial and neural differentiation2. Myxomas produce vascular endothelial growth factor, which probably induces angiogenesis for tumor growth3.\n\nMacroscopically, the tumor surface can be smooth, friable or villous. The tumor diameter varies, ranging from 1 to 15 cm, with a weight typically between 15 and 180 g (mean, 37 g). Friable tumors are more prone to embolization, while larger tumors present with cardiovascular symptoms4.\n\nThe mean age of patients with myxomas is 56 years and 64–70% are females. However, myxomas have been described in patients ranging in age from 3 to 84 years. Approximately 86% of all myxomas occur in the left atrium, and most of the remainder is found in the right atrium. Over 90% are solitary4,5.\n\nThe cardiovascular manifestations depend upon the anatomic location of the tumor. In a series of 112 consecutive cases of left atrial myxoma: (1) cardiovascular symptoms were present in 67%, more commonly in the form of mitral valve obstruction (mostly cardiac failure or malaise). Cardiac auscultation abnormalities occurred in 64%, essentially pseudo-mitral valve disease in 53.5% and more rarely the suggestive tumor plop in 15%. The most frequent electrocardiographic sign was left atrial hypertrophy in 35%, whereas arrhythmias were uncommon. (2) Embolic symptoms were observed in 29%, essentially cerebral emboli with stroke, with men at greater risk. (3) Constitutional symptoms were observed in 34% with fever, weight loss, or symptoms resembling connective tissue disease4. Right atrial tumors are more commonly associated with signs and symptoms of right heart failure. Tumor fragments can embolize to pulmonary vasculature and cause symptoms consistent with pulmonary emboli, or in the presence of a patent foramen ovale or atrial septal defect, hypoxemia or systemic emboli6,7.\n\nEchocardiography is a widely available, simple and noninvasive approach, which in almost all cases precisely locates the tumor and defines its extent. In addition, transesophageal echocardiography (TEE), cardiac magnetic resonance (MRI) and ultrafast computed tomography (CT) have also proved their usefulness in diagnosis8,9.\n\nOnce a presumptive diagnosis of a cardiac myxoma is made, surgical removal is indicated because of the risk of embolization or of sudden cardiac death. The prognosis for patients with solitary myxomas after surgical resection has been excellent with mortality rates of about 4%. Late recurrences are infrequent and reported to occur in 0.4–5% of patients10.\n\nSeveral studies have attempted to estimate the rate of cardiac myxomas with concomitant CAD and we therefore conducted a systematic review and meta-analysis of observational studies to summarize the point prevalence of CAD in adults with these tumors.\n\n\nMethods\n\nWe carried out a systematic review and meta-analysis of prospective and retrospective observational studies following the PRISMA statement (Supplementary Material S1)11. Initially, a search in the main databases (MEDLINE, The Cochrane Library, and LILACS) was performed, searching for studies with similar objectives and methodology. No similar study was found.\n\nA systematic MEDLINE search was performed with the medical subject headings (MeSH) terms (“Myxoma”[MeSH] AND “Coronary Angiography”[MeSH]) OR (“Myxoma”[MeSH] AND “Coronary Disease”[MeSH]), looking for trials in English, Spanish and Portuguese, published until December 2014, that performed coronary angiography in patients with cardiac myxomas. At the same time, a systematic LILACS search was also performed using the same MeSH terms and search strategy.\n\nWe designed a relatively strict set of inclusion and exclusion criteria and considered studies meeting these criteria to be of acceptable quality. The study selection criteria were: (1) observational studies, with prospective or retrospective data collection; (2) studies that provided a measure of CAD prevalence in adult patients with cardiac myxomas; (3) studies that included at least five cases of cardiac myxomas; (4) studies in which at least 75% of the adult cardiac myxoma patients had coronary angiographies; (5) angiographic and demographic data systematically reported.\n\nTwo researchers, according to the previously established inclusion criteria, then independently reviewed the titles returned by the systematic search. Exclusion by duplicity, title, abstract and full text analyses was independently performed and discrepancies in each stage were solved by consensus after discussion. The selected articles were read in full to confirm eligibility and their data was tabulated and reviewed for the statistical analysis. The second researcher independently double-checked the extraction of primary data from every study.\n\nThe meta-analysis of the pooled prevalence data, as well as associated graphic results was performed using the Comprehensive Meta Analysis software, version 2.2.064. Other computations were performed with IBM SPSS Statistics for Macintosh, Version 22.0.\n\nHeterogeneity of accuracy measures was explored with the I2 estimate (inconsistency measure) from Cochran Q according to the formula: I2 = 100% x (Cochran Q – degrees of freedom)/Cochran Q. This describes the percentage of the variability in effect that is due to heterogeneity rather than sampling error (chance)12. Publication bias was graphically assessed using funnel plot, Egger's test and Trim and Fill method13–15.\n\n\nResults\n\nA flow chart of the studies evaluation is shown in Figure 1. These latter studies were excluded because of a lack of angiographic data in at least 75% of patients with cardiac myxomas in the populations examined. Thus, a total of 6 studies evaluating the prevalence of CAD were selected according to the aforementioned criteria.\n\nFor each study, demographic characteristics, the proportion of adult cardiac myxoma patients who underwent coronary angiography and the location of the tumors are listed in Table I. The criteria used to define the presence of CAD and associated treatment when described is listed in Table II. The prevalence rates of clinically confirmed CAD for each of the 6 studies are reported in Figure 2.\n\nAbbreviations: CA, coronary angiography, LA, left atrium; RA, right atrium; NP, not provided.\n\nAbbreviations: CAD, coronary artery disease; LCx, left circumflex coronary artery; CABG, coronary artery bypass graft surgery; LMCA, left main coronary artery; RCA, right coronary artery; LAD, left anterior descending coronary artery; LA, left atrium; RA, right atrium; NP, not provided.\n\nHorizontal lines represent 95% confidence intervals (CIs). Each box represents the prevalence rate point estimate, and its area is proportional to the weight of the study determined by inverse variance weighting. The diamond represents the overall summary estimate, with the 95% CI given by its width.\n\nAs shown, we found an aggregated estimate of 20.7% (95% CI 0.12 to 0.32). The prevalence rates reported across these studies varied from 5.26%16 to 36.26%5 (Figure 2), with low heterogeneity (Q-value = 5.873, P-value = 0.319, I2 = 14.86%). Egger's test (two tailed) was borderline positive for publication bias (P = 0.047). A funnel plot with Trim and Fill method is shown in Figure 3. The close observed and adjusted values of pooled prevalence suggest a small influence of publication bias.\n\nThis is a display of the study’s effect size on a logit scale against its precision for each study included in the meta-analysis. Egger's test P = 0.047. Trim and Fill method showed close observed (20.7% with 95% CI 12.4 to 32.3%) and adjusted (22.7% with 95%CI 13.0 to 36.5%) values.\n\n\nDiscussion\n\nThere is little documented literature on the relationship between CAD and cardiac myxomas8. This article provides the first compilation of available angiographic data on adult patients with cardiac myxomas and CAD. According to this meta-analysis, the estimated prevalence of CAD in adult patients with myxomas is 20.7%, with low heterogeneity. In observational studies, Van Cleemput et al.16, Ergünes et al.17 and Gismondi et al.18 found that CAD prevalence accounted for 5.3%, 7.1% and 16.6% of adult patients with myxoma, respectively. Shapiro et al.19 found a CAD prevalence of 25% and Erdil et al.5 of 36.4% of adult patients with myxomas. However, in these studies, the number of index patients considered was small, 7 and 11, respectively. Rahmanian et al.20 evaluated 23 out of 28 adult myxoma patients with coronary angiography and found a CAD prevalence of 26.1%; nevertheless, this study enrolled the eldest patients of our series of studies (mean age, 61.3 years).\n\nConcomitant coronary artery bypass grafting (CABG) surgery with resection of the tumor can be crucial in patients with critical coronary lesions. Erdil et al.5 identified 4 patients out of 11 with concomitant CAD, 3 of which had adjuvant CABG performed, and a fourth, which had a noncritical lesion in the right coronary artery and was treated medically. Rahmanian et al.20 also reported that in 6 out of 23 patients significant CAD was found leading to percutaneous angioplasty and stent placement in 3 patients, and surgical revascularization during mass excision in the remaining three patients. Indeed, out of the 14 patients with CAD identified in the 6 studies included in this meta-analysis, 12 (86%) were subject to, or were liable to invasive treatment.\n\nThe use of preoperative coronary angiography (CA) in adult myxoma patients is a topic of debate. Some argue that CA should only be performed in selected patients, particularly those aged > 35–40 years, with atherosclerotic risk factors, a positive anginal history or with a previous history of myocardial infarction to rule out concomitant coronary artery disease16,20–23. However, many report that there has been no significant difference in symptoms, age or prevalence of coronary risk factor distribution between myxoma patients who present with CAD and those who do not5,18,24,25. Actually, even patients without any risk factors can present with CAD25. Therefore, others suggest that all adult patients diagnosed with myxomas should undergo CA5,8,18,25,26. In fact, preoperative CA seems to be quite safe; thus far there has been no report of procedure-related complications8,16,18,24–27.\n\nPreoperative CA can yield even more information that may prove useful intraoperatively8. Selective CA occasionally may visualize the tumor by revealing the angiographic sign of ‘tumor vascularity’, first described by Marshall et al.28, which consists of clusters of small and tortuous vessels with blood pooling and tumor blush arising from the coronary arteries supplying the tumor26,27. From the experience of Van Cleemput et al.16 and the data published by Fueredi et al.27 and Chow et al.26, angiographically visible neovascularity is prevalent in around 40% of symptomatic cardiac myxoma patients. This finding suggests a tumoral origin of the mass, however not specific. Systematic performance of preoperative CA has been recommended by some authors in an attempt to identify a large supplying vessel20.\n\nFailure to identify and ligate these vessels may lead to a coronary-cavitary fistula29 or a “steal syndrome”, by re-directing blood from a coronary artery into a cardiac chamber, with consequent myocardial ischemia8.\n\nOur study has limitations. As not all patients in each study had performed coronary angiography, the pooled prevalence can be overestimated by verification bias. Patients not selected to perform CA are probably those with a low risk profile. To overcome this limitation, we decided to not include those studies that reported less than 75% of patients submitted to CA. Since it is a rare condition, samples are small. Also, number of studies is small. On the other hand, a systematic review is one way to gather evidence on rare conditions. Although the publication bias test was borderline positive, the Trim and Fill method suggested small or no influence of publication bias on pooled results. Even if it is still present, prevalence could be higher, since the adjusted value is slightly higher than the observed value.\n\n\nConclusion\n\nRoutine preoperative angiography in all cases of these tumors is still a matter of debate. Pooled prevalence of coronary disease and the potential to disclose angiographically detectable neovascularity are arguments to advocate routine angiography. Patient management and clinical outcomes could be potentially altered, but more studies are needed to answer this question.", "appendix": "Author contributions\n\n\n\nMCS co-conceived the study, participated in the design of the study, search strategy execution, performance of the statistical analysis, and writing the manuscript.\n\nMSC participated in the search strategy execution, performance of the statistical analysis, and writing the manuscript.\n\nJTN participated in the acquisition of data, performance of the statistical analysis, and writing the manuscript.\n\nACBS participated in the acquisition of data, performance of the statistical analysis, and writing the manuscript.\n\nMRS co-conceived the study, participated in the design of the study, search strategy execution, performance of the statistical analysis, and writing the manuscript.\n\n\nCompeting 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\nReferences\n\nLam KY, Dickens P, Chan AC: Tumors of the heart. A 20-year experience with a review of 12,485 consecutive autopsies. Arch Pathol Lab Med. 1993; 117(10): 1027–31. PubMed Abstract\n\nPucci A, Gagliardotto P, Zanini C, et al.: Histopathologic and clinical characterization of cardiac myxoma: review of 53 cases from a single institution. Am Heart J. 2000; 140(1): 134–8. PubMed Abstract \| Publisher Full Text\n\nKono T, Koide N, Hama Y, et al.: Expression of vascular endothelial growth factor and angiogenesis in cardiac myxoma: a study of fifteen patients. J Thorac Cardiovasc Surg. 2000; 119(1): 101–7. PubMed Abstract \| Publisher Full Text\n\nPinede L, Duhaut P, Loire R: Clinical presentation of left atrial cardiac myxoma. A series of 112 consecutive cases. Medicine (Baltimore). 2001; 80(3): 159–72. PubMed Abstract\n\nErdil N, Ates S, Cetin L, et al.: Frequency of left atrial myxoma with concomitant coronary artery disease. Surg Today. 2003; 33(5): 328–31. PubMed Abstract \| Publisher Full Text\n\nKuon E, Kreplin M, Weiss W, et al.: The challenge presented by right atrial myxoma. Herz. 2004; 29(7): 702–9. PubMed Abstract \| Publisher Full Text\n\nDiaz Castro O, Bueno H, Nebreda LA: Acute myocardial infarction caused by paradoxical tumorous embolism as a manifestation of hepatocarcinoma. Heart. 2004; 90(5): e29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJanas R, Jutley RS, Fenton P, et al.: Should we perform preoperative coronary angiography in all cases of atrial myxomas? Catheter Cardiovasc Interv. 2006; 67(3): 379–83. PubMed Abstract \| Publisher Full Text\n\nVander Salm TJ: Unusual primary tumors of the heart. Semin Thorac Cardiovasc Surg. 2000; 12(2): 89–100. PubMed Abstract\n\nKeeling IM, Oberwalder P, Anelli-Monti M, et al.: Cardiac myxomas: 24 years of experience in 49 patients. Eur J Cardiothorac Surg. 2002; 22(6): 971–7. PubMed Abstract \| Publisher 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 healthcare interventions: explanation and elaboration. BMJ. 2009; 339: b2700. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacaskill P, Gatsonis C, Deeks J, et al.: Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy. Version 09 0 London. The Cochrane Collaboration, 2010. Reference Source\n\nSterne JAC, Harbord RM: Funnel plots in meta-analysis. Stata Journal. 2004; 4(2): 127–41. Reference Source\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\nBorenstein M, Hedges LV, Higgins JPT, et al.: Introduction to meta-analysis. Chichester, West Sussex, U.K; Hoboken: John Wiley & Sons. 2009, xxviii. 421p. Publisher Full Text\n\nVan Cleemput J, Daenen W, De Geest H: Coronary angiography in cardiac myxomas: findings in 19 consecutive cases and review of the literature. Cathet Cardiovasc Diagn. 1993; 29(3): 217–20. PubMed Abstract \| Publisher Full Text\n\nErgunes K, Yetkin U, Yilik L, et al.: Diagnosis and surgical treatment modalities in cardiac myxomas. Anadolu Kardiyol Derg. 2008; 8(5): 379–80. PubMed Abstract\n\nGismondi RAOC, Martino H, Soares RdC, et al.: Doença arterial coronariana e mixoma cardíaco: prevalência e associação com os fatores de risco convencionais (Coronary artery disease and cardiac myxoma: prevalence and association with conventional risk factors). Rev SOCERJ. 2007; 20(1): 16–9. Reference Source\n\nShapiro JB, Kronzon I, Winer HE: Diagnosis of left atrial tumors by coronary angiography and left ventriculography. Cathet Cardiovasc Diagn. 1979; 5(1): 41–9. PubMed Abstract \| Publisher Full Text\n\nRahmanian PB, Castillo JG, Sanz J, et al.: Cardiac myxoma: preoperative diagnosis using a multimodal imaging approach and surgical outcome in a large contemporary series. Interact Cardiovasc Thorac Surg. 2007; 6(4): 479–83. PubMed Abstract \| Publisher Full Text\n\nRice PL, Pifarré R: Left atrial myxoma and coronary artery disease: combined surgical treatment. Arch Surg. 1981; 116(3): 353–5. PubMed Abstract \| Publisher Full Text\n\nShimono T, Makino S, Kanamori Y, et al.: Left atrial myxomas. Using gross anatomic tumor types to determine clinical features and coronary angiographic findings. Chest. 1995; 107(3): 674–9. PubMed Abstract \| Publisher Full Text\n\nSharma S, Sundaram U, Loya Y: Selective coronary angiography in intracardiac tumors. J Interv Cardiol. 1993; 6(2): 125–9. PubMed Abstract \| Publisher Full Text\n\nHuang CY, Yu WC, Chen KC, et al.: Coronary angiography of cardiac myxoma. Clin Cardiol. 2005; 28(11): 505–9. PubMed Abstract \| Publisher Full Text\n\nLi AH, Liau CS, Wu CC, et al.: Role of coronary angiography in myxoma patients: a 14-year experience in one medical center. Cardiology. 1999; 92(4): 232–5. PubMed Abstract \| Publisher Full Text\n\nChow WH, Chow TC, Tai YT, et al.: Angiographic visualization of 'tumour vascularity' in atrial myxoma. Eur Heart J. 1991; 12(1): 79–82. PubMed Abstract\n\nFueredi GA, Knechtges TE, Czarnecki DJ: Coronary angiography in atrial myxoma: findings in nine cases. AJR Am J Roentgenol. 1989; 152(4): 737–8. PubMed Abstract \| Publisher Full Text\n\nMarshall WH Jr, Steiner RM, Wexler L: \"Tumor vascularity\" in left atrial myxoma demonstrated by selective coronary arteriography. Radiology. 1969; 93(4): 815–6 passim. PubMed Abstract \| Publisher Full Text\n\nD'Avila AL, Passos LC, Hueb WA, et al.: [Coronary-cavitary fistula after resection of vascularized left atrial myxoma]. Arq Bras Cardiol. 1991; 57(6): 487–8. PubMed Abstract" }	[ { "id": "9393", "date": "24 Aug 2015", "name": "Guilherme Costa", "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\nDefinitely an important and clinically relevant topic. Through accurately performed meta-analysis and systematic review, the authors remind us of a rare and usually underrated - though certainly morbid - condition which could be managed differently if better researched on.", "responses": [] }, { "id": "22954", "date": "31 May 2017", "name": "Christian Röver", "expertise": [ "Reviewer Expertise Statistics" ], "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 meta-analysis seems to be performed properly, I only have a few remarks concerning mostly the presentation of results.\n\nMethods section:\n1) It would suggest mentioning explicitly here that:\n\na) a random-effects model was used.\n\nb) the DerSimonian-Laird heterogeneity estimator was used.\n\nc) a logit- (or log-odds-) transformation was used for the effects.\n\nResults section:\n2) The forest plot (Fig.2) is somewhat confusing:\n\na) the x-axis should not extend beyond 0 and 1.\n\nb) the p-values (and z-values) shown apparently relate to a test of zero log-odds (i.e. a prevalence of 50%), which does not seem to make sense here; these should be dropped.\n\nc) it would be good to also show the numbers of events and totals (CAs and CADs) either here or in one of the tables.\n3) I do not see why a selection effect would be plausible, and why one would perform Eggers' test here. Would it be plausible that certain results would not be accessible because they would exhibit either a very high or very low CAD prevalence? If not, then the test (and the funnel plot) should probably be omitted.\n\nGeneral minor comments:\n4) In order to put the results into context, I would suggest to also hint on the order of magnitude of CAD prevalence in non-myxoma populations, or on potential CAD prevalence thresholds that might warrant consequences.\n5) The studies in tables I and II (last two rows) are in a different order than in Fig.2\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? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [] } ]	1	https://f1000research.com/articles/4-194
https://f1000research.com/articles/4-193/v1	07 Jul 15	{ "type": "Opinion Article", "title": "Problematizing Digital Research Evaluation using DOIs in Practice-Based Arts, Humanities and Social Science Research", "authors": [ "Muriel Swijghuisen Reigersberg" ], "abstract": "This paper explores emerging practices in research data management in the arts, humanities and social sciences (AHSS). It will do so vis-à-vis current citation conventions and impact measurement for research in AHSS. Case study findings on research data inventoried at Goldsmiths’, University of London will be presented. Goldsmiths is a UK research-intensive higher education institution which specialises in arts, humanities and social science research. The paper’s aim is to raise awareness of the subject-specific needs of AHSS scholars to help inform the design of future digital tools for impact analysis in AHSS. Firstly, I shall explore the definition of research data and how it is currently understood by AHSS researchers. I will show why many researchers choose not to engage with digital dissemination techniques and ORCID. This discussion must necessarily include the idea that practice-based and applied AHSS research are processes which are not easily captured in numerical ‘sets’ and cannot be labelled electronically without giving careful consideration to what a group or data item ‘represents’ as part of the academic enquiry, and therefore how it should be cited and analysed as part of any impact assessment. Then, the paper will explore: the role of the monograph and arts catalogue in AHSS scholarship; how citation practices and digital impact measurement in AHSS currently operate in relation to authorship and how digital identifiers may hypothetically impact on metrics, intellectual property (IP), copyright and research integrity issues in AHSS. I will also show that, if we are to be truly interdisciplinary, as research funders and strategic thinkers say we should, it is necessary to revise the way we think about digital research dissemination. This will involve breaking down the boundaries between AHSS and other types of research.", "keywords": [ "Open Access", "Digital Identifiers", "Arts", "Humanities", "Social Science" ], "content": "Introduction\n\nThis paper explores emerging practices in research data management in the arts, humanities and social sciences (AHSS). It will do so vis-à-vis current citation conventions and impact measurement for research in AHSS. Case study findings on research data inventoried at Goldsmiths’, University of London will be presented. Goldsmiths is a UK research-intensive higher education institution which specialises in arts, humanities and social science research.\n\nThe subject of this paper is a topical one in the UK, where research in Universities is publically funded. Government and research council funders are asking that Universities in receipt of research income demonstrate how their funding is used to generate new knowledge and positive impact in all disciplines, including the arts, humanities and social sciences. The impact that this new knowledge creation and its dissemination are having must be recorded and where possible, quantified. This quantitative information can thereafter be used to help inform future research strategies on a variety of levels. It might also be used, some tentatively suggest, to complement peer review in future research excellence frameworks.\n\nSome of this quantifiable information for research assessment might be delivered through various types of metrics, based on digital information made openly available. This could include bibliometrics; a quantitative analysis of research literature and citation rates or altmetrics, which incorporates for example social media analyses and download rates of visual research-related materials, alongside citation rates.\n\nThis way of measuring research impact and excellence however, is not yet as refined as some researchers might like it to be. Firstly, the data on which metrics relies must be available in a digital format and easily accessible. Secondly, data must be unambiguously linked to its creator(s) through unique researcher identifier numbers, such as those provided by ORCID. Some argue metrics works better for those disciplines that have focussed more heavily on digital dissemination strategies and open access publication methods. It is argued that some forms of research, such as practice-based research, do not naturally lend themselves well to digital capturing for impact measuring purposes. The upshot of this is that those disciplines that are less digitally oriented, are likely to obtain unhelpful metric ratings. This in turn it is feared, will lead to reductions in public funding, if metrics were to be used to allocate financial resources in future. Some even suggest that this then in turn, might jeopardise the diversity of UK research, diminishing arts and practice-based research activity and jeopardising the sustainability of smaller specialist higher education institutions. As a result, the Higher Education Funding Council for England (Hefce) conducted two independent reviews on the suitability of metrics for research and impact assessment purposes: one between 2008 – 2009, and another soon to be published in July 2015.\n\nDisciplinary languages also determine whether or not AHSS researchers are likely to engage with metrics. Many AHSS researchers do not define their research outputs as ‘data’. Neither do many communicate their research enquiries in writing, making bibliometrics problematic. Often, when they do write, AHSS researchers publish monographs or book chapters. Monographs as yet are not widely available in an open access format, again impacting on the ability of the research therein to be captured by digital tools. Lastly, questions of authorship, copyright and ownership arise. Often AHSS research is co-created with the help of non-researchers as well as co-investigators. This impacts on individual incomes and research integrity where the sharing and citation of data is concerned. It may also impinge on research data management strategies for interdisciplinary projects where multiple outputs of different kinds are produced.\n\nIt is the above issues that I will explore in this paper. I will argue that if we are to be truly interdisciplinary, it is necessary to revise the way we think about digital research dissemination and how we write and talk about it. This will involve breaking down the boundaries between AHSS and other types of research.\n\nAlthough presented here as an opinion piece, it is not so much an opinion as a record of the state of play with regards to emerging practices, theory and ideas on digital publication, ORCID numbers and digital object identifiers and how these might complement other mechanisms which enhance digital research impact and discoverability of AHSS research. This article does not pretend to offer a complete view of all emerging practices across AHSS disciplines in the UK. Neither does it suggest that digital citation and discoverability are the only ways in which impact can be achieved. Instead it will give an overview of some of the specific debates pertaining to digital dissemination that researchers and administrators are having at Goldsmiths, University of London and how staff are engaging with open access, open data, digital object identifiers (DOIs) and ORCID numbers.\n\nAt Goldsmiths, University of London, not many researchers have created their individualised ORCID numbers yet and neither do they, by and large, label their data with digital object identifiers (DOIs), apart from possibly their journal articles, which of course contain processed data. This is because many researchers active in arts, humanities and social science (AHSS) research do not identify their research outputs (e.g. field notes, monographs, art work sketches, film rushes, interview materials etc.) as research data. The term ‘data’, to them has a much narrower definition, belonging only to the professional jargon used by researchers working in (bio) medical and other scientific fields (e.g. quantitative, numerical data sets, graphs and pie charts). Other AHSS researchers see ‘data’ as only being synonymous with personal data: personal information about research participants, such as names and dates of birth, subject to safeguarding by the data protection act.\n\nThe definition of what might be classified as research data, however, could be much broader if conceived of creatively. The first mission of any digital advocate therefore, is to clarify with the help of researchers themselves, what might count as research data in their disciplines and how this could be labelled appropriately, digitised, archived and maintained as necessary, to document research processes and methods. This process of defining and scoping research data must meet the needs of project-specific research enquiries without letting administrative requirements take the upper hand or imposing non-AHSS data collection models where this is not appropriate.\n\nTo some extent the question of what counts as data in visual arts research was already addressed at Goldsmiths. Between 2011–2013, Goldsmiths participated in a Joint Information Systems Committee (JISC) funded project called KAPTUR together with Glasgow School of Art, University for the Creative Arts and University of the Arts London. KAPTUR built on the work undertaken by the Digital Curation Centre and was led by the Visual Arts Data Service. The KAPTUR project sought to: “investigate the current state of the management of research data in the arts; develop a model of best practice applicable to both specialist arts institutions and arts departments in multidisciplinary institutions; and to apply, test and embed the model with four institutional partners (http://www.vads.ac.uk/kaptur/about.html)”. The project helped question definitions of data and provide examples of what might count as data and how it might be managed.\n\nKAPTUR’s findings, useful toolkits, reports and outcomes have not filtered through to most visual arts researchers however and few therefore have embraced the idea that the definition of research data can be broadened. Those that have, however: are struggling with IP and copyright issues for digitised work; have no time to digitise analogue research outputs to be able to create digital object identifiers; or may not publish in the digital domain, meaning that digital identification and metrics cannot currently be used effectively to assess any potential impact being created.\n\nConsequently, many AHSS researchers feel ORCID numbers and digital object identifiers are simply not relevant to what they currently do. As much AHSS research remains unfunded there are also no contractual or funder obligations to which researchers must adhere which stipulate that research data must be open access or discoverable. This reluctance to engage with digital dissemination and citation practices is especially in evidence when researchers work in disciplines that still rely on the production of practice-based outputs and the publication of monographs as the most important esteem indicators and research outputs in their fields.\n\nGoldsmiths’ researchers publish many monographs. Monographs contain research data: images of art works; creative writing outputs; and (auto) biographical details for example. The author is not always the copyright owner of this data or indeed its creator, and will often have to gain permission to use information for the purposes of publishing their monograph. What is owned by the author is the intellectual theory and often text.\n\nAcademic monographs are still predominantly issued in non-open access formats. If published in the digital domain, proprietary formats and specific software are used. These are unlikely to promote sustainability of digital research data in the open access domain. The challenge is compounded by the current lack of digital research data for monographs that might easily be labelled and referenced electronically. Here it is worth referring to the January 2015 Crossick report on open access monographs generated by the Higher Education Funding Council England (Hefce).\n\nWhilst recognising the importance of monographs to AHSS scholars, the Hefce report identified the limitations of hard copy formats. Those relevant to us here include: a) the fact that video, audio and other examples cannot be embedded in hard-copy monographs; b) text-mining options and easy ways of measuring citation levels and impact rates are absent; c) the fact that hard copy monographs are not ‘living’ documents. Comments and reviews cannot be easily shared, updates require new editions and comparisons of passages and ideas are not quickly communicated. It is these limitations of the monograph that researchers at Goldsmiths are wishing to explore in collaboration with publishers, computing and legal experts.\n\nThe challenges identified by the Hefce report with regards to labelling research data and making monographs open access are several. Those highlighted by the report are, for example, that academically authored exhibition catalogues are part of business models for Independent Research Organisations (IROs) and galleries. Making exhibition catalogues and the research data openly accessible will reduce the vital income received by IROs such as the Tate Galleries. Additionally, data in catalogues and creative writing outputs have often been generated by people other than the catalogue author. This raises copyright, intellectual property and revenue challenges impacting on the licensing and sub-licensing of research data such as images and musical examples if researchers wanted to include certain materials in their open access monograph using digital object identifiers. Careful consideration must therefore be given to labelling research data before joining it to researcher ORCID numbers and making it open access if in monograph or catalogue format.\n\nQuestions of research ethics, integrity and licencing also come in to play when labelling practices are considered. Pictures, images and text may constitute to a representation of something or someone. Not the actual object or person. Practice-based researchers often argue that a (digital) image or recording of their analogue, and possibly temporaneous art work is a different object epistemologically to the actual, physical work and therefore cannot be labelled as being the same item. Similarly, where open access is an option, anthropologists and ethnographers frequently opt for non-derivative licences meaning no materials based data/text can be created derived from the original text. This is important where non-academic research collaborators agree to participate on the basis that they are represented fairly and where these collaborators often have a say in how their interview excerpts, musical materials and images (that is to say, research data collected by the researcher) are used and placed in texts. Re-using research data uncritically, or labelling it with object identifiers often runs contrary to the highly personalised material that is being explored, which belongs to both the researcher and his/her participants at the very least and in some cases to the research participant alone, who is sharing it with the researcher in good faith. AHSS authors therefore choose the most restrictive licensing options in order to do no harm. This ethical priority reduces the re-use and therefore citation options available for their data and as a result the potential for metrical impact.\n\nQuestions of authorship and citation surface as well. In the sciences definitions of what constitutes ‘authorship’ vary across disciplines and between journals. Citation practices are also not standardised. Guidelines do exist, however. By way of contrast, in AHSS research very few, if any, definitions and guidelines of what constitutes to authorship exist. If one were to apply certain biomedical models of authorship definition to AHSS research, many non-academic research participants would technically qualify as co-authors of research papers as they helped shape research data via their active, sometimes non-anonymous, participation in the research enquiry, particularly in applied, bottom-up, process-oriented research enquiries.\n\nWhilst acknowledging research participants’ input and possible co-authorship of research papers and data may be a more accurate and ethical reflection of their role in the research, it could also potentially create logistical challenges in the domains of IP, copyright, ethics and citation. Ethical considerations need to inform citation and author definition practices. Whilst it might seem like a good idea to measure impact through non-academic authorship, citation and engagement in the way hinted at above, this approach should not be recommended without careful ethical screening addressing questions of anonymity and equitable data sharing and ownership that are likely to arise, amongst many other hurdles.\n\nAnother challenge to be overcome is that of the use of metrics in assessing research quality and impact. ORCIDs and DOIs will be especially helpful in collecting statistical information quickly and digitally on how often research is being cited, and accessed and may go some way, so the argument goes, to showing how much impact is being achieved. Metrics however, is only one way in which impact becomes measurable and for AHSS researchers, it is thought to be misleading and ineffective. In a response to Hefce’s consultation on the use of metrics in assessment of research quality and impact, many Goldsmiths staff remained unconvinced that metrics could be used to conclusively prove research excellence or impact. Implicitly therefore, they had little faith in the use of DOIs and ORCID numbers as a way of improving impact analyses through metrics, although some did agree it would improve the visibility of research outputs and data. Most felt though that discoverability should not be equated with quality or impact per se.\n\nMetrics such as citation rates and journal impact factors, they observed, operate differently not only according to discipline (e.g., between psychology and literature), but also differ significantly between varying branches of the same discipline. Psychology, for example, is arguably a more diverse discipline than some, so indices like citation factors need to be interpreted very carefully even within different sub-disciplines to allow for meaningful 'like for like' comparisons (e.g., neuroscience journals typically have much higher impact factors than social psychology journals).\n\nThe Media, Communication and Cultural Studies Association’s (MeCCSA) REF consultation, compiled by Prof Golding and reiterated in the Goldsmiths’ response to the Hefce metrics consultation by MeCCSA members employed at Goldsmiths, yielded various anonymised comments from researchers. It was pointed out that evidence “suggests that variations in citation practices occur within disciplines as much as across disciplines, so the issue of calibration cannot simply be to the average for the subject as is proposed in science subjects. Staff felt it difficult to envisage a reliable way in which to develop disciplinary citation norms in interdisciplinary areas against which to compare individual counts. This would be especially true in fields such as media and communications which encourage publication across a very wide range of outlets in the arts, humanities, and social sciences. The usual suggestion of Web of Science (Thomson Scientific) as the database to be used for calibration of citation counts is acknowledged to be problematic. Web of Science is demonstrably incomplete in many areas (http://www.meccsa.org.uk/pdfs/REF-Consultation.pdf)”. Colleagues were unconvinced any database was complete and therefore figures not necessarily accurate or useful.\n\nOthers Goldsmiths’ colleagues felt that in AHSS research the number of citations was not a conclusive indicator of research merit, value or impact. A colleague commented that in the humanities (literary studies in particular), no publication is ever really superseded or made obsolete, nor is the author-critic irrelevant to the argument; the argument is very often his/her interpretation and appreciation of certain phenomena; in the sciences the assumption is that scientists report hard facts/results of experiments, not their idiosyncratic and poetic take on the set of data. Once a set of data or theory is superseded, scientific research tends no longer be cited or becomes part of what ‘everybody already knows’ (see Latour & Woolgar, 1986). This is not true for a lot of AHSS research and so the citation of data or the linking of outputs with ORCIDs might be of limited use it was felt.\n\nThe intellectual debates surrounding digital dissemination that are flourishing at Goldsmiths will inform practical digital dissemination strategies that the University as a whole will adopt in future. Intellectually these same debates seek to influence emerging theory in digital scholarship and dissemination in AHSS subjects more generally.\n\nDespite this scepticism and caution, new developments have begun to flourish. During recent data management scoping exercise, it became apparent that Goldsmiths researchers are engaging with digital dissemination practices quite effectively. This is especially the case where researchers are working on projects that are interdisciplinary, well-funded and usually include an element of non-AHSS research.\n\nFor example, Goldsmiths hosts a large Arts and Humanities, Research Council (AHRC) grant. Its research team actively seek to use and create scientific computing tools to collect and analyse large amounts of data to help further musicological analyses and practice. One such undertaking is the ‘Transforming Musicology’ project (Box 1).\n\n\n\nThis research project explores how software tools developed by the music information retrieval (MIR) community can be applied in musical study. Specifically the project seeks to:\n\nenhance the use of digitally encoded sources in studying 16th-century lute and vocal music and using such sources to develop new musical pattern matching techniques to improve existing MIR tools;\n\naugment traditional study of Richard Wagner’s leitmotif technique through audio pattern matching and supporting psychological testing;\n\nexplore how musical communities on the Web engage with their music by employing MIR tools in developing a social platform for furthering musical discussion online.\n\nA key technological contribution of Transforming Musicology is the enhancement of Semantic Web provisions for musical study. This involves augmenting existing controlled vocabularies (known as ontologies) for musical concepts, and especially developing such vocabularies for musical discourse (both academic and non-academic). It will also involve developing and promoting methods to improve the quality and accessibility of music data on the Web; especially the accessibility for automatic applications, following techniques known as linked data.\n\nThe project relies heavily on the digital labelling of musical units to help catalogue and identify compositional structures and musical pieces. The research has the potential to inform debates on musical performance, copyright, composition and musical analysis amongst other areas. To this end it will develop open source software tools as well. The process of identifying and labelling musical units with URIs means that this project has a large number of data sets and individual data items which might potentially be cited and accessible in audio format in the planned monograph for this large grant. The grant’s research team are presently considering the possibility of engaging with publishers to explore open access monograph formats so that they might include digital data sets. The creation of digital data sets and an open access monograph, in turn, provide a significant impetus for the research team to consider adopting ORCID numbers so that URIs and DOIs might be linked to their names and the grant. If data sets and digital tools were made available they would also benefit non-academic and amateur music groups, such as the lute-players with which the principal researcher works. For open access sharing to be made a reality however, careful consideration will need to be given to how research data sets are shared in the monograph and whether or not this might contravene existing copyright legislation, for example, as data is compiled from existing musical pieces. Until adequate sharing mechanisms are explored, it may not be possible to freely share the data accumulated during the project’s lifespan.\n\nAnother initiative taken by Goldsmiths’ researcher Joanna Zylinska and her team (Professor Joanna Zylinska, Dr Kamila Kuc, Jonathan Shaw, Ross Varney, Dr Michael Wamposzyc. Project advisor: Professor Gary Hall), includes the creation of an open book, Photomediations. The project redesigns a coffee-table book as an online experience to produce a creative resource that explores the dynamic relationship between photography and other media. Photomediations: An Open Book uses open (libre) content, drawn from various online repositories (Europeana, Wikipedia Commons, Flickr Commons) and tagged with the CC-BY licence and other open licences. In this way, the book showcases the possibility of the creative reuse of image-based digital resources.\n\nThrough a comprehensive introduction and four specially commissioned chapters on light, movement, hybridity and networks that include over 200 images, Photomediations: An Open Book tells a unique story about the relationship between photography and other media. The book’s four main chapters are followed by three ‘open’ chapters, which will be populated with further content over the next 18 months. The three open chapters are made up of a social space, an online exhibition and an open reader. A version of the reader, featuring academic and curatorial texts on photomediations, will be published in a stand-alone book form later in 2015, in collaboration with Open Humanities Press.\n\nPhotomediations: An Open Book’s online form allows for easy sharing of its content with educators, students, publishers, museums and galleries, as well as any other interested parties. Promoting the socially significant issues of ‘open access’, ‘open scholarship’ and ‘open education’, the project also explores a low-cost hybrid publishing model as an alternative to the increasingly questioned traditional publishing structures. Photomediations: An Open Book is a collaboration between academics from Goldsmiths, University of London, and Coventry University. It is part of Europeana Space, a project funded by the European Union's ICT Policy Support Programme under GA n° 621037. It is also a sister project to the curated online site Photomediations Machine: http://photomediationsmachine.net. This example provides a good model of how AHSS researchers in the visual arts might approach the production of open access monographs and arts catalogues, where licencing and copy right issues are very much foregrounded.\n\nA third example of Goldsmiths engagement with open access dissemination and data is that of work led by Dr Jennifer Gabrys and her team on an ERC funded project called Citizen Sense in the Sociology department (Box 2).\n\n\n\nThe project, which runs from 2013–2017, investigates the relationship between technologies and practices of environmental sensing and citizen engagement. Wireless sensors, which are an increasing part of digital communication infrastructures, are commonly deployed for environmental monitoring within scientific study. Practices of monitoring and sensing environments have migrated to a number of everyday participatory applications, where users of smart phones and networked devices are able to engage with similar modes of environmental observation and data collection. Such “citizen sensing” projects intend to democratize the collection and use of environmental sensor data in order to facilitate expanded citizen engagement in environmental issues.\n\nThe team examine how effective citizen sensing practices are in not just providing “crowd-sourced” data sets, but also in giving rise to new modes of environmental awareness and practice. Through intensive fieldwork, study and use of sensing applications, the project areas set out to contextualize, question and expand upon the understandings and possibilities of democratized environmental action through citizen sensing practices.\n\nAs part of their studies the research team on Citizen Sense collect live scientific data on for example air quality, using sensor devices. The team has now developed a website which visualises air quality data so that the general public can view the crowd sourced results. The interdisciplinary scope of this large project therefore means that the research data generated comes in a format that is more akin to data generated in science environments as opposed to AHSS disciplines. Therefore the collection, labelling, storage and archiving of this data using DOIs and attaching these to ORCIDs might therefore usefully draw on practices established in non-AHSS domains. This in turn could potentially enhance the visibility and impact of this research both environmentally and academically.\n\n\nConclusion\n\nWhilst many AHSS researchers at Goldsmiths remain sceptical about the use of ORCID numbers and digital object identifiers to enhance impact, the Goldsmiths examples show that there are distinct possibilities for their ability to enhance the visibility of on-line research outputs such as open access monographs, digital musical data and sociologically inspired scientific data. These examples, however, are sourced from projects that are highly interdisciplinary and well-funded, drawing on collaborations and resources not normally available to AHSS researchers in general. By and large most research grants in AHSS subjects tend to range between £5–£250k in value and many last no longer than between 12–24 months, allowing little time and resources for the development of novel strategies to digital research dissemination. Similarly, not all research enquiries might lend themselves well to digitisation due to the ethical, epistemological and practical concerns referred to above. Questions of authorship, the suitability of metrics for assessing impact and dissemination ethics continue influencing debates on the merits of digital dissemination and shall remain points of contention in the foreseeable future. However, as has been demonstrated above, there are circumstances where employing digital dissemination practices, DOIs and ORCIDs numbers is highly appropriate and could potentially lead to raising the profile of research in AHSS domains, demonstrating that this same research is capable of generating its theory either independently or in true collaboration with science partners.\n\nWhilst this paper has explored some of the (perceived) differences between AHSS and non-AHSS uses of digital approaches to data sharing and management, I would suggest that, based on preliminary discussions had, there are also many similarities between researchers and how they relate to their data, regardless of their disciplinary background. In future it may therefore be useful to explore the commonalities between disciplines alongside differences to help foster interdisciplinary approaches to research data management both practically and epistemologically, using a bottom-up approach.", "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\nAcknowledgements\n\nThis article was written with the help of staff at Goldsmiths, University of London. It uses information provided by: The Chair of The Media, Communication and Cultural Studies Association (MeCCSA); Prof Tim Crawford and Mr Richard Lewis; Pr Joanna Zylinska; Dr Jennifer Gabrys; Ms Caroline Lloyd and Mr Andrew Gray.\n\n\nReferences\n\nLatour B, Woolgar S: Laboratory Life: The Construction of Scientific Facts. 2nd edn. Princeton: Princeton University Press, 1986. Reference Source" }	[ { "id": "9381", "date": "13 Jul 2015", "name": "Anne Galliot", "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 article explores the specific needs of arts, humanities and social sciences (AHSS) data management, whether these are adequately served by existing data management systems, and whether they are appropriately considered in future data management developments. It highlights issues with the quantification and digitisation of information for research assessment and problematizes the concept of data itself, a notion that does not easily translate into AHSS disciplines. The author identifies this as a barrier to AHSS researchers engaging with the digital research information systems such as ORCID: it simply has little relevance to them. Links are made between current practices in research data management, licencing, authorship and ethics, and the myriad considerations that might arise in AHSS projects for which we do not at present have satisfactory frameworks. The article reiterates the inadequacy of metrics to capture research impact, and indeed, excellence, in the AHSS; this is corroborated by their absence in the latest Research Excellence Framework under Panel D, as well as by the recent findings of the Independent Review of the Role of Metrics in Research Assessment and Management, which indicate that it is not currently feasible to assess research outputs or impacts in the REF using quantitative indicators alone (Wilsdon et al., 2015). Finally, the author presents two research projects where some of the issues highlighted are being targeted, specifically by the creation of digital data sets and open access monograph and arts catalogues, in the formulation of research method and output. These projects are interdisciplinary and, crucially perhaps, ‘well-funded and usually include an element of non-AHSS research’. While it is clear that interdisciplinarity can contribute much to exploring these issues, it is uncertain whether tools and findings from these projects will fare better that the KULTUR project, which AHSS researchers are mainly unaware of. This is an opinion article, based on the author’s experience of working, and interviews, with researcher at Goldsmith. It certainly reflects my experience as Research Adviser for a College of Arts and Humanities. It may have been useful to reference some of the arguments referred to in the introduction, although the paucity of literature on these very current issue may have played against this. This article raises important questions about the definitions, ethical dimension, and the process of digitisation of research data, as well as about sector endeavours to quantify research impact and excellence. It works as a thought-provoking piece, begging many follow-on questions: How might we help AHSS researchers expand a definition of research data that will be relevant to them, and how do we enable the sector to acknowledge and redress the generalisation of its definition in favour of STEM disciplines? How might we ensure that the findings of KAPTUR and future projects are taken into consideration and their toolkits used? How might we address the apparent contradictions between copyright and intellectual property considerations and open access policies? How might we ethically define authorship in cases where research participants have contributing to shaping the data? I hope this article leads to many more engaging with these questions in more depth.", "responses": [ { "c_id": "1463", "date": "15 Jul 2015", "name": "Muriel Swijghuisen Reigersberg", "role": "Author Response", "response": "Dear Anne,Thank you for your review and supportive comments. I think based on these I ought to have better contextualized this submission, as it is somewhat unusual. This paper is in fact part of the conference proceedings of the CASRAI-ORCID conference, Barcelona, May 2015 on Research Evaluation, with an emphasis on emerging practice in the Humanities and Social Sciences http://www.orcid-casrai-2015.org/. The content of this paper was broadly discussed in a lively panel entitled: “Beyond Authorship: Recognising all research contributions.\"As such this paper therefore, was given prior to the official launch of the Hefce Metrics review http://www.hefce.ac.uk/pubs/rereports/Year/2015/metrictide/. Hence materials of this review were not included in the first versions of this paper. However, now that the report has been launched I shall be able to include a link + DOI for it and draw on some of the literature it mentions, which, as you suggest, would be very useful indeed. I would recommend that anyone reading or reviewing my submission also reads the Hefce report to contextualise this paper.Secondly, this submission is not actually an opinion piece. However, F1000 - being a predominantly bio-medical journal did not cater to arts, humanities and social science electronic 'templates' (if ever there were any), so the 'opinion' format was the only one suited for my particular submission. Thankfully editorial staff and conference organisers were very understanding about this small logistical hurdle and I am grateful to have been given the opportunity to contribute as part of the special theme on communicating science stream.Hope that's useful." } ] }, { "id": "9377", "date": "17 Jul 2015", "name": "Ben Johnson", "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 article gives an insightful and valuable overview of the challenges and opportunities of adopting new practices of digital scholarship within AHSS research processes. Building on the work of the Wilsdon review and earlier studies into practices within arts, humanities and social sciences research (e.g. the Crossick review), this provides a timely description of the difficulties we face in implementing broad solutions to tricky problems within a diverse research base. The diversity of research is often name-checked by those looking at the whole system and seeking to improve the way it works, but perhaps not fully understood. The specific process-oriented examples and case studies revealed here provide important contextual information to inform the sensible and sensitive roll-out of modern research management tools and approaches – it may be desirable, from a management and assessment perspective, to see universal adoption of ORCIDs, DOIs and so on, but this isn’t as easy as it sounds, and this article helps to explain why this might be the case while providing helpful examples of where it has worked and suggestions for ways forward. A particular problem is the complex and finely balanced nature of the relationship between different facets of the research process. While ethics, IP, copyright, digitisation, licensing, identification, citation, metrics and credit are often thought of as somewhat bounded issues that can be solved by ‘fixing the plumbing’ (e.g. by introducing ORCIDs), this article reminds us how complex their linkages are within the research process and how upsetting just one part of the balance can introduce vulnerabilities into the whole system. The examples given here about data management within arts disciplines are rich and informative, and justify a bottom-up approach to managing this agenda (as called for in the conclusion). It is already clear that ‘data’ means different things to different disciplines; even across (largely STEM) disciplines that generate numerical data as a primary output, one finds large variations in definitions, standards, practices and expectations that tend to muddle us. Extending the meaning of ‘data’ to include all inputs and outputs that inform and support the insights generated from the research process is a laudable aim of those seeking to increase the transparency, robustness, replicability, dissemination and impact of research; doing so in a way that take sufficient account of the complex dependencies between anonymity, confidentiality, intellectual property, ethical propriety and so on is a particular challenge within AHSS research and one that is perhaps not given sufficient attention by those operating at the ‘macro’ level of research administration, assessment and policy development. Beyond data, the particular problems of contributor anonymity, delineating roles within collaborations with non-academic colleagues, the invalidity of digital simulacra of real-world artistic artefacts, the complexity of documentation of data drawn from a wide range of often privately-owned sources… these are problems that are not felt by colleagues in STEM (the group of disciplines from which it is often felt that moves to ‘digitise’ research are flowing). The assignment of DOIs, ORCIDs, OA licences and so on to the outputs of research operating in this environment is tricky and fraught with real dangers that will require careful further investigation. There is a clear need to need to tease out the limitations of these new aspects of the research ‘plumbing’ within disciplines, explore novel solutions, find what works and what doesn’t, and seek a sensible way forward. At the heart of this is the question of ethics. The close dependency between more open and transparent scholarly communication practices and more effective research integrity are not disputed, but this is often used to justify a conclusion that ‘open’ is ‘better’ in all cases. The examples above, particularly of ethnographic research, reveal that ethical limitations within disciplinary practice often inform models of communication in a way that might hinder openness, and that this is entirely appropriate in the disciplinary context. This at first appears to fly in the face of the very idea of “open science”, but in practice it only underlines the need for context-specific approaches to openness that take sufficient accounts of the ethical practices within disciplines. Clear delineation is needed, though, between genuine ethical considerations and those simply borne of more affected academic-cultural norms or resistant to practical change – we need to head off any unfair accusations of ‘special treatment’ being granted to these disciplines purely on political grounds. We need to better understand this problem, so that we can more effectively and sensitively tailor our approaches to achieve open research communication in a way that respects good research practice in all disciplines. Finally, the question of metrics. Central to the arguments made above, and elsewhere, is a concern that the ‘plumbing’ of DOIs, ORCIDs, Web of Science coverage etc. is insufficient to enable the accurate capture of research outputs within AHSS, and therefore the metrics systems that depend on counting research outputs will unfairly discriminate against these disciplines. As the article states, “the upshot of this is that those disciplines that are less digitally oriented, are likely to obtain unhelpful metric ratings.” In my view, this masks a more pressing issue, which is that metrics are most applicable to those disciplines that ‘chunk’ their outputs into easily quantifiable forms, with quantifiable relationships to one another, with quantifiable citation practices, quantifiable(ish) contributions of academics to the research, and so on. It’s clear from the above, and from my own discussions with AHSS researchers, that the problems of quantifying AHSS research are not only related the coverage of DOIs and ORCIDs, and we should be careful not to assume that we entirely fix the issue of metrics by fixing the plumbing (even though we might get a few ‘quick wins’ in a few areas).", "responses": [] } ]	1	https://f1000research.com/articles/4-193
https://f1000research.com/articles/4-192/v1	07 Jul 15	{ "type": "Opinion Article", "title": "Connecting Altmetric: Integrating with Institutional Publications Systems", "authors": [ "Natalia Madjarevic" ], "abstract": "This paper discusses Altmetric tools for institutions and how they are supported by accurate, up-to-date and re-usable research information. We examine the importance of standardised metadata formats across research information management systems in enabling altmetric providers to deliver data in robust, reliable and meaningful ways. We share our experience of collaborating with a range of institutions to report and analyse the attention to their collection of research outputs and surfacing altmetrics data at the author, department and institutional level. This includes working with institutions to ensure we can harvest from or integrate with existing technical infrastructure in order to match outputs with the corresponding altmetrics data in the Altmetric database. We discuss integrations with institutional repositories and publications systems including Symplectic Elements, VIVO and DSpace. Finally, we study motivations for incorporating metrics into workflows and systems across institutions, and how altmetrics can be integrated with existing research support and bibliometrics services.", "keywords": [ "Altmetric", "Metadata", "Article Metrics" ], "content": "Introduction\n\nAltmetrics, or alternative metrics, are increasingly recognised as an emerging signal of research engagement beyond traditional citation metrics (Bornmann, 2014; Dinsmore et al., 2014; Priem, 2013; Thelwall et al., 2013). Altmetrics enable researchers, institutions (Liu & Adie, 2014), publishers and funders (Dinsmore et al., 2014) to measure and report on broader societal engagement with research outputs.\n\nAltmetric, founded by Euan Adie in 2011, monitors attention to research outputs across a range of non-traditional sources including news, policy documents, blogs and social media. Altmetric’s tools, including the Altmetric Explorer, Altmetric for Institutions and Altmetric Badges, support researchers, institutions and organisations in tracking attention to research outputs beyond traditional citations.\n\nIn order for altmetrics providers to build reliable and robust tools based upon high quality data, it is important we work closely with the scholarly community and use standardised metadata to describe and connect research outputs. So how does this process work at Altmetric? We are a data science company, and track attention to research outputs with unique identifiers such as Digital Object Identifiers (DOIs), PubMed IDs (PMIDs), Research Papers in Economics (RePEc) IDs and more. In addition, we text mine policy documents and news sources in order to identify mentions of research papers without a persistent identifier or URL in the body of the text, matching the item up with the corresponding record in the Altmetric database. We also ensure mentions of different versions of the same paper are disambiguated in order to present a single unified record for all versions of an item, such as the publisher, PubMed and institutional repository version, presenting all mentions in a single Altmetric Details Page in the database.\n\nAltmetric’s process of tracking attention to research outputs is further enabled by standardised metadata formats in order to connect mentions for different versions of papers. We track attention to research outputs with persistent identifiers including DOIs, RePEc IDs, arXiv IDs, PubMed IDs, Astrophysics Data System (ADS) Bibcodes, Handle identifiers and unique resource identifiers (URIs). When a link to a research paper is shared across a source of attention that we track, e.g. in a news story or shared on Twitter, we follow the link to the page and search the publication page for item metadata, particularly searching for the persistent identifier such as a DOI. For example, a news story discusses a research paper, and links to the journal article hosted on the publisher homepage. As Altmetric recognises the publisher domain, we follow that link to the content hosting platform, and scrape the persistent identifier and key item metadata; adding that item as a record to our database. The Altmetric technical infrastructure is embedded with an identifier mapping to ensure we recognise and disambiguate across multiple versions. A key challenge is to track a wide range of sources where research is being discussed, while being able to make sense of the disparate mentions of over 3.8 million research outputs as of May 2015.\n\nIn order to collect the basic descriptive information about each item we track, Altmetric automatically harvest metadata in Dublin Core meta tags from platforms such as institutional repositories, e.g. DC.title, DC.identifier and DC.creator. For publisher-hosted content, we harvest the HTML “<meta>” tags such as citation_doi and citation_title from the metadata of the item page (see Table 1).\n\nIt is important that content creators and publishers assign persistent identifiers and standardised metadata as listed above, exposing this in a machine-readable format across platforms, in order to allow altmetrics providers to recognise and disambiguate mentions of research outputs. Many publishers and institutional repositories already include high quality metadata to enable Google Scholar visibility, as described further in Google Scholar’s Inclusion Guidelines for Webmasters. In the future, we’d like to harvest complex, standardised metadata from publisher platforms to describe research outputs at a more granular level, such as author identifiers, author affiliations, contributor roles and funder information.\n\nAltmetric also work with a number of larger publishers to enhance item metadata mined from content platforms. For example, when we track a mentioned item published by Springer, we query the Springer Application Program Interface (API) with the item DOI and pull the subject classification from their database in order to add subject level metadata at the Altmetric record level. In addition, Altmetric collect journal International Standard Serial Numbers (ISSNs) for each item we track and attach Excellence in Research Australia (ERA) Field of Research (FoR) subject codes. This enables end users to slice Altmetric data by subject at the item level. We would, however, like to see wider adoption of a lightweight standardised subject classification system across publishers and repositories.\n\nIn 2015, Altmetric will be introducing support for ORCID, building on the app (http://altmetric-orcid-profiles.herokuapp.com/) launched during the 2014 1:AM altmetrics conference hackathon. We’ll be adding support to search for ORCID IDs in the Explorer and mine for ORCID IDs as standard during automated metadata harvesting.\n\nCollecting, structuring and enhancing altmetrics data is a core part of the work we do behind the scenes at Altmetric. We launched Altmetric for Institutions in 2014 to enable universities and institutions to search, browse and report on attention to their research outputs at the author, department, custom group and item level. As Altmetric already tracks attention to publications across our sources, all we need from the institution is the details of institutional published outputs, associated authors and department groups to structure the altmetrics data according to institutional requirements. In order to build Altmetric for Institutions at the institutional level, we developed a number of connectors to harvest publications information directly from research information management systems, as described in more detail below. This enables us to work with existing technical systems and publications metadata curated by the institution.\n\nAn early Altmetric for Institutions integration was with the research information management system Symplectic Elements. Created by our Digital Science sister company, Symplectic, and based in the same office as the Altmetric team, we built a connector to the Symplectic Elements API in order to pull author, publication and organisational hierarchy metadata through to populate Altmetric for Institutions. The University of Melbourne are an Altmetric early adopter and Symplectic Elements customer. This enabled us to develop the Symplectic Elements connector with the established University of Melbourne Elements instance, which is populated with over 150,000 publications and 26,000 authors. Following an initial import of the above data, we now run regular updates from the University of Melbourne instance of Symplectic Elements to ensure Altmetric for Institutions remains up-to-date with modified or recently claimed publications for all authors. We make use of the persistent identifier metadata fields in Symplectic Elements, which allows us to match publications with the corresponding record and associated attention in the Altmetric database. In addition, the Altmetric badges feature alongside each item record in the Symplectic Elements interface, helping raise the visibility of altmetrics and research engagement with university administrators and researchers in their daily workflows.\n\nOur next major Altmetric for Institutions connector enabled us to harvest publications, author and department metadata from VIVO profiles. Working with Stony Brook University School of Medicine, we built a connector that could pull in the key metadata including relevant authors, associated papers and medical faculties in order to create a searchable, browsable and reportable structure within Altmetric for Institutions. In addition, Stony Brook used the Altmetric API in order to integrate the Altmetric badges in their internal publications system and raise awareness of public attention to research with practitioners and researchers across the organisation.\n\nAn important Altmetric for Institutions integration is working with open source repositories to harvest publications information via the Open Archives Initiative – Protocol for Metadata Harvesting (OAI-PMH). In Spring 2015, we worked with the World Bank Group to develop our tools for harvesting via OAI-PMH and mapping across multiple platforms. Firstly, we connected to the World Bank’s Open Knowledge Repository (OKR), which is based on DSpace, via OAI-PMH and this enabled us to populate Altmetric for Institutions with the relevant publications information and connect to the associated altmetrics. In addition, we set up publication tracking for all World Bank published content hosted across their four dissemination platforms: OKR, World Bank eLibrary, Documents and Reports Archive, and RePEc. As World Bank Group publications are often made available across all of these platforms, it was key that we were able to recognise different versions of the same item.\n\nIn order to set up a process of disambiguation for the World Bank collection, we worked with the World Bank team to harvest relevant identifier mappings added to their repository feed and create this mapping within the Altmetric database. As a result, when a World Bank Group paper from any of their platforms is shared across the sources we track, we will disambiguate within the database and ensure all posts are merged for that item. The World Bank Group identifier mapping infrastructure includes Handle IDs, RePEc IDs, DOIs and internal World Bank IDs, demonstrating how important it is that persistent identifiers are used across publications systems in open formats to enable mapping and reliable altmetrics tracking. In addition, the World Bank embedded the Altmetric badges in both OKR and World Bank eLibrary in order to showcase attention to their published research.\n\nHaving discussed several integrated aspects of the Altmetric technical infrastructure, it is important to consider the practical applications of altmetrics in institutions. Altmetrics are used in practice across the scholarly communications lifecycle (Madjarevic & Davies, 2015) – from librarians, research offices and communications teams, to individual researchers and faculty heads. By analysing the underlying qualitative data – the mentions and discussions – we can further understand successful research dissemination, support researchers making the case for grant funding or promotion and provide a broader, more coherent understanding of the potential impact of research on society.\n\nA key institutional use case is surfacing conversations surrounding research outputs beyond traditional citation analysis, conversations that may have previously been unknown to the institution or researcher, in order to collect evidence of indicators of impact for funders or to present during promotion reviews. By surfacing this qualitative data, Altmetric helps organisations identify potential impact in sources such as policy documents. For example, an intervention review paper produced by University of Melbourne researchers and published in Cochrane Database of Systematic Reviews (Boyle et al., 2008), studies the use of probiotics for treating eczema. It found that probiotics were not an effective treatment. The paper has a relatively low Altmetric score of 13, but received mentions in policy documents from the European Food Safety Authority and the Royal College of Paediatrics and Child Health and is cited on the main Wikipedia page for Dermatitis. This demonstrates the importance of analysing the underlying qualitative mentions beyond score in order to identify indicators of research impact. Furthermore, the World Bank Group used Altmetric for Institutions to track attention to research in developing nations, a key target community for their publication outputs. In a recent World Bank Group report, Addressing Inequality in South Asia (Rama et al., 2014), the publications team were able to observe geographical trends in sharing activity across South East Asia in the Twitter demographics, as visualised on the Altmetric Details Page (Figure 1).\n\nAdditional key use cases include: identifying successful author impact activities to recognise institutional best practice in promoting research impact; encouraging researchers to include in grant applications, CVs and author profiles; enhancing library liaison services; encouraging staff to deposit in research information management systems; integrating data in performance reports; and identifying popular research to further promote.\n\nHere at Altmetric, there are a number of improvements we’d like to see across the scholarly publications and repository communities to drive altmetrics future developments and enable robust, interoperable data. This includes the broad adoption of the following metadata as standard across publisher platforms and repositories:\n\nAuthor identifiers, e.g. ORCID IDs attached to all authors and records.\n\nPersistent identifiers as standard for all research outputs, including grey literature and data sets.\n\nDepartmental groupings and sets to enable organisational analysis.\n\nFunder metadata to identify research papers produced as a result of external funding, e.g. FundRef.\n\nStandardised author affiliation identifiers.\n\nSubject taxonomy, namely the adoption of an agreed subject classification system.\n\nStandardised final date of publication available and distinct in each item record.\n\nThe addition of the above metadata fields to research outputs across publisher platforms and repositories, opens opportunities for the altmetrics community to further develop tools and data analysis in robust and innovative ways.\n\n\nConclusion\n\nAltmetrics providers are keen to work with scholarly publishing and institutional repository communities to develop the quality and breadth of persistent identifiers and standard metadata fields. During Altmetric’s work to develop automated item tracking processes, version disambiguation techniques and integrating with a range of publication systems, we identified a number of benefits and potential improvements to the existing practices as described above. However, an overwhelming finding is by connecting Altmetric to existing institutional publication systems, builders of altmetric tools are able to offer low barrier entry to altmetrics at the institutional level to enable deeper impact analysis and uncover conversations about research across broader society.", "appendix": "Competing interests\n\n\n\nThe author is an employee of Altmetric.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nI would like to thank Euan Adie (Altmetric) for his comments on an early version of this paper.\n\n\nReferences\n\nBornmann L: Do altmetrics point to the broader impact of research? An overview of benefits and disadvantages of altmetrics. J Informetr. 2014; 8(4): 895–903. Publisher Full Text\n\nBoyle RJ, Bath-Hextall FJ, Leonardi-Bee J, et al.: Probiotics for treating eczema. Cochrane Database Syst Rev. 2008; 8(4): CD006135. PubMed Abstract \| Publisher Full Text\n\nDinsmore A, Allen L, Dolby K: Alternative perspectives on impact: the potential of ALMs and altmetrics to inform funders about research impact. PLoS Biol. 2014; 12(11): e1002003. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu J, Adie E: Realising the Potential of Altmetrics within Institutions. Ariadne. 2014; (72). Reference Source\n\nMadjarevic N, Davies F: Altmetrics in Higher Education Institutions: Three Case Studies. The Winnower. 2015. Reference Source\n\nPriem J: Scholarship: Beyond the paper. Nature. 2013; 495(7442): 437–40. PubMed Abstract \| Publisher Full Text\n\nRama M, Béteille T, Li Y, et al.: Addressing Inequality in South Asia. The World Bank. 2014. Reference Source\n\nThelwall M, Haustein S, Larivière V, et al.: Do altmetrics work? Twitter and ten other social web services. PLoS One. 2013; 8(5), e64841. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9375", "date": "21 Jul 2015", "name": "John Dupuis", "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\nOverall, this article very strong, and is certainly publishable. The nature of the Opinion-style article gives the author considerable leeway compared to a research-based article. However, when the article is by an employee of a company and it describes that company’s products, the line between Opinion and Advertorial can be thin. This article certainly does a good job of describing what altmetrics are, what the positives aspects of their use are as well as how Altmetric’s offerings demonstrate those advantages.My review suggestions for this article all center around strengthening that separation for this article and making sure it provides more than just advertisement for a particular company and gives at least some sense of a broader product category as well as a bit more on that product category’s disadvantages as well as its advantages. I don’t believe that this article should “be about something else” just that it needs to provide a bit more diversity of ideas and opinions to make it a little less lopsided.One way to address this issue earlier in the process would have been to collaborate on authorship of the article (and presumably the conference presentation it is based on) with someone at one or more of the institutions that participated as case studies.Introduction, the author goes through some of the strengths of altmetrics to give a more balanced sense of whether or not altmetrics (and Altmetrics) are worth the investment for institutions. The author should also address some of the criticisms of altmetrics that are quite common in the literature. The Bornmann article in the references also addresses disadvantages and they could have been explored. As well in the introduction, some mention could have been made of other companies that offer personal or institutional altmetrics solutions, such as Impact Story.The criticisms brought up in the introduction could have also been addressed in the Practical applications of altmetrics section of the article. This would make the whole article stronger, especially as opinion rather than advertisement. There is some mention of taking the Altmetrics score in context of other measures in the section, but these could be expanded.Finally, one small suggestion for wording. In the conclusion the word “finding” is used in relation to the conclusions drawn from the case studies. “Findings” usually relate to research rather than opinion.One thing I would also like to see is a few more illustrations/screenshots from the various systems that Altmetric built with their partner institutions.", "responses": [] }, { "id": "9911", "date": "07 Aug 2015", "name": "Philip Young", "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 article is informative and well written overview of how the Altmetric product integrates with institutional platforms. The sketch of how Altmetric tracks research in the Introduction I found particularly useful. It is worth noting that the author is an employee of Altmetric, but I feel this is acceptable for an opinion article. I agree with reviewer 1 that the article suffers somewhat from a focus on the company and could be strengthened in some areas by a broader perspective. hese do not need to be significant changes, but can be brief mentions outlined below. First, on the topic of standardizing metadata formats, areas of consensus among altmetrics providers might be noted wherever possible, rather than focusing solely on what Altmetric would like to see. Also, NISO’s current efforts might be acknowledged here. Second, under the heading \"Practical applications of altmetrics,\" the statement beginning \"Altmetrics are used in practice aross the scholarly communications lifecycle...\" might be qualified, as it seems to overstate the extent to which altmetrics are currently in use (which is avoided in the cited article). Granted, this information may be proprietary, but it would be useful to know how widespread altmetrics use is. Third, an significant issue of institutional interest that might be touched upon is the relationship between altmetrics and open access (this came to mind in the discussion of the World Bank reports). This seems to be a missed opportunity given the discussion of repositories and altmetrics (though there is brief mention of the use case “encouraging staff to deposit in research information management systems”). As with numerous studies on the OA citation advantage, do OA articles receive more mentions? Can Altmetric data tell us which articles (or which versions) are OA and which are not? Fourth, I agree with reviewer 1 that some criticisms or limitations of altmetrics might be admitted. It seems unlikely, for example, that altmetrics would provide useful data in all disciplines. I think these four issues could be addressed without significant revisions and would improve this paper. A problem (which I do not expect the author to address here) with Altmetric is the prominent display of scores in the \"donut\" with a simultaneous downplaying of that number. For example, the author says “The paper has a relatively low Altmetric score of 13….” and then emphasizes “the importance of analysing the underlying qualitative mentions beyond score…” If it is more important to look at the sources, then the score should be done away with so it does not become subject to abuse like the journal impact factor (one reason that I prefer the displays of other altmetrics providers). As with the JIF, quantitative measures are too tempting to our short-cutting, convenience-minded human nature.", "responses": [] } ]	1	https://f1000research.com/articles/4-192
https://f1000research.com/articles/4-55/v1	26 Feb 15	{ "type": "Review", "title": "Fenofibrate in cancer: mechanisms involved in anticancer activity", "authors": [ "Tomas Koltai" ], "abstract": "Objective: To review the mechanisms of anti-cancer activity of fenofibrate (FF) and other Peroxisome Proliferator Activator Receptor α (PPARα) agonists based on evidences reported in the published literature.Methods: We extensively reviewed the literature concerning FF as an off target anti-cancer drug. Controversies regarding conflicting findings were also addressed.Results: The main mechanism involved in anti-cancer activity is anti-angiogenesis through down-regulation of Vascular Endothelial Growth Factor (VEGF), Vascular Endothelial Growth Factor Receptor (VEGFR) and Hypoxia Inducible factor-1 α (HIF-1α), inhibition of endothelial cell migration, up-regulation of endostatin and thrombospondin-1, but there are many other contributing mechanisms like apoptosis and cell cycle arrest, down-regulation of Nuclear Factor Kappa B (NF-kB) and Protein kinase B (Akt) and decrease of cellular energy by impairing mitochondrial function. Growth impairment is related to down-regulation of Phospho-Inositol 3 Kinase (PI3K)/Akt axis and down-regulation of the p38 map kinase (MAPK) cascade. A possible role should be assigned to FF stimulated over-expression of Tribbles Homolog-3 (TRIB3) which inhibits Akt phosphorylation. Important anti-cancer and anti-metastatic activities are due to down-regulation of MCP-1 (monocyte chemotactic protein-1), decreased Metalloprotease-9 (MMP-9) production, weak down-regulation of adhesion molecules like E selectin, intercellular adhesion molecules (ICAM) and Vascular Endothelial Adhesion Molecules (VCAM), and decreased secretion of chemokines like Interleukin-6 (IL-6), and down-regulation of cyclin D-1. There is no direct link between FF activity in lipid metabolism and anticancer activity, except for the fact that many anticancer actions are dependent from PPARα agonism. FF exhibits also PPARα independent anti-cancer activities.Conclusions: There are strong evidences indicating that FF can disrupt growth-related activities in many different cancers, due to anti-angiogenesis and anti-inflammatory effects. Therefore FF may be useful as a complementary adjunct treatment of cancer, particularly included in anti-angiogenic protocols like those currently increasingly used in glioblastoma. There are sound reasons to initiate well planned phase II clinical trials for FF as a complementary adjunct treatment of cancer.", "keywords": [ "Fenofibrate", "cancer", "PPAR", "angiogenesis", "metastasis", "prostate", "cancer", "glioblastoma", "melanoma", "nelfinavir" ], "content": "Abbreviations\n\nFF: Fenofibrate /HUVEC: Human vascular endothelial cells /MCP-1: Monocyte chemotactic protein-1 /VCAM-1: Vascular cell adhesion molecule 1 /ICAM-1: Intercellular adhesion molecule-1 /VEGF: Vascular endothelial growth factor /VEGFR: Vascular endothelial growth factor receptor /HIF: Hypoxia inducible factor /PPAR: Peroxisome proliferator activator receptor FGF: Fibroblast growth factor /bFGF: basic fibroblast growth factor /TSP-1: Thrombospondin-1 /RXR: Retinoid X receptor /ATRA: All transretinoic acid /ER: Endoplasmic reticulum /RCC: Renal cell carcinoma /PDGF: Platelet derived growth factor /TRIB-3: Tribbles homolog 3 /IGF: Insulin like growth factor /PEG2: Prostaglandin E2 /CTMP: C-terminal modulator protein (which inhibits AKT phosphorylation) /SCC: Squamous cell carcinoma /SREBP: Steroid regulatory element binding proteins /FAS: Fatty acid synthase.\n\n\nIntroduction\n\nFenofibrate (FF) is a drug of the fibrate class (a fibric acid derivative) that has been used since 1975 to reduce cholesterol (LDL and VLDL) and triglyceride levels and increase HDL in patients at risk of cardiovascular disease and for treatment of atherosclerosis (1 and 47). FF is one of the most commonly prescribed fibrates, and has a well known efficacy and tolerability profile1.\n\nFF seems to lower lipid levels by activating peroxisome proliferator-activated receptor alpha (PPARα), a nuclear receptor which acts as a ligand activated transcriptional factor and activates lipoprotein lipase and reducing apolipoprotein CIII expression. These activities increase lipolysis and eliminate triglyceride-rich particles2.\n\nPPARs are a widely distributed family of nuclear receptors. Three isoforms have been identified: PPARα, PPARβ/δ, and PPARγ. Ligand binding activates these receptors that play key roles in cellular energy homeostasis, modulating glucose and lipid metabolism. PPARα as illustrated in Figure 1, is the molecular target of the fibrate class of drugs, which act as agonistic ligands of PPARα. Other fibrates like clofibrate and bezafibrate are also ligands for this receptor. Poly-unsaturated fatty acids are the natural ligands.\n\nUpon ligand binding, PPARα dimerizes with RXR (retinoic X receptor) and both interact with peroxisome proliferators responsive elements of the target gene. Coactivator proteins and RNA polymerase are recruited and the transcription machinery is set to work (trans-activation). When co-repressor molecules are recruited, trans-repression is unleashed and no transcription is produced.\n\nWe shall not go any deeper into lipid metabolism activities of FF because our goal is to consider the effects of this pharmaceutical in cancer prevention and treatment rather than in cardiovascular risk.\n\nBefore the year 2000, all publications on FF considered anti-lipogenic properties of this drug with no mention of possible anti-cancer activity.\n\nWe only found two publications of FF activities before than that may be related with cancer:\n\n1) Marx et al. (1999) describe that FF reduces the expression of vascular cell adhesion molecule 1 (VCAM-1) in human endothelial cells3.\n\n2) Shu et al. (2000) observed that PPARα or PPARγ reduce secretion of MMP-9 in certain cells (human monocytic THP-1 cells)4.\n\nIn 2002 two findings on FF anti-cancer therapy were important:\n\n1) It was demonstrated that PPARα and PPARδ down-regulate NF-kB induced translocation to the nucleus in endothelial cells5 and\n\n2) PPARα activators inhibited endothelial cell migration by targeting Akt phosphorylation in endothelial cells6.\n\nFF has the capacity to induce hepatocarcinoma in rodents, but this effect seems specific for this species, as in humans it has been shown to have cytotoxic effect on HepG2 hepatoma cell line at high concentrations and in a dose dependent manner7.\n\nVaret et al. (2003)8 demonstrated that FF inhibits angiogenesis in vitro and in vivo.\n\nMCP-1 (monocyte chemotactic protein-1) is a protein that recruits and activates monocytes during inflammatory processes but also plays a role in cancer: it increases proliferation and invasion of CaP cells (prostate cancer)9. FF inhibits expression of MCP-1 on activated endothelial cells10.\n\nPPARα agonists like FF were found to inhibit endothelial VEGFR211 expression.\n\nIn diabetic II hyperlipidemic patients, FF decreased E-selectin by 10% and ICAM-1 by 4% and no change of VCAM-1 was detected12.\n\nBut the first real hint towards a possible anticancer activity of FF was provided by Holland et al. in 200413 who showed by transcriptome analysis of endometrial cancer cells an overexpression of PPARα. This finding led them to the investigation of FF activity on tumor cells. FF reduced proliferation and increased apoptosis of cancer cells. At the same time, a second publication14 showed the FF potential to reduce metastasis of melanoma cells in an experimental setting.\n\nIn ooforectomized rats, treated with estradiol and FF for 30 days, the uterine mass decreased, uterine glands had normal structure and there were no cases of atypical hyperplasia15.\n\nKubota et al.16 found that apoptosis induced by FF in cultured human hepatocytes was due to caspase-dependent apoptosis by inhibiting phosphorylation of Akt, in a PPARα independent manner.\n\nThe role of chemokines produced by different stromal cells stimulating proliferation and angiogenesis in cancer tissues is well known. FF exerts a monocyte suppressing activity and reduces secretion of IL-6 and MCP-117.\n\nStudying the possible anti-rheumatic activity of FF it was observed that this compound inhibits NF-kB18.\n\nAfter these preliminary hints indicating the FF possible anti-cancer activities of FF, great amount of research and publications were dedicated to this issue. We summarized these findings in table 1 to table 15 according to anti-cancer activity disclosed.\n\n\nFF anticancer mechanisms\n\nOn artificial grounds, but for better understanding, we have presented the anti-cancer activity of FF according to the main pro-tumor factor/pathway affected by the drug.\n\nIn the field of angiogenesis there is clinical experience with FF besides the laboratory experimental setting. In the research by Blann et al.19, hyperlipidemic patients treated with FF showed reduced lipidemia and plasmatic VEGF. No changes in VEGFR levels were seen.\n\nThere is a tissue where FF has a pro-angiogenic effect: human retina. An anti-apoptotic property of FF in human retinal endothelial cells was reported by Kim J et al. in 200725. FF potently activated AMP-activated protein kinase (AMPK) and vascular endothelial growth factor (VEGF) mRNA expression. This finding was corroborated by the ACCORD medical study26 in patients with diabetes type 2 where FF was shown to have protective activity in diabetic retinopaty and other diabetic microvascular complications, probably through a decrease of human retinal endothelial cells apoptosis27.\n\nThis may mean that FF has a tissue specific activity that needs further investigations.\n\nProbably the strongest and clearest evidence of the FF anti-angiogenic activity comes from the research of Panigraphy et al.21 of the Judah Folkman team and the research by Dana et al.23 on HUVEC.\n\nAngiogenesis inhibition as described in Table 1 is probably one of the main anti-tumor activities of FF. Nickkho-Amiry et al.28 showed that treatment of human endometrial cells with a PPARα agonist leads to reduced secretion of VEGF in addition to reduced proliferation. This was potentiated by RxR (Retinoid X receptor) agonist like ATRA and inhibited by a PPARα antagonist.\n\nIn a rat model Onalan et al. showed that FF caused regression of new endometriotic implants due to decreased angiogenesis29.\n\nFF was included in many multi-agent anti-angiogenic regimens. One consisted of FF, celecoxib, thalidomide with metronomic low dose cyclophosphamide and etoposide. Patients were less than 21 year old with recurrent or progressive tumors. Half of the patients obtained benefits (CR + PR + SD)30.\n\nOther metronomic anti-angiogenic multidrug protocols included FF as one of the pharmaceuticals, particularly for children with embryonal brain tumors and other malignancies31–33.\n\nThe COMBAT Protocol34 included low-dose daily temozolomide, etoposide, celecoxib, vitamin D, FF and retinoic acid and was used in 74 children with advanced refractory/relapsed solid tumors with two years overall survival of 43%.\n\nThe use of FF as part of anti-angiogenic multidrug protocols especially in pediatric cancer is constantly increasing.\n\nUsing a PPARα agonist like Wy-14643 in mice injected with tumor cells showed that treated animals had a marked reduction in tumor size and vascularization35.\n\nIn summary: FF increases thrombospondin synthesis, endostatin generation, decreases VEGF, COX2 and VEGFR2 expressions and prevents endothelial cells migration21,36.\n\nApoptosis induced by FF is caspase-dependent. In the case of clofibrate, apoptosis occurs through caspase 2 and 3 activation and ER stress in Jurkat cells44. Similar results were observed in Yoshida AH130 hepatoma cells45.\n\nPPARα is increased in high grade renal cell carcinoma (RCC), but this does not provide any information about the functional status of this receptor, because in RCC the inhibition of PPARα induces apoptosis and agonists produce little or no effect46.\n\nIn 1983 Pascal et al.47 investigated the cardiovascular and anti-arteriosclerotic activities of FF and demonstrated that FF inhibited platelet derived growth factor (PDGF) stimulating activity on growth of cultured smooth muscle. Ten years later Munro et al.48 showed that FF is not a specific inhibitor of PDGF because smooth muscle cells growth was equally growth-inhibited by FF when the culture was stimulated with fetal calf serum, PDGF or basic fibroblast growth factor (bFGF). Our conclusion based on these two publications is that FF is a growth inhibitor in general (as least regarding vascular smooth muscle).\n\nAnti-proliferation activity of FF has been found in many non-tumor tissues besides vascular smooth muscle, e.g. mesangial cells49 through inhibition of PI3K/AKT and ERK1/2 signaling pathways or by overexpression of TRIB3 (tribbles homolog 3) which inhibits Akt phosphorylation and slows cell cycle or causes arrest in G1/S50. In lymphocytes, FF also up-regulates TRIB3 causing cell cycle arrest51,52.\n\nEndothelin-1 is a protein that increases cardiac fibroblast proliferation. PPARα agonists inhibit cardiac fibroblast proliferation down-regulating endothelin-153. FF also reduced c-jun expression in cardiac fibroblasts54. Endothelin-1 is an activator of the p38 mitogen activated kinase cascade. FF down-regulation of endothelin-1 also down-regulates the MAPK cascade in cardiomyocites55.\n\nFF reduced the IFNγ and IL-1β-induced cell proliferation of astrocytes in culture56.\n\nTable 3 depicts the anti-proliferation activities of FF in cancer.\n\nThe work by Saidi et al.59 needs further discussion. The authors noticed that in Ishikawa endometrial cancer cells FF enhanced growth inhibition when ATRA was simultaneously used. ATRA by itself had no effect on growth. This is a logical finding because PPARα forms a heterodimer with RXR before binding DNA at the peroxisome proliferators responsive element. So this synergy between FF and ATRA regarding growth inhibition seemed a PPARα-dependent activity. Apoptosis also was increased with the combination of these drugs.\n\nParadoxically, RNAi inhibition of PPARα showed only a minor reduction in FF effect and ATRA combined with FF showed minor differences in growth inhibition with or without PPARα RNAi. After 48 hours of treatment the difference was approximately 40% less viability in cells treated with FF plus ATRA and no RNAi against those with RNAi. We hypothesize that RNAi inhibition of PPARα needs at least 48 hours to make the viability difference. So that growth inhibition seems, at least partially, as PPARα dependent.\n\nThey also found down-regulation of two genes: cyclin D1 (CCND1) and methionine adenosyltransferase 2 A (MAT2A), both are pro-growth genes59. High doses of FF up-regulated p21 (cyclin-dependent kinase inhibitor 1a) and TP53.\n\nUnfortunately FF and FF plus ATRA showed no differences in tumor size and growing in vivo compared with control group receiving no drugs.\n\nThe work by Chang et al.62 suggests that FF may be useful for prevention of oral SCC because in an experimental setting FF was capable of reducing the incidence of tumors and also the progression from pre-neoplastic stage to SCC. FF at low doses lacked anti-tumor activity.\n\nIn spite of the known fact that glucocorticoids induce chemotherapy resistance in most of the solid tumors66, Liang et al.65 found that FF and budesonide had synergistic anti-proliferative effect on lung cancer cells with intact TP53.\n\nInflammation plays a very important role in carcinogenesis and tumor progression. NF-kB pathway is an essential actor of the pro-inflammatory and anti-apoptotic activity67–69.\n\nNF-kB pathway increases angiogenesis, proliferation, anti-apoptosis, metastasis and inhibition of differentiation70.\n\nFF has the capacity to down-regulate NF-kB activity according to evidences gathered in Table 4. Through this PPARα-dependent mechanism, FF exerts anti-inflammatory activity. Besides, it also has non PPARα-dependent anti-inflammatory activity through up-regulation of SHP (small heterodimer partner).\n\nEvidences reported in Table 4 strongly support the FF anti-inflammatory activity mediated through NF-kB down-regulation and also PPARα independent mechanisms.\n\nOne of the proposed mechanisms of FF inhibiting NF-kB activity is depicted in Figure 5.\n\nThe research studies reported in Table 5 are evidence of down-regulation of Akt phosphorylation by FF, but Piwowarczyk et al.83 working with prostate cancer cells (DU-145) and endothelial cells (HUVEC) co-cultures found that FF increased levels of phosphorylated Akt in both HUVEC and DU-145 cells. They found that Akt phosphorylation was essential for FF increase of endothelial barrier (Figure 6).\n\nMitochondrial uncoupling proteins (UCP) are mitochondrial anion carrier proteins that separate oxidative phosphorylation from ATP synthesis with energy lost as heat and reduction of mitochondrial membrane potential90. The main function of UCP2 is the control of mitochondria-derived reactive oxygen species. PPARα modulates UCP2 expression91. Pecker et al. have demonstrated that UCP2 exerts control on proliferation: cells (embryonic fibroblast) where UCP2 expression was down-regulated grew faster than cells expressing UCP292. They also found that loss of UCP2 produced a metabolic change toward glucose metabolism, decreased fatty acid oxidation and increased proliferation.\n\nEvidence supports that FF decreases intracellular energy through inhibition of mitochondrial enzymes in a similar way as metformin. On a theoretical basis, we may assume that there might be synergism with metformin on this ground.\n\nPPARα involvement with cancer metabolism has been extensively reviewed by Grabacka and Reiss93.\n\nAnother enzyme down-regulated by FF is FAS (fatty acid synthase)94 which is highly expressed in many cancer tissues. Fatty acid synthase (FAS) is a multicomplex enzyme that intervenes in endogenous synthesis of fatty acids and particularly palmitate. Abnormal fatty acid (FA) synthesis is one of the common features of many cancer cells and FAS has been identified as part of cancer controlling networks. Human cancers that over-express FAS, are usually associated with poor prognosis95–99.\n\nThe expression of adhesion molecules on the endothelial cell surface is critical for cells rolling in the vascular lumen to achieve tethering and adhesion to the vascular wall and eventually achieving diapedesis and colonization in the case of potentially metastatic cells or leukocyte recruitment to atherosclerotic lesions.\n\nPPARα regulates gene expression of certain adhesion molecules in response to unsaturated fatty acids and fibric acid derivatives like FF. This control is achieved probably through inhibition of TNFα induced NF-kB activation100.\n\nThe research by Marchesi et al.100 that demonstrated reduction in adhesion molecules with FF treatment is important for two reasons:\n\n1) It was performed in humans (10 hypertriglyceridemic patients).\n\n2) The amount of reduction in fasting conditions (near 45% reduction for ICAM and around 33% reduction for VCAM levels).\n\nEmpen et al. (2003) described 10% reduction of E-selectin after six weeks treatment with FF, but found no major changes with VCAM-1 and ICAM-1 levels12.\n\nPiwowarczyc et al.83 demonstrated a new FF effect: increased endothelial cell adhesion to the susbstratum and increased adhesion between endothelial cells by activation of focal adhesion kinase (FAK). These impedes cell diapedesis through the vessel wall, which is an important objective to decrease metastatic risk. Figure 6 depicts this activity.\n\nThe production of the metastatic cascade is a complex process in which there are many successive steps that we shall not analyze in depth in this review.\n\nBut for a better understanding lets remember the main steps107:\n\n1) Primary tumor growth and angiogenesis.\n\n2) Future metastatic cells free themselves from the primary tumor.\n\n3) These cells degrade surrounding matrix.\n\n4) Reach endothelium of vessels.\n\n5) Enter blood vessels.\n\n6) Circulate and survive in circulation.\n\n7) Reach the target organ.\n\n8) Attach to endothelial cells.\n\n9) Migrate through vessel wall.\n\n10) Start growth in the colonized site including angiogenesis.\n\n11) Produce new metastatic cells.\n\n12) Reinitiate the whole process.\n\nCell motility, invasion, angiogenesis and the function of connexins and adhesion molecules are an essential part of this cascade and have already been considered. In table 8 we describe only specific research work relative to metastasis and FF.\n\nCOX-2 is the rate-limiting enzyme in prostaglandin synthesis that catalyzes the production of prostaglandins and thromboxanes from arachidonic acid, and has been associated with growth regulation and carcinogenesis in many tumors. The COX2/PGE2 pathway may be considered a pro-tumor pathway at least in certain cancers where elevated levels of COX2 have been identified. Most colorectal carcinomas and many adenomas exhibit this elevation109–110. One of the postulated mechanisms by which COX2/PGE2 signaling stimulates cell growth is through the activation of β-catenin111. COX2 is implicated in breast cancer progression and invasiveness112,114. In stage III breast cancer, COX2 over-expression is an unfavorable prognostic sign, and according to Kim et al. gives ground for using COX2 inhibitor combinations113. Simeone et al. identified the pathway leading to increased invasiveness in breast cancer: COX2 /protein kinase C/interleukin-8/urokinase-type plasminogen activator pathway115. COX2 is also associated with angiogenesis and metastasis116.\n\nMany cancers harbor increased COX2 activity including lung, colorectal, breast and squamous cell carcinoma of the upper digestive system117–119. COX2 down-regulation is an important issue in many cancers. We have been describing the action of FF on different pro-tumor proteins in an artificially separate manner, but many of these proteins share their activity in the pro-tumor evolution or are part of the same pathway. This is the case of NF-kB and COX2 in the progression towards cancer of Barret’s esophagus in which increased NF-kB activity is linked to increased IL-8 and COX2 expression120. As described above, FF is active against both: NF-kB and COX2.\n\nThe anticancer activities of FF are pleiotropic. Besides proliferation and angiogenesis down-regulation representing the main anti-tumoral effects, there are many others that will be described in the next three tables like ovarian aromatase inhibition, AMPK activation, IGF-I down-regulation, etc.\n\nThe IGF-1 receptor signaling system is a contributing factor in invasion, migration and proliferation of glioblastoma and became a legitimate target in the treatment of this pathology128. FF has experimentally shown to inhibit this system and decrease growth and invasion82, 126,127. A sort of IGF-1 trap was designed by D'Ambrosio et al.129 that inhibited tumor growth in vivo and induced apoptosis.\n\nThese publications reported contrasting results: two of the showed that FF increased radiation sensitivity133,134 and one showed decreased radiation sensitivity132. As we shall see latter, PPARα agonists are tissue-specific and species-specific. This may explain the difference. In the first case the experiments were performed on HeLa cells and in the second and third studies, experiments were performed on squamous cell carcinoma cells. In all three cases human cells were used, so the difference may lay in tissue-specific behavior or may be due to the fact that in the second experiment the environment was particularly hypoxic.\n\nSemaphorins are a large family of axon guidance molecules. They interact with their receptors, plexins and neuropilins, and play important roles in a growing list of diverse biological systems, including cancer progression and tumor angiogenesis. Some semaphorins can activate tumor progression and angiogenesis, while others may have the opposite effect.\n\nThere is abundant literature on semaphorins 3,4 and 7. Little is known about semaphorin-6B. It is known that there is an association between gastric cancer137, gliobastoma135 and certain breast cancers (MCF-7 breast adenocarcinoma cell line136) and semaphoring-6B but the exact nature of this association is still poorly understood.\n\nPPARα agonists clofibrate and fenofibrate can down-regulate semaphorin-6B gene expression.\n\nAccording to Ge137, inhibition of semaphoring-6B expression via RNA interference inhibited the migration, adhesion and invasion abilities of SGC-7901 gastric cancer cells in vitro.\n\nSEMA6B transcript was down-regulated in two human glioblastoma cell lines (T98G and A172) when a prolonged treatment with ATRA was performed138.\n\nThe SEMA6B gen has a PPARα binding site in the promoter region so that the interrelation of PPARα and the gene effectively is present, and this interrelation is a negative one and probably exerts anti-tumoral effects.\n\nIn a large population study in Finland to assess the overall risk of prostate cancer in people taking cholesterol lowering medication, no decrease in risk was seen either with statins nor fibrates142. The statistics is significantly biased regarding fibrates: the population is too small (around 220 patients with prostate cancer receiving fibrates out of a total number of prostate cancer cases of 24.723).\n\nHIV treatment with protease inhibitors has a frequent side effect: lipodystrophy syndrome. FF has been very effective in the treatment of this syndrome143. Nelfinavir and ritonavir are protease inhibitors which have shown many anti-cancer activities. We have postulated nelfinavir as a complementary off target treatment for cancer144. The association of FF with nelfinavir may prevent lipodystrophy, hypertriglyceridemia and elevation of lipoparticules145–147. There is no experimental proof of synergy of this interaction regarding cancer treatment, but both drugs share certain characteristics that make synergy a very plausible feature.\n\nNelfinavir is a proteasome inhibitor, and proteasome is the site where PPARα is dissembled, so nelfinavir should prolong PPARα’s life. Moreover nelfinavir and FF share the following activities: Akt inhibition, anti-angiogenesis, decreased proliferation, increased apoptosis, reduction of MMP2 and MMP9, and increased p21144.\n\nTable 17 shows similarities and differences between nelfinavir and FF. The main purpose of this table is to illustrate the similarities and differences between this two drugs regarding cancer: nelfinavir down-regulates CDK-2 and FF down-regulates cyclin D-1 reinforcing cell cycle slowdown. Both are strong anti-angiogenic agents. FF also down-regulates COX2 which is an important step in decreasing angiogenesis.\n\nNote: Blank cells mean that the activity has not been checked or has not been published.\n\n\nDiscussion\n\nPPARα is highly expressed in certain cancer cells like endometrium27, prostate148, bladder149, certain breast cancer cell lines150, NSCLC152 and others.\n\nIn these cases, the administration of a PPARα agonist like FF increased apoptosis and decreased proliferation.\n\nIn human breast cancer cell lines MCF-7 and MDA-MB-231, PPARα was overexpressed but use of agonists of this receptor increased proliferation150.\n\nIn stark contrast, Li et al.152 tested FF in 12 breast cancer cell lines (including MDA-MB-231) and in all of them FF was effective as proliferation inhibitor. This effectiveness was independent of PPARα expression but was linked to triple negative condition. Paradoxically MDA-MB-231 was the more sensitive to inhibition of proliferation by FF treatment.\n\nIn human colon cancer tissues PPARα is underexpressed when compared with normal tissues151. Using PPARα ligands in APCMin/+ mice to evaluate polyp formation, those treated showed decreased number of polyps and decreased size. We may conclude that overexpression or under-expression of PPARα in cancer tissues is not an indicator of future response to FF or other PPARα agonists.\n\nAnother issue to consider is the species-specific response to PPARα agonists: for instance, FF induces hepatocarcinogenesis in rodents but not in humans, insulin resistance in mouse but not in humans, oxidative stress in mouse heart but not in human heart. Human liver has a lower expression of PPARα than rodents153 (The differences of PPARα in human liver has been extensively described in the review by Roberts154).\n\nThere are important differences between rat and human hepatic cells. Vanden Heuvel et al.155 studied the gene expression differences between rat hepatoma cells (FaO) and human hepatocarcinoma cells (HepG2) when treated with a PPARα agonist like WY14643. A large number of kinases and phosphatases were affected in FaO and not in HepG2 cells. Many of them were implicated in cell cycle control and growth signaling like JAK1, JAK2, GSK3α and MKP-1.\n\nRat peroxisomes contain urate oxidase which is absent in human peroxisomes156.\n\nAcyl Co A oxidase (ACO) is a key enzyme in peroxisomes and according to Roberts et al.154 there are differences between human and rat ACO: the promoter for human ACO has a different sequence and activity from rats ACO.\n\nPPARα agonists like FF are tissue-specific. They may increase VEGF and angiogenesis in retina22,23 and the opposite in tumor cells20,21.\n\nThus this species- and tissue-specificity suggests that we should be cautious when interpreting the results of many of the published investigations. PPARα agonists research results obtained in rodents should not be taken for granted in humans.\n\nThe mechanism of FF anticancer activity may differ in different tumors: in mantle cell lymphoma it seems to induce apoptosis by inhibiting the TNFα/NF-kB axis39 while in triple negative breast cancer it requires activation of NF-kB in order to induce apoptosis41.\n\nPPARα agonists like FF are actively investigated as anti-cancer drugs, but paradoxically, PPARα inhibitors may also work against cancer. This is the case of renal cell carcinoma46 where an inhibitor of this receptor produced cell cycle arrest and apoptosis.\n\nAfter these necessary clarifying precautions, we must consider in detail the hard evidence collected in the medical literature that gives ground for FF as a complementary adjunct pharmaceutical in cancer therapy.\n\nPhanigraphy et al.21 tested 19 human tumor cell lines in vitro and found that all expressed PPARα in tumor cells and endothelium. There were differences regarding levels of PPARα. Fibrates and particularly FF at clinically achievable concentrations showed capacity to inhibit proliferation in these cell lines that included highly malignant ones like melanoma and Lewis lung carcinoma cell lines. FF inhibited 95% of bovine capillary endothelial cells proliferation and migration. There was no inhibition on normal fibroblasts growth. In glioblastoma cells FF reduced VEGF secretion by 50% and increased TSP-1 expression by 3- to 4- fold in a fibrosarcoma cell line. The results of these experiments show that FF exerts anti-proliferative and anti-angiogenic activities in tumor cells in vitro at clinically achievable concentrations. Similar results were observed in subcutaneously implanted tumors in mice. The seminal work of Panigraphy et al. concludes that anti-proliferative and anti-angiogenic properties of FF are PPARα activation dependent.\n\nGlioblastoma treatment has already introduced FF as part of different anti-angiogenic schedules, but FF may have also other anti-tumoral effects in this disease besides anti-angiogenesis, particularly apoptosis157–159.\n\nEndometrial carcinoma usually overexpresses PPARα. When PPARα expression is reduced with siRNA in vitro, cellular proliferation decreased substantially and showed a small increase in apoptosis in Ishikawa cells and HEC-1A cells28. Both cell lines reduced VEGF levels when they were treated with FF. Reduction in VEGF after FF treatment showed differences between the two cell lines: HEC-1A showed potentiation of inhibition of VEGF when an RXR ligand was added; Ishikawa cells did not. This reinforces the concept that there may appear important differences with FF treatment according to the tumor type.\n\nFF also uses the AMPK pathway to produce its anti-tumoral effects160. This is a PPARα independent action that has been demonstrated in human oral squamous cell carcinoma (OSCC) where FF inhibited cell migration and invasion and reduced expression of MMP 1, 2, 7 and 9. LKB1 and AMPK were up-regulated after FF treatment. When AMPK was inhibited (with protein C) the anti-invasive effect was significantly reduced.\n\nMetformin is another activator of the AMPK pathway. It has not been tested with FF but it is quite possible that there might be synergy in this activity. Another coincidence between metformin and FF is the decrease in cellular energy production.\n\nThe molecular mechanism of many of the FF anti-cancer activities is known since 2006161 and can be summarized as inhibition of COX2 and VEGF at the transcription level by interfering with AP-1 binding to DNA and decreased expression of NF-kB. In susceptible cells there is a negative cross talk between PPARα and AP-1161.\n\nAnother mechanism postulated in anti-cancer activity is the disruption of the tumor-host stroma symbiosis due to anti-angiogenesis and anti-inflammatory activity162.\n\nFinally we have to mention that FF effectively reduces nuclear SREBP-2 but not SREBP-1163–166. Nelfinavir and PUFA inhibit mature SREB-1 which is a transcription factor that promotes FAS synthesis, so these products may complement FF anti-cancer activity.\n\n\nConclusion\n\nAs observed in the tables, FF exerts polyvalent anti-cancer activities that deserve further research in the clinical setting.\n\nFF is a PPARα agonist drug developed for treatment of elevated triglycerides and LDL cholesterol reducing cardiovascular risk that may be repurposed to be used in cancer due to its anti-angiogenic, anti-inflammatory, anti-proliferative, anti-metastatic, anti-adhesive, anti-invasive and pro-apoptotic activities in certain cancers.\n\n\nFuture perspectives\n\nFF has already been incorporated in anti-angiogenic protocols for the treatment of glioblastoma. Probably in the future it will form part of protocols based on repurposed drugs directed to inhibit angiogenenesis like celecoxib, nelfinavir, and metformin.\n\nColorectal and prostate cancer seem good candidates for these therapies.", "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\nReferences\n\nYang LP, Keating GM: Fenofibric acid: in Combination therapy in the treatment of mixed dsylipidemia. Am J Cardiovasc Drugs. 2009; 9(6): 401–409. PubMed Abstract \| Publisher Full Text\n\nStaels B, Vu-Dac N, Kosykh VA, et al.: Fibrates downregulate apolipoprotein C-III expression independent of induction of peroxisomal acyl coenzyme A oxidase. A potential mechanism for the hypolipidemic action of fibrates. J Clin Invest. 1995; 95(2): 705–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarx N, Sukhova GK, Collins T, et al.: PPARalpha activators inhibit cytokine-induced vascular cell adhesion molecule-1 expression in human endothelial cells. Circulation. 1999; 99(24): 3125–3131. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShu H, Wong B, Zhou G, et al.: Activation of PPARalpha or gamma reduces secretion of matrix metalloproteinase 9 but not interleukin 8 from human monocytic THP-1 cells. Biochem Biophys Commun. 2000; 267(1): 345–349. PubMed Abstract \| Publisher Full Text\n\nRival Y, Benéteau N, Taillandier T, et al.: PPARalpha and PPARdelta activators inhibit cytokine-induced nuclear translocation of NF-kappaB and expression of VCAM-1 in EAhy926 endothelial cells. Eur J Pharmacol. 2002; 435(2–3): 143–51. PubMed Abstract \| Publisher Full Text\n\nGoetze S, Eilers F, Bungenstock A, et al.: PPAR activators inhibit endothelial cell migration by targeting Akt. Biochem Biophys Res Commun. 2002; 293(5): 1431–7. PubMed Abstract \| Publisher Full Text\n\nJiao HL, Zhao BL: Cytotoxic effect of peroxisome proliferator fenofibrate on human HepG2 hepatoma cell line and relevant mechanisms. Toxicol Appl Pharmacol. 2002; 185(3): 172–9. PubMed Abstract \| Publisher Full Text\n\nVaret J, Vincent L, Mirshahi P, et al.: Fenofibrate inhibits angiogenesis in vitro and in vivo. Cell Mol Life Sci. 2003; 60(4): 810–9. PubMed Abstract \| Publisher Full Text\n\nLu Y, Cai Z, Galson DL, et al.: Monocyte chemotactic protein-1 (MCP-1) acts as a paracrine and autocrine factor for prostate cancer growth and invasion. Prostate. 2006; 66(12): 1311–1388. PubMed Abstract \| Publisher Full Text\n\nSonoki K, Iwase M, Iino K, et al.: Dilazep and fenofibric acid inhibit MCP-1 mRNA expression in glycoxidized LDL-stimulated human endothelial cells. Eur J Pharmacol. 2003; 475(1–3): 139–147. PubMed Abstract \| Publisher Full Text\n\nMeissner M, Stein M, Urbich C, et al.: PPARalpha activators inhibit vascular endothelial growth factor receptor-2 expression by repressing Sp1 dependent DNA binding and transactivation. Circ Res. 2004; 94(3): 324–332. PubMed Abstract \| Publisher Full Text\n\nEmpen K, Frost RJ, Geiss HC, et al.: Differential effects of fenofibrate versus atorvastatin on the concentrations of E-selectin and vascular cellular adhesion molecule-1 in patients with type 2 diabetes mellitus and mixed hyperlipoproteinemia: a randomized cross-over trial. Cardiovascular Diabetology. 2003; 2: 17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolland CM, Saidi SA, Evans AL, et al.: Transcriptome analysis of endometrial cancer identifies peroxisome proliferator-activated receptors as potential therapeutic targets. Mol Cancer Ther. 2004; 3(8): 993–1001. PubMed Abstract\n\nGrabacka M, Placha W, Plonka PM, et al.: Inhibition of melanoma metastases by fenofibrate. Arch Dermatol Res. 2004; 296(2): 54–8. PubMed Abstract \| Publisher Full Text\n\nGunin AG, Bitter AD, Demakov AB, et al.: Effects of peroxisome proliferator activated receptors-alpha and -gamma agonists on estradiol-induced proliferation and hyperplasia formation in the mouse uterus. J Endocrinol. 2004; 182(2): 229–39. PubMed Abstract \| Publisher Full Text\n\nKubota T, Yano T, Fujisaki K, et al.: Fenofibrate induces apoptotic injury in cultured human hepatocytes by inhibiting phosphorylation of Akt. Apoptosis. 2005; 10(2): 349–358. PubMed Abstract \| Publisher Full Text\n\nOkopień B, Kowalski J, Krysiak R, et al.: Monocyte suppressing action of fenofibrate. Pharmacol Rep. 2005; 57(3): 367–72. PubMed Abstract\n\nOkamoto H, Iwamoto T, Kotake S, et al.: Inhibition of NF-kappaB signaling by fenofibrate, a peroxisome proliferator-activated receptor-alpha ligand, presents a therapeutic strategy for rheumatoid arthritis. Clin Exp Rheumatol. 2005; 23(3): 323–30. PubMed Abstract\n\nBlann AD, Belgore FM, Constans J, et al.: Plasma vascular endothelial growth factor and its receptor Flt-1 in patients with hyperlipidemia and atherosclerosis and the effects of fluvastatin or fenofibrate. Am J Cardiol. 2001; 87(10): 1160–1163. PubMed Abstract \| Publisher Full Text\n\nPozzi A, Popescu V, Yang S, et al.: The anti-tumorigenic properties of peroxisomal proliferator-activated receptor alpha are arachidonic acid epoxygenase-mediated. J Biol Chem. 2010; 285(17): 12840–12850. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPanigraphy D, Kaipainen A, Huang S, et al.: PPARalpha agonist fenofibrate suppresses tumor growth through direct and indirect angiogenesis inhibition. Proc Natl Acad Sci U S A. 2008; 105(3): 985–990. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhou J, Zhang S, Xue J, et al.: Activation of peroxisome proliferator-activated receptor α (PPARα) suppresses hypoxia inducible factor-1α (HIF-1α) signaling in cancer cells. J Biol Chem. 2012; 287(42): 35161–35169. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDana N, Javanmard SH, Fazilati M, et al.: A comparison of peroxisome proliferator-activated receptor-α agonist and antagonist on human umbilical vein endothelial cells angiogenesis. Adv Biomed Res. 2013; 2(1): 54. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeisnner M, Stein M, Urbich C, et al.: PPARalpha Activators Inhibit Vascular Endothelial Growth Factor Receptor-2 Expression by Repressing Sp1 Dependent DNA Binding and Transactivation. Circ Res. 2004; 94(3): 324–332. PubMed Abstract \| Publisher Full Text\n\nKim J, Ahn JH, Kim JH, et al.: Fenofibrate regulates retinal endothelial cell survival through the AMPK signal transduction pathway. Exp Eye Res. 2007; 84(5): 886–893. PubMed Abstract \| Publisher Full Text\n\nChew EY, Ambrosius WT, Davis MD, et al.: Effects of medical therapies on retinopathy progression in type 2 diabetes. N Engl J Med. 2010; 363(3): 233–244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNoonan JE, Jenkins AJ, Ma JX, et al.: An update on the molecular actions of Fenofibrate and its clinical effects on diabetic retinopathy and other microvascular end points in patients with diabetes. Diabetes. 2013; 62(12): 3968–3975. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNickkho-Amiry M, McVey R, Holland C: Peroxisome Proliferator–Activated Receptors Modulate Proliferation and Angiogenesis in Human Endometrial Carcinoma. Mol Cancer Res. 2012; 10(3): 441–453. PubMed Abstract \| Publisher Full Text\n\nOnalan G, Zeyneloglu HB, Bayraktar N: Fenofibrate causes regression of endometriotic implants: a rat model. Fertil Steril. 2009; 92(6): 2100–2102. PubMed Abstract \| Publisher Full Text\n\nRobison NJ, Campigotto F, Chi SN, et al.: A phase II trial of a multi-agent oral antiangiogenic (metronomic) regimen in children with recurrent or progressive cancer. Pediatr Blood Cancer. 2014; 61(4): 636–642. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeyrl A, Chocholous M, Kieran MW, et al.: Antiangiogenic metronomic therapy for children with recurrent embryonal brain tumors. Pediatr Blood Cancer. 2012; 59(3): 511–517. PubMed Abstract \| Publisher Full Text\n\nPeyrl A, Chocholous M, Kieran MW, et al.: Metronomic antiangiogenic therapy in children with recurrent brain tumours of different histologies. Magazine of European Medical Oncology. 2011; 4(3): 21–23. Publisher Full Text\n\nSterba J, Pavelka Z, Andre N, et al.: Second complete remission of relapsed medulloblastoma induced by metronomic chemotherapy. Pediatr Blood Cancer. 2010; 54(4): 616–617. PubMed Abstract \| Publisher Full Text\n\nZapletalova D, André N, Deak L, et al.: Metronomic Chemotherapy with the COMBAT Regimen in Advanced Pediatric Malignancies: A Multicenter Experience. Oncology. 2012; 82(5): 249–260. PubMed Abstract \| Publisher Full Text\n\nPozzi A, Ibanez MR, Gatica AE, et al.: Peroxisomal proliferator-activated receptor-alpha-dependent inhibition of endothelial cell proliferation and tumorigenesis. J Biol Chem. 2007; 282(24): 17685–17695. PubMed Abstract \| Publisher Full Text\n\nPozzi A, Capdevila JH: PPARalpha ligands as antitumorigenic and antiangiogenic agents. PPAR Res. 2008; 2008: 906542. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAvis I, Hong SH, Martinez A, et al.: Five-lipoxygenase inhibitors can mediate apoptosis in human breast cancer cell lines through complex eicosanoid interactions. FASEB J. 2001; 15(11): 2007–9. PubMed Abstract \| Publisher Full Text\n\nChinetti G, Griglio S, Antonucci M, et al.: Activation of Proliferator-activated Receptors alpha and gamma Induces Apoptosis of Human Monocyte-derived Macrophages. J Biol Chem. 1998; 273(40): 25573–25580. PubMed Abstract \| Publisher Full Text\n\nZak Z, Gelebart P, Lai RR: Fenofibrate induces effective apoptosis in mantle cell lymphoma by inhibiting the TNFalpha/NF-kappaB signaling axis. Leukemia. 2010; 24(8): 1476–1486. PubMed Abstract \| Publisher Full Text\n\nWang J, Wang T, Zeng J, et al.: Experimental study of fenofibrate combined with cisplatin on inhibition and apoptosis of human lung cancer A549 cell line. Shandong Medical Journal. 2011; 48. Reference Source\n\nLi T, Zhang Q, Zhang J, et al.: Fenofibrate induces apoptosis of triple-negative breast cancer cells via activation of NF-κB pathway. BMC Cancer. 2014; 14: 96. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBinello E, Mormone E, Emdad L, et al.: Characterization of fenofibrate-mediated anti-proliferative pro-apoptotic effects on high-grade gliomas and anti-invasive effects on glioma stem cells. J Neurooncol. 2014; 117(2): 225–234. PubMed Abstract \| Publisher Full Text\n\nMajeed Y, Taba T, Shaheen Y: Molecular basis of the anti-proliferative action of the hypolipidemic drug fenofibrate in vitro in angiosarcoma-forming endothelial cells. Qatar Foundation Annual Research Conference Proceedings Volume 1 HBPP0064. 2014. Reference Source\n\nPenna F, Pin F, Costamagna D, et al.: Caspase 2 Activation and ER Stress Drive Rapid Jurkat Cell Apoptosis by Clofibrate. PLoS One. 2012; 7(9): e45327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPenna F, Reffo P, Muzio G, et al.: Mechanisms of clofibrate-induced apoptosis in Yoshida AH-130 hepatoma cells. Biochem Pharmacol. 2009; 77(2): 169–176. PubMed Abstract \| Publisher Full Text\n\nAbu Aboud O, Wettersten HI, Weiss RH: Inhibition of PPARα Induces Cell Cycle Arrest and Apoptosis, and Synergizes with Glycolysis Inhibition in Kidney Cancer Cells. PLoS One. 2013; 8(8): e71115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPascal M, Sepulchre C, Chazan JB, et al.: Evidence for the inhibition of platelet-derived growth factor induced rat smooth muscle cells DNA synthesis by fenofibric acid at the Go/G1 cell cycle level. Life Sci. 1983; 33(10): 925–33. PubMed Abstract \| Publisher Full Text\n\nMunro E, Patel M, Chan P, et al.: Growth inhibition of human vascular smooth muscle cells by fenofibrate: a possible therapy for restenosis. Cardiovasc Res. 1994; 28(5): 615–20. PubMed Abstract \| Publisher Full Text\n\nZeng R, Xiong Y, Zhu F, et al.: Fenofibrate attenuated glucose-induced mesangial cells proliferation and extracellular matrix synthesis via PI3K/AKT and ERK1/2. PLoS One. 2013; 8(10): e76836. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong Q, Sun J, Wang J, et al.: Role of TRB3 in the inhibitory effect of fenofibrate against high glucose-induced proliferation of glomerular mesangial cells. Nan Fang Yi Ke Da Xue Xue Bao. 2013; 33(3): 391–6. PubMed Abstract \| Publisher Full Text\n\nSelim E, Frkanec JT, Cunard R: Fibrates upregulates TRB3 in lymphocytes independent of PPAR alpha by augmenting CCAAT/enhancer binding protein beta (C/EBP beta) expression. Mol Immunol. 2007; 44(6): 1218–1229. PubMed Abstract \| Publisher Full Text\n\nMorse E, Selim E, Cunard R: PPARalpha ligands cause lymphocyte depletion and cell cycle block and this is associated with augmented TRB3 and reduced Cyclin B1 expression. Mol Immunol. 2009; 46(16): 3454–3461. PubMed Abstract \| Publisher Full Text\n\nOgata T, Miyauchi T, Irukayama-Tomobe Y, et al.: The peroxisome proliferator-activated receptor alpha activator fenofibrate inhibits endothelin-1-induced cardiac fibroblast proliferation. J Cardiovasc Pharmacol. 2004; 44(Suppl 1): S279–82. PubMed Abstract\n\nIrukayama-Tomobe Y, Miyauchi T, Sakai S, et al.: Endothelin-1-induced cardiac hypertrophy is inhibited by activation of peroxisome proliferator-activated receptor-alpha partly via blockade of c-Jun NH2-terminal kinase pathway. Circulation. 2004; 109(7): 904–910. PubMed Abstract \| Publisher Full Text\n\nIrukayama-Tomobe Y, Miyauchi T, Kasuya Y, et al.: Activation of peroxisome proliferator-activated receptor-alpha decreases endothelin-1-induced p38 mitogen-activated protein kinase activation in cardiomyocytes. J Cardiovasc Pharmacol. 2004; 44(Suppl 1): S358–61. PubMed Abstract \| Publisher Full Text\n\nLee JK, Seo EM, Lee SS, et al.: Activation of PPARalpha Attenuates IFNgamma and IL-1beta-induced Cell Proliferation in Astrocytes: Involvement of IL-6 Independent Pathway. Korean J Physiol Pharmacol. 2010; 14(3): 185–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThuillier P, Anchiraico GJ, Nickel KP, et al.: Activators of peroxisome proliferator–activated receptor-alpha partially inhibit mouse skin tumor promotion. Mol Carcinog. 2000; 29(3): 134–142. PubMed Abstract \| Publisher Full Text\n\nGizard F, Amant C, Barbier O, et al.: PPAR alpha inhibits vascular smooth muscle cell proliferation underlying intimal hyperplasia by inducing the tumor suppressor p16INK4a. J Clin Invest. 2005; 115(11): 3228–3238. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaidi SA, Holland CM, Charnock-Jones DS, et al.: In vitro and in vivo effects of the PPAR-alpha agonist fenofibrate and retinoic acid in endometrial cancer. Mol Cancer. 2006; 5: 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYokoyama Y, Xin B, Shigeto T, et al.: Clofibric acid, a peroxisome proliferator-activated receptor alpha ligand, inhibits growth of human ovarian cancer. Mol Cancer Ther. 2007; 6(4): 1379–86. PubMed Abstract \| Publisher Full Text\n\nYamasaki D, Kawabe N, Nakamura H, et al.: Fenofibrate suppresses growth of the human hepatocellular carcinoma cell via PPARα-independent mechanisms. Eur J Cell Biol. 2011; 90(8): 657–64. PubMed Abstract \| Publisher Full Text\n\nChang NW, Tsai MH, Lin C, et al.: Fenofibrate exhibits a high potential to suppress the formation of squamous cell carcinoma in an oral-specific4-nitroquinoline1-oxide/arecoline mouse model. Biochim Biophys Acta. 2011; 1812(4): 558–64. PubMed Abstract \| Publisher Full Text\n\nHuang J, Das SK, Jha P, et al.: The PPARα agonist fenofibrate suppresses B-cell lymphoma in mice by modulating lipid metabolism. Biochim Biophys Acta. 2013; 1831(10): 1555–1565. PubMed Abstract \| Publisher Full Text\n\nWang H, He CH, Bai LP, et al.: Effects of fenofibrate on the growth and migration of ovarian cancer cells in vitro. Sichuan Da Xue Xue Bao Yi Xue Ban. 2014; 45(5): 789–92. PubMed Abstract\n\nLiang H, Kowalczyk P, Junco JJ, et al.: Differential effects on lung cancer cell proliferation by agonists of glucocorticoid and PPARα receptors. Mol Carcinog. 2014; 53(9): 753–763. PubMed Abstract \| Publisher Full Text\n\nZhang C, Beckermann B, Kallifatidis G, et al.: Corticosteroids induce chemotherapy resistance in the majority of tumour cells from bone, brain, breast, cervix, melanoma and neuroblastoma. Int J Oncol. 2006; 29(5): 1295–1301. PubMed Abstract \| Publisher Full Text\n\nBeg AA, Baltimore D: An essential role for NF-kappaB in preventing TNF alfa-induced cell death. Science. 1996; 274(5288): 782–784. PubMed Abstract \| Publisher Full Text\n\nAntwerp DJV, Martin SJ, Kafri T, et al.: Suppression of TNF alfa induced apoptosis by NF-kappaB. Science. 1996; 274(5288): 787–789. PubMed Abstract \| Publisher Full Text\n\nWang CY, Mayo MW, Baldwin AS Jr: TNF- and cancer therapy-induced apoptosis: potentiation by inhibition of NF-kappaB. Science. 1996; 274(5288): 784–787. PubMed Abstract \| Publisher Full Text\n\nBaldwin AS: Control of oncogenesis and cancer therapy resistance by the transcription factor NF-κB. J Clin Invest. 2001; 107(3): 241–246. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStaels B, Koenig W, Habib A, et al.: Activation of human aortic smooth-muscle cells is inhibited by PPARalpha but not by PPARgamma activators. Nature. 1998; 393(6687): 790–3. PubMed Abstract \| Publisher Full Text\n\nXu X, Otsuki M, Saito H, et al.: PPARalpha and GR differentially down-regulate the expression of nuclear factor-kappaB-responsive genes in vascular endothelial cells. Endocrinology. 2001; 142(8): 3332–9. PubMed Abstract \| Publisher Full Text\n\nOgata T, Miyahuchi T, Sakai S, et al.: Myocardial fibrosis and diastolic dysfunction in deoxycorticosterone acetate-salt hypertensive rats is ameliorated by the peroxisome proliferator-activated receptor-alpha activator fenofibrate, partly by suppressing inflammatory responses associated with the nuclear factor-kappa-B pathway. J Am Coll Cardiol. 2004; 43(8): 1481–1488. PubMed Abstract \| Publisher Full Text\n\nYang TL, Chen MF, Luo BL, et al.: Fenofibrate decreases asymmetric dimethylarginine level in cultured endothelial cells by inhibiting NF-kappaB activity. Naunyn Schmiedebergs Arch Pharmacol. 2005; 371(5): 401–7. PubMed Abstract \| Publisher Full Text\n\nXu J, Chavis JA, Racke MK, et al.: Peroxisome proliferator-activated receptor-alpha and retinoid X receptor agonists inhibit inflammatory responses of astrocytes. J Neuroimmunol. 2006; 176(1–2): 95–105. PubMed Abstract \| Publisher Full Text\n\nLi L, Emmett N, Mann D, et al.: Fenofibrate attenuates tubulointerstitial fibrosis and inflammation through suppression of nuclear factor-κB and transforming growth factor-β 1/Smad3 in diabetic nephropathy. Exp Biol Med (Maywood). 2010; 235(3): 383–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang CS, Yuk JM, Kim JJ, et al.: Small heterodimer partner-targeting therapy inhibits systemic inflammatory responses through mitochondrial uncoupling protein 2. PLoS One. 2013; 8(5): e63435. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShen W, Gao Y, Lu B, et al.: Negatively regulating TLR4/NF-κB signaling via PPARα in endotoxin-induced uveitis. Biochim Biophys Acta. 2014; 1842(7): 1109–20. PubMed Abstract \| Publisher Full Text\n\nBinello E, Germano IM: Targeting glioma stem cells: a novel framework for brain tumors. Cancer Sci. 2011; 102(11): 1958–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrabacka M, Plonka PM, Urbanska K, et al.: Peroxisome proliferator-activated receptor alpha activation decreases metastatic potential of melanoma cells in vitro via down-regulation of Akt. Clin Cancer Res. 2006; 12(10): 3028–36. PubMed Abstract \| Publisher Full Text\n\nLi R, Zheng W, Pi R, et al.: Activation of peroxisome proliferator-activated receptor-alpha prevents glycogen synthase 3beta phosphorylation and inhibits cardiac hypertrophy. FEBS Lett. 2007; 581(17): 3311–6. PubMed Abstract \| Publisher Full Text\n\nKuno T, Hata K, Takamatsu M, et al.: The peroxisome proliferator-activated receptor (PPAR) α agonist fenofibrate suppresses chemically induced lung alveolar proliferative lesions in male obese hyperlipidemic mice. Int J Mol Sci. 2014; 15(5): 9160–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPiwowarczyk K, Wybieralska E, Baran J, et al.: Fenofibrate enhances barrier function of endothelial continuum within the metastatic niche of prostate cancer cells. Expert Opin Ther Targets. 2015; 19(2): 163–76. PubMed Abstract \| Publisher Full Text\n\nKeller BJ, Yamanaka H, Thurman RG: Inhibition of mitochondrial respiration and oxygen-dependent hepatotoxicity by six structurally dissimilar peroxisomal proliferating agents. Toxicology. 1992; 71(1–2): 49–61. PubMed Abstract \| Publisher Full Text\n\nZhou Z, Wallace KB: The effect of peroxisome proliferators on mitochondrial bioenergetics. Toxicol Sci. 1999; 48(1): 82–89. PubMed Abstract \| Publisher Full Text\n\nCasas F, Pineau T, Rochard P, et al.: New molecular aspects of regulation of mitochondrial activity by fenofibrate and fasting. FEBS Lett. 2000; 482(1–2): 71–4. PubMed Abstract \| Publisher Full Text\n\nBrunmair B, Lest A, Staniek K, et al.: Fenofibrate impairs rat mitochondrial function by inhibition of respiratory complex I. J Pharmacol Exp Ther. 2004; 311(1): 109–114. PubMed Abstract \| Publisher Full Text\n\nNadanaciva S, Dykens JA, Bernal A, et al.: Mitochondrial impairment by PPAR agonists and statins identified via immunocaptured OXPHOS complex activities and respiration. Toxicol Appl Pharmacol. 2007; 223(3): 277–287. PubMed Abstract \| Publisher Full Text\n\nWilk AM, Urbanska K, Impastato D, et al.: Abstract 3230: Fenofibrate mediated energy crisis and apoptotic cell death of glioglastoma. Cancer Res. 2012; 72: 3230. Publisher Full Text\n\nLedesma A, de Lacoba MG, Rial E: The mitochondrial uncoupling proteins. Genome Biol. 2002; 3(12): REVIEWS3015. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFinck BN, Kelly DP: Peroxisome proliferator-activated receptor alpha (PPARalpha) signaling in the gene regulatory control of energy metabolism in the normal and diseased heart. J Mol Cell Cardiol. 2002; 34(10): 1249–57. PubMed Abstract \| Publisher Full Text\n\nPecqueur C, Bui T, Gelly C, et al.: Uncoupling protein-2 controls proliferation by promoting fatty acid oxidation and limiting glycolysis-derived pyruvate utilization. FASEB J. 2008; 22(1): 9–18. PubMed Abstract \| Publisher Full Text\n\nGrabacka M, Reiss K: Anticancer Properties of PPARalpha-Effects on Cellular Metabolism and Inflammation. PPAR Res. 2008; 2008: 9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSrivastava RA, Jahagirdar R, Azhar S, et al.: Peroxisome proliferator-activated receptor-alpha selective ligand reduces adiposity, improves insulin sensitivity and inhibits atherosclerosis in LDL receptor-deficient mice. Mol Cell Biochem. 2006; 285(1–2): 35–50. PubMed Abstract \| Publisher Full Text\n\nBaron A, Migita T, Tang D, et al.: Fatty acid synthase: a metabolic oncogene in prostate cancer? J Cell Biochem. 2004; 91(1): 47–53. PubMed Abstract \| Publisher Full Text\n\nHunt DA, Lane HM, Yygmont ME, et al.: MRNA stability and overexpression of fatty acid synthase in human breast cancer cell lines. Anticancer Res. 2007; 27(1A): 27–34. PubMed Abstract\n\nGansler TS, Hardman W 3rd, Hunt DA, et al.: Increased expression of fatty acid synthase (OA-519) in ovarian neoplasms predicts shorter survival. Hum Pathol. 1997; 28(6): 686–692. PubMed Abstract \| Publisher Full Text\n\nCamassei FD, Cozza R, Acquaviva A, et al.: Expression of the lipogenic enzyme fatty acid synthase (FAS) in retinoblastoma and its correlation with tumor aggressiveness. Invest Ophthal Vis Sci. 2003; 44(6): 2399–2403. PubMed Abstract \| Publisher Full Text\n\nYe Q, Chung LW, Li S, et al.: Identification of a novel FAS/ER-alpha fusion transcript expressed in human cancer cells. Biochim Biophys Acta. 2000; 1493(3): 373–377. PubMed Abstract \| Publisher Full Text\n\nMarchesi S, Lupattelli G, Lombardini R, et al.: Effects of fenofibrate on endothelial function and cell adhesion molecules during post-prandial lipemia in hypertriglyceridemia. J Clin Pharm Ther. 2003; 28(5): 419–424. PubMed Abstract \| Publisher Full Text\n\nGervois P, Kleeman R, Pilon A, et al.: Global suppression of IL-6-induced acute phase response gene expression after chronic in vivo treatment with the peroxisome proliferator-activated receptor-α activator fenofibrate. J Biol Chem. 2004; 279: 16154–16160. PubMed Abstract \| Publisher Full Text\n\nLefebvre P, Chinetti G, Fruchart JC, et al.: Sorting out the roles of PPAR alpha in energy metabolism and vascular homeostasis. J Clin Invest. 2006; 116(3): 571–580. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRyan KE, McCance DR, Powell L, et al.: Fenofibrate and pioglitazone improve endothelial function and reduce arterial stiffness in obese glucose tolerant men. Atherosclerosis. 2007; 194(2): e123–30. PubMed Abstract \| Publisher Full Text\n\nRosenson RS, Wolff DA, Huskin AL, et al.: Fenofibrate therapy ameliorates fasting and postprandial lipoproteinemia, oxidative stress, and the inflammatory response in subjects with hypertriglyceridemia and the metabolic syndrome. Diabetes Care. 2007; 30(8): 1945–51. PubMed Abstract \| Publisher Full Text\n\nLee JW, Baiwa PJ, Carson MJ, et al.: Fenofibrate represses interleukin-17 and interferon-gamma expression and improves colitis in interleukin-10-deficient mice. Gastroenterology. 2007; 133(1): 108–123. PubMed Abstract \| Publisher Full Text\n\nLi H, Cheng X, Yang X, et al.: The effects of fenobribate on endothelial cell secretion of E selectin and ICAM-1. Jiangsu Medical Journal. 2007; 01. Reference Source\n\nSamatov TR, Tonevitsky AG, Schumacher U: Epithelial-mesenchymal transition: focus on metastatic cascade, alternative splicing, non-coding RNAs and modulating compounds. Mol Cancer. 2013; 12: 107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWybieralska E, Szpak K, Górecki A, et al.: Fenofibrate attenuates contact-stimulated cell motility and gap junctional coupling in DU-145 human prostate cancer cell populations. Oncol Rep. 2011; 26(2): 447–453. PubMed Abstract \| Publisher Full Text\n\nEberhart CE, Coffey RJ, Radhika A, et al.: Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology. 1994; 107(4): 1183–1188. PubMed Abstract\n\nElder DJ, Baker JA, Banu NA, et al.: Human colorectal adenomas demonstrate a size dependent increase in epithelial cyclooxygenase-2 expression. J Pathol. 2002; 198(4): 428–434. PubMed Abstract \| Publisher Full Text\n\nCastellone MD, Teramoto H, Williams BO, et al.: Prostaglandin E2 promotes colon cancer cell growth through a Gs-axin-beta-catenin signaling axis. Science. 2005; 310(5753): 1504–1510. PubMed Abstract \| Publisher Full Text\n\nBocca C, Ievolella M, Autelli R, et al.: Expression of Cox-2 in human breast cancer cells as a critical determinant of epithelial-to-mesenchymal transition and invasiveness. Expert Opin Ther Targets. 2014; 18(2): 121–35. PubMed Abstract \| Publisher Full Text\n\nKim HS, Moon HG, Han W, et al.: COX2 overexpression is a prognostic marker for Stage III breast cancer. Breast Cancer Res Treat. 2012; 132(1): 51–9. PubMed Abstract \| Publisher Full Text\n\nSingh B, Berry JA, Shoher A, et al.: COX-2 overexpression increases motility and invasion of breast cancer cells. Int J Oncol. 2005; 26(5): 1393–9. PubMed Abstract \| Publisher Full Text\n\nSimeone AM, Nieves-Alicea R, McMurtry VC, et al.: Cyclooxygenase-2 uses the protein kinase C/ interleukin-8/urokinase-type plasminogen activator pathway to increase the invasiveness of breast cancer cells. Int J Oncol. 2007; 30(4): 785–92. PubMed Abstract \| Publisher Full Text\n\nCosta C, Soares R, Reis-Filho JS, et al.: Cyclo-oxygenase 2 expression is associated with angiogenesis and lymph node metastasis in human breast cancer. J Clin Pathol. 2002; 55: 429–434. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSoslow RA, Dannenberg AJ, Rush D, et al.: COX-2 is expressed in human pulmonary, colonic, and mammary tumors. Cancer. 2000; 89: 2637–2645. PubMed Abstract \| Publisher Full Text\n\nWang ZM, Liu J, Liu HB, et al.: Abnormal COX2 protein expression may be correlated with poor prognosis in oral cancer: a meta-analysis. Biomed Res Int. 2014; 2014: 364207. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpano JP, Chouahnia K, Morère JF, et al.: Cyclooxygenase 2 inhibitors and lung carcinoma. Bull Cancer. 2004; 91(Suppl 2): S109–12. PubMed Abstract\n\nKonturek PC, Nikiforuk A, Kania J, et al.: Activation of NFkappaB represents the central event in the neoplastic progression associated with Barrett's esophagus: a possible link to the inflammation and overexpression of COX-2, PPARgamma and growth factors. Dig Dis Sci. 2004; 49(7–8): 1075–83. PubMed Abstract \| Publisher Full Text\n\nChen XR, Besson VC, Palmier B, et al.: Neurological recovery-promoting, anti-inflammatory, and anti-oxidative effects afforded by fenofibrate, a PPAR alpha agonist, in traumatic brain injury. J Neurotrauma. 2007; 24(7): 1119–31. PubMed Abstract \| Publisher Full Text\n\nChen YJ, Quilley J: Fenofibrate treatment of diabetic rats reduces nitrosative stress, renal cyclooxygenase-2 expression, and enhanced renal prostaglandin release. J Pharmacol Exp Ther. 2008; 324(2): 658–63. PubMed Abstract \| Publisher Full Text\n\nLee DL, Wilson JL, Duan R, et al.: Peroxisome Proliferator-Activated Receptor-α Activation Decreases Mean Arterial Pressure, Plasma Interleukin-6, and COX-2 While Increasing Renal CYP4A Expression in an Acute Model of DOCA-Salt Hypertension. PPAR Res. 2011; 2011: 502631. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlvarez de Sotomayor M, Mingorance C, Andriantsitohaina R: Fenofibrate improves age-related endothelial dysfunction in rat resistance arteries. Atherosclerosis. 2007; 193(1): 112–120. PubMed Abstract \| Publisher Full Text\n\nRamanan S, Kooshki M, Zhao W, et al.: PPARalpha ligands inhibit radiation-induced microglial inflammatory responses by negatively regulating NF-kappaB and AP-1 pathways. Free Radic Biol Med. 2008; 45(12): 1695–1704. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUrbanska K, Pannizzo P, Grabacka M, et al.: Activation of PPARalpha inhibits IGF-I-mediated growth and survival responses in medulloblastoma cell lines. Int J Cancer. 2008; 123(5): 1015–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDrukala J, Urbanska K, Wilk A, et al.: ROS accumulation and IGF-IR inhibition contribute to fenofibrate/PPARalpha -mediated inhibition of glioma cell motility in vitro. Molecular Cancer. 2010; 9: 159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrojan J, Cloix JF, Ardourel MY, et al.: Insulin-like growth factor type I biology and targeting in malignant gliomas. Neuroscience. 2007; 145(3): 795–811. PubMed Abstract \| Publisher Full Text\n\nD’Ambrosio C, Ferber A, Resnicoff M, et al.: A soluble insulin-like growth factor I receptor that induces apoptosis of tumor cells in vivo and inhibits tumorigenesis. Cancer Res. 1996; 56(17): 4013–4020. PubMed Abstract\n\nToda K, Okada T, Miyaura C, et al.: Fenofibrate, a ligand for PPARalpha, inhibits aromatase cytochrome P450 expression in the ovary of mouse. J Lipid Res. 2003; 44(2): 265–270. PubMed Abstract \| Publisher Full Text\n\nZhao H, Zhu C, Qin C, et al.: Fenofibrate down-regulates the expressions of androgen receptor (AR) and AR target genes and induces oxidative stress in the prostate cancer cell line LNCaP. Biochem Biophys Res Commun. 2013; 432(2): 320–5. PubMed Abstract \| Publisher Full Text\n\nLiu X, Jang SS, An Z, et al.: Fenofibrate decreases radiation sensitivity via peroxisome proliferator-activated receptor α-mediated superoxide dismutase induction in HeLa cells. Radiat Oncol J. 2012; 30(2): 88–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGe Y, Liu J, Yang X, et al.: Fenofibrate enhances radiosensitivity of esophageal squamous cell carcinoma by suppressing hypoxia inducible factor-1α expression. Tumour Biol. 2014; 35(11): 10765–10771. PubMed Abstract \| Publisher Full Text\n\nLiu J, Ge YY, Zhu HC, et al.: Fenofibrate increases radiosensitivity in head and neck squamous cell carcinoma via inducing G2/M arrest and apoptosis. Asian Pac J Cancer Prev. 2014; 15(16): 6649–6655. PubMed Abstract \| Publisher Full Text\n\nCollet P, Domenjoud L, Devignes MD, et al.: The human semaphorin 6B gene is down regulated by PPARs. Genomics. 2004; 83(6): 1141–50. PubMed Abstract \| Publisher Full Text\n\nMurad H, Collet P, Huin-Schohn C, et al.: Effects of PPAR and RXR ligands in semaphorin 6B gene expression of human MCF-7 breast cancer cells. Int J Oncol. 2006; 28(4): 977–84. PubMed Abstract \| Publisher Full Text\n\nGe C, Li Q, Wang L, et al.: The role of axon guidance factor semaphorin 6B in the invasion and metastasis of gastric cancer. J Int Med Res. 2013; 41(2): 284–92. PubMed Abstract \| Publisher Full Text\n\nCorrea RG, Sasahara RM, Bengtson MH, et al.: Human semaphorin 6B [(HSA)SEMA6B], a novel human class 6 semaphorin gene: alternative splicing and all-trans-retinoic acid-dependent downregulation in glioblastoma cell lines. Genomics. 2001; 73(3): 343–8. PubMed Abstract \| Publisher Full Text\n\nMurakami H, Murakami R, Kambe F, et al.: Fenofibrate activates AMPK and increases eNOS phosphorylation in HUVEC. Biochem Biophys Res Commun. 2006; 341(4): 973–978. PubMed Abstract \| Publisher Full Text\n\nJiying W, Tao W, Jinrong Z, et al.: Effect of combined arsenic trioxide and fenofibrate on epithelial-interstitial transformation and E-cadherin/snail transformation factor in human pulmonary carcinoma A 549 cells. Journal of Third Military University. 2012–14. Reference Source\n\nTanaka T, Kohno H, Yoshitani S, et al.: Ligands for peroxisome proliferator-activated receptors alpha and gamma inhibit chemically induced colitis and formation of aberrant crypt foci in rats. Cancer Res. 2001; 61(6): 2424–2428. PubMed Abstract\n\nMurtola TJ, Tammela TL, Lahtela J, et al.: Cholesterol-lowering drugs and prostate cancer risk: a population-based case-control study. Cancer Epidemiol Biomarkers Prev. 2007; 16(11): 2226–32. PubMed Abstract \| Publisher Full Text\n\nRao A, D’Amico S, Balasubramanyam A, et al.: Fenofibrate is effective in treating hypertriglyceridemia associated with HIV lipodystrophy. Am J Med Sci. 2004; 327(6): 315–8. PubMed Abstract \| Publisher Full Text\n\nKoltai T: Nelfinavir and other protease inhibitors in cancer: mechanisms involved in anticancer activity. F1000Res. 2015; 4: 9. Publisher Full Text\n\nBonnet E, Ruidavets JB, Tuech J, et al.: Apoprotein c-III and E-containing lipoparticles are markedly increased in HIV-infected patients treated with protease inhibitors: association with the development of lipodystrophy. J Clin Endocrinol Metab. 2001; 86(1): 296–302. PubMed Abstract \| Publisher Full Text\n\nSamson SL, Pownall HJ, Scott LW, et al.: Heart positive: design of a randomized controlled clinical trial of intensive lifestyle intervention, niacin and fenofibrate for HIV lipodystrophy/dyslipidemia. Contemp Clin Trials. 2006; 27(6): 518–30. PubMed Abstract \| Publisher Full Text\n\nThomas JC, Lopes-Virella MF, Bene VE, et al.: Use of fenofibrate in the management of protease inhibitor-associated lipid abnormalities. Pharmacotherapy. 2000; 20(6): 727–734. PubMed Abstract \| Publisher Full Text\n\nCollett GP, Betts AM, Johnson MI, et al.: Peroxisome proliferator-activated receptor alpha is an androgen-responsive gene in human prostate and is highly expressed in prostatic adenocarcinoma. Clin Cancer Res. 2000; 6(8): 3241–8. PubMed Abstract\n\nFauconnet S, Lascombe I, Chabannes E, et al.: Differential regulation of vascular endothelial growth factor expression by peroxisome proliferator-activated receptors in bladder cancer cells. J Biol Chem. 2002; 277(26): 23534–43. PubMed Abstract \| Publisher Full Text\n\nSuchanek KM, May F, Robinson JA, et al.: Peroxisome proliferator-activated receptor alpha in the human breast cancer cell lines MCF-7 and MDA-MB-231. Mol Carcinog. 2002; 34(4): 165–171. PubMed Abstract \| Publisher Full Text\n\nJackson L, Wahli W, Michalik L, et al.: Potential role for peroxisome proliferator activated receptor (PPAR) in preventing colon cancer. Gut. 2003; 52(9): 1317–1322. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi MY, Yuan H, Ma LT, et al.: Roles of peroxisome proliferator-activated receptor-alpha and -gamma in the development of non-small cell lung cancer. Am J Respir Cell Mol Biol. 2010; 43(6): 674–83. PubMed Abstract \| Publisher Full Text\n\nPalmer CN, Hsu MH, Griffin KJ, et al.: Peroxisome proliferator activated receptor-alpha expression in human liver. Mol Pharmacol. 1998; 53(1): 14–22. PubMed Abstract\n\nRoberts RA, James NH, Hasmall SC, et al.: Apoptosis and proliferation in nongenotoxic carcinogenesis: species differences and role of PPARalpha. Toxicol Lett. 2000; 112–113: 49–57. PubMed Abstract \| Publisher Full Text\n\nVanden Heuvel JP, Kreder D, Belda B, et al.: Comprehensive analysis of gene expression in rat and human hepatoma cells exposed to the peroxisome proliferator WY14,643. Toxicol Appl Pharmacol. 2003; 188(3): 185–98. PubMed Abstract \| Publisher Full Text\n\nRao MS, Reddy JK: Peroxisome proliferation and hepatocarcinogenesis. Carcinogenesis. 1987; 8(5): 631–636. PubMed Abstract \| Publisher Full Text\n\nWilk A, Urbanska K, Grabacka M, et al.: Fenofibrate-induced nuclear translocation of FoxO3A triggers Bim-mediated apoptosis in glioblastoma cells in vitro. Cell Cycle. 2012; 11(14): 2660–2671. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilk A, Wyczechowska D, Zapata A, et al.: Molecular mechanisms of fenofibrate-induced metabolic catastrophe and glioblastoma cell death. Mol Cell Biol. 2005; 35(1): 182–198. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHan DF, Zhang J, Wei W, et al.: Fenofibrate induces G0/G1 phase arrest by modulating the PPARα/FoxO1/p27kip pathway in human glioblastoma cells. Tumor Biol. 2015. PubMed Abstract \| Publisher Full Text\n\nTsao SC, Tsai MH, Chiu CF, et al.: AMPK-dependent signaling modulates the suppression of invasion and migration by fenofibrate in CAL 27 oral cancer through NF-kB pathway. Environ Toxicol. 2014; PubMed Abstract \| Publisher Full Text\n\nGrau R, Punzón C, Fresno M, et al.: Peroxisome-proliferator-activated receptor alpha agonists inhibit cyclo-oxygenase 2 and vascular endothelial growth factor transcriptional activation in human colorectal carcinoma cells via inhibition of activator protein-1. Biochem J. 2006; 395(1): 81–88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVamecq J, Colet JM, Vanden Eynde JJ, et al.: PPARs: interference with Warburg´ effect and clinical anticancer trials. PPAR Res. 2012; 2012: 304760. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSekiya M, Yahagi N, Matsuzaka T, et al.: Polyunsaturated fatty acids ameliorate hepatic steatosis in obese mice by SREBP-1 suppression. Hepatology. 2003; 38(6): 1529–1539. PubMed Abstract \| Publisher Full Text\n\nKim HJ, Takahashi M, Ezaki O: Fish Oil Feeding Decreases Mature Sterol Regulatory Element-binding Protein 1 (SREBP-1) by Down-regulation of SREBP-1c mRNA in Mouse Liver A POSSIBLE MECHANISM FOR DOWN-REGULATION OF LIPOGENIC ENZYME mRNAs. J Biol Chem. 1999; 274(36): 25892–25898. PubMed Abstract \| Publisher Full Text\n\nGuo K, Wang PR, Milot DP, et al.: Regulation of lipid metabolism and gene expression by fenofibrate in hamsters. Biochim Biophys Acta. 2001; 1533(3): 220–232. PubMed Abstract \| Publisher Full Text\n\nKönig B, Koch A, Spielmann J, et al.: Activation of PPARalpha lowers synthesis and concentration of cholesterol by reduction of nuclear SREBP-2. Biochem Pharmacol. 2007; 73(4): 574–585. PubMed Abstract \| Publisher Full Text" }	[ { "id": "8393", "date": "11 May 2015", "name": "Krzysztof Reiss", "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 in an excellent comprehensive review of the literature related to Fenofibrate anticancer effects, which definitely deserves to be indexed after revision.I have three comments which should be implemented: It should be mentioned that fenofibrate is a prodrug, which is processed by blood esterases to fenofibric acid, and that fenofibric acid is potent agonist of PPARa. Therefore, there are two very different substances, fenofibrate and fenofibric acid, which could have very different anticancer properties. As far as I understand the system, mitochondrial and growth factor-mediated effects are rather associated with unprocessed fenofibrate, which in opposite to fenofibric acid is capably of interacting with biological membranes. Reference 89 in table 6 should be updated (see below); and the issue of Fenofibrate and mitochondrial respiration deserves a separate paragraph since it explains how unprocessed fenofibrate can target tumor cells, being at the same time much less harmful to normal cells. Other described mechanisms could be involved as a consequence of this primary mitochondrial action, which actually happens in minutes.Wilk A, Wyczechowska D, Zapata A, Dean M, Mullinax J, Marrero L, Parsons C, Peruzzi F, Culicchia F, Ochoa A, Grabacka M, Reiss K. Molecular mechanisms of fenofibrate-induced metabolic catastrophe and glioblastoma cell death. Mol Cell Biol. 2015 Jan;35(1):182-98. doi: 10.1128/MCB.00562-14. Epub 2014 Oct 20. PubMed PMID: 25332241; PubMed Central PMCID: PMC4295376. Reference 158 has a mistake it should be MCB 2015 instead of MCB 2005", "responses": [ { "c_id": "1355", "date": "13 May 2015", "name": "Tomas Koltai", "role": "Author Response", "response": "Dear Dr. Krzysztof ReissThank you for your kind contribution to this review.A new version is under way, which will include all the issues you remarked:A figure will be introduced which will show the molecular structure of FF and fibric acid. The difference between the pharmacologic actions of both will be discussed. Membrane rigidity due to FF will be explained under a new heading and the importance of this effect will be further discussed. The difference in cell distribution of FA and FF will be also considered. Reference in Table 6 will be updated and reference 158 will be corrected.Remains sincerely yoursTomas Koltai" } ] } ]	1	https://f1000research.com/articles/4-55
https://f1000research.com/articles/4-183/v1	06 Jul 15	{ "type": "Review", "title": "Recent scientific advances in leiomyoma (uterine fibroids) research facilitates better understanding and management", "authors": [ "Darlene K. Taylor", "Kristine Holthouser", "James H. Segars", "Phyllis C. Leppert", "Darlene K. Taylor", "Kristine Holthouser", "James H. Segars" ], "abstract": "Uterine leiomyomas (fibroids) are the most prevalent medical problem of the female reproductive tract, but there are few non-surgical treatment options. Although many advances in the understanding of the molecular components of these tumors have occurred over the past five years, an effective pharmaceutical approach remains elusive. Further, there is currently no clinical method to distinguish a benign uterine leiomyoma from a malignant leiomyosarcoma prior to treatment, a pressing need given concerns about the use of the power morcellator for minimally invasive surgery. This paper reviews current studies regarding the molecular biology of uterine fibroids, discusses non-surgical approaches and suggests new cutting-edge therapeutic and diagnostic approaches.", "keywords": [ "leiomyoma", "uterine fibroids", "uterine leiomyoma", "fibroids", "gynecological cancers" ], "content": "Introduction and context\n\nThe clinical management of uterine leiomyomas has advanced slowly and the current options remain limited. Advances in our understanding of the basic mechanisms of initiation and development over the past 5 years have elucidated the complexity of the molecular biology of leiomyomas. Although reviews of standard medical therapies have recently been published, this paper focuses on current findings in both basic and clinical research that have advanced the field and may open new strategies for treatment. Our goal is to open a dialog between clinicians and scientists to stimulate additional treatment options for women with uterine leiomyomas.\n\n\nBackground\n\nLeiomyomas, also called fibroids due to their abundant fibrotic tissue, have a 70–80% cumulative incidence in the childbearing years1. These benign tumors are known to clinicians as a worldwide public health problem. Estimates of treatment costs of leiomyomas in the US range from $5.9 billion to $34.4 billion annually, and includes the costs of medical and surgical treatment, amount of work time lost and complications attributable to the tumors2. These data suggest the estimated costs contribute more to health care expenditures than breast, colon or ovarian cancer2. The symptoms of uterine leiomyomas include bleeding with possible subsequent severe anemia, symptoms of pain and pressure leading to difficulty with bowel and bladder function and, in some cases, infertility and pregnancy complications. Notably, the tumors may arise anywhere in the uterine myometrium necessitating individualization of therapy. Large tumors may present with few if any symptoms while small fibroids may cause severe bleeding and pain. Hysterectomy is a common treatment that unfortunately negates the possibility of childbearing. Hysterectomy aside, subserosal or intramural fibroids can negatively impact fertility3. Currently, the decisive treatment is hysterectomy, either via abdominal/vaginal route or increasingly through laparoscopic incisions. Myomectomy, the surgical removal of only tumors, is a popular therapeutic option because it preserves childbearing, an important consideration for women of reproductive age. Uterine artery embolization, and MRI guided focused ultrasound and radiofrequency ablation are also suitable for some women.\n\nMany advances have occurred over the past 5 years and are reviewed here briefly as they have changed our understanding of the nature of leiomyoma. Without this critical basic understanding, advances in non-invasive therapies cannot be developed and optimal individualized therapies adopted.\n\n\nAdvances in basic studies\n\nAfrican-Americans develop benign leiomyomas at younger ages than Caucasians. The tumors appear to decrease in their growth rates before menopause in Caucasians but no decrease in growth is apparent in African-Americans4. Although this finding was published in 2008, it has not always been appreciated by active clinicians and researchers. Reported similarities between leiomyomas and keloids5,6 are consistent with recent findings7. Leiomyoma cells secrete high levels of disarrayed and altered collagen fibrils, fibronectin, and other extracellular matrix components and resist apoptosis6,8,9. Fibroids vary in uterine location and size up to 20 cm, or greater. One individual may have only one tumor while another might have multiple tumors. Growth is influenced by female gonadal steroids. However, the steroid-dependent growth is tumor-specific and not systemic as the same individual uterus may present multiple fibroids having differing growth patterns—some grow, some regress, and some are stable in the same time period4. Thus, leiomyomas exhibit complex mechanisms of development and growth.\n\nMechanotransduction, the response of cells to the mechanical forces such as compression and stretch, influences the biochemical pathways in all cells that affect growth at the cellular and tissue level10–12, including wound healing responses5, growth factors, reproductive hormones and cytokines13, and uterine stem cells14,15. Three recent reviews on the topic of mechanotransduction in reproduction expand in detail on this signaling mechanism10–12. Increasing evidence suggests a role for mechanotransduction in leiomyoma initiation and growth. Both biomechanical and biochemical factors, and not merely one paramount molecule, cause changes in uterine smooth muscle and leiomyoma cell behavior10. These changes occur through bi-directional signaling from individual cells to their matrix microenvironment and back to fibroid cells10. Catastrophic genetic alterations called chromothripsis, a sudden episode of chromosomal shattering and rearrangement, have been found in uterine fibroids16. However, not all leiomyomas display these genetic alterations; thus it is not clear whether the genetic defects are a primary cause or only associated with the development of some tumors. A recent study analyzed the genetic abnormalities in 256 fibroid tumors from 120 women17. In this study, 20 (7.8%) of the fibroids had a chromosomal rearrangement of 12q14-15 reflecting the rearrangement of the HMGA2 allele, while 179 (69.9%) of the fibroids exhibited a mutation of mediator complex subunit 12 (MED12), a transcription factor gene17. The remaining 22.3% of the tumors were reported as having either another genetic abnormality or no detectable abnormality17. Similar findings were recently found in a population of 135 women from the Southern United States with 64.33% of the fibroid tissues having MED12 mutations in exon 2 including deletion mutations18. Uterine smooth muscle cells respond to mechanotransduction in a different manner from cardiac muscle, which suggests that their innate qualities are unique19. One interesting aspect of leiomyomas is that they are surrounded by a relatively thick wall, a pseudocapsule, which encapsulates the tumors. Investigations of mutations in the MED12 gene have demonstrated that the pseudocapsule is derived from surrounding myometrium and not the tumor itself20. Understanding of pseudocapsule development may reveal new therapeutic targets.\n\nInterestingly, while fibroids are clonal tumors, each arising from a single cell, they are grossly and molecularly heterogeneous growths, consisting of the considerable extracellular matrix that provides the characteristic property of tumor stiffness noted on clinical palpation21. Leiomyomas are rare in animals and there is no universally-accepted spontaneous animal model. The Eker rat develops tumors that resemble fibroids, but the growths do not exhibit the abundant collagen characteristic of the human tissue22. While murine models have been reported, they have not been widely adopted.\n\nCurrently, research in the field relies on human tissues and cultured cells from surgical specimens, but the tumor or tumor-derived cells being studied might be in a state of active growth, or alternatively senescence at the time of acquisition. This fact is a significant consideration for the field. Because of this complexity, the identification of key molecular pathways in tumor development remains elusive and presents challenges to pharmaceutical development.\n\n\nRecent advances in clinical treatment\n\nClinical management decisions revolve around control of the heavy menstrual bleeding, including anemia which is often severe, chronic pain and pressure, or infertility. These symptoms are severe enough in approximately 25% of women with fibroids to require treatment13. Here we review pertinent advances and suggest areas of further avenues of inquiry. Several recent articles review in detail the treatment options currently available, including herbal medications23–26, and provide clinicians with comprehensive up-to-date information for treatment decision-making. Strategies for prevention or reduction in fibroid growth rate in high-risk women may be possible, as reviewed in Table 127–36. It is worth mentioning that, in addition, multiple in vivo and animal studies suggest that Vitamin D presents an attractive strategy to prevent uterine fibroid formation37–42, and hopefully clinical trials will show the efficacy of this approach.\n\nAbbreviations: CC, case control; COC, combined oral contraceptive pill; DMSO, dimethyl sulfoxide; DMPA, depot medroxyprogesterone acetate; EGCG, epigallocatechin gallate; LNG-IUS, levonorgestrel-releasing intrauterine system; OR, odds ratio; PBAC, Pictorial Pain, Bleeding Assessment Chart; PPARγ, peroxisome proliferator-activated receptor gamma; QOL, quality of life; RR, relative risk; TFV, Total fibroid volume; TVUS, transvaginal ultrasound; UF, uterine fibroid.\n\nCurrently, there is no simple, effective screening method to determine if a uterine tumor is indeed benign and not malignant, prior to treatment. It is known that adenomyosis can present clinically in a manner suggestive of fibroids. Recently, it was reported that experienced physicians using preoperative ultrasonograms interpreted myometrial hyperplasia on tissue histopathology as uterine fibroids43. This study suggests that preoperative ultrasound imaging using current standard technology may be responsible for over diagnosing uterine fibroids. However, the misdiagnosis of leiomyosarcoma is of greater concern. A strategy to determine if a tumor is a leiomyosarcoma is urgently needed. MRI techniques demonstrate the ability to differentiate malignant from benign tumors44 but have not yet been validated indistinguishing leiomyosarcoma from leiomyoma. While important, this approach is clearly not cost-effective. Using shear wave elastography, a leiomyosarcoma was accurately diagnosed pre-operatively, based on the degree of stiffness throughout the tumor45. If this modality were confirmed in larger studies, it would be a major breakthrough for the field. Specifically, power morcellation has been restricted as a modality, even though it reduces complications46–48, but a reliable pre-treatment tool to diagnose leiomyosarcoma would renew interest in that method.\n\nUlipristal acetate was developed as a selective progesterone receptor modulator with pure progesterone receptor antagonistic activity and minimal antiglucocorticoid effects. Ulipristal is currently marketed as Esmya and was approved by the FDA in 2010 for emergency contraception. It is approved in Europe and Canada for pre-surgical treatment of fibroids. One or three month courses of ulipristal acetate has been shown to induce apoptosis and to decrease proliferation of uterine fibroid cells, and to decrease fibroid size by a variable amount. No relevant affinity for estrogen, androgen or progesterone receptors (ER, AR or PR) has been observed. Several randomized trials demonstrated that ulipristal decreased the volume of leiomyomas significantly in comparison to controls49–51. Ulipristal has also been shown to induce amenorrhea49–51. Ulipristal does not induce changes in gonadotropin releasing hormone (GnRH) levels and does not reduce serum estradiol levels below the 50 pg/dl levels necessary to maintain bone mineral density52,53. For many women the advantage of this non-surgical treatment is the ability to preserve fertility. The first study of pregnancy after completed ulipristal treatment was recently published53. Of 21 women who had stopped ulipristal and attempted pregnancy, a pregnancy occurred in 15 women (71%) with 18 pregnancies during the study period with no regrowth of the leiomyomas. Twelve pregnancies produced healthy live infants and 6 resulted in spontaneous abortion (miscarriage)53. Other selective progesterone modulators, such as proellex, are currently being evaluated and Elagolix, an orally administered formulation of GnRH, is currently being studied in clinical trials.\n\n\nPossible new therapies on the horizon\n\nA purified bacterial collagenase from Clostridium histolyticum (CCH) has recently been shown in ex vivo leiomyoma tissue to significantly degrade the altered collagen when injected into tumor tissue. When the concentration of the CCH was increased and the injection volume kept small, the penetration of the CCH into the myometrium was limited and indicates that, on refinement of the dose, penetration into the myometrium could be eliminated. CCH is inhibited by serum proteins, a fact which also mitigates the concern for damage to the myometrium. Most importantly, our group in collaboration with Farshid Guilak and his colleagues at Duke University have shown that this collagenase (already FDA approved for use in the treatment of hand contractures due to collagen cord formation and for a disease of the male penis due to abnormal collagen formation), clearly reduced tissue stiffness in leiomyomas54. This reduced stiffness would not only reduce the bulk of the tumor, it is theoretically capable of altering mechanical signaling pathways in the leiomyoma, overcoming the resistance to apoptosis and allowing the cells to die. Clinical studies have demonstrated that the collagenase does not affect blood vessels or nerves. The use of CCH, alone or with other drugs such as a selective progesterone receptor modulator, could potentially be utilized as an injectable therapy for uterine leiomyomas 3–7 cm in size and could be most useful in treating submucosal leiomyomas, the type most associated with infertility55.\n\nThe development of materials designed to deliver and protect drug therapeutics by direct injection to the tumor site is an area of active research. Several such drug delivery materials that change phase in response to temperature changes are currently in development as they offer many advantages over conventional drug delivery systems23. These thermoresponsive materials form a solution in aqueous media that reversibly transitions to a gel at physiological temperatures. The system often degrades in a defined period of time, thereby eliminating the need for surgical explantation. In its solution state, the delivery system readily mixes with therapeutic agents to afford a drug formulation that can be administered by a single injection. The injected formulation is a stable solution that transforms into a gel depot at the site of injection as a result of an elevation in temperature.\n\nThe marriage of injectable thermoresponsive delivery systems with the unmet need for viable non-surgical options for the management of uterine fibroids offers several advantages. Creating a drug depot inside the fibroid by local injection would impede diffusion and distribution of the drug away from the injected fibroid, prolong release, delay inactivation, and therefore reduce the need for repeat injections. This treatment approach for women wanting to maintain fertility yet seeking relief from fibroid symptoms could be administered by skilled individuals under ultrasound guidance in a doctor’s office. A few examples of the most promising of these thermoresponsive delivery systems are given below (Table 2).\n\nNotes: 1poly(DL-lactide); 2poly(DL-lactide-do-glycolide); 3poly(DL-lactide-co-caprolactone); 4polyanydrides; 5hydroxyethyl methacrylate-polylactide; 6Hyperbranched polyglycerol. Modified from reference 23\n\nOne material developed by our group and listed in Table 2 is particularly worth noting. LiquoGel™ delivers drugs similar to other thermoresponsive delivery systems but distinguishes itself from other materials in that it contains multiple functional groups that enable chemical modifications to covalently link therapeutics. Thus, multiple drugs can be delivered at one time. With the advent of the means to deliver drugs or drug combinations directly to leiomyoma tumors, the potential of reduction and perhaps eradication of tumors prior to the need for surgical or other major interventions (such as focused ultrasounds or uterine artery embolization) could be realized. Multiple drugs could be given as combination chemotherapy, such as an anti-fibrotic agent combined with a selective progesterone receptor modulator, or sequentially, for the benefit of patient care. It could be possible to deliver gene therapy in this manner as well55–58. A number of the more conventional drug therapies for uterine fibroids could be potentially entrapped or covalently linked to LiquoGel™ to afford delivery with potentially reduced side effects, improved efficacy, and controlled release profiles23.\n\n\nImplications for clinical practice\n\nEven though treatments for fibroids can be developed currently without a complete elucidation of their etiology and molecular biology, ultimately, if the molecular mechanisms for fibroid development and of myometrial proliferation are understood, additional nonsurgical therapeutic interventions may be forthcoming. Taken together, we have evidence that uterine leiomyomas grow due to cell proliferation, but even more because of excessive deposition of altered extracellular matrix due to the persistence of secreting cells. There is a growing appreciation of the complex pathways leading to the formation of uterine leiomyomas which will lead to new therapeutic approaches. Could drug therapy, either a single drug or most likely combination chemotherapy, rival the effectiveness of surgical procedures yet preserve the uterine childbearing function? If realized, could such a therapy be administered during a routine visit to the doctor or clinic? Addressing these questions presents unique opportunities at the interface of molecular medicine and clinical care.\n\nThe optimal treatment remains one that reduces the bulk of the leiomyoma and reduces blood loss while preserving the ability to have children. Clinician, doctors, patients, and researchers should continue to work together to develop cost-effective and efficacious solutions to leiomyoma disease that are compatible with the woman’s life-style, reducing or eliminating hospital stay and lengthy recovery time23.", "appendix": "Competing interests\n\n\n\nDKT developed LiquoGel™ with funding from NIH (grant number: NIH K12HD043446-04) and had funding from BioSpecifics Technologies Inc. to optimize the compound; North Carolina Central University has filed a patent for the product, but DKT has no pending obligations to BioSpecifics Technologies or any other company.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBaird DD, Dunson DB, Hill MC, et al.: High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstet Gynecol. 2003; 188(1): 100–7. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nCardozo ER, Clark AD, Banks NK, et al.: The estimated annual cost of uterine leiomyomata in the United States. Am J Obstet Gynecol. 2012; 206(3): 211.e1–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCook H, Ezzati M, Segars JH, et al.: The impact of uterine leiomyomas on reproductive outcomes. Minerva Ginecol. 2010; 62(3): 225–36. PubMed Abstract \| Free Full Text\n\nPeddada SD, Laughlin SK, Miner K, et al.: Growth of uterine leiomyomata among premenopausal black and white women. Proc Natl Acad Sci U S A. 2008; 105(50): 19887–92. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nLeppert PC, Catherino WH, Segars JH: A new hypothesis about the origin of uterine fibroids based on gene expression profiling with microarrays. Am J Obstet Gynecol. 2006; 195(2): 415–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCatherino WH, Leppert PC, Stenmark MH, et al.: Reduced dermatopontin expression is a molecular link between uterine leiomyomas and keloids. Genes Chromosomes Cancer. 2004; 40(3): 204–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarrino DA, Mesiano S, Barker NM, et al.: Proteoglycans of uterine fibroids and keloid scars: similarity in their proteoglycan composition. Biochem J. 2012; 443(2): 361–8. PubMed Abstract \| Publisher Full Text\n\nLeppert PC, Baginski T, Prupas C, et al.: Comparative ultrastructure of collagen fibrils in uterine leiomyomas and normal myometrium. Fertil Steril. 2004; 82(Suppl 3): 1182–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFujisawa C, Castellot JJ Jr: Matrix production and remodeling as therapeutic targets for uterine leiomyoma. J Cell Commun Signal. 2014; 8(3): 179–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThorne JT, Segal TR, Chang S, et al.: Dynamic reciprocity between cells and their microenvironment in reproduction. Biol Reprod. 2015; 92(1): 25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeppert PC, Jayes FL, Segars JH: The extracellular matrix contributes to mechanotransduction in uterine fibroids. Obstet Gynecol Int. 2014; 2014: 783289. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJorge S, Chang S, Barzilai JJ, et al.: Mechanical signaling in reproductive tissues: mechanisms and importance. Reprod Sci. 2014; 21(9): 1093–107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIslam MS, Protic O, Stortoni P, et al.: Complex networks of multiple factors in the pathogenesis of uterine leiomyoma. Fertil Steril. 2013; 100(1): 178–93. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nOno M, Bulun SE, Maruyama T: Tissue-specific stem cells in the myometrium and tumor-initiating cells in leiomyoma. Biol Reprod. 2014; 91(6): 149. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoravek MB, Yin P, Ono M, et al.: Ovarian steroids, stem cells and uterine leiomyoma: therapeutic implications. Hum Reprod Update. 2015; 21(1): 1–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMehine M, Kaasinen E, Mäkinen N, et al.: Characterization of uterine leiomyomas by whole-genome sequencing. N Engl J Med. 2013; 369(1): 43–53. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nMarkowski DN, Helmke BM, Bartnitzke S, et al.: Uterine fibroids: do we deal with more than one disease? Int J Gynecol Pathol. 2014; 33(6): 568–72. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nHalder SK, Laknaur A, Miller J, et al.: Novel MED12 gene somatic mutations in women from the Southern United States with symptomatic uterine fibroids. Mol Genet Genomics. 2015; 290(2): 505–11. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nNorian JM, Owen CM, Taboas J, et al.: Characterization of tissue biomechanics and mechanical signaling in uterine leiomyoma. Matrix Biol. 2012; 31(1): 57–65. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nDi Tommaso S, Massari S, Malvasi A, et al.: Selective genetic analysis of myoma pseudocapsule and potential biological impact on uterine fibroid medical therapy. Expert Opin Ther Targets. 2015; 19(1): 7–12. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nJayes FL, Ma X, Flannery EM, et al.: Biomechanical evaluation of human uterine fibroids after exposure to purified clostridial collagenase. Society for the Study of Reproduction. Montreal July, 2013. Supplement to Biol Reprod. A743: 334.\n\nEveritt JI, Wolf DC, Howe SR, et al.: Rodent model of reproductive tract leiomyomata. Clinical and pathological features. Am J Pathol. 1995; 146(6): 1556–67. PubMed Abstract \| Free Full Text \| F1000 Recommendation\n\nTaylor DK, Leppert PC: Treatment for Uterine Fibroids: Searching for Effective Drug Therapies. Drug Discov Today Ther Strateg. 2012; 9(1): e41–e49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh SS, Belland L: Contemporary management of uterine fibroids: focus on emerging medical treatments. Curr Med Res Opin. 2015; 31(1): 1–12. PubMed Abstract \| Publisher Full Text\n\nPatel A, Malik M, Britten J, et al.: Alternative therapies in management of leiomyomas. Fertil Steril. 2014; 102(3): 649–55. PubMed Abstract \| Publisher Full Text\n\nGuo XC, Segars JH: The impact and management of fibroids for fertility: an evidence-based approach. Obstet Gynecol Clin North Am. 2012; 39(4): 521–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang D, Al-Hendy M, Richard-Davis G, et al.: Green tea extract inhibits proliferation of uterine leiomyoma cells in vitro and in nude mice. Am J Obstet Gynecol. 2010; 202(3): 289.e1–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nRoshdy E, Rajaratnam V, Maitra S, et al.: Treatment of symptomatic uterine fibroids with green tea extract: a pilot randomized controlled clinical study. Int J Womens Health. 2013; 5: 477–86. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nMalik M, Mendoza M, Payson M, et al.: Curcumin, a nutritional supplement with antineoplastic activity, enhances leiomyoma cell apoptosis and decreases fibronectin expression. Fertil Steril. 2009; 91(5 Suppl): 2177–84. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nTsuiji K, Takeda T, Li B, et al.: Inhibitory effect of curcumin on uterine leiomyoma cell proliferation. Gynecol Endocrinol. 2011; 27(7): 512–7. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nLumbiganon P, Rugpao S, Phandhu-fung S, et al.: Protective effect of depot-medroxyprogesterone acetate on surgically treated uterine leiomyomas: a multicentre case--control study. Br J Obstet Gynaecol. 1996; 103(9): 909–14. PubMed Abstract\n\nVenkatachalam S, Bagratee JS, Moodley J: Medical management of uterine fibroids with medroxyprogesterone acetate (Depo Provera): a pilot study. J Obstet Gynaecol. 2004; 24(7): 798–800. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSayed GH, Zakherah MS, El-Nashar SA, et al.: A randomized clinical trial of a levonorgestrel-releasing intrauterine system and a low-dose combined oral contraceptive for fibroid-related menorrhagia. Int J Gynaecol Obstet. 2011; 112(2): 126–30. PubMed Abstract \| Publisher Full Text\n\nQin J, Yang T, Kong F, et al.: Oral contraceptive use and uterine leiomyoma risk: a meta-analysis based on cohort and case–control studies. Arch Gynecol Obstet. 2013; 288(1): 139–48. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSabry M, Halder SK, Allah AS, et al.: Serum vitamin D3 level inversely correlates with uterine fibroid volume in different ethnic groups: a cross-sectional observational study. Int J Womens Health. 2013; 5: 93–100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHalder SK, Sharan C, Al-Hendy O, et al.: Paricalcitol, a vitamin D receptor activator, inhibits tumor formation in a murine model of uterine fibroids. Reprod Sci. 2014; 21(9): 1108–19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-Hendy A, Badr M: Can vitamin D reduce the risk of uterine fibroids? Womens Health (Lond Engl). 2014; 10(4): 353–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-Hendy A, Diamond MP, El-Sohemy A, et al.: 1,25-dihydroxyvitamin d3 regulates expression of sex steroid receptors in human uterine fibroid cells. J Clin Endocrinol Metab. 2015; 100(4): E572–82. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nHalder SK, Osteen KG, Al-Hendy A: 1,25-dihydroxyvitamin d3 reduces extracellular matrix-associated protein expression in human uterine fibroid cells. Biol Reprod. 2013; 89(6): 150. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nHalder SK, Sharan C, Al-Hendy A: 1,25-dihydroxyvitamin D3 treatment shrinks uterine leiomyoma tumors in the Eker rat model. Biol Reprod. 2012; 86(4): 116. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nHalder SK, Goodwin JS, Al-Hendy A: 1,25-Dihydroxyvitamin D3 reduces TGF-β3-induced fibrosis-related gene expression in human uterine leiomyoma cells. J Clin Endocrinol Metab. 2011; 96(4): E754–62. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nBaird DD, Hill MC, Schectman JM, et al.: Vitamin D and the risk of uterine fibroids. Epidemiology. 2013; 24(3): 447–53. PubMed Abstract \| Publisher Full Text\n\nNewcomb PM, Cramer SF, Leppert PC: Myometrial hyperplasia mimics the clinical presentation of uterine fibroids: a report of 3 cases. Int J Gynecol Pathol. 2013; 32(6): 585–91. PubMed Abstract \| Publisher Full Text\n\nThomassin-Naggara I, Dechoux S, Bonneau C, et al.: How to differentiate benign from malignant myometrial tumours using MR imaging. Eur Radiol. 2013; 23(8): 2306–14. PubMed Abstract \| Publisher Full Text\n\nFurukawa S, Soeda S, Watanabe T, et al.: The measurement of stiffness of uterine smooth muscle tumor by elastography. Springerplus. 2014; 3: 294. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nLieng M, Berner E, Busund B: Risk of morcellation of uterine leiomyosarcomas in laparoscopic supracervical hysterectomy and laparoscopic myomectomy, a retrospective trial including 4791 women. J Minim Invasive Gynecol. 2015; 22(3): 410–4. PubMed Abstract \| Publisher Full Text\n\nStine JE, Clarke-Pearson DL, Gehrig PA: Uterine morcellation at the time of hysterectomy: techniques, risks, and recommendations. Obstet Gynecol Surv. 2014; 69(7): 415–25. PubMed Abstract \| Publisher Full Text\n\nAmerican College of Obstetricians and Gynecologists ACOG. Power morcellation and occult malignancy in gynecologic surgery. A Special Report, Task Force and Work Group Reports. Washington, DC, May 2014. Reference Source\n\nRodriguez MI, Warden M, Darney PD: Intrauterine progestins, progesterone antagonists, and receptor modulators: a review of gynecologic applications. Am J Obstet Gynecol. 2010; 202(5): 420–8. PubMed Abstract \| Publisher Full Text\n\nDonnez J, Tatarchuk TF, Bouchard P, et al.: Ulipristal acetate versus placebo for fibroid treatment before surgery. N Engl J Med. 2012; 366(5): 409–20. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nNieman LK, Blocker W, Nansel T, et al.: Efficacy and tolerability of CDB-2914 treatment for symptomatic uterine fibroids: a randomized, double-blind, placebo-controlled, phase IIb study. Fertil Steril. 2011; 95(2): 767-72.e1–2. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nChabbert-Buffet N, Esber N, Bouchard P: Fibroid growth and medical options for treatment. Fertil Steril. 2014; 102(3): 630–9. PubMed Abstract \| Publisher Full Text\n\nChabbert-Buffet N, Pintiaux-Kairis A, Bouchard P: Effects of the progesterone receptor modulator VA2914 in a continuous low dose on the hypothalamic-pituitary-ovarian axis and endometrium in normal women: a prospective, randomized, placebo-controlled trial. J Clin Endocrinol Metab. 2007; 92(9): 3582–9. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nLuyckx M, Squifflet J, Jadoul P, et al.: First series of 18 pregnancies after ulipristal acetate treatment for uterine fibroids. Fertil Steril. 2014; 102(5): 1404–9. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nBrunengraber LN, Jayes FL, Leppert PC: Injectable Clostridium histolyticum collagenase as a potential treatment for uterine fibroids. Reprod Sci. 2014; 21(12): 1452–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNair S, Curiel DT, Rajaratnam V, et al.: Targeting adenoviral vectors for enhanced gene therapy of uterine leiomyomas. Hum Reprod. 2013; 28(9): 2398–406. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHassan MH, Salama SA, Zhang D, et al.: Gene therapy targeting leiomyoma: adenovirus-mediated delivery of dominant-negative estrogen receptor gene shrinks uterine tumors in Eker rat model. Fertil Steril. 2010; 93(1): 239–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-Hendy A, Salama S: Gene therapy and uterine leiomyoma: a review. Hum Reprod Update. 2006; 12(4): 385–400. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9353", "date": "06 Jul 2015", "name": "Donna Day Baird", "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", "responses": [] }, { "id": "9354", "date": "06 Jul 2015", "name": "Ayman Al-Hendy", "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", "responses": [] }, { "id": "9355", "date": "06 Jul 2015", "name": "Alicia Armstrong", "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", "responses": [] }, { "id": "9356", "date": "06 Jul 2015", "name": "Andrea Tinelli", "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", "responses": [] }, { "id": "9486", "date": "14 Jul 2015", "name": "Pasquapina Ciarmela", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-183
https://f1000research.com/articles/4-181/v1	06 Jul 15	{ "type": "Software Tool Article", "title": "An ORCID based synchronization framework for a national CRIS ecosystem", "authors": [ "João Mendes Moreira", "Alcino Cunha", "Nuno Macedo", "Alcino Cunha", "Nuno Macedo" ], "abstract": "PTCRIS (Portuguese Current Research Information System) is a program aiming at the creation and sustained development of a national integrated information ecosystem, to support research management according to the best international standards and practices.\nThis paper reports on the experience of designing and prototyping a synchronization framework for PTCRIS based on ORCID (Open Researcher and Contributor ID). This framework embraces the \"input once, re-use often\" principle, and will enable a substantial reduction of the research output management burden by allowing automatic information exchange between the various national systems.\nThe design of the framework followed best practices in rigorous software engineering, namely well-established principles in the research field of consistency management, and relied on formal analysis techniques and tools for its validation and verification.\nThe notion of consistency between the services was formally specified and discussed with the stakeholders before the technical aspects on how to preserve said consistency were explored. Formal specification languages and automated verification tools were used to analyze the specifications and generate usage scenarios, useful for validation with the stakeholder and essential to certificate compliant services.", "keywords": [ "CRIS", "ORCID-CASRAI", "PT CRIS", "Synchronisation Framework" ], "content": "1 Introduction\n\nPTCRIS (Portuguese Current Research Information System) is a program, officially initiated in May 2014 by FCCN (Fundação para a Computação Científica Nacional), the FCT (Fundação para a Ciência e Tecnologia – the Portuguese Foundation for Science and Technology) unit responsible for planning, management and operation of the national research and education network, a high performance platform for developing and testing advanced communication applications and services. PTCRIS aims to ensure the creation and sustained development of a national integrated information ecosystem, to support research management according to the best international standards and practices.\n\nOne of the goals of PTCRIS is to reduce the burden of research output management, by adopting an “input once, re-use often” principle. In order to achieve this goal, a synchronization framework is being developed that relies on ORCID (http://www.orcid.org/) – a community-based service that aims to provide a registry of unique researcher identifiers and a method of linking research outputs to these identifiers, based on data collected from external sources – as a central hub for information exchange between the various national systems (including CV management systems, open-access repositories, and local CRIS systems) and international systems (WoK, Scopus, Datacite, etc.). Among other features, this framework will enable researchers (or managers) to register a given research output once at one of the interconnected national systems, and have that output automatically propagated to the other ones, thus ensuring global consistency of the stored information. The goal of this paper is to report precisely on the experience of designing and prototyping this synchronization framework.\n\nThe design of the synchronization framework followed well-established principles of rigorous software engineering. The main principle is that one should distinguish the what from the how: in this particular case, what is the desired notion of consistency between ORCID and each of the PTCRIS services, and how can a synchronization procedure be implemented to enforce such consistency. This allowed us to break down the discussion with the various stakeholders, first seeking an agreement concerning the what before dwelling in the technicalities of the how.\n\nThe second principle is that formal analysis methods and tools should be used to verify that the proposed artifacts follow desirable “well-behavedness” properties. Paraphrasing Richard Feynman, “the first principle is that you must not fool yourself, and you are the easiest person to fool”: the usage of a formal specification language and automatic verification tools allowed us to uncover several corner cases not easy to predict otherwise. In particular, we relied on the formal specification language Alloy1 and its automatic Analyzer. This tool was also used to automatically generate usage scenarios that were useful for requirement elicitation and validation with the stakeholders, but will also be of major importance for certifying compliant PTCRIS services, by allowing rigorous testing of the proposed implementations.\n\nThe paper is organized as follows: Section 2 presents a brief overview of (and rationale for) the proposed architecture; Section 3 presents the methodology followed to achieve a trustworthy design for the various components of the framework; Section 4 describes the first prototype that was developed to validate and demo the proposed framework to the community; Section 5 briefly describes some related work; and, finally, Section 6 presents some conclusions and ideas for future work.\n\n\n2 Architecture overview and rationale\n\nFigure 1 presents an overview of the architecture of the PTCRIS synchronization framework, with some PTCRIS services shown in orange and ORCID sources in blue. PTCRIS is composed of several services with distinct objectives. Among those we have, for example:\n\nDeGóis The national academic CV management system (http://www.degois.pt), where information from researchers is stored and, with the proposed synchronization framework, shared across the PTCRIS ecosystem. DeGóis currently hosts around 22,000 academic CVs.\n\nRCAAP The national open-access scientific repository portal (http://www.rcaap.pt), a platform that acts as an OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) aggregator that harvests content from a network of institutional repositories (currently, around 70 in total) and open-access journals. RCAAP currently indexes around 213,000 publications.\n\nSARI A DSpace hosting platform for institutional open-access repository services (it currently hosts 26 repositories).\n\nAs depicted in Figure 1, not all of these services are expected to synchronize bidirectionally with ORCID. For example, RCAAP will only export research outputs to ORCID, so that they can be harvested by other PTCRIS services. In contrast, institutional repositories (namely those hosted in SARI) will just use ORCID to harvest publications, thus liberating researchers from (the often mandatory task of) having to (manually) insert them. The academic CV management service DeGóis will both import and export research outputs. As the figure also depicts, at least in the earlier stages of deployment of the synchronization framework, some services will still synchronize directly with each other, for example the RCAAP aggregation of open-access publications from institutional repositories will still be performed directly.\n\nThere were several reasons that contributed to the choice of ORCID as the central hub for PTCRIS. The main ones are described below:\n\nHigh coverage of predefined requirements. The requirements were grouped into three categories: general, functional and technical, and ORCID scored well in all of them. On the general level, items like documentation and support were considered. The main functional requirements were related with both the set of APIs (get, put, etc) and the completeness of the ORCID profile (it also supports most of the CASRAI academic funding CV elements but “services”). Usability issues were also analyzed, as this is a critical issue for PTCRIS. It was concluded that the interface for the integration of systems with ORCID is not only easy to use, but also becoming a standard.\n\nThe technical requirements were related with the easiness of implementation and infrastructure reliability and resilience (ORCID infrastructure is hosted in a world-class datacenter).\n\nHigh interoperability with external sources. By the time ORCID was being considered to act as the hub for PTCRIS it was already interoperable with some of the most relevant and important sources (Crossref, Datacite, Scopus and WoK). Furthermore, its interoperability tends to increase as more sources are being added.\n\nHigh ORCID coverage of the national research community. In late 2013 – early 2014, due to the research assessment exercise and for the purpose of carrying out a bibliometric study, around 15,000 researchers from Portuguese research units applied for an ORCID iD. These researchers were responsible for more than 90% of the Portuguese scientific output of the 5 years prior to 2013.\n\nSustainability. The costs of using ORCID as a hub are very small when compared with the alternative of developing and maintaining a homegrown hub.\n\nBesides the benefits, risks and mitigation measures were also considered when deciding whether to use ORCID as a hub for PTCRIS. The most relevant risk identified was the collapse of the ORCID organization, but the probability of this event was considered to be low. Nevertheless, two mitigation measures were considered: install the hub locally using the ORCID source code (deposited in GitHub); populate the PTCRIS database with the mensal database copy provided to ORCID premium members.\n\n\n3 Specification of the synchronization framework\n\nThis section presents the specification methodology that was used to achieve a trustworthy design for the PTCRIS synchronization framework. First, a formal specification of the data models and of the desired consistency predicates was developed. Then, synchronization procedures to enforce such consistency were specified and verified for several “well-behavedness” properties. These formal specifications were also used to automatically derive the already mentioned usage scenarios.\n\nThe synchronization framework operates at the user profile level, that is it intends to synchronize user profiles from the different PTCRIS services with the corresponding user profile from ORCID. The matching of users across these systems is a simple matter, since PTCRIS services can simply store (and most already do) the ORCID iD of the researcher locally. As such, the design of the framework focused only on a single user profile.\n\nThe main difficulties in the design of this framework stemmed from fundamental differences between the data model of ORCID and that of most PTCRIS services. As such, before presenting the desired consistency notion and synchronization procedures, we briefly present such data models. Here, we present only a very abstract view of the information stored in such profiles, focusing only on research outputs, namely works, and only on the attributes that are relevant for their synchronization. An ORCID user profile contains additional information, which PTCRIS is also interested in synchronizing among its services. However, some of this information is trivial to synchronize (e.g. education affiliations) while other, albeit not trivial, may be synchronized following the technique presented in this paper for works (e.g. funding information).\n\nFigure 2 presents an abstract model of an ORCID user profile. For our purposes, a profile consists essentially of a set of works, each a record containing: a putcode, that uniquely identifies the work internally; a (possibly empty) set of external unique identifiers (UIDs) of the work; the source of the information in the record (which can be the user himself or any other external source associated with ORCID, such as Scopus, CrossRef, or, from now on, a PTCRIS service); any meta-data associated with the work, such as its title, publication year, publication type, authors, etc; and a boolean attribute marking whether the work is the one preferred by the user among similar ones (this boolean attribute is not directly returned by the ORCID API, but can be inferred from the order in which the works are stored in an ORCID profile, see discussion below).\n\nA distinctive feature of ORCID is precisely the possibility of using different external sources to automatically populate a user profile. This means that a user profile can contain different works that actually describe the same research output (possibly containing different or even contradictory meta-data). The ORCID web interface already groups together works that describe the same output, showing only the preferred one in the overview. The grouping mechanism is quite simple, and just assumes two works w1 and w2 are similar if, and only if, they have a shared UID or there is another work w3 that is similar to both w1 and w2. Essentially, this recursive definition considers two works to be similar if, and only if, they share directly or indirectly (via transitivity) some UID.\n\nORCID imposes several constraints on this data model, such as: there cannot be two works with the same external source with shared UIDs; and among sets of similar works exactly one of them is the preferred one. The ORCID API also forces every work from an external source to have some UIDs assigned, but works added by the user via the web interface may still have an empty set of UIDs. The biggest difference of a user profile in a PTCRIS service (depicted in Figure 3) is that it does not support multiple versions of the same research output, nor the grouping feature of similar versions likewise to ORCID. To avoid confusion with ORCID works we will denote research outputs in PTCRIS as productions. The profile of a user in a PTCRIS service is essentially a set of productions, each a record with the following information: a key that uniquely identifies the production; a (possibly empty) set of UIDs; the associated meta-data; and a boolean field indicating whether the production is currently selected be the user to be exported to ORCID.\n\nThe PTCRIS synchronization framework is semi-automatic and notification-based. As such, each service will be required to support two kinds of notifications in a user profile: creation notifications, to alert the user that a new production has been found in ORCID; and modification notifications, to alert the user that new UIDs for an existing production have been found. The latter will be particularly useful for propagating UIDs between different PTCRIS services, in particular from open access repositories that provide handles for research outputs to academic CV management services, such as DeGóis.\n\nLikewise to ORCID, this data model is subject to several constraints, such as disallowing exported productions to share UIDs or have no UIDs at all (to comply with the above ORCID guideline).\n\nAs stated above, we first specified and validated what is the desired notion of consistency between ORCID and each of the PTCRIS services. Formally, this consistency is a predicate of type ORCID × PTCRIS → Bool, that given two user profiles returns a boolean indicating whether they are consistent with each other. Typically, this consistency predicate is specified as a set of logical rules that must all be satisfied to render the profiles consistent.\n\nThe consistency between ORCID and a PTCRIS service was factorized in two modular consistency predicates whose rules were precisely defined in the design phase:\n\nIMPORTED : ORCID × PTCRIS → Bool This consistency predicate should be enforced by every PTCRIS service that wishes to rely on the synchronization framework to harvest research outputs from ORCID, namely new publications and new UIDs of known publications. The general principle of IMPORTED is that every UID in ORCID should be harvested. The enforcement of this consistency predicate should be semi-automatic, based on a notification system, giving freedom to the user to select which outputs or UIDs he wishes to add to his PTCRIS profile.\n\nEXPORTED : ORCID × PTCRIS → Bool This consistency predicate should be enforced by every PTCRIS service that wishes to be an ORCID source, and export its productions to ORCID, ensuring that other PTCRIS services can harvest them. The general principle of EXPORTED is that every exported production should be stored as a work in ORCID and then automatically kept up-to-date.\n\nThese consistency predicates are logically independent, in the sense that each can either hold or not, independently of the value of the other. A PTCRIS service may also wish to implement the conjunction of both, leading to a consistency predicate we denote as (fully) SYNCED:\n\nSYNCED : ORCID × PTCRIS → Bool\n\nSYNCED(o, p) =˙ IMPORTED(o, p) ∧ EXPORTED(o, p)\n\nSince the PTCRIS services do not support grouping likewise to ORCID, some caution must be exercised to avoid the proliferation of productions and notifications that describe the same research output. In particular, when an ORCID work is unknown to the PTCRIS service, the existence of a single creation notification, grouping all UIDs of its similar works should suffice to ensure consistency. This is just one of the rules that must be satisfied for IMPORTED to hold. IMPORTED is mainly focused on UID harvesting, the consistency of the meta-data being a secondary concern. However, meta-data still needs to be filled in when a creation is notified following the discovery of a group of (unknown) similar works. Since their meta-data can (and often does) differ, it is not clear how this meta-data extraction should be performed. On first glance, the obvious choice would be to pick the meta-data of the preferred work. Unfortunately, the following reasons prevent us from currently enforcing this behavior:\n\nSince all groups of similar works must have a preferred work (essentially the one chosen to be displayed in the user web page), a default preferred is always chosen by ORCID when a new research output is imported or the current preferred one is deleted by the user.\n\nThe ORCID API does not currently distinguish such default preferred works from user-selected ones.\n\nThis means that the user might not have sanctioned the meta-data present in his preferred works at the time they are being imported into a PTCRIS service.\n\nUnfortunately, meta-data is of highly variable quality in ORCID, with some sources currently publishing meta-data with gross mistakes, for example, wrong publication types.\n\nAs such, the choice of how to fill in the meta-data in creations was currently left for each PTCRIS service. Some will ignore the preferred and just allow the user to rank sources according to the perceived quality of their metadata, and then try to choose the meta-data of the work from the highest ranked source.\n\nThe EXPORTED consistency predicate is considerably simpler than IMPORTED. Essentially, the specified consistency rules force that there must exist a one-to-one correspondence between exported productions and works in ORCID whose source is the PTCRIS service.\n\nWhen the user profiles at ORCID and at the PTCRIS service are inconsistent how can they be automatically synchronized to recover the consistency? To achieve that, we have specified two separate synchronization procedures to be used when the service intends to enforce consistency according to IMPORTED or EXPORTED, respectively. These modular synchronization procedures can also be combined in a precise way, to recover the consistency in services that are enforcing both consistency predicates.\n\nIMPORT : ORCID × PTCRIS → PTCRIS This synchronization procedure should be used to enforce the IMPORTED consistency predicate. The main principle is that it does not change the user profile in ORCID. Moreover the only changes it produces to the PTCRIS profile is to add and remove notifications.\n\nEXPORT : ORCID × PTCRIS → ORCID This synchronization procedure should be used to enforce the EXPORTED consistency predicate. The main principle is that it does not change the user profile in PTCRIS. Moreover the only changes it produces to the ORCID profile is to add / delete / modify works whose source is the PTCRIS service.\n\nWith the help of automatic formal verification tools, the specified synchronization procedures were checked for several “well-behavedness” properties. The most important of those is correctness, that ensures that after running the synchronization procedures the user profiles in ORCID and in the PTCRIS service are indeed consistent:\n\nIMPORTED(o, IMPORT(o, p))\n\nEXPORTED(EXPORT(o, p), p)\n\nAnother important “well-behavedness” property is stability, ensuring that if we run the synchronization procedures on already consistent states the result is the same (modulo differences in keys):\n\nIMPORTED(o, p) ⇒ IMPORT(o, p) = p\n\nEXPORTED(o, p) ⇒ EXPORT(o, p) = o\n\nHaving stable synchronization procedures ensures that there is no need to explicitly check the consistency to determine if they should be run. If both user profiles are consistent, running the specified procedures would not affect them. In fact, the checking procedures have the same approximate complexity as the synchronizing procedures, and thus, no significant performance gains would be achieved by running them beforehand.\n\nThe two specified synchronization procedures can be combined to obtain a synchronization procedure that enforces SYNCED, the full consistency of the user profiles according to both IMPORTED and EXPORTED (to be used by services that wish to enforce both):\n\nSYNC : ORCID × PTCRIS → ORCID × PTCRIS\n\nSYNC(o, p) =˙ let o′ = EXPORT(o, p)\n\nin (o′, IMPORT(o′, p))\n\nThe specified order of execution is not arbitrary. In fact, it is the only order that ensures that the resulting procedure is both correct and stable:\n\nSYNCED(SYNC(o, p))\n\nSYNCED(o, p) ⇒ SYNC(o, p) = (o, p)\n\nIn particular, if the user profiles in ORCID and PTCRIS are not consistent according to EXPORTED, running the EXPORT procedure can make them inconsistent according to IMPORTED. As such, IMPORT must be run after EXPORT to ensure that full consistency is attained. A concrete example is presented in the following section.\n\nThe formal specification of the data models, consistency predicates, and synchronization procedures, allowed the usage of automatic analysis tools (namely, the Alloy Analyzer model finder) to generate a large number of diverse usage scenarios. This section presents one of the generated scenarios. Although small, we believe it is interesting enough to convey the usefulness of this process for requirement validation and for implementation testing.\n\nAs the initial state of this scenario, consider the PTCRIS and ORCID user profiles presented below, over which both the IMPORTED and EXPORTED consistency predicates are enforced (simulating, for instance, the DeGóis CV management system). The PTCRIS profile consists of two productions, none selected to be exported, that do not share UIDs, and thus cannot be considered similar, as depicted in Figure 4.\n\nThe ORCID profile contains two groups of similar works that correspond to the two productions, depicted in Figure 5 (preferred works are depicted with round shapes).\n\nEven though the UIDs of Work1 and Work2 are not exact matches (nor their meta-data), they both share the EID1 identifier, and thus are considered similar and grouped by ORCID. These two profiles are IMPORTED-consistent because all UIDs from ORCID are known to the PTCRIS: Production0 contains the identifiers from Work0 while Production1 aggregates the identifiers from Work1 and Work2. Note that the PTCRIS productions actually contain additional UIDs not known to ORCID; this does not affect the consistency of the profiles since the goal of IMPORTED is to harvest information from ORCID to the PTCRIS service. Since no production is selected to be exported, the profiles are also EXPORTED-consistent.\n\nNow imagine that the user, after examining the production’s meta-data concluded that the two productions at the PTCRIS profile actually represent the same research output. To unify them, the user introduces a UID from Production0 in Production1 (e.g., DOI0), rendering them similar. Then, this update can be propagated to other services by exporting Production1 to ORCID that acts as the research hub, as depicted in Figure 6 (productions set to be exported are denoted by bold frames).\n\nAt this point, the profiles are no longer EXPORTED-consistent, so an identical ORCID work must be created from the exported production. After running the EXPORT procedure, the updated ORCID profile is depicted in Figure 7.\n\nThis update reflected the intentions of the user: the introduction of Work3 in the ORCID profile due to the exportation of Production1, unified the two groups under a single group of similar works. (In this scenario, Work2, one of the preferred works in the initial ORCID profile, was preserved as the preferred, while Work0 was demoted. At the moment it is not clear how ORCID would select the preferred in this situation, so our specification also considers as a acceptable possible outcome a profile where Work0 is the preferred one.) However, this has consequences to the IMPORTED-consistency of the profile, since productions related with Work0 need now be updated with the new UIDs. Thus, this update needs to be propagated to all other relevant services (like the SARI repositories), but also to the PTCRIS service that triggered this update, since it was assumed to enforce both IMPORTED and EXPORTED Concretely, when IMPORT is run back to the PTCRIS profile, Production0 is matched with the whole group of works, resulting in the profile depicted at Figure 8, where a modification notification is associated to Production0 to add all harvested UIDs.\n\nSince the EXPORT procedure may introduce IMPORTED-inconsistencies, the service enforcing both consistency predicates should always run IMPORT after EXPORT, that is, the SYNC procedure specified above.\n\n\n4 Prototype implementation\n\nAs proof of concept, a prototype of this synchronization framework was implemented and shown to the FCCN community in its annual meeting, held in February 2015 (http://jornadas.fccn.pt). In this prototype, the following PTCRIS services and systems were involved:\n\nDeGóis The national CV system, already supporting ORCID iDs and a preliminary version of the IMPORT and EXPORT synchronization procedures. This new version of this system is expected to be released to the community in the 3rd Quarter of 2015.\n\nRCAAP The national OAI-PMH aggregator, implementing a preliminary version of the EXPORT synchronization procedure.\n\nSARI The DSpace platform used to provide institutional repository services. Version 5.1 of DSpace was used with some minor adjustments aimed to support ORCID iD (for the purpose of the demo only). Notice that the current version of DSpace, branch JSPUI, still does not support ORCID.\n\nOJS Platform used to provide hosting to open-access journals. Like DSpace, the Open Journal System was used with minor adjustments to support ORCID iD.\n\nThe demo at the aforementioned event involved the following steps:\n\n1. At DeGóis, create a new user profile and run IMPORT to populate it with research outputs harvested from ORCID.\n\n2. Insert a new production at DeGóis (entitled “The gap between technologies and science”) and run EXPORT to send it to ORCID.\n\n3. At OJS submit and approve a new article from the same researcher (entitled “Registo submetido a Revista com OJS 1”).\n\n4. At SARI deposit an article (entitled “Portuguese repositories bloom : the RCAAP project”) in a institutional open-access repository.\n\n5. These two articles are harvested by RCAAP, and then the researcher runs EXPORT to send them to ORCID. The state of the user profile at ORCID after these steps can be seen in Figure 9.\n\n\n\n6. At DeGóis run the IMPORT procedure to harvest these two outputs. The state of the user profile at DeGóis after this step can be seen in Figure 10.\n\n\n\nThis prototype showed, in our opinion, compelling and real use cases for the synchronization framework. The prototype still differs from the upcoming release version of the framework in the following aspects:\n\nThe RCAAP portal still does not implement the final version of the EXPORT procedure, that relies on ORCID Metadata Round Trip functionality to automatically feed updates to the ORCID profile.\n\nThe DSpace based SARI platform does not yet implement the IMPORT procedure. From the end user perspective, this is one of the most expected features, since it will free researchers from manually filling in meta-data. From the FCT perspective, it is also of critical importance, as the notification based synchronization service will increase the output deposit rate, and thus facilitate its open-access mandate.\n\n\n5 Related work\n\nThe laws presented in Section 3.3 are standard “well-behavedness” laws in synchronization frameworks, namely on those for bidirectional transformation, whose goal is precisely to maintain two artifacts consistent by means of two transformations that propagate updates from each to the other (for an overview of this research field please see 2). To be more precise, our formalization is based on the concrete framework of constraint maintainers, proposed by Meertens3, and later used by Stevens4 to formalize the OMG (Object Management Group) standard bidirectional transformation language QVT-R (Query/View/Transformation - Relations)5. In fact, an interesting question is whether the domain specific QVT-R language could be used instead of the general purpose Alloy to formalize the consistency predicates, and later used with a QVT-R engine (for example, the Echo tool6,7) to implement the synchronization procedures.\n\nAlloy and its Analyzer have been previously used in the validation of transformation specifications, namely for transformations specified in QVT-R8 and ATL (ATLAS Transformation Language)9. Likewise to these approaches, we have also used Alloy to verify properties of the specified consistency predicates. However, we also relied on its model finding functionalities to generate scenarios that helped the different stakeholders consensually establishing the system’s requirements.\n\nCV management systems and open-access repositories typically connect with ORCID only in the IMPORT context and only support creation notifications, not allowing the user to EXPORT research outputs back to ORCID. Such is the case of services like Impactstory (http://impactstory.org), ScienceOpen (http://www.scienceopen.com) and Symplectic’s Elements (http://symplectic.co.uk/products/elements). The exception is Thomson Reuters’ ResearcherID (http://www.researcherid.com), which also aims to provide a unique researcher identifier, and that allows the user to export research outputs back to the ORCID profile. Interestingly, some of these services, like Impactstory and Elements, resort to ORCID only to harvest UIDs and then retrieve meta-data from other trusted services, ignoring the actual ORCID works. As a consequence, any grouping of works (possibly enforced by the user) is ignored, contradicting the perspective of ORCID as a central hub for research outputs (in our scenario, there would be a different notification for each supported UID). For ORCID works without UIDs assigned, Impactstory retrieves the ORCID meta-data, while Elements currently ignores such works. In contrast, ResearcherID considers each ORCID work as an independent entry, not embracing its essence as an aggregator of research outputs from varied sources, which may lead to several duplicated entries (in our scenario, there would be a different notification for each work). Our proposed approach sits between these two approaches: while IMPORT does focus on the retrieval of UIDs, it also considers how these UIDs are grouped in the ORCID profile. In the EXPORT context, ResearcherID, unlike our approach, does not keep track of previously exported outputs. Since the ORCID API does not allow sources to introduce works with repeated UIDs, the user is currently not able to update works from ResearcherID without previously deleting them from the ORCID profile. Outputs without UIDs are duplicated in the ORCID profile when exported; our framework forbids the exportation of productions without UIDs to avoid this issue.\n\n\n6 Conclusion\n\nThis paper reported on the experience of developing an ORCID based synchronization framework for PTCRIS. This synchronization framework was recently prototyped and demoed at a national research community event, receiving quite positive feedback. During its design, formal analysis techniques and tools were used with excellent results, in particular to automatically generate usage scenarios (namely, corner cases) that proved very useful to help clarify and validate the requirements with the stakeholders. The first stable and detailed specification of the synchronization framework will be made available soon as an open-access report. We expect to have certified implementations of that specification in DeGóis, RCAAP, SARI, and two local CRIS systems by the end of 2015.\n\nThis paper is based on the ORCID API v1.2. Version 2.0, currently in early stages of development, is expected to affect some details of the described synchronization framework, but not the overall concepts. The most relevant change is that groups of similar works, as well as the preferred among them, will become explicit in the ORCID data model. This will simplify the IMPORT procedure, since it will no longer need to compute the groups of similar works.\n\nIn a future version we intend to design alternative IMPORT and EXPORT procedures with more sophisticated behaviors. For example, depending on user feedback we may consider implementing a notification dismiss feature, to accommodate users that may want to register different research outputs in their ORCID and PTCRIS profiles. Another possible interesting feature would be to allow IMPORT to somehow recognize edits to works in ORCID and automatically incorporate them in the respective PTCRIS profile.", "appendix": "Author contributions\n\n\n\nJoão Mendes Moreira is responsible at the FCCN unit of FCT for the PTCRIS program, the proponent of the synchronization architecture with ORCID as hub (Section 2), and led the prototype implementation both at DeGóis and RCAAP (Section 4). Alcino Cunha and Nuno Macedo are active researchers in the topic of model management and synchronization, and were responsible for the formal specification and verification of the synchronization framework (Section 3).\n\n\nCompeting interests\n\n\n\nThe authors disclosed no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no funds were involved in supporting this research.\n\n\nAcknowledgements\n\nThe authors would like to thank the DeGóis and RCAAP teams, namely Carlos Pinto, Luis Valério, Pedro Lopes, and José Carvalho, for their effort in the prototype implementation and for the fruitful discussions that helped us clarify and improve the specification of the framework. The authors would also like to thank the ORCID team, namely Laura Paglione, for the valuable comments on Section 3 and insight provided about the ORCID API.\n\n\nReferences\n\nJackson D: Software Abstractions: Logic, Language, and Analysis. MIT Press, revised edition, 2012. Reference Source\n\nCzarnecki K, Foster J, Hu Z, et al.: Bidirectional transformations: A cross-discipline perspective. In ICMT’09 of LNCS. 2009; 5563. : 260–283. Publisher Full Text\n\nMeertens L: Designing constraint maintainers for user interaction. In Third Workshop on Programmable Structured Documents. Tokyo University, 2005. Reference Source\n\nStevens P: Bidirectional model transformations in QVT: semantic issues and open questions. Software and System Modeling. 2010; 9(1): 7–20. Publisher Full Text\n\nMOF 2.0 Query/View/Transformation specification (QVT). OMG, version 1.1. 2011. Reference Source\n\nMacedo N, Cunha A: Implementing QVT-R bidirectional model transformations using Alloy. In FASE’13 of LNCS, Springer, 2013; 7793. : 297–311. Publisher Full Text\n\nMacedo N, Guimarães T, Cunha A, et al.: Model repair and transformation with Echo. In ASE’13. IEEE, 2013; 694–697. Publisher Full Text\n\nCabot J, Clarisó R, Guerra E, et al.: Verification and validation of declarative model-to-model transformations through invariants. J Syst Softw. 2010; 83(2): 283–302. Publisher Full Text\n\nBüttner F, Egea M, Cabot J, et al.: Verification of ATL transformations using transformation models and model finders. In ICFEM’12 of LNCS, Springer. 2012; 7635. : 198–213. Publisher Full Text" }	[ { "id": "9346", "date": "27 Jul 2015", "name": "Mikael K. Elbaek", "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 paper by Moreira et al. provides an interesting and relevant proto-type solution to the synchronization problem of several parallel research information systems in a national CRIS (Current Research Information Systems) setting. In the Portuguese CRIS landscape there are no local CRIS-systems (i.e. Pure, Converis, Symplectic etc.) in order to understand the Portuguese implementation it is important to make this distinction that the PTCRIS is an umbrella (or agregator) that collects data from several sources including a network of repositories SARI (DSpace), a national research portal (RCAAP) an OAI-PMH service provider, e.g. aggregator and a national CV management system (DeGóis). PTCRIS in that sense is not one system or piece of software. Even though this architecture may be unique to Portugal the implications of using ORCID as a syncornization framework is interesting and relevant for other national Research Information/CRIS-architectures. The result is interesting and important for developers of national and international research infrastructures – this is why there are also a few points where the paper may be improved.In the paper the authors provides an analytic and generic approach to specifying the relevant synchronization procedures asking first what and then how. The article does not go into the details of the what but describes briefly how this data has been collected and analysed. In this reviewers opinion a more detailed description of the what and a description of the selected use-cases would provide a more comprehensible analysis and clear definition of the selected scenarios. Currently the authors provides abstract descriptions of the synchronization scenarios (see chapter 3.2: Consistency predicates). Because of this level of abstraction it is not completely clear who the end-users are, is it the system managers, system owners, research managers or the researchers (authors and owners of the ORCID profiles). The proposed framework is focussing on the technical implementation of synchronization. But how will the users or researchers e.g. the ORCID profile owners influence the robustness of the synchronization job. One example is that ORCID profile owners may not want to have the same records (works/productions) in their PTCRIS profile as the want in the ORCID profile? The stakeholders of the PTCRIS could be the national government or the university of the researcher who wants all publications listed of a researcher (i.e. the numbers counts in assessments) whereas the researcher may only want a selected number of publications (e.g. the best) to be listed visible. This is briefly discussed in the conclusion. It could be suggested to separate discussion and future perspectives i.e. to a separate section. The metadata quality must be a major challenge to this synchronization framework using ORCID. It is well-known that metadata about research outputs and funding data are often not consistent or of an acceptable quality. It is not clear from the article how the CRIS infrastructure is used in Portugal. If it is used for a research assessment i.e. bibliometric assessment this must be a major issue. It would be relevant to have discussed this issue a bit further in the paper. However since the paper is presenting a prototype framework – it is acceptable that it is not dealt with in details. I hope that this will be explored further in later writings from the authors – also when ORCIDs API version 2.0 will be in production and the metadata roundtrip also will be in production. Annotations – there are some use of annotations like SYNCED (o, p) where o, p could have been described to help the reader. The description and use of PTCRIS sometimes gives the impression of PTCRIS as a single entity or system – while as far as this reviewer understands it is an umbrella that encompasses several systems that together collects and exposes the collected research output, actors and activities of Portugal. It currently includes: DeGóis, RCAAP and SARI. The authors do not discuss issues with researchers having more ORCIDs – is this not an issues for the proposed synchronization framework?Overall the paper is solid and provides an important stepping stone towards a better use and re-use of research information using ORCID as a hub. And I would be happy to accept the paper as is – but I hope that the comments will provide thought for possible revisions or further work in a later article.", "responses": [ { "c_id": "1631", "date": "29 Sep 2015", "name": "Alcino Cunha", "role": "Author Response", "response": "First of all, many thanks for your review and invaluable feedback. Here are some comments on your remarks:Indeed the paper focus more on the design methodology, leaving out many details that would be necessary to fully understand the synchronization framework. The detailed specifications of the proposed framework can now be found at http://ptcris.pt/hub-ptcris-en/. Two versions of the document are available: v0.3 concerning ORCID API 1.2 and v0.4 concerning ORCID API 2.0. The later version was improved and substantially simplified based on comments from the “end users” that are the target of these documents, namely the systems managers that will be responsible for the implementation of the synchronization procedures at their services. Unfortunately, we have still not yet written user manuals for the researchers, the real end users of the system. When we do so, they will be available at the same web site. Maybe it is not clear in the paper, but the user can already select at each PTCRIS service which research outputs (namely publications) he wants to export to ORCID, and thus propagated to other services. Also, the import mechanism is notification based, so the user can choose not to import some of the research outputs. Also at ORCID, he can set the privacy level of the research outputs so that they are visible for importing by ORCID members, but not to the general public. As such, we believe the framework already provides the flexibility you propose. Indeed, metadata quality is a major problem to be considered. We believe this problem will be less severe in the PTCRIS ecosystem, because some of the services (namely OA aggregators, such as RCAAP) will be mandated to export to ORCID (using the metadata roundtrip facility you refer) records with “good enough” metadata. Users will be recommended to give preference to such records in ORCID, and we also intend to have the import procedure biased to prefer metadata from such reliable sources. You are right, and that will be improved in a future version. In this case, o stands for the user ORCID profile and p for the user profile at the PTCRIS service intending to synchronize with ORCID. You are absolutely right, and we intend to incorporate your remark in a future version of the paper. This should not be a problem, because due to the mandated use of ORCID iDs in the recent national research assessment exercises, most Portuguese researchers already have a single ORCID iD registered at the national science foundation, and are thus fully aware of what is their “official” ORCID iD that will be used in the PTCRIS ecosystem." } ] }, { "id": "9347", "date": "28 Jul 2015", "name": "Keith G. Jeffery", "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 paper describes a synchronisation method across multiple research information systems Holding CV information and full-text articles with metadata) in Portugal using ORCID as the essential linking / identification mechanism. The synchronisation protocol is specified formally and a prototype has been constructed which is reported to have been received successfully.The paper provides a useful advance in attempts to ensure consistency in research output publications and thus should be made widely available.There are several aspects of relevance which are not covered - or covered poorly - in the paper.The paper does not 'stand back' and describe that the need for synchronisation arises because of a lack of policy by national research funding organisations in terms of what should be recorded/stored and how it should be recorded/stored. This policy vacuum has led - in all countries not just Portugal - to the provision of multiple independent incompatible systems and inconsistent information; The synchronisation discussed covers only publications; there are many other aspects of research information that may be recorded inconsistently in multiple systems. This restriction should be made clear; The data models do not seem to be recording temporal information about (a) the valid time (interval) relating to the publication and (b) the transaction time relating to the publication appearance and any subsequent changes in the digital stored representation. These temporal elements are necessary for precedence, provenance and version differentiation. The need for and lack of temporal information should be discussed; The mechanisms for grouping 'like' publications are likely to produce inconsistent results. Relying one one or more equivalent identifiers is clearly not sufficient and the paper correctly indicates that metadata may be inconsistent. However, some measure of 'separation distance' based on metadata element values could - and probably should - be used. This would be facilitated by point 5 below; The data models used in the participating systems are very simple and do not really reflect the real world. There exist in the real world complex inter-relationships between (versions of) publications (and indeed between publications and other entities in the research world such as persons, organisations, projects, funding, datasets, software, equipment...) and their correct representation requires a syntax and semantics beyond simple identification. The use of an overarching, superset data model in PCTRIS with formal syntax and declared semantics (thus ensuring referential and functional integrity) and providing a superset over the metadata of the collection of connected systems would give a greater chance of correct assignment of publications to groupings of 'like' publications;However, as indicated, the paper provides a useful advance in the state of the art and forms a good basis for discussion of more complete solutions to a very difficult problem.", "responses": [ { "c_id": "1630", "date": "29 Sep 2015", "name": "Alcino Cunha", "role": "Author Response", "response": "First of all, many thanks for your review and invaluable feedback. Here are some comments on your remarks:You are absolutely right, and in a future version of the paper we will improve the introduction to incorporate your remark. Again, you are right. The present synchronization framework focus only on research outputs that are covered by ORCID and which have UIDs associated, namely publications and projects. We thought we had that clearly stated in the introduction, but apparently not. We will do so in a future version. Indeed, our current version of the synchronization framework does not incorporate such temporal information, but this is one of the key points for future work, namely: improve the synchronization mechanism with versioning info to track changes to information, and thus reliably recognize and propagate changes to metadata. To help us in that task, we have already enrolled in the PTCRIS team some distributed systems researchers, whose research is precisely focused on distributed data aggregation, synchronization, and dissemination. (and 5.) The unique identifier vs. metadata based synchronization was (and still is) a question of much debate inside the PTCRIS team. We are fully aware of the limitations of opting for the the former, but ultimately we opted for that because we believe it is more transparent (and easy to understand) to the end user, in particular because it relies on the same publication matching algorithm that is currently used by ORCID. ORCID plays a central role in this framework, and having a completely different matching algorithm would be quite confusing for a user that frequently uses both ORCID and the PTCRIS services. As soon as the first services implementing the proposed mechanism are released (hopefully until the end of 2015), we will conduct a proper evaluation to assess the reliability and usability of the system. We hope to include the results of such evaluation in a future version of the paper." } ] }, { "id": "9344", "date": "07 Aug 2015", "name": "Simeon Warner", "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 article describes work to use ORCID iDs as the basis for synchronization of data between distributed components of a national CRIS system in Portugal. Portugal has very high ORCID coverage because of the mandated use of ORCID iDs for a 2013/2014 research assessment exercise, and thus provides a good test-bed for ORCID iD based synchronization. The background for this work and the architecture used are well described in the introductory sections.My principal criticism of the article is that it describes a formal modeling process, that was used to assess and refine the synchronization strategy, in insufficient detail for a reader assess the correctness of or repeat the analysis. Yet, the description of this modeling process nonetheless introduces a number of formalisms that distract from the flow of the article. Ideally, the description of the formal modeling process would be separated out and then expanded to a point where, along with the Alloy source files, the reader could repeat the analysis (plans to provide these details are mentioned in the conclusion, the article should be updated to link to them when available). Because this is a descriptive software tools article, I feel that this deficiency does not undermine the article's usefulness or conclusions.The later sections of the article describe the prototype implementation, a modest amount of additional work to be completed before a production release, and related work. A number of other issues would benefit from discussion, such as the handling of edits to works in ORCID and other sources; user testing, understanding and acceptance of the system; and limitations on the type of research output included beyond publications.Overall I feel this is a useful and timely description of new work in an evolving area and thus approve it as is. I hope the authors will follow-up with an article evaluating the system in production.", "responses": [ { "c_id": "1629", "date": "29 Sep 2015", "name": "Alcino Cunha", "role": "Author Response", "response": "First of all, many thanks for your review and invaluable feedback. Concerning your principal criticism, we agree that the detail in the paper is insufficient for the reader to assess the formal design methodology that was followed. The detailed specifications of the synchronization framework can now be found at http://ptcris.pt/en/hub-ptcris-en/. Two versions of the document are available: v0.3 concerning ORCID API 1.2 and v0.4 concerning ORCID API 2.0. The former contains the formal Alloy models in Appendix, but they were removed from the later, which was substantially simplified based on comments from our partners. The formal Alloy models used in the design of v0.4 can be found in the following git repository: https://gitlab.fccn.pt/dev-ptcris/ptcrisync/tree/masterYour remaining suggestions are quite relevant also, and some of them are already on our future work list, namely: improve the synchronization mechanism with versioning info to track changes to information; conduct proper user evaluation once the first implementations are released (hopefully, still in 2015); and extend the synchronization mechanism to handle other research outputs. Concerning the later, for research outputs that also have unique identifiers associated the extension is rather trivial. This is the case of research projects, already stored in ORCID." } ] } ]	1	https://f1000research.com/articles/4-181
https://f1000research.com/articles/3-274/v1	13 Nov 14	{ "type": "Data Note", "title": "Journal subscription expenditure of UK higher education institutions", "authors": [ "Stuart Lawson", "Ben Meghreblian", "Ben Meghreblian" ], "abstract": "The academic libraries of higher education institutions (HEIs) pay significant amounts of money each year for access to academic journals. The amounts paid are often not transparent especially when it comes to knowing how much is paid to specific publishers. Therefore data on journal subscription expenditure were obtained for UK HEIs using a series of Freedom of Information requests. Data were obtained for 139 HEIs’ expenditure with seven publishers over a five-year period. The majority of institutions have provided figures but some are still outstanding. The data will be of interest to those who wish to understand the economics of scholarly communication and see the scale of payments flowing within the system. Further research could replicate the data collection in other jurisdictions.", "keywords": [ "Journals", "subscription", "higher education institution" ], "content": "Introduction\n\nThe amount of money paid by higher education institutions (HEIs) to access academic journals is of high interest to the academic community, and academic libraries in particular as they are responsible for the vast majority of journal purchases. In light of current trends within academic publishing towards open access models rather than subscription models, the economics of the publishing industry have come under increasing scrutiny, but accurate data about the flow of money within the system is difficult to come by. Libraries do not usually publish details of their expenditure with individual publishers and there is no official source of these data. This situation led to undertaking this research to make journal subscription expenditure openly available.\n\nFreedom of Information (FOI) requests were sent to HEIs to obtain the data. While the authors considered using a diplomatic approach and asking individual libraries to publish their data, this would have taken a considerable amount of time, and while some libraries may have been happy to publish the data themselves, others may not have seen the value in it. The situation is also complicated by the fact that some publishers insist on having non-disclosure clauses in their contracts with libraries, which prohibit them from disclosing some aspects of the deals. The UK’s Freedom of Information Act (2000) overrides these clauses and allows full data to be obtained by sending FOI requests.\n\nIt is hoped that the data contained within this dataset will contribute to a better informed discussion surrounding the issue of how scholarly communication could or should be funded. Further research could undertake a similar endeavour in the 100 other countries (McIntosh, 2014) which have FOI laws, in order to work towards understanding the costs of scholarly communication on a global scale.\n\n\nMaterials and methods\n\nA list of HEIs was created based on UK institutions which the Higher Education Statistics Agency (HESA, n.d.) collects data about. The list was not fully comprehensive because it excluded some Welsh universities which have recently undergone restructuring, and the authors were unclear on how to represent the expenditure of merged institutions across different years. In order to obtain data which cover the majority of HEI journal expenditure, seven of the largest publishers of academic journals were chosen (Elsevier, Wiley, Springer, Taylor & Francis, Sage, OUP, and CUP). An individual known to the authors sent similar request to Russell Group universities for Wiley, Springer, and OUP earlier in 2014, so that these requests were not duplicated and the authors hope to incorporate that data at a later stage.\n\nEach institution was then sent three separate FOI requests via the website whatdotheyknow.com, which sends FOI requests on behalf of UK citizens. The site was chosen because it places all correspondence in the public domain indefinitely, thus ensuring that the data will be verifiable. The three requests were grouped as follows: Group 1 - Wiley, Springer, OUP; Group 2 - Taylor & Francis, Sage, CUP; Group 3 - Elsevier. The groupings were chosen to ensure that each request would not be too onerous for an HEI to respond to, as stipulated under the UK’s FOI law. Elsevier data were requested separately because the nature of their contract with libraries means that the institution must contact Elsevier when it receives a request, thus increasing the time burden on institutions.\n\nThe figures should include payments made directly to the publishers as well as any payments made to subscription agents or intermediaries for the purchase of, and/or access to, the publishers' academic journals. Institutions were asked to provide data for the payment for journal packages such as Jisc Collections’ NESLi agreement, as well as for individual journals, and to include VAT where possible. Since the authors are relying solely on data provided by the HEIs it is not possible to independently verify whether all of these aspects of the requests have been adhered to. While this may result in some inaccuracies in individual figures, the authors do not consider that the overall scale will be unduly affected.\n\nData were requested for five calendar years (2010–14). Some institutions provided data in financial years, which for UK academic institutions is from August-July. In these cases the financial year was mapped on to the second of the two years, for example 2009–10 was mapped on to 2010. This is because although during the financial year 2009–10 it is possible that the money was actually transferred during 2009, it will have been used to pay for subscriptions for 2010.\n\nThe dataset is now well-populated but incomplete because at the time of writing, out of the 429 FOI requests that were sent there are still 91 outstanding for which data has not yet been provided. Further data will be incorporated into the dataset as it becomes available.\n\n\nData availability\n\nData can be accessed directly via Figshare at http://figshare.com/articles/Journal_subscription_costs_FOIs_to_UK_universities/1186832, http://dx.doi.org/10.6084/m9.figshare.1186832 (Lawson & Meghreblian, 2014).\n\nData were obtained from each institution sending separate FOI requests via the website whatdotheyknow.com. Requests can be viewed individually at https://www.whatdotheyknow.com/user/stuart_lawson#foi_requests and https://www.whatdotheyknow.com/user/ben_meghreblian#foi_requests.", "appendix": "Author contributions\n\n\n\nStuart Lawson authored the data note and designed, carried out, and recorded the data collection.\n\nBen Meghreblian designed, carried out, and recorded the data collection.\n\n\nCompeting 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\nReferences\n\nFreedom of Information Act 2000. Reference Source\n\nLawson S, Meghreblian B: Journal subscription costs – FOIs to UK universities. Figshare. 2014. Data Source\n\nMcIntosh T: ‘Paraguay is 100th nation to pass FOI law, but struggle for openness goes on’. The Guardian. 2014. Reference Source\n\n‘Overview’. HESA. 2014. Reference Source" }	[ { "id": "6694", "date": "14 Nov 2014", "name": "Theodore Bergstrom", "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 have collected a set of data that is likely to be very useful both to researchers who investigate journal pricing and to librarians who want to know more about what others are paying when they make their own bargains with publishers.\n\nThe methods of collection seem sound and the work appears to be carefully done.There is one thing that I would like to see clarified. The report states the amounts that are paid to each publisher, but do not make it clear whether all of the reported institutions are buying the same thing from the publishers.\n\nFor example, Elsevier markets a \"Freedom package\" which includes a published list of journals constituting most, but not all, of the journals they publish.They also market a \"Complete package\" which does not include the entire Freedom package, but basically a set of journals to which the university has previously subscribed. Elsevier also separately markets the Cell Press journals, a \"clinical medicine package\", a series of monographs called \"advances in ....\" Some of the contract totals that I have seen include all or some of these.Similarly, Springer sells not only its journal package, but also a large number of textbooks and monographs. Some libraries' total expenditure with Springer may include these. It would be good to know if your totals are just for the journal bundle.With some of the other publishers, it also is not entirely clear whether the totals reported are from libraries that do not subscribe to the publisher's entire journal bundle but only to some limited subset. Once again, it would be good to have this clarified.", "responses": [ { "c_id": "1084", "date": "17 Nov 2014", "name": "Stuart Lawson", "role": "Author Response", "response": "Thank you for your comments, you're right that it would be best to clarify what is actually being purchased in each instance, as far as we are able. When one institution pays a different amount to another institution, it is likely that they are not purchasing access to exactly the same 'package' of content. We did not ask institutions to provide this information as part of the FOI requests because we believed that in some cases it would add significantly to the time it would take for them to produce responses, which may have led to refusals. Some institutions did provide this level of details in their response; those who purchase access directly from a publisher rather than through Swets appear to find it easier to provide such data.One way we could highlight this would be to mark the figures in some way to indicate whether an institution has paid for the complete package of content from a given publisher. However, in order to do this consistently, we would need to contact most or all of the institutions again in order to confirm this with them. For some institutions this information can be gleaned from the original FOI response which is viewable on whatdotheyknow.com.The most accurate way to compare between different institutions would be if we asked for each institution to release the title list of purchased content for each publisher during each year. Through a project I am working on at Jisc Collections, which aims to build an 'entitlement registry' of this information for all UK institutions, I have learned that it may be very difficult for all institutions to provide that data." } ] } ]	1	https://f1000research.com/articles/3-274
https://f1000research.com/articles/4-178/v1	01 Jul 15	{ "type": "Review", "title": "Cytochrome P450 enzymes: understanding the biochemical hieroglyphs", "authors": [ "John T. Groves" ], "abstract": "Cytochrome P450 (CYP) enzymes are the primary proteins of drug metabolism and steroid biosynthesis. These crucial proteins have long been known to harbor a cysteine thiolate bound to the heme iron. Recent advances in the field have illuminated the nature of reactive intermediates in the reaction cycle. Similar intermediates have been observed and characterized in novel heme-thiolate proteins of fungal origin. Insights from these discoveries have begun to solve the riddle of how enzyme biocatalyst design can afford a protein that can transform substrates that are more difficult to oxidize than the surrounding protein architecture.", "keywords": [ "Cytochrome P450", "CYP", "cytochrome c peroxidase", "heme-thiolate proteins", "reaction mechansim" ], "content": "Introduction\n\nThe Rosetta Stone of Egyptian antiquity is a trilingual transcription of a Ptolemaic edict of March 27, 196 BCE. The subsequent decoding of the text, rediscovered by Napoleon’s army in 1798, opened the writings of the distant past to historians and to the world. A Rosetta Stone—surely that description applies aptly today to the large superfamily of cytochrome P450 (CYP) enzymes, as well as related peroxidases, that have revealed so much about biochemical and chemical biological oxidation. Here, the languages have been the imaginative and interconnected application of structural, mechanistic, and spectroscopic idioms1,2. CYP proteins have long been known to mediate the oxidative processes involved in phase 1 drug metabolism, which occur in the liver. More than 70% of drug compounds are metabolized in this way. It has become increasingly important to identify these drug metabolites and to determine the extent to which they are toxic or actually the active form of the administered drug. For example, the platelet aggregation inhibitor, Plavix, which contains a thiophene ring, is a prodrug, inactive in its administrated form, which is first transformed by liver P450 enzymes to a thiolactone structure. The active form of the drug evolves subsequently in a second, hydrolytic step (Figure 1). By contrast, acetaminophen is transformed by hepatic P450 proteins to a toxic iminoquinone that is the cause of liver failure because of overdoses of this common over-the-counter drug. P450 enzymes of the adrenal cortex orchestrate the extensive tailoring of cholesterol in the intricate biosynthetic pathways that produce steroid hormones. This understanding has led to the development of effective dual drug strategies in oncology, particularly for the treatment of human breast cancer. One of the leading drugs, tamoxifen, functions by initial P450 metabolism to 4-hydroxytamoxifen, which then blocks estrogen receptors in the proliferating tissue. Estrogen, in turn, is biosynthetically produced in a series of P450-mediated oxidative transformations that include removal of the C19 methyl group to form an aromatic A-ring in the steroid estrogen. Blocking this P450-mediated “aromatase” reaction with aromatase inhibitors further reduces the stimulating effect of estrogen on the cancerous cells. P450 proteins are also used by pathogens such as the Tuberculosis bacillus to erect its waxy, impermeable cell membrane. In this light, pathogen CYP enzymes are obvious drug targets. Finally, there are numerous microorganisms that can derive food from even the most recalcitrant organic molecules. Here, bacterial P450 enzymes, and other iron proteins, are instrumental in consuming petroleum from environmental oil spills3.\n\n\nAnalysis and discussion of the cytochrome P450 reaction cycle\n\nFor all of these reasons, P450 enzymes have received sustained attention for decades. But how do they work? Particularly, what is the role of the unusual coordination of cysteine sulfur to the heme iron center forming the essential heme-thiolate active site (Figure 2)? And why is that sulfur so essential?\n\nThe reaction cycle proceeds clockwise from the upper left. In the resting form of P450 (R), the iron in the heme-thiolate active site is in the ferric (FeIII) oxidation state. Upon binding of the substrate molecule (S-H), molecular oxygen is bound and reduced to a coordinated hydroperoxide (0). A proton relayed from a glutamic acid through a chain of water molecules facilitates cleavage of the peroxide O-O bond (blue arrows). One role of the cysteine sulfur ligation is thought to be electron donation (push) that weakens the O-O bond. The result of peroxide O-O bond scission is to produce a stable water molecule and a highly reactive and strongly oxidizing intermediate (I). Spectroscopic evidence supports an oxo-iron(IV) porphyrin radical cation formulation for I, which is the species responsible for cleaving even very strong substrate C-H bonds. This C-H bond scission is a one-electron process wherein the substrate proton is transferred to the ferryl oxygen (Fe=O) of I to produce II and a substrate-derived radical (S•). Dual roles for the cysteine sulfur in this process are increasing the basicity of the ferryl oxygen that receives the proton while reducing the redox potential of the iron(IV) porphyrin radical cation in I. Finally, collapse of this ensemble [S• HO-FeIV-S-Cys] affords the hydroxylated product (S-OH) and the resting enzyme R to complete the cycle.\n\nIt was recognized early in the development of the P450 field that sulfur coordination to the P450 heme was responsible for the red-shifted UV-vis spectrum that displayed a strong (Soret) maximum at approximately 450 nm. Ferrous-CO adducts of typical heme proteins such as myoglobin absorb at approximately 420 nm. The red-shifted porphyrin absorbance band was instrumental in the discovery of P450 enzymes and the reactions they mediate4–8. Indeed, this spectroscopic signature is the origin of the P450 name, the P referring to the fact that these pigmented P450 proteins were found in the particulate portion of the cell lysate8–10. Confirmation of the cysteine sulfur ligation came from the first crystal structure of a soluble P450 isolated from Pseudomonas putida11. This revelation caused considerable discussion among mechanistic biochemists. Why sulfur?\n\nFor the purpose of this analysis, we will use the mechanistic reaction sequence depicted in Figure 2 as a road map1,12. A more detailed description of each step in this scenario is provided in the legend of Figure 2. For further reading, there are a number of very informative and authoritative reviews2,13–19.\n\nIt seemed counterintuitive to most coordination chemists and heme protein biochemists that sulfur coordination could be advantageous for an enzyme designed to oxidize even aliphatic hydrocarbons that have very high oxidation potentials and very strong C-H bonds. Thiols themselves are easily oxidized. Furthermore, sulfur coordination generally stabilizes higher metal oxidation states. An important break in the case arrived in 2004 with the announcement that chloroperoxidase, a chloride-oxidizing heme-thiolate protein of fungal origin, had an oxidized form that was an unusual and unique hydroxo-iron(IV) species [Cys-S-FeIV-OH] (II in Figure 2) and not a ferryl [Cys-S-FeIV=O]20. Perhaps this curious fact was a clue to the amazing abilities of P450 enzymes to break these strong C-H bonds. Subsequently, both I and II from P450 enzymes were generated and spectroscopically characterized21,22.\n\nSo why is it significant that the hydroxo-iron(IV) species [Cys-S-FeIV-OH] (II) is protonated? As can be seen, II arises from intermediate I during the substrate C-H bond cleavage event. This C-H activation, a hydrogen atom abstraction, can be dissected into two parts: the scissile proton and an electron that was part of the initial C-H bond. For this reason, this kind of hydrogen atom abstraction has been called a proton-coupled electron transfer23–26. This nomenclature emphasizes the fact that the substrate proton ends up on the ferryl oxygen of I to produce the iron(IV)-hydroxide of II while the electron has filled the radical cation hole in the porphyrin π-system. This approach also makes it apparent that the basicity of the iron-bound oxygen (Fe=O) and the redox potential of the porphyrin ring are both important in breaking the C-H bond. An increase of either parameter has the effect of increasing the strength of the FeO-H bond that is formed in II, increasing the driving force for the reaction. In this light, we can understand an important paradigm in hydrocarbon oxidation: Nature breaks strong C-H bonds by making stronger O-H bonds.\n\nIf only it were so simple. All of this remarkable catalytic chemistry is being performed within the confines of an enzyme active site that is surrounded by peptide architecture. How does CYP avoid oxidizing itself? Long-range electron transfer is ubiquitous and essential in biology27,28. If we take inventory of the susceptibility of ordinary amino acid side chains to one-electron oxidation, two of them, tryptophan and tyrosine, stand out in addition to the mysterious cysteine thiolate. Both of these amino acid residues have oxidation potentials near 1 V, as do the π-electron system of porphyrin ring and the heme iron(III). Methionine residues are also potential sites of oxidation, but the oxidation potentials are at least several hundred millivolts higher than those of tyrosine or tryptophan29. The situation for tryptophan is readily illustrated by the fact that in compound I of cytochrome c peroxidase (CCP)30–35, the structure corresponding to I in Figure 2 is a histidine-coordinated oxoiron(IV) tryptophan cation radical. This innovation appears to be a strategy to allow CCP to accept successive long-range, one-electron transfers from its substrate cytochrome c. The catalytic cycle of CYP is also initiated by two long-range electron transfers from a protein reductase partner36–38.\n\nOne solution for CYP might be to isolate the heme center with less easily oxidized amino acid side chains. Indeed, crystal structures do show an abundance of phenylalanine and alkyl chain residues such as leucine and valine nearby13. But this local redox insulation, which is probably arranged to facilitate substrate binding, would not be enough. Long-range electron transfer from protein tyrosines to heme centers over distances of 10 Å or more is still facile27,28. Indeed, the turn-on trigger of prostaglandin synthase relies on just such a tyrosine oxidation39,40. Certainly, placing the substrate (S-H) close to the ferryl oxygen as in [S-H---O=FeIV] would help. Chemists call such an atom transfer between contiguous atoms an inner-sphere process41–43, whereas a long-range electron transfer from some distant amino acid side chain to the heme center would be an outer-sphere process. Generally, inner-sphere processes occur faster than outer-sphere processes even if the two have the same driving force.\n\nSo in what ways are CYP proteins engineered through their amino acid sequence to allow long-range electron transfers to the heme-thiolate center in the early steps of oxygen binding and reduction at relatively low potentials, while at the same time preventing long-range electron transfers at the moment the highly oxidizing I is breaking a strong C-H bond? In the end, it is a balance of competitive rates and redox potential modulation by the axial thiolate ligand. Intriguingly, the cysteine thiolate of P450 proteins, as well as those of chloroperoxidase and newly discovered aromatic peroxygenase (APO) heme-thiolate proteins17, are held in place by a phalanx of peptide backbone N-H---S hydrogen bonds (Figure 3). Various thiolate electron donor parameters such as coulombic effects, σ- and π-trans-axial ligand effects, and field effects of other charges in the active site such as the heme propionate anions all could contribute to the electron push effect44. The net result would be a lowering of the redox potential of I and a compensating increase in the basicity of the ferryl oxygen (Figure 2). The extent of this electron donation effect on the ferryl basicity has been dramatically illustrated in two recent cases. For a thermostable CYP, the pKa of the oxygen-bound proton in II has been measured to be a remarkable value of 11.922. For the heme-thiolate APO from Agrocybe aegerita, that pKa value for II is 1045–47. By contrast, typical ferryl species, such as compound II of myoglobin, resist protonation even at pH 348. This large difference in ferryl basicity of 7–9 pKa units corresponds to 400–500 mV in terms of redox potential. Accordingly, sacrificing redox potential to get a more basic oxygen could indeed facilitate C-H bond cleavage while sparing the enzyme from internally generated oxidative stress.\n\n\nConcluding remarks\n\nIs the story over? By no means! First, there remains considerable uncertainty, over a range as large as 500 mV, as to what the reduction potentials of heme-thiolate compound I intermediates really are21. But here it is still early days and there have been only a handful of measurements and estimates. There is, however, a basic difference between the productive cleavage of the substrate C-H bond in S-H by I and a non-productive, long-range electron transfer from an amino acid side chain. In the productive pathway, a substrate proton arrives at the ferryl oxygen during the reaction. By contrast, the long-range electron transfer process would require a proton from some other source. Perhaps that proton is readily available through the water aqueduct leading to the P450 active site, but perhaps not. Perhaps, also, an acidic proton from the water channel can activate the ferryl oxygen for substrate hydrogen abstraction49,50. Another major point of spirited debate has to do with the extent to which energy barriers for C-H bond scission, and thus the rates of these reactions, are affected by the exact electron configurations in oxidants such as I51,52. Although there has been progress recently in the preparation of synthetic ferryl species in both high-spin and intermediate-spin electronic configurations53–57, what matters is the arrangement of spin density at the transition state [S---H---O-Fe].\n\nCYP research continues to be a rich, vibrant, and important field. Determining and understanding the reaction mechanisms of CYP substrate oxidations over the past several decades have greatly advanced a variety of fields. Numerous spectroscopic techniques and diagnostic reaction probes have been applied to dissecting the mechanism. With this knowledge in hand, drug metabolism pathways can often be anticipated, weeding out poorly performing candidates early in the drug development pipeline. CYP and APO enzymes can now be engineered and evolved for particular purposes58–62. Reaction processes are being developed by using immobilized P450 and APO enzymes. Also, new heme-thiolate proteins are being discovered.\n\n\nAbbreviations\n\nAPO, aromatic peroxygenase; CCP, cytochrome c peroxidase; CYP, cytochrome P450.", "appendix": "Competing interests\n\n\n\nThe author declares that he has no competing interests.\n\n\nGrant information\n\nThis work was supported by the National Institutes of Health (2R37 GM036298).\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 author thanks group members, collaborators, and colleagues in the field for discussion, inspiration, and many insights.\n\n\nReferences\n\nGroves JT: Models and Mechanisms of Cytochrome P450 Action. In Cytochrome P450: Structure, Mechanism, and Biochemistry. 3rd edition; Edited by Ortiz de Montellano, PR. New York: Klewer Academic/Plenum Publishers. 2005: 1–44. Publisher Full Text\n\nOrtiz de Montellano PR: Hydrocarbon hydroxylation by cytochrome P450 enzymes. Chem Rev. 2010; 110(2): 932–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAustin RN, Groves JT: Alkane-oxidizing metalloenzymes in the carbon cycle. Metallomics. 2011; 3(8): 775–87. PubMed Abstract \| Publisher Full Text\n\nHayaishi O, Katagiri M, Rothberg S: Mechanism of the Pyrocatechase Reaction. J Am Chem Soc. 1955; 77(20): 5450–1. Publisher Full Text\n\nMason HS: Mechanisms of oxygen metabolism. Adv Enzymol Relat Subj Biochem. 1957; 19: 79–233. PubMed Abstract\n\nImai Y, Sato R: Substrate interaction with hydroxylase system in liver microsomes. Biochem Biophys Res Commun. 1966; 22(6): 620–6. PubMed Abstract \| Publisher Full Text\n\nCooper DY: Discovery of the function of the heme protein P-450: A systematic approach to scientific research. Life Sci. 1973; 13(9): 1151–61. Publisher Full Text\n\nEstabrook RW: A passion for P450s (rememberances of the early history of research on cytochrome P450). Drug Metab Dispos. 2003; 31(12): 1461–73. PubMed Abstract \| Publisher Full Text\n\nOmura T, Sato R: A new cytochrome in liver microsomes. J Biol Chem. 1962; 237: 1375–6. PubMed Abstract\n\nOmura T: Heme-thiolate proteins. Biochem Biophys Res Commun. 2005; 338(1): 404–9. PubMed Abstract \| Publisher Full Text\n\nPoulos TL, Perez M, Wagner GC: Preliminary crystallographic data on cytochrome P-450CAM. J Biol Chem. 1982; 257(17): 10427–9. PubMed Abstract\n\nGroves JT: High-valent iron in chemical and biological oxidations. J Inorg Biochem. 2006; 100(4): 434–47. PubMed Abstract \| Publisher Full Text\n\nPoulos TL: Heme enzyme structure and function. Chem Rev. 2014; 114(7): 3919–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuengerich FP, Munro AW: Unusual cytochrome p450 enzymes and reactions. J Biol Chem. 2013; 288(24): 17065–73. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nLuthra A, Denisov IG, Sligar SG: Spectroscopic features of cytochrome P450 reaction intermediates. Arch Biochem Biophys. 2011; 507(1): 26–35. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMak PJ, Luthra A, Sligar SG, et al.: Resonance Raman spectroscopy of the oxygenated intermediates of human CYP19A1 implicates a compound i intermediate in the final lyase step. J Am Chem Soc. 2014; 136(13): 4825–8. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nHofrichter M, Ullrich R: Oxidations catalyzed by fungal peroxygenases. Curr Opin Chem Biol. 2014; 19: 116–25. PubMed Abstract \| Publisher Full Text\n\nDenisov IG, Makris TM, Sligar SG, et al.: Structure and chemistry of cytochrome P450. Chem Rev. 2005; 105(6): 2253–77. PubMed Abstract \| Publisher Full Text\n\nShaik S, Cohen S, Wang Y, et al.: P450 enzymes: their structure, reactivity, and selectivity-modeled by QM/MM calculations. Chem Rev. 2010; 110(2): 949–1017. PubMed Abstract \| Publisher Full Text\n\nGreen MT, Dawson JH, Gray HB: Oxoiron(IV) in chloroperoxidase compound II is basic: implications for P450 chemistry. Science. 2004; 304(5677): 1653–6. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nRittle J, Green MT: Cytochrome P450 compound I: capture, characterization, and C-H bond activation kinetics. Science. 2010; 330(6006): 933–7. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nYosca TH, Rittle J, Krest CM, et al.: Iron(IV)hydroxide pKa and the role of thiolate ligation in C-H bond activation by cytochrome P450. Science. 2013; 342(6160): 825–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nWarren JJ, Tronic TA, Mayer JM: Thermochemistry of proton-coupled electron transfer reagents and its implications. Chem Rev. 2010; 110(12): 6961–7001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMayer JM: Hydrogen Atom Abstraction by Metal-Oxo Complexes: Understanding the Analogy with Organic Radical Reactions. Acc Chem Res. 1998; 31(8): 441–50. Publisher Full Text\n\nLind J, Shen X, Eriksen TE, et al.: The one-electron reduction potential of 4-substituted phenoxyl radicals in water. J Am Chem Soc. 1990; 112(2): 479–82. Publisher Full Text\n\nBordwell FG: Equilibrium Acidities in Dimethyl Sulfoxide Solution. Acc Chem Res. 1988; 21(12): 456–63. Publisher Full Text\n\nWinkler JR, Gray HB: Electron flow through metalloproteins. Chem Rev. 2014; 114(7): 3369–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWinkler JR, Gray HB: Long-range electron tunneling. J Am Chem Soc. 2014; 136(8): 2930–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBergès J, de Oliveira P, Fourré I, et al.: The one-electron reduction potential of methionine-containing peptides depends on the sequence. J Phys Chem B. 2012; 116(31): 9352–62. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nGuo M, Bhaskar B, Li H, et al.: Crystal structure and characterization of a cytochrome c peroxidase-cytochrome c site-specific cross-link. Proc Natl Acad Sci U S A. 2004; 101(16): 5940–5. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGumiero A, Metcalfe CL, Pearson AR, et al.: Nature of the ferryl heme in compounds I and II. J Biol Chem. 2011; 286(2): 1260–8. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nVolkov AN, Nicholls P, Worrall JA: The complex of cytochrome c and cytochrome c peroxidase: the end of the road? Biochim Biophys Acta. 2011; 1807(11): 1482–503. PubMed Abstract \| Publisher Full Text\n\nJiang N, Kuznetsov A, Nocek JM, et al.: Distance-independent charge recombination kinetics in cytochrome c-cytochrome c peroxidase complexes: compensating changes in the electronic coupling and reorganization energies. J Phys Chem B. 2013; 117(31): 9129–41. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nCasadei CM, Gumiero A, Metcalfe CL, et al.: Heme enzymes. Neutron cryo-crystallography captures the protonation state of ferryl heme in a peroxidase. Science. 2014; 345(6193): 193–7. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMiner KD, Pfister TD, Hosseinzadeh P, et al.: Identifying the elusive sites of tyrosyl radicals in cytochrome c peroxidase: implications for oxidation of substrates bound at a site remote from the heme. Biochemistry. 2014; 53(23): 3781–9. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMyers WK, Lee Y, Britt RD, et al.: The conformation of P450cam in complex with putidaredoxin is dependent on oxidation state. J Am Chem Soc. 2013; 135(32): 11732–5. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nYasutake Y, Nishioka T, Imoto N, et al.: A single mutation at the ferredoxin binding site of P450 Vdh enables efficient biocatalytic production of 25-hydroxyvitamin D3. Chembiochem. 2013; 14(17): 2284–91. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nHollingsworth SA, Poulos TL: Molecular dynamics of the P450cam-Pdx complex reveals complex stability and novel interface contacts. Protein Sci. 2015; 24(1): 49–57. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMarnett LJ: Cyclooxygenase mechanisms. Curr Opin Chem Biol. 2000; 4(5): 545–52. PubMed Abstract \| Publisher Full Text\n\nvan der Donk WA, Tsai AL, Kulmacz RJ: The cyclooxygenase reaction mechanism. Biochemistry. 2002; 41(52): 15451–8. PubMed Abstract \| Publisher Full Text\n\nCreutz C, Ford PC, Meyer TJ: Henry Taube: inorganic chemist extraordinaire. Inorg Chem. 2006; 45(18): 7059–68. PubMed Abstract \| Publisher Full Text\n\nRosokha SV, Kochi JK: Continuum of outer- and inner-sphere mechanisms for organic electron transfer. Steric modulation of the precursor complex in paramagnetic (ion-radical) self-exchanges. J Am Chem Soc. 2007; 129(12): 3683–97. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nRosokha SV, Kochi JK: Fresh look at electron-transfer mechanisms via the donor/acceptor bindings in the critical encounter complex. Acc Chem Res. 2008; 41(5): 641–53. PubMed Abstract \| Publisher Full Text\n\nGroves JT: Enzymatic C-H bond activation: Using push to get pull. Nat Chem. 2014; 6(2): 89–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang X, Peter S, Ullrich R, et al.: Driving force for oxygen-atom transfer by heme-thiolate enzymes. Angew Chem Int Ed Engl. 2013; 52(35): 9238–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang X, Peter S, Kinne M, et al.: Detection and kinetic characterization of a highly reactive heme-thiolate peroxygenase compound I. J Am Chem Soc. 2012; 134(31): 12897–900. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang X, Ullrich R, Hofrichter M, et al.: Heme-thiolate ferryl of aromatic peroxygenase is basic and reactive. Proc Natl Acad Sci U S A. 2015; 112(12): 3686–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYosca TH, Behan RK, Krest CM, et al.: Setting an upper limit on the myoglobin iron(IV)hydroxide pKa: insight into axial ligand tuning in heme protein catalysis. J Am Chem Soc. 2014; 136(25): 9124–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGroves JT, Boaz NC: Biochemistry. Fishing for peroxidase protons. Science. 2014; 345(6193): 142–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoaz NC, Bell SR, Groves JT: Ferryl protonation in oxoiron(IV) porphyrins and its role in oxygen transfer. J Am Chem Soc. 2015; 137(8): 2875–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaouma CT, Mayer JM: Do Spin State and Spin Density Affect Hydrogen Atom Transfer Reactivity? Chem Sci. 2014; 5(1). PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUsharani D, Lai W, Li C, et al.: A tutorial for understanding chemical reactivity through the valence bond approach. Chem Soc Rev. 2014; 43(14): 4968–88. PubMed Abstract \| Publisher Full Text\n\nMandal D, Ramanan R, Usharani D, et al.: How does tunneling contribute to counterintuitive H-abstraction reactivity of nonheme Fe(IV)O oxidants with alkanes? J Am Chem Soc. 2015; 137(2): 722–33. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nGupta R, Lacy DC, Bominaar EL, et al.: Electron paramagnetic resonance and Mössbauer spectroscopy and density functional theory analysis of a high-spin FeIV-oxo complex. J Am Chem Soc. 2012; 134(23): 9775–84. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMcDonald AR, Que L Jr: High-valent nonheme iron-oxo complexes: Synthesis, structure, and spectroscopy. Coord Chem Rev. 2013; 257(2): 414–28. Publisher Full Text\n\nNam W, Lee YM, Fukuzumi S: Tuning reactivity and mechanism in oxidation reactions by mononuclear nonheme iron(IV)-oxo complexes. Acc Chem Res. 2014; 47(4): 1146–54. PubMed Abstract \| Publisher Full Text\n\nEngland J, Guo Y, Van Heuvelen KM, et al.: A more reactive trigonal-bipyramidal high-spin oxoiron(IV) complex with a cis-labile site. J Am Chem Soc. 2011; 133(31): 11880–3. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nHyster TK, Farwell CC, Buller AR, et al.: Enzyme-controlled nitrogen-atom transfer enables regiodivergent C-H amination. J Am Chem Soc. 2014; 136(44): 15505–8. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRoiban GD, Agudo R, Reetz MT: Cytochrome P450 catalyzed oxidative hydroxylation of achiral organic compounds with simultaneous creation of two chirality centers in a single C-H activation step. Angew Chem Int Ed Engl. 2014; 53(33): 8659–63. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nShoji O, Watanabe Y: Peroxygenase reactions catalyzed by cytochromes P450. J Biol Inorg Chem. 2014; 19(4–5): 529–39. PubMed Abstract \| Publisher Full Text\n\nSingh R, Bordeaux M, Fasan R: P450-catalyzed intramolecular sp3 C-H amination with arylsulfonyl azide substrates. ACS Catal. 2014; 4(2): 546–52. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMolina-Espeja P, Garcia-Ruiz E, Gonzalez-Perez D, et al.: Directed evolution of unspecific peroxygenase from Agrocybe aegerita. Appl Environ Microbiol. 2014; 80(11): 3496–507. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "9276", "date": "01 Jul 2015", "name": "Paul R. Ortiz de Montellano", "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", "responses": [] }, { "id": "9277", "date": "01 Jul 2015", "name": "Stephen G. Sligar", "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", "responses": [] }, { "id": "9278", "date": "01 Jul 2015", "name": "Stephen Benkovic", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-178
https://f1000research.com/articles/3-240/v1	09 Oct 14	{ "type": "Method Article", "title": "Quantitative Amplification of Cleaved Ends (qACE) to assay miRNA-directed target cleavage", "authors": [ "Suresh Damodaran", "Sajag Adhikari", "Marie Turner", "Senthil Subramanian", "Suresh Damodaran", "Sajag Adhikari", "Marie Turner" ], "abstract": "microRNA (miRNA) regulation is crucial to achieve precise spatio-temporal expression patterns of their target genes. This makes it crucial to determine the levels of cleavage of a particular target mRNA in different tissues and under different conditions. We developed a quantitative PCR method “quantitative Amplification of Cleaved Ends (qACE)” to assay levels of specific cleavage products in order to determine the extent of miRNA regulation for a specific target gene. qACE uses cDNA generated from adapter-ligated RNA molecules and relies on a carefully designed fusion primer that spans the adapter-cleaved RNA junction in qPCR to specifically amplify and quantify cleaved products. The levels of full-length transcripts can also be assayed in the same cDNA preparation using primers that span across the miRNA cleavage site. We used qACE to demonstrate that soybean roots over-expressing miR164 had increased levels of target cleavage and that miRNA deficient Arabidopsis thaliana hen1-1 mutants had reduced levels of target cleavage. We used qACE to discover that differential cleavage by miR164 in nodule vs. adjacent root tissue contributed to nodule-specific expression of NAC1 transcription factors in soybean. These experiments show that qACE can be used to discover and demonstrate differential cleavage by miRNAs to achieve specific spatio-temporal expression of target genes in plants.", "keywords": [ "microRNA", "degradome", "PARE", "mRNA cleavage", "Glycine max", "Arabidopsis thaliana", "qPCR", "qACE" ], "content": "Introduction\n\nmiRNAs are short 21-22nt RNA molecules that regulate the expression of cognate target genes primarily through post-transcriptional mechanisms1. The majority of miRNAs fine-tune target gene expression to achieve proper spatio-temporal expression patterns during plant development as well as in response to environmental changes2–4. Adaxial localization of transcripts encoding a set of HD-ZIP III transcription factors by miR166 in leaves5,6 is an excellent example. The expression of miR166 is restricted to abaxial layers in contrast to its target genes7. Mutations in miRNA recognition sites of HD-ZIP III genes (conferring resistance against miR166-mediated decay) resulted in ectopic expression of these genes in abaxial cell layers leading to adaxialized leaves8. Polarized expression of miR166 and its target HD-ZIP IIIs is conserved in other plant species as well, e.g. maize7, and soybean9. Another example is the nodule-meristem specific expression of MtHAP2-1 (alpha subunit of a CCAAT-binding NFY), which is spatially restricted by miR169, which is expressed in the infection zone adjacent to the meristem. Expression of a miRNA resistant MtHAP2-1 under the control of its native promoter led to significantly reduced nodule growth suggesting that spatially restricted expression of MtHAP2-1 is crucial for proper nodule development10. Similarly, miR399 plays an important role during inorganic phosphate (Pi) starvation by down regulating PHO2/UBC24 (a gene encoding a putative ubiquitin conjugating enzyme) mRNA11. PHO2 maintains proper levels of phosphate transporter activities12 and its down regulation increases phosphate acquisition. These examples demonstrate that mere transcriptional regulation is not sufficient to achieve proper spatial and temporal expression of target genes and miRNA regulation is crucial for additional fine-tuning of gene expression.\n\nIdentification, validation and quantification of the extent of regulation of specific target genes by miRNA is crucial for proper understanding of their biological roles. Bioinformatics-based prediction of miRNA targets was developed based on the extent of base-pairing between miRNAs and their targets13 and subsequently improved based on base-pairing in specific “seed” regions as well as the cleavage site11,14. A defining feature of miRNA-guided regulation in plants is that cleavage takes place precisely between 10th and 11th nucleotide from the 5’end of miRNA in the complementary region of the target transcript. A modified RNA Ligase Mediated-5’Rapid Amplification of cDNA Ends (RLM-5’RACE) method is being used to validate cleavage by miRNAs by mapping the site of cleavage15. In this technique an RNA oligo of known sequence is ligated to the mRNA molecules that had an uncapped 5’end arising due to RNase cleavage. OligodT-primed cDNA molecules from adapter ligated RNA is then used to amplify specific target RNA molecules (using a forward primer from the adapter and gene-specific reverse primer). Subsequent sequencing of these amplicons can precisely map the cleavage site. This technique serves as an excellent qualitative method to validate miRNA-mediated cleavage of targets. However, this method cannot quantify the extent of miRNA cleavage i.e. increased cleavage in specific tissue types and/or specific developmental stages/time points. In addition, among the multiple targets of a particular miRNA, some are cleaved more efficiently than the others for example, miR393 mediated reduction in the flag22 elicited wild type seedlings, showed reduction of TIR1, AFB2, AFB3 but not AFB116. Recently, high throughput sequencing-enabled methods were developed for genome-wide validation (and identification) of miRNA targets (“degradome” or PARE (Parallel Analysis of RNA Ends))17,18. Briefly, an RNA adaptor with a MmeI recognition site is ligated to the cleaved ends of polyA RNA molecules. Ligated molecules are reverse-transcribed, amplified linearly and digested with MmeI (which cleaves ~20 nucleotides (nt) away from the recognition site) yielding short DNA molecules corresponding to the junction of the RNA adapter and miRNA-cleavage end. A dsDNA oligo adapter is subsequently ligated to the MmeI end and these molecules are linearly amplified, and sequenced using high throughput methods. This results in ~20 nt cleavage signatures that can be used to map miRNA-directed cleavage sites (e.g. CleaveLand bioinformatics pipeline (Addo-Quaye et al. 2009). Normalized abundance of cleavage signatures from specific target genes can be used to quantify miRNA-directed cleavage. However, miRNA binding sites of target genes are highly conserved. Therefore, these short signature sequences generated by degradome/PARE analyses might be shared by more than one gene making it impossible to distinguish cleavage levels of specific target transcripts. For example, at least two targets of miR319 share the same cleavage signature in Arabidopsis (Table S1). Other examples include targets of miR161 and miR172. This is especially significant in species with extensive genome duplications (e.g. soybean) and polyploidy species (e.g. wheat) that have a number of paralogs with high sequence identity. For instance, in soybean five auxin response factor (ARF) genes targeted by miR160 shared the same cleavage signature in the degradome library and RLM-5’RACE had to be used to confirm cleavage of individual target genes19. In addition, degradome/PARE analysis is a global analysis method and does not allow determining cleavage levels of a small set of miRNA targets (Table S1). A semi-quantitative method to assay miRNA-directed cleavage was reported by Schwab et al. (2005)14. However, the method was not widely used subsequently perhaps due to the lack of validation under different conditions and/or its semi-quantitative nature. We enhanced this method through the use of qPCR and named it “quantitative Amplification of Cleaved Ends” (qACE), and evaluated its ability to assay the extent of miRNA-directed cleavage of specific target genes. qACE allows quantification of specific target mRNA molecules resulting from miRNA-directed cleavage in different tissue-types and/or under different experimental conditions. We demonstrate the applicability of this method to (i) assay increased cleavage of targets in miRNA over-expressing plants, (ii) assay reduced cleavage of targets in miRNA-deficient plants, and (iii) to identify a key role for miR164 in conferring nodule-enriched expression of two of its target genes.\n\n\nMaterials and methods\n\nArabidopsis thaliana wild type Ler and hen1-1 mutant seeds (obtained from ABRC, Columbus, OH: Stock#CS6583) were surface-sterilized and grown in Sunshine mix #1 (Tessman Company, Sioux Falls, SD) under 16h light at 25°C. Leaf, stem and flower tissues were harvested at the same developmental stages, immediately frozen in liquid N2 and stored at -70°C. Soybean seeds Glycine max cv. Williams-82, the genotype used for genome sequencing project20 were surface-sterilized21 and germinated in 4” pots filled with a mixture of vermiculite: perlite (Hummert International, MO) in the ratio of 1:3 and watered with nitrogen free plant nutrient solution22. The plants were grown in a controlled environment vertical growth chamber (Conviron Growth chamber, Manitoba, Canada). Growth conditions used were: 16h light and 8h dark and 50% relative humidity with a day and night temperature of 25°C and 20°C respectively. For nodulation assays 5 days old germinated plants were inoculated with Bradyrhizobium japonicum (USDA110) grown in Vincent’s rich medium prepared and supplemented with Chloramphenicol (antibiotic selection marker) at 30°C23. The plants were inoculated with B. japonicum at a concentration of OD600=0.08 and 14 days post inoculation mature nodules and the root segment adjacent to the nodule were harvested separately.\n\nTo over-express miR164, the precursor of gma-miR164 (miRBaseID MI0007209) was PCR amplified using G. max genomic DNA as template. The PCR product was initially cloned in to an entry vector PCR8/GW/TOPO (Invitrogen, Carlsbad, CA) and subsequently in to the binary vector pCAMGFP-CsVMV: GW24 using a Gateway LR clonase (Invitrogen, Carlsbad, CA) reaction. The vector was transformed in to Agrobacterium rhizogenes K599 cells, using electroporation and composite plant transformation was performed as described25. Transgenic GFP roots were collected on dry-ice and stored at -70°C.\n\nTotal RNA was isolated from the different tissues using TRI reagent (Product#T9424, Sigma Aldrich, St. Louis, MO) as described previously21. The RNA was quantified using Nanodrop spectrophotometer (ND1000, Thermo Scientific, Wilmington, DE) and RNA integrity was verified using agarose gel electrophoresis. Adapter ligated cDNA was prepared using GeneRacer kit (Product#L1502-01, Invitrogen, Carlsbad, CA) as per the manufacturer’s instructions except that calf intestinal phosphatase treatment to remove 5’ phosphate and 5’ Cap removal using tobacco acid pyrophosphatase steps were omitted. For Arabidopsis gene expression analysis 2µg of total RNA each of leaf, stem, and flower was pooled together and subject to adapter-ligated cDNA synthesis. For gene expression analysis in soybean 7µg of total RNA was used in adapter-ligated cDNA synthesis. For miRNA stem-loop cDNA synthesis, 1µg of total RNA was used and a multiplexed cDNA was performed for all miRNAs examined using M-MuLV reverse transcriptase26,27. miR1515 & U6 was used as normalization/house-keeping control for miRNAs20,28.\n\nqPCR assays were performed in MX3000P thermocycler (Stratagene/Agilent technologies, Santa Clara, CA) using SYBR Advantage qPCR premix (Product#639676, Clontech, Mountain View, CA). The data was analyzed using the MxPro software. Relative expression values for full length transcripts were performed using the dCt method29 using Actin for soybean and U6 for Arabidopsis as normalization control. Relative levels of cleaved transcripts (described as RCT in text) were calculated using the dCt method comparing the Ct values of full-length vs. cleaved transcripts in the same sample. For statistical analyses, we calculated the range of possible expression values in each sample based on the deviation between Ct values of replicates. Error bars indicate this range in each sample.\n\n\nResults\n\nqACE is an extension of the modified RLM-5’RACE method15,30 used to qualitatively validate target cleavage and the subsequent semi-quantitative method14 used to examine the levels of cleavage remnants. An RNA adapter is ligated to cleaved 5’-ends of polyA or total RNA preparations (“adapter-ligated RNA”). The adapter gets ligated to any available 5’-phosphate of ribose molecule of RNA arising due to miRNA-directed cleavage or other means (e.g. mRNA degradation), but not to full-length mRNAs, because of the 5-methyl Guanosine/CAP. Oligo-dT primed cDNAs generated from adapter-ligated RNA are subject to real-time PCR where a primer that spans the adapter junction is used as a forward primer (subsequently referred to as qACE forward primer) and a gene-specific reverse primer (Figure 1). The 5’end of the qACE forward primer corresponds to the adapter and the last six nucleotides correspond to the target sequence downstream of the cleavage site (See Discussion for additional details). We expected that the qACE forward primer combined with a gene-specific reverse primer would help specifically amplify ligation products where the adapter is ligated to the predicted/validated cleavage site. Therefore, a qPCR assay using these primers will specifically quantify cleaved mRNAs resulting from miRNA-directed cleavage. We also expected that the qACE forward primer will not amplify full-length cDNA molecules and neither would it efficiently amplify ligation products where the cleaved end does not correspond to the miRNA-directed cleavage site (Figure S4a1). Finally, we also expected that the use of a gene-specific reverse primer would amplify linearly and help distinguish different targets of the same miRNA even if the miRNA-binding sites are identical/highly conserved (Figure S4a–d). In addition, the same cDNA preparation can be used to quantify full-length molecules of target mRNA as well as using appropriate primers that span across the miRNA binding site (subsequently referred to as full-length qPCR primers). Ratio of cleaved transcripts vs. full-length transcripts will serve as a quantitative indicator of the extent of miRNA-directed cleavage of the target mRNA. The level of cleaved target transcripts also depends on the levels of transcriptional activity of the target gene. Therefore, comparing absolute levels of cleaved transcripts (or levels normalized to house-keeping genes) between two conditions or tissue types might not reveal differences in miRNA regulation. We decided to normalize the levels of cleaved transcripts to full-length transcripts using the dCt method to obtain the ratio of cleaved transcripts (RCT).\n\nTotal RNA containing full length transcripts with 7-methyl Guanosine/CAP and cleaved transcripts arising due to miRNA activity is used as starting material. An RNA oligo adapter of known sequence is ligated to the cleaved transcripts (with a 5’-Phosphate group) giving rise to adapter ligated mRNA. Oligo dT-primed cDNA is used to assay the levels of full length mRNA (qPCR) and cleaved transcripts (qACE). The qPCR forward and reverse primers (qPCR-F & qPCR-R) were designed across the miRNA binding site whereas the qACE forward primer (qACE-F) is designed complementary to adapter sequence but with six nucleotides at the 3’end specific to the sequence at 5’end of cleaved transcript, and a gene specific reverse primer (qACE-R).\n\nTo obtain proof of concept for the method, we examined the abundance of cleavage remnants in miR164 over-expressing soybean roots. We isolated total RNA from vector control and miR164 over-expressing roots and examined mature miR164 levels by stem-loop qPCR. We observed an 6-fold increase in miR164 over-expressing roots compared to control roots as expected (Figure 2a)31. We also examined the expression of GmNAC1b, a 5’-RACE-validated target of miR164 in soybean (Figure S1) and a close ortholog of AtNAC1 (Figure S2). We generated oligo-dT-primed cDNA and performed qPCR assays using primers designed across the miR164 binding site of GmNAC1b. Results from qPCR assays indicated that there was a 4.2-fold reduction in the levels of full-length GmNAC1b transcripts in miR164 over-expressing roots compared to control roots (Figure 2b). These results indicated that indeed over-expression of miR164 resulted in a reduction in full-length transcripts of its target.\n\n(a) The relative expression level of miR164 (normalized to that of miR1515) in control soybean roots and those over-expressing miR164 (miR164ox) determined using stem-loop qPCR. (b) The relative expression levels of full-length GmNAC1b (normalized to that of GmActin) in control and miR164ox roots assayed using qPCR. The data shown in (a) and (b) are average of three replicate assays and error bars indicate the range of possible values based on deviation between Ct values of replicate assays. (c) The Ratio of Cleaved Transcripts (RCT) over full-length transcripts of GmNAC1b in control and miR164ox roots assayed using qACE. The data shown are average of three replicate assays and error bars indicate the range of possible values based on deviation between Ct values of three replicate assays.\n\nNext, we used qACE to detect and quantify target cleavage in these roots. We designed qACE primers for GmNAC1b as outlined in Figure 1. First we examined if the qACE primer pair does not amplify full-length molecules of GmNAC1b using the above cDNA from total RNA without any adapter ligation. qPCR assays detected no amplification (Figure S4a1) indicating that indeed qACE primers did not amplify/detect full-length GmNAC1b molecules. Next, we prepared adapter-ligated cDNA by ligating a known RNA adapter (see methods) to total RNA and reverse-transcribing these molecules using an oligo-dT primer. We used this cDNA to assay full length and cleaved molecules of GmNAC1b using the appropriate primer pairs (Figure 1) and calculated RCT values for each NAC1b in control and miR164 over-expressing roots. There was an 11.6-fold increase in the ratio of cleaved GmNAC1b transcripts in miR164 over-expressing roots compared to the control roots (Figure 2c) indicating that indeed miR164 over-expressing roots had increased cleavage of GmNAC1b. This experiment demonstrated that qACE could detect and provide a quantitative indicator of the levels of target cleavage.\n\nNext, we examined the levels of target cleavage in Arabidopsis wild-type (Ler) and miRNA-deficient hen1-1 mutant plants. It was previously demonstrated that hen1-1 mutants accumulated less amounts of miRNAs and that there was an increase in the levels of miRNA targets in these plants25. We used stem-loop qPCR to determine the levels of three different conserved miRNAs: ath-miR160, ath-miR164 and ath-miR159 in Arabidopsis Ler and hen1-1 plants27. As reported earlier, a clear reduction in the levels of mature miR160, miR164 and miR159 was observed in hen1-1 plants in comparison to Ler (control) (Figure 3a). However, it should be noted that the levels of reduction were not uniform for all three miRNAs; fold change of -3.2, -21, and -18 respectively.\n\n(a) Expression of ath-miR160, ath-miR164 and ath-miR159 in wild-type Ler and hen1-1 tissues assayed by stem-loop qPCR. The expression levels of the miRNAs were normalized to U6. The data shown are average and error bars indicate the range of possible values based on deviation between Ct values of three replicate assays. (b) Levels of full-length target transcripts (AtARF17 target of miR160, AtNAC1 & AtCUC2 target of miR164 and AtMYB65 target of miR159) in Ler and hen1-1 tissues analyzed using qPCR. Gene expression levels were normalized to U6 and further confirmed using two additional housekeeping genes, AtEF-α and AtGAPDH (data not shown). The data shown are average and error bars indicate the range of possible values based on deviation between Ct values of two replicate assays. (c) The Ratio of Cleaved Transcripts (RCT) over full-length transcripts of AtARF17, AtNAC1, AtCUC2, AtMYB65 in Ler and hen1-1 tissues assayed by qACE. Data shown are average and error bars indicate the range of possible values based on deviation between Ct values of two replicate assays.\n\nNext, we used qPCR and qACE respectively to determine the levels of full length and cleaved transcripts of selected targets of the above miRNAs. AtARF17 is an auxin response factor post-transcriptionally regulated by ath-miR16032. As reported previously, we also observed an increase in the expression of AtARF17 in hen1-1 plants (~1.4-fold increase; Figure 3b). Similarly, AtNAC1 and AtCUC2, targets of ath-miR164 showed ~3.5 fold increase in gene expression levels in hen1-1 compared to Ler (Figure 3b). AtMYB65, known to be regulated by ath-miR159 had a 4.6-fold increase in transcript levels in hen1-1 compared to Ler (Figure 3b). qACE assays indicated a significant reduction in the abundance of miRNA-directed cleavage products for all these target genes in hen1-1. For example, the levels of cleaved AtARF17 levels were lower in hen1-1 plants resulting in a ~3.5-fold reduction in RCT (Figure 3c). Similarly, AtNAC1 had a 6.3-fold reduction in RCT in hen1-1 compared to Ler. AtCUC2 another target of miR164 had only a 2.5-fold decrease in RCT in hen1-1 vs. Ler. Note that both AtNAC1 and AtCUC2 had a 3.5-fold increase in the levels of full-length transcripts (hen1-1 vs. Ler) but the reduction in RCTs was very different. This data indicated that the extent of gene expression regulated by miR164 is perhaps different between these two targets. Possible reasons include differences in cleavage efficiency as well as differences in the extent of overlap in tissue domains where the miRNA and the targets are expressed33,34 qACE helped identify this difference where as simple comparison of full-length transcript levels by qPCR would not have identified it. Finally, AtMYB65 showed a ~3.5-fold reduction in RCT in hen1-1 vs. Ler (Figure 3c). Evidence from these experiments clearly demonstrated the ability of qACE to provide a quantitative indicator of miRNA-directed cleavage of targets resulting from changes in cognate miRNA levels.\n\nHaving demonstrated the ability of qACE to detect and quantify cleavage of specific miRNA targets, we used the technique to examine the role of miRNAs in governing nodule-specific/enriched gene expression in soybean. We compiled a total of 326 miRNA genes classified in to 134 families from soybean and predicted a total of 596 genes targeted by these miRNAs in soybean35. Among these, a set of miRNA-target pairs had inverse expression pattern between mature symbiotic nodules (MN) and adjacent root tissues (ABMN; data not shown). We identified two NAC transcription factors, GmNAC1a and GmNAC1b (5’-RACE validated targets of miR164) that were expressed in a nodule-enriched manner. While GmNAC1a had ~6.3-fold higher expression in MN vs. ABMN tissues, GmNAC1b had a 3.2-fold higher expression (Figure 4a). Interestingly, miR164 had a nodule-excluded expression pattern i.e. ~40-fold lower expression in MN vs ABMN (Figure 4b).\n\n(a) Levels of full-length GmNAC1a & b transcript levels in mature nodules (MN) and adjacent root tissues (ABMN) assayed by qPCR and expression levels were normalized to GmActin. Data shown are average and error bars indicate the range of possible values based on deviation between Ct values of two replicate assays. (b) Level of gma-miR164 in MN and ABMN tissues assayed by stem-loop qPCR. miRNA expression levels were normalized to that of miR1515. Data shown are average and error bars indicate the range of possible values based on deviation between Ct values of three replicate assays. (c) The Ratio of Cleaved Transcripts (RCT) over full-length transcripts of GmNAC1a & b assayed by qACE. Data shown are average from two biological replicates and error bars indicate ± SD.\n\nBased on such inverse expression of these miRNA-target pairs we hypothesized that nodule-specific expression of these target genes might be at least in part regulated by nodule-excluded expression of miR164. In other words, miR164 might actively cleave these target genes in adjacent root tissues to restrict their expression to the nodules. Such a mechanism combined with transcriptional regulation can be used to generate expression gradients as observed between miR166 and its HD-ZIP III proteins during xylem development36. However, mere inverse expression of miRNAs and their targets is not conclusive evidence for our hypothesis. If indeed, these target genes were preferentially cleaved in ABMN tissues to ensure MN-specific expression, we would expect to observe a larger proportion of cleaved transcripts in ABMN tissues vs. MN tissues.\n\nWe used qACE to determine the cleavage levels of GmNAC1a and GmNAC1b in ABMN and MN tissues and calculated RCT values. It is worth noting that these targets share the same cleavage signature in degradome/PARE analyses (Table S1). We designed gene specific reverse primers that distinguished these genes (Table S2 and Figure S3a & b). Both GmNAC1a and GmNAC1b had increased cleavage in ABMN tissues compared to MN tissues. Interestingly, consistent with the higher enrichment of GmNAC1a in MN tissues, this gene had a higher cleavage in ABMN tissues (~23.5-fold higher than MN tissues). GmNAC1b also had higher cleavage in ABMN tissues (~15.8-fold higher than MN tissues; Figure 4c). Increased cleavage of GmNAC1a in ABMN compared to GmNAC1b was consistent with increased enrichment of full length GmNAC1a transcripts in MN compared to that of GmNAC1b. qACE assays strongly support the hypothesis that increased cleavage of targets in ABMN tissues by miR164 contributes at least in part to determine their nodule-enriched expression.\n\nIn summary, we have demonstrated that qACE can detect and quantify miRNA-directed cleavage of specific target genes using miRNA over-expression and miRNA-deficient tissues. We successfully used the technique to identify differential cleavage of specific target genes between nodule and adjacent root tissues revealing an additional miRNA-directed mechanism that results in nodule-specific/enriched gene expression.\n\n\nDiscussion\n\nqACE provides a quantitative indicator of miRNA-directed cleavage of specific target genes by assaying the levels of cleavage remnants. A number of studies have shown reduced levels of target transcripts as an indirect evidence of miRNA-directed cleavage36,37. Northern hybridization has also been used to estimate miRNA-directed cleavage. For example, a putative cleavage product of NAC1 resulting from miR164 activity was detected using a 3’-specific probe in Arabidopsis33. More recently, degradome/PARE analysis has enabled quantification of global miRNA cleavage17,18. However, this method cannot distinguish target transcripts that share the same cleavage signature (see Introduction). qACE, an enhancement to the semi-quantitative method developed by Schwab et al. 200514, enables quantification of miRNA-directed cleavage of a smaller set of target genes and also can differentiate closely related genes. During the preparation of this manuscript a near identical method was used by Li et al. (2014)38 to distinguish efficiencies of miR159-directed cleavage of different MYB33 targets with specific mismatches. This study validated the usefulness of the qACE method. We have performed proof of concept and control experiments (e.g. demonstrating lack of amplification using qACE primers in non-adapter ligated cDNA preparations) to demonstrate the method’s utility. We have demonstrated this using GmNAC1a and GmNAC1b which share a similar cleavage signature (Table S1). Our results show that using qPCR (to assay reduction in full-length transcripts) and qACE (to assay the levels of cleaved transcripts) on the same cDNA preparation, one can estimate specific levels of cleavage for each target gene. For example, we demonstrate that both NAC1 and CUC2 had similar levels of increase in hen1-1, but different levels of reduction in miR164-directed cleavage.\n\nIt should be noted that qACE relies on prior knowledge about the validity of the cleavage site e.g. obtained through 5’-RACE or degradome/PARE analyses. It cannot be used to validate a predicted cleavage site. In addition, one has to assume that the stability of cleavage products are identical among multiple target genes that are compared. However, such an assumption has to be made even for alternate methods such as degradome/PARE analysis. A potential limitation of qACE is the level of target gene expression and thus the detectable level of cleavage products. Among the different genes we tested, RCTs ranged from 0.0033 (GmNAC1b Figure 2c) to 0.2 (AtNAC1; Figure 3c). As one can imagine, for very poorly expressed genes with a low levels of miRNA cleavage, it can be technically challenging to reliably detect cleavage products. We used linearity and efficiency assays to determine valid range of Ct values that are reliable for each gene of interest (See below). Alternate solutions to this deficiency include the use of xrn4 mutants as done previously18 or the use of linear pre-amplification of the adapter-ligated RNA molecules by incorporating a T7 promoter in the adapter. However, this will amplify only cleaved products and not full-length transcripts. In addition, specific miRNAs also induce generation of secondary sRNAs from the cleaved product making them unavailable for quantification. Another limitation of qACE is its inability to distinguish cleavage products with closely spaced cleavage sites. For example, G. max has twenty miR166 genes that are predicted to target all twelve GmHD-ZIP III genes. We have identified two major cleavage products from several of these genes where the position of cleavage site differs by two nucleotides. Based on the known miR166 variants in soybean, we hypothesized that these two cleavage products arose from the activities of miR166a and miR166h respectively (Figure 5). We attempted to distinguish these cleavage products by qACE. Unfortunately qACE failed to differentiate the cleaved molecules arising due to gma-miR166h owing to loop formation in template which enables amplification of the qACE primer pairs designed to detect cleavage levels arising due to miR166a. Despite these minor limitations, we have demonstrated that qACE is a very useful method that can provide a quantitative indicator of the cleavage levels of specific miRNA target genes.\n\nThe images depict binding of the qACE forward primer (qACE-F) to cleavage products resulting from miR166h and miR166a guided cleavage of GmHD-ZIP III transcripts. The primer-template combination in the top panel resulted in no amplification underscoring the specificity of the primer. However, the primer-template combination shown in the bottom panel resulted in non-specific amplification. We hypothesize that the 6nt binding stretch at the 3’end was the reason for such non-specific amplification. The oligo adapter ligated to the cleaved end is indicated in bold letters and as grey bar.\n\nqACE being a qPCR-based method, specificity and linearity of primers used are key factors to be considered39. The first level of specificity is to ensure that the qACE forward primer does not amplify full-length cDNA molecules or bind non-specifically to other adapter-ligated molecules. We achieved this by careful design such that this primer corresponds to the adapter-cleavage product junction with 6 nucleotides at the 3’end of the primer corresponding to the nucleotides downstream of the cleavage site in the target gene. A shorter design (a 4nt sequence for example), might result in non-specific binding to non-specific cleavage products due to the increased probability of binding sites (data not shown). This can potentially be overcome with a gene-specific reverse primer (See below). On the other hand, longer than 6nt design had a higher thermodynamic stability making it possible to bind to full-length cDNA molecules (depending on the GC content; data not shown). The second level of specificity is distinguishing target genes with identical miRNA cleavage signatures (and therefore having identical qACE forward primers). This is again achieved by carefully designing gene-specific reverse primers. For linearity assays for qACE primers, one has to use adapter-ligated cDNA. Due to the lower abundance of cleavage products, we made cDNA preparations from 5–7µg of DNase treated total RNA. To efficiently utilize these cDNA preparations for qACE assays, we used PCR amplified cDNA for linearity assays. We amplified a small amount of adapter-ligated cDNA using 5’adapter as forward primer and 3’PolydT adapter as reverse primer in a 25 cycle PCR. We reasoned that dilutions of this PCR product can be used to check the linearity of the qACE primer pairs, but cannot be used for qACE assays (due to non-linear amplification). This enabled the use of adapter-ligated cDNA primarily for qACE assays and increased the overall efficiency of the method in terms of amount of RNA requirement.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data qACE to assay miRNA-directed target cleavage, 10.5256/f1000research.5266.d3638340", "appendix": "Author contributions\n\n\n\nSS conceived the study, guided data analysis, and wrote the manuscript. SD, SA and MT performed the experiments. SD and SA analyzed the data and co-wrote the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nResearch in the authors’ laboratory is supported by funds (awarded to S.S) from USDA-AFRI (2010-65116-20514), NSF-PGRP (IOS-1350189), South Dakota Soybean Research and Promotion Council and South Dakota State University and Agricultural Experiment station.\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\nThe use of equipment at the SDSU-Functional Genomics Core Facility supported in part by NSF/EPSCoR Grant No. 0091948, the State of South Dakota is acknowledged.\n\n\nSupplementary information\n\nThe arrow indicates the cleavage site in the miR164a binding site of GmNAC1b. The numbers within parenthesis indicate the frequency of miRNA cleaved product of the total number of sequences analyzed.\n\nThe amino acid sequences were aligned using global alignment in MEGA 5. The aligned sequences were used for constructing a phylogenetic tree using the rooted neighborhood joining method. The numbers in the branches indicate the distance and arrows shows the GmNAC1b presence in AtNAC1 clade.\n\nDissociation curves of qACE amplicons (a) GmNAC1a and (b) GmNAC1b. The melting temperatures (Tm) are different for GmNAC1a and GmNAC1b underscoring the specificity of the gene specific reverse primer although the qACE-F primer is common for both. The data is the average of two replicate assays.\n\nAmplification plots of GmNAC1b in (a1) control and (a2) miR164ox (miR164 over expression) amplification plots showing linear amplification of full length (red squares) and cleaved cDNA (blue circles) by GmNAC1b qPCR-F & R and qACE-F & R primers respectively determined by SYBR fluorescence. (a1) Amplification plot showing no amplification of full length cDNA by GmNAC1b qACE-F & R primers determined by SYBR fluorescence (yellow lines). The data shown are normalized fluorescence from three replicate assays.\n\nAmplification plots of AtNAC1 & AtCUC2 in (b1 & b3) Ler and (b2 & b4) hen1-1 Arabidopsis mutant. Amplification plot showing linear amplification of full length (blue circles) and cleaved cDNA (red squares) by qPCR-F & R and qACE-F & R primers respectively for AtNAC1 (b1 & b2), AtCUC2 (b3 & b4) determined by SYBR fluorescence. The data shown are normalized fluorescence from two replicate assays.\n\nAmplification plots of AtARF17 & AtMYB65 in (c1 & c3) Ler and (c2 & c4) hen1-1 Arabidopsis mutant. Amplification plot showing linear amplification of full length (blue circles) and cleaved cDNA (red squares) by qPCR-F & R and qACE-F & R primers respectively for AtARF17 (c1 & c2), AtMYB65 (c3 & c4) determined by SYBR fluorescence. The data shown are normalized fluorescence from two replicate assays.\n\nAmplification plots of GmNAC1a & b in (d1 & d3) MN (d2 & d4) ABMN tissues of soybean. Amplification plot showing linear amplification of full length (blue circles) and cleaved cDNA (red squares) by qPCR-F & R and qACE-F & R primers respectively for GmNAC1a (d1 & d2), GmNAC1b (d3 & d4) determined by SYBR fluorescence. The data shown are normalized fluorescence from two replicate assays.\n\nThe data was obtained for Arabidopsis (German et al., 2009) & Soybean (Shamimuzzaman and Vodkin 2012). Targets of a particular miRNA family with identical cleavage signatures are denoted by the same letter label (a,b, or c).\n\n\nReferences\n\nReinhart BJ, Weinstein EG, Rhoades MW, et al.: MicroRNAs in plants. Genes Dev. 2002; 16(13): 1616–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJover-Gil S, Candela H, Ponce MR: Plant microRNAs and development. Int J Dev Biol. 2005; 49(5–6): 733–44. PubMed Abstract \| Publisher Full Text\n\nJones-Rhoades MW, Bartel DP, Bartel B: MicroRNAs and their regulatory roles in plants. Annu Rev Plant Biol. 2006; 57: 19–53. PubMed Abstract \| Publisher Full Text\n\nChen X: Small RNAs and their roles in plant development. Annu Rev Cell Dev Biol. 2009; 25: 21–44. PubMed Abstract \| Publisher Full Text\n\nEmery JF, Floyd SK, Alvarez J, et al.: Radial patterning of Arabidopsis shoots by class III HD-ZIP and KANADI genes. Curr Biol. 2003; 13(20): 1768–74. PubMed Abstract \| Publisher Full Text\n\nPrigge MJ, Otsuga D, Alonso JM, et al.: Class III homeodomain-leucine zipper gene family members have overlapping, antagonistic, and distinct roles in Arabidopsis development. Plant Cell. 2005; 17(1): 61–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJuarez MT, Kui JS, Thomas J, et al.: microRNA-mediated repression of rolled leaf1 specifies maize leaf polarity. Nature. 2004; 428(6978): 84–8. PubMed Abstract \| Publisher Full Text\n\nMallory AC, Reinhart BJ, Jones-Rhoades MW, et al.: MicroRNA control of PHABULOSA in leaf development: importance of pairing to the microRNA 5’ region. EMBO J. 2004; 23(16): 3356–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWong CE, Zhao YT, Wang XJ, et al.: MicroRNAs in the shoot apical meristem of soybean. J Exp Bot. 2011. 62(8): 2495–506. PubMed Abstract \| Publisher Full Text\n\nCombier JP, Frugier F, de Billy F, et al.: MtHAP2-1 is a key transcriptional regulator of symbiotic nodule development regulated by microRNA169 in Medicago truncatula. Gene Dev. 2006; 20(22): 3084–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAllen E, Xie Z, Gustafson AM, et al.: microRNA-directed phasing during trans-acting siRNA biogenesis in plants. Cell. 2005; 121(2): 207–21. PubMed Abstract \| Publisher Full Text\n\nKuo HF, Chiou TJ: The role of microRNAs in phosphorus deficiency signaling. Plant Physiol. 2011; 156(3): 1016–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRhoades MW, Reinhart BJ, Lim LP, et al.: Prediction of plant microRNA targets. Cell. 2002; 110(4): 513–20. PubMed Abstract \| Publisher Full Text\n\nSchwab R, Palatnik JF, Riester M, et al.: Specific effects of microRNAs on the plant transcriptome. Dev Cell. 2005; 8(4): 517–27. PubMed Abstract \| Publisher Full Text\n\nLlave C, Xie Z, Kasschau KD, et al.: Cleavage of Scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA. Science. 2002; 297(5589): 2053–6. PubMed Abstract \| Publisher Full Text\n\nNavarro L, Dunoyer P, Jay F, et al.: A plant miRNA contributes to antibacterial resistance by repressing auxin signaling. Science. 2006; 312(5772): 436–9. PubMed Abstract \| Publisher Full Text\n\nAddo-Quaye C, Miller W, Axtell MJ: CleaveLand: a pipeline for using degradome data to find cleaved small RNA targets. Bioinformatics. 2009; 25(1): 130–1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGerman MA, Pillay M, Jeong DH, et al.: Global identification of microRNA-target RNA pairs by parallel analysis of RNA ends. Nat Biotechnol. 2008; 26(8): 941–6. PubMed Abstract \| Publisher Full Text\n\nShamimuzzaman M, Vodkin L: Identification of soybean seed developmental stage-specific and tissue-specific miRNA targets by degradome sequencing. BMC Genomics. 2012; 13: 310. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmutz J, Cannon SB, Schlueter J, et al.: Genome sequence of the palaeopolyploid soybean. Nature. 2010; 463: 178–83. PubMed Abstract \| Publisher Full Text\n\nSubramanian S, Fu Y, Sunkar R, et al.: Novel and nodulation-regulated microRNAs in soybean roots. BMC Genomics. 2008; 9: 160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLullien V, Barker DG, de Lajudie P, et al.: Plant gene expression in effective and ineffective root nodules of alfalfa (Medicago sativa). Plant Mol Biol. 1987; 9(5): 469–78. PubMed Abstract \| Publisher Full Text\n\nVincent JM: A manual for the practical study of root nodule bacteria. IBP handbook Blackwell Scientific Publications, Oxford, United Kingdom. 1970; (15). Reference Source\n\nGraham TL, Graham MY, Subramanian S, et al.: RNAi silencing of genes for elicitation or biosynthesis of 5-deoxyisoflavonoids suppresses race-specific resistance and hypersensitive cell death in Phytophthora sojae infected tissues. Plant Physiol. 2007; 144(2): 728–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCollier R, Fuchs B, Walter N, et al.: Ex vitro composite plants: an inexpensive, rapid method for root biology. Plant J. 2005; 43(3): 449–57. PubMed Abstract \| Publisher Full Text\n\nTurner M, Adhikari S, Subramanian S: Optimizing stem-loop qPCR assays through multiplexed cDNA synthesis of U6 and miRNAs. Plant Signal Behav. 2013; 8(8). PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVarkonyi-Gasic E, Hellens RP: Quantitative stem-loop RT-PCR for detection of microRNAs. Methods Mol Biol. 2011; 744: 145–57. PubMed Abstract \| Publisher Full Text\n\nAdhikari S, Turner M, Subramanian S: Hairpin priming is better suited than in vitro polyadenylation to generate cDNA for plant miRNA qPCR. Mol Plant. 2013; 6(1): 229–31. PubMed Abstract \| Publisher Full Text\n\nLivak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001; 25(4): 402–8. PubMed Abstract \| Publisher Full Text\n\nRapid amplification of 5’ complementary DNA ends (5’ RACE). Nat Methods. 2005; 2(8): 629–30. PubMed Abstract \| Publisher Full Text\n\nMao G, Turner M, Yu O, et al.: miR393 and miR164 influence indeterminate but not determinate nodule development. Plant Signaling Behav. 2013; 8(10). PubMed Abstract \| Publisher Full Text\n\nVazquez F, Gasciolli V, Crete P, et al.: The nuclear dsRNA binding protein HYL1 is required for microRNA accumulation and plant development, but not posttranscriptional transgene silencing. Curr Biol. 2004; 14(4): 346–51. PubMed Abstract \| Publisher Full Text\n\nGuo HS, Xie Q, Fei JF, et al.: MicroRNA directs mRNA cleavage of the transcription factor NAC1 to downregulate auxin signals for arabidopsis lateral root development. Plant Cell. 2005; 17(5): 1376–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhao Y, Yu Y, Zhai J, et al.: The Arabidopsis nucleotidyl transferase HESO1 uridylates unmethylated small RNAs to trigger their degradation. Curr Biol. 2012; 22(8): 689–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTurner M, Yu O, Subramanian S: Genome organization and characteristics of soybean microRNAs. BMC Genomics. 2012; 13: 169. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarlsbecker A, Lee JY, Roberts CJ, et al.: Cell signalling by microRNA165/6 directs gene dose-dependent root cell fate. Nature. 2010; 465(7296): 316–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu PP, Montgomery TA, Fahlgren N, et al.: Repression of AUXIN RESPONSE FACTOR10 by microRNA160 is critical for seed germination and post-germination stages. Plant J. 2007; 52(1): 133–46. PubMed Abstract \| Publisher Full Text\n\nLi J, Reichel M, Millar AA: Determinants beyond both complementarity and cleavage govern microR159 efficacy in Arabidopsis. PLoS Genet. 2014; 10(3): e1004232. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUdvardi MK, Czechowski T, Scheible WR: Eleven golden rules of quantitative RT-PCR. Plant Cell. 2008; 20(7): 1736–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDamodaran S, Adhikari S, Turner M, et al.: Raw data qACE to assay miRNA-directed target cleavage. F1000Research. 2014. Data Source" }	[ { "id": "6570", "date": "19 Nov 2014", "name": "Rebecca Schwab", "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\nTechnical comments: The ‘proof’ that the method can reliably quantify miRNA target cleavage products is based on inferences from previous reports - that increased cleavage is expected when the corresponding miRNA is over-expressed, and that reduced cleavage is to be seen in miRNA biosynthetic mutants such as hen1. An adequate validation should however include an independent methodology, such as regular northern blots for miRNA targets. This is easily done when choosing targets that are reasonably stable. Quantification of the resulting bands (full length and cleaved) will serve as an independent proof of concept. It would be helpful to add a couple of words during the results section about the design of qACE primers and to what extent they require optimisation (overlap adapter/target). Sort of a guideline. Also, the RCT value should be explained a bit better during the results section. Stability of miRNA target cleavage products is an issue - sometimes they are easily detected, sometimes they are not. This is mentioned in the discussion, but should come up before as it is a major point of consideration for the design of the experiments. Materials and Methods:What tissues were used for RT experiments? Which primers were used to over-express miR164 in soybean?Other:I somewhat disagree with the statement ‘…quantification of the extent of regulation of specific target genes by miRNA is crucial for proper understanding of their biological roles.’ To investigate the biological roles of miRNAs, quantifying (and spatially/temporally visualising) target protein levels is very important. Quantifying the extent of miRNA-directed cleavage contributing to target protein levels are however very interesting for mechanistic studies of miRNA-mediated gene regulation. Sentences are sometimes phrased very generally. Since miRNA-directed cleavage is not a main issue in animals, ‘plants’ can come up already in the beginning of sentences.", "responses": [] }, { "id": "8479", "date": "27 Apr 2015", "name": "Patrick Masson", "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 a method that allows quantification of miRNA-directed cleavage of specific target RNAs in plant tissues. Method validation includes an analysis of target cleavage in miRNA over-expressing tissues and in miRNA-deficient samples (hen1 mutant). Usefulness of the approach is illustrated by the identification of a role for miRNA-directed cleavage in the modulation of nodule-specific GmNAC1-isoform expression in Glycine max.This quantitative protocol may reveal itself as very useful in experiments aimed at characterizing the role played by miRNAs in the regulation of target gene expression in plants and their contribution to growth, development and response to the environment. However, under its current form, this manuscript suffers from a lack of independent validation of its quantitative output. Such validation should involve a different approach aimed at quantifying the relative levels of miRNA, full-length and cleaved target transcripts in wild type, overexpressing and miRNA-deficient mutants. For instance, a Northern blot analysis of wild-type, miR164-over-expressing and hen1 plant tissues could be carried out to evaluate relative levels of miR164 as well as full-length AtNAC1 and its miR164-induced cleavage product, and the results compared to those obtained through the qRT-PCR and qAce approaches described in this manuscript. Indeed, we know from the publication by Guo et al. (2005) (referenced in this manuscript) that these various RNAs can be identified and quantified by Northern blot analysis", "responses": [] }, { "id": "8620", "date": "12 May 2015", "name": "Anthony A. Millar", "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\nHere, the authors describe an assay that quantifies miRNA-guided cleavage products. Then, by comparing to the mRNA levels of the corresponding non-cleaved target transcripts, a ratio is obtained that estimates the proportion of transcript present in the cell has been cleaved by the miRNA. They tested this in an overexpression line, a miRNA-deficient mutant, and in an example where an endogenous miRNA is expressed in a tissue specific manner. The discussion nicely points out the advantages and limitations of the approach. Although the assay represents a relatively straight forward modification of a widely used protocol, modifying it into a quantitative-PCR protocol may enable a number of interesting questions to be addressed. For example, it may give an insight into the efficiency of miRNA-mediated regulation, where it could determine if all equally predicted targets are cleaved as efficiently as one another. Minor comments The authors state in the abstract that qACE can “determine the extent of miRNA regulation for a specific target gene”. This method only measures transcript cleavage, but nothing can be inferred about any translational inhibition mechanism, which there is now very strong evidence that many (possibly most) miRNA-target interactions have a translational repression component to it. Maybe this should be mentioned somewhere in the manuscript as it has basically been ignored, and so this method can only measure the efficiency of cleavage but not the “extent of miRNA regulation”. Also in abstract the term “differential cleavage by miR164” makes it sound like the miRNA is cleaving differently in the different tissues, this may turn out to be the case in some examples, but here isn’t more likely due to the fact that miR164 is differentially expressed, having a higher abundance in root tissues compared to nodules? I would be careful with the use of that term. In the discussion; “we demonstrate that both NAC1 and CUC2 had similar levels of increase in hen1-1, but different levels of reduction in miR164-directed cleavage.” Maybe the authors should be more careful with their claims, as qPCR measurements can vary, and the number of biological reps were not indicated, they did mention “replicate assays” but this seems something different, so how repeatable was this difference?", "responses": [] } ]	1	https://f1000research.com/articles/3-240
https://f1000research.com/articles/4-176/v1	01 Jul 15	{ "type": "Review", "title": "Mitochondrial DNA Damage and Diseases", "authors": [ "Gyanesh Singh", "U C Pachouri", "Devika Chanu Khaidem", "Aman Kundu", "Chirag Chopra", "Pushplata Singh", "U C Pachouri", "Devika Chanu Khaidem", "Aman Kundu", "Chirag Chopra", "Pushplata Singh" ], "abstract": "Various endogenous and environmental factors can cause mitochondrial DNA (mtDNA) damage. One of the reasons for enhanced mtDNA damage could be its proximity to the source of oxidants, and lack of histone-like protective proteins. Moreover, mitochondria contain inadequate DNA repair pathways, and, diminished DNA repair capacity may be one of the factors responsible for high mutation frequency of the mtDNA. mtDNA damage might cause impaired mitochondrial function, and, unrepaired mtDNA damage has been frequently linked with several diseases. Exploration of mitochondrial perspective of diseases might lead to a better understanding of several diseases, and will certainly open new avenues for detection, cure, and prevention of ailments.", "keywords": [ "Oxidative damage", "mitochondrial DNA repair", "mitochondrial pathology", "mitochondrial mutations" ], "content": "Introduction\n\nMitochondria, a key organelle of most eukaryotic cells, are not only essential for cellular energy generation but also important for calcium metabolism and apoptotic cell-signaling1. Like the nucleus, both mitochondria and chloroplasts contain their own DNA, and mitochondrial DNA (mtDNA) damage has been frequently implicated in several diseases including neurodegeneration, cancer, stroke, cardiomyopathy, diabetes, and aging-related disorders (2, Figure 1). Unlike nuclear DNA, the mitochondrial genome is circular, contains very few introns, and the number of mtDNA copies in one mitochondrion can be in the range of two to ten. Furthermore, the size of mtDNA is very small (16.6 kb in humans), and mitochondrial codon-usage is also different. The multicopy nature of mtDNA bestows unconventional modes of DNA maintenance such as selective degradation of damaged DNA, and an unusual form of recombination3. mtDNA is maternally inherited, and sperm mitochondria are mostly degraded after fertilization4. Mitochondria synthesize some of its own proteins, and one of the reasons for this could be that all proteins that are translated in cytoplasm might not be able to cross mitochondrial membranes owing to their varied hydrophobicity5. mtDNA encodes 22 tRNAs, 2 rRNAs, and 13 proteins that participate in mitochondrial ATP synthesis6. Reactive oxygen species (ROS) are very reactive oxygen-containing molecules. ROS are produced in all types of cells and can have various harmful effects. mtDNA, like other DNA, can not only be damaged by radiation and genotoxic chemicals but also by ROS that are frequently produced in mitochondria7. mtDNA damage can exaggerate further because of errors during DNA replication, and lack of conventional histone proteins in mitochondria8. ROS can cause various types of oxidative damage including DNA strand breaks, base modification or removal, and cross linking. DNA polymerase γ (pol γ), the only DNA polymerase known to be present in the mitochondria, have low frameshift fidelity, and, is believed to be a major contributor to changes in mtDNA9.\n\n\nConsequences of mitochondrial DNA damage\n\nSeveral studies report the effect of genotoxic agents on mitochondria10,11. However, it is not easy to draw conclusions in these cases, as agents that damage mtDNA also damage nuclear DNA. Therefore, it is suggested that all studies should compare consequences of nuclear and mtDNA damage in such cases, as far as possible. Other than its involvement in cancer and neurological disorders, changes in mtDNA have been shown to be associated with a few hereditary diseases12. mtDNA damage is well known to cause impaired bioenergetics, reduced cell proliferation and apoptosis, hypercholesterolemia, and atherosclerosis12. Interestingly, mtDNA defects are known to cause defective mitochondrial ATP generation that results into compromised organ function and diseases13.\n\nIn case of the most common neurodegenerative disorders including Parkinson's disease (PD), Alzheimer's disease (AD), and amyotrophic lateral sclerosis (ALS) also, mtDNA damage has been implicated as a factor that cause or exaggerate these diseases14. Brain tissues from Alzheimer's patients show greater fragmentation mtDNA. However, similar damage to nuclear DNA is controversial in this case. Increased mtDNA damage was also associated with reduced levels of mitochondrial protein expression13. Interestingly, brain tissues from Alzheimer's patients show higher levels of oxidized bases. In this case, mtDNA was found to have 10-times more oxidized bases compared to nuclear DNA indicating that mtDNA is more succeptible to oxidants14.\n\nIn the case of Huntington disease (HD), higher levels of oxidative stress were observed in the brain tissues of both humans and mice16. In the case of a mouse model of HD, embryonic fibroblasts showed increased mitochondrial matrix Ca2+ loading, and higher superoxide generation. This confirmed that both mitochondrial Ca2+ signaling and superoxide generation are dysregulated in HD, and, reducing mitochondrial Ca2+ uptake can be a therapeutic strategy for HD16. Peripheral blood mononuclear cells (PBMCs) from systemic lupus erythematosus (SLE) patients also exhibited enhanced mtDNA damage indicating potential role of mitochondria in the pathogenesis of SLE17. Apolipoprotein E (ApoE) is known to play a protective role in preventing artery wall thickening in atherosclerosis and ApoE-/- mice show mtDNA damage before significant atherosclerosis18. pol γ-/-/ApoE-/- mice show extensive mtDNA damage, impaired mitochondrial respiration, and increased atherosclerosis, even without increased ROS. Furthermore, pol γ-/-/ApoE-/- monocytes showed increased inflammatory cytokine release18. Aging is often associated with the accumulation of deleterious changes, reduced physiological functions, and increased likelihood of diseases19. In this context, a number of mitochondrial aberrations have been observed with aging. These aberrations are accumulation of mtDNA mutation, inefficient oxidative phosphorylation, increased production of ROS, and disorganized mitochondrial structure20. These mtDNA mutations are often somatic, with variable changes in individual cells. Often, higher levels of these mutations are associated with respiratory chain deficiency. A mosaic pattern of respiratory chain deficiency can be found in different tissues because of uneven distribution of mutations13. The mitochondrial free radical theory of aging has been one of the most supported ideas of aging19. This theory postulates that the production of intracellular ROS is the major determinant during aging. Several invertebrate and mammalian models already support this hypothesis. Oxidative stress, when propagated by active radicals, can damage DNA, phospholipids, proteins and other biomolecules. Reactive oxygen species mediated mtDNA damage can occur directly at the sugar-phosphate backbone, at the bases, or in the form of single and double strand breaks20. Unfortunately, most of the antioxidant-supplementation regimens do not increase longevity, as predicted by the free radical theory of aging. Intracellular ROS are generated in multiple compartments and by multiple pathways. Important contributors in this case are NADPH oxidases, cyclooxygenases, and lipid metabolism enzymes21. Despite several non-mitochondrial contributors, almost 90% of cellular ROS are still generated in mitochondria. In some cases, long-lived species were not only found to produce less ROS but also showed less oxidative damage22. Similarly, various animal and human studies suggest that the decline in muscle mitochondria is a leading factor for muscular abnormalities23.\n\nAged monkeys showed enhanced DNA damage and reduced transcription of mtDNA compared to young ones24. D-gal-induced aging rats are important animal model of aging, and the level of mtDNA deletions was found to be significantly more in the hippocampus of D-gal-treated rats compared to controls25. NADPH oxidase (NOX) generates ROS while transporting electrons across the mitochondrial membrane. Similarly, uncoupling protein 2 (UCP2) transports anions and protons across the mitochondrial membrane, and also controls ROS generation. In case of D-gal-induced animal model of aging, damaged mitochondrial ultrastructure was seen in the hippocampus region along with increased production of NOX and UCP2. Nicotinamide adenine dinucleotide (NAD+) is a key electron transporter in mitochondria. NAD+ depletion may play a prominent role in the aging process, not only by limiting energy production, but also by compromising DNA repair and genomic signaling as NAD+ is an important substrate for the nuclear repair enzymes21. Poly(ADP-ribose) polymerase (PARP) controls inflammatory immune responses, and hyperactivation of PARP-1 is known to activate mitochondrial pathway of apoptosis26. Age-associated increase in oxidative nuclear damage was found to be associated with PARP-induced NAD+ depletion and absence of SIRT1 activity in rodents26. Ercc1 mutant mice, which are deficient in DNA repair pathways, show accelerated aging and progressive memory loss27. Defective oxidative phosphorylation, mutated mtDNA, or mitochondrial ROS have also been documented in cases of tumorigenesis28. Oxidative stress in the cardiovascular system is known to cause accumulation of reactive oxygen and nitrogen species, which increase leukocyte adhesion and endothelial permeability29. NFκB is one of the most important transcription factor that is known to be involved in important signaling pathways, development, and several diseases. Hypoxia-Inducible Factor (HIF-1) is a protein that not only protects from hypoxia-induced damage, but is also important for smooth functioning of immune system and key metabolic pathways. In an interesting study, ROS, NFκB- and HIF1-activation in the tumor microenvironment induced accelerated aging in rodents, which subsequently caused stromal inflammation and altered cancer cell metabolism30. Certain dietary treatments or enrichment of mitochondrial membranes with oxidant-resistant fatty acids were found to increase life span in rodents31. Monounsaturated-fatty-acid-rich diet prevented the accelerated mtDNA mutations in the brain mitochondria from aged animals. Therefore, changes in mtDNA that gradually accumulate in a variety of tissues during aging appear to be involved in onset of various diseases32 and a better understanding of mitochondrial biology is required in this perspective. mtDNA ligase is essential for cell survival particularly because of its role in base excision repair pathway33.\n\n\nConclusions\n\nMitochondria are of central importance in eukaryotic cells. However, mtDNA is more prone to damage, and mtDNA repair pathways are inadequate. Together, these problems might frequently lead to unrepaired mtDNA lesions, and defective energy metabolism. mtDNA damage has been frequently shown to be involved in initiation and progression of several diseases including various types of neurodegenerative disorders, cancer, stroke, heart-diseases, and diabetes. There is an urgent need for detailed investigation in this area, to find out the mitochondrial contribution to various diseases, so that improved prevention measures and cures can be developed.", "appendix": "Author contributions\n\n\n\nGS and PS conceived the study and prepared the first draft of the manuscript. UCP and CC did the analysis of the literature. DCK and AK did cross-checking and referencing.\n\n\nCompeting 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\nReferences\n\nMilane L, Trivedi M, Singh A, et al.: Mitochondrial biology, targets, and drug delivery. J Control Release. 2015; 207: 40–58. PubMed Abstract \| Publisher Full Text\n\nMishra P, Chan DC: Mitochondrial dynamics and inheritance during cell division, development and disease. Nat Rev Mol Cell Biol. 2014; 15(10): 634–646. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaniak-Golik A, Skoneczna A: Mitochondria-nucleus network for genome stability. Free Radic Biol Med. 2015; 82: 73–104. PubMed Abstract \| Publisher Full Text\n\nAanen DK, Spelbrink JN, Beekman M: What cost mitochondria? The maintenance of functional mitochondrial DNA within and across generations. Philos Trans R Soc Lond B Biol Sci. 2014; 369(1646): 20130438. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReithinger JH, Yim C, Park K, et al.: A short C-terminal tail prevents mis-targeting of hydrophobic mitochondrial membrane proteins to the ER. FEBS Lett. 2013; 587(21): 3480–3486. PubMed Abstract \| Publisher Full Text\n\nSuzuki T, Nagao A, Suzuki T: Human mitochondrial tRNAs: biogenesis, function, structural aspects, and diseases. Annu Rev Genet. 2011; 45: 299–329. PubMed Abstract \| Publisher Full Text\n\nSavu O, Sunkari VG, Botusan IR, et al.: Stability of mitochondrial DNA against reactive oxygen species (ROS) generated in diabetes. Diabetes Metab Res Rev. 2011; 27(5): 470–479. PubMed Abstract \| Publisher Full Text\n\nFurda AM, Marrangoni AM, Lokshin A, et al.: Oxidants and not alkylating agents induce rapid mtDNA loss and mitochondrial dysfunction. DNA Repair (Amst). 2012; 11(8): 684–692. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcKinney EA, Oliveira MT: Replicating animal mitochondrial DNA. Genet Mol Biol. 2013; 36(3): 308–315. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWisnovsky SP, Wilson JJ, Radford RJ, et al.: Targeting mitochondrial DNA with a platinum-based anticancer agent. Chem Biol. 2013; 20(11): 1323–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBellance N, Lestienne P, Rossignol R, et al.: Mitochondria: from bioenergetics to the metabolic regulation of carcinogenesis. Front Biosci (Landmark Ed). 2009; 14: 4015–34. PubMed Abstract \| Publisher Full Text\n\nGreaves LC, Reeve AK, Taylor RW, et al.: Mitochondrial DNA and disease. J Pathol. 2012; 226(2): 274–286. PubMed Abstract \| Publisher Full Text\n\nBoczonadi V, Horvath R: Mitochondria: impaired mitochondrial translation in human disease. Int J Biochem Cell Biol. 2014; 48: 77–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartin LJ: Biology of mitochondria in neurodegenerative diseases. Prog Mol Biol Transl Sci. 2012; 107: 355–415. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCassereau J, Codron P, Funalot B, et al.: Inherited peripheral neuropathies due to mitochondrial disorders. Rev Neurol (Paris). 2014; 170(5): 366–374. PubMed Abstract \| Publisher Full Text\n\nMena NP, Urrutia PJ, Lourido F, et al.: Mitochondrial iron homeostasis and its dysfunctions in neurodegenerative disorders. Mitochondrion. 2015; 21: 92–105. PubMed Abstract \| Publisher Full Text\n\nFernandez D, Bonilla E, Phillips P, et al.: Signaling abnormalities in systemic lupus erythematosus as potential drug targets. Endocr Metab Immune Disord Drug Targets. 2006; 6(4): 305–311. PubMed Abstract \| Publisher Full Text\n\nFriedland-Leuner K, Stockburger C, Denzer I, et al.: Mitochondrial dysfunction: cause and consequence of Alzheimer's disease. Prog Mol Biol Transl Sci. 2014; 127: 183–210. PubMed Abstract \| Publisher Full Text\n\nVendelbo MH, Nair KS: Mitochondrial longevity pathways. Biochim Biophys Acta. 2011; 1813(4): 634–644. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGomez-Cabrera MC, Sanchis-Gomar F, Garcia-Valles R, et al.: Mitochondria as sources and targets of damage in cellular aging. Clin Chem Lab Med. 2012; 50(8): 1287–1295. PubMed Abstract \| Publisher Full Text\n\nMassudi H, Grant R, Braidy N, et al.: Age-associated changes in oxidative stress and NAD+ metabolism in human tissue. PLoS One. 2012; 7(7): e42357. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPamplona R, Barja G: Highly resistant macromolecular components and low rate of generation of endogenous damage: two key traits of longevity. Ageing Res Rev. 2007; 6(3): 189–210. PubMed Abstract \| Publisher Full Text\n\nHepple RT: Mitochondrial involvement and impact in aging skeletal muscle. Front Aging Neurosci. 2014; 6: 211. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMao P, Gallagher P, Nedungadi S, et al.: Mitochondrial DNA deletions and differential mitochondrial DNA content in Rhesus monkeys: implications for aging. Biochim Biophys Acta. 2012; 1822(2): 111–119. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDu Z, Hu Y, Yang Y, et al.: NADPH oxidase-dependent oxidative stress and mitochondrial damage in hippocampus of D-galactose-induced aging rats. J Huazhong Univ Sci Technolog Med Sci. 2012; 32(4): 466–472. PubMed Abstract \| Publisher Full Text\n\nBraidy N, Guillemin GJ, Mansour H, et al.: Age related changes in NAD+ metabolism oxidative stress and Sirt1 activity in wistar rats. PLoS One. 2011; 6(4): e19194. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVégh MJ, de Waard MC, van der Pluijm I, et al.: Synaptic proteome changes in a DNA repair deficient ercc1 mouse model of accelerated aging. J Proteome Res. 2012; 11(3): 1855–1867. PubMed Abstract \| Publisher Full Text\n\nLisanti MP, Martinez-Outschoorn UE, Pavlides S, et al.: Accelerated aging in the tumor microenvironment: connecting aging, inflammation and cancer metabolism with personalized medicine. Cell Cycle. 2011; 10(13): 2059–2063. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBonomini F, Rodella LF, Rezzani R: Metabolic syndrome, aging and involvement of oxidative stress. Aging Dis. 2015; 6(2): 109–120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMercier I, Camacho J, Titchen K, et al.: Caveolin-1 and accelerated host aging in the breast tumor microenvironment: chemoprevention with rapamycin, an mTOR inhibitor and anti-aging drug. Am J Pathol. 2012; 181(1): 278–293. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOchoa JJ, Pamplona R, Ramirez-Tortosa MC, et al.: Age-related changes in brain mitochondrial DNA deletion and oxidative stress are differentially modulated by dietary fat type and coenzyme Q10. Free Radic Biol Med. 2011; 50(9): 1053–1064. PubMed Abstract \| Publisher Full Text\n\nVaccaro JA, Huffman FG: Monounsaturated fatty acid, carbohydrate intake, and diabetes status are associated with arterial pulse pressure. Nutr J. 2011; 10: 126. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoriyama T, Sato N: Enzymes involved in organellar DNA replication in photosynthetic eukaryotes. Front Plant Sci. 2014; 5: 480. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "11088", "date": "07 Dec 2015", "name": "Linda Bergersen", "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 short review by Gyanesh Singh et al. provides useful and up-to-date information on the roles of mtDNA in disease. Abnormalities in mtDNA may affect all organs of the body, but cause symptoms primarily in tissues that are dependent on high energy production. Deficient mtDNA maintenance contributes to conditions as diverse as normal aging, neurodegenerative disease, diabetes, cardiovascular disease, and cancer.The authors should give more precise and explicit reference to the repeated assertion that \"mtDNA repair pathways are inadequate\" (e.g. in the Abstract and Conclusions sections), or moderate these statements. The statement reads as indicating that even when the mtDNA repair mechanisms function normally, they are inadequate. Is there direct evidence for this?On p2, the authors correctly point out that it is difficult to distinguish the effects of genotoxic agents on mtDNA, \"as agents that damage mtDNA also damage nuclear DNA\". The authors should reference work on transgenic animal models with damage specifically in mtDNA (e.g., Trifunovic A et al 2004 Nature; Lewis W et al 2007 Lab Invest; Lauritzen KH et al 2010 Mol Cell Biol).Typography: in \"NAD+\" the \"+\" should be corrected to superscript.Please check the text for grammatical errors, e.g., on p2 \"...the only DNA polymerase known to be present in the mitochondria, havelow frameshift fidelity, and, is believed to...\": \"have\" should be corrected to \"has\".", "responses": [] }, { "id": "14777", "date": "05 Jul 2016", "name": "George Shinomol", "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 article (Review) titled \"Mitochondrial DNA damage and diseases\" is a good work. However there are certain shortcomings that the authors need to address.\nIn general there is much refinement needed for language\nThe abstract lacks clarity and it doesn’t actually represent the entire article. It needs to be re written\n\nThe paper is unorganised .The consequences of mitochondrial damage could have been described under various appropriate subheadings, e.g.: Neurodegenrative diseases, Aging, Cardiovascular diseases etc. This could have increased the clarity and quality of this review\n\nThere also could have been mentioned some details on amelioration of the effects using mitochondria targeted therapies\n\nInclude recent references", "responses": [] }, { "id": "14538", "date": "26 Jul 2016", "name": "Ullas Kolthur-Seetharam", "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 review titled “Mitochondrial damage and diseases” has attempted to give an overview of the current literature on the causes of mtDNA damage and its effect in disease and aging. Although the authors seem to have cited a lot of literature in this regard, they have not been able to present their views in a collected manner.\nThe language used is extremely colloquial (e.g. Page 2 “Furthermore, the size of mtDNA is very small (16.6 kb in humans), and mitochondrial codon-usage is also different” and “Reactive oxygen species (ROS) are very reactive oxygen-containing molecules”) and disjointed. In the last paragraph, it is not clear why the authors are suddenly talking about nuclear DNA damage as the review concentrates on mtDNA damage.\nThe authors move from one point to another without clarity of thought and without drawing relevant conclusions supported by multiple references. References are missing in multiple places. There also seem to be some cases of mis-referencing.\nThe authors could discuss a little more in their introduction about the DNA repair pathways that do exist in the mitochondria. From their text, it seems that any damage induced in the mitochondria is never repaired. The authors should also discuss the current techniques that are available for rectifying mtDNA damage like genome editing. This is particularly important as they stress this in their conclusion. Also, they should speculate on the effect of DNA damage, NAD loss, and accelerated aging based on previous literature.", "responses": [] } ]	1	https://f1000research.com/articles/4-176
https://f1000research.com/articles/4-175/v1	30 Jun 15	{ "type": "Case Report", "title": "Case Report: Stevens-Johnson syndrome following a single double dosing of nevirapine-containing regimen once in an HIV-infected woman on long-term antiretroviral therapy.", "authors": [ "Betty Kakande", "Thuraya Isaacs", "Rudzani Muloiwa", "Sipho Dlamini", "Rannakoe Lehloenya", "Betty Kakande", "Thuraya Isaacs", "Rudzani Muloiwa", "Sipho Dlamini" ], "abstract": "A 31-year old HIV-infected African woman on nevirapine, tenofovir and lamivudine for more than 4 years presented with an 8-day history of symptoms and signs of Stevens-Johnson syndrome. She was on no other medication. Her viral load was undetectable and she had maintained a CD4 count of between 356 and 387cells/mm3 in the preceding 2½ years. She missed her antiretrovirals 10 days before the onset of her symptoms and subsequently doubled her daily dose the following day. She had been on no other medication in the preceding 8 weeks.Her ARVs were stopped and she fully re-epithelialized with the exception of the lips, over the following 10 days. She was started on a daily single tablet of Odimune® (a fixed drug combination antiretroviral containing tenofovir, emtricitabine and efavirenz). Nevirapine is the most common offender in cases of antiretroviral-associated SJS in published literature. Lamivudine is very rarely implicated while there are no similar reports with tenofovir. We concluded that nevirapine was by far the most likely offender in this case. Nevirapine toxicity is associated with high CD4 counts, undetectable viral load and high drug plasma level. We postulate that the sudden increase of the plasma levels of nevirapine in a patient with a high CD4 count and undetectable viral load created a perfect storm for the development of SJS in our patient, who had been on the NVP-containing regimen for many years.Clinicians should be aware that severe adverse drug reactions are dynamic and can occur even when the drug has been in use for a long time.", "keywords": [ "HIV", "nevirapine", "double dose", "Stevens-Johnson syndrome" ], "content": "Background\n\nNevirapine (NVP)-based antiretroviral regimens have been used widely in developing countries because of its affordability, availability and efficacy1. NVP is also used to minimize diarrhea and cardiovascular side effects of protease inhibitors and neuropsychiatric side effects of Efavirenz2. However, NVP is associated with severe adverse reactions, including Stevens-Johnson syndrome (SJS).\n\nIn larger series, the incubation period before developing features of SJS ranges from 10–240 days with a median duration of around 12 days3. We report a case of probable nevirapine-associated Stevens-Johnson syndrome occurring more than 4 years after initiation of the drug. This case illustrates that in susceptible persons, a severe drug reaction can occur when there are dose adjustments despite having been exposed to the same drug for a long time.\n\n\nCase report\n\nA 31-year-old HIV-infected black African female presented with an 8-day history of painful swallowing, sore eyes, malaise and a worsening rash. She had been on antiretroviral (ARV) regimen of nevirapine (NVP) 400 mg daily, tenofovir 300 mg daily and lamivudine 300 mg daily for more than 4 years uneventfully. She was on no other medications and had not taken any other medication in the preceding 8 weeks. She had acquired HIV via heterosexual contact – the exact date of HIV infection was unknown. At the time of initiation of ARV therapy, her nadir CD4 cell count was 139 cells/mm3 and her HIV RNA viral load at the time was unknown. Her last CD4 counts, done 31, 21 and 11 months before developing her current symptoms were 373, 356 and 387 cells/mm3 respectively. The last HIV RNA viral load test, done 11 months prior to the onset of her symptoms showed an undetectable viral load.\n\nShe gave a history of forgetting to take her ARV medication for a day. The following day, 10 days before development of her symptoms she took 2 days equivalent of her ARVs in one day, in her own words “to make up for the missed dose”.\n\nOn examination, she was normotensive and had a temperature of 38.8°C. She had conjunctivitis and hemorrhagic cheilitis but no involvement of the genital mucosa. She had epidermal necrosis involving predominantly her trunk and face and to a lesser extent palms, soles and extremities, totaling 10% of her body surface area, 3% of which was stripping. Initial laboratory studies showed normal blood count, except for an elevated eosinophil count of 0.70 × 109/L. The liver and renal function tests were normal. Her ARVs were stopped and her epidermal necrosis did not extend and eosinophil counts normalized. Over the next few days her skin condition improved and she was discharged 10 days later fully re-epithelialized with the exception of the lips, which were still eroded in areas. On review, 2 weeks after discharge, her skin had normalized except for residual hyperpigmentation. She was started on a single daily tablet of Odimune®, a fixed drug combination of tenofovir 300 mg, emtricitabine 200 mg and efavirenz 600 mg. On follow-up, 4 weeks later she was tolerating the new ARV regimen.\n\n\nDiscussion\n\nClinicians routinely blame NVP as the culprit agent in cases of ARV-associated SJS as it is by far the most common offender in published literature. However, lamivudine has infrequently been reported as a cause of SJS, but there have not been similar reports with tenofovir4. Based on these, we concluded that NVP was by far the most likely offender in our case.\n\nNumerous factors have been associated with NVP toxicity, including high current and nadir CD4 counts, undetectable viral load, female sex, abnormally elevated baseline transaminases, history of drug allergy, lower body weight and high drug plasma levels2,5,6. We postulate that the sudden increase of the plasma levels of NVP in a patient with a high CD4 count and undetectable viral load created a perfect storm for the development of SJS in our patient, who had been on the NVP-containing regimen for many years. The impact of NVP drug concentrations on the risk of adverse drug reactions is still controversial. Some studies have shown that above a certain threshold, the risk of severe drug reactions is significantly increased. A recent study has shown that impaired renal function, thus impaired clearance of the drug, is associated with an odds ratio of 8.0 (3.9 to 17) to develop allopurinol-associated SJS in susceptible people7. However other studies have not found this association6,8,9. NVP has a long half-life, 25–30 hours, and the effect of doubling the dose possibly increased the plasma drug concentrations considerably resulting in SJS8.\n\nThe patient had been compliant on treatment for more than 4 years and had undetectable viral load and a high CD4 count the last time they were measured, 11 months earlier. Both of these are independent predictors of NVP-associated SJS2,10. There were previous suggestions that the timing of SJS and other NVP-associated reactions coincided with the increase of the drug dose. However, this has been disproved, because as high as a third of cases in larger series occurred during the lead-in period of the dosing regimen before the dose was increased3,11. Pretreatment with ARVs longer than 1 year has also been reported as an independent predictor of a cutaneous drug reaction. Whether this is truly an independent predictor or a result of low HIV viral load or higher CD4 counts is unclear and further studies are needed to answer this question.\n\nIn summary, this case highlights the potentially life-threatening risk of the development of SJS outside the expected period, probably triggered by a spike in serum levels of NVP. The case also illustrates that development of severe drug eruptions, like SJS, is a dynamic processes that may evolve with changing environment in the body. Clinicians should warn their patients against unregulated dosing changes of their NVP-containing regimen.\n\n\nConsent\n\nWritten informed consent for publication of the clinical details was obtained from the patient.", "appendix": "Author contributions\n\n\n\nBK, TI, RJL, and SD managed the patient. RJL, BK, RM and SD contributed to the preparation of the manuscript and revision of the draft manuscript. TI revised the final manuscript. All authors have agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors have no competing interests to declare.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nTansuphaswadikul S, Aung SE, Phonrat B, et al.: Predisposing factors for nevirapine toxicity among AIDS patients with low baseline CD4 count. Asian Pac J Allergy Immunol. 2007; 25(2–3): 147–154. PubMed Abstract\n\nKesselring AM, Wit FW, Sabin CA, et al.: Risk factors for treatment-limiting toxicities in patients starting nevirapine-containing antiretroviral therapy. AIDS. 2009; 23(13): 1689–1699. PubMed Abstract \| Publisher Full Text\n\nFagot JP, Mockenhaupt M, Bouwes-Bavinck JN, et al.: Nevirapine and the risk of Stevens-Johnson syndrome or toxic epidermal necrolysis. AIDS. 2001; 15(14): 1843–1848. PubMed Abstract \| Publisher Full Text\n\nModak D, Guha SK: Severe skin rash with lamivudine in HIV infected patients: some unusual case reports. Indian J Pharmacol. 2013; 45(3): 298–300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRatanasuwan W, Jariyasethpong T, Anekthananon T, et al.: Association of Nevirapine Levels with Rash or Hepatotoxicity Among HIV-Infected Thai Women. Open AIDS J. 2012; 6: 266–273. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Maat MM, ter Heine R, Mulder JW, et al.: Incidence and risk factors for nevirapine-associated rash. Eur J Clin Pharmacol. 2003; 59(5–6): 457–462. PubMed Abstract \| Publisher Full Text\n\nChung WH, Chang WC, Stocker SL, et al.: Insights into the poor prognosis of allopurinol-induced severe cutaneous adverse reactions: the impact of renal insufficiency, high plasma levels of oxypurinol and granulysin. Ann Rheum Dis. 2014. PubMed Abstract \| Publisher Full Text\n\nKappelhoff BS, van Leth F, Robinson PA, et al.: Are adverse events of nevirapine and efavirenz related to plasma concentrations? Antivir Ther. 2005; 10(4): 489–498. PubMed Abstract\n\nDong BJ, Zheng Y, Hughes MD, et al.: Nevirapine pharmacokinetics and risk of rash and hepatitis among HIV-infected sub-Saharan African women. AIDS. 2012; 26(7): 833–841. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKiertiburanakul S, Sungkanuparph S, Charoenyingwattana A, et al.: Risk factors for nevirapine-associated rash among HIV-infected patients with low CD4 cell counts in resource-limited settings. Curr HIV Res. 2008; 6(1): 65–69. PubMed Abstract \| Publisher Full Text\n\nDziuban EJ, Hughey AB, Stewart DA, et al.: Stevens-Johnson syndrome and HIV in children in Swaziland. Pediatr Infect Dis J. 2013; 32(12): 1354–1358. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9555", "date": "20 Jul 2015", "name": "Neil Shear", "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 Title, Abstract and all content of the article are clear and relevant. This case could seem to be simple but it actually adds a valuable perspective to our understanding of what might trigger severe drug reactions and adds clinical insights and clarity to what is often an overly complex environment.", "responses": [] }, { "id": "10651", "date": "20 Oct 2015", "name": "Carla Ferrándiz-Pulido", "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 an interesting case of Stevens-Johnson syndrome probably due to a single double dosing of nevirapine-containing regimen in an HIV-infected woman. There have been a lot of case reports and short series of SJS/TEN due to antiretroviral drugs. However, the singularity of this case is that the drug had been taken for a long time (4 years), and the triggering was an increase in the dosage. This side effect due to an increase in the dosage has also been reported for allopurinol.Most of Lyell's syndrome/TEN are caused by drugs; however, SJS may be triggered by a \"new\" drug or also an infection such as VHS/VVZ or Mycoplasma pneumoniae. In my opinion, it is always important to rule out such infections, when a SJS in suspected. This should be included in the article.Some clinical images of the skin and mucosal involvement would improve the quality of the article.", "responses": [] }, { "id": "10650", "date": "06 Nov 2015", "name": "Ronald Kiguba", "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 authors highlight an important report of Stevens-Johnson syndrome (SJS) in a patient on nevirapine-based antiretroviral therapy for more than two years. The median duration to the onset of nevirapine-associated SJS is two weeks, thus the case report stimulates exploration of the relationship between sudden rise in plasma nevirapine concentration and increased risk of SJS. However, it was not clear whether the authors performed any formal adverse drug reaction causality assessment, even though they stated that the reaction was \"probable\".", "responses": [] } ]	1	https://f1000research.com/articles/4-175
https://f1000research.com/articles/4-174/v1	29 Jun 15	{ "type": "Data Note", "title": "High-resolution 7-Tesla fMRI data on the perception of musical genres – an extension to the studyforrest dataset", "authors": [ "Michael Hanke", "Richard Dinga", "Christian Häusler", "J. Swaroop Guntupalli", "Michael Casey", "Falko R. Kaule", "Jörg Stadler", "Richard Dinga", "Christian Häusler", "J. Swaroop Guntupalli", "Michael Casey", "Falko R. Kaule", "Jörg Stadler" ], "abstract": "Here we present an extension to the studyforrest dataset – a versatile resource for studying the behavior of the human brain in situations of real-life complexity (http://studyforrest.org). This release adds more high-resolution, ultra high-field (7 Tesla) functional magnetic resonance imaging (fMRI) data from the same individuals. The twenty participants were repeatedly stimulated with a total of 25 music clips, with and without speech content, from five different genres using a slow event-related paradigm. The data release includes raw fMRI data, as well as precomputed structural alignments for within-subject and group analysis. In addition to fMRI, simultaneously recorded cardiac and respiratory traces, as well the complete implementation of the stimulation paradigm, including stimuli, are provided. An initial quality control analysis reveals distinguishable patterns of response to individual genres throughout a large expanse of areas known to be involved in auditory and speech processing. The present data can be used to, for example, generate encoding models for music perception that can be validated against the previously released fMRI data from stimulation with the “Forrest Gump” audio-movie and its rich musical content. In order to facilitate replicative and derived works, only free and open-source software was utilized.", "keywords": [ "functional magnetic resonance imaging", "music perception", "natural sounds", "7 Tesla", "auditory features" ], "content": "Background\n\nPreviously, we have released a large, high-resolution, 7 Tesla fMRI dataset on the processing of natural auditory stimuli – a two-hour audio movie1. Recently, we have extended this initial release with a detailed annotation of the emotional content of the stimulus2 to broaden the range of research questions that could be addressed with these data. Here we further amend this dataset with additional high-resolution fMRI data from the same participants on the perception of musical genres. We employed a proven paradigm and stimuli that have been previously shown to enable investigation of distributed population codes of musical timbre in bilateral superior temporal cortices3.\n\nThe present data release enables comparative studies of the representation of musical genres (spectrum, timbre, vocal content) with ultra high-field, high resolution fMRI data from a larger sample of participants. In conjuction with the previous data releases, it will also further expand the continuum of research question that can be approached with the joint dataset. For example, the development of encoding models for cortical representations of music in complex auditory stimuli (the audio-movie contains several dozen musical excerpts from a broad range of genres). To this end, we include extracted audio features that represent the time-frequency information of each stimulus in four different views. The views are mapped to different perceptually-motivated scales (mel and decibel scales) and via a decorrelating linear transformation (DCT-II). It is hoped that providing these example features will catalyze discoveries of auditory stimulus codes in neural populations.\n\nLastly, these data can also serve as a public resource for benchmarking algorithms for functional alignment [e.g., 4], or other analyses, and thus, further the availability of resources for the investigation of real-life cognition5.\n\n\nMaterials and methods\n\nAcquisition of the data described herein was part of a previously published study1, and took place in close temporal proximity (no more than a few weeks apart). The participants in this data release are identical to those previously reported. They were fully instructed about the nature of the study and were paid a total of 100 EUR for their participation, which included the previously reported data acquisitions, as well as the one described herein. All data acquisitions were jointly approved by the ethics committee of the Otto-von-Guericke-University of Magdeburg, Germany (approval reference 37/13).\n\nAll stimuli employed in this study are identical to those used in a previous study [for details refer to 3]. They were five natural, stereo, high-quality music stimuli (6 s duration; 44.1 kHz sampling rate) for each of five different musical genres: 1) Ambient, 2) Roots Country 3) Heavy Metal, 4) 50s Rock’n’Roll, and 5) Symphonic (see Figure 1 for details).\n\nEach stimulus was a six second excerpt from the middle of a distinct musical piece. Excerpts were normalized so that their root-mean-square power values were equal, and a 50 ms quarter-sine ramp was applied at the start and end of each excerpt to suppress transients. Most prominent are the differences between music clips with and without vocal components.\n\nThe setup for audio-visual presentation was as previously reported1. Participants listened to the audio using custom-built in-ear headphones, and an LCD projector displayed visual instructions on a rear-projection screen that they saw via a mirror attached to the head coil.\n\nAt the start of each recording session, during the preparatory MR scans, participants listened to a series of longer excerpts of musical pieces and songs from the five different genres. During this phase participants were instructed to request adjustments of the stimulus volume in order to guarantee optimal perception of the stimuli against the noise pedestal emitted by the scanner. There was no overlap between the songs presented in this phase and those used as stimuli in the main experiment.\n\nEight scanning runs followed the initial sound calibration. Each run was started by the participant with a key-press ready signal. There were 25 trials, with five different stimuli (Figure 1) for each of the five genres per run (see Figure 2 for details on the experiment design). At the end of each run participants were given the opportunity for a break of variable length until they indicated readiness for the next run. Most participants started the next run within a minute.\n\n(A) Trial configuration. The start of each trial was synchronized with the MRI volume acquisition trigger. When the trigger was received the permanently displayed white fixation cross turned green, a 6 s music stimulus was presented, and, immediately afterwards, the fixation cross turned white again. Stimulation was followed by a variable delay (minimum delay 4 s). For the five trials of a genre, a 4 s and 8 s delay occurred once, while the remained three trials included a 6 s delay period. Thereby all trials had 4–8 s of uniform stimulation (no audio, white fixation cross) after each musical stimulus. The order of delays was randomized within a run. During trials with an 8 s second delay participants were presented with a yes/no question four seconds after the end of the music stimulus. The content of the question was randomized and asked for particular features of the stimulus that had just ended (e.g., “Was there a female singer?”, “Did the song have a happy melody?”). Participants had to indicate their response by pressing one of two buttons with the index or middle finger of their right hand corresponding to the response alternative presented on the screen. “Yes” was always mapped to the left side (index finger), “No” always to the right side (middle finger). The question had the purpose of keeping the participants attentive to the stimuli and counteract the effect of increasing familiarity across multiple runs. (B) Run configuration. The 25 stimuli were identical across runs and presented exactly once per run. Order of stimulus genres within each run was counter-balanced using De Bruijn cycles6 (alphabet size = 5, counter-balancing level = 2), hence each genre was followed by any other genre equally often and exactly once. Eight unique genre order sequences were generated and used for all participants, while randomizing the order of run sequences across participants. This was done in order to enable the application of the hyperalignment algorithm4. Data acquisition for two participants showed anomalies with respect to this procedure (see Table 2 for details).\n\nStimulus presentation and response logging were implemented using PsychoPy7 running on a computer with the (Neuro)Debian operating system8.\n\nThe acquisition protocol for functional MRI was largely identical with the one previously reported1, hence only differences and key facts are listed here.\n\nImportantly, the same landmark-based procedure for automatic slice positioning that was used to align the scanner field-of-view between acquisition sessions, was used again to align the field-of-view for this acquisition with the one in the previous study1. As the exact same alignment target was used, this led to a very similar field-of-view configuration across acquisitions.\n\nEach acquisition run consisted of 153 volumes (repetition time of 2.0 seconds with no inter-volume gaps).\n\nThe cardiac and respiratory traces were recorded for the full duration of all eight runs. The acquisition setup for physiological was identical with the one previously reported1.\n\n\nDataset content\n\nThe released data comprises raw and pre-processed fMRI data, physiological recordings, behavioral log files, and auditory stimuli (total ≈95 GB). Table 1 provides an overview of the location of individual data components. The following sections briefly describe important properties.\n\nFile paths and descriptions for all available content.\n\nLog files are available as plain text files with comma-separated value markup. All enumerations are zero-based. Each lines represents a trial. Columns for the following information are present: order of run in sequence (run), ID of trial sequence for this run (run_id; see Figure 2B), fMRI volume corresponding to stimulation start (total: volume, in the current run: run_volume), stimulus file name (stim), music genre label (genre), inter-stimulus interval in seconds (delay), flag whether a control question with presented (catch), measured asynchrony between MRI trigger and sound onset in seconds (sound_soa), and time stamp of the corresponding MRI trigger with respect to the start of the experiment in seconds (trigger_ts).\n\nInformation on the stimulus timing is also available in per-subject, per-run, per-condition plain-text files in FSL’s EV3 format: one line per stimulation event, three columns with stimulus onset and duration (both in seconds relative to the start of a scan), as well as a third column with an arbitrary intensity weight that is always set to 1.\n\nAll functional MRI data were converted from the DICOM format into the NIfTI format for publication using the same procedure as in 1.\n\nfMRI data are available in three different flavours, each stored in an individual 4D image for each run separately. Raw BOLD data are stored in bold.nii.gz. While raw BOLD data are suitable for further analysis, they suffer from severe geometric distortions. BOLD data that have been distortion-corrected9 at the scanner console are provided in bold_dico.nii.gz. In addition, distortion-corrected data that have been anatomically aligned to a per-subject BOLD template image are available: bold_bold7Tp1_to_subjbold7Tp1.nii.gz.\n\nHead movement correction was performed with respect to a dedicated reference scan at the start of the recording session within scanner online reconstruction as part of the distortion correction procedure. The associated motion estimates are provided in a whitespace-delimited 6-column text file (translation X, Y, Z in mm, rotation around X, Y, Z in deg) with one row per fMRI volume for each run separately.\n\nPhysiological data were truncated to start with the first MRI trigger pulse and to end one volume acquisition duration after the last trigger pulse. Data are provided in a four-column (MRI trigger, respiratory trace, cardiac trace and oxygen saturation), space-delimited text file for each run. A log file of the automated conversion procedure is provided in the same directory (conversion.log). Sampling rate for the majority of all participants is 200 Hz (see Table 2 for exceptions).\n\nRecent experiments have shown that audio features can be predicted via regression models from fMRI signals to test stimulus coding hypotheses3,10. To facilitate this activity with the current data we extracted four audio features from down-mixed mono stimuli. Feature extraction used a front-end windowed short-time Fourier transform, with window size 16384 samples (371.52 ms) and hop size 4410 samples (100 ms) yielding 63 overlapping feature vectors per stimulus file. Window parameters were chosen to trade temporal for spectral acuity, yielding frequency samples spaced linearly at 2.69 Hz intervals from 0–22.05 kHz. The four features extracted from this representation are described below.\n\nMel-Frequency Spectrum (mfs) – 48 dimensions. Motivated by human auditory perception the mel scale organizes frequency by equidistant pitch locations as determined by psychophysical experiments. We used the essentia open source audio processing library11 to extract the mel-frequency spectrum, which yielded energy in mel bands by applying a frequency-domain filterbank12 to the short-time Fourier spectrum. Frequency-domain filtering consisted of applying equal area overlapping triangular filters to the Fourier spectrum spaced according to the mel scale and normalized such that the sum of coefficients for every filter equals one.\n\nMel-Frequency Cepstral Coefficients (mfcc) – 48 dimensions. Cepstral features have been widely reported to perform well in speech recognition and music classification systems13, where the task is required to be sensitive to timbre. Typically, only the lower 10–20 cepstral coefficients (low quefrency) are retained; these encode the shape of the broad spectral envelope – an acoustic correlate of timbre. However, when sensitivity to timbre is not required, utilizing the upper coefficients (high quefrency), that encode fine spectral structure such as pitch, makes the feature robust to timbral changes14. We extracted the full set of 48 cepstral coefficients from the mel-frequency spectrum, by mapping the mel spectrum to a decibel amplitude scale and multiplying by the discrete cosine transform (DCT-II) matrix. It is expected that any application would first remove the constant first column and retain either the subsequent 13–20 coefficients or the remaining upper coefficients after those, depending on whether sensitivity or robustness to timbral difference is required. The remaining two features yield such a separation into low and high quefrency spectral components.\n\nLow-Quefrency and High-Quefrency Mel-Frequency Spectrum (lq_mfs, hq_mfs). Although proven to be useful in machine classification tasks, cepstral coefficients are in a different domain than the spectrum. The last two features map selected cepstral coefficients back to the spectrum domain by reconstructing the 48 mel-frequency spectrum bands using the low-quefrency and high-quefrency mfcc coefficients respectively. In each case, the non-selected coefficients were zeroed and the resulting feature mapped back to the spectral domain using the inverse (transposed) DCT-II matrix and then inverting the decibel amplitude scale. These two sets of features represent broad-spectrum information (timbre) and fine-scale spectral structure (pitch) respectively. The product of these two spectra yields the mel-frequency spectrum.\n\nThe source code for descriptive statistics in Figure 1 and Figure 3, as well as the implementation for the analysis presented in Figure 4 is available in a Git repository at https://github.com/psychoinformatics-de/paper-f1000_pandora_data. Source code for the implementation of the stimulation paradigm and audio feature extraction are included in the data release. Additional scripts for data conversion and quality control are available at: https://github.com/hanke/gumpdata.\n\nThese estimates indicate relative motion with respect to a dedicated reference scan at the beginning of each scan session. The area shaded in light gray depicts the range across participants, while the medium gray area indicates the 50% percentile around the mean, and the dark gray area shows ± one standard error of the mean. The black line indicates the median estimate. Dashed vertical lines indicate run boundaries where participants had a brief break. The red lines indicate motion estimate time series of outlier participants. An outlier was defined as a participant whose motion estimate exceeded a distance of two standard deviations from the mean across participants for at least one fMRI volume in a run. For a breakdown of detected outliers see Table 2.\n\n(A) Voxel-wise genre-selectivity label. Random-effects GLM group analysis (n=20) were computed using the FEAT component of FSL16. Individual contrasts were evaluated for each genre to identify voxels showing a BOLD response to this particular genre that is larger than the average response to all other genres. For all voxel clusters that show a significant difference at the group-level (cluster forming threshold Z=3.1, cluster probability threshold p<0.05) for any genre, the selectivity label was determined by the maximum Z statistic across all genres. No significant selective activation was found for the ambient genre. The majority of all voxels were labeled selective for one of the musical genres where stimuli contained vocals (country, rock’n’roll, heavy metal). Only a small cluster in BA44 R (Broca’s area) was labeled selective for symphonic music, despite the lack of speech content in these stimuli. (B) For comparison, the location of voxel clusters with above-chance classification accuracy for predicting the genre of a music stimulus (colors only indicate individual clusters, not association with particular genres). The associated areas are largely overlapping with the results of the GLM analysis. However, genre-discriminating signals were identified in a number of additional areas. For details on the MVP analysis and cluster statistics see Table 3. Unthresholded maps for GLM and MVP analyses are available at NeuroVault.org17 collection 308.\n\nThe table lists statistics (size, mean/max/std accuracy) as well as localization information (coordinates in mm MNI152) for clusters with above-chance classification performance in the group (cluster-level probability p<0.05; FWE-corrected). Clusters are depicted in Figure 4B. Statistical evaluation was implemented using a bootstrapped permutation analysis, as described by Stelzer and colleagues18 and implemented in PyMVPA19, using 50 permutation searchlight accuracy maps per subject, 10000 bootstrap samples, voxel-wise cluster forming threshold of p<0.001). Apart from two large clusters covering the majority of bilateral area for auditory perception and speech processing, additional clusters with genre-discriminating signals were identified. These include the bilateral medial geniculate bodys, as well as smaller regions on the ventral visual pathway, frontal orbital cortex, and the cerebellum. For these regions the NeuroSynth database20 reports high posterior probabilities for the topics: counting, motor, naming, phonology, prosody, visual, and vocal (as determined with the Neurosynth term atlas shipped with NeuroDebian8).\n\n\nDataset validation\n\nIn order to assess data quality, we investigated whether different BOLD response patterns associated with the five musical genres could be discriminated, using either univariate statistical parametric mapping or multivariate pattern (MVP) classification accuracy (searchlight-based analysis, radius of two voxels, sparse spatial sampling with sphere-centers spaced by two voxels, leave-one-run-out cross-validated classification analysis with a support vector machine, accuracy mapped on a voxel reflects the average across all sphere-analysis a voxel participated in). Inspection of the participant motion estimates revealed a median translation of less than a voxel size, and a maximum rotation of about 1 deg (see Figure 3 for outliers).\n\nDespite the variable magnitude of motion, no participant was excluded from the subsequent analysis.\n\nThe results of the univariate analysis (Figure 4A) and the MVP analysis (Figure 4B and Table 3) identify largely congruent areas. MVP analysis generally detects larger and more numerous areas, either due to higher sensitivity or a comparably more liberal statistical threshold. Noteably, clusters of above-chance classification accuracy not only contain auditory cortex and other cortical fields related to speech and music processing, but also the subcortical bilateral medial geniculate bodies, a neural relay station immediately prior to the primary auditory cortex in the auditory pathway15.\n\nGiven the confirmed wide-spread availability of genre-discriminating signal we conclude that these data are suitable for studying the representation of music and auditory features. Table 2 contains a list of all known data anomalies that may help potential data consumers to select appropriate subsets of this dataset.\n\n\nUsage notes\n\nThese data are part of a larger public dataset available at http://www.studyforrest.org. The website includes information on all available resources, data access options, publications that employ this dataset, as well as source code for data conversion and data processing.\n\nAll data are made available under the terms of the Public Domain Dedication and License (PDDL; http://opendatacommons.org/licenses/pddl/1.0/). All source code is released under the terms of the MIT license (http://www.opensource.org/licenses/MIT). In short, this means that anybody is free to download and use this dataset for any purpose as well as to produce and re-share derived data artifacts. While not legally required, we hope that all users of the data will acknowledge the original authors by citing this publication and follow good scientific practice as laid out in the ODC Attribution/Share-Alike Community Norms (http://opendatacommons.org/norms/odc-by-sa/).\n\n\nData availability\n\nOpenFMRI.org: High-resolution 7-Tesla fMRI data on the perception of musical genres: ds000113b21\n\nZENODO: Article sources for 7-Tesla fMRI data on the perception of musical genres, doi: 10.5281/zenodo.1876722\n\nZENODO: “Forrest Gump” data release source code, doi: 10.5281/zenodo.1877023\n\n\nConsent\n\nWritten informed consent for publication of acquired data in a de-identified form was obtained from all participants.", "appendix": "Author contributions\n\n\n\nMH conducted the study, implemented the stimulation paradigm, performed dataset validation analysis, and wrote the manuscript. RD implemented and performed the dataset validation analyses. CH converted the stimulation protocol into the OpenFMRI format. JSG contributed to the implementation of the stimulation paradigm, and contributed to the validation analysis. MC contributed the stimuli. FRK contributed to quality control analysis and to the manuscript. JS was the data acquisition lead.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by the German Federal Ministry of Education and Research (BMBF) as part of a US-German collaboration in computational neuroscience (CRCNS; awarded to James Haxby, Peter Ramadge, and Michael Hanke), co-funded by the BMBF and the US National Science Foundation (BMBF 01GQ1112; NSF 1129855). Work on the data-sharing technology employed for this research was supported by US-German CRCNS project awarded to Yaroslav O. Halchenko and Michael Hanke, co-funded by the BMBF and the US National Science Foundation (BMBF 01GQ1411; NSF 1429999). Michael Hanke was supported by funds from the German federal state of Saxony-Anhalt, Project: Center for Behavioral Brain Sciences.\n\nI confirm that the 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 Stefan Pollmann and André Brechmann for their feedback on the results of the classification analysis. We acknowledge the support of the Combinatorial NeuroImaging Core Facility at the Leibniz Institute for Neurobiology in Magdeburg. Only open-source software was employed in this study. We thank their respective authors for making it publicly available.\n\n\nReferences\n\nHanke M, Baumgartner FJ, Ibe P, et al.: A high-resolution 7-Tesla fMRI dataset from complex natural stimulation with an audio movie. Sci Data. 2014; 1: 140003. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLabs A, Reich T, Schulenburg H, et al.: Portrayed emotions in the movie “Forrest Gump” [v1; ref status: indexed, http://f1000r.es/55o]. F1000Res. 2015; 4: 92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasey M, Thompson J, Kang O, et al.: Population Codes Representing Musical Timbre for High-Level fMRI Categorization of Music Genres. In Mach Learn Interpret Neuroimaging. Springer Science and Business Media. 2012; 7263. : 34–41. Reference Source\n\nHaxby JV, Guntupalli JS, Connolly AC, et al.: A common, high-dimensional model of the representational space in human ventral temporal cortex. Neuron. 2011; 72(2): 404–416. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHanke M, Halchenko YO: A communication hub for a decentralized collaboration on studying real-life cognition [v1; ref status: not peer reviewed, http://f1000r.es/55n]. F1000Res. 2015; 4: 62. PubMed Abstract \| Publisher Full Text\n\nAguirre GK, Mattar MG, Magis-Weinberg L: de Bruijn cycles for neural decoding. Neuroimage. 2011; 56(3): 1293–1300. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeirce JW: PsychoPy--Psychophysics software in Python. J Neurosci Methods. 2007; 162(1–2): 8–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHalchenko YO, Hanke M: Open is Not Enough. Let’s Take the Next Step: An Integrated, Community-Driven Computing Platform for Neuroscience. Front Neuroinform. 2012; 6: 22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIn MH, Speck O: Highly accelerated PSF-mapping for EPI distortion correction with improved fidelity. MAGMA. 2012; 25(3): 183–192. PubMed Abstract \| Publisher Full Text\n\nToiviainen P, Alluri V, Brattico E, et al.: Capturing the musical brain with Lasso: Dynamic decoding of musical features from fMRI data. Neuroimage. 2014; 88C: 170–180. PubMed Abstract \| Publisher Full Text\n\nBogdanov D, Wack N, Gómez E, et al.: Essentia: an open-source library for sound and music analysis. In Proc of the 21st ACM Int Conf Multimed. ACM, 2013; 855–858. Publisher Full Text\n\nGanchev T, Fakotakis N, Kokkinakis G: Comparative evaluation of various MFCC implementations on the speaker verification task. In Proceedings of the SPECOM. 2005; 1: 191–194. Reference Source\n\nLogan B, Salomon A: A music similarity function based on signal analysis. In Proc IEEE Int Conf Multimed Expo. IEEE, 2001; 745–748. Publisher Full Text\n\nMüller M, Ewert S, Kreuzer S: Making chroma features more robust to timbre changes. In Acoust Speech Signal Proc, 2009. ICASSP 2009. IEEE Int Conf on. IEEE, 2009; 1877–1880. Publisher Full Text\n\nKrumbholz K, Schönwiesner M, Rübsamen R, et al.: Hierarchical processing of sound location and motion in the human brainstem and planum temporale. Eur J Neurosci. 2005; 21(1): 230–238. PubMed Abstract \| Publisher Full Text\n\nSmith SM, Jenkinson M, Woolrich MW, et al.: Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage. 2004; 23(Suppl 1): S208–19. PubMed Abstract \| Publisher Full Text\n\nGorgolewski KJ, Varoquaux G, Rivera G, et al.: NeuroVault.org: a web-based repository for collecting and sharing unthresholded statistical maps of the human brain. Front Neuroinform. 2015; 9: 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStelzer J, Chen Y, Turner R: Statistical inference and multiple testing correction in classification-based multi-voxel pattern analysis (MVPA): random permutations and cluster size control. Neuroimage. 2013; 65: 69–82. PubMed Abstract \| Publisher Full Text\n\nHanke M, Halchenko YO, Sederberg PB, et al.: PyMVPA: A Unifying Approach to the Analysis of Neuroscientific Data. Front Neuroinform. 2009; 3: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYarkoni T, Poldrack RA, Nichols TE, et al.: Large-scale automated synthesis of human functional neuroimaging data. Nat Methods. 2011; 8(8): 665–670. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHanke M, Dinga R, Häusler C, et al.: High-resolution 7-Tesla fMRI data on the perception of musical genres – an extension to the studyforrest dataset. OpenfMRI. 2015. Data Source\n\nHanke M, Dinga R, Häusler C, et al.: Article sources for 7-Tesla fMRI data on the perception of musical genres. Zenodo. 2015. Data Source\n\nHanke M: “Forrest Gump” data release source code. Zenodo. 2015. Data Source" }	[ { "id": "9304", "date": "28 Jul 2015", "name": "Cristiano Micheli", "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\nGeneral commentsI am not an fMRI expert myself; therefore I deemed it necessary to invite two co-reviewers to add comments on the manuscript. Please find a set of suggested changes hereafter.The manuscript describes a recent rich amendment of an existing fMRI open dataset (documented in studyforrest.org), aimed at increasing the amount of shared data from the same subjects, stimuli and procedures for 7T fMRI slab recordings, and 3T whole-head anatomical scans, but with a different cognitive task consisting of listening to music in 5 different genres (vocalized or not) and answering a dual-choice question (‘did the clip have a happy melody?’ [yes/no]). In addition, the dataset includes recordings of several physiological parameters.The dataset is extremely rich and very well documented and as such has the potential to be used as a reference dataset for various research questions from auditory to music and emotional processing. The choice of freely available software packages for data processing and the inclusion of the processing scripts greatly enhance reproducibility and augment the documentation of the processing steps and analyses provided in the paper. The idea of an open-source sharing framework for replication of results and follow up projects is greatly appreciated, together with the suggestion of having fMRI data available for benchmarking.The authors describe the experimental procedures and the data processing techniques that led to a pre-processing of the data, with the goal to run a preliminary quality check and validation on the material that will afterwards be available for future studies (the authors name this part ‘quality control’). Secondarily, the authors provide a pre-processing of the stimuli set, too, with the aim to provide elementary descriptors which could be used to relate fMRI brain responses to delivered music snippets for encoding models estimation. Thirdly, the authors mention the goal to validate the encoding models generated by the previous dataset (studyforrest.org). Alongside the main goals, the authors mention the possibility to use the latest fMRI corpus as a resource for benchmarking algorithms for functional alignment. Could the authors make examples of such algorithms, ideally with references?Firstly, in my opinion the stimuli and the experimental procedures are generally clearly described, but the pre-processing of fMRI analyses could be further commented in order to allow appropriate replication. In fact, some procedures are mentioned in other papers and that would require the knowledge of previous work done by the group. To ease the readers into the topic could the authors add one/two sentences for each procedure or method that is relevant for the points made in the present manuscript? For example at page 3 a landmark-based procedure is mentioned. Could the authors briefly mention what was used as a landmark and which software or metric allowed this? Additionally, one shortcoming of the dataset is the lack of a characterization of the audio presentation. The transforms were computed on the source waveforms. However, it is not so clear what waveform arrives at the ear and how the gain adjustments scaled them. This information would be important for the investigations of neural sound encoding with this dataset. In addition, this information would be of great use for users who intend to calculate other transforms on the source waveforms.Secondarily, while the authors’ premises are valid in only providing a quick and preliminary analysis of the data as a quality check, it would be helpful to briefly introduce the research context behind it. A couple of sentences in the introduction would be sufficient to explain the expected activated networks for those readers that are unfamiliar with neuroimaging studies on music.Finally, it would be helpful to briefly outline which ones of the many tools are made available in the paper directly, and which ones are preparatory for future studies. For example the computation of cepstral coefficients is often used in speech recognition, to separate vocal tract from speech descriptors and to create features for recognition, but is not clear if and how they are used in the current manuscript. We are aware of the amount of work carried out by the authors, in documenting and making the material available, however the achieved goals should not be mixed with future ones. Please amend in order to outline the current state of the research framework to the general reader. Detailed comments:Abstract. What is the advantage of a slow event-related paradigm (long ISI) for this experiment? Please comment on it in the text. p.4. Some more information about the acquisition procedures and the procedures applied to the preprocessed dataset should be provided, or at least stated where they can be found (e.g. scripts in the dataset). For example a short description of the FOV, (e.g. covering temporal and inferior frontal areas of the brain) may be added in the article itself. p.5. See previous comment. A better description of the physiological recordings is needed. Are they acquired with standard Siemens sensors? p.5. Preprocessing of fMRI data. Similarly to the previous comments, it is our opinion that the authors should thoroughly describe the set of steps that lead to the subsequent analyses, despite them being already documented in other papers. Critically, having missed on this point will not facilitate the reader in the choice among three different flavors of BOLD datasets at disposal. Please add more operative details, a flow chart figure or indicate where to find them. For example, a distortion correction technique is mentioned, along with a realignment of the fMRI volumes. A template image is also mentioned. How is the template generated? Is there a slice-time correction step? Is the distortion correction applied before, after or during the realignment? p.5. In the description of the mel coefficients could the authors briefly explain the difference between DCT_II and the other flavors of the DCT algorithm (or cite a representative reference)? p.7. What is the contrast of the univariate analysis? I understand that this is a standard analysis, but in our opinion its description is missing some crucial information on how it was performed. For example:a) I’d like some basic information on how the validation analyses were done: is the source code for the classification provided too?b) Given this statement \"Given the confirmed wide-spread availability of genre-discriminating signal we conclude that these data are suitable for studying the representation of music and auditory features.” it would be good to have a brief overview of previous studies (with references) in this field and the brain areas they usually associate with music processing + processing of different genres to facilitate understanding for readers without a background in the neurobiology of music.p.8. Figure 4 is a bit confusing. As I understand it the color code on top only refers to Figure 4a. Maybe label a) statistical analysis and b) classification analysis and use different colors in b. The legend refers to a small symphonic music cluster in RIFG which I cannot identify on the actual brain images. Minor commentsAbstract and p.2. The term ‘quality control’ might be slightly misleading. One could just state that basic GLM and MVPA analyses were conducted as a ‘proof of principle’ of effective classification on the data. p.3 Correct the sentence ‘continuum of research question’ into ‘… questions’ p.3 Participants: which of the functional scans was done first? p.4 Were the sound adjustments done only once at the beginning of the experiment or for every functional run? Is the level of the sound adjustment documented in the material? p.5 Correct the sentence ‘flag whether a control question with presented’ into ‘…was presented’ p.5 “As the exact same alignment target was used, this led to a very similar field-of view configuration across acquisitions.” Somewhat unclear. Does this mean that the subjects’ head was in the same position in both experiments (i.e. voxels match functionally) or does it mean that the FOV was in the same scanner’s coordinates? It is unlikely that both conditions are met simultaneously. p.5 “While raw BOLD data are suitable for further analysis, they suffer from severe geometric distortions. BOLD data that have been distortion-corrected9 at the scanner console are provided in bold_dico.nii.gz.“ Please provide more information about the correction (e.g. k-space or voxel space correction). Is it documented? Is all information that was used available in the dataset? Users may want to apply their own correction p.5 “A log file of the automated conversion procedure is provided in the same directory (conversion.log)”. Conversion of what? p.5 “Audio features” Are the sound files provided with the dataset? You might want to state that here. Otherwise state how the user could obtain them to apply their own transforms. It would be important to characterize the sound presentation system. The transforms were calculated on the sound files but it is not so clear what actually reached the ear and how the genres might have differed in energy etc. This is information could greatly extend the usefulness of this rich dataset. p.5 “frequency domain filter bank” Are the parameters provided? If yes, mention that here. p.5 The description of the audio features reports a highest signal frequency of ~22KHz but the signals in figure 1 show a clear decrease in gain for all stimuli after ~16KHz. Could the authors document what the sampling frequency of the audio waves was and if a low pass filter was applied?", "responses": [] }, { "id": "9305", "date": "04 Aug 2015", "name": "Karsten Müller", "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 paper presents a very valuable functional MRI (fMRI) data set, investigating brain function on the perception of musical genres. It is a very interesting study with music perception of the human brain. The paper is well written and the data presentation is in an excellent fashion. This presentation will allow a further usage of the data even by other scientists.1. Although it's a 'data paper', I would suggest including at least a few sentences about the motivation of the study. Why is it interesting to study the presented type of stimuli? In which context would it fit into the current state of research?2. MRI acquisition protocol and subject selection is referred to in a previous paper. As the other reviewer suggests, I would also prefer to read acquisition parameters and subject data in this paper because it plays a central role.3. As the current manuscript is a data paper, I would suggest to include a characterization of the MRI data quality. I would be happy to see noise parameters of the BOLD images over time. Is the noise stable? It should be possible to present signal-to-noise and contrast-to-noise values within a subject over time but also between subjects. This would be very helpful to characterize the data set. I assume that the numbers are excellent compared with 3T data.4. There's nothing written about normalization quality. Would it be possible to demonstrate that the normalization was successful for all subjects? It would be great to read some numbers reflecting a normalization quality assessment.5. I further suggest to present the individual foreground voxel mask images. When they are normalized, the masks can be shown in a single image with different colors showing the number of subjects having foreground in a selected voxel.6. Are the audio files available? It would be a good idea to offer the stimulus material.", "responses": [] } ]	1	https://f1000research.com/articles/4-174
https://f1000research.com/articles/4-172/v1	29 Jun 15	{ "type": "Research Article", "title": "Following specific podocyte injury captopril protects against progressive long term renal damage", "authors": [ "Yu S Zhou", "Ihmoda A Ihmoda", "Richard G Phelps", "Christopher OS Bellamy", "A Neil Turner", "Ihmoda A Ihmoda", "Richard G Phelps", "Christopher OS Bellamy" ], "abstract": "Background: Angiotensin converting enzyme inhibitors (ACEi) reduce proteinuria and preserve kidney function in proteinuric renal diseases. Their nephroprotective effect exceeds that attributable to lowering of blood pressure alone. This study examines the potential of ACEi to protect from progression of injury after a highly specific injury to podocytes in a mouse model.Methods: We created transgenic (Podo-DTR) mice in which graded specific podocyte injury could be induced by a single injection of diphtheria toxin. Transgenic and wild-type mice were given the ACEi captopril in drinking water, or water alone, commencing 24h after toxin injection. Kidneys were examined histologically at 8 weeks and injury assessed by observers blinded to experimental group.Results: After toxin injection, Podo-DTR mice developed acute proteinuria, and at higher doses transient renal impairment, which subsided within 3 weeks to be followed by a slow glomerular scarring process. Captopril treatment in Podo-DTR line 57 after toxin injection at 5ng/g body weight reduced proteinuria and ameliorated glomerular scarring, matrix accumulation and glomerulosclerosis almost to baseline (toxin: 17%; toxin + ACEi 10%, p<0.04; control 7% glomerular scarring). Podocyte counts were reduced after toxin treatment and showed no recovery irrespective of captopril treatment (7.1 and 7.3 podocytes per glomerular cross section in water and captopril-treated animals compared with 8.2 of wild-type controls, p<0.05).Conclusions: Observations in Podo-DTR mice support the hypothesis that continuing podocyte dysfunction is a key abnormality in proteinuric disease. Our model is ideal for studying strategies to protect the kidney from progressive injury following podocyte depletion. Demonstrable protective effects from captopril occur, despite indiscernible preservation or restoration of podocyte counts, at least after this degree of relatively mild injury.", "keywords": [ "ACEi", "angiotensin", "podocyte", "proteinuria" ], "content": "Introduction\n\nPodocytes are terminally differentiated, highly specialised epithelial cells which cover the outer surface of the glomerular basement membrane and form the final barrier to protein loss during glomerular filtration. Podocyte dysfunction and subsequent loss plays a major role in the initiation and progression of glomerular diseases1,2. Podocyte injury is characterised by leakage of protein into urine, which can occur even without morphological changes detectable by light microscopy. The close association between damaged podocytes and proteinuria is supported by the observations that numerous congenital causes of substantial proteinuria are due to mutations influencing podocyte-specific molecules such as molecules involved in the cytoskeleton3 and slit diaphragm (podocin, nephrin and CD2AP4–7).\n\nKriz and colleagues1 proposed that progressive podocyte damage might lead to chronic renal failure in a number of renal diseases, and that progression might arise because “podocyte damage damages podocytes” (directly or through loss of inter-cell support), leading to a vicious cycle that drives progressive glomerular injury and scarring8,9. If so, interventions that reduce the disruption by rescuing susceptible podocytes next to injured ones are potential therapies to restore podocyte phenotype and therefore ameliorate renal damage and/or protect the kidney from progressive damage.\n\nThere is strong evidence that proteinuria reduction with angiotensin converting enzyme inhibitors (ACEi) and angiotensin receptor blockers (ARB) can arrest deterioration in renal function in proteinuric kidney diseases of any aetiology, both in animal models10–13 and in man14,15. This protection was first attributed to haemodynamic effects. Recent evidence suggests additional mechanisms16,17.\n\nWe developed a model of targeted podocyte injury by constructing a transgenic mouse (Podo-DTR) in which the diphtheria toxin receptor is expressed on podocytes. Murine cells are naturally 1000-fold less susceptible to diphtheria toxin than human cells18. Transgenic mice with podocytes fully susceptible to diphtheria toxin were made by expressing the human diphtheria toxin receptor (hDTR, also known as human heparin binding epidermal growth factor receptor, HB-EGFR) under the control of a fragment of the nephrin promoter that was previously shown to be expressed solely in podocytes when coupled to β-galactosidase19. This technique was first applied to hepatocytes20 but has subsequently been successfully applied to a number of other cell types8,21.\n\nUnlike models of podocyte injury that involve use of toxins of uncertain specificity, such as adriamycin in mice22,23, puromycin in rats24, and possibly pamidronate toxicity and HIV infection in humans25, our Podo-DTR model permits a graded, specific podocyte injury that can be delivered by a single injection of diphtheria toxin. Here we describe application of Podo-DTR to investigate the nephroprotective potential of ACEi in podocyte injury.\n\n\nMethods\n\nThe HB-EGF receptor was expressed as a transgene under the control of the murine nephrin promoter. A murine 1.25kb nephrin gene fragment19 was generated from murine genomic DNA. Oligonucleotide PCR primers (MWG Oligo synthesis) for the mouse nephrin gene with added NotI and BamHI restriction sites (primer sequences: 5'-ATGGCCCAGGGATTCAGGTGC-3' and 5'-GCTTGGACCCAGTGTGAACTC-3') were used to clone the gene fragment. The thermal cycling protocol of the PCR machine (Thermal Cycler Apollo ATC201) comprised an initial denaturation step at 75ºC for 10 minutes followed by 30 cycles of 95ºC for 1minute (denaturation), 60ºC for 1 minute (annealing) and 72ºC for 1.5 minute (elongation). The final cycle consisted of a re-annealing at 72ºC for 10 minutes.\n\nThis nephrin fragment was used to replace the albumin promoter in the pMS7 plasmid, which was kindly donated by Dr. Saito20. The resulting plasmid consisted of the murine nephrin promoter-fragment and the human HB-EGF cDNA19. Expression driven by this nephrin promoter fragment was shown by Moeller et al. (2002) to achieve podocyte specific expression in kidneys without detectable expression outside the kidney by chemiluminescence assay.\n\nTransgenic mice were generated by male pronuclear microinjection of murine fertilized ova derived from B6CBAF1 mice with the linearised pIN plasmid (Figure 1). These mice originated from Harlan UK, and were offspring of a cross between the C57BL/6JolaHsd inbred female and the CBA/CaOlaHsd inbred male.\n\n(a) Schematic representation of the transgene. Human diphtheria toxin receptor (DTR), or human heparin binding-epidermal growth factor (HB-EGF) receptor, is expressed as a transgene under the control of the murine nephrin promoter. (b) Agarose gel showing positive bands for the PCR products of a 243bp fragment from podocyte nephrin promoter and rabbit-ß-globin intron of transgenic Podo-DTR mice. M, marker Hae ladder; Tg, trasgenic; WT, wild type.\n\nTransgenic offspring were identified by PCR analysis of ear notch DNA using the following primers: 5'-GGA AGA GAG AAG GGC GAG TT-3' and 5'-GGG TCC ATG GTG ATA CAA GG-3' for a 243bp nephrin gene/intron product; and 5'-GGT GGT GCT GAA GCT CTT TC-3' and 5'-GCT TGT GGC TTG GAG GAT AA-3' for a human HB-EGF gene product. Thermal Cycler Apollo ATC201 PCR machine was used with the same thermal cycling protocol as described above except for the annealing temperature (set to 50ºC for 1 minute per cycle).\n\nAll animal studies conformed to local ethical guidelines and the Home Office (UK) Animals Scientific Procedures Act (1986) and were approved by the University of Edinburgh Ethical Review Committee (Ethical Review Number PL23-07). Mice were housed in laboratory cages (n=1–6) in a room with 12:12 hour light-dark cycle and allowed free access to standard dry pellet diet and water (unless otherwise specified) in accordance with the institutional guidelines.\n\nPodo-DTR mice and wild-type littermates as controls were injected with a single dose of diphtheria toxin (DT) (Product no. 150, Lot 15023A1; List Biological Laboratories Inc., California, USA) ranging from 0.1–166ng/g body weight (bw). At various time points, urine and blood samples were collected for albuminuria and creatinine, and urea measurement respectively. Animals were sacrificed (aged 7 to 11 months) by intraperitoneal injection of terminal anaesthesia (1mg/ml medetomidine (Dormitor) (Ref. VD DOM02, made by Orion Pharma, supplied by Henry Schein Medical) and 100mg/ml ketamine (Vetalar) (Ref. PD VET10, made by Pfizer, supplied by Henry Schein Medical) at approximately 0.1ml/10g of body weight and their kidneys analysed histologically.\n\nGroups of 16 transgenic (Tg) mice and eight wild-type (WT) littermates were given captopril (200mg/L) (Product no. C4042, Sigma Aldrich, UK) in their drinking water or placebo (water alone) 24h after a single intraperitoneal (i.p.) injection of diphtheria toxin at 5ng/g bw (Product no. 150, Lot 15023A1; List Biological Laboratories Inc., California, USA). The size of the experimental group was based on a power calculation in order to have 80% power to detect a histological score difference of 1 on a 5 point scale with p≤0.05, assuming SD for score is 1.\n\nAnimals of both sexes (1:3 female to male ratio) aged 3 to 12 months (mean: 7.4, median: 8 months) were allocated to three groups:\n\nTg DT+H2O received water alone;\n\nTg DT+ACEi were treated with captopril;\n\nWT DT+ACEi were wild-type littermate controls treated with captopril.\n\nUrine samples were collected over 24h in metabolic cages at days 0, 14, 42, 49 and 56 and analysed for albumin:creatinine ratio (ACR). Systolic blood pressures were measured by tail cuff plethysmography on trained conscious animals during week 7 after toxin injection, and the mean of 3–4 measurements was recorded for each animal. Untreated Podo-DTR mice (n=5) (transgenic not given diphtheria toxin or captopril) were also included to assess baseline blood pressure.\n\nTerminal blood samples were collected at week 8 from intraperitoneally anaesthetised animals (injected with medetomidine and ketamine).\n\nKidneys from each animal were bisected sagitally and fixed as required by overnight incubation at 4ºC in fixative (10% neutral formalin (VWR Brand P/L-Chemicals), Methyl-Carnoy fixative (60% absolute methanol, 30% chloroform, 10% glacial acetic acid (Fisher Scientific UK Ltd), or Karnovsky’s glutaraldehyde (700mOsm) (Ref. G5882-100ml, Sigma Aldrich) or snap frozen in liquid nitrogen. For light microscopy, formalin-fixed samples were embedded in paraffin-wax and 2µm sections cut and stained with haematoxylin and eosin (H&E) or periodic acid-Schiff (PAS) (Fisher Scientific UK Ltd).\n\nUrine and serum creatinine concentrations were measured using the creatinase reaction, with the exception of the first cohort of the model evaluation studies where Jaffe reaction was used (for the Podo-DTR line 47 given 1ng/g DT). Serum urea was measured using the urease reaction (Alpha Laboratory Ltd, Poole, UK). An immunoturbidimetric assay was developed to measure urinary mouse albumin concentration using a commercial diagnostic Microalbumin Kit (Olympus Diagnostic Ltd, Watford, UK) standardised against purified mouse albumin (Sigma Chemical Co. Poole, UK). All the assays were adapted for use on a Cobas Fara centrifugal analyser (Roche diagnostics Ltd, Welwyn Garden City, UK) according to manufacturer’s instructions.\n\nSclerosis was defined as collapse and/or obliteration of glomerular capillary tuft accompanied by presence of hyaline material and/or an increase in matrix12. Glomerulosclerosis was graded on 2µm thick PAS-stained sections, adopting the semi-quantitative scoring system proposed by El Nahas et al.26. The severity of glomerulosclerosis was expressed on a scale of 0 to 2. The scoring system used was as follows: 0: normal glomerulus or no lesion; 1: <50% sclerosis; 2: 50–100% sclerosis of glomerular tuft area. Using light microscopy at a magnification of x40 (Olympus CX40), 100 glomeruli per animal were scored for glomerulosclerosis by an observer blinded to experimental group.\n\nParaffin-embedded, formalin-fixed kidneys were cut at 3µm thickness. After deparaffinising and hydrating, the sections were treated with the antigen retrieval solution Borg Decloaker RTU (Ref. BD1000MM Biocare Medical) for 2 minutes after reaching pressure according to manufacturer’s instructions. The sections were incubated with rabbit polyclonal anti-Wilm’s tumour 1 (WT1) sc-192 IgG antibody (1:50; Santa Cruz Biotechnology, Inc) for 1h at room temperature. Immunoperoxidase staining was performed according to the Vectastain ABC kit (Vector Laboratories). Diaminobenzidine (DAB) (Vector Laboratories, Inc Burlingame, CA) was used as the immunoperoxidase detection system.\n\nWT-1 is a marker that specifically stains podocyte nuclei. Podocytes were counted in 50 consecutive glomerular cross-sections per animal viewed at x40 magnification (Olympus CX40), and the mean podocyte per glomerulus count calculated for each animal, n=16 (transgenic), n=8 (wild-type).\n\nTo investigate for any change in glomerular size following treatment (with diphtheria toxin and/or ACEi treatment), 50 consecutive glomerular cross section (GCS) per animal were measured using ImageJ 1.4r software on pictures taken at x10 magnification using QCapturePro (QImaging Micropublisher 3.3 RTV, Zeiss Axioskop, Germany).\n\nResults are expressed as mean ± standard error of the mean (SEM) or median where specifically stated. Statistical differences between groups were tested by the Student t-test or Wilcoxon Rank Sum as appropriate (Graphpad Prism 4 version 1.0 or R www.r-project.org version 2.1). A p-value of <0.05 was considered to be significant.\n\n\nResults\n\nThree transgenic lines with positive PCR results for the transgene (Figure 1b) were established (Podo-DTR 47, 57 and 21). Earlier studies in total of 10, 17 and 24 mice of Podo-DTR line 47, 57 and 21 respectively (male and female) aged 6.5 to 9 months showed that two of the lines (47 and 57) were susceptible to diphtheria toxin. All three lines were entirely healthy and had normal glomerular morphology. Non-transgenic animals and also animals from line 21 had normal glomerular morphology regardless of the dose of toxin received. Mice from lines 47 and 57 developed renal injury in response to low doses of toxin but no pathology outside the kidneys was detected.\n\nIn dose-ranging studies, animals were injected intraperitoneally with diphtheria toxin at 0.1–166ng/g bw. Mice from line 47 were most susceptible and developed fatal acute renal injury in response to doses of 2.5ng/g or greater (n=3–6/group, aged 2–10 months), and a dose-dependent, slowly progressive (over 6 months) glomerulosclerosis in response to lower doses (n=3–6/group, aged 2–6 months). Mice from line 57 developed only proteinuria even at doses as high as 20ng/g bw (n=6/group, aged 2–10 months). Mice from line 21 were unaffected by administration of toxin despite carrying the transgene (n=3–6/group at 1ng and 50ng/g bw, aged 6–9 months).\n\nThe time course of sub-lethal toxin treatment in line 47 was: development of proteinuria (6mg/mmol at 48h vs 2mg/mmol baseline, p=0.031) within days which increased in severity over 2 weeks (2048mg/mmol) then declined (57.6mg/mmol at 5 weeks) but never to baseline (12.4mg/mmol at 26 weeks (n=3–6) (Figure 2a). No significant albuminuria was seen in transgenic animals not given toxin or wild-type animals treated with diphtheria toxin (0.9–6.3mg/mmol) (Supplementary Table 1).\n\n(a) Albumin:creatinine ratio (ACR) measured in urine of Podo-DTR line 47 at intervals after injection with 1ng/g bw. Acute proteinuria was highest at week 2 (mean value 2047.9mg/mmol) falling to much lower levels by 5 weeks (57.6mg/mmol) and remaining low at 26 weeks (12.4mg/mmol). No substantial change was seen in control groups (0.9–6.3mg/mmol) (data not visualized). (b) Glomerulosclerosis score of Podo-DTR line 47 mice at 6 and 26 weeks post DT injection at 1ng/g bw. 100 glomeruli were scored per animal (Uncropped gel image in Supplementary Figure 1).\n\nSerum urea showed a similar acute profile to ACR (Supplementary Table 2). Although the number of experimental animals was small (n=3–7), it appeared that there was an acute rise in blood urea at day 14 (10.9mmol/L vs 5.4mmol/L at d0), followed by temporary recovery at week 5 (5.1mmol/L) and then by slow deterioration (7.9, 8.1, 9.2mmol/L at week 6, 8 and 26 respectively).\n\nToxin treated transgenic (Podo-DTR) mice showed a progressive reduction in the number of normal glomeruli with time (74% at 6 weeks, 53% at 26 weeks, Figure 2b). At 1 month, early focal segmental glomerulosclerosis (FSGS) and chronic renal damage was observed (Figure 3C–D). Importantly, additional glomeruli became morphologically abnormal even after 6 weeks post-toxin-treatment. In contrast, almost all of sampled glomeruli in control group animals (wild-type toxin-treated and transgenic untreated animals) were morphologically normal (96%) (Figure 2b).\n\n(A) Normal glomerulus in untreated transgenic mouse. (B) Damaged glomerulus with abnormal morphology and vacuolated cytoplasm () from treated Podo-DTR line 47 (25ng/g bw DT) at D7. (C–D) Sclerosed glomeruli from treated Podo-DTR line 57 (1ng/g bw DT) at D28 with (C) partially damaged glomerulus with normal morphology at 4 o’clock, with tuft/capsular adhesion in the area of segmental scar at 9 o’clock, and (D) almost completely sclerosed glomerulus. PAS staining, 40x magnification. p=0.002; ** p<0.0001; DT, diphtheria toxin; bw, body weight; D, day; NS, non significant; PAS, Periodic Acid Schiff.\n\nThe number of podocytes per glomerular cross section (GCS) was reduced in toxin-treated susceptible animals compared to the controls in Podo-DTR line 47 (6.2 at 2 weeks, 5.3 at 26 weeks vs 10.0 podocyte per GCS in controls, p<0.02) (Figure 4). The reduction in podocyte count was dose-dependent in line 57 animals. At higher toxin dose (20ng/g bw DT) at 14 days, the mean podocyte number dropped to 6.1 podocyte/GCS compared with 7.2 podocyte/GCS in mice treated with lower dose at 5ng/g bw toxin at week 8 versus 8.5 podocyte/GCS of Tg saline treated controls (p<0.04). (Supplementary Table 3).\n\nPodocyte numbers were significantly reduced at 2 and 26 weeks after toxin injection compared to the controls (6.2 and 5.3 versus 10.0 podocyte/GCS respectively, p<0.02). bw, body weight; GCS, glomerular cross section; * p=0.015; ** p=0.001; ♦, wild-type control mice injected with DT; ♦, transgenic mice injected with saline.\n\n\nDrug intervention study\n\nACEi-treatment lowered systolic blood pressure in toxin-treated line 57 and wild-type mice from a mean of 114 to 84±1.7 and from 114 to 73±1.9 mmHg respectively (Figure 5a).\n\n(a) Tail cuff blood pressure (BP) of Podo-DTR line 57 mice. At 7 weeks post DT injection (1ng/g bw), the BP of ACEi treated mice, whether Tg or WT (84 and 73mmHg respectively) were significantly lower (p<0.001) than the untreated groups (114mmHg). (b) Urine albumin:creatinine ratio (ACR) of Podo-DTR line 57 mice. At day 0, prior to DT injection, mice from all 3 groups had baseline level of urine ACR (range: 0.0–6.8mg/mmol). The DT+H2O treated group peaked at week 2 (271.5±128mg/mmol), the DT+ACEi treated group had the urinary ACR level blunted substantially (39.1±9mg/mmol). The long-term albuminuria was lowered in both DT+ACEi treated and DT+ H2O groups. The WT controls had baseline level of urine ACR throughout the experiment (mean range value: 3.0–4.6mg/mmol). DT, diphtheria toxin; ACEi, angiotensin converting enzyme inhibitor; Tg, transgenic; WT, wild type; bw, body weight; vs, versus; , p<0.0001 vs Tg DT+H2O; #, p≤0.0001 vs Tg DT+ACEi; +, p=0.002 vs WT DT+ACEi.\n\nACEi-treatment also substantially reduced proteinuria in toxin-treated line 57 mice, although not to the levels observed prior to toxin treatment, or in wild type mice (range: 0.0–6.8mg/mmol) (Figure 5b). The peak level of proteinuria (at week 2) was reduced from 272±128mg/mmol in ACEi-treated mice compared to 39.1±9mg/mmol in mice treated with the diphtheria toxin only, and was substantial at all measurement times (Figure 5b).\n\nThe proportion of glomeruli showing scarring and matrix accumulation in toxin-treated line 57 mice was substantially reduced in ACEi-treated mice (10% vs 17%, 10%, p<0.04), almost to the levels observed in wild-type control mice (7%) (Figure 6 & Figure 7).\n\nGlomerular scarring was reduced almost to baseline level by ACEi captopril (Tg DT+H2O: 17%, Tg DT+ACEi: 10%, p<0.04; wild-type control: 7%).\n\n(A) Cluster of 4 glomeruli with varying degree of scar from a Tg DT+ H2O treated mouse injected with 5ng/g bw DT; (B) Improved glomerular histology of a Tg DT+ACEi treated mouse injected with 5ng/g bw DT. PAS staining, 20x magnification. DT, diphtheria toxin, ACEi: angiotensin converting enzyme inhibitor, bw: body weight; Tg, transgenic; , p<0.02; *, p≤0.001; NS, non significant (p=0.33); PAS, Periodic Acid Schiff.\n\nPodocyte counts at week 8 were lower in toxin-treated susceptible mice than wild-type controls (median podocytes per GCS was 7.1 for Tg DT+H2O, and 8.2 for wild-type mice, p<0.05 by Wilcoxon Rank Sum). However, the counts in ACEi-treated, toxin-treated line 57 mice were not significantly higher (7.3 for Tg DT+ACEi) than in toxin-treated line 57 mice, and were significantly lower than in wild type controls (10% Wilcoxon Rank Sum) (Figure 8a). These results suggest that ACEi-treatment was not advantageous for podocyte preservation in this model (Figure 8b).\n\nBox plots of the podocyte count. (a) or podocyte count per unit glomerular cross section area (b) for Podo-DTR mice treated with toxin with or without ACEi compared with wild-type (toxin + ACEi treated) mice at week 8. Significance was assessed using the 10% Wilcoxon Rank Sum test. Tg, transgenic; WT, wild-type; ACEi, angiotensin converting enzyme inhibitor; GCS, glomerular cross section; DT, diphtheria toxin; NS, non significant; , p=0.03; **, p≤0.003.\n\nA lack of benefit of ACEi treatment for podocyte preservation continued to be apparent when podocyte counts were expressed as counts per unit in glomerular cross sectional area, in order to compensate for any change in glomerular size. Indeed, mean glomerular size measured in 50 consecutive glomerular cross-sections using ImageJ 1.4r software was similar in the experimental groups irrespective of toxin-treatments at 9676, 10708, and 9693µm2 in WT DT+ACEi, Tg DT+H2O and Tg DT+ACEi respectively, p>0.08 (Supplementary Table 4).\n\n\nDiscussion and conclusions\n\nOur results demonstrate utilisation of mice engineered to be susceptible to podocyte damage to evaluate the capacity of podocyte-protective drugs such as captopril to modulate self-perpetuating mechanisms of podocyte damage. Other groups have developed toxin receptor-mediated conditional podocyte knockout models8,27 similar to ours, but the application of such models in drug studies is novel.\n\nOur Podo-DTR mouse model has a number of advantages over the existing animal models. Wiggin et al. have employed similar technology8,27 but in rats rather than mice. A mouse model offers greater potential for further analysis because of the rich availability of mouse-specific reagents and large number of existing transgenic mouse lines. Moreover, breeding turnover is greater with a higher litter number, maintenance costs are lower and smaller volumes of toxins or drugs are required for studies.\n\nUnlike the human (h) CD25 mouse model generated by Ichikawa’s group in Japan8,27, where only relatively short-term timepoints (up to four weeks) have been presented, even after low dose of LMB2 immunotoxin treatment at 0.625ng/g bw, our Podo-DTR mice survived up to six months after 1ng/g bw DT injection in line 47. This allowed analysis of long-term timepoints for histological changes and sclerosis development.\n\nACEi and ARB were originally thought to effect renal protection via haemodynamic effects and reduced glomerular filtration pressure. Although blood pressure reduction by any means has been shown to protect renal excretory function in proteinuric diseases28, this accounts for only a part of the activity of ACEi and ARB agents. Proposed additional mechanisms of renal protection include modulation of the toxicity of filtered protein by influences upon non-glomerular cells29: it is thought that filtered protein may be toxic to renal tubular cells contributing to interstitial fibrosis30.\n\nAn alternative hypothesis is that ACEi and ARB exert a beneficial influence on the rate of continuing podocyte damage/loss thought to be central to the progression of proteinuric renal diseases and development of chronic renal failure8,9,31. There is evidence - from elegant work utilising genetic chimeras in which a podocyte subpopulation expresses the hCD25 toxin receptor and may be selectively injured - that podocyte injury initially focused on a subset of podocytes can spread to other podocytes32. The initial insult caused by the toxin exposure to hCD25-positive podocytes lead to cell death and massive proteinuria within 4 days, followed by a secondary wave of injury to podocytes lacking the toxin receptor at 6 weeks post-toxin administration, with foot process effacement and downregulation of podocyte specific markers such as nephrin, podocin and VEGF in conjunction with an increase of the injury marker desmin. These results parallel the findings of our study where additional glomerular damage and possibly further reduction in podocyte number was observed from up to 26 weeks after acute injury, long after the initial injury induced by diphtheria toxin (Figure 2b & Figure 4). This suggests that we may also be seeing the phenomenon of propagation of podocyte injury, “podocyte damage damages podocytes” leading to progressive glomerular injury and scarring. This finding may open up the possibility that drugs that protect podocytes may also be generally nephroprotective.\n\nAny broader implication of podocyte health for renal preservation is important to define as many drugs employed in renal disease have been shown to have direct effects on podocytes. As well as ACEi and ARB10,16,17, direct effects on podocytes have been reported for: immunosuppressive drugs including corticosteroids33,34, calcineurin inhibitors such as cyclosporine35 and tacrolimus36, and mizoribine37,38; and non-immune drugs including the endothelin A receptor antagonist (ETA-RA)39,40, and peroxisome proliferator-activated receptor gamma (PPARγ) agonists41,42.\n\nThe Podo-DTR mouse model described here has advantages for studying podocyte rescue as it is possible - through selection of the dose of toxin - to produce a consistent non-lethal degree of podocyte-specific injury. Mild degrees of injury were shown to superficially heal over 8 weeks with normalisation of glomerular morphology observable under light microscopy and reduction of proteinuria, but also with continuing reduction in podocyte numbers and low level proteinuria. On this background we were able to demonstrate a benefit of captopril in arresting histopathological progression of glomerular damage despite a reduction in podocyte number. This is in contrast to studies that have shown a protective effect of ACEi upon podocyte number in certain settings10,17, but in agreement with two other chronic models of renal disease in subtotally nephrectomised rats13 or antibody-induced nephritic mice42.\n\nComparison of the change in podocyte number between studies is complicated by the range of indirect measurement techniques that have been used. Our study employed quantification of podocytes per glomerular tuft in a 2D image similar to the approach of many other groups43–45. Another approach is to estimate the total number of podocytes per whole glomerulus by extrapolating a volume from a series of 2D images2, which may have some advantages but is elaborate, very difficult to apply to experiments with large numbers of animals, and assumes glomeruli have a consistent size and spherical shape. The best technique with which to quantify podocytes is still a matter of debate, but whilst this difficulty complicates comparison between studies, it does not prevent conclusions from the comparisons made within the various studies where measurements have been made using a consistent technique.\n\nOur results suggest that protection against podocyte loss is not the only mechanism by which ACEi achieve long-term nephroprotection: it is likely that an influence on podocyte phenotype or function is also important. Our model could be used to identify or screen new compounds to reduce podocyte damage and preserve renal function.", "appendix": "Author contributions\n\n\n\nANT, YSZ conceived and designed the experiments. YSZ performed the experiments. YSZ, ANT, RGP, COB analyzed the data. The transgenic animal model was created by IAI. YSZ, ANT and RGP wrote and edited the paper. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a Medical Research Council UK studentship to YSZ and the Edinburgh Renal Research Fund.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nYSZ was supported by a Medical Research Council Studentship. The authors would like to thank Dr Forbes Howie for developing the urine albumin immunoturbidimetric assay, Spike Clay for assistance with injections and Lauren Melrose for technical support.\n\n\nSupplementary materials\n\n24h urine analysis measuring albumin and creatinine of untreated control Podo-DTR line 47 mice and treated with saline or DT at 1ng/g bw\n\nClick here to access the data.\n\nSerum creatinine and urea concentration of experimental Podo-DTR line 47 mice at various time points\n\nClick here to access the data.\n\nPodocyte quantification of Podo-DTR mice treated with 20ng/g bw DT at week 2 (14d) and 5ng/g bw DT at week 8 (56d). At higher toxin dose (20ng/g bw DT) at 14 days, the mean podocyte number dropped to 6.1 podocyte/GCS compared with 7.2 podocyte/GCS in mice treated with lower dose at 5ng/g bw toxin at week 8 versus 8.5 podocyte/GCS of Tg saline treated controls (p<0.04).\n\nClick here to access the data.\n\nGlomerular area and diameter measurements of Podo-DTR line 57 treated with 5ng/g bw DT and ±ACEi captopril (200mg/L). Glomerular area of 50 consecutive GCS was measured using ImageJ 1.4r software. The results showed no significant difference in mean glomerular area between the WT DT+ACEi, Tg DT+H2O and Tg DT+ACEi groups (group mean 9676, 10708, 9693µm2 respectively), p>0.08.\n\nDT, Diphtheria toxin; ACEi, angiotensin converting enzyme inhibitor; GCS, glomerular cross section; WT, wild type; Tg, transgenic.\n\nClick here to access the data.\n\nRaw gel image for Figure 1b.\n\nClick here to access the data.\n\n\nReferences\n\nKriz W, Gretz N, Lemley KV: Progression of glomerular diseases: is the podocyte the culprit? Kidney Int. 1998; 54(3): 687–697. PubMed Abstract \| Publisher Full Text\n\nSanden SK, Wiggins JE, Goyal M, et al.: Evaluation of a thick and thin section method for estimation of podocyte number, glomerular volume, and glomerular volume per podocyte in rat kidney with Wilms’ tumor-1 protein used as a podocyte nuclear marker. J Am Soc Nephrol. 2003; 14(10): 2484–2493. PubMed Abstract \| Publisher Full Text\n\nTryggvason K, Patrakka J, Wartiovaara J, et al.: Hereditary proteinuria syndromes and mechanisms of proteinuria. N Engl J Med. 2006; 354(13): 1387–1401. PubMed Abstract \| Publisher Full Text\n\nBoute N, Gribouval O, Roselli S, et al.: NPHS2, encoding the glomerular protein podocin, is mutated in autosomal recessive steroid-resistant nephrotic syndrome. Nat Genet. 2000; 24(4): 349–354. PubMed Abstract \| Publisher Full Text\n\nKestila M, Lenkkeri U, Mannikko M, et al.: Positionally cloned gene for a novel glomerular protein--nephrin--is mutated in congenital nephrotic syndrome. Mol Cell. 1998; 1(4): 575–582. PubMed Abstract \| Publisher Full Text\n\nShih NY, Li J, Karpitskii V, et al.: Congenital nephrotic syndrome in mice lacking CD2-associated protein. Science. 1999; 286(5438): 312–315. PubMed Abstract \| Publisher Full Text\n\nShih NY, Li J, Cotran R, et al.: CD2AP localizes to the slit diaphragm and binds to nephrin via a novel C-terminal domain. Am J Pathol. 2001; 159(6): 2303–2308. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWharram BL, Goyal M, Wiggins JE, et al.: Podocyte depletion causes glomerulosclerosis: diphtheria toxin-induced podocyte depletion in rats expressing human diphtheria toxin receptor transgene. J Am Soc Nephrol. 2005; 16(10): 2941–2952. PubMed Abstract \| Publisher Full Text\n\nIchikawa I, Ma J, Motojima M, et al.: Podocyte damage damages podocytes: autonomous vicious cycle that drives local spread of glomerular sclerosis. Curr Opin Nephrol Hypertens. 2005; 14(3): 205–210. PubMed Abstract \| Publisher Full Text\n\nGross ML, Ritz E, Schoof A, et al.: Renal damage in the SHR/N-cp type 2 diabetes model: comparison of an angiotensin-converting enzyme inhibitor and endothelin receptor blocker. Lab Invest. 2003; 83(9): 1267–1277. PubMed Abstract \| Publisher Full Text\n\nRemuzzi A, Gagliardini E, Donadoni C, et al.: Effect of angiotensin II antagonism on the regression of kidney disease in the rat. Kidney Int. 2002; 62(3): 885–894. PubMed Abstract \| Publisher Full Text\n\nMa LJ, Nakamura S, Aldigier JC, et al.: Regression of glomerulosclerosis with high-dose angiotensin inhibition is linked to decreased plasminogen activator inhibitor-1. J Am Soc Nephrol. 2005; 16(4): 966–976. PubMed Abstract \| Publisher Full Text\n\nAdamczak M, Gross ML, Krtil J, et al.: Reversal of glomerulosclerosis after high-dose enalapril treatment in subtotally nephrectomized rats. J Am Soc Nephrol. 2003; 14(11): 2833–2842. PubMed Abstract \| Publisher Full Text\n\nLewis EJ, Hunsicker LG, Rodby RA, et al.: A clinical trial in type 2 diabetic nephropathy. Am J Kidney Dis. 2001; 38(4 Suppl 1): S191–194. PubMed Abstract \| Publisher Full Text\n\nPraga M, Gutiérrez E, González E, et al.: Treatment of IgA nephropathy with ACE inhibitors: a randomized and controlled trial. J Am Soc Nephrol. 2003; 14(6): 1578–1583. PubMed Abstract \| Publisher Full Text\n\nFukuda A, Wickman LT, Venkatareddy MP, et al.: Angiotensin II-dependent persistent podocyte loss from destabilized glomeruli causes progression of end stage kidney disease. Kidney Int. 2012; 81(1): 40–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNemeth Z, Kokeny G, Godo M, et al.: Increased renoprotection with ACE inhibitor plus aldosterone antagonist as compared to monotherapies--the effect on podocytes. Nephrol Dial Transplant. 2009; 24(12): 3640–3651. PubMed Abstract \| Publisher Full Text\n\nMitamura T, Higashiyama S, Taniguchi N, et al.: Diphtheria toxin binds to the epidermal growth factor (EGF)-like domain of human heparin-binding EGF-like growth factor/diphtheria toxin receptor and inhibits specifically its mitogenic activity. J Biol Chem. 1995; 270(3): 1015–1019. PubMed Abstract \| Publisher Full Text\n\nMoeller MJ, Sanden SK, Soofi A, et al.: Two gene fragments that direct podocyte-specific expression in transgenic mice. J Am Soc Nephrol. 2002; 13(6): 1561–1567. PubMed Abstract \| Publisher Full Text\n\nSaito M, Iwawaki T, Taya C, et al.: Diphtheria toxin receptor-mediated conditional and targeted cell ablation in transgenic mice. Nat Biotechnol. 2001; 19(8): 746–750. PubMed Abstract \| Publisher Full Text\n\nDuffield JS, Tipping PG, Kipari T, et al.: Conditional ablation of macrophages halts progression of crescentic glomerulonephritis. Am J Pathol. 2005; 167(5): 1207–1219. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZheng Z, Pavlidis P, Chua S, et al.: An ancestral haplotype defines susceptibility to doxorubicin nephropathy in the laboratory mouse. J Am Soc Nephrol. 2006; 17(7): 1796–1800. PubMed Abstract \| Publisher Full Text\n\nGuo J, Ananthakrishnan R, Qu W, et al.: RAGE mediates podocyte injury in adriamycin-induced glomerulosclerosis. J Am Soc Nephrol. 2008; 19(5): 961–972. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInokuchi S, Shirato I, Kobayashi N, et al.: Re-evaluation of foot process effacement in acute puromycin aminonucleoside nephrosis. Kidney Int. 1996; 50(4): 1278–1287. PubMed Abstract \| Publisher Full Text\n\nDijkman HB, Weening JJ, Smeets B, et al.: Proliferating cells in HIV and pamidronate-associated collapsing focal segmental glomerulosclerosis are parietal epithelial cells. Kidney Int. 2006; 70(2): 338–344. PubMed Abstract \| Publisher Full Text\n\nel Nahas AM, Bassett AH, Cope GH, et al.: Role of growth hormone in the development of experimental renal scarring. Kidney Int. 1991; 40(1): 29–34. PubMed Abstract \| Publisher Full Text\n\nMatsusaka T, Xin J, Niwa S, et al.: Genetic engineering of glomerular sclerosis in the mouse via control of onset and severity of podocyte-specific injury. J Am Soc Nephrol. 2005; 16(4): 1013–1023. PubMed Abstract \| Publisher Full Text\n\nKlahr S, Levey AS, Beck GJ, et al.: The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. Modification of Diet in Renal Disease Study Group. N Engl J Med. 1994; 330(13): 877–884. PubMed Abstract \| Publisher Full Text\n\nRosenberg ME, Smith LJ, Correa-Rotter R, et al.: The paradox of the renin-angiotensin system in chronic renal disease. Kidney Int. 1994; 45(2): 403–410. PubMed Abstract \| Publisher Full Text\n\nBaines RJ, Brunskill NJ: The molecular interactions between filtered proteins and proximal tubular cells in proteinuria. Nephron Exp Nephrol. 2008; 110(2): e67–71. PubMed Abstract \| Publisher Full Text\n\nGassler N, Elger M, Kränzlin B, et al.: Podocyte injury underlies the progression of focal segmental glomerulosclerosis in the fa/fa Zucker rat. Kidney Int. 2001; 60(1): 106–116. PubMed Abstract \| Publisher Full Text\n\nMatsusaka T, Sandgren E, Shintani A, et al.: Podocyte injury damages other podocytes. J Am Soc Nephrol. 2011; 22(7): 1275–1285. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShibata S, Nagase M, Fujita T: Fluvastatin ameliorates podocyte injury in proteinuric rats via modulation of excessive Rho signaling. J Am Soc Nephrol. 2006; 17(3): 754–764. PubMed Abstract \| Publisher Full Text\n\nRansom RF, Lam NG, Hallett MA, et al.: Glucocorticoids protect and enhance recovery of cultured murine podocytes via actin filament stabilization. Kidney Int. 2005; 68(6): 2473–2483. PubMed Abstract \| Publisher Full Text\n\nFaul C, Donnelly M, Merscher-Gomez S, et al.: The actin cytoskeleton of kidney podocytes is a direct target of the antiproteinuric effect of cyclosporine A. Nat Med. 2008; 14(9): 931–938. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBhimma R, Adhikari M, Asharam K, et al.: Management of steroid-resistant focal segmental glomerulosclerosis in children using tacrolimus. Am J Nephrol. 2006; 26(6): 544–551. PubMed Abstract \| Publisher Full Text\n\nNakajo A, Khoshnoodi J, Takenaka H, et al.: Mizoribine corrects defective nephrin biogenesis by restoring intracellular energy balance. J Am Soc Nephrol. 2007; 18(9): 2554–2564. PubMed Abstract \| Publisher Full Text\n\nZhou YS, Turner AN: New ways of thinking about proteinuria and progression of renal damage. Nephron Exp Nephrol. 2010; 116(1): e1–2. PubMed Abstract \| Publisher Full Text\n\nBuelli S, Rosanò L, Gagliardini E, et al.: β-arrestin-1 drives endothelin-1-mediated podocyte activation and sustains renal injury. J Am Soc Nephrol. 2014; 25(3): 523–533. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDhaun N, Ferro CJ, Davenport AP, et al.: Haemodynamic and renal effects of endothelin receptor antagonism in patients with chronic kidney disease. Nephrol Dial Transplant. 2007; 22(11): 3228–3234. PubMed Abstract \| Publisher Full Text\n\nKanjanabuch T, Ma LJ, Chen J, et al.: PPAR-gamma agonist protects podocytes from injury. Kidney Int. 2007; 71(12): 1232–1239. PubMed Abstract \| Publisher Full Text\n\nMiyazaki Y, Cersosimo E, Triplitt C, et al.: Rosiglitazone decreases albuminuria in type 2 diabetic patients. Kidney Int. 2007; 72(11): 1367–1373. PubMed Abstract \| Publisher Full Text\n\nVaughan MR, Pippin JW, Griffin SV, et al.: ATRA induces podocyte differentiation and alters nephrin and podocin expression in vitro and in vivo. Kidney Int. 2005; 68(1): 133–144. PubMed Abstract \| Publisher Full Text\n\nLogar CM, Brinkkoetter PT, Krofft RD, et al.: Darbepoetin alfa protects podocytes from apoptosis in vitro and in vivo. Kidney Int. 2007; 72(4): 489–498. PubMed Abstract \| Publisher Full Text\n\nPippin J, Kumar V, Stein A, et al.: The contribution of podocytes to chronic allograft nephropathy. Nephron Exp Nephrol. 2009; 111(1): e1–10. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9900", "date": "24 Aug 2015", "name": "Taiji Matsusaka", "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\nIn this study, the authors established a new mouse model of inducible podocyte injury. They also showed that ACEi treatment effectively attenuated proteinuria and glomerulosclerosis in this model. Similar to previous reported models, this model also reproducibly develops progressive glomerulosclerosis. The data are confirmative, but this model will be an alternative useful model for podocyte injury. Compared to hCD25 transgenic mouse model (NEP25), podocyte injury is relatively mild. The time course of ACR in line 47 (Figure 2a) appears to be similar to that of NEP25 injected with 0.625 ng/g BW of LMB2. Generally, the study is well done, and the data are properly presented. There are only some minor concerns.It would be better to present HB-EGF mRNA or protein expression data. Is the transgene actually expressed selectively in podocytes? Is the sensitivity of toxin parallel to transgene expression among three lines? Data among three or more groups should be analyzed by ANOVA or some other methods to reduce type 1 error. Generally, urinary ACR shows logarithmic normal distribution. These data should be properly transformed before analysis. Figure 5b data should be analyzed by mixed effect model or other appropriate methods. In wild-type C57BL/6 mice younger than 10 months of age, glomerulosclerosis is very rare. The wild-type control mice treated with ACEi, 7% of glomeruli showed segmental sclerosis. Is this senile change? If so, more detailed information of mouse age would be presented, or consider deleting aged mice from the experiment. Please indicate what the bottom and top of the box and whisker represent in Figure 8. The authors state that ACEi was not advantageous for podocyte preservation. I do not agree on this statement. It is just a matter of statistical power. The protective effect by ACEi is small relatively to the variance of podocyte number in each group.", "responses": [] }, { "id": "10200", "date": "26 Oct 2015", "name": "Johannes Schlondorff", "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 their manuscript, the authors adapt the use of podocyte-specific ablation using transgenic diphtheria toxin receptor (DTR) expression to the mouse. The murine model functions similarly to the previously described rat model of Wiggens et al. Using the model, the authors examine the role of ACE inhibition (captopril) as a protective therapy. Overall, the conclusions are well supported by the experimental results, though the experiments are neither as detailed nor as compelling as those previously published by others. The two major conclusions of the manuscript, that podocyte depletion can lead to the initiation of a feed-forward podocyte pathology and that ACE inhibition can slow or ameliorate this process, have been well documented by others.The major concerns about the interpretation of the results stem from the fact that: 1. two different transgenic lines are used without clear side-by-side comparison of both lines for the most relevant experiments; and 2. the animals used in the captopril experiment are mixed sex and of a very large range of ages (3-12 months), likely leading to the large variability in the data, limiting power, and making it difficult to lend much credence to the negative results regarding captopril’s effect on podocyte numbers. Analysis of the data using matching for age and sex, if possible, would be of interest. Overall, the 57 transgenic lines used in the ACEi experiment show only very modest glomerular sclerosis compared to controls (Figure 6), and the control animals in this experiment have surprisingly lower number of podocytes per glomerulus (8.2; figure 8) compared to control animals in an earlier experiment (10.0, figure 4). Some explanation should be provided. Minor points:Figure 3 legend refers to p values, body weight to which there are no corresponding data presented. Figure 5a: Is there a difference between the “++” and “+++”? For figure 5b, the lines connecting the time points should be removed. In addition, including individual values and means (as in other figures) or at least including the n’s would be helpful. Number of animals/samples should be consistently provided in the figures (eg figure 8, 5b).", "responses": [] }, { "id": "10890", "date": "02 Nov 2015", "name": "Gavin I. Welsh", "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 an interesting and well-presented paper that uses transgenic (Podo-DTR) mice to demonstrate (1) that continuing podocyte dysfunction is a key abnormality in proteinuric disease and (2) following specific podocyte injury captopril protects against progressive long-term renal damage. However there are a number of issues with the manuscript.Three transgenic lines with positive PCR results for the transgene were established (Podo-DTR 47, 57 and 21) by the authors two of which (47 and 57) were susceptible to the DT. However instead of comparing the 2 lines with wild type mice, different lines are used for different experiments. The model characterization uses mainly Line 47 but experiments using captopril are done exclusively in Line 57. What happens to mice from Line 47 treated with captopril-does it have the same protective effect to that seen in Line 57? The authors should confirm that the DTR protein is indeed exclusively expressed in the podocyte. The authors do not explain why there is such a wide spread in the ages of the mice (3-12 months) used. Does not ageing introduce unneeded variability into the study that may confound some of the effects seen? There is variability in the doses of DT used in each experiment that makes the results difficult to compare. For example in figure 5 for tail cuff blood pressure and urine albumin:creatine ratio 1ng/g bw was used but for glomerulosclerosis and kidney histology studies animals were treated with 5ng/g bw. Furthermore kidney histology for Line 57 in figure 3 uses 1ng/g bw but in figure 7 it is 5ng/g bw. Why the lack of consistency? In figure 4 control mice for PodDTR line 47 are reported to have 10 podocytes/GCS whilst those for PodDTR line 47 in figure 8 have 8.2 podocytes/GCS. Why the discrepancy between the two lines? At what age were the samples taken for figure 6 and 7 and what dose of DT was used in figure 8?", "responses": [] } ]	1	https://f1000research.com/articles/4-172
https://f1000research.com/articles/4-170/v1	26 Jun 15	{ "type": "Research Article", "title": "Characterization of an APC Promoter 1B deletion in a Patient Diagnosed with Familial Adenomatous Polyposis via Whole Genome Shotgun Sequencing", "authors": [ "Ted Kalbfleisch", "Pamela Brock", "Angela Snow", "Deborah Neklason", "Gordon Gowans", "Jon Klein", "Pamela Brock", "Angela Snow", "Deborah Neklason", "Gordon Gowans", "Jon Klein" ], "abstract": "Recently, deletions have been identified and published as causal for Familial Adenomatous Polyposis in the 1B promoter region of the APC gene. Those deletions were measured using multiplex ligation-dependent probe amplification. Here, we present and characterize an ~11kb deletion identified by whole genome shotgun sequencing. The deletion occurred in a patient diagnosed with Familial Adenomatous Polyposis, and was located on chr5, between bases 112,034,824 and 112,045,845, fully encompassing the 1B promoter region of the APC gene.\n\nResults are presented here that include the sequence evidence supporting the presence of the deletion as well as base level characterization of the deletion site. These results demonstrate the capacity of whole genome sequencing for the detection of large structural variants in single individuals.", "keywords": [ "APC", "Familial Adenomatous Polyposis", "Clinical Sequencing", "Next Generation Sequencing" ], "content": "Introduction\n\nFamilial Adenomatous Polyposis (FAP) is an autosomal dominant condition characterized by the development of hundreds to thousands of polyps in the colon. This condition results in colon cancer in adult individuals in their late 20s to early 30s with nearly 100 percent penetrance. Mutations in two genes, the adenomatous polyposis coli (APC) and mutY homolog (MUTYH) loci, have been identified as causative for this disease. The majority of the mutations occur in the APC locus. The APC mutations often take the form of single nucleotide substitutions or small insertions or deletions in the coding region of the gene that produce premature stop codons, or frame shifts respectively. These result in a change of function. The exact mechanism by which these mutations affect the disease is unknown. However, deletions of APC promoter 1B are known to cause a significant change in transcription levels of the APC RNA marked by allele specific differences in transcription1,2. Several mutations have been reported in the promoter region of the APC gene1–4, identified either by sequencing, or by multiplex ligation-dependent probe amplification (MLPA).\n\nThe patient analyzed in this work is a 50 year-old Caucasian female who has a personal and maternal family history of FAP. She developed colon polyps at 14 years of age and underwent a partial colectomy at 16 years. The patient had a complete colectomy and a Whipple procedure in her 20’s. Her mother and multiple avunculars and cousins on the maternal side are affected. One sibling has a clinical diagnosis of FAP and three siblings are unaffected. The patient’s maternal grandfather died of colon cancer later in life, but a diagnosis of FAP was not confirmed.\n\nPreviously, DNA testing in family members had failed to identify a causative mutation. Therefore, the patient and her family participated in a linkage analysis project through the Mayo Clinic in Rochester, Minnesota to identify at-risk family members. The FAP in the family showed linkage to the APC locus on chromosome 5. The patient underwent molecular testing of the APC gene (sequence analysis and Southern blot) and MUTYH gene (analysis for 2 common mutations) in 2008. No mutations were detected. A variant of unknown significance (referred to as Glu1317Gln) was found in the APC gene. However, this variant was absent in other affected family members and was present in the patient’s unaffected child. It was later classified as likely benign5. The multiplex ligation-dependent probe amplification (MLPA) assays for the APC locus in use at the time did not characterize the APC promoters, and was negative for APC mutations for this patient.\n\nIn an effort to comprehensively search for potential mutations, the patient’s genomic DNA was sent to Illumina whole genome sequencing. A deletion of ~11kb encompassing the APC promoter 1B was identified, and is consistent with the deletion identified recently by Snow et al.2 via an updated MLPA assay for APC that now includes promoter 1B and by Lin et al.4.\n\nIn this work, we present a comprehensive characterization of this deletion using Illumina short reads, including base level resolution of the deletion site. Further, it is demonstrated that this deletion is detectable using the MLPA assay for the APC locus current at the submission of this article, and would be ambiguous if this, or any single patient were analyzed solely via whole exome sequencing.\n\n\nMethods\n\nThe whole blood sample for this study was collected under a protocol approved by the University of Louisville IRB (IRB tracking number 11.0659, approval date 1/30/2012). Written informed consent for publication of clinical details was obtained from the patient/next of kin. The blood was sent to the Illumina Clinical Services Laboratory for paired end sequencing of 100 bp reads from fragments with a target length of 300bp. The reads produced were mapped via CASAVA (CASAVA-1.9.0a1_110909) to the human reference genome build 37.1 at an average depth of coverage 37.51X.\n\nThe pipeline employed in our lab for read mapping and variant detection uses the Burrows-Wheeler Alignment6 algorithm, and the Genome Analysis Toolkit7 respectively. To be consistent with other work in our lab, reads for the regions of interest were extracted from the bam file produced by Illumina, and run through our pipeline.\n\nMapped reads were extracted from the binary alignment map file for remapping using Samtools8 version 0.1.18 from the individual’s full binary alignment map file (provided by Illumina) corresponding to 50,000 bases upstream and downstream of the APC, and MUTYH loci defined respectively by the mapping of accession NM_001127511.1 and NM_001293192.1 to human genome build 37.1 (chr5:111,993,219-112,231,936 and chr1:45,744,915-45,856,143). Reads mapping to other chromosomes, or positions on chromosomes 1 and 5 outside of the target region would have also been extracted if their mate mapped within the target regions. Reads in these extraneous regions were not considered in variant detection. To be consistent with the remainder of our work, the FASTQ files corresponding to the first and second reads of the pair (R1 and R2) were re-derived via BEDTools9 from the BAM file provided by Illumina, and remapped using the BWA algorithm for short read alignment. Duplicates were marked, indels were realigned, base quality scores recalibrated, and variants identified and simultaneously genotyped for our trace data by applying the GATK MarkDuplicates, IndelRealigner, BaseRecalibrator, and HaplotypeCaller algorithms respectively10,11.\n\nThe deletion was identified by visual inspection within the Integrative Genomics Viewer (IGV)12 of the mapped next generation sequence data set as well as the variation reported in the accompanying variant call format file. This deletion is characterized by a loss of heterozygosity of variants measured relative to the reference, a cluster of 11 paired end reads (target length 500 bases) whose mates map in excess of 11kb from one another, as well as 15 reads that span the junction of the deletion that were soft trimmed by the mapping algorithm. The option “Show soft-clipped bases” within View/Preferences/Alignments was turned on and revealed soft trimming that began in several reads at positions 112,034,824 and 112,045,845 on chromosome 5. Bases from these reads were copied from within the IGV user interface for subsequent analysis in BLAT13 to confirm the position of the deletion.\n\nPrimers were designed to specifically interrogate this deletion with one primer pair flanking the deletion, and one primer pair with one primer located in the deleted region. The primer located 3’ to the deleted region was common to both pairs. Full description of the primers is in provided in Table 1.\n\nWhole blood was fractionated by spinning at 5,000 rpm for 10 minutes at room temperature. White cells were transferred to sterile, nuclease free microcentrifuge tubes and stored at -20°C until processing. Genomic DNA was isolated from 250uL buffy coat with Gentra Puregene Genomic DNA purification buffers (Qiagen, Valencia, CA). Separate amplification of the wild type or deletion APC fragments were performed in a 20uL reaction containing 0.4uL Phusion HF DNA Polymerase (Thermo Fisher Scientific, Pittsburg, PA), 1x Phusion Reaction Buffer, 200uM dNTP’s (Promega Corporation, Madison, WI), 200ng gDNA, and 0.5uM each primer. The cycling conditions were as follows: 98°C for 30s followed by 35 cycles of 98°C for 10s, 60°C for 30s, and 72°C for 60s, ending with a final extension of 72°C for 7min.\n\nThe amplicons were sequenced with BigDye® Terminator v3.1 (Life Technologies Corporation, Carlsbad, CA) utilizing the PCR primers and standard sequencing conditions. The sequence reactions were purified with Performa DTR Ultra 96-well filtration plates (Edge Biosystems, Gaithersburg, MD) and processed on the ABI 3130xl Genetic Analyzer (Life Technologies Corporation, Carlsbad, CA).\n\nThe resulting gel for the PCR products is shown in Figure 1, and the sequencing results are shown in Figure 2, rendered in Geospiza’s FinchTV, (http://www.geospiza.com/Products/finchtv.shtml).\n\nThe relationship between the lanes and the nine kindreds defined in Snow et al. are Lane 1: Ladder, Lane 2: Kindred 8, Lane 3: Kindred 43, Lane 4: Kindred 44, Lane 5: Kindred 256, Lane 6: Kindred 509. Lane 7: Kindred 685, Lane 8: Kindred 691, Lane 9: Kindred 353 (APC c.426_427delAT) And Lane 10: Kindred 6699 (APC c.532–941G>A). These images demonstrate heterozygous deletions in eight of the samples analyzed. PCR products corresponding to the bands circled in red were sequenced using Sanger technology. Those results are shown in Figure 2.\n\nThe top two traces indicate the nucleotide sequence of the wild type APC locus in the Promoter 1B region, and the deletion site. The bottom trace demonstrates that the deletion detected by Snow et al. is identical to the deletion to the individual described in this work.\n\n\nResults\n\nPaired end whole genome sequence data was generated at ~40X coverage for the patient, and mapped to the human reference assembly Build-37.1. Given the clinical phenotype our initial analysis of the data was limited to the APC and MUTYH loci. Variation analysis was performed in the region defined by the 5’ and 3’ most exons of the longest reported transcript for APC and MUTYH, plus and minus 50,000 bases respectively (described in detail in Methods). The resulting counts of single nucleotide variations (SNVs) and small indels are shown in Table 2–Table 4. The corresponding VCF file, along with mapped reads for these regions are available for download or visualization at http://dx.doi.org/10.13013/J6QN64N8. When viewing in IGV, navigate to the APC locus by entering APC in the text box at the top of the frame.\n\nSingle Nucleotide Polymorphism database (dbSNP) accession numbers and Human Genome Variation Society (HGVS) names for the gene, including the amino acid change and position are also listed.\n\nAll missense variants identified had corresponding records in dbSNP and are listed in Table 4. None are reported as deleterious. There were no non-sense SNVs or frame shifting small insertions or deletions identified. The search was then turned toward larger structural variants. Visual inspection of the VCF file for the APC locus revealed a region of approximately 10kb with 17 measured SNVs or small insertions relative to the reference. None of their respective genotypes were classified as heterozygous. This loss of heterozygosity suggested a deletion. Upon further inspection, there were other signatures characteristic of a deletion, that included a cluster of paired end reads whose mate mapped ~11kb from their respective start, and several mates that were soft trimmed because they spanned the deletion site. These soft trimmed mates were identified (described in methods), and aligned via BLAT to hsBuild-37.1, revealing the deleted region to be of length 11,020 bases, located on chr5, between bases 112,034,824 and 112,045,845, spanning the annotated APC promoter 1B. This deletion is illustrated in Figure 3, along with the positions of commercial probe sets, and other annotation relevant to this work. Given that this deletion was consistent with the deletion reported by Snow et al., the primers used for verification in this work, were run on the kindreds studied in that work. It was verified that the deletion reported there was identical to the one reported here. Also, this deletion is identical to a deletion published by Lin et al.,4 identified in kindreds from Missouri, Illinois, and Idaho not known to be related to each other.\n\nRecords from the VCF file for the patient described here are displayed in the top track indicating a region with a loss of heterozygosity consistent with a deletion. We also render the exon identified as APC promoter 1B, the MLPA probes used commercially to analyze this locus, the region selected for pull-down in the TruSeq exome capture kit, the deletion reported by Rohlin et al. in 2008, and the position of the deletion described here relative to all these features.\n\nThe Illumina paired end short read data that provides evidence for the deletion relative to the reference has been isolated from the larger dataset, and is made available in its own binary alignment map file for inspection at the DOI included above.\n\nIn order to confirm the deletion, PCR primers were designed to specifically interrogate it. These primers produce a product of approximately 1kb for individuals with no deletion, and a second pair of primers was designed that flank the deletion site. This placement produces a product of 0.6kb from chromosomes with the deletion, and 11.7kb in chromosomes without. As the NGS data suggests a heterozygous deletion, the expectation was a single band with the first primer pair, and two bands, one strong from the 0.6kb amplicon, and one weak (if detectable) for the 12kb amplicon. This was confirmed in the gel represented in Figure 1. The ~1kb and .6kb bands were cut from the gel and sequenced using Sanger technology. The trace images are shown for the two different alleles in Figure 2. One read shows the deletion, and the second allele is consistent with the reference. The deletion is confirmed by the Sanger sequence data, and the primers are provided as a definitive Sanger sequencing assay for it. The second PCR image in Figure 1, and third read included in Figure 2 confirmed that our respective kindred shares the same deletion as the seven families reported by Snow et al. We predict that all families descend from a common founder.\n\nAlthough this deletion was identified by visual inspection, the binary alignment map file for the region was analyzed by the application BreakDancer14 to determine if the deletion could be identified algorithmically from whole genome sequence data. BreakDancer identifies putative deletions by identifying read pairs, clustered by genomic coordinate, that have similar inferred insert sizes which are either much larger or smaller than the standard distribution of insert sizes measured for mapped pairs. Using this algorithm, a deletion was identified on chr5 and was approximated to lie between bases 112,034,793 and 112,045,844, corroborating the finding presented here.\n\nThe methods of Snow et al. used multiplex ligation-dependent probe amplification (MLPA) assays. These are described in a document from MRC-Holland, available at the time of publication at (http://www.mlpa.com/WebForms/WebFormDBData.aspx?FileOID=McLO2Mc0V%5Cc%7C). Information for those probes, including the partial sequence adjacent to the ligation site, as well as the genomic coordinate derived from a BLAT search using the partial sequence information is reproduced in Table 5, and rendered in Figure 3 relative to the deletion identified in this work. These coordinates are contained within the region deleted for this patient, and as such result in a deletion of the signals corresponding to these probes. The next probe in the set, APC 142, which is outside the deleted region, did not indicate a deletion.\n\n\nDiscussion/Conclusion\n\nSeveral years ago, a female patient of the University of Louisville Weisskopf Child Evaluation Center presented with Familial Adenomatous Polyposis (FAP). Whole genome shotgun sequencing on the Illumina platform revealed a deletion on chromosome 5 between bases 112,034,824 and 112,045,845, fully encompassing promoter 1B of the APC locus. Deletions that include this promoter have been demonstrated to affect the expression of the full length APC transcript.\n\nIn other work by Snow et al., a deletion was identified via MLPA that is consistent with the deletion characterized here. An investigation via PCR of their seven kindreds with the primers used in this work establishes that the deletion is identical to the deletion reported here. Furthermore, this deletion is also reported by Lin et al., in three kindred not known to be related to each other, or these families. It is likely that this mutation descends from an ancestor common to each of these reported families.\n\nExome capture has become a popular tool for mutation screening in clinical genetics. The deletion reported here extends several kilobases beyond the region captured by one of the more popular exome capture products (Figure 3). This deletion would have been very difficult to identify by exome capture since the only practical measurements that could have been employed would have been read density and loss of heterozygosity in the captured region.\n\nThe whole genome sequencing approach taken here produces an information rich dataset capable of resolving large deletions in individuals. These structural variants result in a number hallmarks that are easily detected. Specifically, the loss of heterozygosity over a large region, a collection of read pairs whose mates consistently map much further apart than the majority of the read pairs, and soft trimmed reads all pinpoint the deletion site unequivocally. We have demonstrated that whole genome sequencing is both a sensitive and accurate approach for the detection and characterization of deletions of this size.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw Gel electrophoresis image for Figure 1, 10.5256/f1000research.6636.d5027615", "appendix": "Author contributions\n\n\n\nTK and JK conceived of and led the project. TK performed primary and secondary data analyses, and wrote the manuscript. PB and GG served as clinical liaisons. AS and DN performed PCR and background on their sample sets. All provided input during the preparation of the manuscript. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nT.K. serves as the CEO of Intrepid Bioinformatics.\n\n\nGrant information\n\nThe Next Generation Sequencing work was supported by DOE grant DE-EM0000197 (Kalbfleisch, Rouchka co-PI). Dr. Kalbfleisch received additional financial support from the National Institute of General Medical Sciences of the National Institutes of Health Grant# P20GM103436 (Cooper PI). University of Utah work is supported by PO1CA073992.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe PCR and Sanger Sequencing work were performed in the UofL Center for Genetics and Molecular Biology core facility by Ms. Elizabeth Hudson. The alignment and analysis work for the next generation sequencing data was performed on the University of Louisville Cardinal Research Cluster.\n\n\nReferences\n\nRohlin A, Engwall Y, Fritzell K, et al.: Inactivation of promoter 1B of APC causes partial gene silencing: evidence for a significant role of the promoter in regulation and causative of familial adenomatous polyposis. Oncogene. 2011; 30(50): 4977–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSnow AK, Tuohy TM, Sargent NR, et al.: APC promoter 1B deletion in seven American families with familial adenomatous polyposis. Clin Genet. 2014. PubMed Abstract \| Publisher Full Text\n\nKadiyska TK, Todorov TP, Bichev SN, et al.: APC promoter 1B deletion in familial polyposis--implications for mutation-negative families. Clin Genet. 2014; 85(5): 452–7. PubMed Abstract \| Publisher Full Text\n\nLin Y, Lin S, Baxter MD, et al.: Novel APC promoter and exon 1B deletion and allelic silencing in three mutation-negative classic familial adenomatous polyposis families. Genome Med. 2015; 7(1): 42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKerr SE, Thomas CB, Thibodeau SN, et al.: APC germline mutations in individuals being evaluated for familial adenomatous polyposis: a review of the Mayo Clinic experience with 1591 consecutive tests. J Mol Diagn. 2013; 15(1): 31–43. PubMed Abstract \| Publisher 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\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): 1297–303. 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\nQuinlan AR, Hall IM: BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010; 26(6): 841–2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDePristo MA, Banks E, Poplin R, et al.: A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet. 2011; 43(5): 491–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVan der Auwera GA, Carneiro MO, Hartl C, et al.: From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best practices pipeline. Curr Protoc Bioinformatics. 2013; 11(1110): 11.10.1–11.10.33. 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): 24–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKent WJ: BLAT--the BLAST-like alignment tool. Genome Res. 2002; 12(4): 656–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen K, Wallis JW, McLellan MD, et al.: BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat Methods. 2009; 6(9): 677–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKalbfleisch T, Brock P, Angela S, et al.: Dataset 1 in: Characterization of an APC Promoter 1B deletion in a Patient Diagnosed with Familial Adenomatous Polyposis via Whole Genome Shotgun Sequencing. F1000Research. 2015. Data Source" }	[ { "id": "9222", "date": "08 Jul 2015", "name": "Nicholas Davidson", "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\nMutations in the APC gene have been established as a cause of classical FAP. However, in a small subset (~20%) of affected families, mutations in the coding regions of APC or other polyposis-associated genes such as MUTYH cannot be identified. In this study, the authors perform whole-genome sequencing on one such subject and identify a heterozygous 11kb deletion in the exon 1B / promoter region of APC. The three non-synonymous mutations identified in APC and MUTYH were not felt to be causal mutations. The authors confirmed the presence of the heterozygous deletion through differential PCR and Sanger sequencing.The authors then went on to investigate nine APC mutation-negative kindreds which were previously found to have exon 1B / promoter deletions using multiplex ligation-dependent probe amplification assays. In all nine kindreds, the same coordinates of the 11kb deletion were inferred.The study design, presentation of results and conclusions are straightforward. The authors find an APC promoter deletion through whole-genome sequencing and establish the deletion coordinates through Sanger sequencing. Ideally, further investigation of a number of affecteds and unaffected members in the kindred of this study would provide convincing evidence that this 11kb deletion is associated with the affected state. However, considered in conjunction with the results from Lin et al. and Snow et al., this study shows that whole-genome sequencing is a suitable method for the detection of non-coding mutations in APC-mutation negative FAP individuals. This builds upon mounting evidence that associates APC promoter deletions with FAP. Interestingly, these studies all show that affected members of eleven FAP kindreds in the United States share the promoter deletion with identical coordinates. It might be worth noting in the discussion that the original Snow et al paper using MPLA identified a deletion that was thought to be much larger (>33kb) but that approach did not map the exact coordinates.Two minor points: (1) In the Results section (p 6 left column second paragraph), it is stated: “these primers produce a product of approximately 1kb for individuals with no deletion.” More accurately, the 1kb product is produced from chromosomes without the deletion (as the authors later state); the 1kb product is also produced in all individuals with the deletion (Fig. 1).(2) In the discussion, the authors discuss potential difficulties in using targeted capture techniques to discover or assess larger deletions such as the one described. However, we successfully used such a targeted capture assay to discover the promoter deletion described in our study. The analysis methods for detecting individual reads and read pairs that straddle the deletion are applicable and to targeted sequencing strategies.", "responses": [] }, { "id": "9224", "date": "15 Jul 2015", "name": "Anna Rohlin", "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\nKalbfleisch et al. present an interesting article on identification and characterization of an 11kb APC promoter deletion from whole genome sequencing (WGS) data. The design of the study, the methods used and the presentation of the results are presented in a conclusive and way suitable to the investigation. They have analyzed the WGS data only for the APC and MUTYH locus and identified the deletion by visual inspection of the data. The identified deletion is confirmed and breakpoints identified with sanger sequencing. The deletion is also found to be identical to the previously identified deletions by Snow et al. and Lin et al., and they all descent from a common founder.Minor points for revision and comments:When a coding DNA reference sequence is used the following recommendations can be followed according to HGVS; the coding DNA reference sequence should be complete and preferably derived from the RefSeq database (format NM_033337.2). In general the longest transcripts are used, and for the MUTYH gene this is NM_001128425.1 (16 exon 549 aa). It would be more appropriate to at least name the variants found according to the NM reference sequence that has been used in table 4. For example the variant rs3219484 if the NM_001128425.1 would have been used the variant would preferable be named NM_001128425.1:c.64A>G, p.Val22Met. For the APC gene NM_000038.5 is usually used. In the tables and text when discussing genomic coordinates, the genome build used would be good too include for example (GRCh37/hg19). When discussing MLPA, it would be appropriate to mentioned the versions of the MLPA kit, like P043 (version C1), this is a common way to present which kit has been used, a link to the document describing the assay is not necessary. Probes for the APC promoter 1B were included in the MLPA kit in 2011. In the result section findings of missense and nonsense variants in coding region and small indels are mentioned, what about variants in splice acceptor or donor-sites are these included in the analyses done? It would be nice to mention this in the text and also in the tables (2 and 4) if anyone’s are found as these variants are important and often constitute disease-causing mutations. Regarding the deleteriousness of the variants found how was this interpreted? It would be nice to mentioned which databases and/or prediction tools has been used since all variants in dbSNP are not benign. I would be recommended to use several prediction tools like for example SIFT, Polyphen-2, Mutation taster, Condel and Combined Annotation-Dependent Depletion (CADD) among others for missense interpretation and also looking at conservation between species.Other important tools used for interpretation of variants found are databases were variants are reported and sometimes also classified often in a 1-5 scale (1 is benign and 5 pathogenic). The InSiGHT database (http://insight-group.org/variants/database/) is commonly used for variants in coloncancer genes, where the APC and MUTYH gene can be found among others, clinvar and HGMD professional (Human Genome Mutation Database) are also databases too use. Information about the minor allele frequency of the variant found can also be found in ExAc (Exome Aggregation Consortium) and ESP (Exome Sequencing Project) and can be included when reporting variants. This deletion was found by manual inspection of the region were the reads maps further apart than expect and by looking at the soft-clipped bases as well as identifying the region having LOH. It would be interesting to discuss the limitation with this visual method regarding sizes and type of rearrangement that can be detected. Some discussion comparing (pros and cons) regarding this method with different algorithms methods like BreakDancer (read-pair methods) and others including read-depth methods and split reads methods for example would also be valuable for an interesting discussion useful for readers trying to find methods to analyze for these types of mutations.", "responses": [] } ]	1	https://f1000research.com/articles/4-170
https://f1000research.com/articles/4-169/v1	25 Jun 15	{ "type": "Opinion Article", "title": "Naturopathic medicine: Nine parts negative, one part positive", "authors": [ "Norman Temple" ], "abstract": "Naturopathic medicine, also known as naturopathy, is a type of complementary and alternative medicine. It appeals to many people, especially those who desire a “holistic” approach to both prevention and treatment. While there is much variation in the types of treatment used by different naturopaths, commonly used ones include acupuncture, herbalism, and homeopathy. These types of treatment often lack sound supporting evidence of efficacy. But at the same time naturopaths are often hostile to conventional drugs, even those that are of proven effectiveness and pose little risk of harmful side effects. Many naturopaths employ treatments, such as diet, herbs, fasting, and colonic irrigation that are claimed to “detoxify” the body and thereby lead to improved health. There is a complete absence of supporting evidence for this type of therapy. Some aspects of the theories and practices employed by naturopaths are well supported by the evidence. In particular, the emphasis that naturopaths place on leading a healthy lifestyle so as to prevent disease is entirely consistent with modern concepts in this area. Overall, the positive aspects of naturopathy are greatly outweighed by the negative aspects.", "keywords": [ "Alternative medicine", "Complementary and alternative medicine", "Naturopathy" ], "content": "Introduction\n\nNaturopathic medicine – naturopathy – is a popular type of complementary and alternative medicine (CAM) that has steadily evolved over the last 100 years1. Naturopaths believe that the human body strives toward health and is its own best healer. Naturopaths claim to treat the whole person using natural therapeutics and cures2. However many aspects of naturopathy suffer from a serious lack of solid supporting evidence.\n\nPatients choose to visit naturopaths for a variety of reasons. The most important of these is an attraction to the overall philosophy of naturopathy, including the desire for a more “natural” and “holistic” approach that (supposedly) addresses the root of the problem2. Other reasons include general dissatisfaction with the care provided by conventional health-care providers, wanting more time and attention, and having had a previous positive experience with a naturopath. The patients who most often visit a naturopath are white, middle-aged, female, and have a chronic condition3.\n\nNaturopathic physicians (NDs) are trained as primary-care physicians in four-year, accredited doctoral-level naturopathic medical schools. There are several such schools in the USA and Canada4. The legal status of naturopaths varies between states and provinces. Some jurisdictions permit registered naturopaths to carry out minor surgery, write prescriptions for at least some drugs, give vaccinations, and carry out spinal manipulations.\n\n\nGuiding principles of naturopathy\n\nNaturopathy embraces the concept of prevention which is best accomplished by educating their patients to lead a healthy lifestyle. In that respect naturopathy resembles health promotion. A generally healthy lifestyle is now recognized as the ideal way to prevent many diseases including coronary heart disease, type 2 diabetes, and hypertension5.\n\nNaturopaths take a holistic approach to treatment and thereby aim to treat the whole person4,6. They see this approach as being superior to that of conventional medicine which takes a mechanistic view of disease (reductionism) and then focuses on symptoms. With some diseases a whole-person (holistic) approach can make good sense. For example, coronary disease is the result of a generally unhealthy lifestyle that causes dysfunction in several body systems. A generally healthy lifestyle is effective for not only the prevention of coronary heart disease but also as a treatment5. This is also the case with type 2 diabetes and hypertension5. A holistic approach may also be of value in patients with cancer as they need nutritional support to help the body recover as well as social and emotional support. A holistic approach to coronary disease and cancer is hardly controversial and most cardiologists and oncologists would probably support this, provided, of course, that effective treatments that target the specific disorder are also employed.\n\nAn altogether different story presents itself with other disorders. With arthritis and depression, for example, there can be a specific dysfunction in a single body system. Conventional medicine has treatments of proven effectiveness that target the problem. In these cases treating the whole body will likely lead to poorer outcomes.\n\nOne of the core principles of naturopathic medicine is to use treatments that minimize risk to the patient. Of course, conventional physicians also claim to follow this principle. However, there is no doubt that on many occasions physicians have indeed given inappropriate treatments causing harm. The obvious solution to this is to improve the quality of treatment given by conventional physicians. But we must recognize that harmful side effects are often the price patients must pay in order to receive benefit. For example, cancer therapy often involves the use of drugs, radiation, and/or surgery which while helpful in the treatment of the disease can have harmful side effects.\n\nMany naturopaths believe the suppression of symptoms should be avoided because such an action interferes with the healing process6. This leads to a situation where naturopaths may refuse to give analgesic medications to a patient with arthritis or anti-depressants to one with depression, thereby depriving the patient of potentially helpful treatments6. Moreover, there is no evidence that these drugs prevent the healing of the joints in arthritic patients or impede the brain from normalizing the neurochemical imbalances causing some forms of depression.\n\n\nThe practice of naturopathy\n\nPractitioners of CAM use a wide range of therapies, several of which have been adopted by naturopathy. These include herbalism, homeopathy, acupuncture, hydrotherapy, physical therapy, spinal manipulation, lifestyle counseling, nutrition (including the use of vitamin and mineral supplements), and psychological counseling. Some of these treatments do not stand up to close scrutiny. There is much variation in the eclectic choice of therapies used by individual naturopaths. As part of their practice naturopaths carry out patient assessment and diagnosis using standard approaches including physical examination, lab tests, and clinical assessment.\n\nAs noted above, the logic of naturopathy dictates that practitioners should whenever possible avoid giving pain killers to a patient with arthritis or anti-depressants to one with depression. But most naturopaths strongly support the use of herbal medicine3,6. For example, naturopaths sometimes recommend the use of herbal medicines for treating cervical dysplasia6, despite a lack of supporting evidence. By contrast, the standard medical procedure involves minor surgery, which is safe and effective. Similarly, mistletoe may be recommended by naturopaths for the treatment of hypertension6, even though it has not been properly tested in clinical trials and is toxic. Yet, drugs are often avoided despite being reasonably safe and effective.\n\nHerbal medicines are merely plant extracts that contain chemicals with drug-like action. Indeed, many of today’s drugs started life in previous centuries as herbal treatments. The attitude of naturopaths towards herbal medicines and drugs is therefore irrational.\n\nSimilar irrationality is seen with homeopathy, a type of CAM that is highly controversial and far from being proven as effective7,8. Nevertheless, the therapy is an integral part of the practice of many naturopaths1,3,9. Acupuncture is another modality often used by naturopaths as a treatment for many disorders3,9 but its proven value is very limited beyond the control of pain10.\n\nOne of the fundamental principles of naturopathy since its birth a century ago is the general claim that much sickness is caused by an accumulation of toxins in the body1. Accordingly, an important therapy in curing disease is the application of treatments that help eliminate these toxins. This concept – detoxification – is still the basis for various naturopathic therapies today6. Fasting is often employed in the belief that it induces detoxification. A variation of this approach is autotoxicity where the focus is toxins in the colon. This is commonly treated by colonic irrigation, a procedure that is potentially harmful as it can hyper-extend the colon11,12.\n\nThere is no credible evidence that detoxification treatments, such as dietary changes, consumption of herbs and supplements, fasting, or colonic irrigation, can remove toxins from the body or lead to improved health. These beliefs seem to be based on little more than speculation. Nevertheless, a survey of naturopaths in the USA found that 92% reported using detoxification therapies13. Similarly, a survey of naturopaths in Canada found more than half were using fasting as a treatment for various conditions9.\n\nHydrotherapy is another healing modality that naturopathy inherited from previous times. This treatment was very popular in central Europe where many people would go to a spa and sit in a pool containing spring water that was credited with healing properties. Indeed, this therapy is still popular in parts of Europe. The therapy was imported into the USA. Hydrotherapy is advocated by many naturopaths, often with the claim that it aids in detoxification and helps strengthen immune function1. Certainly, spending time in a sauna, hot tub, spa, or sanatorium is relaxing, but there is little evidence that hydrotherapy has any more direct therapeutic value. Interestingly, a recent observational study carried out on men in Finland reported a protective association between use of a sauna and reduced risk of cardiovascular disease and of all-cause mortality14.\n\nIridology is a technique where practitioners examine a patient’s eye and then make a diagnosis based on changes in the iris. The technique rests on the claim that many disorders can be diagnosed using the technique but it has no scientific basis15. Nevertheless, 23% of naturopaths in Canada reported using iridology in their practice9.\n\n\nPositive aspects of naturopathic medicine\n\nSome aspects of naturopathic medicine compare favorably with conventional medicine. As was stated earlier, naturopathic medicine places a strong emphasis on avoiding treatments that pose a risk. Unfortunately, what this can often mean is that patients are denied treatments that are effective while posing an acceptably low risk of harm. But conventional physicians often go too far in the opposite direction. In particular, there have been many stories over the past several decades of medications being prescribed that cause significant harm. A recent example is OxyContin, an opioid widely used for the relief of pain. The over-prescribing of this medication by physicians has led to an epidemic of addiction and resulted in many deaths16,17.\n\nNaturopaths place a strong emphasis on preventive medicine by means of encouraging their patients to live a healthy lifestyle. Moreover, they have been critical of conventional medicine for failing to be active in this area. There may have been much truth to this in the past but over the last several decades conventional medicine has given much attention to this crucial area. This includes strong support for health promotion and various efforts to prevent disease at the population level. In addition, conventional physicians routinely screen middle-aged patients for such conditions as diabetes, hypertension, and high blood cholesterol. Interventions are then made where appropriate so as to prevent disease. However, what conventional physicians do, more often than not, is to write a prescription rather than encourage their patients to follow a healthy lifestyle. For example, one study on obesity showed that less than one third of Canadian physicians advised their overweight patients to lose weight18.\n\nThe obvious explanation for why conventional physicians routinely ignore lifestyle when treating their patients is lack of time. Counseling patients on making lifestyle changes takes far more time than writing a prescription for drugs that treat hypertension or high blood cholesterol. In a system where physicians generally spend no more than 10 or 15 minutes on each consultation, it is simply not possible to make a serious effort to assess a patient’s lifestyle and then deliver appropriate counseling. Naturopaths, by contrast, spend much more time with their patients; a typical first office visit to a naturopath takes one hour6 while an average visit takes around 42 minutes3.\n\n\nConclusions\n\nA challenge in analyzing the practice of naturopathy and of critically evaluating its advantages and disadvantages is that relatively few research studies have been carried out. There is an urgent need for more research.\n\nNaturopathic medicine offers an approach to health that appeals to many people: it is considered to be “holistic”, based on prevention, promises treatments that “detoxify” the body, and avoids the potential hazards often seen with conventional medicine. However, naturopathic medicine has serious negative features including the use of treatments that have little or no supporting evidence, such as homeopathy and treatments intended to bring about detoxification. Similarly, naturopaths use treatments where the supporting evidence is weak, such as many uses of acupuncture and herbalism. Naturopaths frequently fail to prescribe drugs where the benefit strongly outweighs the risk.\n\nNaturopathic medicine does have some positive features, especially its strong emphasis on encouraging patients to prevent disease by living a healthy lifestyle. Conventional physicians, by contrast, are much less likely to dispense lifestyle advice, mainly due to lack of time. However, this is a rather weak advantage of naturopathy as the benefits of a healthy lifestyle are already well known by the majority of the population, especially by people who are most likely to visit a naturopath.\n\nOverall, the positive aspects of naturopathy are greatly outweighed by the negative aspects.", "appendix": "Competing interests\n\n\n\nThe author declared no competing interests.\n\n\nGrant information\n\nNo funding was involved in supporting this work.\n\n\nReferences\n\nPizzorno JE, Snider P, Micozzi MS: Nature Cure, Naturopathy and Natural Medicine. In: Micozzi M, ed. Fundamentals of complementary and alternative medicine. St Louis, MO: Elsevier (Saunders). 2015; 347–65. Reference Source\n\nElder CR: Integrating naturopathy: can we move forward? Perm J. 2013; 17(4): 80–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoon HS, Cherkin DC, Erro J, et al.: Practice patterns of naturopathic physicians: results from a random survey of licensed practitioners in two US States. BMC Complement Altern Med. 2004; 4: 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFleming SA, Gutknecht NC: Naturopathy and the primary care practice. Prim Care. 2010; 37(1): 119–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTemple NJ, Wilson T, Jacobs DR Jr: Nutritional health: strategies for disease prevention. 3rd ed. New York: Humana Press, 2012. Publisher Full Text\n\nPizzorno JE, Snider P: Contemporary naturopathic medicine. In: Micozzi M, ed. Fundamentals of complementary and alternative medicine. St Louis, MO: Elsevier (Saunders). 2015; 366–86. Reference Source\n\nBewley S, Ross N, Braillon A, et al.: Clothing naked quackery and legitimising pseudoscience. BMJ. 2011; 343: d5960. PubMed Abstract \| Publisher Full Text\n\nFisher PA: What about the evidence base for homeopathy? BMJ. 2011; 343: d6689. PubMed Abstract \| Publisher Full Text\n\nVerhoef MJ, Boon HS, Mutasingwa DR: The scope of naturopathic medicine in Canada: an emerging profession. Soc Sci Med. 2006; 63(2): 409–17. PubMed Abstract \| Publisher Full Text\n\nErnst E, Pittler MH, Wider B, et al.: The desktop guide to complementary and alternative medicine: an evidence-based approach. Toronto: Mosby. 2006. Reference Source\n\nAcosta RD, Cash BD: Clinical effects of colonic cleansing for general health promotion: a systematic review. Am J Gastroenterol. 2009; 104(11): 2830–6, quiz 2837. PubMed Abstract \| Publisher Full Text\n\nMishori R, Otubu A, Jones AA: The dangers of colon cleansing. J Fam Pract. 2011; 60(8): 454–7. PubMed Abstract\n\nAllen J, Montalto M, Lovejoy J, et al.: Detoxification in naturopathic medicine: a survey. J Altern Complement Med. 2011; 17(12): 1175–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaukkanen T, Khan H, Zaccardi F, et al.: Association between sauna bathing and fatal cardiovascular and all-cause mortality events. JAMA Intern Med. 2015; 175(4): 542–8. PubMed Abstract \| Publisher Full Text\n\nKnipschild P: Looking for gall bladder disease in the patient’s iris. BMJ. 1988; 297(6663): 1578–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWebster PC: Medically induced opioid addiction reaching alarming levels. CMAJ. 2012; 184(3): 285–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDhalla IA, Mamdani MM, Sivilotti ML, et al.: Prescribing of opioid analgesics and related mortality before and after the introduction of long-acting oxycodone. CMAJ. 2009; 181(12): 891–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKirk SF, Tytus R, Tsuyuki RT, et al.: Weight management experiences of overweight and obese Canadian adults: findings from a national survey. Chronic Dis Inj Can. 2012; 32(2): 63–9. PubMed Abstract" }	[ { "id": "10440", "date": "23 Sep 2015", "name": "Jonathan L. Wardle", "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\nInitially it is only right that I completely disclose my potential conflict of interest up front. In addition to my qualifications in public health and law I am also qualified as a naturopath. This would suggest that I – automatically – would disagree with the premise of the article. However, I will review the article below on its merits alone. This article is also clearly labelled as a commentary, so my criticism will also bear this in mind. This relates to the author’s opinion, I will only highlight where I think this opinion is not necessarily argued well or on fact. Title and abstract: I would suggest that the title, whilst certainly poetic, does not really reflect the content of the article. I do not think the author has offered ‘nine negatives’ to each ‘positive’ (in fact, there appears to be nearly as much word count supporting naturopathy as there is arguing against it). I think that as it stands the article’s title reflects a strength of argument that isn’t there. I think that there are less ‘quantitatively driven’ titles the author could choose, or the author could modify content to address the current mismatch between title and content. There is far more positive discussion of aspects of naturopathic care than the title would suggest. The abstract tends to describe naturopaths as a homogenous group, and I would suggest that qualifiers such as ‘many naturopaths are hostile’ or ‘a proportion of naturopaths are hostile’ would be more accurate than blanket statements. It is undeniably true that these opinions exist, but the homogeneity of opinion certainly does not. Article content: I give specific comments below, in the order that they appear in text:In the introduction the author does not appear to define naturopathy correctly. It is a principles-based system of medicine (just as Chinese medicine and Ayurveda are). Detoxification is not one of these principles. The principles – adopted by the World Naturopathic Foundation and recognized by the World Health Organization – are summarized here - http://aanmc.org/naturopathic/6principles/ . Given that the author has used ‘guiding principles’ as a sub-heading this omission appears particularly noticeable. I think the author over-simplifies clinical care when he suggests that it ‘is a different story altogether’ in conditions like arthritis and depression. Holistic management of these conditions is considered important even in conventional care (e.g. depression associated with arthritis, functionality, and fatigue in rheumatoid arthritis). Moreover, he author makes the critically incorrect assumption that naturopaths ignore symptomatic treatment, with a whole body approach ‘leading to worse outcomes’. Like conventional medicine, naturopathic medicine has a spectrum of healing approaches individualized to patient need. A cursory glance of the depression and arthritis chapters of naturopathic texts would have confirmed this (such as Sarris J; Wardle J (eds) (2014) Clinical Naturopathy: An Evidence-Based Guide To Practice, 2nd edition. Churchill Livingstone: Sydney ISBN: 978-0-7295-4173-2 – used as a core text in over a dozen countries including the author’s own). Moreover, I find the choice of depression particularly interesting, given the high level of evidence of a specific naturopathic treatment in that condition (St John’s Wort), which is superior to many conventional medicines. The author appears to assume that naturopathic and conventional medicine do not interact or integrate, and discussed them as completely separate entities. One of the glaring examples of this is the discussion of cancer treatments – specifically the discussion of the ‘harmful effects price pay-off’ of cancer treatment. However, part of the core focus of modern naturopathic oncology treatment is to integrate with conventional oncology and to support this treatment through positive interactions (which increase the effect of chemotherapeutics), reduction of side effects from conventional treatment and to improve quality of life throughout. This is the treatment taught in naturopathic schools (and in Janet Schloss’ cancer chapter in the aforementioned Sarris and Wardle text) and practiced (see Dugald Seely’s research papers in PubMed, or view the work at the Ottawa Integrative Cancer Centre – and external teaching clinic of the Canadian College of Naturopathic Medicine). The author overplays naturopathic resistance to conventional medicine. The approach taken by naturopaths is classed as the Therapeutic Order. Suppressing Pathology is not discouraged completely, simply only encouraged to be used sparingly and not as first-line treatment. A full explanation can be found here: https://www.researchgate.net/publication/43328689_A_hierarchy_of_healing_the_therapeutic_order or in the relevant clinical chapters mentioned previously. It is probably worthwhile, in the introduction to the section on practice, to highlight the fact that naturopaths are defined by their philosophy, and not by their tools of trade. The logic used to suggest that naturopaths would ‘whenever possible’ not give pain-killers or anti-depressants. Many of the herbal or other therapeutic agents have very similar profiles (e.g. herbal analgesics or capsacain, St John’s Wort is essentially a SSRI). In severe depression and other mental health disorders naturopathic texts discuss – at length – co-management with pharmaceuticals. In some US states naturopaths have prescribing rights and prescribe these treatments (in fact, in Washington about 12% of naturopaths administer vaccines). In the entire practice section the author appears to cherry-pick treatments – including some (such as iridology) that are considered fringe even by naturopaths. The author uses the example of mistletoe in hypertension, which misses the point not only because it is rarely used but also because whole-system naturopathic care does have evidence for improvement in hypertension (e.g. see Bradley R, Kozura E, Kaltunas J, Oberg EB, Probstfield J, et al. (2011) Observed Changes in Risk during Naturopathic Treatment of Hypertension. Evid Based Complement Alternat Med 2011: 826751 and Seely, Dugald, et al. \"Naturopathic medicine for the prevention of cardiovascular disease: a randomized clinical trial.\" Canadian Medical Association Journal 185.9 (2013): E409-E416) – with treatments not including mistletoe. Herbal medicines are far more complex than the author suggests (e.g. see the importance of ‘whole herb’ hyperforin, hypericin etc reactions in St John’s Wort – or the work of the World Health Organization on collection, harvesting and manufacturing herbal medicines) and therefore is not irrational. It is difficult to see, without any compelling argument of any kind, how the author can suggest naturopaths using herbal medicines simply because they are chemical as being irrational either. The author conflates fasting (of which there is much evidence – especially in chronic conditions such as RA) with detoxification, overstates the importance of detoxification (probably through an overly simplified interpretation of toxic load theory – which is more to do with a mnemonic for reducing the impact of metabolic wastes). ‘Detoxification therapies’ in naturopathic care can include anything from undoubtedly beneficial therapies such as encouraging more water intake, more regular bowel motions, increased fibre to less likely to be beneficial treatments such as foot spas. Given this breadth it is not surprising that 92% stated using them. The author is also confused on hydrotherapy – confusing balneotherapy with hydrotherapy at one stage. Hydrotherapy is the therapeutic application of pressure or heat via water – ice compresses are probably the most commonly known of these, but there are numerous other treatments. Leon Chaitow’s text ‘Naturopathic Physical Medicine’ published by Elsevier has a good chapter authored by Eric Blake on this topic should the author require more detail. The author highlights one study on sauna therapy – but there are many more he could find if he searched PubMed. The author states that the beneficial aspects of naturopathic care can easily be resolved by giving conventional physicians more time. I think this is an overly simplistic interpretation. There are socio-cultural and philosophical reasons for these differences, suggesting they are merely time-based is disingenuous, and not borne out from the data. Observational studies of naturopathic practice have found that health promotion counseling on diet, physical activity, and stress management is incorporated into almost every clinical encounter (80%–100%) and is reinforced over successive patient visits. This finding diverges substantially from the low rates of health promotion in conventional care (< 35%–40%). These REFS can be found in the following article – Wardle J; Oberg E (2011) “The intersecting paradigms of naturopathic medicine and public health: opportunities for naturopathic medicine” Journal of Alternative and Complementary Medicine 17(11); pp1079-84 Conclusions:I feel the statement that naturopaths use treatments where the evidence is weak is too general and too directive. The same can be said of any profession (there is very little evidence for physiotherapy – see last year’s BMJ review – and conventional physicians continue to perform knee arthroscopy more than a decade after it has been unequivocally demonstrated to have little clinical value). I agree that more research is needed, but there is also areas in which naturopathic care has demonstrable value in the evidence base. I feel as though, rather than attempting to discredit an entire profession through a few cherry-picked examples, the author would make a better argument by highlighting some of the issues that may make naturopathic care sub-optimal. I have written many of these myself. The very nature of naturopathic medicine may attract people distrustful of conventional medicine with conflicting scientific and philosophical worldviews (see Heather Boon’s work or Amie Steel’s work on naturopaths engaging with evidence). The lack of regulation in many jurisdictions means that naturopaths can practice with little or no training. The ‘broad church’ and inclusive nature of naturopathy may make it easier for fringe therapies to establish. Gort and Coburn suggest that the marginalization of naturopathic medicine itself may encourage it adopt marginal doctrines that bear little philosophical relation to naturopathic practice. However, I feel as though the few cherry-picked and poorly researched topics chosen by the author do little to advance their argument. I feel as though the author has spoken beyond the data on numerous occasions. However, as a commentary piece this should reflect their views. There are some, I believe, errors of fact that need to be addressed. Overall, most of the problems with content I put down to poor knowledge of the topic. Complementary medicine is a highly politically charged topic that can lead to strong opinions on ‘both sides’ of the debate. I think that this article needs to be either more measure if it is to keep its current form, or if it is to keep its current tone the arguments need to be made far more compelling. As it stands it is a commentary which clearly highlights one authors view of the topic, but it does not convince a reader that this view is a ‘rational’ one (to use the author’s terms).", "responses": [] }, { "id": "11703", "date": "29 Dec 2015", "name": "Charles R. Elder", "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 timely topic and the report is well organized however I have a number of major concerns.\n\nFirst, the tone of the report is not well balanced, and provides harsh criticism of naturopathy but with insufficient exploration of potential benefits. Secondly, there are numerous criticisms of naturopathy which are offered as statements of fact that in reality are unsubstantiated, or incorrect. For example, the author states:“With arthritis and depression, for example, there can be a specific dysfunction in a single body system. Conventional medicine has treatments of proven effectiveness that target the problem. In these cases treating the whole body will likely lead to poorer outcomes.” In reality, these two conditions are often recalcitrant to conventional therapies, and may be appropriate areas for including complementary therapies. As another example, the author states:“There is no credible evidence that detoxification treatments, such as dietary changes, consumption of herbs and supplements, fasting, or colonic irrigation, can remove toxins from the body or lead to improved health. These beliefs seem to be based on little more than speculation.”Yet there is evidence in the biomedical literature of reductions in inflammatory markers associated w/ fasting, and Ayurvedic medicine provides a long tradition, therapeutic rationale, and vast body of anecdotal evidence supporting the use of such detoxification procedures. Finally, it is unclear what the real purpose of the article might be. Is this an attempt to encourage dialogue between naturopaths and conventional clinicians? Is the purpose to educate conventional clinicians about the role, and limits, or naturopathy? Is the purpose to simply criticize naturopathy? What could be a constructive message, or contribution, for the essay? I think the project has excellent potential once this issue is clarified.", "responses": [] } ]	1	https://f1000research.com/articles/4-169
https://f1000research.com/articles/4-168/v1	25 Jun 15	{ "type": "Data Note", "title": "Comprehensive knowledge base of two- and three-dimensional activity cliffs for medicinal and computational chemistry", "authors": [ "Ye Hu", "Norbert Furtmann", "Dagmar Stumpfe", "Jürgen Bajorath", "Ye Hu", "Norbert Furtmann", "Dagmar Stumpfe" ], "abstract": "Activity cliffs are formed by pairs or groups of structurally similar or analogous active compounds with large differences in potency. They can be defined in two or three dimensions by comparing graph-based molecular representations or compound binding modes, respectively. Through systematic analysis of publicly available compound activity data and ligand-target X-ray structures we have in a series of studies determined all currently available two- and three-dimensional activity cliffs (2D- and 3D-cliffs, respectively). Furthermore, we have systematically searched for 2D extensions of 3D-cliffs. Herein, we specify different categories of activity cliffs we have explored and introduce an open access data deposition in ZENODO (doi: 10.5281/zenodo.18490) that makes the entire knowledge base of current activity cliffs freely available in an organized form.", "keywords": [ "Active compounds", "X-ray structures", "activity cliffs", "data mining", "structure-activity relationships", "computational methods" ], "content": "Introduction\n\nThe activity cliff concept has experienced increasing interest in chemical informatics and medicinal chemistry1–5. A consensus definition of activity cliffs1–4 refers to pairs or groups of structurally similar or analogous active compounds with large differences in potency4,5. For the definition of activity cliffs, the specification of similarity and potency difference criteria is required. Two-dimensional activity cliffs (2D-cliffs) have mostly been defined on the basis of Tanimoto similarity6 comparing molecular fingerprint representations2. More recently, 2D-cliffs have also been defined on the basis of substructure relationships, preferably employing the matched molecular pair (MMP) formalism7,8, leading to the introduction of MMP-cliffs9. An MMP is defined as a pair of compounds that are only distinguished by a structural change at a single site7, i.e., the exchange of a substructure, termed a chemical transformation8. For the definition of MMP-cliffs, transformation size restrictions have been introduced to limit transformations to small chemical changes typically observed in analog series9. Applying well-defined similarity and potency difference criteria, 2D-cliffs can be systematically extracted from compound databases10.\n\nThe vast majority of 2D-cliffs (i.e., close to or more than 95%, depending on the molecular representations and similarity measure used) are not formed in isolation (i.e., in the absence of structural neighbors with significant potency variations), but rather in a coordinated manner involving series of compounds with varying potency forming multiple and overlapping cliffs4,5,11. In activity cliff network representations where nodes represent compounds and edges activity cliffs, coordinated cliffs emerge as individual clusters of varying composition and size11, which can be isolated for further analysis.\n\nIn addition to 2D-cliffs, three-dimensional activity cliffs (3D-cliffs) can also be defined by comparing compound binding modes in X-ray structures12. This requires the superposition of structures of a given target available in different crystallographic ligand-target complexes and the assessment of the 3D similarity of bound ligands12. Three-dimensional activity cliffs can be further extended by taking 2D ligand information into account. This can be accomplished by systematically searching compound activity classes for analogs of 3D-cliff partners13. For example, for each cliff partner, MMPs with database compounds sharing the same activity can be determined and qualifying analogs can be assigned to the 3D-cliff13, leading to what we term herein a 3D-cliff-MMP extension. Figure 1 shows an exemplary 2D-cliff (MMP-cliff), activity cliff cluster, 3D-cliff, and 3D-cliff-MMP extension.\n\nDifferent categories of activity cliffs are shown formed by inhibitors of tyrosine kinase ABL. MMP-cliffs are used to represent 2D-cliffs. For each compound, the ChEMBL or Protein Data Bank (PDB) ID and its negative logarithmic potency value are reported. (a) An exemplary MMP-cliff (structural modification highlighted in red) taken from an activity cliff cluster (dashed blue box) is shown. In an activity cliff network, nodes represent compounds and edges cliffs. Nodes are colored according to potency values using a continuous color spectrum from red (lowest potency) via yellow (intermediate) to green (highest potency). In network representations, coordinated activity cliffs emerge as clusters. (b) An exemplary 3D-cliff and its 2D extension are shown (3D-cliff-MMP). The extension results from MMP-based mapping of analogs from ChEMBL to 3D-cliff compounds. Structural differences between 3D-cliff compounds and their 2D (MMP) partners are highlighted in red.\n\nIn medicinal chemistry, 2D-cliffs are often considered in the context of structure-activity relationship (SAR) analysis and compound design2,3. For SAR exploration, activity cliff clusters are of particular interest because they provide more SAR information than 2D-cliffs studied individually. Furthermore, 3D-cliffs are of prime interest for structure-based design and also for computational chemistry applications including, for example, the calibration of scoring functions or free energy (perturbation) calculations. Last but not least, 2D extensions of 3D-cliffs bridge different applications in medicinal and computational chemistry and help to identify candidate compounds for further analysis.\n\n\nMethods and materials\n\nThe activity cliff information provided herein is the result of recent surveys and systematic analyses of 2D-cliffs including clusters14, 3D-cliffs15, and extensions of 3D-cliffs16. Table 1 summarizes the different activity cliff categories. All 2D-cliffs reported herein originated from the most recent release of ChEMBL (version 20)17,18 and all 3D-cliffs from the Protein Data Bank (PDB; accessed December, 2014)19. MMP extensions of 3D-cliffs were identified in ChEMBL (version 19).\n\nReported are the number of 2D-cliffs belonging to three different categories (FP, fingerprint; Tc, Tanimoto coefficient), corresponding activity cliff clusters (comprising at least three compounds), 3D-cliffs for different potency measurement-dependent data sets, and corresponding 3D-cliff-MMPs (giving the total number of MMPs detected for 3D-cliffs from each data set).\n\nFor all activity cliffs, an at least 100-fold difference in potency between cliff partners was consistently required. For 2D-cliffs, only (assay-independent) Ki values were considered as potency measurements. For 3D-cliffs, Ki and IC50 measurements were separately considered (using a Ki and IC50 value-based data set, respectively). In addition, 3D-cliffs were also determined in a combined Ki/IC50 data set (taking into consideration that 3D-cliffs provide a much smaller knowledge base than 2D-cliffs; vide infra). For 3D-cliff analysis, a crystallographic resolution limit of 3.0 Å was applied.\n\nTwo-dimensional activity cliffs were determined using three different molecular representations including the extended connectivity fingerprint with bond diameter 4 (ECFP4)20, molecular access system (MACCS) structural keys21, and transformation size-restricted MMPs (MMP-cliffs)9. As similarity criteria for ECFP4- and MACCS-based activity cliffs, Tanimoto coefficient threshold values of 0.55 and 0.85 were applied, respectively2,14. For an MMP-cliff, our preferred 2D-cliff definition3,4, the formation of a transformation size-restricted MMP served as a similarity criterion. By definition 2D-cliffs do not contain stereochemical information.\n\nFor the identification of 3D-cliffs, the normalized overlap of atomic property density functions calculated for a pair of bound ligands was used as a measure of 3D similarity, taking conformational, positional, and atomic property differences into account12. An at least 80% calculated 3D similarity was required as a threshold for 3D-cliff formation12,15.\n\n\nData description\n\nActivity cliff statistics are reported in Table 1. A total of 17,111 MMP-cliffs, 31,975 ECFP4-, and 34,813 MACCS-based 2D-cliffs were identified formed by compounds active against more than 300 targets in each case. The corresponding number of activity clusters (comprising at least 3 compounds) was 1267, 1462, and 1402, respectively. Therefore, a very large knowledge base of well-defined 2D-cliffs is currently available. In addition, on the basis of Ki and IC50 measurements, 236 and 292 3D-cliffs were detected and were formed by crystallographic ligands of 26 and 43 targets, respectively. The combined Ki/IC50 data set yielded 595 3D-cliffs for 58 targets. Although many more 2D- than 3D-cliffs are currently available, as one would expect, the number of 3D-cliffs is larger than we anticipated, hence providing substantial opportunities for structural and computational studies. Table 2 provides details for the 61 different targets for which 3D-cliffs were detected. Furthermore, more than 1000 3D-cliff-MMPs were identified for each of the Ki and IC50 data sets and 2608 for the combined set (Table 1). Hence, for many 3D-cliffs, active analogs are available whose SAR characteristics and possible interaction patterns can be explored based upon 3D-cliff information, for example, by superposing them onto 3D-cliff compounds with which they form transformation size-restricted MMPs.\n\nA total of 61 targets are listed for which 3D-cliffs were detected. For each target, the ChEMBL ID, UniProt accession ID (UniProtID)22, and the number of available 3D-cliffs are reported. 3D-cliffs were separately determined on the basis of only Ki or IC50 measurements (available for active compounds) as well as for the combined Ki and IC50 data set (Ki/IC50). In addition, the number of MMP-cliffs (if available; defined on the basis of Ki values) is also reported for each target.\n\n\nData availability\n\nThe activity cliff information described above is made freely available in four separate data files containing 2D-cliffs, 3D-cliffs, 3D-cliff-MMP extensions, and superpositions of complex X-ray structures and 3D ligands for selected targets:\n\n(1) 2D-Cliffs_and_Cliff-Clusters.xlsx (Excel format): 2D-cliffs and clusters belonging to different categories are separately recorded using ChEMBL IDs.\n\n(2) 3D-Cliffs.xlsx (Excel): 3D-cliffs from the Ki, IC50, and Ki/IC50 data sets are separately provided using PDB IDs for compounds and UniProt22 IDs for targets.\n\n(3) 3D-Cliff_Extension.xlsx (Excel): Analogs of 3D-cliff compounds identified by MMP search are reported. For each of the three data sets, all 3D-cliff-MMPs are listed.\n\n(4) Superpositions.zip (all files in MOL2 format): For each target in Table 2, superpositions of complex X-ray structures and 3D ligands are provided.\n\nThese data sets are contained in an open access ZENODO deposition23. The deposition also contains a README document that details the data organization and information provided.\n\n\nConclusions\n\nIn this study, we have discussed different categories of activity cliffs (including cliff extensions) and reported the distribution of cliffs belonging to these categories. Given the cliff definitions applied herein, the activity cliff information we provide as an open access deposition is up-to-date and comprehensive. We hope that this large knowledge base of activity cliffs will be helpful in the practice of medicinal chemistry and structure-based drug design as well as in further evaluating and advancing computational methods.\n\n\nData availability\n\nZENODO: Knowledge base of two- and three-dimensional activity cliffs, doi: 10.5281/zenodo.1849023", "appendix": "Author contributions\n\n\n\nJB designed the study, NF, YH, and DS collected, organized, and deposited the data, JB wrote the manuscript, all authors examined the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests declared.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nN.F. was supported by a fellowship from the Jürgen Manchot Foundation, Düsseldorf, Germany.\n\n\nReferences\n\nMaggiora GM: On outliers and activity cliffs--why QSAR often disappoints. J Chem Inf Model. 2006; 46(4): 1535. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Bajorath J: Exploring activity cliffs in medicinal chemistry. J Med Chem. 2012; 55(7): 2932–2942. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Hu Y, Dimova D, et al.: Recent progress in understanding activity cliffs and their utility in medicinal chemistry. J Med Chem. 2014; 57(1): 18–28. PubMed Abstract \| Publisher Full Text\n\nHu Y, Stumpfe D, Bajorath J: Advancing the activity cliff concept [v1; ref status: indexed, http://f1000r.es/1wf]. F1000Research. 2013; 2: 199. Publisher Full Text\n\nStumpfe D, de la Vega de León A, Dimova D, et al.: Follow up: Advancing the activity cliff concept, part II [v1; ref status: indexed, http://f1000r.es/34p]. F1000Research. 2014; 3: 75. Publisher Full Text\n\nWillett P, Barnard JM, Downs GM: Chemical similarity searching. J Chem Inf Comput Sci. 1998; 38(6): 983–996. Publisher Full Text\n\nKenny PW, Sadowski J: Structure modification in chemical databases. In Chemoinformatics in Drug Discovery; Oprea, T. I., Ed.; Wiley-VCH: Weinheim, Germany, 2005; 271–285. Publisher Full Text\n\nHussain J, Rea C: Computationally efficient algorithm to identify matched molecular pairs (MMPs) in large data sets. J Chem Inf Model. 2010; 50(3): 339–348. PubMed Abstract \| Publisher Full Text\n\nHu X, Hu Y, Vogt M, et al.: MMP-cliffs: systematic identification of activity cliffs on the basis of matched molecular pairs. J Chem Inf Model. 2012; 52(5): 1138–1145. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Bajorath J: Frequency of occurrence and potency range distribution of activity cliffs in bioactive compounds. J Chem Inf Model. 2012; 52(9): 2348–2353. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Dimova D, Bajorath J: Composition and topology of activity cliff clusters formed by bioactive compounds. J Chem Inf Model. 2014; 54(2): 451–461. PubMed Abstract \| Publisher Full Text\n\nHu Y, Furtmann N, Gütschow M, et al.: Systematic identification and classification of three-dimensional activity cliffs. J Chem Inf Model. 2012; 52(6): 1490–1498. PubMed Abstract \| Publisher Full Text\n\nFurtmann N, Hu Y, Bajorath J: Comprehensive analysis of three-dimensional activity cliffs formed by kinase inhibitors with different binding modes and cliff mapping of structural analogues. J Med Chem. 2015; 58(1): 252–264. PubMed Abstract \| Publisher Full Text\n\nStumpfe D, Bajorath J: Monitoring global growth of activity cliff information over time and assessing activity cliff frequencies and distributions. Future Med Chem. 2015; 7: in press.\n\nFurtmann N, Hu Y, Gütschow M, et al.: Identification and analysis of the currently available high-confidence three-dimensional activity cliffs. RSC Adv. 2015; 5(54): 43660–43668. Publisher Full Text\n\nHu Y, Furtmann N, Bajorath J: Extension of three-dimensional activity cliff information through systematic mapping of active analogs. RSC Adv. 2015; 5(54): 43006–43015. Publisher Full Text\n\nGaulton A, Bellis LJ, Bento AP, et al.: ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2011; 40(Database issue): D1100–D1107. 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\nBerman HM, Westbrook J, Feng Z, et al.: The protein data bank. Nucleic Acids Res. 2000; 28(1): 235–242. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRogers D, Hahn M: Extended-connectivity fingerprints. J Chem Inf Model. 2010; 50(5): 742–754. PubMed Abstract \| Publisher Full Text\n\nMACCS Structural keys. Accelrys, San Diego, CA, USA. Reference Source\n\nUniProt Consortium: The Universal Protein Resource (UniProt) in 2010. Nucleic Acids Res. 2010; 38(Database issue): D142–D148. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu Y, Furtmann N, Stumpfe D, et al.: Knowledge base of two- and three-dimensional activity cliffs. Zenodo. 2015. Publisher Full Text" }	[ { "id": "9335", "date": "06 Jul 2015", "name": "Alexandre Varnek", "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 short but nice paper describing the data related to 2D and 3D activity cliffs for large variety of biological targets. The data can freely be downloaded from ZENODO which opens a way for numerous computational experiments. They also may become an important support of particular medchem projects. I recommend this paper for indexation", "responses": [] }, { "id": "9206", "date": "09 Jul 2015", "name": "Veer Shanmugasundaram", "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\nRecently there has been a lot of interest and effort in curating databases and providing MMP information for analysis of various properties by both academic and industrial groups as noted by some references below http://dx.doi.org/10.1021/ci5005256 http://dx.doi.org/10.1021/jm400223y http://dx.doi.org/10.1021/jm500317a Further, a few companies have started providing content and are in the data-provisioning business accessing corporate information through consortia like the one below from Medchemica: http://www.medchemica.com/index.html http://www.astrazeneca.com/Research/news/Article/260613-roche-and-astrazeneca-launch-medicinal-chemistry-datasha Or developing capabilities to use such information in drug discovery applications such as NextMove: http://www.nextmovesoftware.com The data note and open access data submitted here from the Bajorath Group based on their comprehensive work using 2D and 3D activity cliff information provides a wealth of information that directly will address many questions that medicinal chemistry project teams typically tend to ask. Very relevant, timely and valuable effort that should be communicated broadly to the research community.", "responses": [] }, { "id": "9508", "date": "16 Jul 2015", "name": "Gerhard Müller", "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\nY. Hu, N. Furtmann, D. Stumpfe, and J. Bajorath report on a conceptual extension of their well-exemplified activity cliff concept from the previously established graph-based two-dimensional representation to a three-dimensional version by factoring crystallographically solved high-resolution complex structures of respective ligands into the comparative analyses of sub-structural molecular changes linked to changes in activity. With that extension into three-dimensional space they account for ligand-target interactions, thus providing activity cliff-forming compound sets a design-relevant context that is of immediate assistance when embedded in a molecular design campaign. This clearly is a useful enrichment of the toolbox of computational, as well as medicinal chemists. The authors provide all the relevant molecular similarity and potency difference criteria underlying this analyses that are essential for the definition of activity cliff-forming compound clusters, together with a sound description of all methodological details. While the traditional 2D cliffs were based on e.g. Tanimoto similarity and the size-restricted matched molecular pair concept, introduced by the Bajorath group earlier, 3D cliffs are based on mutual molecular similarity for ligands for which high-resolution complex structures are available. Obtained 3D activity cliffs were further enriched to 3D-cliff-MMP extended sets by including molecular similarity considerations to active compounds for which no x-ray structure is available. In total, app. 17.000 MMP cliffs have been identified at more than 300 distinct biological targets revealing close to 1.300 activity clusters. Cumulatively, 600 3D cliffs have been detected involving app. 60 distinct protein targets, respectively. In Table 2, the authors provide a comprehensive overview of the target landscape and underlying statistics for the detected 3D activity cliffs. In addition, all relevant data are accessible and an open access data deposition in ZENODO has been established. It is especially the extended 3D-cliff-MMP datasets in which ligands with experimentally determined binding modes and interaction patterns serve as probe compound for closely related active analogues with available SAR information that bears a huge potential to extrapolate medicinal chemists’ understanding of structure-activity relationships from a pure comparative framework into a 3D direct design concept. An immediate interrogation of the binding site’s functionalities becomes amenable to a previously restricted indirect design approach. It will be interesting to see as to whether the 3D activity concept introduced in this contribution can be extended to a better understanding of structure-selectivity relationships of compound sets acting e.g. at different isoforms of densely populated target families. Summarizing, this contribution laid the basis for migrating a formerly indirect design-restricted tool for comprehensive SAR analysis into 3D space, actually into the binding pocket of investigated ligand sets, thus increasing the interpretability and the feasibility of identified cliff information for the community of practicing medicinal chemists", "responses": [] } ]	1	https://f1000research.com/articles/4-168
https://f1000research.com/articles/4-167/v1	24 Jun 15	{ "type": "Software Tool Article", "title": "Constellation Map: Downstream visualization and interpretation of gene set enrichment results", "authors": [ "Yan Tan", "Felix Wu", "Pablo Tamayo", "W. Nicholas Haining", "Jill P. Mesirov", "Yan Tan", "Felix Wu", "Pablo Tamayo", "W. Nicholas Haining" ], "abstract": "Summary: Gene set enrichment analysis (GSEA) approaches are widely used to identify coordinately regulated genes associated with phenotypes of interest. Here, we present Constellation Map, a tool to visualize and interpret the results when enrichment analyses yield a long list of significantly enriched gene sets. Constellation Map identifies commonalities that explain the enrichment of multiple top-scoring gene sets and maps the relationships between them. Constellation Map can help investigators take full advantage of GSEA and facilitates the biological interpretation of enrichment results. Availability: Constellation Map is freely available as a GenePattern module at http://www.genepattern.org.", "keywords": [ "gene set enrichment analysis", "GSEA", "gene expression", "signature", "pathway", "visualization", "mutual information" ], "content": "Introduction\n\nGene set enrichment analysis (GSEA) (Mootha et al., 2003; Subramanian et al., 2005) is widely used to analyze transcription data by identifying sets of genes that are coordinately up- or down-regulated in a phenotype of interest. By focusing on cumulative changes in the expression of multiple genes, GSEA can detect biologically meaningful processes (e.g., groups of genes in the same pathway) that differ significantly between phenotypes. The broad use of GSEA has, however, resulted in a rapid increase in the number of gene sets available for analysis. This presents a new challenge, because, depending on the collection(s) of sets employed, GSEA may yield tens or hundreds of significantly enriched gene sets. Thus, investigators may face the difficult task of sifting through multiple high scoring gene sets to find biologically relevant relationships between them. To address this need we developed Constellation Map, a network-based visualization tool, to facilitate the downstream analysis of enrichment results.\n\n\nDescription & case study\n\nConstellation Map presents gene set enrichment results generated by GSEA as a radial plot. Each node of the plot represents a significantly enriched gene set. Nodes that are closer to the origin (i.e., with shorter radial distance) are more highly associated with the phenotype of interest. The angular distance between two nodes represents the per-sample similarity of their respective gene sets’ enrichment. We use a normalized mutual information (NMI) score to measure both these associations (see Workflow & Methods). Edges between nodes denote an overlap between sets’ member genes, while edge thickness captures the relative size of the overlap.\n\nThese elements are all presented via a JavaScript-powered browser environment for interactive exploration. Investigators can quickly, visually identify tight clusters of connected nodes, i.e., gene sets with similar enrichment patterns that may represent different aspects of the same biological process, and assess how similar each node is to the others in that cluster. Identified clusters may be further interrogated by selecting them, extracting overlapping genes, and querying those genes using a variety of functional annotation tools (MSigDB, GeneMANIA, and DAVID) (Dennis et al., 2003; Subramanian et al., 2005; Warde-Farley et al., 2010) all within the tool. Constellation Map thus accelerates the biological interpretation of enrichment results by clarifying the relationships of high scoring gene sets relative to the phenotype and relative to each other.\n\nWe previously demonstrated these advantages by applying our tool to uncover gene sets that characterize the transcriptional response to trivalent inactivated influenza vaccine (TIV) (Tan et al., 2014). We analyzed expression profiles of peripheral blood mononuclear cells (PBMCs) from 24 subjects vaccinated with TIV and performed enrichment analysis to discriminate high and low responders. We used Constellation Map to project 13 gene sets significantly associated with high response (FDR < 0.25) (Figure 1). We identified two distinct clusters of gene sets enriched for immunoglobulin (labeled A) and proliferation genes (labeled B) and showed that these sets are tightly associated with the immune response to TIV. Visualizing and annotating with Constellation Map was crucial to our identification of the common biological processes that resulted in enrichment of these gene sets.\n\nHere we show the JavaScript powered Constellation Map visualization of the top 13 gene sets significantly associated with the transcriptional response in PBMC to vaccination with trivalent inactivated influenza vaccine (TIV). Nodes represent gene sets, and edges indicate overlap of member genes with thickness proportional to the amount of overlap. Gene sets radially closer to the origin are more highly associated with the high response phenotype. Gene sets in close angular proximity have similar enrichment patterns. Visually identified clusters were enriched for (A) immunoglobulin and (B) proliferation genes. Proliferation cluster nodes have been selected (highlighted in red), and the relevant gene set names, overlapping genes, and other metadata are displayed in the side panel.\n\n\nAdvantages of Constellation Map\n\nSeveral visualization and interpretation tools have been developed over the last few years to address the challenge of downstream interpretation of enrichment results. Unlike some of these tools, which are designed to use Gene Ontology (GO) or other hierarchically organized gene sets (Grossmann et al., 2007; Lewin & Grieve, 2006), Constellation Map can also perform well with gene sets derived from larger, less structured collections, such as the pathways and experiment signatures found in the popular MSigDB collections (http://www.msigdb.org). The network-based visualizer, Enrichment Map (EM) (Merico et al., 2010), is somewhat similar to Constellation Map in that it displays gene set enrichment results using a network representation where nodes represent sets and edges represent gene overlap between sets. However, EM clusters gene sets based on member gene overlap regardless of their relationship to the phenotype of interest. This ignores the possibility of gene sets having similar enrichment profiles despite little member gene overlap. Conversely, EM could highlight gene sets with some overlap that are different in their enrichment profiles across a group of samples. Constellation Map, on the other hand, takes similar per-sample enrichment profiles into account, providing this information to the investigator as an intuitive angular distance.\n\n\nWorkflow & methods\n\nA user begins the Constellation Map workflow (Figure 2) by either: (1) identifying a group of top-scoring gene sets using GSEA, or some other preferred enrichment analysis approach, and utilizing single sample gene set enrichment analysis (ssGSEA) (Barbie et al., 2009) to project samples into the space of top-scoring gene sets; or (2) directly projecting data into the space of all gene sets of interest using ssGSEA and later choosing to display only those most associated with a phenotype. The gene set enrichment projection result from this module is used as the input for Constellation Map. ssGSEA is an extension of GSEA, available as a GenePattern module ssGSEAProjection (http://www.genepattern.org), that generates an enrichment-ranked list of gene sets for each sample.\n\nTwo options exist. Users may either (1) analyze their whole genome transcript expression data using a preferred analysis method (e.g., GSEA) to identify a group of top-scoring gene sets, project samples into the space of these top-scoring gene sets using ssGSEA, and visualize the results using Constellation Map or (2) directly project their data into the space of all gene sets of interest using ssGSEA and choose only a small group of these gene sets to display with Constellation Map.\n\nUsing the gene set enrichment scores obtained via ssGSEA, Constellation Map estimates the probability density functions of gene set and phenotypic class variables using a kernel density estimation. These density functions are subsequently used to calculate NMI scores for each gene set, which capture the association between each gene set’s enrichment scores and phenotypic classes (Equation 1). The NMI of two variables is their mutual information (Equation 2) divided by their joint entropy (Equation 3) (Shannon, 1948). We chose to use the NMI metric because it is independent of the sample distribution and more sensitive to nonlinear associations than the more commonly used correlation coefficients. As NMI is unidirectional, we created a signed version (SNMI) using the sign of the Pearson correlation to distinguish between positive and negative associations (Equation 4).\n\n\n\nAfter calculating the NMI scores, gene sets that significantly associate with phenotypes of interest can be selected (using an FDR or NMI score cutoff) and projected onto a radial plot. A second set of NMI scores is calculated pairwise across the N selected gene sets to estimate the similarity between their ssGSEA enrichment profiles. These pairwise NMI scores are converted into dissimilarity scores, d = 1 - NMI, which provides a true distance metric (Vinh et al., 2010). Constellation Map uses this property to construct an N-by-N distance matrix D containing the distances d between all pairs of gene sets. Constellation Map then projects the distance matrix onto a radial plot using the multidimensional scaling projection R package “SMACOF,” version 1.5-0 (Leeuw & Mair, 2009). An angular distance matrix Δ is calculated by minimizing the objective function (Equation 5), where δij is the angular distance and dij is the original distance (stored in D) between gene sets i and j. The gene sets are plotted as points distributed about the origin. Angular distance between two gene sets is determined from Δ and is proportional to the similarity of the gene sets’ enrichment profiles. Radial distance (i.e., distance to the origin) indicates the gene set’s association with respect to the phenotype (1 - NMI).\n\nThe final step of Constellation Map projection involves calculating pairwise Jaccard indices across the gene sets. The Jaccard index is equal to the number of genes shared by two sets divided by the number of genes in their union. For pairs with Jaccard indices greater than a given threshold, edges are drawn connecting the respective nodes where the thickness of each edge is proportional to the Jaccard index (Jaccard, 1901; Merico et al., 2010).\n\n\nSummary\n\nConstellation Map is a powerful and intuitive tool in that it allows investigators to determine the relevance and relationships of their gene sets with relative ease. The visualizer evaluates a large set of gene set enrichment profiles using a variety of comparison metrics and presents these metrics in an understandable manner. This uncluttered, simple presentation reduces an investigator’s workload by easing the complex task of having to interpret the enrichment profiles of many gene sets. Just as Constellation Map aided us in identifying subgroups of gene sets with distinct immunologic biologies in our TIV vaccination case study (see above), we believe that investigators can similarly enhance their enrichment analyses by leveraging Constellation Map across their own data, helping them to draw meaningful biology from their many gene sets.\n\nAs the scientific community continues discovering new regulatory pathways, perturbation signatures, etc. and casting them into lists of genes, gene set collections will continue to expand. This growth may complicate the historically straightforward enrichment analysis when results contain thousands of gene sets, many of which may be redundant or related. Thus, there is a real need for downstream tools that can elucidate the major biological processes represented in these results and present them in an informative, exploratory manner. Constellation Map, with its mutual information-based layout, interactive visualizer, and connection to annotation services is well suited to meet this need.\n\n\nData availability\n\nTIV vaccine gene expression datasets are available from the NCBI Gene Expression Omnibus; accession number GSE29619 (Nakaya et al., 2011). Gene sets are contained in MSigDB version 3.0, collection C2 (Subramanian et al., 2005) available at the MSigDB download page (http://www.msigdb.org).\n\n\nSoftware availability\n\nConstellation Map is freely available as a GenePattern module (http://genepattern.broadinstitute.org/gp/pages/index.jsf?lsid=urn:lsid:8080.gpbroad.broadinstitute.org:genepatternmodules:345). Module source code is available at Zenodo (doi: 10.5281/zenodo.18586) and is maintained at the GenePattern community module archive, GParc (http://www.gparc.org); this module may be installed on a private GenePattern server (R-3.0 and Java required). Proper rendering of the visualizer requires a modern, JavaScript-enabled web browser; the authors recommend using the latest versions of Firefox or Chrome. Constellation Map is distributed under the open source MIT License.\n\nConstellation Map is freely available as a GenePattern module (http://genepattern.broadinstitute.org/gp/pages/index.jsf?lsid=urn:lsid:8080.gpbroad.broadinstitute.org:genepatternmodules:345). Module source code is available at Zenodo (doi: 10.5281/zenodo.18586) and is maintained at the GenePattern community module archive, GParc (http://www.gparc.org); this module may be installed on a private GenePattern server (R-3.0 and Java required). Proper rendering of the visualizer requires a modern, JavaScript-enabled web browser; the authors recommend using the latest versions of Firefox or Chrome. Constellation Map is distributed under the open source MIT License.", "appendix": "Author contributions\n\n\n\nYT and PT designed and developed the Constellation Map algorithm and visualization. FW implemented the software and refined the design of the user interface of the GenePattern module and interactive JavaScript visualizer. YT performed the analysis in the use case. WNH and JPM supervised software and algorithm development as well as the use case project. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNone declared.\n\n\nGrant information\n\nNational Human Genome Research Institute, award number U41HG007517, and National Institute of General Medical Sciences, award number R01GM074024, of the National Institutes of Health to JPM. Bill & Melinda Gates Foundation, grant number OPP50092 to JPM. National Institute of Allergy and Infectious Diseases, U19AI090023, of the National Institutes of Health to WNH.\n\nThe authors confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors thank the GenePattern development team for their help in implementing Constellation Map.\n\n\nReferences\n\nBarbie DA, Tamayo P, Jesse S, et al.: Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009; 462(7269): 108–112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDennis G Jr, Sherman BT, Hosack DA, et al.: DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 2003; 4(5): P3. PubMed Abstract \| Publisher Full Text\n\nGrossmann S, Bauer S, Robinson PN, et al.: Improved detection of overrepresentation of Gene-Ontology annotations with parent child analysis. Bioinformatics. 2007; 23(22): 3024–3031. PubMed Abstract \| Publisher Full Text\n\nJaccard P: Étude comparative de la distribution florale dans une portion des Alpes et des Jura. Bulletin de la Société Vaudoise des Sciences Naturelles. 1901; 37(142): 547–579. Publisher Full Text\n\nLeeuw Jd, Meir P: Multidimensional Scaling using Majorization: SMACOF in R. J Statist Software. 2009; 31(3): 1–30. Reference Source\n\nLewin A, Grieve IC: Grouping Gene Ontology terms to improve the assessment of gene set enrichment in microarray data. BMC Bioinformatics. 2006; 7: 426. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMerico D, Isserlin R, Stueker O, et al.: Enrichment map: A network-based method for gene-set enrichment visualization and interpretation. PLoS One. 2010; 5(11): e13984. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMootha VK, Lindgren CM, Eriksson KF, et al.: PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet. 2003; 34(3): 267–273. PubMed Abstract \| Publisher Full Text\n\nNakaya HI, Wrammert J, Lee EK, et al.: Systems biology of vaccination for seasonal influenza in humans. Nat Immunol. 2011; 12(8): 786–795. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShannon CE: A Mathematical Theory of Communication. Bell Syst Tech J. 1948; 27: 379–423 & 623–656. Reference Source\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005; 102(43): 15545–15550. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTan Y, Tamayo P, Nakaya H, et al.: Gene signatures related to B-cell proliferation predict influenza vaccine-induced antibody response. Eur J Immunol. 2014; 44(1): 285–295. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVinh NX, Epps J, Bailey J: Information theoretic measures for clusterings comparison: variants, properties, normalization and correction for chance. J Mach Learn Res. 2010; 11: 2837–2854. Reference Source\n\nWarde-Farley D, Donaldson SL, Comes O, et al.: The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res. 2010; 38(Web Server issue): W214–W220. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9357", "date": "07 Jul 2015", "name": "Hilary Ann Coller", "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 software that Mesirov and colleagues developed is designed to help scientists interpret the results of their gene set enrichment analysis results. They have developed a useful tool for visualizing the different datasets that emerge and organizing them with relation to each other. The software organizes the gene sets so that those that are most relevant are close to the origin and those that have similar patterns are close to each other. Those with overlapping datasets are indicated with lines between the dots. The metrics used by the software were well-selected, and the visualization approach should make it intuitive for users to gather valuable information about their data. Use of this software will likely allow scientists to gain more biological insights from GSEA analyses of their datasets.", "responses": [] }, { "id": "10711", "date": "07 Oct 2015", "name": "Sayan Mukherjee", "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\nExploratory data analysis, visualization, and hypothesis generation in genomics are essential tools. Methods and software that are clear, robust, and easily accessible are rare and very important to biomedical researchers. I found the approach described in Constellation Map intuitive and visually appealing. The statistical methods were reasonable, I would have liked to also see the possibility of metrics based on likelihood models, the kernel density estimation does go a bit in that direction. I was able to use the software. I would have liked for there to be an R version to download from cran as well as some vignettes independent of GenePattern. That said this work is useful and in my opinion some thought went into the visualization.", "responses": [] } ]	1	https://f1000research.com/articles/4-167
https://f1000research.com/articles/4-165/v1	23 Jun 15	{ "type": "Research Article", "title": "Immunogenicity of 10-valent pneumococcal conjugate vaccine among infants attending Mbagathi District Hospital, Kenya", "authors": [ "Michael Walekhwa", "Margaret Muturi", "Elizabeth Bukusi", "Margaret Muturi", "Elizabeth Bukusi" ], "abstract": "Introduction: This study aimed to determine the serum concentration of IgG antibodies as an indicator of immunogenicity, alongside the assessment of socio-demographic factors that affect IgG antibody levels in infants immunized with 10-valent pneumococcal conjugate vaccine (PCV-10) at the Mbagathi District Hospital in Kenya. Materials and methods: This cross-sectional study measured serum IgG antibodies among infants who had completed a 3-dose course of PCV-10. IgG antibodies to pneumococcal serotype-specific capsular polysaccharide were measured through enzyme-linked immunosorbent assay (ELISA). Results: The majority (83%) of infants who completed the required dose of pneumococcal conjugate vaccine had serum titres of pneumococcal disease- (PD) specific IgG antibodies of between 0.34 mg/dl and 0.36 mg/dl. 4% of infants had serum titres of 0.30 mg/dl to 0.33 mg/dl. The remaining 2% had IgG antibody titres of either ≤0.25 mg/dl, or between 0.25 mg/dl to 0.29 mg/dl. Additionally, there was multi-collinearity among the IgG antibody levels of the infants studied and several variables that had an effect on these levels. These included: alcohol consumption by infants’ biological mothers during pregnancy (r =.595, p ≤ 0.05); maternal diet during pregnancy (r =.137, p ≤ 0.05); breastfeeding frequency (r =.220, p ≤ 0.05); proximity to other children (r =.133, p ≤ 0.05); child hospitalization (r =.131, p ≤ 0.05) and chronic illness (r =.154, p ≤0.01). Conclusion: PCV-10 is immunogenic against PD four weeks after completion of 3-doses among the infants attending the Child Welfare clinic at the Mbagathi District Hospital in Kenya. Socio-demographic factors which include consumption of alcoholic drinks by infant’s biological mother during pregnancy and study infant chronic illness negatively affect the development of PD specific IgG. A balanced maternal diet during pregnancy and a breastfeeding frequency superior to three times per day have a significant positive effect on serum pneumococcal IgG levels among infants.", "keywords": [ "Serum concentration", "IgG antibodies", "PCV-10" ], "content": "Introduction\n\nStreptococcus pneumoniae (pneumococcus) is still the number one cause of morbidity, among infants and the elderly globally1. It is estimated that 150.7 million cases of pneumonia occur annually among infants2. Out of these, over 30% are usually severe enough to require admission to a hospital. Pneumonia accounts for >4 million deaths annually in under-developed countries3. It also contributes 20% of the total child deaths in Kenya annually4. Although there are antibiotic interventions, deaths caused by Pneumococcal Disease (PD) still remain high.\n\nAccording to the World Health Organization (WHO) over 70% of child deaths in 2008 were as a result of pneumococcal disease complications5. In developing countries, pneumococcal septicemia and meningitis account for 20% and 50% of severe PD cases, respectively6. The surge in anti-biotic resistant Streptococcus pneumoniae serotypes is an increasing global concern posing serious treatment challenges7. Because of the relatively high level of success of pneumococcal conjugate vaccines, the WHO has recommended that vaccines with broader serotype coverage be developed8.\n\nAccessibility to a safe and universally relevant sero-type inclusive vaccine is the surest way of curbing morbidity and mortality due to pneumococcal disease9. Although the WHO has recommended inclusion of PCVs in various national infant immunization programs10, it has not been very effective for >90% of under-developed countries. This is largely due to missing Streptococcus pneumoniae serotypes known to be circulating in developing countries11. PCV-10 contains serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F and 1, 5 and 7F12. Pneumococcal infections, irrespective of serotype, can be successfully treated with antibiotics, but the increasing resistance among pneumococci to antibiotics has highlighted the need for prevention, which can be achieved by vaccination13. Both 23-valent pneumococcal polysaccharide vaccine (PPV) and 10-valent pneumococcal conjugate vaccines are currently being used in Kenya14 yet no study has been done to confirm whether they are immunogenic against PD or not. PPV and PCV-10 were not formulated on the basis of exclusive serotypes found in Kenya and also despite their continued use, pneumonia remains the most killer of children in Kenya15. This has excited an interest in development of new type of vaccines that include all serotypes circulating in Kenya.\n\nConsidering the antibiotic resistance levels, prevention of pneumococcal disease early in life is necessary and thus so is the need for maternal and early infant immunization. According to the WHO, the effectiveness of pneumococcal conjugate vaccines can be evaluated on the basis of serum immunoglobulin G (IgG) levels17. A worldwide threshold level of 0.35 mg/dl after three doses at 6 weeks, 10 weeks, 14 weeks and in some cases a booster dose after 12 months has been recommended by WHO as a reference value for assessing immunogenicity of pneumococcal conjugate vaccines18.\n\nOn 14th February 2011, Kenya rolled-out the PCV-10 which is the vaccine that includes >50% of the serotypes circulating in Kenya19. Since the launch of the vaccine in February 2011, there are a rising number of children enrolling, yet it is not clear to what extent the vaccine is protective. Studies have not been conducted on the effectiveness of the pneumococcal vaccine in Kenyan children. Therefore there is still a need to evaluate the immunogenicity to determine the effectiveness of the vaccine in Kenyan children20.\n\nS. pneumoniae has over 90 serotypes, eight of which are contained in the PCV-10 (Synflorix) currently included in the Kenya Expanded Program on Immunization (KEPI). Pneumococcus serotypes vary a lot and their epidemiology is based on age, time period and geographical area21. Currently, there is no information on the pneumococcal serotypes present in Nairobi and other pneumococcal disease endemic areas in Kenya. It is possible that they are different from those included in the vaccine, putting the effectiveness of the vaccine in question. Over 70% of global annual deaths caused by PD is among children from developing countries22. Since the launch for free public use of PCV-10 in February 2011, there is an increase in the number of children being immunized yet its safety and immunogenicity has not been established. No data have been published on the immunogenicity after each dose in a series, or on the relative immunogenicity in Kenyan children of different ages. We therefore evaluated the concentration of PD immune antibodies as indicators of immunogenicity.\n\n\nMaterials and methods\n\nA cross-sectional study to evaluate PCV-10 immunogenicity among Kenyan infants was conducted between April 2013 and September 2013 among infants aged 1–12 months attending the Mbagathi District Hospital Child Welfare Care Clinic (Nairobi, Kenya). This study was funded by the National Council for Science, Technology and Innovation (NACOSTI) and ethical approval was obtained from the Kenyatta University Ethics Review Committee (ERC).\n\nThe majority of the patients treated at the hospital came from the neighbouring Kibera slum – an informal settlement. Most of these children presented with PD complications.\n\nTo determine the minimum sample size, the following formula was used: n=z2p^(1−p^)m2 Sample Size (n) = 38423\n\nA minimum of 384 infants from among those who received the vaccine were required for the study with a precision level of 5%. p^ is the assumed prevalence of pneumococcal disease among vaccinated childrenm, is the desired margin of error around the estimated prevalence, herein taken to be 5% and for 95% confidence z=1.96.\n\nThe subjects were enrolled by simple consecutive and convenient sampling. This entailed determination of the sample size as above and the enrolment of subjects as they were admitted to the clinic for administration of Vitamin A supplements. A total of 318 infants were recruited on a ‘first come, first sampled’ basis and this represented a response rate of 83%24. An equivalent number of mothers were enrolled along with their infants to investigate herd immunity.\n\n\nInclusion criteria\n\nInfants who attended the Child Welfare Clinic of Mbagathi District Hospital for vaccination and who had completed 3-doses of PCV-10 at least 4 weeks earlier were recruited. Participating infants were aged 12 months and below. Study infants’ biological mothers were approached at the clinic for informed consent. A total of 318 biological mothers to infants who were eligible for this study were also recruited to investigate if the vaccine confers herd immunity.\n\n\nExclusion criteria\n\nInfants with a medical condition (e.g. one that requires frequent visits to hospital) which would interfere with the assessment of the study objectives were excluded from the study. All subjects not meeting the general inclusion criteria or whose mothers refused to sign the consent form were excluded.\n\n\nInformed consent\n\nAll biological mothers of the infants involved in the study were approached at the Child Welfare Clinic of Mbagathi District Hospital by the researcher as they brought their children for weight check and vitamin A administration to sign a written informed consent document. This was done after reading the document and receiving a study explanation in Kiswahili as appropriate (Appendix I, translated to English in Appendix II). They also had opportunity to ask questions relevant to the study which facilitated the informed participation in the study.\n\n\nResearch methodology\n\nStructured questionnaires were formulated by the main researcher and moderated by the study supervisors. Piloting was done at the Kiambu District Hospital and corrections made before the main study. Standardized questionnaires were thereafter used to collect quantitative data from the participants’ mothers.\n\n\nCollection and storage of study samples\n\nSerum samples were collected by qualified and government registered phlebotomists working at the hospital. Coagulant-free vials were used to store the collected sample25. After child preparation (entailing the presence of the mother to ease tension), a 5ml capillary blood sample (heel stick sampling) was collected aseptically into the vials. Venous blood samples were collected from the mothers. Specimens were stored at room temperature until a clot formed (usually 15–45 minutes), then centrifuged to obtain serum specimens for assay. The serum was analysed within 24 hours then stored at 4°C. Serum samples that needed to be analysed out of this range of time were stored at -20°C. Samples were then transported to the Human Diagnostics World Laboratory on dry ice.\n\n\nSerosurvey\n\nAll serum samples were analysed using standard ELISA technique26. A pneumococcal antibody threshold concentration limit of 0.35 mg/dl indicated sero-protection27. Samples with antibody levels below this value were considered as lacking sero-protection.\n\n\nData collection\n\nResponses to the well-structured questionnaire were collected. All the study questionnaires remained in safe custody of the principal investigator to ensure confidentiality. Data on IgG levels were collected through assaying serum samples for children and mothers enrolled. Data were entered, and stored on Statistical Package for Social Scientists version 20 (SPSS) until analysed.\n\n\nEthical review\n\nScientific and ethical approval and authorization to conduct the study was sought from Kenyatta University (KU/R/COMM/51/37-2). Signed informed consent was sought from study participants after a clear explanation of the study and its purpose. The data collected and reported were in a form that did not allow identification of individual participants. Study numbers (Barcodes) and not personal identification was used in this study.\n\n\nStatistical methods\n\nSPSS version 20 was used28. Cross-tabulations and correlation analysis were used to generate relationships between dependent and independent variables, respectively. Ninety-five percent confidence intervals were used. Univariate analysis was done to determine the most significant variable that affects the PCV-10 vaccine immunogenicity.\n\n\nResults and discussion\n\nMost of the infants (30.8%) involved in this study were aged between 6–8 months. The second largest group who constituted 24.5% of the study group were aged between 3–5 months (Figure 1).\n\nOur results show that the majority of infants (83%) presented antibody levels between 0.34–0.36 mg/dl after completion of 3-doses of PCV-10 vaccination. 8% of infants presented with 0.37–0.39 mg/dl antibody levels (mg/dl) (Table 1).\n\nThis was followed by 4% of infants presenting 0.30- mg/dl to 0.33 mg/dl serum concentration of PD-specific IgG antibodies. The remainder had IgG antibody titres ranging between 0.25 mg/dl and 0.29 mg/dl and ≤0.25 mg/dl respectively.\n\nThe results from the majority of the subjects (83%) correlate with the threshold recommended by the WHO29. These data were also consistent with the data obtained from studies on the level of bacterial polysaccharide immune globulin needed to prevent pneumococcal otitis media and IPD30 among Ugandan children residing in Kampala. The findings are relatively higher than those found in the study done among Brazilian infants whose IgG antibody titres between 0.30 mg/dl to 0.36 mg/dl were ≤50% of the total subjects studied31.\n\nThe data from our study shows that use of alcoholic drinks, maternal diet during pregnancy and breastfeeding frequency affect serum antibody titres. The IgG antibody titres for subjects who consumed alcoholic drinks during pregnancy were relatively lower as compared to those who did not (Table 2). The results also show that the majority (>68%) of the study subjects had antibody titres between 0.34–0.36 mg/dl (Figure 2) which means that majority of them had developed immunity against PD. This study also found out that there was multi-collinearity between IgG antibody levels (mg/dl) for mothers and several variables. These variables were: consumption of alcoholic drinks by study infants’ biological mother (r =.595, p > 0.01); maternal diet during pregnancy (r =.137, p > 0.05); breast feeding frequency (r =.220, p > 0.01); gap with other children (r =.133, p > 0.05); chronic illness (r =.154, p > 0.01); and IgG antibody levels (mg/dl) for study infant’s biological mothers (r =.675, p > 0.01). Study by CDC shared similar results with this study particularly when comparing IgG antibody levels (mg/dl) for study infant’s biological mothers and IgG antibody levels (mg/dl) for infants. The study found out that there was a relationship between levels of reactive IgG among mothers and IgG antibody levels (mg/dl) for infants. Other studies however, did not find a significant correlation between levels of IgG among mothers and socio-demographic factors.\n\nIn addition, most of the infants in this study who were breast-fed more than five times a day and whose maternal diet was balanced during pregnancy had their IgG levels higher or equal to the recommended threshold (Table 3). It should be noted that the total percentage of infants with a serum concentration of PD-specific IgG between 0.34 mg/dl – 0.36 mg/dl was 76.1% in our study, relatively higher than the score of 24% within a six month period in a study done among infants in South Africa32. This is probably because of the Streptococcus strains included in the vaccine formulation. Out of the ten strains included in the vaccine, more than half are found in Kenya whereas these strains may not be found in South Africa33. Socio-demographic patterns may have also played a role in causing the disparity in immunogenicity of the vaccine between the two studies.\n\n\nConclusion\n\nThe PCV-10 vaccine was found to be immunogenic against PD four weeks after three doses among infants attending the Child Welfare Clinic at Mbagathi District Hospital in Nairobi. Socio-demographic factors which include use of alcoholic drinks by infants’ biological mothers during pregnancy and study infant chronic illness negatively affect the development of PD specific IgG. Balanced maternal diet during pregnancy, breastfeeding frequency (more than three times per day), was shown to have a significant positive effect on serum pneumococcal IgG levels among infants.\n\n\nData availability\n\nF1000Research: Dataset 1. Immunogenicity of 10-valent pneumococcal conjugate vaccine among infants at Mbagathi District hospital, 10.5256/f1000research.6087.d4903535\n\n\nConsent\n\nWritten informed consent for publication of clinical details was obtained from all participants and the mothers of the infants enrolled in the study.", "appendix": "Author contributions\n\n\n\nMichael Walekhwa was the main researcher, directly involved in the collection and coordination of data and sample collection, analysis and report writing. The study was supervised by Dr. Margaret Muturi of Kenyatta University who was very instrumental in proposal and thesis development. Prof. Elizabeth Bukusi was very helpful with regard to technical aspects of laboratory work.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was funded by the National Commission for Science, Technology and Innovation to Michael Walekhwa (NCST/5/003/3rd CALL, MSc/118). The funds entirely facilitated all research work.\n\n\nAcknowledgements\n\nI sincerely thank Dr. Margaret Muturi for her professional input that has seen this research reach this level, not to mention her unmatched patience and motherly treatment. Prof. Elizabeth Bukusi’s courtesy, organization and professionalism were incredible. I thank the National Council of Science and Technology (NACOSTI) for the funding. Mbagathi Hospital staff offered me incredible support and advice.\n\n\nSupplementary materials\n\nStudy questionnaire in Kiswahili\n\nClick here to access the data.\n\nStudy questionnaire in English\n\nClick here to access the data.\n\n\nReferences\n\nNuorti JP, Whitney CG: Prevention of pneumococcal disease among infants and children. Recommendations and Reports. 2010; 59(RR11): 1–18. Reference Source\n\nWHO. Prevalence and distribution of Pneumococcus in Sub-Saharan Africa. Geneva Switzerland: World Health Organization. 2009.\n\nWHO, UNICEF. Global immunization vision and strategy 2006-2015. Geneva, Switzerland: World Health Organization. 2010; 2005. Reference Source\n\nGupta S, Sarosi L: Pneumococcal Disease Epidemiology, Control and Prevention. US Department of Health and Human Services. 2001; 249–263.\n\nChien YW, Klugman KP, Morens DM: Efficacy of whole-cell killed bacterial vaccines in preventing pneumonia and death during the 1918 influenza pandemic. J Infect Dis. 2010; 202(11): 1639–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO. 23-valent pneumococcal polysaccharide vaccine. WHO position paper. Wkly Epidemiol Rec. 2008; 83(42): 373–84. PubMed Abstract\n\nPark IH, Pritchard DG, Cartee R, et al.: Discovery of a new capsular serotype (6C) within serogroup 6 of streptococcus pneumoniae. J Clin Microbiol. 2007; 45(4): 1225–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO. Therapeutic challenges as a result of penicillin-resistance. WHO position paper. Wkly Epidemiol Rec. 2010; 91(24): 244–63.\n\nJohansson N, Kalin M, Hedlund J: Clinical impact of combined viral and bacterial infection in patients with community-acquired pneumonia. Scand J Infect Dis. 2011; 43(8): 609–615. PubMed Abstract \| Publisher Full Text\n\nNuorti JP, Butler JC, Farley MM, et al.: Cigarette smoking and invasive pneumococcal disease. Active Bacterial Core Surveillance Team. N Engl J Med. 2000; 342(10): 681–9. PubMed Abstract \| Publisher Full Text\n\nWhitney GC, Farley MM, Hadler J, et al.: Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med. 2003; 348(18): 1737–46. PubMed Abstract \| Publisher Full Text\n\nKnoll MD, Moïsi JC, Muhib FB, et al.: Standardizing surveillance of pneumococcal disease. Clin Infect Dis. 2009; 48(Suppl 2): S37–S48. PubMed Abstract \| Publisher Full Text\n\nCynthia. Approval of 10-valent pneumococcal conjugate vaccine by European Commission. Cambridge Med J. 2010; 243(27): 284–98.\n\nLaible G, Spratt BG, Hakenbeck R: Interspecies recombinational events during the evolution of altered PBP 2x genes in penicillin-resistant clinical isolates of Streptococcus pneumoniae. Mol Microbiol. 1991; 5(8): 1993–2002. PubMed Abstract \| Publisher Full Text\n\nWHO. Effectiveness of pneumococcal vaccine in Kenya. Nairobi, Kenya: World Health Organization. 2004; 2008.\n\nWorld Health Organization. WHO/UNICEF joint reporting process. Geneva. 2009. Reference Source\n\nCasey W: Launch of PCV-10 vaccine in developing countries and the effect it has on children. J Pneumococcal Dis Med. 2010; 178(13): 261–83.\n\nEngvall E, Perlman P: Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of immunoglobulin G. Immunochemistry. 1971; 8(9): 871–4. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization. Position paper on PCV-10 in developing countries. Geneva. 2011.\n\nAnadiotis M: Evaluation of the immunogenicity to determine the effectiveness of the PCV-10 vaccine in Kenya. J Infect Dis. 2002 / 2003; 180(5): 1569–1576.\n\nWuorimaa T, Dagan R, Eskola J, et al.: Tolerability and immunogenicity of an eleven-valent pneumococcal conjugate vaccine in healthy toddlers. Pediatr Infect Dis J. 2001; 20(3) 272–277. PubMed Abstract \| Publisher Full Text\n\nAvery OT, Goebel WF: Chemo-Immunological Studies on Conjugated Carbohydrate-Proteins : V. The Immunological Specifity of an Antigen Prepared by Combining the Capsular Polysaccharide of Type III Pneumococcus with Foreign Protein. J Exp Med. 1931; 54(3): 437–47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHicks LA, Harrison LH, Flannery B, et al.: Incidence of pneumococcal disease due to non-pneumococcal conjugate vaccine (PCV7) serotypes in the United States during the era of widespread PCV7 vaccination, 1998–2004. J Infect Dis. 2007; 196(9): 1346–1354. PubMed Abstract \| Publisher Full Text\n\nWhitney CG, Pilishvili T, Farley MM, et al.: Effectiveness of seven-valent pneumococcal conjugate vaccine against invasive pneumococcal disease: a matched case-control study. Lancet. 2006; 368(9546): 1495–502. PubMed Abstract \| Publisher Full Text\n\nMillar EV, Watt JP, Bronsdon MA, et al.: Indirect effect of 7-valent pneumococcal conjugate vaccine on pneumococcal colonization among unvaccinated household members. Clin Infect Dis. 2008; 47(8): 989–996. PubMed Abstract \| Publisher Full Text\n\nSiemieniuk RA, Gregson DB, Gill MJ: The persisting burden of invasive pneumococcal disease in HIV patients: an observational cohort study. BMC Infect Dis. 2011; 11: 314. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarocchi MA, Censini S, Rappuoli R: Vaccines in the era of genomics: the pneumococcal challenge. Vaccine. 2007; 25(16): 2963–73. PubMed Abstract \| Publisher Full Text\n\nGrijalva CG, Nuorti JP, Arbogast PG, et al.: Decline in pneumonia admissions after routine childhood immunisation with pneumococcal conjugate vaccine in the USA: a time-series analysis. Lancet. 2007; 369(9568): 1179–86. PubMed Abstract \| Publisher Full Text\n\nTsai CJ, Griffin MR, Nuorti JP, et al.: Changing epidemiology of pneumococcal meningitis after the introduction of pneumococcal conjugate vaccine in the United States. Clin Infect Dis. 2008; 46(11): 1664–72. PubMed Abstract \| Publisher Full Text\n\nO'Brien KL, Wolfson LJ, Watt JP, et al.: Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet. 2009; 374(9693): 893–902. PubMed Abstract \| Publisher Full Text\n\nJohns Hopkins Bloomberg School of Public Health, International Vaccine Access Center. VIMS Report: Global vaccine introduction, 2013.\n\nWhitney CG, Farley MM, Hadler J, et al.: Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med. 2003; 348(18): 1737–46. PubMed Abstract \| Publisher Full Text\n\nAmerican Academy of Paediatrics. Committee on Infectious Diseases. Policy statement: recommendations for the prevention of pneumococcal infections, including the use of pneumococcal conjugate vaccine (Prevnar), pneumococcal polysaccharide vaccine, and antibiotic prophylaxis. Paediatrics. 2000; 106(2 pt 1); 362–6. PubMed Abstract\n\nPoehling KA, Talbot TR, Griffin MR, et al.: Invasive pneumococcal disease among infants before and after introduction of pneumococcal conjugate vaccine. JAMA. 2006; 295(14) 1668–74. PubMed Abstract \| Publisher Full Text\n\nWalekhwa M, Muturi M, Bukusi E: Dataset 1 in: Immunogenicity of 10-valent pneumococcal conjugate vaccine among infants attending Mbagathi District Hospital, Kenya. F1000Research. 2015. Data Source" }	[ { "id": "9459", "date": "14 Jul 2015", "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\nThe paper by Walekhwa et al. aimed to determine the serum concentration of IgG antibodies as an indicator of immunogenicity, alongside the assessment of socio-demographic factors that affect IgG antibody levels in infants immunized with 10-valent pneumococcal conjugate vaccine (PCV-10) at the Mbagathi District Hospital in Kenya. Title needs to include biological mothers; the abstract requires some correction on the results section. In particular the authors say “the remaining 2%” while the actual remaining is not 2%. Article content Sample sizeThe author did a sample size calculation of 384 participants. However in their study, they enrolled 318 individuals. This has an impact on the conclusions they can make whether they have enough statistical power. DataThe authors are comparing antibody levels between the alcoholic mothers versus mothers who do not take alcohol. Antibody levels are usually not normally distributed. Was this the case? If yes, then the authors need to compare the levels using a Mann Whitney test. The authors use a cut off or 0.35 mg/dl for seropositivisity, however they do not indicate the number of infants above this cut-off. This is because they categorise 0.34–0.36 mg/dl antibody levels in the same group. The authors report on maternal diet during pregnancy; gap with other children and chronic illness. However this data is not easily visible, requiring installation of specific software. It would be better if the data were visible within the article, or attached in a more common formatAs such there are fundamental concerns with the analysis. Results and discussion Have sub sections with titles for each section. Each section of the results need to be focussed for instance: Demographic characteristics of the participants (infants and mothers): the authors need to present information on characteristics of the participants in their study. This include information on mean or median age, sex, weight, breastfeeding status, etc. that they collected. This is a standard practice in most scientific papers. Antibody levels in infants Antibody levels in biological mothers This will make the reader “walk” with the authors as they reads the article. Acknowledgements Should not include co-authors as they are already coauthors Conclusion Based on all these factors, the authors should revise this manuscript", "responses": [ { "c_id": "1603", "date": "17 Sep 2015", "name": "Michael Walekhwa", "role": "Author Response", "response": "I have read through the corrections posted by Dr. Ondigo and I am incorporating in this article as appropriate.Thank you so much for your wonderful input." } ] }, { "id": "12496", "date": "18 Feb 2016", "name": "Paul Licciardi", "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 by Walekhwa and colleagues presents data on the immunogenicity of PCV10 among infants and their biological mothers attending the Mbagathi District Hospital in Kenya. This is an interesting study but I have several concerns as it is currently written: The authors should reconcile the sample size of 384 calculated with 318 actually recruited to determine whether the study is adequately powered for the primary outcomes of the study. The IgG data is presented in a confusing way – first the units mg/dl is different to how researchers in the field would report this (as µg/ml or mg/L). The 0.35 mg/dl cut-off the authors state as protective equates to 3.5µg/ml which is 10-fold above the actual protective threshold for PCVs (0.35µg/ml). This also needs to be corrected in the Introduction section (paragraph 4, line 6). The data should be reflected in this format to improve clarity. Secondly, there is no serotype-specific IgG values reported which suggests this may be a total anti-pneumococcal IgG assay? This may explain the IgG values reported. However, the reference cited by the authors for the pneumococcal ELISA should be different as this does not include any ELISA data or method. The authors should provide information on what type of ELISA was performed on these samples. I also suggest reporting the data as GMC +/- 95% CI and the % responding ≥ 0.35 µg/ml. In some instances in the results section, the p-values are reported as p>0.05 or p>0.01 when it should be p<0.05 and p<0.01. Please check throughout for consistency and in the abstract. Demographics of the infants and their mothers also need to be included. A Table or summary of the results for the statistical analyses, especially for the data on alcohol consumption and breastfeeding should be included in the Results section.", "responses": [] } ]	1	https://f1000research.com/articles/4-165
https://f1000research.com/articles/4-161/v1	23 Jun 15	{ "type": "Research Article", "title": "Proximal Femur Size and Geometry in Cementless Total Hip Arthroplasty Patients", "authors": [ "Darrell L. Moulton", "Ronald W. Lindsey", "Zbigniew Gugala", "Darrell L. Moulton", "Ronald W. Lindsey" ], "abstract": "Introduction: Accurate femoral prosthesis press-fit is essential for successful cementless total hip arthroplasty (cTHA) and dependent upon proximal femur size and geometry. Study objectives were to determine the variability of proximal femur size and geometry in primary cTHA patients and correlate them with patient demographics and body mass index (BMI).Methods: Medical records of 127 consecutive primary cTHA patients were reviewed retrospectively. The demographic (ethnicity, sex, age) and BMI data were collected. Intertrochanteric (IT) distance, inner/outer proximal femur diameters and cortical thickness for the subtrochanteric (ST) and cortical diaphyseal (DP) regions were measured from anteroposterior radiographs. Descriptive statistics were used to correlate patient demographics and BMI with radiographic measurements.Results: The study included 96 cTHA patients (mean age 60 years, range 22-91 years; 34 females; 72 Caucasian, 18 Black, and six Hispanic) with four underweight; 13 normal; 34 overweight, and 45 obese BMI. No correlation existed for patient age or race with radiographic measurements. Males had significantly larger proximal femur dimensions and cortical thickness than females. No BMI correlations existed for IT distance; BMI was directly proportional to outer diameter and cortical thickness in ST and DP regions, and inversely proportional to inner diameter in these regions.Discussion: Greater proximal femur size appears to correlate with gender, but not with age or race. Larger subtrochanteric and diaphyseal outer diameters are significantly associated with higher BMI. A trend exists for larger subtrochanteric and diaphyseal inner diameters to be associated with lower BMI. These findings may have implications for optimal cTHA femoral component design.", "keywords": [ "cementless total hip arthroplasty", "proximal femur morphology", "femoral stem", "body mass index" ], "content": "Introduction\n\nCementless total hip arthroplasty (cTHA) is one of the most common orthopaedic procedures performed today. Much of its success is due to modern technological advances in implant design that permit cementless press-fit femoral component fixation1–18. Studies have shown that appropriate cementless femoral component size is critical for optimal implant initial stability, and maximizing femoral implant medullary canal fill enhances the endosteal contact between implant and bone to promote bone ingrowth, and therefore improved long-term outcome19–21.\n\nStandard tapered cTHA stems typically rely on three-point fixation, which achieves more proximal load transfer and thereby decreases risk for stress shielding. Nonetheless, subsidence may occur owing to inadequate distal press-fit. More recent designs of anatomic stems aim to reproduce the normal contours of the intramedullary cavity to allow a more natural load distribution over the proximal femur without relying on a specific fixation point22. The clinical relevance of these evolving designs is the inhibition of aseptic loosening and stress shielding, and the optimization of implant longevity.\n\nBecause the quality of cementless femoral component press-fit is dependent upon matching the implant size to the dimensions and geometry of the proximal femur, determining the relationship between these parameters and basic patient characteristics such as age, race, gender, and body habitus may assist with operative planning. However, the priority of selecting the largest possible femoral stem size to achieve a tight fit often results in a substantial alteration of the implant’s biomechanical properties. The bending stiffness of a cylindrical stem increases exponentially to the fourth power with increasing a stem diameter, and this phenomenon could be a specific concern in cTHA patients with a proximal femur size and geometry inversely related to their body mass, and when proximal femur size and geometry are incommensurate with body habitus, cTHA patients are potentially subjected to femoral implants that are too stiff or too flexible.\n\nWhile several investigators have reported significant proximal femur anatomic variations in regard to patient gender, a correlation of body habitus with proximal femur size and geometry has not been reported. The objective of this study was to determine the variability of proximal femur size and geometry in primary cTHA patients, and determine its correlation with patient age, gender, ethnicity, and body habitus.\n\n\nMaterials & methods\n\nOne hundred twenty-seven patients with primary cTHAs performed at a single institution (University of Texas Medical Branch, Galveston, TX) from 2004 to 2008 were studied retrospectively. Inclusion criteria mandated that the patient was an adult who underwent elective unilateral total hip arthroplasty with a cementless femoral stem and had adequate postoperative radiographs of the involved hip and proximal femur. Exclusion criteria were patients with THAs that were bilateral, revisions, post-infection, acquired (post-traumatic) or due to congenital deformity of the proximal femur, or cemented THA. The study was conducted in compliance with the University of Texas Medical Branch policies and regulations regarding human subject research following study protocol review and approval by the Institutional Review Board (approval IRB #08-156).\n\nAge, gender, ethnicity, and body mass index (BMI) at the time of surgery were documented for each cTHA patient. BMI was defined as the ratio of the patient’s weight (kilograms) to the square of the height (meters). Patients were sub-categorized in accordance with their BMI as follows: underweight (BMI less than 18.5), normal weight (BMI between 18.5 and 24.9), overweight (BMI between 25 and 29.9), obese (BMI greater than 30).\n\nThe medullary canal size and cortex thickness of the proximal femur were determined by a series of measurements of the involved hip and proximal femur measured on a postoperative anteroposterior (AP) plain radiograph completed within two weeks of the surgery. The hip/proximal femur radiograph was performed using a standard technique with the x-ray beam centered on the hip with neutral pelvic and lower extremity rotation. All study radiographs were accessed using the digital Picture Archiving and Communication System (IntelliSpace PACS; Philips, Andover, MA, USA) at our institution.\n\nProximal femur radiographic measurements included the intertrochanteric distance, the inner and outer femur cortical diameters of the femur at the subtrochanteric and diaphyseal regions, as well as the subtrochanteric and diaphyseal cortical thickness. The intertrochanteric distance (ITD) was measured from the tip of the greater trochanter to the tip of the lesser trochanter (Figure 1). The subtrochanteric inner cortical diameter (STID) and outer cortical diameter (STOD) were measured from just proximal to the distal inferior edge of the lesser trochanter perpendicular to the long axis of the diaphysis. The subtrochanteric cortical thickness (STCT) was calculated by subtracting the inner cortical diameter from the outer cortical diameter, and then dividing that figure in half (which assumes that the femur cross-section at that level is circular). The diaphyseal inner cortical diameter (DID), diaphyseal outer cortical diameter (DOD), and diaphyseal cortical thickness (DCT) were determined at a level of 4 cm proximal to the tip of the cementless femoral stem by using the same calculations used for the subtrochanteric region.\n\nProximal femur size was established from measurements of intertrochanteric distance (ITD), outer and inner distance, and cortical thickness in the subtrochanteric region (STOD, STID, STCT, respectively), and the same in the diaphyseal region (DOD, DID, DCT, respectively) on AP plain radiographs. The intertrochanteric distance was measured between the tips of the greater and lesser trochanters; the subtrochanteric measurements were taken at the level immediately below the lesser trochanter, and the diaphyseal measurements were taken at the level of 4 cm above the tip of the cementless femoral stem. A magnification factor was established for each radiograph by measuring the outer diameter of the acetabular cup and comparing it to the known outer diameter of the implant obtained from the operative report, and all measurements were adjusted accordingly.\n\nIn order for these measurements to reflect the true dimensions of the osseous anatomy, a magnification factor was established for each radiograph. The magnification factor was determined by digitally measuring the outer diameter of the acetabular cup and comparing that measurement to the known outer diameter of the implant obtained from the operative report. This magnification calculation was used to adjust all radiographic measurements.\n\nProximal femur geometry was assessed using the intertrochanteric-to-subtrochanteric and subtrochanteric-to-diaphyseal ratios suggested in previous reports23,24 as a better measure of proximal femur morphology compared with individual anatomic dimensions. Also, mean total, cortical and medullary cross-sectional areas of subtrochanteric and diaphyseal regions were calculated and compared with patient gender and BMI.\n\nDescriptive statistics (SAS 9.3; SAS Institute Inc, Cary, NC, USA) were used to describe the sample characteristics and distribution of the proximal femur size and geometry (outcome variables). Pearson correlation test and ANOVA were used to test the bivariate association between outcome variable and sample characteristics (age, sex, race, and BMI) depending on the whether the sample characteristics variable was continuous or categorical. Multiple regression models were used to test the association between BMI and outcome variables, adjusting for age, sex, and race. The unadjusted and adjusted means of the outcome variables were also generated using regression models. A p-value threshold of less than 0.05 was considered statistically significant. Post hoc power analysis confirmed that the study sample size has sufficient power to detect a 20% difference in the designated parameters of the proximal femur morphology.\n\n\nResults\n\nThe study consisted of 96 cTHA patients who met the inclusion/exclusion criteria. The patients had the following demographic characteristics: mean age of 60.1 years ranging between 22 and 91 years, 34 patients were females (35.4%) and 62 males (64.6%); 72 patients were Caucasian (75%), 18 Black (18.7%), and 6 Hispanic (6.3%) (Table 1). The BMI data included four (4.2%) patients who were underweight (BMI<18.5); 13 (13.5%) normal (BMI 18.5–24.9); 34 (35.4%) overweight (BMI 25–29.9), and 45 (46.9%) obese (BMI>30) (Figure 2). Proximal femur mean regional dimensions and standard deviations of the outcome variables for all patients in the study are depicted in Table 2.\n\n$Mean (SD) for continuous variables and N (%) for categorical variables\n\nPatients were categorized as underweight (BMI less than 18.5), normal weight (BMI between 18.5 and 24.9), overweight (BMI between 25 and 29.9), obese (BMI greater than 30).\n\nITD, Intertrochanteric Distance; STOD/STID, Subtrochanteric Outer/Inner Distance; STCT, Subtrochanteric Cortical Thickness; DOD/DID, Diaphyseal Outer/Inner Distance; DCT, Diaphyseal Cortical Thickness; SD, standard deviation.\n\nNo significant correlation existed for patient age or race and all radiographic measurements (Table 3). Males had statistically significant larger proximal femur intertrochanteric (p<0.0001), subtrochanteric (p<0.008), and diaphyseal diameters (p<0.001) as well as cortical thickness compared to females (p=0.002). BMI was directly proportional to outer diameter and cortical thickness in both subtrochanteric and diaphyseal regions; however, BMI was inversely proportional to inner diameter in these same regions. BMI groups exhibited statistically significant correlation for measurements obtained for the subtrochanteric and diaphyseal, but not for the intertrochanteric region. Age and race were not significantly associated with any of the outcome variables (Table 4).\n\n$Adjusted for age, gender and ethnicity using multiple regression analysis\n\nIncreasing BMI was significantly associated with higher STOD, STCT, DOD, and DCT values (P=0.02, 0.008, 0.004, and 0.0001, respectively, in unadjusted tests). These significant associations persist after adjusting for age, sex and race (Table 5). Figure 3–Figure 5 graphically represent the associations between BMI and the outcome variables. Table 6 demonstrates a significant linear trend between BMI and STOD, STCT, DOD and DCT. No significant correlation was detected between BMI and intertrochanteric distance, subtrochanteric inner diameter, or diaphyseal inner diameter.\n\n*Adjusted for age, gender, and ethnicity\n\nError bars represent standard deviation.\n\nError bars represent standard deviation.\n\nError bars represent standard deviation.\n\nThe proximal femur geometry assessed form determining the intertrochanteric-to-subtrochanteric and subtrochanteric-to-diaphyseal did not demonstrate a statistically significant association with patient gender (Table 6). The mean cross-sectional areas of subtrochanteric and diaphyseal regions versus gender and BMI group are depicted in Table 7 and Table 8, respectively. No statistical significant associations were observed between these parameters of the proximal femur geometry and patient gender. The normal-weight group had the largest average canal width and on average received the largest femoral stems, whereas the overweight and obese patients had smaller canals, and received smaller femoral components (Table 8).\n\n$ Percentage of total cross-sectional area\n\n\nDiscussion\n\nSeveral cTHA studies have emphasized the importance of femoral fit in achieving primary implant stability, reducing early prosthetic loosening, and encouraging bony ingrowth2–9. Understanding the bony anatomy of the proximal femur is a prerequisite for optimizing femoral implant press-fit and determining the impact of factors such as ethnicity, gender, age, and body habitus10,12. At present, clinical priority is given to selecting the largest possible femoral stem to accomplish a tight fit, despite increased implant size association with a considerable change in the stem biomechanical properties (four-power increase in bending stiffness in relation to stem diameter). Hence, the size and geometry of the proximal femur may have significant implications for the cementless femoral stem biomechanical characteristics, especially in patients for whom a large discrepancy exists between their proximal femur dimensions and their body habitus.\n\nPrevious anthropometric studies have compared the variability of femoral bony architecture among different races. Travison et al.25 used bone densitometry to show that African-Americans have larger intertrochanteric and diaphyseal outer diameters compared to Hispanics and Caucasians. Marshall et al.26 determined that African-Americans and Asian men have greater femoral neck and shaft mean cortical thickness compared to Caucasian men. Other studies27,28 demonstrated that African-American women had a larger cross-sectional area, larger outer diameter, smaller inner diameter, and thicker cortices in the femoral neck and shaft compared to Caucasian women, and also have smaller inner and outer diameters, but thicker cortices in the intertrochanteric region. This is in contrast with the results of the present study, which did not establish a statistically significant correlation between race and the proximal dimensions. These findings corroborate those of Peacock et al.29 and Saeed et al.30, who reported computed tomography measurements of the proximal femur to be independent of race for both sexes. These studies, however, have substantiated significant differences in bone mineral density among these ethnic groups25,26,29,31.\n\nThe literature has shown that men generally have a larger bony architecture compared to women, even when adjusted for confounding variables29,32,33. These differences have promoted the development of gender-specific arthroplasty implant designs. Kostamo et al.34 in a large primary THA cohort demonstrated that differences in clinical outcome scores were found only in the WOMAC pain score in favor of the female cohort (39.4 versus 36.1), whereas the survivorship and revision rate were not significantly different. Men required larger femoral stems with greater stem lengths, neck offset, and neck lengths. Despite the gender variations in proximal femur anatomy, the authors concluded that current implant systems in their versatility sufficiently address the different size and offset needs of both male and female patients. Similar results have been reported for gender-specific total knee arthroplasty components35.\n\nIn the present study a statistically significant difference in BMI was determined between males and females for all outcome variables (Table 4). The proximal femur morphology assessment involving ratios of geometric measures as opposed to individual anatomic dimensions demonstrated that, although the single measures of the specific proximal femur regions exhibited statistically significant differences (Table 2), the intertrochanteric-to-subtrochanteric and subtrochanteric-to-diaphyseal ratios did not (Table 6 and Table 7). These findings suggest that although the bony anatomy of men is larger than that of women, the proportional increase in proximal femur size, and subsequent shape, is similar for both. Also, the cortical and medullary cross-sectional areas as a percentage of the total cross-sectional area in the subtrochanteric and diaphyseal regions were similar for both men and women (Table 8). Lamellar bone stiffness is largely determined by its radius and cortical content. Therefore, although women may exhibit less lamellar bone stiffness, this difference is in proportion to that of men. Furthermore, a disparity did not exist between anatomic regions for either sex, and this may suggest that although men may require larger femoral implants compared to women, gender-specific implant geometry may not be necessary as the proximal femur regional anatomic proportions are equivalent.\n\nBone mass decreases with age after its peak in the third decade of life; however, studies have shown that the outer diameter of bone slightly increases with age, possibly owing to continued periosteal appositional growth25,33,36,37. But to produce an age-associated decrease in bone density while its outer diameter is increasing would require thinning of the cortex. Our study did not demonstrate a statistically significant correlation between age and proximal femur geometry, although the aforementioned trends may be seen (Figure 6 and Figure 7). Bone adaptation to mechanical loading (Wolff’s law) is site specific because of the unique magnitude, direction, and type of load in each anatomic region. Among the major determinants of the loads experienced at the hip joint and subtrochanteric region of the femur is body weight. In a bone densitometry study, Petit et al.38,39 demonstrated that the femoral neck and shaft in children and adolescents who were overweight due to a higher percentage of lean body mass had larger diameters, cross-sectional areas, and cortical thicknesses compared to normal-weight individuals. Van der Meulen et al.40 indicated that of age, pubertal stage, body mass, and height, body mass is a stronger predictor of femoral cross-sectional properties than age, pubertal stage, or height. Moreover, these authors concluded that the correlation of body mass with femoral cross-sectional structure is independent of gender. Interestingly, other authors have indicated that body weight and femoral dimensions may possess common genetic factors41–43.\n\nIn our study, there were significant differences in subtrochanteric (STOD, STCT) and diaphyseal (DOD and DCT) measurements versus individual BMI groups (Table 4), and versus BMI as a continuous variable (Table 5). This correlation demonstrated the significant linear trends between STOD, STCT, DOD and DCT and BMI. The subtrochanteric inner diameter (STID) versus BMI groups demonstrated statistical significance only after being adjusted for age, sex, and race; the same was observed for the diaphyseal inner diameter region (Table 5). In both regions, the outer diameters increased with larger BMI, but the cortical thickness also increased, resulting in fairly consistent inner diameters among BMI groups. On the other hand, the differences in intertrochanteric measurement (ITD) were not statistically significant. Intertrochanteric distance can be highly variable on anterior-posterior plain radiography with varying degrees of hip rotation. Although AP hip radiographs in our institution are routinely attempted with the femur in neutral alignment, many patients with longstanding hip disease may have had contractures causing external proximal femur rotation. Our study showed a less than 5% variation in the ITD measurements.\n\nOne concern with press-fit femoral components is the disparity that frequently exists between the anatomy/configuration of the human proximal femur and the specific design and size limitations of any implant system. Our study confirms that a wide variation exists in proximal femur geometry among hip arthroplasty patients. The notion that current off-the-shelf femoral components do not accurately accommodate this wide variation is supported by reports of persistent bone stress shielding or implant subsidence despite advances in stem design. All cTHA study patients received a tapered stem, and the average proximal and distal cross-sectional area of the implanted femoral component according to BMI group is depicted in Table 8. The normal-weight group had the largest average canal width and, on average, received the largest femoral stems. By comparison, overweight and obese patients had smaller canals, and received smaller femoral components. These findings suggest that the normal-weight patients may receive implants which for their body habitus, are excessively stiff; conversely, obese patients may actually be receiving implants that are not stiff enough. Although, to our knowledge, there has not been a study that relates implant stiffness and fatigue resistance to implant longevity among different BMI groups, several studies suggest no differences in the clinical or radiographic outcomes among different BMI groups44,45.\n\nA correlation of body weight with the incidence or severity of thigh pain has not been demonstrated. Several studies have also attempted to correlate thigh pain with stem size; however, their results have been inconclusive7,46. Stem size and/or geometry clearly influences thigh pain, as a stem with a better fit will result in more stability and bony ingrowth, and theoretically, less thigh pain47,48. Similarly, stem stiffness may influence thigh pain as a stem with a stiffness closer to that of the host femur would result in more uniform stress transfer48,49. Body weight could also have a great influence on the incidence of thigh pain, especially if there is a considerable mismatch of host bone to implant stiffness. In the present study, normal-weight patients had smaller outer diameters with thinner cortices compared to obese patients, but actually received larger implants owing to their larger canals. Theoretically, this mismatch in stiffness may the cause of thigh pain in normal-weight individuals, however, this has not been proven clinically44.\n\nThe present study has several limitations. Selection bias may have been introduced into the study - the typical study patients were older and obese, while younger, underweight or normal-weight patients were underrepresented. The study demographic data may also be inconsistent or incomplete regarding ethnicity and body mass. The designation of ethnicity was used as recorded in the medical records without a standard definition or verification. Also, the study design did not account for physical activity, or the lack thereof, as well as any differences between low and high body mass as previously discussed. In addition, each patient underwent cTHA with a press-fit femoral component, and the amount of endosteal cortical bone removed through reaming and/or broaching was undetermined. Finally, the complexity of the proximal femur geometry was assessed only with plain radiography.\n\nIn summary, the present study shows a strong correlation between the proximal femur dimension and gender, but not with age or ethnicity. This correlation is statistically significant for outer diameter and cortical thickness in both the subtrochanteric and diaphyseal regions of the proximal femur. Conversely, there is a non statistically significant reverse trend between BMI and inner medullary diameter in both regions. The study findings suggest that gender-specific femoral implants based on presumed differences between male and female proximal femur size but not geometry may be warranted. The development of cTHA femoral components that more accurately address the proximal femur variations specifically reflected by the patient overall and local biomechanical demands (BMI, bone quality, etc.) may have considerable merit in extending implant longevity.\n\n\nConclusions\n\nThe present study demonstrates a strong correlation between gender and the size of the proximal femur, but no correlation with age or race. There is a direct correlation between BMI and proximal femur size; this correlation is statistically significant for outer diameter and cortical thickness in both the subtrochanteric and diaphyseal regions. Conversely, there is a non statistically significant reverse trend between BMI and inner medullary diameter in both regions.\n\nStudy results suggest that gender-specific cTHA femoral component matching male versus female proximal femur size may be indicated, and the development of femoral stems that addresses the proximal femur anatomic variations together with its matched biomechanical properties may have considerable merit.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for ‘Proximal femur size and geometry in cementless total hip arthroplasty patients’, 10.5256/f1000research.6554.d4979650", "appendix": "Author contributions\n\n\n\nDLM - data collection, data analysis, manuscript preparation; RWL - study design, manuscript preparation; ZG - study design, data analysis, manuscript preparation. All authors have seen and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declared no competing interests.\n\n\nGrant information\n\nThe authors declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe thank Alai Tan, PhD, for assistance in statistical analysis and Suzanne Simpson, BA, for editing the manuscript.\n\n\nReferences\n\nBourne R, Mukhi S, Zhu N, et al.: Role of obesity on the risk for total hip or knee arthroplasty. Clin Orthop Relat Res. 2007; 465: 185–8. PubMed Abstract\n\nBurke DW, O’Connor DO, Zalenski EB, et al.: Micromotion of cemented and uncemented femoral components. J Bone Joint Surg Br. 1991; 73(1): 33–7. PubMed Abstract\n\nCallaghan JJ, Fulghum CS, Glisson RR, et al.: The effect of femoral stem geometry on interface motion in uncemented porous-coated total hip prostheses. Comparison of straight-stem and curved-stem designs. J Bone Joint Surg Am. 1992; 74(6): 839–48. PubMed Abstract\n\nChristie M, Brinson MF: Proximal/distal mismatch: type A and C femurs. Orthopedics. 2005; 28(9 Suppl): s1033–6. PubMed Abstract\n\nEffenberger H, Graf R, May M, et al.: Anatomically adapted, HA coated SBG stem--ten years of successful implantation. Biomed Tech (Berl). 2004; 49(10): 290–4. PubMed Abstract \| Publisher Full Text\n\nEngh CA Jr, Mohan V, Nagowski JP, et al.: Influence of stem size on clinical outcome of primary total hip arthroplasty with cementless extensively porous-coated femoral components. J Arthroplasty. 2009; 24(4): 554–9. PubMed Abstract \| Publisher Full Text\n\nHaraguchi K, Sugano N, Nishii T, et al.: Comparison of fit and fill between anatomic stem and straight tapered stem using virtual implantation on the ORTHODOC workstation. Comput Aided Surg. 2001; 6(5): 290–6. PubMed Abstract\n\nHelm CS, Greenwald AS: The rationale and performance of modularity in total hip arthroplasty. Orthopedics. 2005; 28(9 Suppl): s1113–5. PubMed Abstract\n\nHoward JL, Hui AJ, Bourne RB, et al.: A quantitative analysis of bone support comparing cementless tapered and distal fixation total hip replacements. J Arthroplasty. 2004; 19(3): 266–73. PubMed Abstract \| Publisher Full Text\n\nHusmann O, Rubin PJ, Leyvraz PF, et al.: Three-dimensional morphology of the proximal femur. J Arthroplasty. 1997; 12(4): 444–50. PubMed Abstract \| Publisher Full Text\n\nKim YH, Kim VE: Cementless porous-coated anatomic medullary locking total hip prostheses. J Arthroplasty. 1994; 9(3): 243–52. PubMed Abstract \| Publisher Full Text\n\nLaine HJ, Lehto MU, Moilanen T: Diversity of proximal femoral medullary canal. J Arthroplasty. 2000; 15(1): 86–92. PubMed Abstract \| Publisher Full Text\n\nLaine HJ, Puolakka TJ, Moilanen T, et al.: The effects of cementless femoral stem shape and proximal surface texture on 'fit-and-fill' characteristics and on bone remodeling. Int Orthop. 2000; 24(4): 184–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMalchau H, Wang YX, Kärrholm J, et al.: Scandinavian multicenter porous coated anatomic total hip arthroplasty study. Clinical and radiographic results with 7- to 10-year follow-up evaluation. J Arthroplasty. 1997; 12(2): 133–48. PubMed Abstract \| Publisher Full Text\n\nNishihara S, Sugano N, Nishii T, et al.: Comparison of the fit and fill between the Anatomic Hip femoral component and the VerSys Taper femoral component using virtual implantation on the ORTHODOC workstation. J Orthop Sci. 2003; 8(3): 352–60. PubMed Abstract \| Publisher Full Text\n\nNoble PC, Alexander JW, Lindahl LJ, et al.: The anatomic basis of femoral component design. Clin Orthop Relat Res. 1988; 235: 148–65. PubMed Abstract\n\nVan der Wal BC, de Kramer BJ, Grimm B, et al.: Femoral fit in ABG-II hip stems, influence on clinical outcome and bone remodeling: a radiographic study. Arch Orthop Trauma Surg. 2008; 128(10): 1065–72. PubMed Abstract \| Publisher Full Text\n\nViceconti M, Pancanti A, Dotti M, et al.: Effect of the initial implant fitting on the predicted secondary stability of a cementless stem. Med Biol Eng Comput. 2004; 42(2): 222–9. PubMed Abstract \| Publisher Full Text\n\nEngh CA Jr, Culpepper WJ 2nd, Engh CA: Long-term results of use of the anatomic medullary locking prosthesis in total hip arthroplasty. J Bone Joint Surg Am. 1997; 79(2): 177–84. PubMed Abstract\n\nCameron HU, Pilliar RM, MacNab I: The effect of movement on the bonding of porous metal to bone. J Biomed Mater Res. 1973; 7(4): 301–11. PubMed Abstract \| Publisher Full Text\n\nPanisello JJ, Herrero L, Canales V, et al.: Long-term remodeling in proximal femur around a hydroxyapatite-coated anatomic stem: ten years densitometric follow-up. J Arthroplasty. 2009; 24(1): 56–64. PubMed Abstract \| Publisher Full Text\n\nØstbyhaug PO, Klaksvik J, Romundstad P, et al.: An in vitro study of the strain distribution in human femora with anatomical and customised femoral stems. J Bone Joint Surg Br. 2009; 91(5): 676–82. PubMed Abstract \| Publisher Full Text\n\nGregory JS, Testi D, Stewart A, et al.: A method for assessment of the shape of the proximal femur and its relationship to osteoporotic hip fracture. Osteoporos Int. 2004; 15(1): 5–11. PubMed Abstract \| Publisher Full Text\n\nO’Neill TW, Grazio S, Spector TD, et al.: Geometric measurements of the proximal femur in UK women: secular increase between the late 1950s and early 1990s. Osteoporos Int. 1996; 6(2): 136–40. PubMed Abstract \| Publisher Full Text\n\nTravison TG, Beck TJ, Esche GR, et al.: Age trends in proximal femur geometry in men: variation by race and ethnicity. Osteoporos Int. 2008; 19(3): 277–87. PubMed Abstract \| Publisher Full Text\n\nMarshall LM, Zmuda JM, Chan BK, et al.: Race and ethnic variation in proximal femur structure and BMD among older men. J Bone Miner Res. 2008; 23(1): 121–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelson DA, Barondess DA, Hendrix SL, et al.: Cross-sectional geometry, bone strength, and bone mass in the proximal femur in black and white postmenopausal women. J Bone Miner Res. 2000; 15(10): 1992–97. PubMed Abstract \| Publisher Full Text\n\nNelson DA, Pettifor JM, Barondess DA, et al.: Comparison of cross-sectional geometry of the proximal femur in white and black women from Detroit and Johannesburg. J Bone Miner Res. 2004; 19(4): 560–65. PubMed Abstract \| Publisher Full Text\n\nPeacock M, Buckwalter KA, Persohn S, et al.: Race and sex differences in bone mineral density and geometry at the femur. Bone. 2009; 45(2): 218–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaeed I, Carpenter RD, Leblanc AD, et al.: Quantitative computed tomography reveals the effects of race and sex on bone size and trabecular and cortical bone density. J Clin Densitom. 2009; 12(3): 330–6. PubMed Abstract \| Publisher Full Text\n\nMikhail MB, Vaswani AN, Aloia JF: Racial differences in femoral dimensions and their relation to hip fracture. Osteoporos Int. 1996; 6(1): 22–4. PubMed Abstract \| Publisher Full Text\n\nKarasik D, Dupuis J, Cupples LA, et al.: Bivariate linkage study of proximal hip geometry and body size indices: the Framingham study. Calcif Tissue Int. 2007; 81(3): 162–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRiggs BL, Melton Iii LJ 3rd, Robb RA, et al.: Population-based study of age and sex differences in bone volumetric density, size, geometry, and structure at different skeletal sites. J Bone Miner Res. 2004; 19(12): 1945–54. PubMed Abstract \| Publisher Full Text\n\nKostamo T, Bourne RB, Whittaker JP, et al.: No difference in gender-specific hip replacement outcomes. Clin Orthop Relat Res. 2009; 467(1): 135–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacDonald SJ, Charron KD, Bourne RB, et al.: The John Insall Award: gender-specific total knee replacement: prospectively collected clinical outcomes. Clin Orthop Relat Res. 2008; 466(11): 2612–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeta M, Lu Y, Keyak JH, et al.: Young-elderly differences in bone density, geometry and strength indices depend on proximal femur sub-region: a cross sectional study in Caucasian-American women. Bone. 2006; 39(1): 152–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeeman E: From density to structure: growing up and growing old on the surfaces of bone. J Bone Miner Res. 1997; 12(4): 509–21. PubMed Abstract \| Publisher Full Text\n\nPetit MA, Beck TJ, Shults J, et al.: Proximal femur bone geometry is appropriately adapted to lean mass in overweight children and adolescents. Bone. 2005; 36(3): 568–76. PubMed Abstract \| Publisher Full Text\n\nPetit MA, McKay HA, MacKelvie KJ, et al.: A randomized school-based jumping intervention confers site and maturity-specific benefits on bone structural properties in girls: a hip structural analysis study. J Bone Miner Res. 2002; 17(3): 363–72. PubMed Abstract \| Publisher Full Text\n\nVan der Meulen MC, Ashford M Jr, Kiratli J, et al.: Determinants of femoral geometry and structure during adolescent growth. J Orthop Res. 1996; 14(1): 22–9. PubMed Abstract \| Publisher Full Text\n\nDemissie S, Dupuis J, Cupples LA, et al.: Proximal hip geometry is linked to several chromosomal regions: genome-wide linkage results from the Framingham Osteoporosis Study. Bone. 2007; 40(3): 743–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang YJ, Dvornyk V, Jian WX, et al.: Genetic and environmental correlations between bone phenotypes and anthropometric indices in Chinese. Osteoporos Int. 2005; 16(9): 1134–1140. PubMed Abstract \| Publisher Full Text\n\nMalkin I, Ermakov S, Kobyliansky E, et al.: Strong association between polymorphisms in ANKH locus and skeletal size traits. Hum Genet. 2006; 120(1): 42–51. PubMed Abstract \| Publisher Full Text\n\nMcLaughlin JR, Lee KR: The outcome of total hip replacement in obese and non-obese patients at 10- to 18-years. J Bone Joint Surg Br. 2006; 88(10): 1286–92. PubMed Abstract \| Publisher Full Text\n\nAndrew JG, Palan J, Kurup HV, et al.: Obesity in total hip replacement. J Bone Joint Surg Br. 2008; 90(4): 424–9. PubMed Abstract \| Publisher Full Text\n\nVresilovic EJ, Hozack WJ, Rothman RH: Incidence of thigh pain after uncemented total hip arthroplasty as a function of femoral stem size. J Arthroplasty. 1996; 11(3): 304–11. PubMed Abstract \| Publisher Full Text\n\nBrown TE, Larson B, Shen F, et al.: Thigh pain after cementless total hip arthroplasty: evaluation and management. J Am Acad Orthop Surg. 2002; 10(6): 385–92. PubMed Abstract\n\nBurkat BC, Bourne RB, Rorabeck CH, et al.: Thigh pain in cementless total hip arthroplasty. A comparison of two systems at 2 years' follow-up. Orthop Clin North Am. 1993; 24(4): 645–53. PubMed Abstract\n\nSkinner HB, Curlin FJ: Decreased pain with lower flexural rigidity of uncemented femoral prostheses. Orthopedics. 1990; 13(11): 1223–8. PubMed Abstract\n\nMoulton DL, Lindsey RW, Gugala Z: Dataset 1 in: Proximal Femur Size and Geometry in Cementless Total Hip Arthroplasty Patients. F1000Research. 2015. Data Source" }	[ { "id": "9274", "date": "01 Jul 2015", "name": "Loren Latta", "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 only suggestion I have is to either call the “cortical thickness” measurement exactly what it is, or explain its limitations correctly. It is a measure of the average thickness of the medial and lateral cortices from the radiographic view in the coronal plane. Or, it is an estimate of the average thickness under the assumptions that the bone is circular in cross-section at this location and the thickness of the cortex is uniform throughout the cross-section.I think #1 is the best choice because neither of the assumptions given in #2 are true, and it can be misleading to label that measurement an average of the total cortical thickness.", "responses": [] }, { "id": "10680", "date": "22 Oct 2015", "name": "Philip C Noble", "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 an interesting paper which addresses an important issue in joint replacement.My comments /questions are as follows:With several confounding variables it is not possible to draw robust generalizable conclusions without a much larger dataset (ie 100+ cases). This casts doubt on some of the comparisons that are deemed not statistically significant. The authors state that a difference of at least 20% must be present between group averages for the comparison to be valid (ie adequately powered). Such a large value makes many of the comparisons impossible to generalize to other populations and reports. The challenge with using clinical radiographs is guaranteeing the rotation of the femur projected on each AP exposure. As canal dimensions are sensitive to rotation, how was this achieved in this study? Gender, age, and to some extent, race, have been widely documented as fundamentally affecting the shape of the femur. Therefore these variables must be treated separately as nominal variables and cannot be simply assumed to be “pseudo-continuous” in the analysis. This is particularly challenging due to the differential interactions of age and gender that occur because of the effect of menopause on female cortical morphology. Please describe how this challenge was handled in the statistical analysis which seems to lump all cases together.", "responses": [] } ]	1	https://f1000research.com/articles/4-161
https://f1000research.com/articles/3-135/v1	26 Jun 14	{ "type": "Case Report", "title": "Case Report: Neuropathic pain in a patient with congenital insensitivity to pain", "authors": [ "Daniel W. Wheeler", "Michael C.H. Lee", "E. Katherine Harrison", "David K. Menon", "C. Geoffrey Woods", "Michael C.H. Lee", "E. Katherine Harrison", "David K. Menon", "C. Geoffrey Woods" ], "abstract": "We report a unique case of a woman with Channelopathy-associated Insensitivity to Pain (CIP) Syndrome, who developed features of neuropathic pain after sustaining pelvic fractures and an epidural hematoma that impinged on the right fifth lumbar (L5) nerve root. Her pelvic injuries were sustained during painless labor, which culminated in a Cesarean section. She had been diagnosed with CIP as child, which was later confirmed when she was found to have a null mutation of the SCN9a gene that encodes the voltage-gated sodium channel Nav1.7. She now complains of troubling continuous buzzing in both legs and a vice-like squeezing in the pelvis on walking. Quantitative sensory testing showed that sensory thresholds to mechanical stimulation of the dorsum of both feet had increased more than 10-fold on both sides compared with tests performed before her pregnancy. These findings fulfill the diagnostic criteria for neuropathic pain. Notably, she only experiences the negative symptoms (such as numbness and tingling) and she has not reported sharp, burning or electric shock sensations, although the value of verbal descriptors is somewhat limited in a person who has never felt pain before. However, her case strongly suggests that at least some of the symptoms of neuropathic pain can persist despite the absence of the Nav1.7 channel. Pain is a subjective experience and this case sheds light on the transmission of neuropathic pain in humans that cannot be learned from knockout mice.", "keywords": [ "There has been an explosion of interest in Nav1.7 as a potential therapeutic target for novel analgesics", "as mutations in SCN9A are associated with profoundly altered pain thresholds1. Perhaps the greatest level of interest has been reserved for those very rare individuals with autosomal recessive mutations that truncate the protein Nav1.7 resulting in a complete lack of expression of the ion channel. The result is Channelopathy-associated Insensitivity to Pain (CIP) Syndrome: a complete absence of pain sensation", "while all other sensory modalities apart from the sense of smell remain intact. Here we describe the experiences of a Caucasian 37-year-old patient with CIP whose older sister", "but neither of her parents or other family members", "is also affected. Other than a variety of injuries to the cornea and tongue", "burns and relatively minor fractures sustained during childhood and now ascribed to CIP", "there was no other medical history of note. Nonetheless", "after childbirth she developed a symptom that she now readily describes as pain", "and which has neuropathic features. We believe that this case report provides insights into the mechanisms of neuropathic pain", "dissecting “positive” from “negative” symptomatology", "and shows that it is possible to experience neuropathic pain in the absence of prior experience of acute pain." ], "content": "Case\n\nThere has been an explosion of interest in Nav1.7 as a potential therapeutic target for novel analgesics, as mutations in SCN9A are associated with profoundly altered pain thresholds1. Perhaps the greatest level of interest has been reserved for those very rare individuals with autosomal recessive mutations that truncate the protein Nav1.7 resulting in a complete lack of expression of the ion channel. The result is Channelopathy-associated Insensitivity to Pain (CIP) Syndrome: a complete absence of pain sensation, while all other sensory modalities apart from the sense of smell remain intact. Here we describe the experiences of a Caucasian 37-year-old patient with CIP whose older sister, but neither of her parents or other family members, is also affected. Other than a variety of injuries to the cornea and tongue, burns and relatively minor fractures sustained during childhood and now ascribed to CIP, there was no other medical history of note. Nonetheless, after childbirth she developed a symptom that she now readily describes as pain, and which has neuropathic features. We believe that this case report provides insights into the mechanisms of neuropathic pain, dissecting “positive” from “negative” symptomatology, and shows that it is possible to experience neuropathic pain in the absence of prior experience of acute pain.\n\nOur patient had been recognized as having CIP aged 7, diagnosed at the same time as her older sister. This diagnosis was confirmed 18 years later by finding bi-allelic heterozygous null mutations of SCN9A in exon 29 (c.4975T>A p.K1659X) and exon 22 (c.3699-3709delATGGATAGCAT p.I1235LfsX2). The SCN9A gene on chromosome 2q24.3 encodes the alpha-subunit of the Nav1.7 voltage-gated sodium channel, which is expressed at high levels in small-diameter peripheral nociceptive neurons2.\n\nThe patient sustained painless pelvic fractures, presumably during labor, which were not recognized for two months. By then, examination revealed significant weakness in both legs, worse on the right, with absence of both ankle reflexes. We subsequently compared the results of formal quantitative sensory testing three months post-injury to those obtained four years pre-injury. Sensory thresholds to heat and cold in the foot dorsum were broadly similar on both sides (Table 1), and should be interpreted in the context of someone who has never felt pain. However, thresholds to mechanical (von Frey) stimulation of the dorsum of both feet were increased more than 10-fold bilaterally. Imaging studies revealed multiple fractures of both sacral wings and of the superior and inferior pubic rami bilaterally. Furthermore, there was an extensive hematoma extending into the left iliopsoas, right obturator externus and spinal canal, causing occlusion of the thecal sac at the level of the fifth lumbar (L5) and first sacral (S1) intervertebral space (Figure 1). The fractures were attributed to transient osteoporosis of pregnancy, and their severity to her continued walking in the face of CIP. However, shortly after the fractures were diagnosed, bone densitometry studies and all serum bone profile results were found to be normal.\n\nThe X-ray shows multiple fractures of the superior and inferior pubic rami (left) and an axial magnetic resonance image (right) plus magnification showing a hematoma impinging upon the right L5 nerve root at the exit foramen (arrow).\n\nTwo months later, four months after delivery, she reported troubling continuous buzzing in both legs and a vice-like squeezing in the pelvis when she walked: symptoms that are consistent with neuropathic pain3. These two symptoms did not respond to the anti-neuropathic drug gabapentin, and persist two years after the delivery. Further treatment has focused on physiotherapy and conservative measures such as pacing and activity management.\n\nNeuropathic pain arises as a direct consequence of a lesion or disease affecting the somatosensory system, and is characterized according to four criteria: pain distribution; the link between distribution and history; confirmatory tests of neurologic status demonstrating sensory signs confined to the territory of the lesioned nerve, and further confirmatory diagnostic tests to identify the lesion or disease entity underlying the neuropathic pain3. The history, examination and investigations that we have described fulfill these criteria. We therefore believe that this patient has definite neuropathic pain, although it is manifested only by ‘negative’ symptoms, such as numbness and tingling, with an absence of the ‘positive’ symptoms such as stabbing or burning, but is nonetheless becoming increasingly debilitating.\n\nThe Nav1.7 channel plays a crucial role in pain transmission. However this case shows that neuropathic pain can be initiated and maintained in its absence in humans, as well as in knockout mice4, although we cannot rule out that Nav1.7 may mediate sharp, burning or electric shock sensations. Our data provide a further rational basis for seeking specific molecular substrates for neuropathic pain, some of which could act as mechanistic targets for new therapies for patients with symptoms of neuropathic pain.\n\n\nPatient consent\n\nWritten consent for the publication of this case report was obtained from the patient.", "appendix": "Author contributions\n\n\n\nAll authors conceived the paper. MCL undertook the post-injury quantitative sensory testing and CGW the SCN9a genotyping. DWW wrote the first draft, which was reviewed, revised and approved by all authors.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in funding this work.\n\n\nAcknowledgments\n\nWe are indebted to the late Dr David Bowsher, who kindly shared his pre-injury quantitative sensory testing data. We are grateful to our patient, who gave written consent for the publication of this report.\n\n\nReferences\n\nFischer TZ, Waxman SG: Familial pain syndromes from mutations of the NaV1.7 sodium channel. Ann N Y Acad Sci. 2010; 1184: 196–207. PubMed Abstract \| Publisher Full Text\n\nCox JJ, Reimann F, Nicholas AK, et al.: An SCN9A channelopathy causes congenital inability to experience pain. Nature. 2006; 444(7121): 894–8. PubMed Abstract \| Publisher Full Text\n\nTreede RD, Jensen TS, Campbell JN, et al.: Neuropathic pain: redefinition and a grading system for clinical and research purposes. Neurology. 2008; 70(18): 1630–5. PubMed Abstract \| Publisher Full Text\n\nNassar MA, Levato A, Stirling LC, et al.: Neuropathic pain develops normally in mice lacking both Na(v)1.7 and Na(v)1.8. Mol Pain. 2005; 1: 24. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "5741", "date": "11 Sep 2014", "name": "Frank van Eijs", "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 presented aspects of the reported case are indeed interesting, however...One cannot talk about pain if no pain is reported. Buzzing and a vice like squeezing of the pelvis is not the same as pain. The vice like squeezing of the pelvis may also have been due to the observed fractured pelvis. It is unclear if there indeed is subjective pain. If there is a report of pain there should also be an assessment of the visual analogue or numerical pain rating scale. Unfortunately this is not mentioned. Another clinical aspect is the missing of a DN4 score (douleur neuropathique 4 questionnaire). As I can see the score may not be more than 3 in which case the presence of neuropathic pain is unlikely.The MRI scan of the L5-S1 interspace is not clear. It lacks a saggital view. The cross-sectional view should be more clearly and preferentially showing more than 1 slice. Suggestions:Add the VAS or NRS of the pain intensity (probably 0?) Specify if pain scores are for the feet, legs or pelvis. If pain scores are not 0 then indeed there may possibly be neuropathic pain. In that case add the DN 4 score. If there is no pain, symptoms should be mentioned as neuropathic or neuropathy instead of neuropathic pain. In that case also the title needs to be changed (e.g. neuropathy in a patient with congenital insensitivity to pain). Add the saggital view of the lumbar MRI. Show more than 1 cross-sectional view of the MRI. Specify if views are T1, T2 or contrast dye enhanced.", "responses": [ { "c_id": "1422", "date": "19 Jun 2015", "name": "Daniel W. Wheeler", "role": "Author Response", "response": "The presented aspects of the reported case are indeed interesting; however, one cannot talk about pain if no pain is reported. Buzzing and a vice like squeezing of the pelvis is not the same as pain. The vice like squeezing of the pelvis may also have been due to the observed fractured pelvis. It is unclear if there indeed is subjective pain. If there is a report of pain there should also be an assessment of the visual analogue or numerical pain rating scale. Unfortunately this is not mentioned. Another clinical aspect is the missing of a DN4 score (douleur neuropathique 4 questionnaire). As I can see the score may not be more than 3 in which case the presence of neuropathic pain is unlikely.The MRI scan of the L5-S1 interspace is not clear. It lacks a sagittal view. The cross-sectional view should be more clearly and preferentially showing more than 1 slice. Suggestions:1. Add the VAS or NRS of the pain intensity (probably 0?) Specify if pain scores are for the feet, legs or pelvis. If pain scores are not 0 then indeed there may possibly be neuropathic pain. In that case add the DN 4 score.2. If there is no pain, symptoms should be mentioned as neuropathic or neuropathy instead of neuropathic pain. In that case also the title needs to be changed (e.g. neuropathy in a patient with congenital insensitivity to pain)ResponseThe International Association for the Study of Pain (IASP) defines pain as an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage. The IASP also acknowledges that application of the word ‘pain’ is learnt in early life (http://www.iasp-pain.org/Taxonomy?navItemNumber=576#Pain). It is unclear what patients have been diagnosed with congenital insensitivity to pain (CIP) understand by the word ‘pain’. However, there is psychometric and neuroimaging evidence they understand what the word ‘pain’ means, as they are clearly able to empathise with behavioural and verbal expressions of pain in normal individuals (Danziger et al., 2009).Our patient reports that her right hip and pelvis still “hurt” a great deal, using descriptors such as “tight” and “aching”, and that she “suffers” if she walks too far or doesn’t wear an orthotic heel raise. At rest, this “pain” resolves, but she is left with “tingling”, “buzzing” and “electric shocks”. The patient in our report labelled her sensations as painful and we contend that description alone is adequate to fulfil the IASP definition of pain. Furthermore she also describes headaches that respond to paracetamol, “the sting of a graze”, “the sharpness of an exposed gum”, and “back aches”, “period pains and stomach cramps” that arose after pregnancy.Our patient has also exhibited behavior consistent with a person in pain. She sought treatment in the local Pain Clinic (with DWW) for the sensations, which suggests that these “buzzing” and “vice-like” descriptors had strong aversive-motivational qualities. We believe that the simple verbal use of the word pain and behavioral response sufficed and have acknowledged and managed her report of pain as such. The visual analog or numeric rating scale (NRS) is undoubtedly useful in normal individuals, and our patient rates her current pain intensity in the right hip and leg as between 0 and 4 on a 10-point NRS. However, the scale should not be required to validate the report of pain by this patient.The DN4 score aims to increase the specificity of a diagnosis of neuropathic pain, with a score >4 providing 90% specificity for neuropathic pain. The sensitivity of the questionnaire is not well documented and the specificity of lower scores for neuropathic pain is unclear. The score in our patient was 5, but as the scale was validated in individuals with premorbid normal nociceptive physiology, we judge that the diagnosis of neuropathic pain stands independently of the DN4 score.We have added several additional paragraphs to the manuscript to address these points, which we hope you will find satisfactory. Add the sagittal view of the lumbar MRI. Show more than 1 cross-sectional view of the MRI. Specify if views are T1, T2 or contrast dye enhanced.ResponseThe MRI scans are T2-weighted and non-contrast enhanced. An additional cross-sectional view and a sagittal view are included below (included as Figure 1c and 1d in the revised manuscript). The arrows indicate the presence of hematoma material in the psoas and adjacent to the cauda equina." } ] }, { "id": "6088", "date": "17 Sep 2014", "name": "Juan D. Ramirez", "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 an exciting report by Wheeler et al., assessing the meaning of pain after a traumatic labour in a subject who has never experienced anything similar. This case is paramount for highlighting the importance of a detailed study of patients with inherited channelopathies.Neuropathic pain has been established by IASP as pain arising as a consequence of damage to the somatosensory system. In this case the patient describes positive sensory phenomena in the form of buzzing and squeezing which is troubling and is accompanied by numbness which is often regarded as negative phenomena (positive and negative sensory abnormalities often co-exist in neuropathic pain states) so I would recommend changing the terms ‘sensory loss’ for both ‘sensory gain and loss’.One issue is whether what the subject describes is more akin to paraesthesia rather than pain as understood by healthy subjects in whom NaV 1.7 is functional. The IASP definition of pain is actually fairly broad including the term ‘unpleasant sensory experience’. In the current case this sensory disturbance is an anatomically plausible distribution with evidence of a lesion to the somatosensory system.The evidence comes from MRI that shows nerve root compression as a consequence of a haematoma. It would be helpful to have more sequences in order to demonstrate bilateral involvement and neurophysiology to define the extent of the nerve damage (sensory and motor).The thermal and mechanical sensory testing is thorough and I would only recommend the authors to add the baseline temperature at which they started the assessment e.g. “32°C” added as part of the figure legend.Finally, I would vouch for the use of the neuropathic pain symptom inventory for assessing in more detail the symptomatology of the subject (Bouhassira et al., 2004).In general I believe this a relevant case that illustrates the need to carefully delineate new sensory symptoms in patients with congenital inability to experience pain and emphasises the distinct nature of neuropathic versus nociceptive pain states.", "responses": [ { "c_id": "1421", "date": "19 Jun 2015", "name": "Daniel W. Wheeler", "role": "Author Response", "response": "This is an exciting report by Wheeler et al., assessing the meaning of pain after a traumatic labour in a subject who has never experienced anything similar. This case is paramount for highlighting the importance of a detailed study of patients with inherited channelopathies.Thank you for your positive comments about our manuscript. Neuropathic pain has been established by IASP as pain arising as a consequence of damage to the somatosensory system. In this case the patient describes positive sensory phenomena in the form of buzzing and squeezing which is troubling and is accompanied by numbness which is often regarded as negative phenomena (positive and negative sensory abnormalities often co-exist in neuropathic pain states) so I would recommend changing the terms ‘sensory loss’ for both ‘sensory gain and loss’.We agree, and the text has been amended as suggested. One issue is whether what the subject describes is more akin to paraesthesia rather than pain as understood by healthy subjects in whom NaV 1.7 is functional. The IASP definition of pain is actually fairly broad including the term ‘unpleasant sensory experience’. In the current case this sensory disturbance is an anatomically plausible distribution with evidence of a lesion to the somatosensory system.The evidence comes from MRI that shows nerve root compression as a consequence of a haematoma. It would be helpful to have more sequences in order to demonstrate bilateral involvement and neurophysiology to define the extent of the nerve damage (sensory and motor).The patient was seen in the clinical setting. Routine neurological examination revealed sensory loss to touch already suggesting neuropathy. EMG and ENG were not performed, as neither would have altered the clinical management in this case.Please see our response to Dr van Eijs for our comment on the MRI scan. The thermal and mechanical sensory testing is thorough and I would only recommend the authors to add the baseline temperature at which they started the assessment e.g. “32°C” added as part of the figure legend.The table caption has been amended to indicate that the baseline temperature was 32°C. Finally, I would vouch for the use of the neuropathic pain symptom inventory for assessing in more detail the symptomatology of the subject (Bouhassira et al., 2004).Please see our response to Dr van Eijs regarding the value of pain questionnaires in patients with CIP." } ] } ]	1	https://f1000research.com/articles/3-135
https://f1000research.com/articles/4-158/v1	18 Jun 15	{ "type": "Research Article", "title": "CPIRD: A successful Thai programme to produce clinically competent medical graduates", "authors": [ "Yi Yanhua", "Virasakdi Chongsuvivatwong", "Hutcha Sriplung", "Chulalak Rueanarong", "Hutcha Sriplung", "Chulalak Rueanarong" ], "abstract": "The programme titled “Collaborative Project to Increase Production of Rural Doctors” (CPIRD) is a rural medical education project launched in 1994 in Thailand. This study aimed to compare the academic performances in medical study over five years and the pass rates in national medical license examinations (MLE) between students enrolled in CPIRD and two other tracks.Grade point average (GPA) over five years and results of MLEs for four cohorts of students enrolled from 2003 to 2006 in Prince of Songkla University were collected from the registration department. A longitudinal analysis was used to compare the GPA over time for medical students enrolled in CPIRD and those from the national and direct regional tracks through generalized estimating equation (GEE) models. The MLE pass rates were compared using chi-square and fisher's exact tests as appropriate.Female students dominated the CPIRD group. GPAs in the first three years in the CPIRD group were significantly lower than those of the other two groups, this disparity narrowed in the fourth and fifth years. For step one of the MLE (basic sciences), cohorts 2003 and 2006 of the CPIRD group had a significantly lower pass rate than the other two groups but there was no significant difference in cohort 2004 and cohort 2005. The CPIRD step two and three MLE pass rates were not significantly different from the national track in all cohorts and lower than the direct track only for step two in cohort 2003 and step three in cohort 2006. The step three pass rate of the CPIRD group in cohort 2004 was significantly higher than the other two tracks.Despite weaker competency in basic science, the CPIRD was successful in forming clinical competency.", "keywords": [ "rural", "grade point average (GPA)", "license examination", "clinical competency", "CPIRD" ], "content": "Introduction\n\nIn 2010, the World Health Organization (WHO) recommended sixteen interventions to improve health force retention in rural areas. These included education strategies to recruit students of rural origin, locating medical schools outside major cities, bringing students to rural communities and matching curricula with rural health needs1.\n\nThailand is well known for its emphasis on rural health development2. Since 1972, all medical graduates must spend at least three years of compulsory service in rural areas. In the same year, the medical school of Prince of Songkla University (PSU) was established in southern Thailand, the most remote part of the country, in order to strengthen the local capacity in medical services. From the initial establishment period, PSU had two kinds of enrollment methods. The first is a national entrance examination (hereafter abbreviated to national track), which allows students from all over Thailand to sit the examination for a chance to study3. For geographic and socio-cultural reasons, this medical school in the south has not been a popular choice for candidates from high schools in other regions of the country. The local medical school compensates for this by using a second method of recruitment called the direct admission programme (hereafter abbreviated to direct track). This method recruits students from the southern regional provinces exclusively based on an institution-specific examinations4, which take place a few months before the announcement of the national track examinations. This earlier announcement makes the programme popular to local candidates because they get admitted earlier and naturally have no difficulties acclimatizing to the different culture in the south of the country4. For decades, direct track students are known to have a better average academic performance than the national track students4–6.\n\nTo further ensure adequate supply of medical doctors to the rural region, especially the potential insurgent areas of southern Thailand, a third track was introduced in 2003. Under the “Collaborative Project to Increase Production of Rural Doctors” (CPIRD), rural students from the region were recruited with a longer period (six years) of obligatory service in specific areas where there were a shortage of doctors. Later, the “One District One Doctor (ODOD)” programme was brought in as the fourth track4. ODOD students were not included in this analysis as the programme was considered too new.\n\nStudents of all tracks complete the first three years of medical study together. The national track and direct track students take their following three years of clinical study in university hospitals and CPIRD students in accredited regional and provincial hospitals of the Ministry of Public Health (MOPH)4. Grade Point Average (GPA) was used to assess the student’s performance in all years and was based on the same standard set assessed by the regional medical university.\n\nSince 2002, the Thai Medical Council has required all medical students who matriculated from the year of 2003 to pass all three parts of the Medical Licensing Examination (MLE) before getting their medical licenses7. The three steps of the MLE are taken at the end of the third, the fifth and sixth year, respectively. The first step of the MLE focuses on basic science knowledge, the second step on clinical science knowledge, and the third step on both knowledge and clinical skills evaluation. This is to standardize the basic competencies of graduate physicians and to assure health consumers have a standard health care service8.\n\nWhile the idea of recruiting medical students from rural areas and training them at hospitals close to the rural population is highly advocated based on the findings that it had positive implication on rural retention9–12, but competency of graduates from such programmes have rarely been investigated.\n\nThe main objective of this study was therefore to compare the academic performance of the students recruited from different tracks as reflected by their GPA over five years and the pass rate at each step of the MLE.\n\n\nMethods\n\nSouthern Thailand, where this study was conducted, is a region of the country with the highest levels of heterogeneity of the population and continuous ethnic unrest13.\n\nA retrospective cohort study based on the records of the performance of all medical students enrolled in 2003 to 2006 was used.\n\nThe data was retrieved from the student registry of Faculty of Medicine, PSU. All personal identification was encrypted. The study protocol was approved by the Ethics Committee of the Faculty of Medicine, PSU (Permit No: 56-317-18-5).\n\nAll data analyses were performed using R version 3.1.3 (http://www.r-project.org) and Epicalc package 2.15.1.0 (http://CRAN.R-project.org/package=epicalc). A longitudinal data analysis was used to compare the GPA over five years for medical students enrolled in three different programmes through generalized estimating equation (GEE) models. The results for the pass rates in MLE were analyzed using chi-square and fisher's exact test as appropriate. Statistical significance was set at 5%.\n\n\nResults\n\nTable 1 compares baseline characteristics of the students from the three tracks. Female students had a larger percentage in the CPIRD group compared with direct track and national track students. The number of students in the CPIRD increased from 19 in 2003 to 72 in 2006, whereas students from other two programmes remained stable.\n\nNumbers in bracket are percent unless otherwise stated. * p-value <0.001, p-value <0.01, * p-value <0.05\n\nCPIRD: Collaborative Project to Increase Production of Rural Doctors\n\nFigure 1 shows how the GPA changed over five years of time. In the first three years, the mean GPA of students from three enrollment programmes was significantly different. Students from the direct track performed best. Followed by national track students. Students from CPIRD had lower GPAs than the others. However, the GPAs of the last two years from the three groups tended to converge. Table 2 summarizes results of the GEE with ‘year’ as a continuous variable and ‘track’ as a categorical variable. Their interaction was statistically significant; therefore the interaction terms were included in the final model. Non-significant positive coefficient for the main effect ‘year’ indicates that the tendency of increment of average performance scores of the reference group (CPIRD) was not significant. The other two main effects ‘direct track’ and ‘national track’ were both significantly higher than that of CPIRD in the reference year (first year). Both interaction coefficients are negative indicating that over the years, the difference between the two tracks and CPIRD was reduced significantly.\n\n† Coef: Coefficient; SE: Standard error\n\nTable 3 shows the association between enrollment programmes and results of the MLE for the four cohorts of students enrolled from 2003 to 2006. In step one of the MLE, CPIRD students were weaker than students in the other tracks for cohort 2003 and cohort 2006. In the remaining two MLE steps, CPIRD students’ pass rate was not statistically different from that of the national track students. Direct track students had a higher pass rate only in step two of the MLE in cohort 2003 and step three of the MLE in cohort 2006 compared with CPIRD students. In fact, CPIRD had the highest pass rate in step three of the MLE in cohort 2004.\n\nN: the number of students … indicates referent group\n\n\nDiscussion\n\nCPIRD students had a lower GPA on average in pre-clinical years and lower pass rates of the MLE in basic science parts than students of the other two tracks. Their GPA tended to catch up with their peers in clinical years and the pass rate of the MLE in the clinical parts were more or less comparable with their peers.\n\nThe selection process of medical students in Thailand could explain the fact that CPIRD students had the lowest GPA in the first three years. Direct track students were those students in southern Thailand with good academic records who sat for the entrance examination at Prince of Songkla University. National track students selected Prince of Songkla University as an alternative choice because of its geographic distance from Bangkok. CPIRD students were those unable to get through by direct track examination and finally selected by the CPIRD route. As a result, direct track students had the highest academic performance in high school, followed by national track students, while CPIRD students were weakest4. The first three years was the pre-medical and pre-clinical study. It has been shown in other medical schools that the pre-clinical stage including second and third year, had a high correlation with the first year pre-medical stage3. The lower performance in these first three years for CPIRD students reflected their weaker background and performance in science and thus these students need support to reduce the dropout rate14,15.\n\nA previous study suggested that CPIRD students had more opportunities to practice in regional hospitals and thus displayed more capable clinical skills in the fourth and fifth year3. In addition, the successful application of problem-based learning (PBL) in clinical study reduced the difference in academic performance and fostered a self-motivated study atmosphere among medical students16.\n\nThe findings that the CPIRD students could perform as good as those normal tracks of students in step two and step three MLE has important implications. Good clinical education does not need to be confined to a conventional teaching hospital. Decentralized medical education requires enhancement of existing hospitals. The byproducts of this strengthening include increasing service capacity and quality of health services to local populations, which reduces the inequality problems due to geographical barriers. Other studies also reported that Thai CPIRD doctors were more likely to stay longer in rural areas than their peers17,18. Most low and middle-income countries (LMICs) have a serious rural–urban disparity of health service and the clinical education is mostly based in university teaching hospitals in large cities19. The experience from the Thai CPIRD should therefore be carefully reviewed for potential adaptation to other low LMICs.\n\nThis was a retrospective study; other factors influencing academic performance could not be determined and taken into account. Examination questions and behavioral performance of the students may differ with time and place. However, the MLE were rigorously standardized national examinations and comparisons across student groups were made mainly within the same cohort. Thus, this limitation has been minimized.\n\n\nConclusion\n\nThe CPIRD was successful in creating clinically competent doctors despite lower GPAs in the pre-clinical year.\n\n\nData availability\n\nF1000Research: Dataset 1. Medical students academic performance and MLE results for four cohorts, 10.5256/f1000research.6638.d4993020", "appendix": "Author contributions\n\n\n\nYY was principal investigator of the study, conceptualized the research, collected the data, performed data analysis, and drafted the manuscript. VC, HS and CR conceived the study, assisted in development of data analysis, manuscript writing, and provided supervision and suggestions. All authors have seen and agreed to the content of the final manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nFinancial support for data collection was obtained from the Epidemiology Unit at Prince of Songkla University under the support of the China Medical Board.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThis study is a part of the first author’s thesis in partial fulfillment of the requirements for a Ph.D. in the Epidemiology Unit at Prince of Songkla University, Thailand.\n\n\nSupplementary material\n\nData analysis using R and Epicalc.\n\nData analysis and output for Table 1, Figure 1, Table 2 and Table 3.\n\nClick here to access the data.\n\n\nReferences\n\nDolea C, Stormont L, Shaw D, et al.: Increasing access to health workers in remote and rural areas through improved retention. First expert meeting to develop evidence-based recommendations to increase access to health workers in remote and rural areas through improved retention. World Health Organization Geneva, 2009. Reference Source\n\nWibulpolprasert S, Pengpaibon P: Integrated strategies to tackle the inequitable distribution of doctors in Thailand: four decades of experience. Hum Resour Health. 2003; 1(1): 12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPinyopornpanish M, Wongsawasdi L, Panjaisee N, et al.: Comparison of the academic achievement of Chiang Mai graduate medical students which selected by quota, entrance and rural project. Chiang Mai Med Bull. 2004; 43(2): 77–86. Reference Source\n\nPutthasri W, Suphanchaimat R, Topothai T, et al.: Thailand special recruitment track of medical students: a series of annual cross-sectional surveys on the new graduates between 2010 and 2012. Hum Resour Health. 2013; 11: 47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSuphanchaimat R, Boonthai N, Tangthasana S, et al.: A survey of manual vacuum aspiration’s experiences among the new medical graduates in Thailand. Reprod Health. 2013; 10(1): 49. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEkpanyaskul C, Sithisarankul P, Wattanasirichaigoon S: Overweight/Obesity and related factors among Thai medical students. Asia Pac J Public Health. 2013; 25(2): 170–80. PubMed Abstract \| Publisher Full Text\n\nWanvarie S, Sathapatayavongs B: Logistic regression analysis to predict Medical Licensing Examination of Thailand (MLET) step1 success or failure. Ann Acad Med Singapore. 2007; 36(9): 770–3. PubMed Abstract\n\nTangjitgamol S, Tanvanich S, Pongpatiroj A, et al.: Factors related to the achievement of the National License Examination Part 1 of medical students in Faculty of Medicine Vajira Hospital, Navamindradhiraj University. South-East Asian Journal of Medical Education. 2013; 7(1): 51. Reference Source\n\nRabinowitz HK: Recruitment, retention, and follow-up of graduates of a program to increase the number of family physicians in rural and underserved areas. N Engl J Med. 1993; 328(13): 934–9. PubMed Abstract \| Publisher Full Text\n\nBrooks RG, Walsh M, Mardon RE, et al.: The roles of nature and nurture in the recruitment and retention of primary care physicians in rural areas: a review of the literature. Acad Med. 2002; 77(8): 790–8. PubMed Abstract \| Publisher Full Text\n\nCurran V, Rourke J: The role of medical education in the recruitment and retention of rural physicians. Med Teach. 2004; 26(3): 265–72. PubMed Abstract \| Publisher Full Text\n\nRabinowitz HK, Diamond JJ, Hojat M, et al.: Rural health research: demographic, educational and economic factors related to recruitment and retention of physicians in rural Pennsylvania. J Rural Health. 1999; 15(2): 212–8. PubMed Abstract \| Publisher Full Text\n\nThe Deep South Relief and Reconciliation (DSRR) Foundation and the Rugiagli Initiative (tRI). Healing Under Fire The Case of Southern Thailand. Bangkok, 2014. Reference Source\n\nSuwanthawee T: Problems and impacts on medical students and methods of prevention at the Faculty of Medicine of Ramathibodi Hospital, Mahidol University. Proceedings of the Thai Medical Education Conference, 1995.\n\nLakakul A: Backgrounds of medical students effect on academic achievement. Proceedings of the Thai Medical Education Conference, 1995.\n\nUdomratn P: Psychiatric Education In Thailand: A Focus On Undergraduate Curriculum. 2002. Reference Source\n\nPagaiya N, Kongkam L, Sriratana S: Rural retention of doctors graduating from the rural medical education project to increase rural doctors in Thailand: a cohort study. Hum Resour Health. 2015; 13(1): 10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKiernan ML: Reducing Inequities in Doctor Distribution: Literature Review, Thai Case Study and Policy Recommendations. Reference Source\n\nHarris B, Goudge J, Ataguba JE, et al.: Inequities in access to health care in South Africa. J Public Health Policy. 2011; 32(Suppl 1): S102–S23. PubMed Abstract \| Publisher Full Text\n\nYi Y, Chongsuvivatwong V, Sriplung H, et al.: Dataset 1 in: CPIRD: A successful Thai programme to produce clinical competent medical graduates. F1000Research. 2015. Data Source" }	[ { "id": "9104", "date": "01 Jul 2015", "name": "Supasit Pannarunothai", "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 paper is concise and clear, providing evidence that medical students from Collaborative Project to Increase Production of Rural Doctors (CPIRD) tract finally watched up with medical students from local direct admission and national admission.There are some points to improve the quality of this paper:The text explaining the CPIRD programme in paragraph 3 of Introduction is not accurate. The obligatory service period of service is still kept constant at three years not six years as written. The One District One Doctor (ODOD) programme coerces 12 years of obligatory service. Methods section should explain clearer on selection of students especially about drop-out between years. It's unlikely that there were no dropouts during the 6-year MD programme. How the study handle this issue? What bias is likely to occur? Please discuss. The GPA presentation in figure 1 shows that students of the national tract best performed since the first year, but table 2 and text in result section explained that students of direct tract best performed. Please check. Table 3 shows very high fail rate of step 3 exam as compared to others. There is no explanation how the systems interacted with this high fail rate.", "responses": [] }, { "id": "9408", "date": "09 Jul 2015", "name": "Weerasak Putthasri", "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 manuscript generates concrete evidence to support the WHO 2010 recommendation focusing on the graduates' competencies. To disseminate this academic proof is important and useful for both national and global audiences. However, there are some minor points could be improved for paper quality, as followsTable 1, please check 'Percentage' of the national track column. The total percent must be 100.0%. As for the MLE pass rate, the 'Fail' was students who did not pass the exam on 'only the first attempt of that test', correct? Please describe more. Discussion:Authors introductory mentioned the WHO global policy recommendations on increasing access to health workers in remote and rural areas through improved retention and want to prove the competencies or performance of that implementation. In order to link this evidence to policy recommendation, authors may consider to add some sentences using this key finding to support the global recommendation implementation, esp. Students from rural backgrounds and Clinical rotations in rural areas during studies which are the key characteristic of CPIRD. Further more, authors could also consider to mention the WHO recommendations on transforming and scaling up health professionals' education and training, 2013 which has recommended for targeted admissions policies to increase the socio-economic, ethnic and geographical diversity of students and the expansion of faculty through recruitment of community-based clinicians and health workers as educators, as well.", "responses": [] }, { "id": "9407", "date": "20 Jul 2015", "name": "Shama Virani", "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\nCPIRD: A successful Thai programme to produce clinically competent medical graduates This research describes differences in GPA and MLE scores of medical students that enter the program through three different mechanisms. These mechanisms seem to have an impact on initial scores, but become comparable at the end of the program. Comments: This paper is integral in future implementation of intervention programs designed to retain health forces in rural areas. Through this analyses, the authors have shown that clinical experience is as valuable as the curriculum. Minor revisions are suggested below. Page 1 Is there a rationale as to why step one MLE scores of the CPIRD group was significantly lower for only cohorts 2003 and 2006 but not 2004 and 2005? Figure 1This figure shows that national track students have the best GPAs overall. It seems to contradict the statements in the introduction and discussion that mention direct track students do better than national track. It would be useful to address this discrepancy in the discussion.Page 3:Is there information on how many students in CPIRD first try to attend medical school through the national or direct tracks? For the study site, please include the university where the study took placePage 4:Is there any particular reason why the CPIRD group contains more females? Did this change over time? If so, it might be useful to include gender as a covariate in the model as this might influence the trajectory of the scores. Table 3: Please indicate the test used to obtain p-values in the table legend. It might also be useful to bold the “Cohort 2003”..etc to make the table a bit easier to readLevel of Interest: An article of importance in its field", "responses": [] } ]	1	https://f1000research.com/articles/4-158
https://f1000research.com/articles/4-156/v1	18 Jun 15	{ "type": "Research Article", "title": "1-Octen-3-ol – the attractant that repels", "authors": [ "Pingxi Xu", "Fen Zhu", "Garrison K. Buss", "Walter S. Leal", "Pingxi Xu", "Fen Zhu", "Garrison K. Buss" ], "abstract": "Since the discovery in the early 1980s that 1-octen-3-ol, isolated from oxen breath, attracts tsetse fly, there has been growing interest in exploring the use of this semiochemical as a possible generic lure for trapping host-seeking mosquitoes. Intriguingly, traps baited with 1-octen-3-ol captured significantly more females of the malaria mosquito, Anopheles gambiae, and the yellow fever mosquito, Aedes aegypti, than control traps, but failed to attract the southern house mosquito, Culex quinquefasciatus. Additionally, it has been demonstrated that this attractant is detected with enantioselective odorant receptors (ORs) expressed only in maxillary palps. On the basis of indoor behavioral assays it has even been suggested that 1-octen-3-ol might be a repellent to the southern house mosquito. Our approach was two-prong, i.e., to isolate 1-octen-3-ol-sensitive ORs expressed in maxillary palps and antennae of southern house female mosquito, and test the hypothesis that this semiochemical is a repellent. An OR with high transcript levels in maxillary palps, CquiOR118b, showed remarkable selectivity towards (R)-1-octen-3-ol, whereas an OR expressed in antennae, CquiOR114b, showed higher preference for (S)-1-octen-3-ol than its antipode. Repellency by a surface landing and feeding assay showed that not only racemic, but enantiopure (R)- and (S)-1-octen-3-ol are repellents at 1% dose thus suggesting the occurrence of other (S)-1-octen-3-ol-sensitive OR(s). Female mosquitoes with ablated maxillary palps were repelled by 1-octen-3-ol, which implies that in addition to OR(s) in the maxillary palps, antennal OR(s) are essential for repellency activity.", "keywords": [ "Culex quinquefasciatus", "odorant receptors", "chiral discrimination", "antennae", "maxillary palps", "CquiOR118b", "CquiOR114b", "repellency assay" ], "content": "Introduction\n\n1-Octen-3-ol (Figure 1) is a natural product derived from linoleic acid, which was first isolated from the matsutake pine mushroom1 and thereafter from plants and other fungi. It is approved by US Food and Drug Administration (ASP 1154, Regnum 172.515) as a food additive and also considered a wine fault – an unpleasant characteristic of wine. Since it was discovered as an emanation from oxen breath that attracts tsetse fly2, there has been growing interest in using 1-octen-3-ol as an insect attractant. Indeed, it was demonstrated earlier on that 1-octen-3-ol synergizes with CO2 and thus increase mosquito trapping efficacy3. Intriguingly, field experiments demonstrated that the effect of 1-octen-3-ol on mosquito captures is species specific4. Of notice, 1-octen-3-ol seem to have little or no effect on trapping of the southern house mosquito, Culex quinquefasciatus, although being undoubtedly an attractant (kairomone) for Anopheles and Aedes mosquitoes4.\n\n(S) and (R)-1-octen-3-ol, (S) and (R)-1-octyn-3-ol, (S) and (R)-3-octanol.\n\nClub-shaped olfactory basiconic sensilla (peg sensilla) in the maxillary palps of the southern house mosquito harbor three types of olfactory receptor neurons (ORNs) identifiable by their spike amplitudes5. The second largest neurons, ORN-B, responded to 1-octen-3-ol with very high sensitivity. Cell B also showed a remarkable selectivity between the two enantiomers of 1-octen-3-ol, with the (R)-(-)-isomer eliciting robust responses at 10 ng dose (256.6 ± 12 spikes/s), whereas the (S)-(+)-antipode eliciting only 115.5 ± 23 spikes/s even when challenged with 100x higher does, i.e., 1 µg5. It was also demonstrated that neuron-B in the maxillary palps of the malaria mosquito, Anopheles gambiae, responds to 1-octen-3-ol6, and chiral discrimination was also observed with electrophysiological recordings from the neuron B in the maxillary palps of the yellow fever mosquito, Aedes aegypti7,8. Additionally, odorant receptors housed in the maxillary palps of the malaria mosquito6 and yellow fever mosquito9, AgamOR8 and AaegOR8, respectively, showed significant preference for the (R)-enantiomer when co-expressed in Xenopus oocytes along with the obligatory co-receptor Orco.\n\nIn-door behavioral studies demonstrated that at two doses (R)-(-)1-octen-3-ol caused an increase in activation for Cx. quinquefasciatus, and at seven of the doses tested (R:S)-1-octen-3-ol mixture (84:16) caused significantly more mosquitoes to sustain their flight and reach the capture chambers in a two-choice, Y-tube olfactometer thus suggesting that the isomeric mixture has an excitatory effect7. Additionally, they observed that at the highest concentration, mosquitoes that reached the capture chambers moved towards the control chamber rather than the arm containing (R)-(-)-1-octen-3-ol per se or in mixtures, i.e., a reduced attraction response mediated by the (R)-enantiomer7. Since the ability of the olfactory system to detect the two enantiomers at this close ratio (approximately 5:1) was not observed in our electrophysiological recordings from peg sensilla5, we aimed at testing chiral discrimination at the receptor level. Using cDNA template from the maxillary palps, we cloned the Culex ortholog of AgamOR8 and AaegOR8, co-expressed it along with CquiOrco in Xenopus oocytes, and observed an ability to discriminate enantiomers that reflects our previous findings with single sensillum recordings. Additionally, we cloned a paralogous odorant receptor (OR) from antennae, which responded to both enantiomers of 1-octen-3-ol. These findings provide evidence that peripheral reception of 1-octen-3-ol is enantioselective at the maxillary palps, but random (racemic) at the antennae.\n\n\nMaterials and methods\n\nCulex quinquefasciatus mosquitoes used in this study were from a laboratory colony10, maintained for the last 5 years at 27 ± 1°C under a photoperiod of 12:12 h (light:dark). Our Davis colony was derived from mosquitoes collected in Merced, California, in the 1950s and maintained by Dr. Anthon Cornel in the Kearney Agricultural Center, University of California. Twenty pairs of antennae and maxillary palps of 9-day old gravid female adults were dissected on ice under a light microscope. Total RNA was extracted using RNeasy Micro Kit (QIAgen, Valencia, CA). Before synthesizing first-strand cDNA, RNA concentrations from antennae and maxillary palps extracts were adjusted (normalized). First-strand cDNA was synthesized with oligo (dT) primer (BioRad, Hercules, CA) and GoScript Reverse Transcriptase (Promega, Madison, WI) following the manufacturer’s protocol.\n\nFull-length sequences of CquiOR114b and CquiOR118b were amplified from female antennae and maxillary palps cDNAs, respectively. The In-Fusion cloning strategy was taken by using In-Fusion® HD Cloning Kit (Clontech® Laboratories, Mountain View, CA). Briefly, the PCR primers were designed with 16 overlapped nucleotides at 5’-end homologous to the linearized ends of the destination vector (pGEMHE), which was double digested by XbaI and XmaI. The primers for CquiOR114b were: forward, AGATCAATTCCCCGGGaccATGGCTACGAAGAAGGTTGCATTC; and two reverse primers, reverse-1: TCAAGCTTGCTCTAGATTACGATCCTTCATAAACCGCCTT and reverse–2: TCAAGCTTGCTCTAGATTACAACTCAAAGGAAACTCTGCTAACTCC. Low case “acc” stands for Kozak sequence.\n\nFor CquiOR118b two forward and one reverse primers were used; forward-1: AGATCAATTCCCCGGGATGAACGACCTGGTGCGGTTCGAG and forward-2: AGATCAATTCCCCGGGATGCATGTGGGCAACTCCAAGATTTCG; reverse, TCAAGCTTGCTCTAGATTATTTCTCGCTGGGATCATAAATAGTTTTCAGCAG. Underline denotes homologous sequence for In-Fusion reaction. PfuUltra II Fusion HS DNA Polymerase (Agilent Technologies, Santa Clara, CA) was used for PCR. PCR products were directly cloned into pGEMHE by using In-Fusion® HD Cloning Kit, following the manufacturer’s protocol. In brief, a mix of PCR product, In-Fusion HD Enzyme Premix, pGEMHE vector was incubated at 50°C for 15 min. One microliter of the reaction was added to Stellar™ competent cells for transformation. Plasmids were purified by plasmid mini prep columns SpinSmart (Denville Scientific, South Plainfield, NJ) and sequenced by Davis Sequencing Inc. (Davis, CA).\n\nSsoAdvanced™ SYBR® Green from BioRad was used for qPCR. The reactions were carried out in a BioRad C1000 thermal cycler with CFX96 detection module. The detection primers for CquiOr114b were: forward, TTAGCGGGA GAAAACATGGG; reverse, ACTGACTTTGGTACAC GTGG. For CquiOr118b, they were: forward, GTCGTTGCTTTTCCTGATGG; reverse, CACGGCATT CTCATATTTTACACT. The following primers were used for a reference gene, CquiOrco: forward, GCCGGATACGTTTTCTCCTTC; reverse, GCGCATAATTCCCTTCAGATG. The reaction system (total volume, 20 µl) included SsoAdvanced SYBR green mix (2x) 10 µl, cDNA 100 ng, paired primer mix 350 nM, and double distilled H2O. The qPCR program was 95°C for 30 s, 95°C for 5 s, 62°C for 30 s, 72°C for 30 s, and 40 cycles. The melt curves were made from 65°C to 95°C with increment of 5°C, 5 s.\n\n\nElectrophysiology\n\nThe two-electrode voltage-clamp (TEVC) technique was used to measure odorant-induced currents in Xenopus oocytes at a holding potential of −80 to −70 mV. Oocytes on stage V or VI were purchased from Ecocyte Bioscience (Austin, TX). Signals were amplified with an OC-725C amplifier (Warner Instruments), low-pass–filtered at 50 Hz, and digitized at 1 kHz. Data acquisition and analysis were conducted with Digidata 1440A and software pCLAMP10 (Molecular Devices). Traces were collected from same batches and same age of oocytes to make data consistent. Data were analyzed with GraphPad Prism 6 (La Jolla, CA). The following chiral compounds were gifts from Bedoukian Research Inc.: (R)-(-)-1-octen-3-ol (CAS# 3687-48-7), (S)-(+)-1-octen-3-ol (CAS# 24587-53-9); (R)-(-)-1-octyn-3-ol (CAS#32556-70-0), (S)-(+)-1-octyn-3-ol (CAS#322556-71-1); (R)-(-)-3-octanol (CAS#70492-66-9). Racemic 1-octen-3-ol (CAS # 3391-86-4) and (S)-(+)-3-octanol (CAS# 22658-92-0) were acquired from Fluka and Aldrich, respectively (Sigma-Aldrich, Milwaukee, WI).\n\nRepellence was measured by using a previously described surface-landing and feeding assay10. In short, 3–5 days-old female mosquitoes (30–40 mosquitoes per assay) were placed on a two-choice arena designed to attract host-seeking mosquitoes. For physical stimuli, water at 37°C was circulating inside of Dudley tubes, which were painted black in the internal surfaces. Chemical stimuli were provided by stream of CO2 at 50 ml/min and dental cotton rolls impregnated with defibrinated sheep blood, which were placed on the top of the Dudley tubes. For each test, filter paper rings freshly treated at the outer perimeter with 200 μl of hexane only or 200 μl of a tested compound in hexane were placed to surround each Dudley tube. Mosquito activity was observed and recorded for 5 min with a camcorder equipped with Super NightShot Plus infrared system (Sony Digital Hanycam, DCR-DVD 810). Control and treatment sides were rotated between trials. For all experiments except the concentration screen (used twice), the rings were prepared fresh for each assay. The number of mosquitoes responding to control (hexane only) and treatment were counted in real time and the information also retrieved from video recordings. When testing repellency by racemic and enantiomers, experiments were carried out by using all three compounds, R, S, and racemic in a single set of assays. Each compound was tested interactively, such that R was followed by S and S was followed by racemic, giving rise to a “block” of 3 trials. Three blocks (n=9 for each compound) were conducted per assay. In about half of the trials (n=9 repetitions per compound) test compounds were placed in one of the two sides of the arena. Data were arcsin-transformed before paired two-tailed Student t test comparisons.\n\n\nResults and discussion\n\nWe aimed at cloning CquiOR118, the Cx. quinquefasciatus ortholog of AgamOR8 and AaegOR8. Despite several attempts, we were unable to clone the full length cDNA (VectorBase, CPIJ013954). On the basis of our RNA-Seq findings suggesting a shorter N-terminal amino acid sequence11, we designed a new forward primer considering the starting codon as the next ATG. Indeed, this led to the full length sequence, which was cloned and confirmed by DNA sequencing. We named the shorter version of this gene CquiOR118b, which encodes a protein with 391 amino acid residues and is predicted to have seven transmembrane topology (OCTOPUS, http://octopus.cbr.su.se/). While this manuscript was in preparation it has been reported2 that a longer version of CquiOR118, as predicted in VectorBase, was cloned from another strain of Cx. quinquefasciatus. Thus, both CquiOR118b from the Merced strain (see below) and CquiOR118 from the Thai strain are functional2. Out of 5 clones we sequenced, we obtained two isoforms of CquiOR118b, which differed only in the residue 163 predicted to be in the external cellular loop-2: Ile vs. Val.\n\nOur previous differential expression analysis11 suggested that transcript levels of two genes from the same clade, CquiOR114 and CquiOR117, are significantly higher in antennae than control tissues (legs). Because of the predicted longer C-terminus amino acid sequences encoded by these genes11, we designed primers that would allow us to clone the short and longer versions of these genes. No PCR product was generated with primers for the short sequences, but we cloned and sequenced cDNAs (CquiOR114b) encoding proteins with 405 amino acid residues (longer C-terminus) and predicted seven transmembrane topology. Out of 5 cloned sequenced, we found two isoforms, which differed in 3 amino acid residues. We named them CquiOR114b-1 (Leu-63, Gly-122, and Asp-129) and CquiOR114b-2 (Trp-63, Glu-122, and Asn-129). These residues are predicted to be part of the first transmembrane segment formed by residues 60 to 80, and the internal cellular loop-2 formed by residues 112–150.\n\nQuantitative PCR analysis showed that indeed CquiOR114b was expressed in antennae but not in the maxillary palps, whereas CquiOR118b was expressed in the maxillary palps but not in antennae (Figure 2). Next, we used the Xenopus oocyte recording system to compare the responses of the newly cloned ORs and their isoforms.\n\nData show high transcription levels of CquiOR114b and Cqui118b in antennae and maxillary palps, respectively.\n\nInitially, the responses of oocytes co-expressing one of the isoform of CquiOR118 and the obligatory co-receptor CquiOrco were compared. Since there were no significant difference in the responses elicited by Ile-163-CquiOR118b and Val-163-CquiOR118b, we used only colony 1 (GenBank, KT022418) in subsequent analysis. Next, we challenged CquiOR118b·CquiOrco-expressing oocytes with enantiomers of 1-octen-3-ol and C8 analogs, namely, 1-octyn-3-ol and 3-octanol. While robust responses were elicited by (R)-1-octen-3-ol in a dose-dependent manner, currents generated by its antipode, (S)-1-octen-3-ol, were relatively very small (Figure 3). The remarkable ability of CquiOR118b expressed in a heterologous system to discriminate enantiomers of 1-octen-3-ol is in line with the observations with the intact olfactory system5. Likewise, CquiOR118b showed dramatic enantioselectivity towards (R)-as compared to (S)-1-octyn-3-ol. The receptor showed reduced selectivity towards the saturated analog, 3-octanol. Although at higher doses it preferred (S)-3-octanol, the responses to (R)-3-octanol were relatively high. It is worth mentioning that the hydroxyl group in the (S)-enantiomer of the saturated analog has the orientation as in the (R)-isomers of the unsaturated counterparts (Figure 1), their nomenclature differing (S vs. R) because of the IUPAC rules, not the orientation of the polar moiety expected to fit in the binding cavity of the receptor. Our findings suggest that with CquiOR118b per se the mosquito olfactory system is unlikely to be able to detect the behaviorally relevant ratio of the isomers of 1-octen-3-ol, i.e., R/S, 84:167. It is, therefore, likely that 1-octen-3-ol is also detected by other receptor(s).\n\nCquiOR118b-CquiOrco-expressing oocytes were challenged with enantiomers of 1-octen-3-ol, 1-octyn-3-ol, and 3-octanol at 0.01, 0.1, and 1 µM doses. (N = 3)\n\nWe first compared the two isoforms of CquiOR114b by challenging with 1-octen-3-ol, 1-octyn-3-ol, and 3-octanol oocytes expressing each isoform, CquiOR114b-1 (GenBank, KT022419) or CquiOR114b-2 (GenBank, KT022420) along with CquiOrco (Figure 4). Traces comparing these ligands at three different doses were almost indistinguishable, with the responses recorded from CquiOR114b-1·CquiOrco-expressing oocytes being slightly higher than those from CquiOR114b-2. We, therefore, used CquiOR114b-1 to obtain dose-dependent curves.\n\nCompounds were delivered in the following order: 1-octyn-3-ol, 1-octen-3-ol, and 3-octanol from 1 to 10 µM (left to right).\n\nCquiOR114b·CquiOrco-expressing oocytes gave robust responses to 3-octanol, with responses to the (R)- and (S)-stereoisomers being almost indistinguishable (Figure 5). Likewise Cqui114b responded to the unsaturated compounds, with a slightly preference for (S)-isoforms. Of notice, currents elicited by (R)-1-octen-3-ol were significantly lower than those obtained with its antipode, (S)-1-octen-3-ol, particularly at 0.1 mM (Figure 5). It is, therefore, likely that this antennal receptor contributes to the overall reception of (S)-enantiomers of unsaturated C8 alcohols.\n\nCquiOR118-CquiOrco-expressing oocytes were challenged with enantiomers of 1-octen-3-ol, 1-octyn-3-ol, and 3-octanol at 1, 10, and 100 µM doses. (N = 3)\n\nPreviously, Cook and collaborators observed an intriguing reduced relative attraction response elicited by (R)-1-octen-3-ol in Y-tube olfactometer, but they were unable to conclude if the effect was true repellency as the design of their arena did not allow repellency measurement. With a recently designed surface landing and feeding assay10, we tested the hypothesis that 1-octen-3-ol is a repellent. Although at very low concentrations of racemic 1-octen-3-ol (0.01 and 0.1%) (Figure 6) mosquitoes were attracted to both sides of the arena, at higher doses (1 and 10%) they were repelled by 1-octen-3-ol. Next, we compared repellency elicited by enantiomers and racemic 1-octen-3-ol. Surprisingly, both (R)- and (S)-1-octen-3-ol were repellent at the 1% dose (Figure 7). We then surmised on the basis of dose dependence curves obtained with CquiOR118b (Figure 3) that other odorant receptor(s) must mediate repellency elicited by (S)-1-octenol-3-ol, possible candidates being CquiOR114b, which we identified from antennae and the recently reported CquiOR113 from maxillary palps12.\n\nIn the surface-landing and feeding assay, females of the southern house mosquito were significantly repelled by racemic 1-octen-3-ol at 1 and 10% doses, but not at lower doses. Filled bars represent control.\n\nFemale mosquitoes were repelled not only by racemic but also enantiopure isomers, (R)- and (S)-1-ocen-3-ol at 1% dose. Filled bars represent control.\n\nWe then attempted to combine surgery with behavioral measurement to determine if the maxillary palps are the only olfactory tissues involved in reception of this repellent. Mosquitos with ablated antennae show little or no flight activity. It might be that impairing a significant component of the olfactory system, and possibly hygroscopic and thermal detectors, may render mosquitoes completely inactive. By contrast, ablating one or two of the maxillary palps had little effect on mosquito activity. Interestingly, mosquitoes with single or double ablated maxillary palps were still repelled by 1-octen-3-ol (Figure 8). We, therefore, concluded that the maxillary palps are not sufficient for repellency by 1-octen-3-ol. Other appendages, most likely antennae, are involved in the reception of this repellent.\n\nThe effect of surgery on response of female Culex mosquitoes to 1-octen-3-ol was minimal given that mosquitoes with one or two maxillary palps ablated were repelled by 1-octen-3-ol. Filled bars represent control.\n\n\nConclusion\n\nWe have isolated and cloned two odorant receptors from the southern house mosquito sensitive to 1-octen-3-ol and related compounds. CquiOR118b, which is expressed in the maxillary palps, showed remarkable selective and sensitivity towards (R)-1-octen-3-ol and the related alkyne, (R)-1-octnyl-3-ol. To a much lower extent, CquiOR118b-CquiOrco-expressing oocytes discriminated enantiomers of 3-octanol. By contrast, antennal CquiOR114b responded equally to enantiomers of 3-octanol and showed preference for (S)-isomers of 1-octen-3-ol and 1-octyn-3-ol. Repellency assays showed that both isomers of 1-octen-3-ol, a known attractant for Anopheles and Aedes mosquitoes, were indeed repellents to Cx. quinquefasciatus. However, the maxillary palps alone are not enough for detection of this repellent.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data, 10.5256/f1000research.6646.d4987813", "appendix": "Author contributions\n\n\n\nPX and WSL designed the experiments. PX, FZ, and GKB carried out the research. PX, FZ, GKB, and WSL analyzed the data. WL wrote the manuscript. All authors have agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported in part by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health under award R01AI095514. FZ sabbatical leave at UC Davis was supported in part by the Chinese Scholarship Council.\n\nI confirm that the 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 Dr. Anthon Cornel (University of California, Department of Entomology & Nematology) for providing mosquitoes that allowed us to duplicate his colony at the Davis campus, Dr. Nik Nikbakht, for maintaining the Davis colony, supplying mosquitoes for behavioral assays, and commenting on an earlier version of the manuscript, Dr. Robert Bedoukian (Bedoukian Research Inc.) for providing chemicals used in this research.\n\n\nReferences\n\nMurahashi S: About the fragrances of matsutake. Sci Pap Inst Phys Chem Res. 1936; 30: 263–71.\n\nHall DR, Beevor PS, Cork A, et al.: 1-Octen-3-ol. A potent olfactory stimulant and attractant for tsetse isolated from cattle odours. Insect Sci Applic. 1984; 5(5): 335–9. Publisher Full Text\n\nTakken W, Kline DL: Carbon dioxide and 1-octen-3-ol as mosquito attractants. J Am Mosq Control Assoc. 1989; 5(3): 311–6. PubMed Abstract\n\nKline DL, Allan SA, Bernier UR, et al.: Evaluation of the enantiomers of 1-octen-3-ol and 1-octyn-3-ol as attractants for mosquitoes associated with a freshwater swamp in Florida, U.S.A. Med Vet Entomol. 2007; 21(4): 323–31. PubMed Abstract \| Publisher Full Text\n\nSyed Z, Leal WS: Maxillary palps are broad spectrum odorant detectors in Culex quinquefasciatus. Chem Senses. 2007; 32(8): 727–38. PubMed Abstract \| Publisher Full Text\n\nLu T, Qiu YT, Wang G, et al.: Odor coding in the maxillary palp of the malaria vector mosquito Anopheles gambiae. Curr Biol. 2007; 17(18): 1533–44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCook JI, Majeed S, Ignell R, et al.: Enantiomeric selectivity in behavioural and electrophysiological responses of Aedes aegypti and Culex quinquefasciatus mosquitoes. Bull Entomol Res. 2011; 101(5): 541–50. PubMed Abstract \| Publisher Full Text\n\nGrant AJ, Dickens JC: Functional characterization of the octenol receptor neuron on the maxillary palps of the yellow fever mosquito, Aedes aegypti. PloS One. 2011; 6(6): e21785. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBohbot JD, Dickens JC: Characterization of an enantioselective odorant receptor in the yellow fever mosquito Aedes aegypti. PloS One. 2009; 4(9): e7032. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXu P, Choo YM, De La Rosa A, et al.: Mosquito odorant receptor for DEET and methyl jasmonate. Proc Natl Acad Sci U S A. 2014; 111(46): 16592–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeal WS, Choo YM, Xu P, et al.: Differential expression of olfactory genes in the southern house mosquito and insights into unique odorant receptor gene isoforms. Proc Natl Acad Sci U S A. 2013; 110(46): 18704–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHill SR, Majeed S, Ignell R: Molecular basis for odorant receptor tuning: a short C-terminal sequence is necessary and sufficient for selectivity of mosquito Or8. Insect Mol Biol. 2015. PubMed Abstract \| Publisher Full Text\n\nLeal W, Xu P, Zhu F, et al.: Dataset 1 in: 1-Octen-3-ol – the attractant that repels. F1000Research. 2015. Data Source" }	[ { "id": "9112", "date": "29 Jun 2015", "name": "Joseph C. Dickens", "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 detail studies aimed at understanding the olfactory detection of 1-octen-3-ol in the southern house mosquito, Culex quinquefasciatus. Behavioral studies showed mosquitoes to be repelled by high concentrations of 1-octen-3-ol. An odorant receptor (OR) identified in the maxillary palps responded selectively to (R)-1-octen-3-ol in a manner similar to ORs in other mosquitoes. However, when the palps were ablated, the repellent effects of octenol at high concentrations remained. The authors then demonstrated a second OR expressed primarily in antennae that responded to high concentrations of (S)-1-octen-3-ol. The involvement of this antennal OR in the repellency of octenol at high concentrations is postulated. This is an intriguing paper that expands our knowledge octenol reception in the southern house mosquito. Since the antennal OR postulated for octenol reception is activated only at relatively high concentrations (10-5M and above), it would be interesting to discover its natural ligand.", "responses": [] }, { "id": "9109", "date": "01 Jul 2015", "name": "Jeffery K. Tomberlin", "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 very much enjoyed reading this article. And, I believe it makes a significant positive contribution to our understanding of mosquito responses to volatiles associated with hosts. I believe the title and abstract are appropriate for the publication. With regards to the experiment design, were the trials conducted on the same day from the same population of mosquitoes (could there be a generation effect)? Would the authors consider using logistic regression to analyze the data? Such an approach would determine if there is an interaction between dose, response, and trial?Furthermore, was there much movement between the treatments (i.e., were mosquitoes flying from one treatment to the next)? If so, how did these responses vary across concentrations? These data would determine if there was increased activity as a response to the treatment.Another interesting aspect of the study would be to look at the raw data in conjunction with percent response as these data would also lend towards appreciating the compound \"exciting\" the mosquitoes.The discussion and conclusions are well developed. One aspect that would be important to consider is the relationship between the volatile and its source (most likely a fungus) as this compound operates similarly to quorum sensing molecules (concentration of the compound dictates \"behavior\" of the microbe). By building a bridge between the role of the compound with its source, one would be able to tie together the ecology of the mosquito with the source (i.e., microbe) and host health. Such an approach could provide greater explanation as to why certain doses are attractant while others are repellent.", "responses": [] }, { "id": "9110", "date": "02 Jul 2015", "name": "Wynand van der Goes van Naters", "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\nXu et al. report experiments on the reception and behavioral response in the southern house mosquito Culex quinquefasciatus to enantiomers and analogs of 1-octen-3-ol. Specifically, the authors find that the maxillary palp receptor Or118b is especially sensitive to the R(-) enantiomer while the antennal receptor Or114b mediates stronger responses to the S(+) than to the R(-) enantiomer. The authors show that both enantiomers repel female Cx. quinquefasciatus, as does a racemic mixture. Females in which the maxillary palps have been ablated are similarly repelled by racemic 1-octen-3-ol, proving that the maxillary palps are not solely responsible for the response to this chemical. The paper makes an important contribution within the context of the literature on insect reception of 1-octen-3-ol, which is an attractive kairomone for several haematophagous insects. Experiments are compelling and the conclusions follow from the data. I have only a few minor suggestions:Could the authors please provide more detail on the stimulus method in the oocyte recordings? How long was each stimulus pulse? Black lines above traces sometimes indicate the timing of the stimulus pulse, but this is not the case in Figure 4. Legend of Figure 4: please check the concentration range (to 100 micromolar). Legend of Figure 5, below the title of the legend, starts with \"CquiOR118...\" instead of \"CquiOR114...\" Please indicate the meaning of the error bars in the figures (standard deviation or SEM or CI?). While the reader can infer that Or118 was identified by BLAST as an ortholog of Or8 from An. gambiae, it is not obvious how Or114 was identified. Could the authors perhaps include a paragraph on the bioinformatics that underlies this work? It is interesting that 1-octen-3-ol repels this mosquito species, but is attractive to several other species. Is there a possibility that 1-octen-3-ol could be attractive for Culex quinquefasciatus when combined with other chemicals?", "responses": [] }, { "id": "9111", "date": "02 Jul 2015", "name": "Kenneth F. Haynes", "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 first report that connects olfactory receptors to 1-octen-3-ol in a Culex mosquito to behavioral response. In addition, both enantiomers were shown to be repellents. Interestingly the response in the Culex quinquefasciatus contrasts with that seen in Anopholes and Aedes mosquitoes. The adaptive explanation for the contrast between species remains to be explained. This paper reports many important steps towards understanding the mechanisms underlying the behavioral response.", "responses": [] } ]	1	https://f1000research.com/articles/4-156
https://f1000research.com/articles/4-81/v1	30 Mar 15	{ "type": "Software Tool Article", "title": "Kvik: three-tier data exploration tools for flexible analysis of genomic data in epidemiological studies", "authors": [ "Bjørn Fjukstad", "Karina Standahl Olsen", "Mie Jareid", "Eiliv Lund", "Lars Ailo Bongo", "Bjørn Fjukstad", "Karina Standahl Olsen", "Mie Jareid", "Eiliv Lund" ], "abstract": "Kvik is an open-source system that we developed for explorative analysis of functional genomics data from large epidemiological studies. Creating such studies requires a significant amount of time and resources. It is therefore usual to reuse the data from one study for several research projects. Often each project requires implementing new analysis code, integration with specific knowledge bases, and specific visualizations. Existing data exploration tools do not provide all the required functionality for such multi-study data exploration. We have therefore developed the Kvik framework which makes it easy to implement specialized data exploration tools for specific projects. Applications in Kvik follow the three-tier architecture commonly used in web applications, with REST interfaces between the tiers. This makes it easy to adapt the applications to new statistical analyses, metadata, and visualizations. Kvik uses R to perform on-demand data analyses when researchers explore the data. In this note, we describe how we used Kvik to develop the Kvik Pathways application to explore gene expression data from healthy women with high and low plasma ratios of essential fatty acids using biological pathway visualizations. Researchers interact with Kvik Pathways through a web application that uses the JavaScript libraries Cytoscape.js and D3. We use Docker containers to make deployment of Kvik Pathways simple.", "keywords": [ "Functional genomics", "Epidemiological studies", "Data exploration", "On-demand data analysis", "Open-source software", "Kvik" ], "content": "Introduction\n\nVisual explorative analysis is essential to an understanding of biological functions in large-scale ‘omics’ datasets. However, enabling the inclusion of ‘omics’ data in large epidemiological studies requires collecting samples from thousands of people at different biological levels over a long period of time. It is therefore usual to reuse the data for different research questions and projects. Although an existing tool may be useful for one project, no tool provides the required functionality for several different projects.\n\nWe have therefore implemented Kvik, a framework that makes it easy to develop new applications to explore different research questions and data. We have identified five requirements for such applications:\n\nInteractive The applications should provide interactive exploration of datasets through visualizations and integration with relevant information.\n\nFamiliar They should use familiar visual representations to present information to researchers.\n\nSimple to use Researchers should not need to install software to explore their data through the applications.\n\nFlexible Applications should provide support for data from several study designs. This requires the framework to adapt to the statistical analyses used by the applications.\n\nLightweight Applications and statistical analyses should be separated to make it possible for researchers to explore data without having to have the computational power to run the analyses.\n\nThere are several tools for exploring biological data in the context of pathways, such as VisANT (available online at visant.bu.edu) by1, VANTED (available online at vanted.ipk-gatersleben.de)2, enRoute by3 or Entourage by4 (both available online at caleydo.org). However, these tools do not provide the adaptability needed for exploration of multi-exposure datasets. Many existing tools place the visualization, data analysis and storage on the user’s computer, making it necessary to have a powerful computer. In addition, the tools are often stand-alone applications that require users to install them and keep both application and data up to date. In this article we describe how we used Kvik to implement Kvik Pathways, a tool for exploring gene expression in the context of biological pathways. It solves the above requirements as follows:\n\nInteractive Kvik Pathways provides interactive pathway visualizations and information from the popular Kyoto encyclopedia of genes and genomes (KEGG)5 database (available online at kegg.jp).\n\nSimple to use Kvik Pathways uses HTML5 and modern JavaScript libraries to provide an interactive application that runs in any modern web browser.\n\nFamiliar Kvik Pathways uses the familiar pathway representations from KEGG and graphical user interfaces found in modern web applications.\n\nFlexible It uses the R programming language for statistical analyses (r-project.org) so that researchers can tailor analyses to fit the specific research question in each project.\n\nLightweight Kvik Pathways uses a powerful backend provided by the Kvik framework to perform statistical analyses.\n\nBoth Kvik and Kvik Pathways are open-sourced at github.com/fjukstad/kvik. We provide an online version of Kvik Pathways at kvik.cs.uit.no and a Docker image at registry.hub.docker.com/u/fjukstad/kvik to run Kvik Pathways in a local Docker instance or on a cloud service such as Amazon Web Services (aws.amazon.com) or Google Compute Engine (cloud.google.com/compute).\n\n\nMethods\n\nKvik Pathways allows users to interactively explore a molecular dataset, such as gene expression, through a web application. It provides pathway visualizations and detailed information about genes and pathways from the KEGG databases (Figure 1). The Kvik framework provides a flexible statistics back-end where researchers can specify the analyses they want to run to generate data to be used for later visualization. For example, in Kvik Pathways we retrieve fold change for single genes every time a pathway is viewed in the application. This function is run ad-hoc on the back-end servers and generates output that is displayed in the pathways in the client’s web browser. All of these functions are implemented in a simple R script and can make use of all available libraries in R, such as Bioconductor (bioconductor.org).\n\nThe user has selected the gene CPA3 for further exploration.\n\nResearchers modify this R script to, for example, select a normalization method, or to tune the false discovery rate (FDR) used to adjust the p-values that Kvik Pathways uses to highlight significantly differentially expressed genes. Since Kvik Pathways is implemented as a web application and the analyses are run ad-hoc, researchers get an updated application by simply refreshing the Kvik Pathways webpage.\n\nWe implemented interactive visualizations using the Cytoscape.js (cytoscape.github.com/cytoscape.js) library to generate the interactive pathway visualizations, and D3 (d3js.org) for Document Object Model (DOM) manipulation such as generating bar charts with svg elements. We integrate these with the popular Bootstrap front-end framework (getbootstrap.com) to provide a familiar and aesthetically pleasing user interface.\n\nKvik Pathways has a three-tiered architecture of independent layers (Figure 2). The browser layer consists of the web application for exploring gene expression data and biological pathways. A front-end layer provides static content such as HTML pages and stylesheets, as well as an interface to the data sources with dynamic content such as gene expression data or pathway maps to the web application. The back-end layer contains information about pathways and genes, as well as computational and storage resources to process genomic data such as the NOWAC data repository. The Kvik framework provides the components in the back-end layer.\n\nIn our setup the Data Engine in the back-end layer provides an interface to the NOWAC data repository stored on a secure server on our local Stallo Supercomputer Table 1 provides the interfaces). In Kvik Pathways all gene expression data is stored on the computer that runs the Data Engine. The Data Engine runs an R session accessible over remote procedure calls (RPCs) from the browser layer using RPy2 (rpy.sourceforge.net) to interface with R.\n\nAll URLs are relative to the hostname where the Data Engine server runs. On our public installation the Data Engine runs on kvik.cs.uit.no:8888. For example, use kvik.cs.uit.no:8888/genes/ to retrieve all available genes in our dataset.\n\nTo create pathway visualizations the Kvik backend retrieves and parses the KEGG Markup Language (KGML) representation and pathway image from KEGG databases through its REST API (rest.kegg.jp). This KGML representation of a pathway is an XML file that contains a list of nodes (genes, proteins or compounds) and edges (reactions or relations). Kvik parses this file and generates a JSON representation that Kvik Pathway uses to create pathway visualizations. Kvik Pathways uses the Javascript visualization library Cytoscape.js (js.cytoscape.org) to create a pathway visualization from the list of nodes and edges and overlay the nodes on the pathway image. To reduce latency when using the KEGG REST API, we cache every request on our servers locally. We use the average fold change between the groups in the sample set to color the genes within the pathway maps. To highlight p-values, the pathway visualization shows an additional colored frame around the node. We visualize fold change values for individual samples as a bar chart in a side panel. This bar chart gives researchers a global view of the fold change in the entire dataset.\n\nKvik Pathways runs in all modern web browsers and does not require any third-party software.\n\n\nUse case\n\nWe used Kvik Pathways to repeat the analyses in a previous published project (6, doi: 10.1371/journal.pone.0067270) that compared gene expression in blood from healthy women with high and low plasma ratios of essential fatty acids. Gene expression differences between groups were assessed using t-tests (p-values adjusted with the Benjamini-Hochberg method). There were 184 differentially expressed genes significant on the 5% level. When exploring this gene list originally, functional information was retrieved from GeneCards and other repositories, and the list was analyzed for overlap with known pathways using MSigDB (available online at broadinstitute.org/gsea/msigdb). The researchers had to manually maintain overview of single genes, gene networks or pathways, and gather functional information gene by gene while assessing differences in gene expression levels. With this approach, researchers are limited by manual capacity, and the results may be prone to researcher bias.\n\nInitially, Kvik Pathways was implemented to explore gene expression data from a not yet published dataset. To use Kvik Pathways to explore the data from the analyses in6, we only needed to make small modifications to the R script used by the Data Engine. (The modified R script is found at github.com/fjukstad/kvik/blob/master/dataengine/data-engine.r). Instead of loading the unpublished dataset, we could load the dataset from6 and reuse the functions that are accessible over RPC. Currently this script is less than 30 lines, consisting of four functions to retrieve data and a simple initialization step that reads the dataset. These functions are: get(genes), genes(), f c(genes) and pvalues(genes). get retrieves all information available for the given genes. genes() returns a list of all of the genes in the dataset. f c(genes) returns the fold change for the selected genes. pvalues(genes) returns the p-values for the given genes. After updating the R script in the Data Engine researchers using Kvik Pathways only had to reload a web page to get updated Kvik Pathways.\n\nAs an example of practical use of Kvik Pathways, we chose one of the significant pathways from the overlap analysis, the renin-angiotensin pathway (Supplementary table S5 in6). The pathway contains 17 genes, and in the pathway map we could instantly identify the two genes that drive this result. The color of the gene nodes in the pathway map indicates the fold change, and the statistical significance level is indicated by the color of the node’s frame. We use this visual image of a biological process to see how these two genes (and their expression levels) are related to other genes in that pathway, giving a biologically more meaningful context as compared to merely seeing the two genes on a list.\n\n\nSummary\n\nKvik Pathways is an open-source system for explorative analyses of functional genomics data from epidemiological studies. It uses R to perform on-demand data analyses providing a flexible back-end that can expand to new analyses and research projects. It uses modern visualization libraries and a powerful back-end for performing on-demand statistical analyses. Epidemiological researchers have used Kvik Pathways to analyze gene expression data. Kvik Pathways is open-sourced at github.com/fjukstad/kvik and is available as a Docker image at registry.hub.docker.com/u/fjukstad/kvik.\n\n\nData availability\n\nData used in the use case is available in the Gene Expression Omnibus (ncbi.nlm.nih.gov/geo), under accession number GSE15289.\n\n\nSoftware availability\n\nhttps://github.com/fjukstad/kvik\n\nhttps://github.com/F1000Research/kvik/releases/tag/1.0\n\nhttp://dx.doi.org/10.5281/zenodo.16375\n\nThe MIT license", "appendix": "Author contributions\n\n\n\nLAB and BF designed the architecture of the system. BF implemented. All conducted the requirements analysis. EL, MJ, KSO contributed case study. BF drafted manuscript. All authors read, revised and approved the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a grant from the European Research Council, under the title “Transcriptomics in cancer epidemiology - TICE”.\n\n\nAcknowledgements\n\nGene expression profiles were analyzed at the Microarray Resource Center Tromsø, UiT – The Arctic university of Norway.\n\n\nReferences\n\nHu Z, Chang YC, Wang Y, et al.: VisANT 4.0: Integrative network platform to connect genes, drugs, diseases and therapies. Nucleic Acids Res. 2013; 41(Web Server issue): W225–W231. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJunker BH, Klukas C, Schreiber F: VANTED: a system for advanced data analysis and visualization in the context of biological networks. BMC Bioinformatics. 2006; 7(1): 109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPartl C, Lex A, Streit M, et al.: enRoute: Dynamic path extraction from biological pathway maps for in-depth experimental data analysis. In Biological Data Visualization (BioVis), 2012 IEEE Symposium on, pages 107–114. Publisher Full Text\n\nLex A, Partl C, Kalkofen D, et al.: Entourage: visualizing relationships between biological pathways using contextual subsets. IEEE Trans Vis Comput Graph. 2013; 19(12): 2536–2545. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKanehisa M, Goto S: KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 2000; 28(1): 27–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOlsen KS, Fenton C, Frøyland L, et al.: Plasma fatty acid ratios affect blood gene expression profiles--a cross-sectional study of the Norwegian Women and Cancer Post-Genome Cohort. PLoS One. 2013; 8(6): e67270. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8165", "date": "23 Apr 2015", "name": "Paul Klemm", "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 work presented by Fjukstad et al. pursues in pushing the notion of open science in epidemiology. It describes Kvik, a web-based tool for analyzing genomic pathways. I really like the ideas behind it and value the detailed implementation section as well as the state-of-the-art techniques used. On the other hand I think that the paper should focus more on describing the epidemiological context, associated requirements and target groups to communicate the design choices for Kvik. Go into detail on the application workflow.General FeedbackI like the approach the authors take with the paper. The tool they describe seems to be well suited for analyzing genomic pathways in the epidemiological context.I miss a clear statement on the papers contribution. Maybe you can put a bullet list in the Introduction section to tell the reader what things can be done with your software, which could not be done before!In my understanding, epidemiology is a very interdisciplinary in terms of associated experts. You have your clinicians deriving hypotheses from their day to day practice, statisticians deriving statistically sound conclusion as well as biologists and computer scientists associated with such projects. Which of these are your target group?When you described your target users, describe what they are trying to find out. How does your tool help them doing that? Does it allow them do their work faster? Do they derive insights they could not get before? The latter would be a huge contribution! Please give more details on the workflow of you system!I miss a clear distinction from the NIK-2014 paper \"Kvik: Interactive exploration of genomic data from the NOWAC postgenome biobank\". The paper was also not cited in this work. It seems to me that the majority of the content of the presented paper can already be found in the NIK-2014 paper. Please elaborate on the differences and cite the paper. If you can not state clear differences, there is, in my opinion, no point in publishing this paper and I will rate it 'Not Approved'.TitleThe title of the work is appropriate.AbstractThe abstract motivates the need for new tools, which allow to assess the vast amount of epidemiological data well. In my opinion it can be improved by:reduce the amount of implementation detail. You tell the reader later on which frameworks and libraries you useexplain who are your users.what can be done with your tool now, which could not be done before?Minor comments on the abstract:\"Existing data exploration tools do not provide all the required functionality for such multi-study data exploration.\" This is a dangerous statement, since you do not say anything about what the required functionality is! I think I know what you are trying to get at, in the introduction you describe it better with: \"Although an existing tool may be useful for one project, no tool provides the required functionality for several different projects.\"IntroductionThe introduction can be improved by clearly stating the contributions (e.g., as bullet points).I would like to see some reference or a method on how the five requirements were acquired. These are all things, which are important in almost all applications. Where is the difference of software in the epidemiological context towards other context and how does Kvik adapt to the arising requirements? You answer many of these questions, later on when you repeat all the requirements again, but to me it is not structured well.MethodsThe method section is written well. I would like to know how the users modify the R scripts (beginning second paragraph). Do they do this inside Kvik or do they have to switch into another software for it?Figure 1 caption: What can the user do now after he or she selected the gene? The workflow is not clear to me.Figure 2 was already presented very similarly in the NIK'14 paper.Minor: Closing parenthesis in sentence \"In our setup the Data Engine in the back-end layer provides an interface to the NOWAC data repository stored on a secure server on our local Stallo Supercomputer Table 1 provides the interfaces).\"Use CaseThe use case section can be strengthened by reducing the amount of implementation details (in my opinion mentioning the individual function names is not necessary to comprehend the functionality) and focusing more on the involved actors and tasks and contexts associated with the use case. What feedback was given by the user(s)?ReusabilityThe effort of the authors to make the software publicly available is worth a special note. Modern state of the art techniques are combined with powerful back-end systems, which scale well on different application scenarios.", "responses": [ { "c_id": "1404", "date": "16 Jun 2015", "name": "Bjørn Fjukstad", "role": "Author Response", "response": "We would first like to thank the reviewer Paul Klemm for his thorough feedback and comments.Difference between the NIK paper and the Application NoteWhen we wrote the NIK paper, Kvik was only a system for exploring genomic data in the context of pathways. Since then we have realized that exploring genomic data in pathways is not enough, that we need a framework that allows us to build different applications for exploring data from different studies with different designs. With this in mind we have refined the requirement analysis from our initial Kvik system and developed more general requirements that these applications should satisfy.From our initial Kvik implementation we have now decoupled the application (Kvik Pathways) from the framework, allowing fast development of new applications. The Kvik framework provides interfaces to a Data Engine that provides statistical analyses, and interfaces to online databases such as KEGG. Kvik Pathways is the first application that we developed using the Kvik Framework. Using Kvik we have developed several other applications that will be published in the near future.In the NIK paper we described from a computer science point of view the features of Kvik, both looking at the application itself, and the backend features that are now a apart of the Kvik Framework. The NIK paper was written to give a more in detail view of how the system works and performs, while in this application note we want to describe how our epidemiology researchers helped to develop the application and how they used it to reproduce results they found in an already published dataset.Using an already published dataset was important to us since it allows us to provide a publicly available Kvik installation for others to use. We will revisit the second paragraph of the Use case section where we discussed how we used the initial Kvik system to explore different data from a different study design.To sum up, our new contributions in the application note are as follows:Publicly available application Publicly available Docker containers that researchers can use to set up local installations of Kvik Pathways. Reproduced the results from an already published dataset to make the system publicly accessible at kvik.cs.uit.no A more refined requirement analysis that reflects our experiences after publishing the NIK paper. The important changes are:i) emphasis on integration of online knowledge bases (interactive requirement),ii) emphasis on the system being flexible to adapt to data and different statistical analyses,iii) we removed the security since we believe that data should be publicly available, andiv) put emphasis on separating computation and visualization (lightweight).We have cited the NIK paper in the application note and improved the text to highlight the differences between the framework and the Kvik Pathways application.We have included a list of contributions in the Introduction section.We agree that epidemiology is a very interdisciplinary. Kvik Pathways has been developed in a team of epidemiologists, biologists, statisticians and computer scientists. The application note targets such groups of researchers working together to develop systems for gaining biological insights in genomic data. We have re-written parts of the note to clarify how researchers have used the application.Abstract : We agree and have reduced the amount of implementation detail and made it more specific what our users have done with the application.Introduction: We have revisited the requirements and specified how these are different from regular applications. We believe that it is best to separate the requirements and how we solved them in two different lists.Methods: As of today users modify R scripts outside Kvik. We have made it clear in the methods section. We will expand Figure 1 caption to clearify the workflow. Regarding figure 2. We chose to include it since it highlights the important three-tiered architecture with applications that use the Kvik framework. We have modified the figure to highlight connection between the application and the framework.Use case: As mentioned we will expand this section with a more detailed workflow. Since it is an app note targeted towards users we will reduce implementation details and refer to the source code." } ] }, { "id": "8533", "date": "30 Apr 2015", "name": "Zhenjun Hu", "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 presents Kvik as an open-source system for explorative analysis of functional genomics data from large epidemiological studies. The authors seem have excellent ideas, but the implementation of the tool is far behind these ideas. I would like to approve the manuscript if the following points can be addressed: The target of the tools. There are in general two type of tools: data provider and tool provider. The two of course can be combined. The prior in general provides knowledge, and the later provides functions to analyze users' own data. Kvik however seems lack the data to be a knowledge provider, and also does not provide enough functionality to be the later. To be the former, I will recommend authors to add more epidemiological data, to be the later, the author need to give clear instruction how user's own data can be analyzed using Kvik. For an example, the idea to connect to the cloud service is excellent, but how can Kvik to achieve this? The implementation of Kvik seems to be improved, especially the performance. When I tried the Kvik, the browser tells me several time that the page is not responding. Yet I know the page is responding, but just take too much time. In addition, from data security point of view, it is not good to use RPCs from the browser layer to data engine directly, it shall be avoid in general in the three-tier architecture. The manuscript need to focus more on the functionality of the tool. The current manuscript has too many technical details.", "responses": [ { "c_id": "1405", "date": "16 Jun 2015", "name": "Bjørn Fjukstad", "role": "Author Response", "response": "We would first like to thank the reviewer Zhenjun Hu for his thorough feedback and comments.Since we have open-sourced the application we believe that Kvik Pathways can provide knowledge in itself, but also be used by other researchers to gain knowledge in their data. We have modified the use case to make it more clear how users can do so themselves. We also refer to the online repository where this information is also available. As we state in the Introduction we provide Docker images for running Kvik Pathways in a cloud service. We agree, the current implementation can be improved. We are working on a second version where we have reduced the latency and that gives more feedback to the user if he has to wait. Regarding using RPCs from the browser layer we agree. We have updated the note with more details on how the system is implemented. The Data Engine provides an HTTP REST API that the browser layer queries. It does not send RPC requests from the browser layer, but from the frontend when it receives a request to the HTTP REST API. Also, the only RPCs that are allowed to run are defined in the R analysis script. We agree. We have reduced the number of implementation details to make the functionality stand out more." } ] } ]	1	https://f1000research.com/articles/4-81
https://f1000research.com/articles/3-295/v1	05 Dec 14	{ "type": "Research Article", "title": "The PDB database is a rich source of alpha-helical anti-microbial peptides to combat disease causing pathogens", "authors": [ "Sandeep Chakraborty", "My Phu", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Abhaya M. Dandekar", "My Phu", "Basuthkar J. Rao", "Bjarni Asgeirsson", "Abhaya M. Dandekar" ], "abstract": "The therapeutic potential of α-helical anti-microbial peptides (AH-AMP) to combat pathogens is fast gaining prominence. Based on recently published open access software for characterizing α-helical peptides (PAGAL), we elucidate a search methodology (SCALPEL) that leverages the massive structural data pre-existing in the PDB database to obtain AH-AMPs belonging to the host proteome. We provide in vitro validation of SCALPEL on plant pathogens (Xylella fastidiosa, Xanthomonas arboricola and Liberibacter crescens) by identifying AH-AMPs that mirror the function and properties of cecropin B, a well-studied AH-AMP. The identified peptides include a linear AH-AMP present within the existing structure of phosphoenolpyruvate carboxylase (PPC20), and an AH-AMP mimicing the properties of the two α-helices of cecropin B from chitinase (CHITI25). The minimum inhibitory concentration of these peptides are comparable to that of cecropin B, while anionic peptides used as control failed to show any inhibitory effect on these pathogens. Substitute therapies in place of conventional chemotherapies using membrane permeabilizing peptides like these might also prove effective to target cancer cells. The use of native structures from the same organism largely ensures that administration of such peptides will be better tolerated and not elicit an adverse immune response. We suggest a similar approach to target Ebola epitopes, enumerated using PAGAL recently, by selecting suitable peptides from the human proteome, especially in wake of recent reports of cationic amphiphiles inhibiting virus entry and infection.", "keywords": [ "The abundance of alpha helical (AH) structures present within proteins bears testimony to their relevance in determining functionality1. AHs are key components in protein-protein interaction interfaces2", "DNA binding motifs3", "proteins that permeate biological membranes4", "and anti-microbial peptides (AMP)5", "6. Not surprisingly", "these AHs are the targets for antibody binding7", "8 and therapeutic agents9. These therapies in turn use AH peptides against both viral10–12 and bacterial pathogens13." ], "content": "Introduction\n\nThe abundance of alpha helical (AH) structures present within proteins bears testimony to their relevance in determining functionality1. AHs are key components in protein-protein interaction interfaces2, DNA binding motifs3, proteins that permeate biological membranes4, and anti-microbial peptides (AMP)5,6. Not surprisingly, these AHs are the targets for antibody binding7,8 and therapeutic agents9. These therapies in turn use AH peptides against both viral10–12 and bacterial pathogens13.\n\nSome AHs have unique characteristics, which are strongly correlated to their significance in the function of a protein7. For example, hydrophobic residues aligned on one surface (characterized by a hydrophobic moment14), is critical for virus entry into host cells15, and in the permeabilizing abilities of AH-AMPs16. Often, AHs have cationic residues on the opposite side of the hydrophobic surface, which helps them target bacterial membranes17,18. We have previously implemented known methods19 of evaluating these properties, and provided this as open source software (PAGAL)20. PAGAL was used to characterize the proteome of the Ebola virus7, and to correlate the binding of the Ebola protein VP2421 to human karyopherin22 with the immune suppression and pathogenicity mechanisms of Ebola and Marburg viruses23.\n\nPlant pathogens, like Xylella fastidiosa (Xf)24, Xanthomonas arboricola (Xa)25 and Liberibacter crescens (Lc)26 are a source of serious concern for economic27 and humanitarian reasons28. Specifically, we have been involved in developing novel strategies to counter the Pierce’s disease causing Xf, having previously designed a chimeric protein with anti-microbial properties that provides grapevines with enhanced resistance against Xf29. Cecropin B (CECB) is the lytic component of this chimeric protein30,31. However, the non-nativeness of CECB raises concerns regarding its viability in practical applications32.\n\nIn an effort to replace CECB with an equivalent peptide from the grapevine/citrus genome, we present a design methodology to select AH-AMPs from any given genome - Search characteristic alpha helical peptides in the PDB database and locate it in the genome (SCALPEL). CECB consist of two AHs, joined by a small loop. The N-terminal AH is cationic and hydrophobic, while the C-terminal AH consists of primarily hydrophobic residues. Characterizing all available AHs from plant proteins in the PDB database allowed us to identify a peptide with a large hydrophobic moment and a high proportion of positively charged residues, present in both grapevine and citrus (our organisms of interest), mirroring the linear cationic CECB N-terminal AH. One such match was a twenty residue long AH from phosphoenolpyruvate carboxylase in sunflower33. The sequence of this peptide was used to find homologous peptides in the grapevine and citrus genome (PPC20). Subsequently, we used the SCALPEL algorithm to detect two contiguous AHs connected with a loop, mirroring the properties of CECB in a chitinase (CHITI25) from Nicotiana tobaccum (PDBid:3ALG)34. Subsequently, we demonstrate through bioassay experiments that PPC20 from the grapevine and citrus genome, and CHITI25 from the tobacco genome, inhibit Xf, Xa and Lc growth. The minimum inhibitory concentration of these peptides are comparable to that of CECB, while anionic peptides used as controls failed to show any inhibitory effect with these pathogens. Further, we observed variation in the susceptibility of the pathogens to these peptides.\n\n\nMaterials and methods\n\nThe PDB database was queried for the keyword ‘plants’, and proteins with the exact same sequences were removed. This resulted in a set of ~2000 proteins (see list.plants.txt in Dataset 1). These proteins were analyzed using DSSP35 to identify the AHs, and AHs with the same sequence were removed. This resulted in ~6000 AHs (see ALPHAHELICES.zip in Dataset 1). PAGAL was applied to this set of AHs (see RawDataHelix.txt in Dataset 1). This data was refined to obtain peptides with different characteristics. We also computed the set of all pairs of AHs that are connected with a short (less than five residues) loop (see HTH in Dataset 1). This set is used to extract a pair of AHs, such that one of them is cationic with a large hydrophobic moment, while the other comprises mostly of hydrophobic residues. The PAGAL algorithm has been detailed previously20. Briefly, the Edmundson wheel is computed by considering a wheel with centre (0,0), radius 5, first residue coordinate (0,5) and advancing each subsequent residue by 100 degrees on the circle, as 3.6 turns of the helix makes one full circle. We compute the hydrophobic moment by connecting the center to the coordinate of the residue and give it a magnitude obtained from the hydrophobic scale (in our case, this scale is obtained from Jones et al.19). These vectors are then added to obtain the final hydrophobic moment. The color coding for the Edmundson wheel is as follows: all hydrophobic residues are colored red, while hydrophilic residues are colored in blue: dark blue for positively charged residues, medium blue for negatively charged residues and light blue for amides. All protein structures were rendered by PyMol (http://www.pymol.org/). The sequence alignment was done using ClustalW36. The alignment images were generated using Seaview37. Protein structures have been superimposed using MUSTANG38.\n\nSynthesized chemical peptides were obtained from GenScript USA, Inc. The protein molecular weight was calculated per peptide then diluted to 2000µM or 3000µM stock solutions with phosphate buffered saline. Stock solutions were stored in -20°C and thawed on ice before use.\n\nUsing the stock solutions, we made dilute solutions of 300µM, 250µM, 200µM, 150µM, 100µM, 75µM, 50µM, 30µM, 25µM, and 10µM to a final volume of 100µl of phosphate buffered saline. Dilute peptide solutions were stored in -20°C and thawed on ice before use.\n\nXylella fastidosa 3A2 (PD3)39, Xanthomonas arboricola 417 (TYS)40, and Liberibacter crescens BT-1 (BM7)41 media were prepared and autoclaved at 121°C for 15–30 minutes, then cooled and poured into 100 × 15mm sterile petri dishes. Kanamycin (50µg/ml) was added to PD3.\n\nBacteria were inoculated and allowed to grow in liquid medium at 28°C: Xf (5 days), Xa (3 days), and Lc (3 days) to reach the exponential phase. The inoculum was diluted to a working OD of 0.5 (1×107 cells/ml). 10µl of the OD 0.5 was plated with 90µl of liquid media and spread on the pre-made agar plates to create a confluent lawn of bacteria. The bacteria were given an hour to set at room temperature. 10µl of each peptide concentration was spotted onto a plate of agar preseeded with a layer of bacterium. After spotting the plates were incubated at 28°C for 2 to 10 days till zones of clearance were clearly visible and the plates were scored for the minimum inhibitory concentration (MIC) as that beyond which no visible clearance was observed. Data presented is in triplicate, and were identical.\n\n\nResults\n\nCecropin B (CECB) was used as a positive control, as it is known to target membrane surfaces and creates pores in the bacterial outer membrane30,31. CECB consists of an cationic amphipathic N-Terminal with a large hydrophobic moment (Figure 1a), and a C-Terminal comprising mostly of hydrophobic residues, which consequently has a low hydrophobic moment, (Figure 1b) joined by a short loop. Another positive control was a linear AH-AMP consisting of the residues 2-22 of the N-Terminal in CECB (CBNT21) (Figure 1a). The sequences of these are shown in Table 1.\n\nThe color coding for the Edmundson wheel is as follows: all hydrophobic residues are colored red, while hydrophilic residues are colored in blue: dark blue for positively charged residues, medium blue for negatively charged residues and light blue for amides. The hydrophobic moment arrow is not to scale. (a) N-terminal of Cecropin B (CECB) shows its amphipathic nature, with one side being cationic and the other side hydrophobic. (b) C-terminal of CECB consists of mostly hydrophobic residues, and thus has a low hydrophobic moment. (c) Edmundson wheel for PPC20. (d) Edmundson wheel for 3ALGA.α4, which corresponds to the C-terminal of CECB and comprises mostly of hydrophobic residues (low hydrophobic moment). (e) Edmundson wheel for 3ALGA.α5, which corresponds to the cationic, N-terminal of CECB with a large hydrophobic moment.\n\nCO: control peptides SC: SCALPEL generated peptides.\n\nLinear AH-AMPs. In order to choose a peptide mimicking CBNT21 (cationic, amphipathic, large hydrophobic moment), we directed our search to ‘locate a small peptide with a large hydrophobic moment and a high proportion of positively charged residues’ on the raw data computed using PAGAL (See RawDataHelix.txt in Dataset 1). A small peptide is essential for quick and cost effective iterations. Table 2 shows the best matching AHs. Next, we used the sequence of these AHs to search the grapevine and citrus genomes, choosing only those that are present in both genomes. This allowed us to locate an AH from phosphoenolpyruvate carboxylase from sunflower, a key enzyme in the C4-photosynthetic carbon cycle which enhances solar conversion efficiency (PDBid:3ZGBA.α11)33. Figure 2a shows the specific AH located within the protein structure, marked in green and blue. Although DSSP marks the whole peptide stretch as one AH, we chose the AH in blue due to the presence of a small π helix preceding that. We named this peptide PPC20 (Figure 2, Table 1). This peptide is fully conserved (100% identity in the 20 residues) in both grapevine (Accession id:XP_002285441) and citrus (Accession id:AGS12489.1). Figure 2b,c shows the Pymol rendered AH surfaces of PPC20. The Asp259 stands out as a negative residue in an otherwise positive surface (Figure 2c). Since previous studies have noted dramatic transitions with a single mutation on the polar face, it would be interesting to find the effect of mutating Asp259 to a cationic residue42.\n\nAll AHs in plant proteins are analyzed using PAGAL, and the data is pruned for AHs with a high proportion of positive residues, and finally sorted based on their hydrophobic moment. The first match is present in both grapevine and citrus (PDBid:3ZGBA.α11, which is a phosphoenolpyruvate carboxylase from sunflower). We ignored a small π AH in the beginning of this peptide comprising four residues. This peptide has been named PPC20. HM: Hydrophobic moment, RPNR: Relative proportion of positive residues among charged residues, Len: length of the α, NCH: number of charged residues.\n\n(a) 3ZGBA.α11 is marked in green and blue. We ignore the π AH, and also the small AH preceding it (marked in green). PPC20 is marked in blue. (b) Hydrophobic surface of PPC20. (c) Charged surface of PPC20. Asp259 stands out as a negative residue in an otherwise positive surface.\n\nNon-linear AH-AMPs consisting of two AHs. Next, we located two AHs within chitinase from Nicotiana tobaccum (PDBid:3ALGA.α4 and 3ALGA.α5)34 connected by a short random coil such that one of the AHs is cationic and hydrophobic, while the other AH is comprised mostly of hydrophobic, uncharged residues (CHITI25, Figure 3a, Table 1). This peptide mimics the complete CECB protein (Figure 3b). While the properties of the AHs in CHITI25 is reversed from that of CECB, the order in which these AHs occur is not important for functionality. The multiple sequence alignment of CHITI25 from grapevine, citrus and tobacco is shown in Figure 3c. CHITI25 from tobacco is the most cationic (five), followed by citrus (four) and grapevine (three). Thus, it is possible that the antimicrobial properties of CHITI25 from grapevine would be lower than CHITI25 from tobacco. These peptides can be subjected to mutations to enhance their natural anti-microbial properties in such a scenario43.\n\n(a) PDBid:3ALGA.α4 in green, loop in magenta and 3ALGA.α5 in blue. (b) Superimposing CECB (PDBid:2IGRA) in red with CHITI25 in green using MUSTANG38. Note, that the order of the AHs are reversed. (c) Multiple sequence alignment of CHITI25 from grapevine (CHITIVit), citrus (CHITICit) and tobacco (CHITITob). CHITITob is the more cationic than CHITIVit or CHITICit.\n\nNegative control - an anionic AH-AMP. We also located an anionic AH-AMP using a similar strategy - a 13 residue peptide present within the structure of isoprene synthase from gray poplar (PDBid:3N0FA.α18)44. We also used phosphate buffered saline as a negative control. We have extended this helix on both terminals by including one adjacent residue from both terminals to obtain ISS15 (Table 1).\n\nWe have validated our peptides using plating assays (Table 3, Figure 4). CECB, the well-established AH-AMP, is the most potent among all the peptides tested, having minimum inhibitory concentrations of between 25µM (for Xa) to 100µM (for Xf and Lc). This shows the variations in susceptibilities of different organisms. Understanding this differential susceptibility would require a deeper understanding of the underlying mechanism by which these AH-AMPs work45, as well as the difference in the membrane composition of these gram-negative pathogens46. Mostly, CBNT21 has a slightly lower potency, indicating a role for the C-terminal AH in CECB, which comprises of mostly hydrophobic residues for Xf and Lc. This results corroborates a plausible mechanism suggested by others in which the anionic membranes of bacteria is targeted by the cationic N-terminal, and followed by the insertion of the C-terminal AH into the hydrophobic membrane creating a pore. PPC20 and CHITI25 have comparable potencies with CECB and CBNT21, although Lc appears to be resistant to CHITI25. Finally, the anionic peptide used as a negative control shows no effect on these pathogens.\n\nIt can be seen that CECB is the most efficient among all the peptides for all three pathogens, while the anionic ISS15 does not show any effect even at higher concentrations. However, while CHITI25 is almost as effective as CECB for Xf, it fails to inhibit Lc growth. Also, Xa is much more susceptible to these peptides compared to the other two pathogens. Finally, the anionic ISS15 has no effect on these pathogens. Data is in triplicate, and were identical.\n\nPlating assay to determine minimum inhibitory concentration (MIC) of SCALPEL identified peptides for Xanthomonas arboricola. Counter-clockwise: 300µM, 250µM, 200µM, 150µM, 100µM 75µM, 50µM, 30µM, 25µM, 10µM, PBS. CECB: MIC 25, CBNT2: MIC 10, PPC20: MIC 50, CHITI25: MIC 150, ISS15: MIC >300.\n\n\nDiscussion\n\nThe repertoire of defense proteins available to an organism is being constantly reshaped through genomic changes that confer resistance to pathogens. Genetic approaches aim at achieving the same goal of enhancing immunity through rational design of peptides13,47, which are then incorporated into the genome29,31,48. Also, it is important to ensure that these non-endogenous genomic fragments have minimal effect on humans for their commercial viability32. Identifying peptides from the same genome helps allay these concerns to a significant extent. The key innovation of the current work is the ability to identify peptides with specific properties (cationic AHs with a hydrophobic surface, linear or otherwise) from the genome of any organism of interest. Such peptides also present less likelihood of eliciting an adverse immune response from the host.\n\nAlternate computational methods for finding such new AMPs based on known AMPs could be of two kinds, although neither method is as effective in obtaining our results. Firstly, a sequence search using BLAST can be done to find a corresponding peptide in the genome, say for cecropin B. However, a BLAST of the cecropin sequence does not give any significant matches in the grapevine or citrus genomes, and is a dead end. In principle, what we need is a peptide with cecropin B like properties - and that information is not encoded in the linear sequence, but in the Edmundson wheel of the AH. The second method for such a search is to find structural homology in the PDB database through a tool like DALILITE49. However, AHs are almost indistinguishable structurally, and the results will give rise to many redundancies. Thus, there are no existing methods tailored to incorporate the quantifiable properties of AHs in the search. We, for the first time, have proposed such a method in SCALPEL.\n\nComputer-assisted design strategies have also been applied in designing de novo AMPs50,51. Other hand curated comprehensive databases for ‘for storing, classifying, searching, predicting, and designing potent peptides against pathogenic bacteria, viruses, fungi, parasites, and cancer cells’52 do not enjoy the automation and vastness of available data elucidated in the SCALPEL methodology.\n\nThere are several caveats to our study. We are yet to ascertain the hemolytic nature of the identified peptides, and will be performing these experiments in the near future. In fact, the selective cytotoxicity against human cancer cells, might be used as a substitute therapy in place of conventional chemotherapy53,54. Although, we have not measured the lipid permeabilizing abilities of our peptides, a recent study has found that potency in permeabilizing bacteria-like lipid vesicles does not correlate with significant improvements in antimicrobial activity, rendering such measurements redundant55. The electrostatic context of an peptide is known to have a significant bearing on its propensity to adopt an AH structure. The ability to predict the folding of peptides requires significant computational power and modelling expertise56. Peptides often remain in random coil conformations, and achieve helical structures only by interacting with anionic membrane models57. It is also possible to measure peptide helicity through circular dichroism spectroscopy58. However, our results have been all positive based on selected choices of peptides arising from our search results, and suggest a high likelihood of getting anti-microbial activity from these peptides. Additionally, we may have to resort to other innovative techniques that have been previously adopted to overcome thermodynamic instability or proteolytic susceptibility59–62.\n\n\nConclusion\n\nTo summarize, we establish the presence of a large number of AH-AMPs ‘hidden’ in the universal proteome. We have designed a methodology to extract such peptides from the PDB database - the ‘Big Data’ center in proteomics. We demonstrate our results on well known plant pathogens - Xf, Xa and Lc. The feasibility of using such peptides in cancer therapies is also strong54. The ability to choose a peptide from the host itself is an invaluable asset, since nativeness of the peptide allays fears of eliciting a negative immune response upon administration. The problem of antibiotic resistance is also increasing focus on peptide based therapies9,63, since it is ‘an enigma that bacteria have not developed highly effective cationic AMP-resistance mechanisms’64. Lastly, in face of the current Ebola outbreak65,66, we strongly suggest the possibility of developing peptides derived from the human genome to target viral epitopes, such as those enumerated for the Ebola virus recently7. A recent study has reported the inhibition of the Ebola virus entry and infection by several cationic amphiphiles67, suggesting the SCALPEL generated cationic peptides with the aid of cell penetrating peptides68 could achieve similar results.\n\n\nData availability\n\nF1000Research: Dataset 1. Data used for SCALPEL search methodology to identify plant alpha helical - antimicrobial peptides in the PDB database, 10.5256/f1000research.5802.d3982369", "appendix": "Author contributions\n\n\n\nSC wrote the computer programs. MP performed the in vitro experiments. All authors analyzed the data, and contributed equally to the writing and subsequent refinement of the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAMD wishes to acknowledge grant support from the California Department of Food and Agriculture PD/GWSS Board. BJ acknowledges financial support from Tata Institute of Fundamental Research (Department of Atomic Energy). Additionally, BJR is thankful to the Department of Science and Technology for the JC Bose Award Grant. BA acknowledges financial support from the Science Institute of the University of Iceland.\n\n\nAcknowledgements\n\nThe pathogen strains used in our study were kindly provided by Steven E. Lindow, University of California, Berkeley (Xylella fastidiosa 3A2), James E. Adaskaveg, University of California, Riverside (Xanthomonas arboricola 417) and Eric Triplett, University of Florida, Gainesville (Liberibacter crescens BT-1).\n\n\nReferences\n\nAzzarito V, Long K, Murphy NS, et al.: Inhibition of α-helix-mediated protein-protein interactions using designed molecules. Nat Chem. 2013; 5(3): 161–173. PubMed Abstract \| Publisher Full Text\n\nLee JH, Zhang Q, Jo S, et al.: Novel pyrrolopyrimidine-based α-helix mimetics: cell-permeable inhibitors of protein-protein interactions. J Am Chem Soc. 2011; 133(4): 676–679. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLandschulz WH, Johnson PF, McKnight SL: The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins. Science. 1988; 240(4860): 1759–1764. PubMed Abstract \| Publisher Full Text\n\nDathe M, Wieprecht T: Structural features of helical antimicrobial peptides: their potential to modulate activity on model membranes and biological cells. Biochim Biophys Acta. 1999; 1462(1–2): 71–87. PubMed Abstract \| Publisher Full Text\n\nWang G: Structures of human host defense cathelicidin LL-37 and its smallest antimicrobial peptide KR-12 in lipid micelles. J Biol Chem. 2008; 283(47): 32637–32643. PubMed Abstract \| Publisher Full Text\n\nWang G: Human antimicrobial peptides and proteins. Pharmaceuticals (Basel). 2014; 7(5): 545–594. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rao B, Asgeirsson B, et al.: Characterizing alpha helical properties of Ebola viral proteins as potential targets for inhibition of alpha-helix mediated protein-protein interactions [v1; ref status: approved with reservations 1, http://f1000r.es/4lg]. F1000Res. 2014; 3: 251. Publisher Full Text\n\nLee JE, Fusco ML, Hessell AJ, et al.: Structure of the Ebola virus glycoprotein bound to an antibody from a human survivor. Nature. 2008; 454(7201): 177–182. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHancock RE, Chapple DS: Peptide antibiotics. Antimicrob Agents Chemother. 1999; 43(6): 1317–1323. PubMed Abstract \| Free Full Text\n\nJudice JK, Tom JY, Huang W, et al.: Inhibition of HIV type 1 infectivity by constrained alpha-helical peptides: implications for the viral fusion mechanism. Proc Natl Acad Sci U S A. 1997; 94(25): 13426–13430. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChampagne K, Shishido A, Root MJ: Interactions of HIV-1 inhibitory peptide T20 with the gp41 N-HR coiled coil. J Biol Chem. 2009; 284(6): 3619–3627. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHong W, Li T, Song Y, et al.: Inhibitory activity and mechanism of two scorpion venom peptides against herpes simplex virus type 1. Antiviral Res. 2014; 102: 1–10. PubMed Abstract \| Publisher Full Text\n\nZeitler B, Herrera Diaz A, Dangel A, et al.: De-novo design of antimicrobial peptides for plant protection. PLoS One. 2013; 8(8): e71687. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEisenberg D, Weiss RM, Terwilliger TC: The helical hydrophobic moment: a measure of the amphiphilicity of a helix. Nature. 1982; 299(5881): 371–374. PubMed Abstract \| Publisher Full Text\n\nBadani H, Garry RF, Wimley WC: Peptide entry inhibitors of enveloped viruses: the importance of interfacial hydrophobicity. Biochim Biophys Acta. 2014; 1838(9): 2180–2197. PubMed Abstract \| Publisher Full Text\n\nChen Y, Guarnieri MT, Vasil AI, et al.: Role of peptide hydrophobicity in the mechanism of action of alpha-helical antimicrobial peptides. Antimicrob Agents Chemother. 2007; 51(4): 1398–1406. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrogden KA: Antimicrobial peptides: pore formers or metabolic inhibitors in bacteria? Nat Rev Microbiol. 2005; 3(3): 238–250. PubMed Abstract \| Publisher Full Text\n\nHuang Y, Huang J, Chen Y: Alpha-helical cationic antimicrobial peptides: relationships of structure and function. Protein Cell. 2010; 1(2): 143–152. PubMed Abstract \| Publisher Full Text\n\nJones MK, Anantharamaiah GM, Segrest JP: Computer programs to identify and classify amphipathic alpha helical domains. J Lipid Res. 1992; 33(2): 287–296. PubMed Abstract\n\nChakraborty S, Rao B, Dandekar A: PAGAL - Properties and corresponding graphics of alpha helical structures in proteins. [v2; ref status: indexed, http://f1000r.es/4e7] F1000Res. 2014; 3: 206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhang AP, Abelson DM, Bornholdt ZA, et al.: The ebolavirus VP24 interferon antagonist: know your enemy. Virulence. 2012; 3(5): 440–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXu W, Edwards MR, Borek DM, et al.: Ebola virus VP24 targets a unique NLS binding site on karyopherin alpha 5 to selectively compete with nuclear import of phosphorylated STAT1. Cell Host Microbe. 2014; 16(2): 187–200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Rao B, Asgeirsson B, et al.: Correlating the ability of VP24 protein from Ebola and Marburg viruses to bind human karyopherin to their immune suppression mechanism and pathogenicity using computational methods [v1; ref status: approved with reservations 1, http://f1000r.es/4o3]. F1000Res. 2014; 3: 265. Publisher Full Text\n\nHopkins D, Purcell A: Xylella fastidiosa: cause of Pierce’s disease of grapevine and other emergent diseases. Plant Disease. 2002; 86: 1056–1066. Publisher Full Text\n\nRyan RP, Vorholter FJ, Potnis N, et al.: Pathogenomics of Xanthomonas: understanding bacterium-plant interactions. Nat Rev Microbiol. 2011; 9(5): 344–355. PubMed Abstract \| Publisher Full Text\n\nLeonard MT, Fagen JR, Davis-Richardson AG, et al.: Complete genome sequence of Liberibacter crescens BT-1. Stand Genomic Sci. 2012; 7(2): 271–283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlston JM, Fuller KB, Kaplan JD, et al.: Assessing the returns to r&d on perennial crops: the costs and benefits of Pierce’s disease research in the California winegrape industry. Aust J Agric Resour Econ. 2014. Publisher Full Text\n\nStrange RN, Scott PR: Plant disease: a threat to global food security. Annu Rev Phytopathol. 2005; 43: 83–116. PubMed Abstract \| Publisher Full Text\n\nDandekar AM, Gouran H, Ibanez AM, et al.: An engineered innate immune defense protects grapevines from Pierce disease. Proc Natl Acad Sci U S A. 2012; 109(10): 3721–3725. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoore AJ, Beazley WD, Bibby MC, et al.: Antimicrobial activity of cecropins. J Antimicrob Chemother. 1996; 37(6): 1077–1089. PubMed Abstract \| Publisher Full Text\n\nSharma A, Sharma R, Imamura M, et al.: Transgenic expression of cecropin B, an antibacterial peptide from Bombyx mori, confers enhanced resistance to bacterial leaf blight in rice. FEBS Lett. 2000; 484(1): 7–11. PubMed Abstract \| Publisher Full Text\n\nShelton AM, Zhao JZ, Roush RT: Economic, ecological, food safety, and social consequences of the deployment of bt transgenic plants. Annu Rev Entomol. 2002; 47: 845–881. PubMed Abstract \| Publisher Full Text\n\nPaulus JK, Schlieper D, Groth G: Greater efficiency of photosynthetic carbon fixation due to single amino-acid substitution. Nat Commun. 2013; 4: 1518. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOhnuma T, Numata T, Osawa T, et al.: Crystal structure and mode of action of a class V chitinase from Nicotiana tabacum. Plant Mol Biol. 2011; 75(3): 291–304. PubMed Abstract \| Publisher Full Text\n\nJoosten RP, te Beek TA, Krieger E, et al.: A series of PDB related databases for everyday needs. Nucleic Acids Res. 2011; 39(Database issue): D411–419. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLarkin MA, Blackshields G, Brown NP, et al.: Clustal W and Clustal X version 2.0. Bioinformatics. 2007; 23(21): 2947–2948. PubMed Abstract \| Publisher Full Text\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\nKonagurthu AS, Whisstock JC, Stuckey PJ, et al.: MUSTANG: a multiple structural alignment algorithm. Proteins. 2006; 64(3): 559–574. PubMed Abstract \| Publisher Full Text\n\nIonescu M, Zaini PA, Baccari C, et al.: Xylella fastidiosa outer membrane vesicles modulate plant colonization by blocking attachment to surfaces. Proc Natl Acad Sci U S A. 2014; 111(37): E3910–E3918. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindow S, Olson W, Buchner R: Colonization of Dormant Walnut Buds by Xanthomonas arboricola pv. juglandis Is Predictive of Subsequent Disease. Phytopathology. 2014; 104(11): 1163–1174. PubMed Abstract \| Publisher Full Text\n\nFagen JR, Leonard MT, Coyle JF, et al.: Liberibacter crescens gen. nov., sp. nov., the first cultured member of the genus Liberibacter. Int J Syst Evol Microbiol. 2014; 64(Pt 7): 2461–6. PubMed Abstract \| Publisher Full Text\n\nJiang Z, Vasil AI, Hale JD, et al.: Effects of net charge and the number of positively charged residues on the biological activity of amphipathic alpha-helical cationic antimicrobial peptides. Biopolymers. 2008; 90(3): 369–383. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang G, Hanke ML, Mishra B, et al.: Transformation of human cathelicidin LL-37 into selective, stable, and potent antimicrobial compounds. ACS Chem Biol. 2014; 9(9): 1997–2002. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoksal M, Zimmer I, Schnitzler JP, et al.: Structure of isoprene synthase illuminates the chemical mechanism of teragram atmospheric carbon emission. J Mol Biol. 2010; 402(2): 363–373. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShai Y: Mechanism of the binding, insertion and destabilization of phospholipid bilayer membranes by alpha-helical antimicrobial and cell non-selective membrane-lytic peptides. Biochim Biophys Acta. 1999; 1462(1–2): 55–70. PubMed Abstract \| Publisher Full Text\n\nKoebnik R, Locher KP, Van Gelder P: Structure and function of bacterial outer membrane proteins: barrels in a nutshell. Mol Microbiol. 2000; 37(2): 239–253. PubMed Abstract \| Publisher Full Text\n\nHancock RE, Sahl HG: Antimicrobial and host-defense peptides as new anti-infective therapeutic strategies. Nat Biotechnol. 2006; 24(12): 1551–1557. PubMed Abstract \| Publisher Full Text\n\nGray D, Li Z, Hopkins D, et al.: Transgenic grapevines resistant to Pierce’s disease. HortScience. 2005; 40(4): 1104–1105. Reference Source\n\nHolm L, Kaariainen S, Rosenstrom P, et al.: Searching protein structure databases with DaliLite v.3. Bioinformatics. 2008; 24(23): 2780–2781. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrecer V, Ho B, Ding JL: De novo design of potent antimicrobial peptides. Antimicrob Agents Chemother. 2004; 48(9): 3349–3357. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFjell CD, Hiss JA, Hancock RE, et al.: Designing antimicrobial peptides: form follows function. Nat Rev Drug Discov. 2011; 11(1): 37–51. PubMed Abstract \| Publisher Full Text\n\nWang G: Database-Guided Discovery of Potent Peptides to Combat HIV-1 or Superbugs. Pharmaceuticals (Basel). 2013; 6(6): 728–758. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMader JS, Hoskin DW: Cationic antimicrobial peptides as novel cytotoxic agents for cancer treatment. Expert Opin Investig Drugs. 2006; 15(8): 933–46. PubMed Abstract \| Publisher Full Text\n\nDouglas S, Hoskin DW, Hilchie AL: Assessment of antimicrobial (host defense) peptides as anti-cancer agents. Methods Mol Biol. 2014; 1088: 159–170. PubMed Abstract \| Publisher Full Text\n\nHe J, Krauson AJ, Wimley WC: Toward the de novo design of antimicrobial peptides: Lack of correlation between peptide permeabilization of lipid vesicles and antimicrobial, cytolytic, or cytotoxic activity in living cells. Biopolymers. 2014; 102(1): 1–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPiana S, Klepeis JL, Shaw DE: Assessing the accuracy of physical models used in protein-folding simulations: quantitative evidence from long molecular dynamics simulations. Curr Opin Struct Biol. 2014; 24: 98–105. PubMed Abstract \| Publisher Full Text\n\nMishra B, Epand RF, Epand RM, et al.: Structural location determines functional roles of the basic amino acids of KR-12, the smallest antimicrobial peptide from human cathelicidin LL-37. RSC Adv. 2013; 3(42): 19560–19571. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuang YB, He LY, Jiang HY, et al.: Role of helicity on the anticancer mechanism of action of cationic-helical peptides. Int J Mol Sci. 2012; 13(6): 6849–6862. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChapman RN, Dimartino G, Arora PS: A highly stable short alpha-helix constrained by a main-chain hydrogen-bond surrogate. J Am Chem Soc. 2004; 126(39): 12252–12253. PubMed Abstract \| Publisher Full Text\n\nBird GH, Madani N, Perry AF, et al.: Hydrocarbon double-stapling remedies the proteolytic instability of a lengthy peptide therapeutic. Proc Natl Acad Sci U S A. 2010; 107(32): 14093–14098. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBird GH, Boyapalle S, Wong T, et al.: Mucosal delivery of a double-stapled RSV peptide prevents nasopulmonary infection. J Clin Invest. 2014; 124(5): 2113–2124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarrison RS, Shepherd NE, Hoang HN, et al.: Downsizing human, bacterial, and viral proteins to short water-stable alpha helices that maintain biological potency. Proc Natl Acad Sci U S A. 2010; 107(26): 11686–11691. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOyston PC, Fox MA, Richards SJ, et al.: Novel peptide therapeutics for treatment of infections. J Med Microbiol. 2009; 58(Pt 8): 977–987. PubMed Abstract \| Publisher Full Text\n\nPeschel A, Sahl HG: The co-evolution of host cationic antimicrobial peptides and microbial resistance. Nat Rev Microbiol. 2006; 4(7): 529–536. PubMed Abstract \| Publisher Full Text\n\nPiot P: The F1000Research: Ebola article collection [v1; ref status: not peer reviewed, http://f1000r.es/4ot]. F1000Res. 2014; 3: 269. Publisher Full Text\n\nPiot P: Ebola’s perfect storm. Science. 2014; 345(6202): 1221. PubMed Abstract \| Publisher Full Text\n\nShoemaker CJ, Schornberg KL, Delos SE, et al.: Multiple cationic amphiphiles induce a Niemann-Pick C phenotype and inhibit Ebola virus entry and infection. PLoS One. 2013; 8(2): e56265. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontrose K, Yang Y, Sun X, et al.: Xentry, a new class of cell-penetrating peptide uniquely equipped for delivery of drugs. Sci Rep. 2013; 3: 1661. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChakraborty S, Phu M, Rao BJ, et al.: Dataset 1 in: The PDB database is a rich source of α-helical anti-microbial peptides to combat disease causing pathogens. F1000Research. 2014. Data Source" }	[ { "id": "8820", "date": "01 Jun 2015", "name": "Jean-Marc Berjeaud", "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\nTitle and Abstract:The Title of the article, \"The PDB database is a rich source of alpha-helical anti-microbial peptides to combat disease causing pathogens\", is appropriate for the content of the article. However because the detected peptides were solely tested toward plant pathogens the term of \"plant\" in the title could be considered. The abstract represent rather well the work presented in the article except the two last sentences which are expectations of the authors but were not studied in this work. Particularly the sentence : \"The use of native…\" assert that peptide structure extracted from native proteins will be without adverse effect to the host but it has to be proved in my opinion.Article content:The paper describes the use of a software and an in silico method developed by the authors to screen the protein database (PDB) to find new antimicrobial peptides on the basis of the secondary structures of these peptides. The method is innovative and very interesting. Moreover the authors proved the efficacy of their method as they synthesized peptides from portions of protein sequences presenting secondary structures resembling cecropin B and demonstrated the antimicrobial activity of these new peptides.However I did not understand why they used an anionic peptide as the negative control. Indeed it is well known that the global positive charge of the peptides is required for their initial stacking on the membrane of target cells. Thus it is predictable that any anionic peptide will be inactive.Conclusions:The main problem concerns the conclusions of the article. Indeed there are not sufficient experimental evidences in the article to assert that the alpha-helical peptides with antimicrobial activity have a strong potency to act toward cancer cells. In my experience I tested several alpha-helical peptides which were solely antimicrobial. Thus I strongly suggest to moderate this conclusions part of the manuscript.", "responses": [ { "c_id": "1412", "date": "05 Jun 2015", "name": "Sandeep Chakraborty", "role": "Author Response", "response": "Dear Dr Berjeaud,We would like to thank you for taking the time to review this paper. Please find our responses below.Title and Abstract: The Title of the article, ”The PDB database is a rich source of alpha-helical anti-microbial peptides to combat disease causing pathogens”, is appropriate for the content of the article. However because the detected peptides were solely tested toward plant pathogens the term of ”plant” in the title could be considered.We are in the process of testing these peptides on other pathogens. Furthermore, considering that the mechanism of the inhibitory effect of these peptides is independent of the pathogen host, we are quite confident of replicating our results on other ‘non-plant’ pathogens.The abstract represent rather well the work presented in the article except the two last sentences which are expectations of the authors but were not studied in this work. Particularly the sentence : ”The use of native” assert that peptide structure extracted from native proteins will be without adverse effect to the host but it has to be proved in my opinion.We believe this is an important hypothesis, although unproven, which differentiates SCALPEL from other methods of identifying anti-microbial peptides. However, we have modified the statement in the abstract to make this less of an assertion, and more of a hypothesis.Article content: The paper describes the use of a software and an in silico method developed by the authors to screen the protein database (PDB) to find new antimicrobial peptides on the basis of the secondary structures of these peptides. The method is innovative and very interesting. Moreover the authors proved the efficacy of their method as they synthesized peptides from portions of protein sequences presenting secondary structures resembling cecropin B and demonstrated the antimicrobial activity of these new peptides.We appreciate the positive comments.However I did not understand why they used an anionic peptide as the negative control. Indeed it is well known that the global positive charge of the peptides is required for their initial stacking on the membrane of target cells. Thus it is predictable that any anionic peptide will be inactive.We have used this as a negative control. If our experimental setup in the process of adding the peptides had any undesired conditions which was inhibiting the pathogens, this anionic peptide would show positive results. So, this is slightly different from a null negative control, as it involves adding a peptide.Conclusions: The main problem concerns the conclusions of the article. Indeed there are not sufficient experimental evidences in the article to assert that the alpha-helical peptides with antimicrobial activity have a strong potency to act toward cancer cells. In my experience I tested several alpha-helical peptides which were solely antimicrobial. Thus I strongly suggest to moderate this conclusions part of the manuscript.We appreciate your concern regarding the lack of confirmatory evidence of AH peptides as anti-cancer therapeutics. However, this continues to be an active front in research, and we hope that SCALPEL will provide further avenues for testing this hypothesis. We have cited two recent papers- http://www.ncbi.nlm.nih.gov/pubmed/25270878 and http://www.ncbi.nlm.nih.gov/pubmed/24101917in this context. Also, we have modified the text to reflect the lack of conclusiveness in such studies.Once again, we are thankful for your insightful comments, and hope to have addressed your concerns.Thanking you,Sincerely,Sandeep ChakrabortyPlant Sciences Department,University of California, Davis, CA 95616, USA." } ] } ]	1	https://f1000research.com/articles/3-295
https://f1000research.com/articles/4-127/v1	26 May 15	{ "type": "Opinion Article", "title": "agINFRA: a research data hub for agriculture, food and the environment", "authors": [ "Andreas Drakos", "Vassilis Protonotarios", "Nikos Manouselis", "Andreas Drakos", "Nikos Manouselis" ], "abstract": "The agINFRA project (www.aginfra.eu) is a European Commission funded project under the 7th Framework Programme that aimed to introduce agricultural scientific communities to the vision of open and participatory data-intensive science. Working on enhancing the interoperability between heterogeneous data sources, the agINFRA project has left a set of grid- and cloud- based services that can be reused by future initiatives and adopted by existing ones, in order to facilitate the dissemination of agricultural research, educational and other types of data. On top of that, agINFRA provided a set of domain-specific recommendations for the publication of agri-food research outcomes. This paper discusses the concept of the agINFRA project and presents its major outcomes, as adopted by existing initiatives activated in the context of agricultural research and education.", "keywords": [ "agri-food", "cloud-based", "grid-based" ], "content": "Introduction\n\nagINFRA was an innovative Integrated Infrastructure Initiative (I3) project (2012–2015) that aimed to introduce agricultural scientific communities to the vision of open and participatory data-intensive science. To achieve this aim, agINFRA designed and developed a scientific data infrastructure for agricultural sciences that facilitates the development of policies and services that promote the sharing of data among agricultural scientists in a manner that develops trust within and amongst their communities. The key goals and corresponding project objectives are as follows:\n\nIncreased sharing and federation of agricultural data (scientific objective)\n\n– Successfully deploy an open infrastructure for the sharing of digital agricultural content;\n\n– Include in the infrastructure resources ranging from raw observational and experimental data through to publications;\n\n– Promote the sharing of data amongst the wider scientific community to build trust;\n\n– Ensure outcomes significantly advance the state-of-the art in agricultural e-infrastructures;\n\nEfficient data management in the agricultural research process (scientific objective)\n\n– Ensure stakeholder needs are met with regards to data management and sharing;\n\n– Involve as many agricultural data sources as possible to provide maximum value;\n\n– Facilitate easier curation, certification, annotation, navigation and management of data;\n\n– Create new opportunities for data intensive research in the agricultural domain\n\nDeployment of robust European service infrastructure for scientific agricultural data (technical objective)\n\n– Deploy tools (exchange standards, software, methodologies) for collaboration between European institutions in data-intensive research;\n\n– Validate the approach by enabling users to interact with each other and the data\n\n– Identify gaps in standards that will be needed to guarantee the integrity/authenticity of data\n\n– Establish specific and realistic indicators to measure success\n\nHigh interoperability between agricultural and other data resources (technical objective)\n\n– Improve interoperability between existing e-infrastructures;\n\n– Successfully interconnect agricultural data repositories through extended metadata;\n\n– Advanced implementation and adoption of European standards and specifications\n\n\nagINFRA outcomes\n\nThe agINFRA project has been seen as a project that would bring a number of advances in the area of agricultural data management and re-usage. Its main goals as they had been set up from the beginning of the project and revised during the project lifetime can be summarized as1:\n\nSet up FAO and its stakeholders as the managing and promoting stakeholder of a data infrastructure for agricultural scientists, enhancing their current reach with projects as the CIARD Routemap to Information Nodes and Gateways (RING - http://ring.ciard.net), AGRIS (http://agris.fao.org/) and the Agriculture Information Management Standards (AIMS - http://aims.fao.org/);\n\nConnect existing networks of data repositories and institutional/national repositories through this data infrastructure, enabling new possibilities for retrieval and data analysis;\n\nConnect this data infrastructure to other infrastructures (such as LifeWatch - http://www.lifewatch.eu/ and ViBRANT - http://vbrant.eu/) and data repository networks (such as VOA3R - http://voa3r.eu/ and Organic.Edunet - http://www.organic-edunet.eu, Integrated capture information system by D4Science - ICIS http://www.d4science.eu/icis) to aggregate and expose their data resources through a flexible infrastructure;\n\nProvide mediating capabilities for heterogeneous, distributed data sets in agriculture, including the necessary scalability and high performance support to handle complex queries and data extraction processes;\n\nProvide machine-processable interfaces to the data resources and systems integrated, enabling the development of semantics-aware applications and using the recommendations widely adopted for open linked data;\n\nFurther integrate the educational content repository stakeholders, allowing agricultural scientists to have access to and share their data resources that they used for education/extension. To achieve its aims, the agINFRA project took upon a number of initiatives to help shape a new area around agricultural data management.\n\nOne of the key elements of the agINFRA analysis that took place during the first years of the project was the stakeholders analysis. We have in several project reports stated that the main prospective “users” of the agINFRA infrastructure are data managers, information service managers and information system developers dealing with agriculture-related data and more precisely:\n\nResearch projects, i.e. researchers and their IT/information managers: This group is of interest mainly as end-users who will use portals supported by agINFRA to access content and data/datasets. Researchers are also providers of content/data which, however, should be deposited in their institutional repositories or national/international subject or domain based repositories from which the metadata is collected into the agINFRA data layer;\n\nIT/information managers of individual research organisations (e.g. university departments, research centres, research libraries, etc.) who manage content/data repositories and information services, from which metadata is collected in the agINFRA data layer;\n\nIT/information managers of shared discipline- or subject-based content/data repositories and services, which are content/data aggregators and service providers on top of the local institutional repositories. This category also includes national or international agencies that aggregate research data and make it available for purposes such as agricultural geographic and other information systems (e.g. for land use planning, conservation of natural resources, risk assessment, etc.);\n\nDevelopers/providers of software-based applications for research and content/data management tasks who are interested to customize or further develop agINFRA components, tools and services. They may become involved in agINFRA-related, open source software development or simply use agINFRA Application Programming Interfaces (APIs) to create useful applications such as mash-ups of research data;\n\nThis stakeholder analysis was essential to understand the potential users of agINFRA, understand their needs and proceed with the deployment of an infrastructure that will serve the actual needs of the agricultural community. The detailed analysis that was done, helped the project partners shape the agINFRA common vision1,2.\n\nFrom the beginning of the agINFRA project, it was promised to develop and deliver a shared technology infrastructure. During the lifetime of the project, agINFRA partners designed and deployed research infrastructure components and APIs (Application Profile Interfaces) that support the agINFRA applications and layers in a federated way. The agINFRA Virtual Organisation (VO) has been created to organise the infrastructure provided by the agINFRA partners (Institute of Physics Belgrade - IPB, Instituo Nazionale di Fisica Nucleare - INFN, Computer and Automation Research Institute of the Hungarian Academy of Science - SZTAKI) but also to connect with external infrastructure providers as for example the Greek Research and Technology Network (GRNET), the National Centre for Scientific Research ‘Demokritos’ (NCSR) or other National Grid Initiatives through the European Grid Initiative network. The “vo.aginfra.eu” Virtual Organization (VO) supporting the agINFRA community is operated on the European Grid Infrastructure (EGI) infrastructure, it is supported by all Grid sites participating in the project (AEGIS01-IPB-SCL, INFN-CATANIA, INFN-ROMA3 and SZTAKI), and it is delivering access to more than 2000 CPU cores and 900 TB of storage space to its users. The primary agINFRA Virtual Organization Membership Service (VOMS) server (voms.ipb.ac.rs) is installed at the IPB. The name vo.aginfra.eu is chosen with a format recommended by EGI, and allows global VO registration at CIC portal that will certainly increase VO visibility3. On top of the agINFRA infrastructure a number of application have been ported to the Grid infrastructure (Grid modules) with a RESTful interface on top that enables configuration of application input parameters and output retrieval. For each of the Grid modules, the applications had to be gridified and exposed through a RESTful API (Figure 1). All the ported applications are related with the aggregation and management of data (e.g. agHarvester) and the list of the services is available through the API section of the agINFRA site (www.aginfra.eu/api).\n\nSource: agINFRA website (www.aginfra.eu).\n\nIn addition agINFRA took advantage of two different Science Gateways (SG) architectures for the management of complex job workflows and on user membership management through the adoption of the Identity Federations and Identity Providers approach for access to Grid and/or Cloud resources. Both of the generic versions of the SG have been customized to support the needs of the agricultural research data community. Finally, one of the biggest risks in e-infrastructure projects is the sustainability and compatibility of the infrastructure elements. From the beginning of the project, the agINFRA infrastructure was created in a way to be fully compatible with the EGI Federated Cloud. Being compliant with such an initiative allows the easy of new resources to be committed to the project. In addition, EGI itself has recognized the work and the needs of the community and during the upcoming EGI-ENGAGE project (the flagship project of EGI) a Virtual Team for the Agricultural Sciences is to be created. At the same time, additional initiatives have shown the will to provide resources and support the work of the agINFRA as for example the pledge for a Linked Open Data layer for the agricultural community.\n\nSource: CIARD RING (http://ring.ciard.net).\n\nOne of the main advances during the agINFRA project was the revamp of the CIARD RING (http://ring.ciard.net) global directory of web-based information services and datasets for agriculture. The CIARD RING, a project maintained within the CIARD initiative and is led by the Global Forum on Agricultural Research (GFAR), has become a machine-readable hub/switchboard to information sources such as search engines, databases, repositories, Open Archives, feeds, data sheets etc (Figure 2). All information stored/registered in the RING i.e. metadata about services and datasets, can be queried via SPARQL and via a simpler REST API. In addition, during the agINFRA project, it has become evident that in order for the RING to become the ‘central broker of the information’ around the agricultural community, even more additions had to be made. Today, the RING includes a registry of organizations and networks with the vision of being part of the future AgriProfiles list of people, networks and organizations in the agricultural research. Starting from the agINFRA project, software are also registered in the RING as a way to easily find and retrieve software services (tools, web services) that can process datasets. Parallel with the RING evolution, the AIMS Vocabularies, Metadata Sets and Tool (VEST) registry (http://aims.fao.org/vest-registry) was also enhanced as a catalog to retrieve datasets, metadata sets, tools and Knowledge Organization Systems (KOS) used in the broader agricultural information management community. Last but not least, during the lifetime of the project, project partners registered their own datasets or collections in the RING enhancing the number of available datasets especially for the case of bibliographic, educational, germplasm and soil datasets.\n\nThe agINFRA project specifically aimed to create an environment to support data interoperability. One of the main outcomes of the data interoperability work was the generic aggregation workflow and the common metadata model. As it has been show cased in the project, a very well designed, domain specific aggregation workflow can be used to aggregate data from different sources. This generic version of the aggregation workflow was used and tested under the prism of the agINFRA services (i.e. AGRIS, Green Learning Network - GLN). As it was specified, part of the workflow was the transformation of the aggregated metadata records to an internal format. The usage of such format allows easily managing and re-using the data. But the work around the data interoperability didn’t stop in the aggregation of metadata. During the agINFRA project a lot of effort was made on the implementation of a Linked Open Data layer, especially for the case of germplasm and soilmaps data (http://vocabularies.aginfra.eu). The work done in the agINFRA project set the first steps for the Global agricultural Concept Scheme (GACS) project that aims to create a hub for thesauri in the agricultural field, in multiple languages, for use in Linked Data, a project supporter by The Food and Agricultural Organization of the United Nations (FAO) responsible for the AGROVOC Concept Scheme, CAB International (CABI) responsible for the CAB Thesaurus (CABT), and the National Agricultural Library of the USA (NAL) managing the NAL Thesaurus.\n\nOne of the obligations of the agINFRA project was to participate in the FP7 pilot program that requires all project publications to be openly available and delivered in OpenAIRE2. OpenAIRE (https://www.openaire.eu), a three-year project, establish the infrastructure for researchers to support them in complying with the European Commission Open Access pilot and the European Resuscitation Council Guidelines on Open Access. It delivers an electronic infrastructure and supporting mechanisms for the identification, deposition, access, and monitoring of FP7 and ERC funded articles. With the pilot program during FP7 becoming mandatory for all projects funded under the Horizon 2020 framework, agINFRA saw the need and opportunity to provide its data management services for projects and researchers in the agri-food community. Hence, in collaboration with OpenAIRE, the thematic agINFRA European (EU) Node was created as an aggregator that will collect all information for the agri-food sector and supply it to OpenAIRE relieving its contributors from the duty from registering themselves. To support the agINFRA EU Node, the agINFRA project in collaboration and with the support of AIMS and CIARD introduced a set of recommendations applying to agri-food research community for data management, sharing and dissemination aiming among others to provide a framework for the research community of European agri-food research institutions that need to follow the H2020 Open Access mandate and share their metadata. These recommendations are: 1) Metadata repository for data sources and publications following international metadata standards and classification schemes:\n\nSpecific metadata standard (e.g. AGRIS metadata application profile, Dublin Core) describing publications (e.g. author, abstract, keywords);\n\nSpecific metadata standard (e.g. CERIF: Common European Research Information Format) describing datasets (e.g. processed data of experiments);\n\nSpecific classification schemes according to the needs of each research community (e.g. AGROVOC terms (http://aims.fao.org/agrovoc)\n\n2) Repository registration in an agricultural aggregator (like the CIARD RING: http://ring.ciard.net) 3) Institutions’ researchers’ profiles and/or Institutions’ profiles need to be publicly available and accessibly by the agri-food community (using tools like registration in the AgriVIVO: http://www.agrivivo.net or the future AgriProfiles) 4) Data and publications accompanied by licenses using an international common standard (like Creative Commons: http://creativecommons.org) 5) Sharing the (metadata) descriptions of institutions data/publications with the agri-food research community (like the European AGRIS node “agINFRA”: http://aginfra.eu, OpenAIRE: https://www.openaire.eu). 6) Storing and preserving metadata over a data infrastructure (e.g. agINFRA e-infrastructure), provided with persistent identifiers compliant with known standards (e.g EUDAT or DataCite recommendations).\n\nAfter the kick-off meeting of the agINFRA project, it was obvious that the expectations were set very high. The agINFRA project and the work performed quickly got in the spotlight of the EC and global initiatives as the Research Data Alliance (RDA; https://rd-alliance.org). Up to now, the agINFRA project has been invited in a number of EC events (e.g. European Union Delegation to Australia on Research Infrastructure) and has an important role in the RDA leading the Interest Group on Agricultural Data and supporting the Wheat Data Interoperability Working Group. In the meantime, the movement of open data for the agricultural domain was spreading with a highlight event the ‘G-8 International Conference on Open Data for Agriculture’ that took place in Washington, D.C., US, on April 2013, just in the middle of the agINFRA project. Following this event, the Global Open Data for Agriculture and Nutrition (GODAN; http://www.godan.info) initiative was launched in October 2013. GODAN is an initiative that seeks to support global efforts to make data relevant to agriculture and nutrition available, accessible, and usable for unrestricted use worldwide. The initiative focuses on building high-level policy and public and private institutional support for open data, encouraging collaboration and cooperation among existing agriculture and open data activities, without duplication, bringing together all stakeholders to solve long-standing global problems. The agINFRA project was among the first to support GODAN and both the project and the project’s partners are actively involved in the initiative. The agINFRA services and outcomes in many times have set the pace for the GODAN strategy towards agricultural open data. In collaboration with CIARD, agINFRA supported the GODAN initiative and collaborated for the organization of its annual meetings and workshops.\n\nagINFRA provides data infrastructure and services for agricultural and related scientific communities, relevant data repositories in Europe, as well as virtual collaboration with researchers and repositories from other countries around the globe. This effort was realized with a European approach, involving partners from EU member states and European and international initiatives. agINFRA had identified a strategy to deliver socio-economic impacts across Europe. Towards this end, two different aspects of agINFRA work implemented this strategy: the open availability of services to create new products and the organisation of hackathons events to train users and create new opportunities2. Firstly all services created and data managed by agINFRA are openly shared with the community so anyone can use them. Today business models around open data and open services and are more and more common. In agINFRA one of the partners, Agro-Know, took advantage of the provided technology to create and launch a new product, the Agro-Know Stem (http://www.akstem.com), a solution that help users open, track, monitor and disseminate their research data. The Agro-Know Stem makes use of many of the agINFRA technologies (e.g. CIARD-RING, micro-finders, Grid powered aggregation workflows, etc.) and experience (e.g. metadata formats, vocabularies) to provide a high end product to their clients (Figure 3).\n\nSource: Agro-Know Stem website (www.akstem.com).\n\nSimilarly and in order to stimulate the community for the creation of new products/applications, the agINFRA project organized, sponsored or supported a number of hackathons events around the world. A hackathon is an event, which aims to use datasets from various sources (agricultural sciences in the case of agINFRA) in order to provide alternative, useful applications or enhance existing ones with new functionalities. It brings together web/software developers, domain and usability experts who are interested in applying their technical skills in the agricultural context and use the data sets suggested by the organizers in order to develop useful applications. Typical hackathons focus on the provision of working prototypes for one or more problems proposed by the organizers. The outcomes from hackathon events are propriety of the participants who are up to them to use this chance to evolve their idea to actual products. In the context of the agINFRA hackathons, new ideas and applications were conceptualized with some of them aiming to become products (e.g. the Skalidi project, an outcome of the Athens Green City Hackathon).", "appendix": "Author contributions\n\n\n\nAD authored the first draft of the paper, VP critically revised it for content and NM was the main contributor to the vision of the project. All authors approved the final manuscript for publication.\n\n\nCompeting interests\n\n\n\nAndreas Drakos is a consulting project manager responsible for the coordination of the agINFRA project. Vassilis Protonotarios is a consulting researcher, one of the main contributors in the agINFRA project work on data interoperability. Nikos Manouselis is the CEO of Agro-Know and technical leader of the agINFRA project.\n\n\nGrant information\n\nThis work is funded with the support by European Commission and more specifically the FP7 project agINFRA “A data infrastructure to support agricultural scientific communities” (http://aginfra.eu), which is funded by the schema “Combination of Collaborative Project and Coordination and Support Action: Integrated Infrastructure Initiative (I3)” under the work programme topic “INFRA-2011-1.2.2: data infrastructures for e-Science”. This publication reflects the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein.\n\nI confirm that the 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 all the agINFRA consortium partners for their contribution in the developments of the outcomes described in this publication.\n\n\nReferences\n\nGavrilut L, Thanopoulos C, Drakos A, et al.: Deliverable D1.3.3 agINFRA Scientific Vision White Paper Part A: of the agINFRA project. 2015. Reference Source\n\nManouselis N, Drakos A, Sicilia MA: Deliverable D1.3.3 \"agINFRA Scientific Vision White Paper Part B\" of the agINFRA project\". 2015. Reference Source\n\nCave B, Pesce V, Geser G, et al.: Deliverable D8.7.2 \"Sustainability Plan and Perspectives\" of the agINFRA project\". 2015. Reference Source" }	[ { "id": "8763", "date": "03 Jun 2015", "name": "George Adamides", "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\nIn general I find the article very interesting and its tackling a very interesting topic sharing open scientific data.The article is well organized and provides sufficient information on the goals and outcomes of the agInfra project. I would expect to see more on related work and other background information related to the sharing of agricultural information among scientist. For example what other practices exist and how the agINFRA project outcomes complement or differ from other sharing tools for scientific research (not necessarily only in agriculture).", "responses": [ { "c_id": "1419", "date": "09 Jun 2015", "name": "Andreas Drakos", "role": "Author Response", "response": "Dear George,thank you very much for your comments. The main concept of the agINFRA project was to re-use main infrastructure elements, connect with existing generic solutions and create something specified for the agricultural research domain. While in this article we tried to summarize the work of agINFRA,the specific aspects of each element has been presented in the numerous publications made from the partners (they can all be found at http://publications.aginfra.eu/ ).The agINFRA vision was to improve research collaboration and data exchange, but during the lifetime of the project, this had to be limited in creating the necessary building blocks (e-infrastructure) for services to built upon (as for example the new AGRIS). Hopefully this effort will continue and in the future we will be able to present a number of different end-user services specialized for the agricultural research community." }, { "c_id": "1420", "date": "09 Jun 2015", "name": "Vassilis Protonotarios", "role": "Author Response", "response": "Dear George,We would like to thank you for your comments. As Andreas mentioned earlier, the agINFRA project aimed at reusing existing components and provide the necessary interconnections or missing links in order to provide meaningful services that would facilitate agricultural researchers in conducting, publishing and disseminating their research.Despite the fact that we could indeed elaborate on the existing similar efforts, previous experiences and use cases, we decided to focus on the agINFRA approach and how potential end-users could benefit from that. However, based on your idea, we could work on an additional article that will focus on the background of the agINFRA project and the services that the project made use of." } ] } ]	1	https://f1000research.com/articles/4-127
https://f1000research.com/articles/4-152/v1	15 Jun 15	{ "type": "Study Protocol", "title": "Study Protocol: The influence of Running Therapy on executive functions and sleep of prisoners", "authors": [ "Jesse Meijers", "Joke Harte", "Gerben Meynen", "Pim Cuijpers", "Joke Harte", "Gerben Meynen", "Pim Cuijpers" ], "abstract": "Background: Executive dysfunction appears to be related to increased recidivism. Of note is that sleep disturbances, which are highly prevalent in prisons, may attenuate executive functions. Thus, improving executive functions, either directly or indirectly through the improvement of sleep, may reduce recidivism. It is hypothesised that physical exercise, in the form of Running Therapy, has a direct positive effect on executive functions as well as an indirect effect through the improvement of sleep.Methods/Design: Seventy two (N = 72) detainees in various penitentiary institutions in the Netherlands will be recruited in this study. A baseline measurement, including six neuropsychological tests of the Cambridge Neuropsychological Test Automated Battery (CANTAB), an assessment of sleep quality and duration using the Actiwatch (Actiwatch 2, Philips Respironics, Murrysville, PA, USA) and various other measurements will be administered before the start of the treatment. After 3 months of Running Therapy, participants will be assessed again with the same tests for neuropsychological and physical functioning. Primary outcomes are executive functioning and various sleep variables.Discussion: This study will be the first to investigate the possible influence of Running Therapy on the cognitive functioning, sleep and aggression in prisoners.", "keywords": [ "prison", "offenders", "executive functions", "CANTAB", "running therapy", "physical activity", "Actiwatch", "sleep" ], "content": "Background\n\nInternational studies report that 35 to 67 percent of released prisoners are detained for reoffending within approximately 2 to 3 years1–4. These percentages show that the reduction of recidivism is of great importance to society, since crime carries a great (financial) burden5.\n\nThe risk of reoffending appears to be negatively related to executive functioning6–8. Executive functions are higher order cognitive functions including planning, working memory, taking initiatives, set-shifting, attention, and impulse control9,10, and are crucial for self-regulation11. Planning and goal-directed behaviour are essential for successful re-entry into society, as ex-prisoners face complex challenges such as finding housing and employment12. Another important executive function is impulse control, which enables us to regulate and suppress aggressive behaviour, for example9; reduced impulse control thus increases the risk of aggressive behaviour.\n\nExecutive functions may be improved by physical activity13,14. For example, brisk walking was found to improve impulse control in older adults14. An increase in impulse control after participation in a physical activity programme is quite a consistent finding13,15,16 and appears especially effective in sedentary people13. Prison life consists mostly of passive leisure activities such as watching television17, and physical inactivity is a hallmark of prison life in various countries18–21. So, a large percentage of the prison population has a sedentary lifestyle.\n\nClosely related to both executive functioning and physical (in)activity is sleep. Sleep disturbances may diminish executive functioning22, while increased physical activity may improve sleep23,24. Sleep disturbances are highly common in prison and are responsible for a relatively large part of prison health care use25,26.\n\nBesides sleep disturbances, various psychiatric disorders are also highly prevalent in prisons, such as depression27, anxiety disorders and ADHD28–30. These psychiatric disorders may also negatively affect sleep31. Exercise, e.g. Running Therapy, may indeed reduce symptoms of depression and anxiety32 and behavioural symptoms of ADHD33,34.\n\nIn sum, physical activity may positively influence executive functions directly, but also indirectly, through improved sleep. We therefore hypothesize that physical exercise, in the form of supervised Running Therapy, will improve executive functioning and sleep, and will reduce aggression in prisoners. Although a randomized controlled trial (RCT) would be the best option to test this hypothesis, our current study could be considered a pilot behavioural intervention trial. Conducting an RCT directly is not possible, as the participating prisons were currently not willing to withhold treatment for some prisoners (i.e. the control group). However, this pilot study may eventually lead to an opportunity to conduct an RCT.\n\n\nMethods and analysis\n\nThis study concerns a prospective cohort study, measuring neuropsychological performance and sleep of detainees before and after receiving 3 months of Running Therapy. No further experimental intervention will take place, and this study does not interfere with or alter daily programs, treatments or any other environmental factors.\n\nPrisoners (male adults) of two Penitentiary Institutions in the Netherlands (i.e. PI Leeuwarden and PI Ter Apel) that are referred to Running Therapy by the psycho-medical staff will be recruited in this study. Various complaints or disorders may result in referral to Running Therapy, e.g. ADHD, sleep disturbances, anxiety disorders and depression. All participants in Running Therapy who speak sufficient Dutch or English will be approached and asked to participate in this study. In addition, we will attempt to approach those who do not speak Dutch or English by using a translation phone service often used in Penitentiary Institutions by the psycho-medical staff. As we are not allowed to interfere with daily practice in prison, participants will continue to receive treatment-as-usual from the prison care, which may entail changes in drug regimen or participation in other interventions. Participants may be excluded from Running Therapy when they sustain an injury or when they behave aggressively. Such decisions are made by the running therapist in concordance with the prison staff.\n\nRunning therapy is already part of the regular care in the Penitentiary Institutions where this study will be conducted. Prisoners are referred to Running Therapy by the psycho-medical staff, which consists of the psychologist, psychiatrist, physician and nursing staff. Of note is that this study does not make any modifications to the Running Therapy, and studies the intervention as is.\n\nRunning Therapy. This aerobic exercise group-therapy consists of supervised running for 2 days per week and 1 day per week of unsupervised running. During the therapy, the therapist encourages participants to run at a moderate pace and keeps track of the goals and sub-goals of the participants. The main goal of the therapy is to be able to run consecutively for 30 minutes at a moderate pace, three times per week, which will be achieved after 15 weeks of Running Therapy (see Appendix A for the complete program). Of note is that the therapy does not solely consist of running, but also contains a social aspect (running in a group, talking with participants or the therapist) and a reward aspect (achieving sub goals).\n\nThe Running Therapy is currently offered in two Penitentiary Institutions in the Netherlands, PI Ter Apel and PI Leeuwarden, and the study will take place in both these institutions.\n\nParticipants will undergo two measurements: a baseline measurement before starting with Running Therapy and a post measurement after 3 months of Running Therapy. The baseline measurement takes place over a single day and will take approximately 90 minutes in total. At the baseline measurement, anamnesis takes place and the participant will be assessed with a neuropsychological test battery, the Cambridge Neuropsychological Test Automated Battery (CANTAB). In addition, participants are given four questionnaires (see “Secondary Outcome Measures” for more information) to fill out in their own time, and are instructed to bring these with them to the first day of Running Therapy. Optionally, an Actiwatch (Actiwatch 2, Philips Respironics, Murrysville, PA, USA) will be handed to the participant, which will be worn for 7 consecutive days (for more information about the Actiwatch, refer to “Primary outcome measures – Sleep”.\n\nAfter three months of participating in Running Therapy, the participant will undergo the same tests as taken when assessing the baseline measurement. The four questionnaires and the Actiwatch will again be handed out to the participant.\n\nAccording to the policy of the Dutch Custodial Institutions Agency (DJI), we are not allowed to provide the participant with an incentive. However, the measurements will be planned during the moments in which participants are usually spending time in their cell. We have experienced that prisoners regard the measurements as a welcome change.\n\nExecutive functions\n\nThe CANTAB35 is a computerised neuropsychological test battery and is used to assess the executive functions of the participants. A 12.1 inch touchscreen tablet is used to administer the test. In two tests, a two-buttoned press pad is used. Few studies have assessed the reliability and validity of the CANTAB. One older study, that included some, but not all tests included in the current study, showed that test-retest correlations are above 0.6 for most of the subtests36. The main reason that we chose this battery is a practical one; neuropsychologically testing a prisoner can sometimes be logistically challenging. Using this highly portable test battery, that can be used in any room with a table and two chairs, provides us with the necessary flexibility to test participants anywhere the prison staff wants us to.\n\nThe following six CANTAB tests will be administered.\n\nSOC. Stockings of Cambridge measures planning, and is similar to the commonly used Tower of London (TOL). The main outcome measure is the number of problems solved in the minimum number of moves.\n\nSWM. Spatial Working Memory measures the ability to retain and manipulate information in the working memory and heuristic strategy. The main outcome measure is total number of errors made.\n\nSST. The Stop Signal Task measures response inhibition. The main outcome measure is SSRT, which is calculated by subtracting the mean stop-signal delay (the time between the stimulus and the beep) from the median go reaction time (the median response time on trials without a beep).\n\nIED. Intra-Extra Dimensional Set-Shift measures set-shifting and is similar to the commonly used Wisconsin Card Sorting Test. The main outcome measure is the total number of errors (adjusted for the number of trials conducted).\n\nCRT. Choice Reaction Time is a simple reaction time assessment and is used to measure reaction time and attention. The main outcome measures are mean reaction time, and the difference in mean reaction time between the second and the first half.\n\nCGT. The Cambridge Gambling Task measures decision-making impulsivity and risk-taking behaviour. The main outcome variables are measures of risk-taking, risk-adjustment, quality of decision making, delay aversion and overall proportion bet.\n\nSleep\n\nActiwatch. The Actiwatch AW2 (Respironics, Philips) will be used to measure the rest-activity rhythm, specifically sleep. Actiwatches are small activity monitors that are worn on the wrist, like a wristwatch, 24 hours a day. The Actiwatch gathers data by measuring the amount and intensity of movements made within every 5 seconds (which is the chosen epoch length). The Actiwatch stores the movement data of each epoch separately, which is retrieved later using a reader connected to a PC. The device's wristwatch-like design reduces physical discomfort to a minimum.\n\nThe Actiwatch provides information on the (in)stability of the rest-activity rhythm from one day to another (Inter-daily Stability; IS), and on the fragmentation of the rest-activity rhythm within the day, i.e. changes from periods of rest to activity and vice versa (Intra-daily Variability; IV). It also indicates the difference between maximal activity and maximal rest (Relative Amplitude; RA), and provides data on the 10 most active hours (M10) and the 5 least active hours (L5; for detailed information about these measurements, see the following references37,38).\n\nSleep analysis software (Philips Actiware 6.0.4, Respironics Inc.) will be used in order to analyse sleep. Sleep analysis will produce variables such as time spent in bed, sleep efficiency, sleep onset latency and total sleep duration.\n\nSleep-wake diary. Sleep-wake diaries are used to assess several events/activities such as time of medication use, use of caffeine/nicotine, bed time, wake time and time spent exercising. The sleep-wake diary is mainly used to improve the accuracy of sleep analysis.\n\nSCL-90-R. The Symptom Checklist-90-Revised39 is used to assess a broad spectrum of complaints, such as pain, depression and hopelessness, and with the SCL-90-R, we are able to reliably assess clinically significant change40. Test-retest correlations, for a 10-week interval, were found to range from r = .68 to .8041.\n\nSDL. The Sleep Diagnosis List42 is a self-report questionnaire that consists of 75 statements related to sleep and symptoms of sleep disorders. It will be used as a subjective measure of sleep and to control for disorders associated with sleep disturbances. The SDL is based on the Sleep Diagnosis Questionnaire43,44 and has been validated in a large Dutch population with sleep disorders45.\n\nPSQI. The Pittsburgh Sleep Quality Index46 is a commonly used self-report questionnaire and is used to assess e.g. sleep quality and sleep duration. Internal consistency of the PSQI ranges from 0.70 to 0.80 (Cronbach’s alpha) and the PSQI is known to have a good construct validity47; PSQI scores are more highly correlated to sleep disturbances (r = .69 to .77) than to e.g. mood and depression (r = .22 to .65). In patients with primary insomnia, the PSQI has been shown to have a high test-retest reliability, with r = 0.8748.\n\nAVL. The Aggression Questionnaire49 is a self-report questionnaire that consists of 29 statements that are related to aggressive thoughts and behaviour. While internal consistency for the global aggression scale and three subcomponents are sufficient (Cronbach’s alpha > 0.7), the internal consistency of the verbal aggression subcomponent is insufficient (Cronbach’s alpha = 0.5). Test-retest correlations for all subcomponents and the total score are high, r > 0.76.\n\nActivities. Participants are asked to provide an estimate of their participation in activities such as sports, labour and outdoor time on a 4-point Likert scale (never, sometimes, often, always). In addition, participants are asked to estimate the average hours per day spent on activities such as watching television and reading.\n\nDemographic and control variables. At baseline, demographic variables such as age, level of education, type of crime, number of previous incarcerations and current medication use are gathered.\n\nFor the main research questions (i.e. does running therapy improve cognition and sleep?), a repeated measures AN(C)OVA will be conducted for the results on the CANTAB and for the various sleep variables, comparing the baseline results with the post-measurement results. The main analyses will be conducted according to the intention-to-treat principle. Missing data will be imputed using the multiple imputation function as provided in SPSS 21 (IBM Corp, Armonk, NY, USA).\n\nA sample size calculation was made using GPower version 3.1.350. Since no similar studies have been conducted, the effect size was set small-to-moderate (f = 0.15). The lowest test-retest correlation of one of the used CANTAB tests was used as input for the correlation between repeated measures (r = 0.6).\n\nIn sum, effect size was set to f = 0.15, with an alpha error probability of p = 0.05, power of 1-β = 0.8, 1 group, 2 moments of measurement and correlation among repeated measures of r = 0.6. In GPower, this resulted in a total sample size of N = 72.\n\nThis study protocol has been reviewed by CERCO (Committee of Ethics in studies of Law and Criminology), the ethical committee of the faculty of Law at VU University, which declared it saw no objections to the study. In addition, the accredited medical ethical committee of the VU medical centre provided an official declaration (reference number 2014.399) that this study does not need further medical ethical approval, because of the low burden and non-medical non-interventional nature of the study (i.e. the intervention is already part of the institutional care and therefore requires no additional ethical approval). This study has been submitted for registration in the Dutch Trial Register (Nederlands Trial Register); the identification number will be made available in a revised version of this article.\n\nInformation letters and Informed Consent forms will be translated in various languages, to ensure that participants can read them in their own language. Data will be coded using a chronological number in combination with an identifier for the institution; the first participant in PI Ter Apel, for example, will be coded as “TA_001”. All non-anonymous data, such as the Informed Consent forms will be stored in the Penitentiary Institutions, as required by the regulations of the Custodial Institutions Agency (DJI). Anonymous data will be stored and analysed at the VU University Amsterdam. Any researcher who works with the participants (i.e. any researcher who could violate the privacy of the participants) is obliged to sign a confidentiality agreement, as required by the regulations of DJI. Students of VU University who will work on this dataset, for example for a thesis project, are required to sign a confidentiality agreement provided by VU University.\n\nThe researchers of this study, as well as their students working on this project, will have unlimited access to the dataset. The dataset will be made available to colleagues and peer-reviewers upon request; official approval of DJI may be required beforehand. Variables containing sensitive information may be removed before sharing the dataset, such as date of birth. Participants will not be allowed to see the final data.\n\nThis study will be finished in the second half of 2016, and the results will be published in international peer reviewed scientific journals.\n\n\nDiscussion\n\nThe main purpose of this study will be to investigate the influence of Running Therapy on executive functioning, sleep and aggression of prisoners. To our knowledge, this will be the first study of its kind in a prison population, which may provide results that are relevant to prison administrations, policymakers, and prison clinicians.\n\nDifferent aspects of Running Therapy may have a positive influence on the prisoner population. Firstly, regular exercise may positively influence executive functions, in particular impulse control13,14, which could be of importance in reducing recidivism. Secondly, acute bouts of exercise may also improve executive functions such as attention, memory, reasoning and planning15,51, also making Running Therapy of added value in the short term. For example, it may be useful to plan Running Therapy sessions right before other activities or therapy sessions to improve attentional performance and increase participation of the prisoners. Furthermore, although investigating the influence of Running Therapy on depressive symptoms is a secondary objective of the current study, Running Therapy may have significant clinical impact on these symptoms as well. In general, aerobic exercise has a small-to-moderate effect on depressive symptoms32,52, which could improve cognition, sleep, and general well-being.\n\nAs antisocial personality disorder is highly prevalent in prison (65%)53, this study may also be the first to examine the effects of exercise on the cognition of people with antisocial personality disorder. As impulsivity and/or failure to plan ahead is a clinical hallmark of these patients54, improving executive functions such as impulse control or planning may be of added value in treatment.\n\nStudying an intervention such as Running Therapy in prison comes with challenges and considerable limitations. One such limitation is that it we will be conducting a Phase I/II study instead of an RCT, as the participating prisons were not ready to directly allow an RCT. A follow-up RCT would be needed to confirm any positive results. Another limitation is that rewards for participation are not allowed, making it more challenging for us to recruit participants. An example of a logistic issue is the difficulty in planning a meeting with a participant with sufficient time to administer all the tests and questionnaires. Our solution is to ask participants to fill out a number of the questionnaires in their own time and bring these with them to therapy a week later, which might result in a somewhat lower compliance rate for these questionnaires. Another limitation concerns the use of actigraphy to measure sleep. Polysomnography is considered the gold standard in sleep research, with actigraphy being the second most valid method. However, as it not within our possibilities to temporarily transport all the participants to a sleep research lab for polysomnographic research, for this study, we consider actigraphy to be the most suitable and valid method; a clear advantage of actigraphy is the possibility to study the participants in the environment they reside in. One limitation of actigraphy is that we know from previous experience in prison, a number of the participants choose not to wear the Actiwatch, as they fear it might affect their social status with their fellow inmates. These limitations are examples of the challenges one faces when conducting research in a prison environment. In our view, however, prisoners are an important population with great impact on society, and researchers should not be discouraged to conduct research in this setting. Instead, we should try to build expertise conducting research in this particular environment, and hopefully, the results of our study will help to pave the way to the possibility to conduct RCTs in prison in the future.\n\nIn sum, if the results are indeed in accordance with our hypotheses, Running Therapy may eventually prove to be a useful approach to improve executive function, and possibly reduce aggressive behaviour and psychiatric symptoms of depression, ADHD, sleep disturbances and anxiety in prisoners.", "appendix": "Author contributions\n\n\n\nJM conceived the idea of the study. JM is the main author of the manuscript and will collect the data. JH and GM contributed to all sections of the manuscript. PC also contributed to all sections of the manuscript, and is the supervisor of this project.\n\nAll authors have agreed to the final content of the article and to its submission for publication.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work is supported by a grant of the Phoolan Devi Institute, an interdisciplinary research institute that funds criminological-related research. This work is also supported by a grant of the Arnold Oosterbaan Hersenstichting, a fund for neuropsychological research. Stichting Koningsheide funded the CANTAB, for this study specifically. These three grants were assigned to the first author. Lastly, GGZ Drenthe provided funding for the Actiwatches that are used in this study.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nFederal Ministry of the Interior. Second Periodical Report on Crime and Crime Control in Germany. 2006. Reference Source\n\nLangan PA, Levin DJ: Recidivism of Prisoners Released in 1994. 2002; 15(1): 58–65. Publisher Full Text\n\nSpicer K, Glicksman A, Britain G: Adult reconviction: results from the 2001 cohort. 2004. Reference Source\n\nWartna BSJ, Tollenaar N, Blom M, et al.: Recidivebericht 2012. 2012. Reference Source\n\nMcCollister KE, French MT, Fang H: The cost of crime to society: new crime-specific estimates for policy and program evaluation. Drug Alcohol Depend. 2010; 108(1–2): 98–109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHancock M, Tapscott JL, Hoaken PN: Role of executive dysfunction in predicting frequency and severity of violence. Aggressive Behav. 2010; 36(5): 338–49. PubMed Abstract \| Publisher Full Text\n\nLangevin R, Curnoe S: Psychopathy, ADHD, and brain dysfunction as predictors of lifetime recidivism among sex offenders. Int J Offender Ther Comp Criminol. 2011; 55(1): 5–26. PubMed Abstract \| Publisher Full Text\n\nRoss EH, Hoaken PN: Executive cognitive functioning abilities of male first time and return Canadian federal inmates. Can J Criminol Crim. 2011; 53(4): 377–403. Publisher Full Text\n\nDiamond A: Executive functions. Annu Rev Psychol. 2013; 64: 135–68. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJurado MB, Rosselli M: The elusive nature of executive functions: a review of our current understanding. Neuropsychol Rev. 2007; 17(3): 213–33. PubMed Abstract \| Publisher Full Text\n\nHofmann W, Schmeichel BJ, Baddeley AD: Executive functions and self-regulation. Trends Cogn Sci. 2012; 16(3): 174–80. PubMed Abstract \| Publisher Full Text\n\nLuther JB, Reichert ES, Holloway ED, et al.: An exploration of community reentry needs and services for prisoners: a focus on care to limit return to high-risk behavior. AIDS Patient Care STDS. 2011; 25(8): 475–81. PubMed Abstract \| Publisher Full Text\n\nColcombe S, Kramer AF: Fitness effects on the cognitive function of older adults: a meta-analytic study. Psychol Sci. 2003; 14(2): 125–30. PubMed Abstract \| Publisher Full Text\n\nKramer AF, Hahn S, Cohen NJ, et al.: Ageing, fitness and neurocognitive function. Nature. 1999; 400(6743): 418–9. PubMed Abstract \| Publisher Full Text\n\nBudde H, Voelcker-Rehage C, Pietrabyk-Kendziorra S, et al.: Acute coordinative exercise improves attentional performance in adolescents. Neurosci Lett. 2008; 441(2): 219–23. PubMed Abstract \| Publisher Full Text\n\nTomporowski PD, Davis CL, Miller PH, et al.: Exercise and Children’s Intelligence, Cognition, and Academic Achievement. Educ Psychol Rev. 2008; 20(2): 111–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nElger BS: Prison life: television, sports, work, stress and insomnia in a remand prison. Int J Law Psychiatry. 2009; 32(2): 74–83. PubMed Abstract \| Publisher Full Text\n\nCashin A, Potter E, Butler T: The relationship between exercise and hopelessness in prison. J Psychiatr Ment Health Nurs. 2008; 15(1): 66–71. PubMed Abstract \| Publisher Full Text\n\nIreland JL, Culpin V: The relationship between sleeping problems and aggression, anger, and impulsivity in a population of juvenile and young offenders. J Adolesc Health. 2006; 38(6): 649–55. PubMed Abstract \| Publisher Full Text\n\nPlugge EH, Foster CE, Yudkin PL, et al.: Cardiovascular disease risk factors and women prisoners in the UK: the impact of imprisonment. Health Promot Int. 2009; 24(4): 334–43. PubMed Abstract \| Publisher Full Text\n\nYoung M, Waters B, Falconer T, et al.: Opportunities for health promotion in the Queensland women’s prison system. Aust N Z J Public Health. 2005; 29(4): 324–7. PubMed Abstract \| Publisher Full Text\n\nKamphuis J, Meerlo P, Koolhaas JM, et al.: Poor sleep as a potential causal factor in aggression and violence. Sleep Med. 2012; 13(4): 327–34. PubMed Abstract \| Publisher Full Text\n\nMorgan K: Daytime activity and risk factors for late-life insomnia. J Sleep Res. 2003; 12(3): 231–8. PubMed Abstract \| Publisher Full Text\n\nYoungstedt SD, Kline CE: Epidemiology of exercise and sleep. Sleep Biol Rhythms. 2006; 4(3): 215–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeron JM, Paulus D, Tonglet R, et al.: Substantial use of primary health care by prisoners: epidemiological description and possible explanations. J Epidemiol Community Health. 2005; 59(8): 651–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNesset MB, Rustad AB, Kjelsberg E, et al.: Health care help seeking behaviour among prisoners in Norway. BMC Health Serv Res. 2011; 11(1): 301. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFazel S, Seewald K: Severe mental illness in 33,588 prisoners worldwide: systematic review and meta-regression analysis. Br J Psychiatry. 2012; 200(5): 364–73. PubMed Abstract \| Publisher Full Text\n\nBulten E, Nijman H, van der Staak C: Psychiatric disorders and personality characteristics of prisoners at regular prison wards. Int J Law Psychiatry. 2009; 32(2): 115–9. PubMed Abstract \| Publisher Full Text\n\nRösler M, Retz W, Retz-Junginger P, et al.: Prevalence of attention deficit-/hyperactivity disorder (ADHD) and comorbid disorders in young male prison inmates. Eur Arch Psychiatry Clin Neurosci. 2004; 254(6): 365–71. PubMed Abstract \| Publisher Full Text\n\nGinsberg Y, Hirvikoski T, Lindefors N: Attention Deficit Hyperactivity Disorder (ADHD) among longer-term prison inmates is a prevalent, persistent and disabling disorder. BMC Psychiatry. 2010; 10(1): 112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSutton EL: Psychiatric disorders and sleep issues. Med Clin North Am. 2014; 98(5): 1123–43. PubMed Abstract \| Publisher Full Text\n\nWegner M, Helmich I, Machado S, et al.: Effects of exercise on anxiety and depression disorders: review of meta- analyses and neurobiological mechanisms. CNS Neurol Disord Drug Targets. 2014; 13(6): 1002–14. PubMed Abstract \| Publisher Full Text\n\nGapin JI, Labban JD, Etnier JL: The effects of physical activity on attention deficit hyperactivity disorder symptoms: the evidence. Prev Med. 2011; 52(Suppl 1): S70–S4. PubMed Abstract \| Publisher Full Text\n\nBerwid OG, Halperin JM: Emerging support for a role of exercise in attention-deficit/hyperactivity disorder intervention planning. Curr Psychiatry Rep. 2012; 14(5): 543–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSahakian B, Owen A: Computerized assessment in neuropsychiatry using CANTAB: discussion paper. J R Soc Med. 1992; 85(7): 399. PubMed Abstract \| Free Full Text\n\nLowe C, Rabbitt P: Test/re-test reliability of the CANTAB and ISPOCD neuropsychological batteries: theoretical and practical issues. Cambridge Neuropsychological Test Automated Battery. International Study of Post-Operative Cognitive Dysfunction. Neuropsychologia. 1998; 36(9): 915–23. PubMed Abstract \| Publisher Full Text\n\nVan Someren EJ, Swaab DF, Colenda CC, et al.: Bright light therapy: improved sensitivity to its effects on rest-activity rhythms in Alzheimer patients by application of nonparametric methods. Chronobiol Int. 1999; 16(4): 505–18. PubMed Abstract \| Publisher Full Text\n\nCarvalho-Bos SS, Riemersma-van der Lek RF, Waterhouse J, et al.: Strong association of the rest-activity rhythm with well-being in demented elderly women. Am J Geriatr Psychiatry. 2007; 15(2): 92–100. PubMed Abstract \| Publisher Full Text\n\nDerogatis LR: SCL-90-R: Symptom Checklist-90-R: administration, scoring, and procedures manual. NCS Pearson, 1996. Reference Source\n\nSchmitz N, Hartkamp N, Franke GH: Assessing clinically significant change: application to the SCL-90-R. Psychol Rep. 2000; 86(1): 263–74. PubMed Abstract \| Publisher Full Text\n\nDerogatis LR, Savitz KL: The SCL–90–R and Brief Symptom Inventory (BSI) in primary care. In: Maruish ME, editor. Handbook of psychological assessment in primary care settings. Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers. 2000; 297–334. Reference Source\n\nSweere Y, Kerkhof G, De Weerd A, et al.: The Sleep Diagnosis List (SDL): a questionnaire for screening general sleep disorders. Sleep-Wake Res Netherlands. 1998; 9: 117–21. Reference Source\n\nDouglass AB, Bornstein R, Nino-Murcia G, et al.: The Sleep Disorders Questionnaire. I: Creation and multivariate structure of SDQ. Sleep. 1994; 17(2): 160–7. PubMed Abstract\n\nSweere Y, Kerkhof GA, De Weerd AW, et al.: The validity of the Dutch Sleep Disorders Questionnaire (SDQ). J Psychosom Res. 1998; 45(6): 549–55. PubMed Abstract \| Publisher Full Text\n\nSweere Y, Kerkhof GA, de Weerd AW, et al.: De Slaap Diagnose Lijst (SDL), een Nederlandse vragenlijst voor het diagnosticeren van slaapstoornissen. [The Sleep Diagnosis Checklist, a Dutch questionnaire for the diagnosis of sleep disorders.]. Ned Tijdschr Psychol. 1999; 54(1): 47–53. Reference Source\n\nBuysse DJ, Reynolds III CF, Monk TH, et al.: The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989; 28(2): 193–213. PubMed Abstract \| Publisher Full Text\n\nCarpenter JS, Andrykowski MA: Psychometric evaluation of the Pittsburgh Sleep Quality Index. J Psychosom Res. 1998; 45(1): 5–13. PubMed Abstract \| Publisher Full Text\n\nBackhaus J, Junghanns K, Broocks A, et al.: Test-retest reliability and validity of the Pittsburgh Sleep Quality Index in primary insomnia. J Psychosom Res. 2002; 53(3): 737–40. PubMed Abstract \| Publisher Full Text\n\nMeesters C, Muris P, Bosma H, et al.: Psychometric evaluation of the Dutch version of the Aggression Questionnaire. Behav Res Ther. 1996; 34(10): 839–43. PubMed Abstract \| Publisher Full Text\n\nFaul F, Erdfelder E, Lang AG, et al.: G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007; 39(2): 175–91. PubMed Abstract \| Publisher Full Text\n\nNanda B, Balde J, Manjunatha S: The acute effects of a single bout of moderate-intensity aerobic exercise on cognitive functions in healthy adult males. J Clin Diagn Res. 2013; 7(9): 1883. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCooney GM, Dwan K, Greig CA, et al.: Exercise for depression. Cochrane Database Syst Rev. 2013; 9: CD004366. PubMed Abstract \| Publisher Full Text\n\nFazel S, Baillargeon J: The health of prisoners. Lancet. 2011; 377(9769): 956–65. PubMed Abstract \| Publisher Full Text\n\nAmerican Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed.). 2013. Publisher Full Text" }	[ { "id": "11147", "date": "13 Nov 2015", "name": "Joseph HR Maes", "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 research proposal describes the details of a pilot study to assess the effect of a running therapy on the quality of sleep and executive functioning (EF) in prison inmates. The plan is based on previous literature suggesting: 1) a negative relationship between executive functioning and aggression and chance of recidivism in offenders; 2) a positive relationship between physical activity and EF; 3) a positive association between physical activity and sleep quality; 4) a negative association between sleep disturbances and EF; 5) a positive association between imprisonment and sleep disturbances. Given that prison life is characterized by physical inactivity, the points 1‒5 suggest that a running therapy may improve EF either directly or indirectly by improving sleep quality. The improvement in EF in turn might imply a decrease in aggression and recidivism rate. The idea is to have the research embedded within ongoing practices in two penitentiary institutions in the Netherlands. Specifically, the study is proposed to be performed in prisoners that are referred to a running therapy by the psycho-medical staff of the institutions as part of regular treatment interventions. Basically, the main study involves performing pre- and post-running therapy measurements using a variety of instruments to assess aspects of EF and sleep.The basic idea behind the proposal seems to logically follow from the existing pieces of knowledge concerning the interrelationships between imprisonment, aggression, recidivism, physical activity, EF, and sleep. Moreover, the proposal is well-written and carefully addresses practical, ethical, and legal considerations, and may set the stage for future, more controlled research within the prison context. For these reasons, I approve the manuscript.However, the practical limitation of not being able to perform a randomized controlled trial automatically implicates the major methodological weakness of the proposed study. That is, the absence of a control group, not receiving the target intervention, prevents us from determining whether any differences found in EF indices between the two measurements are due to the running therapy as such or to a general practice or loss-of-novelty effect, or test-retest unreliability (measurement error). The various measures from the CANTAB may differ in their susceptibility to the latter type of non-treatment-related effects. For example, especially the IED test from the CANTAB may be particularly vulnerable for practice effects (participants may have become aware of the ‘trick’ of shifting relevant stimulus dimensions). Also with respect to potential changes in sleep parameters, the absence of a control group prevents us from unequivocally associate them with a running therapy effect. Given this weakness, the authors could try to get the most out of their data by considering the following methodological and analytical points.From the description under Participants, I understand that not all prisoners are eligible to take part in the running therapy. Would there be any objections by the participating prisons to ask to have these individuals also perform the EF tests, wearing the Actiwatch, and filling out the questionnaires? In this way, these participants could serve as a control group, being helpful in interpreting the cause of any changes observed in EF or sleep parameters in the ‘experimental’/running therapy group. Of course, such ‘control group’ would be far from ideal, probably not being matched on many relevant characteristics and potentially showing large baseline differences with the treatment group, but it would be better than not having a control group at all. There are various ways to try to diminish the influence of practice and test-retest reliability issues. Although each of the available techniques has its own problems and limitations, one could for example use the technique of dual basement assessment to decrease the influence of practice effects. Alternatively, one could use statistical means to try to minimize their influence (e.g., see Collie et al., 2002). The proposed statistical analysis does not (explicitly) address the issue of sleep mediating a possible effect of the running therapy on (parameters of) EF. Perhaps one could try to perform a formal mediation analysis, in which some index of the success of the running training (e.g., difference in mean speed or distance run in a given period of time at baseline and at the end of training) as predictor, some (compound) index of change in sleep quality as mediator, and some (compound) measure of EF improvement as criterion. In this way, one could also get (preliminary) information on causal relationships. For example, it could, at least theoretically, also be the case that sleep problems are caused by poor EF (e.g., not being able to inhibit rumination causes sleep problems), suggesting that improved EF mediates the association between the effect of the running therapy on sleep rather than the other way around.", "responses": [] }, { "id": "11146", "date": "17 Nov 2015", "name": "Thomas Kleinsorge", "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 present a protocol for a study-in-preparation that aims to investigate effects of Running Therapy on executive functions and sleep quality in prisoners. Study participants in two Dutch prisons will receive three months of Running Therapy consisting of supervised running for 2 days per week and 1 day per week of unsupervised running. Before starting with Running Therapy the prisoners will be assessed with a neuropsychological test battery examining executive functions (CANTAB) and will be given four questionnaires (and optionally an Actiwatch) measuring a broad spectrum of complaints, sleep characteristics and aggression. After three months of participating in Running Therapy, the participants will be re-assessed with the same tests and questionnaires. The application of Running Therapy is supposed to positively influence executive functioning, sleep and to reduce aggression as well as symptoms of depression, anxiety and behavioural symptoms of ADHD. General evaluation The authors propose an innovative study that tackles a highly relevant applied issue, namely to identify measures that possibly reduce recidivism among prisoners. As the authors are aware, this relevance in practical terms goes along with some limitations in terms of meeting scientific standards. These limitations are imposed by severe restrictions regarding the study design as well as the examination procedures. Because the investigators are not allowed to interfere with any daily practice in prison, the authors are obliged to resort to a pre-post design instead of conducting a randomized controlled trial. Furthermore, the authors are not allowed to control the referral of potential participants to Running Therapy but will include any participants referred to this therapy by the psycho-medical staff of the respective institution. Despite these limitations, the proposed study promises to yield important data that will hopefully pave the way to a more rigorous study of this issue. Specific comments One of the major problems associated with a pre-post design is that practice effects may be erroneously considered as treatment effects. Would it be possible to run a control group of not imprisoned subjects that is matched with respect to other relevant characteristics in order to estimate the magnitude of to-be expected practice effects? The authors discuss both direct and indirect routes by which Running Therapy may improve executive functions. Maybe it would be possible to distinguish among possible modes of action by running mediational analyses? The measures obtained with Actiwatch offer the opportunity to analyze effects of Running Therapy on sleep quality and other activity related measures in a temporally more fine-grained manner than the primary outcome measures related to executive functions, for example, by using multi-level analyses. Overall, given that prisons are a very hard to access field of observation, it should be worthwhile to make use of more than standard statistical procedures to gather as much information as possible from any data obtained in this setting.", "responses": [] } ]	1	https://f1000research.com/articles/4-152
https://f1000research.com/articles/4-151/v1	15 Jun 15	{ "type": "Review", "title": "Echocardiography in the evaluation of athletes", "authors": [ "Gonzalo Grazioli", "Maria Sanz", "Silvia Montserrat", "Bàrbara Vidal", "Marta Sitges", "Gonzalo Grazioli", "Maria Sanz", "Silvia Montserrat", "Bàrbara Vidal" ], "abstract": "Echocardiography is currently a widely available imaging technique that can provide useful data in the field of sports cardiology particularly in two areas: pre-participation screening and analysis of the cardiac adaptation induced by exercise.The application of pre-participation screening and especially, the type and number of used diagnostic tests remains controversial. Echocardiography has shown though, higher sensitivity and specificity as compared to the ECG, following a protocol adapted to athletes focused on ruling out the causes of sudden death and the most common disorders in this population. It is still a subject of controversy the actual cost of adding it, but depending on the type of sport, echocardiography might be cost-effective if added in the first line of examination.Regarding the evaluation of cardiac adaptation to training in athletes, echocardiography has proved to be useful in the differential diagnosis of diseases that can cause sudden death, analysing both the left ventricle (hypertrophy cardiomyopathy, dilated cardiomyopathy, left ventricle non compaction) and the right ventricle (arrhythmogenic right ventricular cardiomyopathy).The aim of this paper is to review the current knowledge and the clinical practical implications of it on the field of echocardiography when applied in sport cardiology areas.", "keywords": [ "echocardiography", "athletes", "pre-participation screening", "athlete’s heart" ], "content": "Contributions of echocardiography in athletes\n\nThe number of people practicing sport has increased about five fold over the past 30 years1. The benefits of sport practice in improving cardiovascular health are unquestionable2, but an increase in cardiovascular events has also been demonstrated during its practice3. As a consequence, the absolute number of people at risk of sudden cardiac death (SCD) during exercise is also increasing4. Sports activity is not a cause of the increased mortality per se; rather, it might act as a trigger of cardiac arrest in athletes with structural or electrical heart abnormalities that generating malignant arrhythmias. Thus, it seems reasonable that every effort should be made for early recognition of any disease that may put the athlete at risk, keeping in mind the perspective that inadequate disqualification of individuals might also pose a risk.\n\nTherefore, a pre-participation screening (PPS) protocol seems to be of interest. Consequently, the European Society of Cardiology5 has proposed an exam which emphasizes three points or steps: a) family and personal history, b) physical examination and c) 12-lead electrocardiogram (ECG). The ECG has demonstrated a 70% sensitivity to detect the most frequent causes of SCD in young athletes6,7. However, about a third of these athletes with an anomalous origin of coronary arteries, aortic diseases and incipient forms of cardiomyopathies will present with a normal ECG.\n\nThe echocardiogram might be a useful, non-invasive and accessible tool to increase sensitivity of screening8. Our group reported the echocardiographic findings among 2688 competitive athletes; most of the echocardiographic evaluations were normal and only 203 (7.5%) showed abnormalities9. Cessation of athletic activity was indicated in 4 athletes: 2 for hypertrophic cardiomyopathy (electrocardiography had shown changes that did not meet diagnostic criteria), 1 pectus excavatum with compression of the right ventricle, and 1 significant pulmonary valve stenosis; the other minor alterations in echocardiography (7.5% of the total population) did not entail cessation of athletic activity and only indicated periodic monitoring.\n\nAlthough rare, some cardiac structural changes can be missed on physical examination and electrocardiography; in contrast, they are easily recognized with echocardiography. These findings suggest the use of echocardiography in at least the first PPS of competitive athletes to improve the effectiveness of programs aimed at preventing SCD in athletes.\n\nCurrently, there is no consensus on what kind of echo scan has to be included in the PPS in order to detect the most prevalent cardiac abnormalities related to SCD. Some studies have suggested a quick 5 minute echocardiogram protocol10 while other studies have proposed longer protocols performing a complete echocardiogram11. In our group, we carry out the standard transthoracic echocardiographic views suggested by the European Society of Echocardiography12; we consider that the long-axis parasternal, the short-axis parasternal, the apical 4-chamber views and 2-chamber views, suprasternal and parasternal right view provide a high sensitivity to diagnose the most prevalent causes of SCD (summarized in Table 1).\n\nThe most common abnormalities detected in athlete’s echocardiograms can be divided into two different groups: physiological structural and functional cardiac adaptive changes that result in what is called the athlete’s heart, and echocardiographic signs of different cardiomyopathies that can induce SCD (Table 2).\n\nHypertrophic cardiomyopathy constitutes the leading cause of SCD in young athletes13. The ECG has demonstrated a high sensitivity in the diagnosis of this entity, but there is still around 10% of patients with hypertrophic cardiomyopathy, who have abnormal ECGs14–16. On the other hand, 9% of the athletes with mild adaptive left ventricular (LV) hypertrophy show pathological changes in the ECG17. In both situations an echo scan would help to achieve the correct diagnosis.\n\nThe anomalous origin of the coronary arteries was considered to be a rare cause of SCD in athletes, but nowadays it has been demonstrated that it can be related to up to half of the previously asymptomatic SCD cases15. It was the second cause of sudden death associated with sports in the largest register of SCD in athletes13. The resting ECG of these athletes is normal, so this entity, if asymptomatic, cannot be detected in a regular PPS based on anamnesis, physical examination and ECG. It is known that an echocardiogram performed by physicians with adequate training can differentiate coronary anomalies with high sensitivity18, and therefore echo is again a key tool to unmask these asymptomatic patients.\n\nAortic root diseases are an infrequent cause of SCD in young people17, although they seem to be a more prevalent cause of SCD in athletes19. Echocardiogram allows for the diagnosis as well as the follow up of these patients. Similarly, bicuspid aortic valve without significant functional abnormalities would not be diagnosed in a regular PPS, and again, the echocardiogram would allow for an early diagnosis and a proper follow up20.\n\nFinally, an echo scan is recommended in congenital heart disease patients, with special focus on ventricular morphology and function, pulmonary pressure and aortic diameters before starting an exercise protocol21.\n\n\nUsefulness of echocardiography in pre-participation screening\n\nTo evaluate the cost effectiveness of the addition of a complementary study to the standard PPS, we have to take into account three basic parameters: a) incidence of SCD related to sport practice in the population b) cost of the study; c) years of potential life saved.\n\nTo date, there is no consensus which is the real incidence of SCD related to sports practice, because it depends on the analyzed population. Studies in Italy22 and Israel23 have reported an incidence of 2 cases of sudden death per 100,000 athletes per year, while other studies in France have documented an incidence of < 1 per 100,000 athletes4. On the other hand, the cost to perform an echocardiography is significantly different between American countries7 and Europe24. Finally, the quantification of the years of potential life lost depends on which population is considered, from school age25 or all young adults up to master athletes engaged in sports26. The weight of each of these three factors (incidence of SCD, cost of studies, years of life saved) in a given population to determine the cost-effectiveness of adding echocardiography in PPS represent the challenge that we will have to face in the upcoming years.\n\nCurrently, it is still controversial if the inclusion of an ECG in the regular PPS is cost effective27 in the USA, while in Europe this recommendation was established more than 10 years ago5 and it has been adopted by the different sport committees28 and international federations29. Several studies have demonstrated the cost-effectiveness of including an ECG in the PPS7,30, but to date, only a small population study in school-aged athletes has analyzed the cost-effectiveness of adding an echocardiography in the screening31; the study showed that adding echocardiography increased both the cost and the sensitivity. In our opinion, the echocardiography provides a higher sensitivity of the PPS, especially in some special populations with greater amount of cardiac disorders described such as competitive athletes9, sports with high static component32 or long distance endurance athletes33. Recommendations for PPS and use of echocardiography used by our group according to the level of sport practice and training are summarized in Table 3.\n\nF&P: family and personal, PE: physical exam.\n\n\nDifferentiating physiologic adaptation from pathology\n\nThe structural and functional adaptive changes that the heart develops in response to exercise, classically called “athlete’s heart“, has intrigued clinicians and scientists for more than a century. In the 19th century, Henschen described for the first time sport induced cardiac enlargement by auscultation and percussion34. Seventy years after the first athlete’s electrocardiographic features were described35,36 and a few years after the first 2-dimensional echocardiography images showed the characteristic chamber enlargement and myocardial hypertrophy of the athlete’s heart. Finally, the current advanced echocardiography techniques and magnetic resonance imaging (MRI) have begun to clarify the mechanism involved in these athlete’s heart adaptive features. The study of the athlete’s heart is thus essential, not only to understand how cardiac adaptation contributes to improved athletic performance, but also to differentiate the athlete’s heart from important disease states which may share similar morphologic features. We briefly review the physiologic and morphologic features associated with athletic training and the keys to differentiate normal adaptive athlete’s heart features from mild or initial expression forms of cardiac diseases such as hypertrophy cardiomyopathy (HCM), dilated cardiomyopathy (DCM), left ventricle non compaction (LVNC) and arrhythmogenic right ventricular cardiomyopathy (ARVC).\n\n\nFactors influencing cardiac remodeling in athletes\n\nDifferent forms of exercise impose different loads on the cardiovascular system. Classically, two forms have been described according to their hemodynamic effect. Endurance exercise results in an increased cardiac output due to the rise in heart rate and stroke volume, reduced peripheral resistance and moderate increment in systemic blood pressure, leading to a volume overload. On the other hand, strength exercise is characterized by a maintained or a slightly increased cardiac output and peripheral vascular resistance, which results in increased blood pressure and thus an increased LV afterload. Cycling or running are examples of endurance exercise while weightlifting is an example of strength exercise, but there are also overlapped sports combining endurance and strength hemodynamic conditions in different proportions such as soccer or hockey. These different hemodynamic conditions will result in different cardiac adaptive structural and functional changes. Moreover, cardiac remodeling is not a continuous response to exercise; it is influenced by individual genetic factors, gender and race. Thus, a proper athlete’s evaluation should be individual and take into account these potential influencers.\n\n\nThe left ventricle\n\nEndurance exercise LV remodeling is typically described as LV chamber enlargement with increased wall thickness resulting in an eccentric LV hypertrophy, while strength remodeling is described as a thickening of the LV wall with a slight increase in the size of the LV cavity resulting in a concentric hypertrophy. This dichotomous view, first described by Morganroth et al.37 is currently controversial. A meta-analysis by Pluim et al.38 initially confirmed this model; in contrast, a recent meta-analysis by Utomi et al.33 did not find this classic concentric remodeling in strength athletes and only found a slight LV dilatation and similar LV wall thickness as in endurance athletes. An increase in LV wall thickness is a typical feature of the athlete’s heart, however it is usually minimal and within normal range. In a cohort of 947 elite athletes Pelliccia et al.39 found a LV wall thickness > than 13 mm in only 1.7% of the athletes. Sharma et al.40 in a cohort of 720 elite athletes also reported a low incidence with only 0.4% of subjects showing LV wall thickness > 12 mm among elite junior athletes. However, this small number of extreme cases of exercise-induced remodeling may be difficult to differentiate from mild forms of hypertrophic cardiomyopathy. Various studies have tried to find out the key to differentiate these two entities, but to date there is no pathognomonic sign available, and a combination of clinical and family history, electrocardiographic and echocardiographic features is recommended. The increase in LV wall thickness in athletes is an adaptation to increase stroke volume so it has to be accompanied by chamber enlargement. Thus, a LV end-diastolic diameter > 54 mm41, an increased LV volume and particularly LV volume/mass ratios by MRI42 have been proposed to differentiate athlete’s heart from disease41.\n\nRecent advances in echocardiographic techniques including Tissue Doppler Imaging (TDI) and Speckle Tracking Imaging (STI) permit an accurate assessment of the myocardial function, helping us to differentiate adaptation from disease. Numerous studies have demonstrated normal or even supranormal diastolic LV function in athletes43; instead, the pathological forms of LV hypertrophy are typically associated with an impaired diastolic dysfunction characterized by lower early diastolic mitral annulus velocity44. Thus, the evaluation of diastolic LV function by TDI is nowadays mandatory in the evaluation of LV hypertrophy16. Figure 1 illustrates this echocardiographic differential diagnosis. In cases where differential diagnosis is unclear, MRI is useful. MRI offers a more accurate assessment of LV wall thickness, cardiac volumes, and tissue composition. Furthermore, adding gadolinium for late enhancement, the presence and location of myocardial fibrosis can be determined.\n\nPatients with hypertrophic cardiomyopathy present smaller LV end-diastolic diameters, reduced radial strain values and reduced velocities of the mitral annulus as compared to athletes.\n\nAs previously described, LV cavity enlargement is part of the cardiac remodeling observed in response to exercise, but this LV dilatation is in most cases minimal and indexed LV cavity dimensions are below pathologic limits. However, in the selected population of endurance elite-athletes this LV remodeling can be extreme. In a study by Pelliccia et al.45 more than 10% of elite ultra-endurance athletes had LV cavity end-diastolic dimensions > 60 mm, simulating forms of DCM. LV systolic function is described to be in a normal range among athletes46, but again, studies including high intensity endurance athletes have revealed a slight LV systolic dysfunction with LV ejection fraction around 45–49%47. In these extreme cases, the new advanced echocardiographic techniques can also help us. Although slightly lower ejection fraction of the LV might be found, the adaptive cardiac remodeling shows normal or even supranormal values of strain and strain rate by TDI, and normal values of longitudinal strain assessed by STI48. In contrast, in DCM patients these values are reduced49. The effect of endurance training on ventricular deformation, torsion and untwisting rate needs further investigation, but promising findings report exercise-induced supranormal LV untwisting rates50, confirming again the physiological LV response to exercise.\n\nFinally, parallel to the improvement in echocardiographic techniques and image resolution, a surprising high prevalence of LV hypertrabeculation has been described in athletes. Gati et al.51 in a cohort of 1146 athletes studied by echocardiography, reported trabeculations in 20% of the athletes, and even more, around 8% fulfilled conventional criteria for the diagnosis of LV non-compaction cardiomyopathy; this prevalence raised to 13% when only black athletes were considered. LV non-compaction cardiomyopathy is a rare cardiomyopathy thought to be secondary to the arrest of normal myocardial development, resulting in multiple deep ventricular trabeculations52. This entity has a wide clinical expression from asymptomatic patients to advanced cases with three characteristic symptoms: heart failure, thromboembolic events and fatal arrhythmias53, and has indeed been related to exercise-related SCD in young athletes54. The mechanisms implicated in LV hypertrabeculation in athletes are still unknown, but the reported high prevalence suggests that it might be another expression of cardiac adaptation to increased preload and afterload influenced by genetic and ethnical factors51. Structural echocardiographic features that could help to differentiate cardiac adaptive remodeling from disease are: the location of trabeculations (apical region in LV non-compaction cardiomyopathy versus mid-cavity region in athletes) and the evidence of late enhancement in cardiac MRI following gadolinium in LV non-compaction cardiomyopathy54. Furthermore, LV non-compaction cardiomyopathy patients may have reduced systolic and diastolic function while athletes with LV hypertrabeculation normally have no systolic or diastolic dysfunction. In the few cases with slightly low LV ejection fraction, a normal or even supranormal increase in LV systolic function with exercise could help us to distinguish pathology from physiological adaptation51.\n\n\nThe right ventricle\n\nDuring exercise, both ventricles have to increase stroke volume in response to the increased cardiac output demanded during exercise. This workload imposes high stress to all myocardial structures, which seems to be especially important in the right ventricle (RV) that typically works at low pressures in physiological conditions55. Classically, the study of the athlete’s heart was focused on the LV; but in the last two decades, with the introduction of advanced echocardiographic techniques and MRI, the RV exercise-remodeling has started to be described. Structurally, endurance exercise has been related to RV enlargement, typically balanced with LV dilatation56. Functionally, high intensity endurance exercise has been related to lower global RV peak systolic longitudinal strain values at rest as compared to controls; the RV basal is the segment most affected in this change57. To date, whether these lower strain values are the result of myocardial damage58 or are only an adaptive response demonstrated by an increased reserve after exercise provocation59 is still controversial. So far, few studies have focused on the RV in strength athletes but such RV remodeling seems to be less pronounced60. Extreme RV remodeling cases in elite ultra-endurance athletes may be indistinguishable from arrhythmogenic right ventricular dysplasia (ARVD). ARVD is a desmosomal cardiomyopathy characterized by progressive adipose and fibrosis myocardial infiltration with potential bad prognosis and constitutes one of the most important causes of sudden death in young athletes13. Various studies have demonstrated more rapid disease progression in patients that practice moderate-high intensity exercise, making the differential diagnosis between disease and cardiac remodeling even more challenging61. As previously mentioned, endurance exercise can induce adaptive physiologic biventricular dilatation, where the ratio of LV/RV remains unchanged. On the other hand, a reduced LV/RV ratio could be a warning sign of underlying disease62. In addition, athlete’s RV remodeling has proved to be global as opposed to that observed in ARVD patients who show a disproportionate enlargement of the RV outflow tract. Functionally, no motion abnormalities have been described in athletes62 despite having a lower deformation in the basal segment of the RV; abnormal motion of the RV is essential data to fulfil the ARVD diagnostic criteria63. Figure 2 illustrates this echocardiographic differential diagnosis. Finally in doubtful cases, MRI can provide us an accurate structural and functional RV evaluation, distinguishing segments with dyskinesia, fibrosis or outflow tract microaneurysms64. However, to date, there is no single sign available to differentiate both entities and consequently, a combination of clinical and family history, electrocardiographic and echocardiographic features are recommended. Table 4 summarizes the main echocardiographic features used to differentiate athlete’s heart from early stages of myocardial disease.\n\nPatients with arrhythmogenic right ventricular dysplasia in early stage can present only mild RV dilatation but the relationship between LV and RV cavity tend to be less than 1 and the RVOT is at least mildly dilated and they present reduced RV global and segmental strain values as compared to athletes.\n\nLV GLS: Left ventricle Global Longitudinal Strain.\n\n\nConclusions\n\nIn summary, echocardiography is a useful imaging tool to detect underlying heart disease that may imply a risk for people practicing sport and at the same time is a non-expensive and non-invasive technique to evaluate cardiac adaptation to training. The challenge remains the diagnosis and differentiation of extreme adaptation to training that very much resembles early stages of some myocardial diseases. Recently developed tools to better quantify cardiac performance have improved this issue but still more knowledge on the pathophysiology of cardiac adaptation to training is needed to optimize the identification of subjects at risk for sudden death or irreversible cardiac damage.", "appendix": "Author contributions\n\n\n\nGonzalo Grazioli: reviewed the literature, prepared tables wrote the manuscript.\n\nMaria Sanz: reviewed the literature, prepared tables wrote the manuscript.\n\nSilvia Montserrat: reviewed the literature and critically revised the manuscript at all stages.\n\nBàrbara Vidal: reviewed the literature and critically revised the manuscript at all stages.\n\nMarta Sitges: coordinated and critically revised the manuscript at all stages.\n\nAll authors revised the manuscript and agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was partially funded by grants from Generalitat de Catalunya (FI-AGAUR 2014–2017 (RH 040991, M. Sanz) and Consell Catala de l’Esport 11/2014, and from the Spanish Government (Plan Nacional I+D+i, Ministerio de Innovación y Ciencia DEP 2010-20565, Intensificación Actividad Investigadora, Instituto de Salud Carlos III (M Sitges), Plan Nacional I+D, Ministerio de Economia y Competitividad DEP2013-44923-P).\n\n\nAcknowledgements\n\nTo the Mémora Group for supporting research into the prevention of sport-related sudden death.\n\n\nReferences\n\nO’Keefe JH, Schnohr P, Lavie CJ: The dose of running that best confers longevity. Heart. 2013; 99(8): 588–591. PubMed Abstract \| Publisher Full Text\n\nLee DC, Pate RR, Lavie CJ, et al.: Leisure-time running reduces all-cause and cardiovascular mortality risk. J Am Coll Cardiol. 2014; 64(5): 472–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim JH, Malhotra R, Chiampas G, et al.: Cardiac arrest during long-distance running races. N Engl J Med. 2012; 366(2): 130–40. PubMed Abstract \| Publisher Full Text\n\nMarijon E, Tafflet M, Celermajer DS, et al.: Sports-related sudden death in the general population. Circulation. 2011; 124(6): 672–81. PubMed Abstract \| Publisher Full Text\n\nCorrado D, Pelliccia A, Bjørnstad HH, et al.: Cardiovascular pre-participation screening of young competitive athletes for prevention of sudden death: proposal for a common European protocol. Consensus Statement of the Study Group of Sport Cardiology of the Working Group of Cardiac Rehabilitation and Exercise Physiology and the Working Group of Myocardial and Pericardial Diseases of the European Society of Cardiology. Eur Heart J. 2005; 26(5): 516–24. PubMed Abstract \| Publisher Full Text\n\nPerez M, Fonda H, Le VV, et al.: Adding an electrocardiogram to the pre-participation examination in competitive athletes: a Systematic Review. Curr Probl Cardiol. 2009; 34(12): 586–662. PubMed Abstract \| Publisher Full Text\n\nWheeler MT, Heidenreich PA, Froelicher VF, et al.: Cost-effectiveness of preparticipation screening for prevention of sudden cardiac death in young athletes. An Intenal Med. 2010; 152(5): 276–286. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRizzo M, Spataro A, Cecchetelli C, et al.: Structural cardiac disease diagnosed by echocardiography in asymptomatic young male soccer players: implications for pre-participation screening. Br J Sports Med. 2012; 46(5): 371–3. PubMed Abstract \| Publisher Full Text\n\nGrazioli G, Merino B, Montserrat S, et al.: Usefulness of echocardiography in preparticipation screening of competitive athletes. Rev Esp Cardiol (Engl Ed). 2014; 67(9): 701–705. PubMed Abstract \| Publisher Full Text\n\nWyman RA, Chiu RY, Rahko PS: The 5-minute screening echocardiogram for athletes. J Am Soc Echocardiogr. 2008; 21(7): 786–8. PubMed Abstract \| Publisher Full Text\n\nWeiner RB, Wang F, Hutter AM Jr, et al.: The feasibility, diagnostic yield, and learning curve of portable echocardiography for out-of-hospital cardiovascular disease screening. J Am Soc Echocardiogr. 2012; 25(5): 568–75. PubMed Abstract \| Publisher Full Text\n\nEvangelista A, Flachskampf F, Lancellotti P, et al.: European Association of Echocardiography recommendations for standardization of performance, digital storage and reporting of echocardiographic studies. Eur J Echocardiogr. 2008; 9(4): 438–48. PubMed Abstract \| Publisher Full Text\n\nMaron BJ, Doerer JJ, Haas TS, et al.: Sudden deaths in young competitive athletes: analysis of 1866 deaths in the United States, 1980-2006. Circulation. 2009; 119(8): 1085–92. PubMed Abstract \| Publisher Full Text\n\nGersh BJ, Maron BJ, Bonow RO, et al.: 2011 ACCF/AHA Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Developed in collaboration with the American Association for Thoracic Surgery, American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol. 2011; 58(25): e212–60. PubMed Abstract \| Publisher Full Text\n\nRowin EJ, Maron BJ, Appelbaum E, et al.: Significance of false negative electrocardiograms in preparticipation screening of athletes for hypertrophic cardiomyopathy. Am J Cardiol. 2012; 110(7): 1027–1032. PubMed Abstract \| Publisher Full Text\n\nElliott PM, Anastasaki sA, Borger MA, et al.: 2014 ESC Guidelines on diagnosis and management of hypertrophic cardiomyopathy: the Task Force for the Diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J. 2014; 35(39): 2733–79. PubMed Abstract \| Publisher Full Text\n\nFuller CM, McNulty CM, Spring D a, et al.: Prospective screening of 5,615 high school athletes for risk of sudden cardiac death. Med Sci Sports Exerc. 1997; 29(9): 1131–1138. PubMed Abstract \| Publisher Full Text\n\nLytrivi ID, Wong AH, Ko HH, et al.: Echocardiographic diagnosis of clinically silent congenital coronary artery anomalies. Int J Cardiol. 2008; 126(3): 386–393. PubMed Abstract \| Publisher Full Text\n\nPelliccia A, Di Paolo FM, De Blasiis E, et al.: Prevalence and clinical significance of aortic root dilation in highly trained competitive athletes. Circulation. 2010; 122(7): 698–706. PubMed Abstract \| Publisher Full Text\n\nStefani L, De Luca A, Maffulli N, et al.: Speckle tracking for left ventricle performance in young athletes with bicuspid aortic valve and mild aortic regurgitation. Eur J Echocardiogr. 2009; 10(4): 527–31. PubMed Abstract \| Publisher Full Text\n\nBudts W, Börjesson M, Chessa M, et al.: Physical activity in adolescents and adults with congenital heart defects: individualized exercise prescription. Eur Heart J. 2013; 34(47): 3669–74. PubMed Abstract \| Publisher Full Text\n\nCorrado D, Basso C, Pavei A, et al.: Trends in sudden cardiovascular death in young competitive athletes after implementation of a preparticipation screening program. JAMA. 2006; 296(13): 1593–1601. PubMed Abstract \| Publisher Full Text\n\nSteinvil A, Chundadze T, Zeltser D, et al.: Mandatory electrocardiographic screening of athletes to reduce their risk for sudden death proven fact or wishful thinking? J Am Coll Cardiol. 2011; 57(11): 1291–6. PubMed Abstract \| Publisher Full Text\n\nMenafoglio A, Di Valentino M, Segatto JM, et al.: Costs and yield of a 15-month preparticipation cardiovascular examination with ECG in 1070 young athletes in Switzerland: implications for routine ECG screening. Br J Sports Med. 2014; 48(15): 1157–61. PubMed Abstract \| Publisher Full Text\n\nMaron BJ, Haas TS, Ahluwalia A, et al.: Incidence of cardiovascular sudden deaths in Minnesota high school athletes. Heart Rhythm. 2013; 10(3): 374–7. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Baggish AL, Knuuti J, et al.: Cardiac imaging and stress testing asymptomatic athletes to identify those at risk of sudden cardiac death. JACC Cardiovasc Imaging. 2013; 6(9): 993–1007. PubMed Abstract \| Publisher Full Text\n\nMaron BJ, Friedman RA, Kligfield P, et al.: Assessment of the 12-Lead ECG as a Screening test for detection of cardiovascular disease in healthy general populations of young People (12-25 Years of Age): a Scientific Statement From the American Heart Association and the American College of Cardiology. Circulation. 2014; 130(15): 1303–34. PubMed Abstract \| Publisher Full Text\n\nLjungqvist A, Jenoure P, Engebretsen L, et al.: The International Olympic Committee (IOC) Consensus Statement on periodic health evaluation of elite athletes March 2009. Br J Sports Med. 2009; 43(9): 631–643. PubMed Abstract \| Publisher Full Text\n\nDvorak J, Kramer EB, Schmied CM, et al.: The FIFA medical emergency bag and FIFA 11 steps to prevent sudden cardiac death: setting a global standard and promoting consistent football field emergency care. Br J Sports Med. 2013; 47(18): 1199–202. PubMed Abstract \| Publisher Full Text\n\nFuller CM: Cost effectiveness analysis of screening of high school athletes for risk of sudden cardiac death. Med Sci Sports Exerc. 2000; 32(5): 887–90. PubMed Abstract\n\nKoch S, Cassel M, Linne K, et al.: ECG and echocardiographic findings in 10-15-year-old elite athletes. Eur J Prev Cardiol. 2014; 21(6): 774–781. PubMed Abstract \| Publisher Full Text\n\nSitges M, Gutiérrez JA, Brugada J, et al.: Consensus for the prevention of sudden cardiac death in athletes. Apunt Med Esport. 2013; 48(177): 35–41. Reference Source\n\nUtomi V, Oxborough D, Whyte GP, et al.: Systematic review and meta-analysis of training mode, imaging modality and body size influences on the morphology and function of the male athlete’s heart. Heart. 2013; 99(23): 1727–33. PubMed Abstract \| Publisher Full Text\n\nHenschen S: Skidlauf und Skidwettlauf. Eine medizinische Sportstudie. Mitt Med Klin Upsala Jena. 1899; 2: 15–18.\n\nHunt EA: Electrocardiographic study of 20 champion swimmers before and after 110-yard sprint swimming competition. Can Med Assoc J. 1963; 88: 1251–1253. PubMed Abstract \| Free Full Text\n\nVan Ganse W, Versee L, Eylenbosch W, et al.: The electrocardiogram of athletes. Comparison with untrained subjects. Br Heart J. 1970; 32(2): 160–164. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorganroth J, Maron BJ, Henry WL, et al.: Comparative left ventricular dimensions in trained athletes. Ann Intern Med. 1975; 82(4): 521–524. PubMed Abstract \| Publisher Full Text\n\nPluim BM, Zwinderman AH, van der Laarse A, et al.: The athlete's heart. A meta-analysis of cardiac structure and function. Circulation. 2000; 101(3): 336–344. PubMed Abstract \| Publisher Full Text\n\nPelliccia A, Maron BJ, Spataro A, et al.: The upper limit of physiologic cardiac hypertrophy in highly trained elite athletes. N Engl J Med. 1991; 324(5): 295–301. PubMed Abstract \| Publisher Full Text\n\nSharma S, Maron BJ, Whyte G, et al.: Physiologic limits of left ventricular hypertrophy in elite junior athletes: relevance to differential diagnosis of athlete’s heart and hypertrophic cardiomyopathy. J Am Coll Cardiol. 2002; 40(8): 1431–1436. PubMed Abstract \| Publisher Full Text\n\nCaselli S, Maron MS, Urbano-Moral JA, et al.: Differentiating left ventricular hypertrophy in athletes from that in patients with hypertrophic cardiomyopathy. Am J Cardiol. 2014; 114(9): 1383–9. PubMed Abstract \| Publisher Full Text\n\nLuijkx T, Cramer MJ, Buckens CF, et al.: Unravelling the grey zone: cardiac MRI volume to wall mass ratio to differentiate hypertrophic cardiomyopathy and the athlete’s heart. Br J Sports Med. 2013; 992: 1–7. PubMed Abstract \| Publisher Full Text\n\nPelà G, Bruschi G, Montagna L, et al.: Left and right ventricular adaptation assessed by Doppler tissue echocardiography in athletes. J Am Soc Echocardiogr. 2004; 17(3): 205–211. PubMed Abstract \| Publisher Full Text\n\nMatsumura Y, Elliott PM, Virdee MS, et al.: Left ventricular diastolic function assessed using Doppler tissue imaging in patients with hypertrophic cardiomyopathy: relation to symptoms and exercise capacity. Heart. 2002; 87(3): 247–251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPelliccia A, Culasso F, Di Paolo FM, et al.: Physiologic left ventricular cavity dilatation in elite athletes. Ann Intern Med. 1999; 130(1): 23–31. PubMed Abstract \| Publisher Full Text\n\nGilbert CA, Nutter DO, Felner JM, et al.: Echocardiographic study of cardiac dimensions and function in the endurance-trained athlete. Am J Cardiol. 1977; 40(4): 528–533. PubMed Abstract \| Publisher Full Text\n\nAbergel E, Chatellier G, Hagege AA, et al.: Serial left ventricular adaptations in world-class professional cyclists: Implications for disease screening and follow-up. J Am Coll Cardiol. 2004; 44(1): 144–149. PubMed Abstract \| Publisher Full Text\n\nBaggish AL, Yared K, Wang F, et al.: The impact of endurance exercise training on left ventricular systolic mechanics. Am J Physiol Hear Circ Physiol. 2008; 295(3): H1109–H1116. PubMed Abstract \| Publisher Full Text\n\nLakdawala NK, Thune JJ, Colan SD, et al.: Subtle abnormalities in contractile function are an early manifestation of sarcomere mutations in dilated cardiomyopathy. Circ Cardiovasc Genet. 2012; 5(5): 503–510. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNotomi Y, Martin-Miklovic MG, Oryszak SJ, et al.: Enhanced ventricular untwisting during exercise: a mechanistic manifestation of elastic recoil described by doppler tissue imaging. Circulation. 2006; 113(21): 2524–2533. PubMed Abstract \| Publisher Full Text\n\nGati S, Chandra N, Bennett RL, et al.: Increased left ventricular trabeculation in highly trained athletes: do we need more stringent criteria for the diagnosis of left ventricular non-compaction in athletes? Heart. 2013; 99(6): 401–8. PubMed Abstract \| Publisher Full Text\n\nPaterick TE, Tajik AJ: Left ventricular noncompaction: a diagnostically challenging cardiomyopathy. Circ J. 2012; 76(7): 1556–62. PubMed Abstract \| Publisher Full Text\n\nChin TK, Perloff JK, Williams RG, et al.: Isolated noncompaction of left ventricular myocardium. A study of eight cases. Circulation. 1990; 82(2): 507–513. PubMed Abstract \| Publisher Full Text\n\nWan J, Zhao S, Cheng H, et al.: Varied distributions of late gadolinium enhancement found among patients meeting cardiovascular magnetic resonance criteria for isolated left ventricular non-compaction. J Cardiovasc Magn Reson. 2013; 15(1): 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLa Gerche A, Heidbüchel H, Burns AT, et al.: Disproportionate exercise load and remodeling of the athlete’s right ventricle. Med Sci Sports Exerc. 2011; 43(6): 974–81. PubMed Abstract \| Publisher Full Text\n\nScharf M, Brem MH, Wilhelm M, et al.: Cardiac magnetic resonance assessment of left and right ventricular morphologic and functional adaptations in professional soccer players. Am Heart J. 2010; 159(5): 911–918. PubMed Abstract \| Publisher Full Text\n\nTeske AJ, Prakken NH, De Boeck BW, et al.: Echocardiographic tissue deformation imaging of right ventricular systolic function in endurance athletes. Eur Heart J. 2009; 30(8): 969–77. PubMed Abstract \| Publisher Full Text\n\nHeidbüchel H, Hoogsteen J, Fagard R, et al.: High prevalence of right ventricular involvement in endurance athletes with ventricular arrhythmias. Role of an electrophysiologic study in risk stratification. Eur Heart J. 2003; 24(16): 1473–1480. PubMed Abstract \| Publisher Full Text\n\nLa Gerche A, Burns A, D’Hooge J, et al.: Exercise Strain Rate Imaging Demonstrates Normal Right Ventricular Contractile Reserve and Clarifies Ambiguous Resting Measures in Endurance Athletes. J Am Soc Echocardiogr. 2012; 25(3): 253–62. PubMed Abstract \| Publisher Full Text\n\nPagourelias ED, Kouidi E, Efthimiadis GK, et al.: Right atrial and ventricular adaptations to training in male Caucasian athletes: an echocardiographic study. J Am Soc Echocardiogr. 2013; 26(11): 1344–1352. PubMed Abstract \| Publisher Full Text\n\nDewilde W, Boersma L, Delanote J, et al.: Symptomatic arrhythmogenic right ventricular dysplasia/cardiomyopathy. A two-centre retrospective study of 15 symptomatic ARVD/C cases and focus on the diagnostic value of MRI in symptomatic ARVD/C patients. Acta Cardiol. 2008; 63(2): 181–189. PubMed Abstract \| Publisher Full Text\n\nBauce B, Frigo G, Benini G, et al.: Differences and similarities between arrhythmogenic right ventricular cardiomyopathy and athlete’s heart adaptations. Br J Sports Med. 2010; 44(2): 148–154. PubMed Abstract \| Publisher Full Text\n\nMarcus FI, McKenna WJ, Sherrill D, et al.: Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the Task Force Criteria. Eur Heart J. 2010; 31(7): 806–814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nte Riele AS, Tandri H, Bluemke DA: Arrhythmogenic right ventricular cardiomyopathy (ARVC): cardiovascular magnetic resonance update. J Cardiovasc Magn Reson. 2014;16: 50. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "9040", "date": "26 Jun 2015", "name": "Andreas Müssigbrodt", "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 review summarizes some important aspects of echocardiography as a screening tool for the screening of athletes and should therefore be of interest for sports physicians and cardiologists. The potential advantages of this method and the technical aspects are well described.I would nevertheless recommend to mention potential problems too, mainly overdiagnosis and misdiagnosis with important consequences for the career of athletes and additional costs. In order to prevent overdiagnosis and misdiagnosis, intensive training of sports cardiologists seems to be of utmost importance.Furthermore, as research in this area often focuses on young, competitive athletes, I would recommend to discuss the potential value of echocardiography for older, non-competitive, recreational athletes.I recommend indexing after minor revision.", "responses": [] }, { "id": "9496", "date": "20 Jul 2015", "name": "Flavio D'Ascenzi", "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 paper by Grazioli and colleagues provides a comprehensive review on the current knowledge and the clinical implications of echocardiography when applied to the evaluation of athlete’s heart in the context of pre-participation screening. The paper is currently focused on right and left ventricles, is well written, and analyses not only standard echocardiography, but also the novel echocardiographic techniques applied to athlete’s heart, with a specific focus on the differential diagnosis between exercise-induced physiological remodelling and the diseases most commonly cause of sudden cardiac death. Major revisionsAlthough the clinical implications of biatrial analysis has to be understood, in my opinion a brief paragraph on the analysis of biatrial morphology and function could be potentially of interest, also considering the some authors have demonstrated its utility in the differential diagnosis between athlete’s heart and cardiomyopathies. An interesting point of using echocardiography in the context of a screening has not been analysed. Although some authors debate about the utility and the costs of echocardiography, in my opinion the fundamental question of the application of echocardiography to the pre-participation screening in athletes is WHEN performing this exam. Considering that a frequent echocardiographic evaluation is not feasible in low-risk populations and dramatically increases the cost in absence of real clinical advantages, one has to suppose that one or two exams can be performed during the career of a non-professional athlete. Although some congenital defects can be detected in the early phase of the life of an individual, some of the most common cause of sudden cardiac death are the genetic myocardial diseases, such as HCM, DCM, and ARVC, all known to have a delayed phenotype expression. Thus, performing an echocardiogram at the beginning of the athletic career of a child could detect congenital defects, but could be misleading in excluding the cardiomyopathies most commonly associated with sudden cardiac death, with a typical expression in the II-III (or even IV) decades of life of a young athlete. I think that the authors should analyse this clinical question and should provide a personal opinion based on the current available evidences and on the experience of their Center. Minor revisionsIn table 1 I think that “aortic coartation” should be included in the section “suprasternal view”. Table 2: exercise has not proven to be a determinant of a physiological dilation of the aorta; conversely, it has been demonstrated that aortic dimensions are comparable with those of the general population. Accordingly, I suggest removing “mild aortic dilation” from table 2, section “athlete’s heart”. Atherosclerotic coronary artery disease cannot be detected by echocardiography, although indirect signs such as left ventricular wall motion abnormalities can suggest its presence. However, considering the signs are uncommonly observed in athletes, I suggest to remove “Atherosclerotic coronary artery disease” from table 2, section “sudden cardiac death”. Left ventricular non compaction: “it might be another expression of cardiac adaptation to increased preload and afterload influenced by genetic and ethnical factor”: ref. #51 is not correct because does refer to an hypothesis rather than an explanation of the phenomenon. I suggest to cite the article by Gati S. et al (Circulation 2014) and the work by D’Ascenzi F. et al (Int J Cardiol 2015) both explaining the physiological determinants of benign hypertrabeculation in the athletes. Table 3: the definition “competitive athletes, high static and dynamic component endurance training, > 10 hours/week” is not clear: which is the opinion of the authors? Although it is clear that echocardiography is not recommended in subjects engaged in recreational sports, which are the athletes to be evaluated by echocardiography? All the competitive athletes? Athletes with abnormal findings at resting ECG or at history/physical examination? Athletes engaged in high-volume training programs?...Please specify. Table 4At the present time, LV circumferential and radial strains cannot be definitively suggested to distinguish between cardiomyopathies and athlete’s heart. LV longitudinal strain is highly sensitive for myocardial disorders and more reproducible than circumferential and radial strains ((Feigenbaum H, Circ J 2012). Because of the LV shortens from base to apex in systole, the fixed short-axis tracking that is required for circumferential and radial strain is more difficult than longitudinal tracking. Particularly, radial strain is intrinsically more vulnerable to measurement variability because strain assessed in these orientations represents an aggregate measure of endocardial, mesocardial and epicardial tissue motion, encompassing both myocardial fiber shortening and global LV torsion in systole. The best inter- and intra-observed reproducibility is observed for LV longitudinal strain and particularly data based on measurements of LV radial strain should be interpreted with caution (Cheng S, J Am Soc Echocardiogr 2013). A limited reliability has been observed also for LV circumferential strain measurements. A possible explanation is that circumferential deformation twist can hardly be assessed in the same view, because the base descends to the apex. This movement makes the correct measurement in the short axis view impossible in two different time-points or even in two different days.Finally, LV longitudinal strain measurements are known to be more robust than radial ones (Mor-Avi, Eur J Echocardiogr 2011). According to these considerations and to the fact that LV longitudinal strain is the only form of strain that is used routinely (Feigenbaum H, Circ J 2012), I suggest to remove both radial and circumferential strains from table 4 and to stress in the text that future studies are needed to evaluate whether these deformation parameters could be used in the clinical practice. “motion abnormalities” -> please correct with “wall motion abnormalities” Table 4: according to the current available data on deformation analysis applied to the athlete’s heart, I suggest to describe LV GLS at rest as “normal” and not “normal or slightly reduced”.", "responses": [] }, { "id": "9498", "date": "27 Jul 2015", "name": "Roman Leischik", "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\nEchocardiography is an informative method for examination of cardiac structures and not dispensable in examination of athletes. Dr Grazioli focused his paper on analysis of left and right ventricle and did it well. Right and left atrial dimensions1-3 (and function3, 4), aortic dimension5, 6 (and stiffness5, 7) or pulmonary dimensions and function8 were not focused in this initial article of the sports cardiology channel because it is problematic to start with and to write a complete update with the burden of complete echocardiographic potential/possibilities. In one of his own papers he described alterations of aortic valve/pulmonary valve/mitral valve in athletes and aortic root dilatation or severe pulmonary stenosis9.Echocardiography was recommended for screening of athletes under different circumstances10, 13 and particularly for follow up14, 15 examinations. The possibilities of strain echocardiography were discussed in the past16, 17 and we look forward to new. The global longitudinal strain has better reproducibility18, the potential of radial strain seems to be limited in examination of training effects19.\n\nI recommend indexing of this article with the option to write an update with further possibilities and importance of echocardiography in the examinations of atrial function/valve and aortic function including 3D echocardiography and the further analysis of recreational athletes20 and importance of follow-up examinations21, 22. The advantage of this sports cardiology channel is the possibility for updating papers with new literature and information and to actualize the growing knowledge in sports cardiology.", "responses": [] } ]	1	https://f1000research.com/articles/4-151
https://f1000research.com/articles/4-118/v1	13 May 15	{ "type": "Research Article", "title": "Promiscuity progression of bioactive compounds over time", "authors": [ "Ye Hu", "Swarit Jasial", "Jürgen Bajorath", "Ye Hu", "Swarit Jasial" ], "abstract": "In the context of polypharmacology, compound promiscuity is rationalized as the ability of small molecules to specifically interact with multiple targets. To study promiscuity progression of bioactive compounds in detail, nearly 1 million compounds and more than 5.2 million activity records were analyzed. Compound sets were assembled by applying different data confidence criteria and selecting compounds with activity histories over many years. On the basis of release dates, compounds and activity records were organized on a time course, which ultimately enabled monitoring data growth and promiscuity progression over nearly 40 years, beginning in 1976. Surprisingly low degrees of promiscuity were consistently detected for all compound sets and there were only small increases in promiscuity over time. In fact, most compounds had a constant degree of promiscuity, including compounds with an activity history of 10 or 20 years. Moreover, during periods of massive data growth, beginning in 2007, promiscuity degrees also remained constant or displayed only minor increases, depending on the activity data confidence levels. Considering high-confidence data, bioactive compounds currently interact with 1.5 targets on average, regardless of their origins, and display essentially constant degrees of promiscuity over time. Taken together, our findings provide expectation values for promiscuity progression and magnitudes among bioactive compounds as activity data further grow.", "keywords": [ "Polypharmacology", "compound promiscuity", "pharmaceutical targets", "publicly available activity data", "data growth", "data confidence levels", "promiscuity progression" ], "content": "Introduction\n\nPolypharmacology is an emerging theme in pharmaceutical research and refers to the property of many bioactive compounds or drugs to act on multiple physiological targets, modulate different signaling pathways, and elicit multi-target-dependent pharmacological effects1–3. The molecular basis of polypharmacology is provided by compound promiscuity, which is defined as the ability of small molecules to specifically interact with multiple targets4,5. It should be emphasized that this form of “specificity pattern promiscuity” is distinct from non-specific interactions or assay artifacts6–8. In light of the latter problems, it is important to identify compound classes that are frequently responsible for artificial activity readouts7,8, e.g. through reactivity under assay conditions. Even in the absence of interaction artifacts, the experimental assessment of promiscuity, e.g. by systematic compound profiling on target sets or families, might be affected by assay confidence limits or detection techniques9, as is the case with any screening experiment. Hence, it might sometimes be difficult to clearly distinguish between “assay promiscuity” and true target promiscuity.\n\nIn addition to experimental studies, promiscuity can also be assessed computationally by mining the rapidly increasing amounts of compound activity data that become available and systematically collecting target annotations for compounds3–5. For computational analysis, it is also of critical importance to carefully consider activity data integrity and confidence levels to arrive at reliable promiscuity estimates5. For compound data mining, public repositories are essential including ChEMBL10, the major public source of data from medicinal chemistry, PubChem’s BioAssay database11, the major source of screening data, and DrugBank12, which collects target annotations for drug candidates and drugs. Systematic computational analysis of promiscuity has been largely dependent on these resources (although proprietary pharmaceutical data have also been used).\n\nIn recent years, computational investigations have provided different promiscuity estimates, depending on the specific aims, study design, and data selection criteria that were applied. Drugs have been the major focal point of these studies. Early estimates on the basis of drug-target networks have suggested that a drug interacts with two targets on average13. Recently, it has been proposed that drugs directed against different target families bind to an average of two to seven targets, depending on their primary target families, and that more than 50% of current drugs bind to more than five targets3. For bioactive compounds, analysis of high-confidence activity data indicated that they interact with an average of one to two targets, with most promiscuous compounds being annotated with two to five targets from the same target family5,14. Moreover, the analysis of high-confidence activity data from 1085 PubChem confirmatory bioassays for 439 targets revealed that a confirmed hit interacted with only two targets on average, although nearly 80% of these active PubChem compounds were tested in more than 50 different assays15. Taken together, computational analyses of bioactive compounds from medicinal chemistry and screening sources indicated the presence of lower degrees of promiscuity overall than was detected for drugs.\n\nThese findings could be rationalized based on the assumption that drugs might often be more extensively tested against different targets than average bioactive compounds. However, this would not explain the relatively low degree of promiscuity observed for active compounds from screening libraries, many of which are extensively tested. Furthermore, promiscuity estimates from computational analysis are occasionally questioned in light of data sparseness16, referring to the fact that available active compounds have not been tested against all targets, which represents the vision and ultimate goal of chemogenomics17. Data incompleteness might principally lead to an underestimation of the degree of promiscuity. However, it remains unclear how significant such deviations might be. In fact, if one considers that millions of activity annotations are already available at present, it should be possible to deduce statistically meaningful trends from such large data samples. Such promiscuity trends might be detected by monitoring promiscuity over time as activity data grow. In a recent study, this type of analysis has been carried out for approved drugs18. For a set of 518 drugs, promiscuity was quantified over different time intervals considering activity data at different confidence levels. When only high-confidence activity records were considered, an increase in the average degree of promiscuity from 1.5 to 3.2 targets per drug was detected over a period of 14 years (from 2000 and 2014). By contrast, when all available activity data were considered, regardless of confidence levels, partially unrealistic increases in promiscuity were observed, ranging from six targets per drug on average in 2000 to more than 28 targets in 201418. For individual high-profile drugs, literally hundreds of target annotations were detected when no confidence criteria were applied. This study showed how dramatic the influence of data confidence levels on promiscuity assessment could be. Furthermore, when considering the results obtained on the basis of high-confidence activity data, the findings also corroborated conclusions drawn from earlier studies discussed above, which indicated that detectable promiscuity of active compounds and drugs might be lower overall than often assumed (and that these observations might not be largely determined by data incompleteness).\n\nTo further refine current promiscuity estimates, we report herein a detailed analysis of the degree of promiscuity of current bioactive compounds monitored over time, spanning a period of 39 years. Special attention was paid to compounds that were first recorded many years ago and are still available. Promiscuity was viewed in light of data growth and monitored using high- and low-confidence activity data. A large number of compounds qualified for this analysis and clear trends were detected. The results of our analysis are presented in the following.\n\n\nMaterials and methods\n\nThe ChEMBL database10 that was analyzed collects large numbers of compounds and activity data, mainly from the medicinal chemistry literature and the PubChem BioAssay database11. The current ChEMBL version (v.20) contains 1,463,270 structurally distinct compounds with activity against 10,774 targets. From 1,148,942 assays, a total of 13,520,737 activity records originated, as reported in Table 1. To systematically explore data growth over time, our analysis focused on data for which release dates were available, which included 913,972 compounds, 10,142 targets, 872,577 assays, and 5,258,052 activity records (Table 1). The growth of these data was monitored on an annual basis. For each year, the number of new entries that became available and the total (cumulative) number of entries was recorded.\n\nFor ChEMBL v.20 and subsets for which specific release dates were available, the total number of compounds, targets, assays, and activity records (activities) is shown.\n\nIn order to investigate compound promiscuity over time as well as the effect of data confidence levels on promiscuity degrees, two data sets with different confidence were assembled from ChEMBL v.20. For the high-confidence data set, a series of selection criteria was applied. Compounds with direct interactions (i.e. assay relationship type “D”) with human single-protein targets at the highest confidence level (i.e. assay confidence score 9) were collected. The two ChEMBL parameters ‘assay relationship type’ and ‘assay confidence score’ qualitatively and quantitatively describe, respectively, the level of confidence that the activity against a given target is evaluated in a relevant assay system. Accordingly, type “D” and score 9 represent the highest level of confidence for activity data. In addition, two types of activity measurements were considered; assay-independent equilibrium constants (Ki values) and assay-dependent IC50 values. To ensure a high level of data integrity, only compounds with explicitly defined Ki and/or IC50 values were selected. Hence, approximate measurements such as “>”, “<”, and “~” were disregarded. Furthermore, activity records including the comments “inactive”, “inconclusive”, or “not active”, were discarded. Thus, this compound set exclusively contained high-confidence activity data. By contrast, the low-confidence data set comprised all compounds with reported interactions against human single-protein targets, regardless of their confidence levels and activity measurement types.\n\nOn the basis of the high- and low-confidence data sets, the progression of compound promiscuity was quantified. Activity records with release dates were assigned to individual compounds. For each year, activity records were assembled. For instance, if a compound was reported to be active against target A in 1990, targets B and C in 2000, and target D in 2005, the cumulative activity records for this compound consisted of target A in 1990, targets A, B and C in 2000, and targets A, B, C, and D in 2005. Thus, the degree of promiscuity of this compound increased from 1 over 3 to 4. For a given year, the average degree of promiscuity was calculated over all qualifying compounds. In addition, subsets of compounds for which activity data first became available in 1994 (20 year activity history) or 2004 (10 year history) were separately monitored.\n\n\nResults and discussion\n\nIn ChEMBL v.20, release dates were reported for 913,972 compounds, 10,142 targets, 872,577 assays, and 5,258,052 activity records (Table 1). Initially, the growth of these source data was analyzed over time. Figure 1 reports the number of new entries that became available each year since 1976 and the total (cumulative) number of entries for each year. As shown in Figure 1a, only 3188 compounds were reported in 1976. In 1977, 6496 compounds were released, yielding a total of 9684 compounds. Since then steady growth in compound numbers was observed until 2006 when the growth rate became nearly exponential, with ~50,000–80,000 compounds becoming available in 2007 and subsequent years. The number of compounds released in 2014 was much lower, probably due to the likely situation that not all new compounds and activity data published in 2014 would have been deposited in the database by the end of the year. Similar growth trends were observed for targets (Figure 1b), assays (Figure 1c) and activity records (Figure 1d).\n\nThe growth of compounds (a), targets (b), assays (c), and activity records (d) is reported. In (a), the number of new compounds becoming available each year is provided using blue bars (scale on the left vertical axis) and the cumulative number of compounds is given as a red line (scale on the right). Corresponding representations are used in (b)–(d).\n\nIn Table 2, the numbers of compounds, targets, assays, and activity records available in 1976 and 2014 are compared. Within this 39-year period, available activity records increased most significantly from 13,999 to 5,258,052 (by a factor of ~376). For compounds and assays, growth factors were comparable (~287 and ~261, respectively). The number of targets increased by a factor of ~79.\n\nThe numbers of compounds, targets, assays, and activity records available in 1976 and 2014 are compared.\n\nOverall, significant increases in the number of compounds, targets, assays, and activity records were observed, especially from 2007 on, thus providing a sound basis for the analysis of compound promiscuity progression over time.\n\nBased on the selection criteria detailed above, two sets of compounds with high- and low-confidence activity data were assembled. In the low-confidence set, compounds with any reported activities against human single-protein targets were included, without applying additional data confidence criteria. By contrast, for the high-confidence set, additional criteria were applied including assay confidence levels as well as the type and integrity of potency measurements. As reported in Table 3, the high-confidence set contained 154,062 compounds active against 1449 targets, yielding a total of nearly 234,000 activity records with release dates. In the low-confidence set, 361,159 compounds active against 2552 targets were available, yielding a total of nearly 782,000 activity records. Data sets of this magnitude were expected to reveal statistically relevant trends in promiscuity progression.\n\nThe numbers of compounds, targets, assays, and activity records with available release dates are reported for the high- and low-confidence data sets, respectively.\n\nGlobal estimate. For compounds in the high- and low-confidence data sets, the average degree of compound promiscuity was determined over the years, as reported in Figure 2. Early on, compounds from both data sets were mostly associated with single-target activities (corresponding to a promiscuity degree of 1). Beginning in 2004, a difference in promiscuity between the high- and low-confidence sets became apparent. However, only a limited increase in promiscuity was observed for compounds from both data sets. From 1976 to 2014, the average degree of promiscuity increased from 1 to 1.5 for the high- and from 1 to 2.2 for the low-confidence data set, thus indicating an overall low degree of promiscuity among bioactive compounds. More interestingly, the average degree of promiscuity for compounds in the high-confidence set only increased by 0.4 (i.e. by less than one target) after 1994 and essentially remained constant between 2004 and 2014, although the amount of available compounds and activity data dramatically increased after 2006 (Figure 1).\n\nFor compounds in the high- and low-confidence data sets, the average degree of compound promiscuity is reported over different years.\n\nPromiscuity on a per-compound basis. In addition to the global assessment of compound promiscuity, progression of promiscuity was also monitored for individual compounds. Table 4 reports the number of compounds with increasing degrees of promiscuity over time. Strikingly, a total of 151,786 (i.e. 98.5%; high-confidence set) and 352,466 (97.6%; low-confidence set) compounds displayed constant degrees of promiscuity over time. Exemplary compounds are shown in Figure 3. These compounds were active against varying numbers of targets. Yet their degrees of promiscuity remained constant until 2014. It is unlikely that subsets of large numbers of compounds with a constant degree of promiscuity over many years have not been tested in various assays. For example, the compound shown at the bottom left in Figure 3 (CHEMBL340211) was reported to be active against two targets in 1993. However, no additional high-confidence activity data became available for this compound during the following 21 years. An abundance of such examples exists for compounds active across current targets.\n\nThe number of compounds with increasing degrees of promiscuity (∆Promiscuity) is reported for the high- and low-confidence data sets. For example, “0” indicates that the degree of promiscuity remained constant over time and “5” that the degree of promiscuity increased by five target annotations.\n\nShown are eight exemplary compounds from the high-confidence data set that displayed a constant degree of promiscuity over different time periods. For each compound, its ChEMBL ID, the degree of promiscuity, and the first year in which target-specific activities were reported are given. For example, “2 \| 1993” (lower left) indicates that this compound was first reported in 1993 to be active against two targets and that this degree of promiscuity (i.e., 2) has remained constant until 2014.\n\nIncreases in promiscuity were only observed for 2276 and 8693 compounds in the high- and low-confidence sets, respectively (Table 4). Moreover, only 181 (high-confidence set) and 1354 (low-confidence set) compounds - a minute fraction of all monitored compounds - gained more than five target annotations over the years.\n\nCompounds with 20 year activity history. Subsets of compounds reported to be active since 1994 were assembled. From the high- and low-confidence sets, 1040 and 19,351 qualifying compounds were obtained, respectively. Promiscuity progression over the subsequent 20 years was separately analyzed for these compound subsets. Figure 4a shows that the degree of promiscuity of the 1040 compounds from the high-confidence data set essentially remained constant, with an increase from 1.1 (1994) to only 1.2 (2014), hence representing lower promiscuity than the global degree of promiscuity determined for the high-confidence set. For the 19,351 compounds from the low-confidence set, the degree of promiscuity only increased from 1.3 to 1.6, which was also lower than the global degree of promiscuity for this set (Figure 4b). Hence, on the basis of activity data monitored over the course of 20 years, compound promiscuity only slightly increased and promiscuity rates were lower than might have been anticipated, although large amounts of activity data became available over time.\n\nThe average degree of promiscuity was compared for all high- (a) and low-confidence (b) set compounds (solid lines) and subsets of compounds reported to be active beginning in 1994 (dashed lines).\n\nUp-to-date promiscuity levels were determined for all qualifying compounds, the subsets of compounds for which activity data first became available in 1994 (20 year activity history), and compound subsets for which activity data first became available in 2004 (10 year history). The results are reported in Table 5. The degree of promiscuity was consistently low in all cases and differences in promiscuity were only marginal. For the high-confidence set, the average degree of promiscuity ranged from 1.3 (20 year activity history) over 1.5 (all compounds) to 1.7 (10 year activity history). For the low-confidence set, it ranged from 1.6 (20 year history) over 2.0 (10 year history) to 2.2 (all compounds). Thus, bioactive compounds generally displayed only a low degree of promiscuity, regardless of the data set from which they originated.\n\nFor the high- and low-confidence data sets, the current average degree of promiscuity is reported for all compounds and compound subsets with activity records available since 1994 and 2004, respectively.\n\n\nConclusions\n\nCurrently available activity data provide an unprecedented source of information for the analysis of bioactive compounds. To assess the promiscuity of bioactive compounds in detail, available activity data have been assigned on the basis of release dates to individual years, thus enabling the study of data growth and compound promiscuity on a time scale and in context. Monitoring compound promiscuity over time was expected to reveal sound trends concerning promiscuity progression and evolving magnitudes. Furthermore, to take data confidence explicitly into account, high- and low-confidence compound data sets were separately generated and analyzed. Data growth and promiscuity progression were ultimately monitored over nearly 40 years (beginning in 1976), both at a global level, as well as focusing on individual compounds or compound subsets of compounds (from the high- and low-confidence sets) with a 20 year or 10 year activity history. The analysis provided a perhaps unexpectedly clear picture and revealed generally low degrees of promiscuity for bioactive compounds, regardless of their activities and origins. Moreover, only minor increases in promiscuity over time were detected for compounds from all sets and subsets, although activity data dramatically increased since 2007. For the high-confidence set, the average degree of promiscuity only increased from 1 to 1.5 over time. Furthermore, even for the low-confidence set, an increase in the degree of promiscuity to only 2.2 was detected. Interestingly, in both cases, promiscuity was constant over time for most compounds. Moreover, for the high-confidence set, the degree of promiscuity essentially remained constant between 2004 and 2014, despite massive data growth. Given the extensive time course followed, the large data volumes accumulated, and the consistent trends detected, these findings could hardly be solely attributed to data incompleteness (although conclusions drawn from data mining might well be affected by data integrity and/or sparseness issues). In our systematic analysis, bioactive compounds were found to display only low degrees of promiscuity, with a surprisingly small influence of data confidence levels, and very limited promiscuity progression over time. The observed trends are anticipated to remain stable as compounds and activity data continue to grow at high rates and provide reference points for future studies of compound and drug promiscuity as the molecular basis of polypharmacology.\n\n\nData availability\n\nThe data selection criteria specified in the Materials and methods section make it possible to reproduce all data sets from ChEMBL v.20, including release dates. The resulting data set statistics are provided in the first part of the Results and discussion section.", "appendix": "Author contributions\n\n\n\nJB conceived the study, YH and SJ planned and performed the analysis, YH and JB wrote the manuscript. All authors agreed to the final content of the manuscript.\n\n\nCompeting 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\nReferences\n\nPaolini GV, Shapland RH, van Hoorn WP, et al.: Global mapping of pharmacological space. Nat Biotechnol. 2006; 24(7): 805–815. PubMed Abstract \| Publisher Full Text\n\nBoran AD, Iyengar R: Systems approaches to polypharmacology and drug discovery. Curr Opin Drug Discov Devel. 2010; 13(3): 297–309. PubMed Abstract \| Free Full Text\n\nJalencas X, Mestres J: On the origins of drug polypharmacology. Med Chem Comm. 2013; 4(1): 80–87. Publisher Full Text\n\nHu Y, Bajorath J: Compound promiscuity: what can we learn from current data? Drug Discov Today. 2013; 18(13–14): 644–650. PubMed Abstract \| Publisher Full Text\n\nHu Y, Bajorath J: High-resolution view of compound promiscuity. [v2; ref status: indexed, http://f1000r.es/1ig]. F1000Res. 2013; 2: 144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcGovern SL, Caselli E, Grigorieff N, et al.: A common mechanism underlying promiscuous inhibitors from virtual and high-throughput screening. J Med Chem. 2002; 45(8): 1712–1722. PubMed Abstract \| Publisher Full Text\n\nBaell JB, Holloway GA: New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J Med Chem. 2010; 53(7): 2719–2740. PubMed Abstract \| Publisher Full Text\n\nBaell J, Walters MA: Chemistry: Chemical con artists foil drug discovery. Nature. 2014; 513(7519): 481–483. PubMed Abstract \| Publisher Full Text\n\nDimova D, Hu Y, Bajorath J: Matched molecular pair analysis of small molecule microarray data identifies promiscuity cliffs and reveals molecular origins of extreme compound promiscuity. J Med Chem. 2012; 55(22): 10220–10228. PubMed Abstract \| Publisher Full Text\n\nGaulton A, Bellis LJ, Bento AP, et al.: ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2012; 40(Database issue): D1100–D1107. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Y, Xiao J, Suzek TO, et al.: PubChem’s BioAssay Database. Nucleic Acids Res. 2012; 40(Database issue): D400–D412. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaw V, Knox C, Djoumbou Y, et al.: DrugBank 4.0: Shedding new light on drug metabolism. Nucleic Acids Res. 2014; 42(Database issue): D1091–D1097. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYildirim MA, Goh KI, Cusick ME, et al.: Drug-target network. Nat Biotechnol. 2007; 25(10): 1119–1126. PubMed Abstract \| Publisher Full Text\n\nHu Y, Bajorath J: Promiscuity profiles of bioactive compounds: potency range and difference distributions and the relation to target numbers and families. Med Chem Commun. 2013; 4: 1196–1201. Publisher Full Text\n\nHu Y, Bajorath J: What is the likelihood of an active compound to be promiscuous? Systematic assessment of compound promiscuity on the basis of PubChem confirmatory bioassay data. AAPS J. 2013; 15(3): 808–815. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMestres J, Gregori-Puigjané E, Valverde S, et al.: Data completeness--the Achilles heel of drug-target networks. Nat Biotechnol. 2008; 26(9): 983–984. PubMed Abstract \| Publisher Full Text\n\nRognan D: Chemogenomic approaches to rational drug design. Br J Pharmacol. 2007; 152(1): 38–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu Y, Bajorath J: Monitoring drug promiscuity over time [v2; ref status: indexed, http://f1000r.es/4oa]. F1000Res. 2014; 3: 218. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8662", "date": "14 May 2015", "name": "Georgia B. McGaughey", "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 believe this paper should be indexed as I have not seen such a methodical and quantified examination of promiscuity before. The article is well written and easy to read. Figures are compelling. Although there is compelling data included suggesting that over time, promiscuity generally doesn’t increase for a given compound, I'm a bit skeptical on concluding that promiscuity may not have markedly increased over the past few decades based on merely ChEMBL. I, however, recognize that those in academia (or in a biotech company where large receptor screening may not be part of the business model) may not have access to an orthogonal data set. Frankly, other than ChEMBL, I’m not sure where else one would go to look for off target data. There are purchasable databases (e.g. Integrity). However, more public data is relatively sparse. Even a (young) small-ish biotechnology company would not have enough data to utilize. One topic that has been raised in the literature the past few years is the concept of phenotypic versus target based drug discovery approaches to developing new medicines. I would have liked to see some differentiation between promiscuity of target based versus phenotypic based projects. Is that something the reviewers can go back to and annotate their data set? Additionally, discussion around the differences between promiscuity and polypharmacology should be elaborated upon. I realize this is raised in the \"introduction\", but I would have liked to see more attention paid to this topic.\n\nFinally, will the data sets be publicly available with annotations?", "responses": [ { "c_id": "1400", "date": "15 Jun 2015", "name": "Jürgen Bajorath", "role": "Author Response F1000Research Advisory Board Member", "response": "1. We absolutely agree with the referee that it would be highly desirable to have more pharmaceutically relevant data sets available in the public domain, for instance, large profiling data sets, which we would love to analyze. We also note that ChEMBL data have been, and continue to be, incorporated in a number of different public databases. Concerning commercial databases, we have made a decision not to analyze (and publish about) databases that are not publicly available (although they are occasionally offered to us). 2. It is also true that publicly available phenotypic data sets would be of great help for the field moving forward. Unfortunately, very little phenotypic data is currently available. For ChEMBL activity records, it is not possible to trace potential phenotypic origin at present (we assume very little target-based activity data derived from phenotypic assays is currently available, if any). 3. The discussion on promiscuity and polypharmacology has been further extended in our revision, as suggested. 4. The high- and low-confidence data sets will be made publicly available as an open access deposition, as stated in the revised manuscript." } ] }, { "id": "8663", "date": "26 May 2015", "name": "Christopher Southan", "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 analysis presented in this paper is of considerable interest and should be indexed, However, I think there are many confounding factors within the ChEMBL data that the authors have not addressed sufficiently. I will pick up some of these below.Polypharmacology usually implies the affects mediated via the multiple targets are therapeutically “positive”. Is this the authors' implication also? Otherwise the term implicitly extends to toxicity and side effects. Figure 1 should include the distribution that underlies the other three, namely papers per-year. While the databases used were different, a published tracking of compound output from papers showed much less increase over 20 years than in figure 1 (PMID:24204758) although the target growth pattern was similar. Have the authors checked that ChEMBL did not pick up new journal coverage from 2008 that would spike the increases? I would like more detail on how the filtration methodology in the paper is used to extract and score (a flow chart would help). Let me pose a hypothetical case of two compounds. The first ranks target A at an IC50 of 20nM and target B at 30 nM. The second compound is 1nM and 500 nM for the same two targets. Do the two cases get the same promiscuity score? (It would be confounding if they did.) What happens when compound-target-assay values are identical for different publication years (not uncommon in ChEMBL) - Do you score only the first year ? I’m confused by use of “release date” (as for a database) surely “publication date” is meant? For fig.3 I suggest the dominant explanation for apparently constant promiscuity is simply “publish-and-forget” (i.e. researchers typically do not re-test compounds published by others). As we know re-testing leading to the publication of new results (promiscuous or not) will be largely dependent as to whether structures become reference compounds, are advanced into development, or become drugs. So could the “papers-per-compound” relationship be plotted to provide insight into this? There are other confounding trends that could be tested for, for example targets-per-paper (i.e. < cross-screening over the years might correlate with apparent promiscuity <) and orthologue vs paralogoue cross screening (i.e. if the average human:rodent ratio changes over time for the low confidence set). Why not select kinase inhibitors as a control subset? We would expect these to exhibit highest promiscuity and they would thus be an important methodological cross-check. In terms of other obvious hypothesis checks why not split by LogP (as might increase promiscuity) and Mw (as might decrease it) ? While appreciating the academic imperative I do wish this team could have merged some of their previous papers that appear to address essentially the same theme. For example, comparing drug (ref.18) vs non-drug promiscuity in the same standardised study is better (easier to review even :) than splitting the result sets.", "responses": [ { "c_id": "1399", "date": "15 Jun 2015", "name": "Jürgen Bajorath", "role": "Author Response F1000Research Advisory Board Member", "response": "1. This view is shared by us. 2. The number of papers published per year was analyzed and found to essentially parallel the growth of assays (Figure 1c of the revision). 3. The study (PMID: 24204758) published by Southan et al. is now referenced but not directly comparable to our current analysis. The authors analyzed the commercial GOSTAR database that includes the GVKBIO Medicinal Chemistry Database and Target Class Database. Figure 1 in this study reports the number of compounds linked to human protein targets in journals published over different years. Promiscuity analysis is not reported. On a close look, the trends of compound growth over time reported in the Southan et al. paper and our current study are actually rather similar. In addition, we note that GVKBIO apparently covers 120 journals while ChEMBL covers 47. 4. The data filtering criteria for the collection of high- and low-confidence data sets have been detailed in the section “Data sets of varying confidence levels”. No promiscuity score was calculated taking potency into account. The degree of compound promiscuity was assessed on the basis of qualifying activity records, as stated in our paper. In the hypothetical example given by the referee, the two compounds would share the same degree of promiscuity (2). 5. We did not score compound-target-assay values, only recorded them. Combination of compounds, targets, and assays analyzed in our study were unique for individual years. Thus, there were no identical values for the same compound-target-assay combination in different years. However, there were cases where a compound was tested in various assays against the same target in different years, yielding the same or different potency values. In these cases, we only recorded the first year, as specified in the revision. 6. Yes, “release date” means “publication date”, as clarified in the revision. 7. This is a (quite plausible) hypothesis, like others put forward in trying to rationalize promiscuity. Our view is data-centric, as commented on in our study. The papers-per-compound ratio in ChEMBL typically is close to one (with relatively small sets of standards/references used in drug target assays being an exception). However, we cannot deduce much from this ratio because inactivity records are sparse in the literature and are not considered in ChEMBL (and other repositories). However, they are available in PubChem. As reported previously (PMID: 23605807), approximately 77% of screening hits in PubChem have been tested in at least 50 confirmatory assays. Yet, the detectable average degree of promiscuity for screening hits is only 2.5, thus only slightly larger than for ChEMBL compounds. 8. There certainly is always more one could possibly do. 9. The promiscuity of kinase inhibitors has been explored in previous studies.For example, in reference 5, compound promiscuity for five well-known target families including kinases was reported. It was observed that kinase inhibitors did not display a higher degree of promiscuity than compounds directed against other major therapeutic targets; a trend we have consistently observed (please, also consider PMID: 25051177). In the context of promiscuity analysis, kinase inhibitors are a good example for cases where expectations/hypotheses are not necessarily consistent with results obtained on the basis of current data. 10. As suggested, the degree of promiscuity was further analyzed for compounds with varying logP values and molecular weight, as reported in new Figure 5 of our revision. 11. The in part unexpected results reported in reference 18 had inspired us to also have a close look at promiscuity of bioactive compounds on a time course; as we know it takes some thought and time to conceptualize such studies and understand which questions to ask next." } ] }, { "id": "8802", "date": "29 May 2015", "name": "John A. Lowe III", "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 investigate the potential growth of off-target activity over time as new assays become available. They control for multiple potential confounds, and possibly the most important is data quality enabling confidence in the results. They note that prior data indicated screening compounds typically bind to at least two targets, while drugs may bind up to seven, but this result might be skewed by insufficient data quality. Their study reported here uses very large datasets and controls for the growth of compound number as well as new data over time. They also deliberately analyze high- and low-quality (or confidence) data separately. This careful analysis gives them a much clearer picture of the changes in compound promiscuity over time, and reveals a low level of off-target activity and only a slight increase with time. Despite new assays becoming available at an increasing rate, compound promiscuity has not increased significantly, a result that will surprise many readers, but which the authors have documented admirably. I highly recommend this manuscript for indexing.", "responses": [ { "c_id": "1398", "date": "15 Jun 2015", "name": "Jürgen Bajorath", "role": "Author Response F1000Research Advisory Board Member", "response": "We thank the reviewer very much for these encouraging comments (that ease the pain often felt when going through massive amounts of compound activity data while striving for a rigorous analysis …)." } ] } ]	1	https://f1000research.com/articles/4-118
https://f1000research.com/articles/4-150/v1	11 Jun 15	{ "type": "Research Note", "title": "Relationships between activation level, knowledge, self-efficacy, and self-management behavior in heart failure patients discharged from rural hospitals", "authors": [ "Van Do", "Lufei Young", "Sue Barnason", "Hoang Tran", "Sue Barnason", "Hoang Tran" ], "abstract": "Non-adherence to self-management guidelines accounted for 50% of hospital readmissions in heart failure patients. Evidence showed that patient activation affects self-management behaviors in populations living with chronic conditions. The purpose of this study was to describe patient activation level and its relationship with knowledge, self-efficacy and self-management behaviors in heart failure patients discharged from rural hospitals. Our study populations were recruited from two hospitals in rural areas of Nebraska. We found that two-thirds of the participants reported low activation levels (e.g., taking no action to manage their heart failure condition). In addition, low patient activation levels were associated with inadequate heart failure knowledge (p=.005), low self-efficacy (p<.001) and low engagement in heart failure self-management behaviors (p<.001) after discharge from hospital.", "keywords": [ "Rural Populations", "Cardiac Failure", "Patient Activation", "Patient Engagement", "Self-Efficacy", "Self-Management", "Knowledge" ], "content": "\n\nHeart failure is a major public health problem in the United States and worldwide. In the United States, heart failure affected 5.1 million patients with 12–15 million office visits, 6.5 million hospital days and cost approximately 32 billion U.S. dollars in 20101. Despite the recent declining trend of cardiovascular disease-related mortality in the US, the rehospitalization rate for heart failure patients remains 30% within 60–90 days after discharge2. Rural populations exhibit higher prevalence of heart failure3 and rural patients are more likely to be readmitted due to heart failure4,5 compared to those in urban areas.\n\nAmong all causes for heart failure-related readmission, non-adherence to self-management is the most common reason, accounting for 50% of readmissions in heart failure patients6,7. Patient self-management is one of the key concepts in the Chronic Care Model developed by Edward H. Wagner8. Self-management behaviors refer to the practice of activities that individuals initiate and perform on their own behalf in the interest of maintaining life, health, continuing personal development, and well-being9. Self-management behaviors in heart failure patients primarily involve monitoring daily weight, following a restricted sodium diet, fluid restriction, taking prescribed medications, exercising regularly, and keeping scheduled follow-up appointments4.\n\nPrior nationwide studies among adults with chronic diseases showed that a high level of patient knowledge, efficacy and activation level were associated with good self-management10–12 and ultimately led to fewer hospitalizations and emergency department visits10,13. Few studies have reported the relationship between disease specific knowledge, self-efficacy, patient activation and self-management behaviors in rural heart failure patients. Self-efficacy for heart failure self-management is defined as how confidently individuals can achieve specific functions or control various aspects of their heart failure14.\n\nPatient engagement with self-management is critical for heart failure patients. Patient activation is the concept that can be applied as a means to determine patient engagement with self-management. The patient’s activation level, which is measured by Hibbard’s Patient Activation Measure (PAM), reflects the degree to which the person is ready, willing and able to engage in managing her or his health conditions12,15. Based on the PAM score, a person’s activation level is graded into 4 levels, from low to high: (1) Patients in level one believe they are responsible for managing their health; (2) patients in level two feel confident and knowledgeable regarding managing their health; (3) patients in level three actively engage in managing their health; and (4) patients in level four consistently engage in activities to manage their health and maintain those actions even under stress. The advancement in patient activation levels reflects the progress of the patient from being a passive care receiver to a more confident care manager12,13,15.\n\nConsidering the low engagement in self-management of patients living with chronic diseases in the rural areas16,17, it is critical to examine whether there are positive relationships between knowledge, self-efficacy, patient activation and self-management behaviors. To our knowledge, these relationships have not been well studied in the rural heart failure population. Our study will contribute knowledge in understanding the impact of patient activation on self-management, which could inform the development of effective interventions to promote self-management in rural heart failure population. For this purpose, our study populations were recruited from hospitals in rural areas of Nebraska.\n\n\nConceptual framework\n\nAs a cultural aspect, many rural patients endorse the importance of personal responsibility, productivity, and self-reliance in terms of health practice18. Based on the rural culture belief and health practice, Bandura’s social cognitive theory19, Hibbard’s Patient Activation Theory12,20, and the Chronic Care Model8,21, we proposed a conceptual framework (Figure 1). Our framework assumes that by attaining self-management knowledge and efficacy, patients will advance their activation to higher levels, leading to long-term engagement in self-management behaviors.\n\n\nMethods\n\nWe conducted a secondary analysis to evaluate the relationships between levels of patient activation and heart failure self-management knowledge, self-efficacy, and self-management behaviors in heart failure patients discharged from rural critical access hospitals to home. The data for this paper formed the baseline data from a NIH-funded randomized controlled trial titled “Patient AcTivated Care at Home (PATCH)” which aims to examine the feasibility of a 12-week home-based intervention to improve heart failure self-management adherence. This trial can be found on https://www.clinicaltrials.gov/ct2/show/NCT01964053.\n\nParticipants were recruited from October 2013 to December 2014 from two rural critical access hospitals in southeast Nebraska. Patients were eligible for the study if they: 1) were age 21 or older; 2) had heart failure as one of their discharge diagnoses; 3) had New York Heart Association (NYHA) class II to IV heart failure or had NYHA class I heart failure and had at least one heart failure-related hospitalization or emergency department visit in the previous year; 4) were discharged to home; 5) passed the Mini-Cog screen test screening for dementia22; 6) understood English; and 7) had access to a phone.\n\nWe excluded patients who: 1) had depressive symptoms (received a score of 3 or above on the Patient Health Questionnaire-2 (PHQ-2)23; 2) were diagnosed with liver cirrhosis; 3) were diagnosed with chronic renal failure; and 4) were diagnosed with other end stage and/or terminal illness (e.g. cancer) which limited the patient’s ability to perform self-management behaviors. The study setting is described in more detail in the study protocol24.\n\nWe collected socio-demographic characteristics including age, gender, educational attainment, race/ethnicity, annual household income, marital status, and smoking status with a structured questionnaire. Clinical characteristics included comorbidities, echocardiographic ejection fraction (EF), and New York Heart Association (NYHA) Functional Classification.\n\nWe measured patient activation using the Short Form of the Patient Activation Measure (13-item version, available at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1361231/table/tbl1/) which has similar reliability and validity to the long form (22 item version) across different ages, genders and health condition status15. Each item of the form was scored on the 5-point Likert response scale. For the ease of interpretation, the raw scores were transformed from the original metric to a 0–100 metric with higher scores indicating higher activation levels. Based on the patient activation score, patients were categorized into four levels: level 1 (score <47.0), level 2 (score 47.1–55.1), level 3 (score 55.2–67.0), and level 4 (score >67.0)12,15.\n\nWe measured self-efficacy using the Self-Care of Heart Failure Index (SCHFI) which has six questions on a 4 point Likert scale14. The SCHFI was used to access the degree of achievement in self-care maintenance and management25. The raw scores were standardized to a 100 point scale, with higher scores indicating higher efficacy. This measurement tool is reliable to assess self-efficacy with Cronbach’s alpha coefficient of .8314.\n\nWe measured patients’ heart failure knowledge using the 27 multiple-choice questions of the Atlanta Heart Failure Knowledge Test (AHFKT)26. Each correct answer scored 1 point and the total score ranged from 0 to 27. This questionnaire established high reliability to evaluate heart failure knowledge with a Cronbach alpha coefficient of .8426.\n\nWe used the 29-item Revised Heart Failure Self-Care Behavior Scale (RHFSCS) to assess patients’ behavior in six heart failure care domains: (1) seeking medical help, (2) being aware of the effects of heart failure, (3) prevention of complications, (4) awareness of deleterious effects of medical care, (5) accepting heart failure, and (6) learning to live with heart failure27. Each response is granted a score from 0 (none of the time) to 5 (all of the time)27. The internal reliability of this questionnaire is consistent with a Cronbach alpha coefficient of .8427.\n\nStatistical analyses were conducted using SPSS version 20, with a p value less than .05 considered significant. We reported mean and standard deviation for continuous variables. Frequency and percentage were used to report categorical variables. Since patient activation level was an ordered variable, the Chi-square test for trend was used to assess correlations between four activation levels and the categorical variables (e.g. patient’s demographic and clinical characteristics). To compare continuous variables (e.g. behaviors, heart failure knowledge, Self-efficacy) across four activation levels, we used the ANOVA test for normally-distributed variables and the Kruskal-Wallis test for the non-normally distributed variables.\n\nThe study protocol was approved by the University of Nebraska Medical Center Institutional Review Board (IRB) and hospital ethical committees (IRB PROTOCOL # 228-13-EP).\n\n\nResults\n\nThere were 101 patients enrolled in this study. Table 1 describes socio-demographic and clinical characteristics. The mean age was 70±12.1 years. The majority of participants were women (63%). Two-thirds of participants had an annual household income of less than $50,000 or an educational attainment of high school graduate or lower. All the patients had multiple comorbidities including more than 2 of these chronic conditions: Hypertension (98.0%), Coronary artery disease (94.1%), Dyslipidemia (83.2%), Diabetes mellitus (41.6%), COPD/asthma (39.4%), CVA/stroke (17.8%), and cancer (6.9%). The majority of participants are at NYHA class II and III with the average EF of 55.8±11.1.\n\nSignificant results in bold face\n\n$missing data from 1 participant, n=100\n\n†missing data from 7 participants, n=94\n\n‡missing data from 12 participants, n=89\n\n--: not applicable\n\nThe mean patient activation score was 57 and the median was 52 (95% CI 53–61). The patient activation score did not distribute normally with a positively skewed pattern. About 40% of participants believed that they were responsible for caring for their health and illness but failed to take action to manage their heart failure condition (level 1). A quarter of participants felt confident and knowledgeable enough to manage their health but failed to take action (level 2). Less than 40% of participants actually took some actions to manage their condition (level 3 and 4).\n\nPatient activation levels were not significantly different across age group, gender, marital status, household income or smoking status. There was an association between activation level and educational attainment. Participants with educational attainment beyond high school had higher patient activation scores on the PAM compared to patients who only graduated high school or lower (62 vs. 54, p=.03).\n\nUsing the Chi-square test for trend, we found an increasing trend between knowledge and self-efficacy across the four levels of patient activation. Patients with higher heart failure knowledge or greater self-efficacy were at a higher level of patient activation (p=.005 and p<.001, respectively). Patients who were at a higher level of patient activation also had higher scores of self-management behavior (p<.001). Table 2 describes the average score of self-management knowledge, efficacy, and behavior among 4 levels of patient activation level.\n\nSignificant results in bold face\n\n* df=1, Chi-square for trend\n\n\n\n\nDiscussion\n\nThis is one of the first studies to assess self-management knowledge, self-efficacy, activation level, and self-management behavior in American rural heart failure patients. The mean patient activation score of our study was 57, lower than that in Shively’s study (61.7) of heart failure patients or Green’s in the general chronic patient population (66.4)10,28. In addition, fewer of our rural heart failure participants were actively managing their condition compared to participants in other studies of both rural and urban areas10,28. According to our study, a lower percentage of heart failure patients (38%) had high activation levels and consistently took action to manage their health compared to diabetic patients in Begum’s (69.9%) and Rask’s studies (82.9%)29,30. However, patient activation scores in our study were higher than that in Evangelista’s study (37.3–39.3) in which the population consisted of heart failure patients in a palliative care setting31. Similarly to Marshall’s study32, the distribution of patient activation scores in our study was not different across various socio-demographic groups except for educational attainment. A higher educational attainment was associated with a higher level of activation in our patients. This finding was consistent with other studies which also found educational attainment to be the most powerful predictor for patient activation in the chronic disease population33–36. It is plausible that patients with higher educational levels are more likely to achieve better health literacy, giving them more awareness, skills and confidence to take self-management actions. In contrast to several studies, we did not find a variance of patient activation score among different age groups31,37. A possible explanation is that our participants seemed to be older and have narrower age variation (mean age 70, rank from 40 to 93) compared other studies (mean 53.731 and age range from 18 to older than 7537). Evidence shows that an increase in patient activation level results in improved self-management behavior, leading to better health outcomes13,38. The overall low patient activation score in our heart failure participants indicates the need of developing interventions to enhance activation and self-management behavior in rural heart failure patients.\n\nThe associations between patient activation and heart failure knowledge, self-efficacy, and self-management behavior support our conceptual framework, as well as previous studies34,37. Patients with higher activation levels tended to have more knowledge to manage their heart failure, more confidence in self-management of heart failure, and were more likely to engage in self-management behaviors such as regular exercise, watching their diet and fluid intake, and adherence to medication. From a qualitative study, Dixon et al. reported that patients at lower activation levels indicated a lack of knowledge and lack of confidence as barriers for them to self-manage their health conditions39. Our findings indicate that we can potentially boost patients’ activation by increasing their self-management knowledge and efficacy to take care of their own health. These findings confirmed our original hypothesis that strategies to enhance activation levels should be included in the intervention to promote heart failure self-management behaviors.\n\nOur study had some limitations. We used only baseline data; therefore, we are unable to demonstrate the changes of patient activation scores, knowledge scores, self-efficacy and self-management behaviors over time. As a result, we could not establish the temporal relationship between knowledge, efficacy, patient activation level and their impact on behavior in this article, which will be reported in future manuscripts. Secondly, the patients who agreed to participate in the original randomized controlled trial might be more motivated than patients who refused to participate, therefore creating a pool of heart failure patients with higher activation levels compared to the general rural heart failure population. If this assumption is valid, the overall activation level in general rural heart failure population might be even lower than what is reported in this article, which indicates a great need to conduct interventions to enhance rural heart failure activation to engage in self-management behaviors.\n\nUsing the short version of PAM (13 items) is a feasible way to assess a patient’s activation level and identify those with low levels in both inpatient and outpatient settings. The assessment results allow clinicians to develop tailored interventions to support those high risk patients. Additionally, the patients with high activation levels could be encouraged and recruited as behavior coaches and/or peer supporters for those with low activation scores.\n\n\nConclusions\n\nSelf-management plays a vital role in improving health outcomes and reducing healthcare costs in the heart failure population. Patient activation level is significantly associated with self-management behavior. However, the activation level in our rural heart failure patients was relatively low compared to patients in other studies, which suggest the need for interventions to improve activation levels in the rural heart failure population.\n\n\nData availability\n\nF1000Research: Dataset 1. Heart failure (HF) patients discharged from rural hospitals: demographic, clinical, patient activation level, HF knowledge and self-management behavior characteristics, 10.5256/f1000research.6557.d4920540\n\n\nConsent\n\nWritten informed consent for publication of clinical details was obtained from all the participants.", "appendix": "Author contributions\n\n\n\nDr. Do and Dr. Young developed the research question and the conceptual framework. Dr. Do and Dr. Tran analyzed data and prepared for tables and figures. Dr. Do, Dr. Tran, Dr. Young and Dr. Barnason prepared the manuscript. The co-author, Dr. Young, has full access to the study data and takes responsibility for the integrity and the accuracy of the data analysis. All authors were involved in the review of the final draft and agreed to the content of the final submission.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nResearch reported in this publication was supported by the National Institutes Nursing Research of the National Institutes of Health under award number 1R15NR 13769-01A1. The sponsor had no role in conducting the study, preparing and disseminating the study results. The authors are the recipients of the funding provided by the National Institutes Nursing Research of the National Institutes of Health.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nGo AS, Mozaffarian D, Roger VL, et al.: Heart disease and stroke statistics--2013 update: a report from the American Heart Association. Circulation. 2013; 127(1): e6–e245. PubMed Abstract \| Publisher Full Text\n\nGheorghiade M, Vaduganathan M, Fonarow GC, et al.: Rehospitalization for heart failure: problems and perspectives. J Am Coll Cardiol. 2013; 61(4): 391–403. PubMed Abstract \| Publisher Full Text\n\nPearson TA, Lewis C: Rural epidemiology: insights from a rural population laboratory. Am J Epidemiol. 1998; 148(10): 949–957. PubMed Abstract\n\nJessup M, Abraham WT, Casey DE, et al.: 2009 focused update: ACCF/AHA Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation. 2009; 119(14): 1977–2016. PubMed Abstract \| Publisher Full Text\n\nJin Y, Quan H, Cujec B, et al.: Rural and urban outcomes after hospitalization for congestive heart failure in Alberta, Canada. J Card Fail. 2003; 9(4): 278–285. PubMed Abstract \| Publisher Full Text\n\nEvangelista LS, Shinnick MA: What do we know about adherence and self-care? J Cardiovasc Nurs. 2008; 23(3): 250–257. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan der Wal MH, Jaarsma T, Moser DK, et al.: Qualitative examination of compliance in heart failure patients in The Netherlands. Heart Lung. 2010; 39(2): 121–130. PubMed Abstract \| Publisher Full Text\n\nBodenheimer T, Wagner EH, Grumbach K: Improving primary care for patients with chronic illness. JAMA. 2002; 288(14): 1775–1779. PubMed Abstract \| Publisher Full Text\n\nOrem DE, Taylor SG, Renpenning KM: Nursing: Concepts of Practice. 6th ed., Mosby St. Louis, 2001. Reference Source\n\nGreene J, Hibbard JH: Why does patient activation matter? An examination of the relationships between patient activation and health-related outcomes. J Gen Internal Med. 2012; 27(5): 520–526. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMosen DM, Schmittdiel J, Hibbard J, et al.: Is patient activation associated with outcomes of care for adults with chronic conditions? J Ambul Care Manage. 2007; 30(1): 21–29. PubMed Abstract \| Publisher Full Text\n\nHibbard JH, Stockard J, Mahoney ER, et al.: Development of the patient activation measure (PAM): Conceptualizing and measuring activation in patients and consumers. Health Serv Res. 2004; 39(4 Pt 1): 1005–1026. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHibbard JH, Greene J, Tusler M: Improving the outcomes of disease management by tailoring care to the patient's level of activation. Am J Manag Care. 2009; 15(6): 353–360. PubMed Abstract\n\nRiegel B, Lee CS, Dickson VV, et al.: An update on the self-care of heart failure index. J Cardiovasc Nurs. 2009; 24(6): 485–497. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHibbard JH, Mahoney ER, Stockard J, et al.: Development and testing of a short form of the patient activation measure. Health Serv Res. 2005; 40(6 Pt 1): 1918–1930. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBell RA, Arcury TA, Snively BM, et al.: Diabetes foot self-care practices in a rural triethnic population. Diabetes Educ. 2005; 31(1): 75–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCoronado GD, Thompson B, Tejeda S, et al.: Sociodemographic factors and self-management practices related to type 2 diabetes among Hispanics and non-Hispanic whites in a rural setting. J Rural Health. 2007; 23(1): 49–54. PubMed Abstract \| Publisher Full Text\n\nBardach SH, Tarasenko YN, Schoenberg NE: The role of social support in multiple morbidity: self-management among rural residents. J Health Care Poor Underserved. 2011; 22(3): 756–771. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBandura A: Perceived self-efficacy in cognitive development and functioning. Educational Psychologist. 1993; 28(2): 117–148. Publisher Full Text\n\nHibbard JH, Mahoney ER, Stock R, et al.: Do increases in patient activation result in improved self-management behaviors? Health Serv Res. 2007; 42(4): 1443–1463. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWagner EH, Austin BT, Davis C, et al.: Improving chronic illness care: translating evidence into action. Health Aff (Millwood). 2001; 20(6): 64–78. PubMed Abstract \| Publisher Full Text\n\nBorson S, scanlan J, Brush M, et al.: The mini-cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry. 2000; 15(11): 1021–1027. PubMed Abstract \| Publisher Full Text\n\nLi C, Friedman B, Conwell Y, et al.: Validity of the Patient Health Questionnaire 2 (PHQ-2) in identifying major depression in older people. J Am Geriatr Soc. 2007; 55(4): 596–602. PubMed Abstract \| Publisher Full Text\n\nYoung L, Barnason S, Do V: Promoting self-management through adherence among heart failure patients discharged from rural hospitals: a study protocol [v2; ref status: indexed, http://f1000r.es/5c7]. F1000Res. 2015; 3. Publisher Full Text\n\nRiegel B, Carlson B, Moser DK, et al.: Psychometric testing of the self-care of heart failure index. J Card Fail. 2004; 10(4): 350–360. PubMed Abstract \| Publisher Full Text\n\nReilly CM, Higgins M, Smith A, et al.: Development, psychometric testing, and revision of the Atlanta Heart Failure Knowledge Test. J Cardiovasc Nurs. 2009; 24(6): 500–509. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArtinian NT, Magnan M, Sloan M, et al.: Self-care behaviors among patients with heart failure. Heart Lung. 2002; 31(3): 161–172. PubMed Abstract \| Publisher Full Text\n\nShively MJ, Gardetto NJ, Kodiath MF, et al.: Effect of patient activation on self-management in patients with heart failure. J Cardiovasc Nurs. 2013; 28(1): 20–34. PubMed Abstract \| Publisher Full Text\n\nBegum N, Donald M, Ozolins IZ, et al.: Hospital admissions, emergency department utilisation and patient activation for self-management among people with diabetes. Diabetes Res Clin Pract. 2011; 93(2): 260–267. PubMed Abstract \| Publisher Full Text\n\nRask KJ, Ziemer DC, Kohler SA, et al.: Patient activation is associated with healthy behaviors and ease in managing diabetes in an indigent population. Diabetes Educ. 2009; 35(4): 622–630. PubMed Abstract \| Publisher Full Text\n\nEvangelista LS, Liao S, Motie M, et al.: On-going palliative care enhances perceived control and patient activation and reduces symptom distress in patients with symptomatic heart failure: a pilot study. Eur J Cardiovasc Nurs. 2014; 13(2): 116–123. PubMed Abstract \| Publisher Full Text\n\nMarshall R, Beach MC, Saha S, et al.: Patient activation and improved outcomes in HIV-infected patients. J Gen Intern Med. 2013; 28(5): 668–674. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLubetkin EI, Lu WH, Gold MR: Levels and correlates of patient activation in health center settings: Building strategies for improving health outcomes. J Health Care Poor Underserved. 2010; 21(3): 796–808. PubMed Abstract \| Publisher Full Text\n\nNijman J, Hendriks M, Brabers A, et al.: Patient activation and health literacy as predictors of health information use in a general sample of dutch health care consumers. J Health Commun. 2014; 19(8): 955–69. PubMed Abstract \| Publisher Full Text\n\nRademakers J, Nijman J, van der Hoek L, et al.: Measuring patient activation in The Netherlands: translation and validation of the American short form Patient Activation Measure (PAM13). BMC Public Health. 2012; 12: 577. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaha S, Koley M, Mahoney ER, et al.: Patient activation measures in a government homeopathic hospital in India. J Evid Based Complementary Altern Med. 2014; 19(4): 253–259. PubMed Abstract \| Publisher Full Text\n\nHendriks M, Rademakers J: Relationships between patient activation, disease-specific knowledge and health outcomes among people with diabetes; a survey study. BMC Health Serv Res. 2014; 14(1): 393. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHibbard JH, Greene J: What the evidence shows about patient activation: better health outcomes and care experiences; fewer data on costs. Health Aff (Millwood). 2013; 32(2): 207–214. PubMed Abstract \| Publisher Full Text\n\nDixon A, Hibbard J, Tusler M: How do People with Different Levels of Activation Self-Manage their Chronic Conditions? Patient. 2009; 2(4): 257–268. PubMed Abstract \| Publisher Full Text\n\nDo V, Young L, Barnason S: Dataset 1 in: Relationships between activation level, knowledge, self-efficacy, and self-management behavior in heart failure patients discharged from rural hospitals. F1000Research. 2015. Data Source" }	[ { "id": "8999", "date": "22 Jun 2015", "name": "Hoa L. Nguyen", "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 paper potentially provides important information about the relationship between activation level, knowledge, self-efficacy and self-management behavior in the rural population with heart failure to inform future interventions. I have several minor comments that the authors may consider in the revised version:Abstract: The authors may consider adding sample size, and several basic characteristics of the study population (mean age, sex distribution, etc.) in the abstract. Since the sample size of the study is relatively small (n=101), and the authors stated that the distribution of the patient activation score was skewed (page 3, last paragraph), they may consider reporting medians (inter quartile ranges) in the tables and text in addition to means (SDs). Statistical analysis: The authors mentioned Chi-square, Anova, and Kruskal-Wallis tests were used to compare patients’ characteristics across 4 activation levels, but it is unclear where the results were reported in the results section.In table 1, patients’ characteristics for all patients were presented (not according to activation levels) and means (SDs) PAM score by patients characteristics were reported.The authors may consider stating clearly in the statistical analysis how they treated the activation score (the main outcome of interest) as continuous or as categorical variable or both approaches and report the results accordingly. Discussion: The author wrote “the distribution of patient activation scores in our study was not different across various socio-demographic groups except for educational” (page 5). This may be due to small sample size issue; the study may be underpowered to detect potential differences (e.g. the mean activation score difference by race/ethnicity was 14 (58 vs. 44), however the difference was not statistically significant). This is one of the study limitations.", "responses": [] }, { "id": "9428", "date": "09 Jul 2015", "name": "George Sokos", "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\nPage 2, paragraph 1: Consider adding information about CMS limited reimbursement for heart failure readmissions as this is a significant current concern in US hospitals. Figure 1 and throughout entire paper: Consider simplifying terms “self-management knowledge” to “Knowledge”, “Self management efficacy” to “efficacy”. Many different variations of these terms are used throughout which reduces the readability of the paper. Table 2 on page 5 is organized well and has these terms outlined more simply. Page 2, paragraph 3, last sentence: Consider changing “self-efficacy for heart failure self-management “ to “Efficacy for heart failure self-management” as the terms can be confusing if the reader is trying to understand the concept using Figure 1. This may be considered throughout the paper. Page 2, paragraph 4: It might be more helpful for the reader to see the PAM scores 1-4 in a table to allow for easy referencing when reading the paper. Page 2, paragraph 5, first sentence: Consider changing “self-efficacy” to “efficacy” (see comment number 2 above). Page 2, paragraph 6, last sentence: Consider removing “long-term” as this specific study make correlations but does not address the length of time the patient’s behavior will be impacted. Page 3, paragraph 1, sentence 1: Consider changing term “self efficacy” to efficacy (see comment number 2 above). Page 3, paragraph 6, sentence 1: Consider changing term “self efficacy” to efficacy (see comment number 2 above). Page 3, Measurement section: Consider arranging paragraphs to match flow in Figure 1 (Knowledge, Efficacy, Patient Activation, and Self-management behavior). Page 3, paragraph 8: Change “patients behavior” to “patients self-management behavior”. Table 1: Consider additional category of stage of heart failure (A,B,C,D) if available. Page 5, paragraph 3 (first paragraph of discussion section): The statement regarding this study’s similarity to other findings which show educational attainment as the “most powerful predictor for patient activation” may led readers to believe that knowledge, efficacy, and behavior were compared and that education outweighed the others. I would consider re-wording and removing the term “most powerful.”. Page 5, paragraph 4, sentence 2 (second paragraph of discussion section): Can “confidence” be changed to “efficacy” here to keep language consistent? Page 5, paragraph 4, sentence 4: This sentence infers that the intervention in this study was to increase knowledge and efficacy to improve patient activation, but from my understanding this was a correlation and not an intervention.", "responses": [] } ]	1	https://f1000research.com/articles/4-150
https://f1000research.com/articles/4-149/v1	11 Jun 15	{ "type": "Research Note", "title": "Psoriasis is characterized by deficient negative immune regulation compared to transient delayed-type hypersensitivity reactions", "authors": [ "Nicholas Gulati", "Mayte Suárez-Fariñas", "Joel Correa da Rosa", "James G. Krueger", "Nicholas Gulati", "Mayte Suárez-Fariñas", "Joel Correa da Rosa" ], "abstract": "Diphencyprone (DPCP) is a hapten that causes delayed-type hypersensitivity (DTH) reactions in human skin, and is used as a topical therapeutic for alopecia areata, warts, and cutaneous melanoma metastases. We examined peak DTH reactions induced by DPCP (3 days post-challenge) by comprehensive gene expression and histological analysis. To better understand how these DTH reactions naturally resolve, we compared our DPCP biopsies to those from patients with psoriasis vulgaris, a chronic inflammatory disease that does not resolve. By both microarray and qRT-PCR, we found that psoriasis lesional skin has significantly lower expression of many negative immune regulators compared to peak DPCP reactions. These regulators include: interleukin-10, cytotoxic T lymphocyte-associated 4 (CTLA4), programmed cell death 1 (PD1), programmed cell death 1 ligand 1 (PDL1), programmed cell death 1 ligand 2 (PDL2), and indoleamine 2,3-dioxygenase (IDO1). Their decreased expression was confirmed at the protein level by immunohistochemistry. To more completely determine the balance of positive vs. negative immune regulators in both DPCP reactions and psoriasis, we developed one comprehensive gene list for positive regulatory (inflammatory) genes, and another for negative regulatory (immunosuppressive) genes, through Gene Ontology terms and literature review. With this approach, we found that DPCP reactions have a higher ratio of negative to positive regulatory genes (both in terms of quantity and expression levels) than psoriasis lesional skin. These data suggest that the disease chronicity that distinguishes psoriasis from transient DTH reactions may be related to absence of negative immune regulatory pathways, and induction of these is therefore of therapeutic interest. Further study of these negative regulatory mechanisms that are present in DPCP reactions, but not in psoriasis, could reveal novel players in the pathogenesis of chronic inflammation. The DPCP system used here thus provides a tractable model for primary discovery of pathways potentially involved in immune regulation in peripheral tissues.", "keywords": [ "psoriasis", "delayed-type hypersensitivity", "immune regulation", "diphencyprone" ], "content": "Introduction\n\nDiphencyprone (DPCP) is a hapten that induces delayed-type hypersensitivity (DTH) reactions in human skin, and is used therapeutically for alopecia areata1, warts2, and cutaneous melanoma metastases3. The mechanisms by which DPCP decreases pathogenic immunity for the promotion of hair growth in alopecia areata are incompletely understood. DPCP has been shown to alter the cytokine profile in treated alopecic scalp, in particular increasing interleukin (IL)-2 and IL-10 expression4. This increased IL-10 expression has been hypothesized to inhibit the lesional T cells of alopecia areata, but a comprehensive evaluation of other negative immune regulators induced by DPCP is lacking. We have previously shown that human skin responses to DPCP evolve from an inflammatory/effector peak at 3 days post-challenge to a more regulated immune response, with diminished markers of T cell activation, at 14 days. This study included comprehensive gene expression profiling, by microarray and qRT-PCR approaches, of biopsies from DPCP-challenged healthy volunteer skin at 3 days (peak reaction), 14 days (actively resolving reaction), and 120 days (4–8 months; fully resolved reaction) compared to placebo-treated skin5. We have also previously performed similar transcriptomic profiling of psoriasis vulgaris lesional vs. non-lesional skin. This resulted in a meta-analysis derived transcriptome (MAD3) which combined the results of 3 individual microarray experiments, in an effort to address the variability in differentially expressed genes observed between experiments6. In this study, we expand our previous characterizations of transient DTH reactions and chronic psoriasis biopsies by directly comparing them to each other, particularly in relation to positive and negative immune regulation.\n\n\nMethods\n\nFor diphencyprone (DPCP) reaction microarray, qRT-PCR, and immunohistochemistry studies, skin biopsies were obtained from 11 volunteers under a protocol approved by The Rockefeller University’s Institutional Review Board (IRB Number JKR-0742). Written, informed consent was obtained from all subjects and the study adhered to the Declaration of Helsinki principles. This trial is registered at clinicaltrials.gov under NCT01452594 (https://clinicaltrials.gov/ct2/show/NCT01452594). For each volunteer, biopsies were taken of placebo-treated skin as well as DPCP reactions 3, 14, and 120 days after challenge, as previously described5.\n\nFor psoriatic lesional vs. non-lesional skin microarray data, we used the meta-analysis derived (MAD3) transcriptome described in 6. Psoriatic lesional tissue for qRT-PCR and immunohistochemistry studies were from deidentified residual samples of plaque-type psoriasis vulgaris from previous studies for whom no clinical characteristics are available; a psoriasis area severity index of more than 12 (moderate-to-severe psoriasis vulgaris with >10% body surface area involvement) was required for entry into these trials.\n\nTotal RNA was extracted using the miRNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s protocol with on-column DNase digestion. The amount of RNA was assessed by NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific Inc., Wilmington, DE). The quality of extracted RNA was examined using Agilent Bioanalyzer 2100 (Agilent Technologies, Palo Alto, CA). RNA was hybridized to HGU133 Plus 2.0 chips (Affymetrix, Santa Clara, CA) to measure relative gene expression.\n\nStatistical analysis. Microarray data were analyzed using R/Bioconductor packages (http://www.r-project.org). The Harshlight package7 was used to scan Affymetrix chips for spatial artifacts. Expression values were normalized using the GeneChip Robust Multi-array Average (GCRMA) algorithm. Genes with low variation and low expression in most samples were filtered out prior to the analysis. Batch effect due to hybridization date was adjusted using ComBat8. To identify differentially expressed genes, we fitted by REML (Restricted Maximum Likelihood) a linear mixed effect model with treatment (placebo/DPCP) and day (3/14) as fixed effects and a random intercept for each patient. Hypotheses of interest were tested using contrasts in R’s limma package framework. The p-values resultant from the moderated paired Student’s t-tests were adjusted for multiple hypotheses using the Benjamini-Hochberg procedure, which controls for the false discovery rate. The DPCP data discussed in this publication have been deposited in the National Center for Biotechnology Information’s Gene Expression Omnibus (GSE accession number GSE52360, http://www.ncbi.nlm.nih.gov/geo/). Psoriasis data were derived from6.\n\nPre-amplification quantitative RT-PCR technique was used for measuring various genes in total RNA extracted from skin biopsy samples according to the company’s instructions. Briefly, 5 ng of total RNA was subjected to first-strand cDNA synthesis using High Capacity cDNA Reverse Transcription kits (Applied Biosystems, Carlsbad, CA). The resulting cDNA was subjected to 14 cycles of pre-amplification using TaqMan PreAmp Master Mix Kit (Applied Biosystems) with desired pooled assay mix. The Gene Amp PCR System 9700 (Applied Biosystems) was used for the pre-amplification reaction with the following thermal cycler conditions: 10 min at 95°C and 14 cycles of 15 seconds at 95°C followed by 4 min at 60°C. 12.5 μl of pre-amplified cDNA was then used for quantitative RT-PCR reaction using TaqMan Gene Expression Master Mix (Applied Biosystems). The 7900HT Fast Real-Time PCR System was used for PCR reactions, and the thermal cycler conditions were as follows: 2 minutes at 50°C, 5 minutes at 95°C, and 40 cycles of 15 seconds at 95°C followed by 60 seconds at 60°C. Data were analyzed by the Applied Biosystems PRISM 7700 software (Sequence Detection Systems, ver. 1.7) and normalized to human acidic ribosomal protein (hARP) housekeeping gene.\n\nAll assays were from Applied Biosystems and inventoried assays used in this study were as follows: IL10 (Hs00961622_m1), CTLA4 (Hs03044418_m1), PDCD1 (PD1) (Hs01550088_m1), CD274 (PDL1) (Hs01125301_m1), PDCD1LG2 (PDL2) (Hs01057777_m1), IDO1 (Hs00984148_m1), and LAG3 (Hs00158563_m1). For RPLP0/hARP, a custom primer/probe set was used (Forward: CGCTGCTGAACATGCTCAA, Reverse: TGTCGAACACCTGCTGGATG, Probe: 6-FAM-TCCCCCTTCTCCTTTGGGCTGG-TAMRA).\n\nFrozen sections of skin biopsies were dried at room temperature and then fixed for 2 minutes in acetone. Next, the samples were blocked with 10% normal serum of the species in which the secondary antibody was made and then the samples were incubated overnight at 4°C with the appropriate primary antibody. Biotin-labeled secondary antibodies (Vector Laboratories, Burlingame, CA) were amplified with avidin-biotin complex (Vector Laboratories) and developed with chromogen 3-amino-9-ethylcarbazole (Sigma Aldrich, St. Louis, MO) to produce a red color indicative of positive staining.\n\nPrimary antibodies used in this study are as follows (all are mouse monoclonal): IL10 (Life Technologies, Clone 945A2A5, IgG1, 1:50 dilution), CD95 (FAS) (BD Biosciences, Clone DX2, IgG1, 1:50), LAG3 (Enzo Life Sciences, Clone 17B4, IgG1, 1:50), PD1 (eBioscience, Clone MIH4, IgG1, 1:50), PDL1 (eBioscience, Clone MIH1, IgG1, 1:50), PDL2 (eBioscience, Clone MIH18, IgG1, 1:100), IDO1 (LifeSpan Biosciences, Inc., Clone 10.1, IgG3, 1:100), and CTLA4 (abcam, Clone BNI3, IgG2a, 1:100).\n\n\nResults\n\nSince DTH reactions naturally resolve, we sought to compare our DPCP biopsies (from 5) to those taken from patients with psoriasis vulgaris (from 6), a chronic T cell-mediated inflammatory disease that does not resolve and which, in many ways, represents amplifications of background immune circuits that exist in normal human skin9. To globally assess the balance of positive vs. negative immune regulators in both DPCP reactions and psoriasis using our microarray data, we developed one comprehensive gene list for positive regulatory or inflammatory genes and another gene list for negative regulatory or immunosuppressive genes (through Gene Ontology terms and literature review previously discussed in 5, Table 1 has “negative regulator” list and fold change values for DPCP day 3 and psoriasis transcriptomes, “positive regulator” list is derived from GO term 0002684 “positive regulation of immune system process” but with genes removed that are in common with GO term 0002683 “negative regulation of immune system process”). Our microarray data showed increased fold changes of many negative regulators in DPCP day 3 biopsies vs placebo-treated skin (DPCP day 3 transcriptome) compared to psoriasis lesional vs non-lesional skin (psoriasis transcriptome). For instance, CTLA4 expression was significantly increased 21.6-fold in the DPCP day 3 transcriptome, but non-significantly increased 3.7-fold in the psoriasis transcriptome. Venn diagrams show that the psoriasis transcriptome only has seven genes from the negative regulator list, while the DPCP day 3 transcriptome has 52 (Figure 1a). Although the DPCP day 3 transcriptome also has more genes from the positive regulator list than psoriasis, the odds ratio for the positive regulator list was not significantly different between these two transcriptomes. The odds ratio for the negative regulator list, however, was significantly different (Figure 1b). The altered balance between positive vs. negative regulatory transcripts in psoriasis compared to DPCP reactions can also be seen in Figure 1c which shows that DPCP transcriptomes at all time points (days 3, 14, and 120) have a higher ratio of negative to positive regulator genes than the psoriasis transcriptome in terms of expression levels for each gene set as a whole (as opposed to number of genes as indicated in the Venn diagrams). This is despite the fact that the DPCP day 3 transcriptome has comparable expression levels of the MAD3 psoriasis transcriptome genes to actual psoriasis samples, and therefore highlights the negative regulator expression that is unique to DPCP reactions.\n\nFCH, fold change; FDR, false discovery rate.\n\n(a) Venn diagrams showing overlap of MAD3 psoriasis transcriptome (left) and DPCP day 3 transcriptome (right) with both positive regulatory (Pos) and negative regulatory (Neg) gene lists (common gene lists applied to both transcriptomes). The percentages of the MAD3 and DPCP day 3 transcriptomes comprised of the positive regulatory gene list are 5.7% and 7.3%, respectively. On the other hand, the percentages comprised of the negative regulatory gene list are 0.7% and 1.5%, respectively. (b) Odds ratios (OR) of negative regulatory (red bars) and positive regulatory (blue bars) gene lists in DPCP day 3 and psoriasis transcriptomes. (c) Black bars represent negative regulator genes, gray bars represent positive regulator genes, and white bars represent all MAD3 psoriasis transcriptome genes. The y-axis shows log2 (fold change) of all genes in the given gene set. DPCP day 3 and MAD3 samples have comparable MAD3 transcriptome expression levels but there is a substantial difference between all DPCP time points (days 3, 14, and 120 or “late”) and MAD3 samples in terms of the relative levels of negative and positive regulator gene list expression. This is quantified as the “ratio of negative to positive regulator genes.”\n\nTo confirm some of our microarray findings, we performed qRT-PCR and found that psoriasis lesional skin biopsies have significantly lower expression of many negative immune regulators compared to peak DPCP biopsies. These regulators include lymphocyte activation gene 3 (LAG3), cytotoxic T lymphocyte-associated 4 (CTLA4), indoleamine 2,3-dioxygenase (IDO1), programmed cell death 1 (PD1), programmed cell death 1 ligand 1 (PDL1), programmed cell death 1 ligand 2 (PDL2), and IL-10 (Figure 2a). We confirmed the decreased expression of these and FAS (which by gene expression had 9.3- and 1.1-fold changes in the DPCP day 3 and psoriasis transcriptomes, respectively) at the protein level by immunohistochemistry (Figure 2b).\n\n(a) qRT-PCR analysis for negative regulators LAG3, CTLA4, IDO1, PD1, PDL1, PDL2, and IL-10. Shown are average normalized expression values for DPCP day 3 samples (n=11, purple bars) and psoriasis lesional skin (LS) samples (n=11, brown bars). All except PDL1 are p<0.05 by unpaired two-tailed t-test assuming equal variance. Error bars represent standard errors of the mean. (b) Immunohistochemistry showing increased protein expression of negative regulators in DPCP day 3 samples compared to psoriasis LS. Shown are stains with antibodies specific to the indicated targets. Scale bar = 100 μm (applies to all images).\n\n\nDiscussion/conclusions\n\nThese data suggest that disease chronicity in psoriasis could be related to absence of several negative immune regulatory pathways, with the implication that strategies to obtain stable clearance/restore tolerance in skin lesions may need to focus on increasing these negative pathways. These negative immune mechanisms may be of more general importance for maintaining skin homeostasis as non-inflammatory in the presence of a large population of effector memory T cells that normally reside in skin10. In addition, these negative immune regulators are likely involved in the therapeutic applications of DPCP, particularly alopecia areata where IL-10 has already been implicated4. Further study of these regulatory mechanisms that are present in DPCP reactions, but not in psoriasis, could reveal novel factors in the pathogenesis of chronic inflammation. The DPCP system used here provides a tractable model for primary discovery of pathways potentially involved in immune regulation in peripheral tissues.", "appendix": "Author contributions\n\n\n\nNG and JGK conceived the study and designed the experiments. NG carried out the research. MS-F and JCR contributed to the design of experiments and provided expertise in genomics. NG prepared the first draft of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by National Institutes of Health (NIH) grant UL1 RR024143 from the National Center for Research Resources and the Milstein Medical Program. NG was supported by a Medical Scientist Training Program grant from the National Institute of General Medical Sciences of the NIH under award number T32GM07739 to the Weill Cornell/Rockefeller/Sloan-Kettering Tri-Institutional MD-PhD Program. The content of this study is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nFreyschmidt-Paul P, Happle R, McElwee KJ, et al.: Alopecia areata: treatment of today and tomorrow. J Investig Dermatol Symp Proc. 2003; 8(1): 12–7. PubMed Abstract \| Publisher Full Text\n\nUpitis JA, Krol A: The use of diphenylcyclopropenone in the treatment of recalcitrant warts. J Cutan Med Surg. 2002; 6(3): 214–7. PubMed Abstract \| Publisher Full Text\n\nDamian DL, Saw RP, Thompson JF: Topical immunotherapy with diphencyprone for in transit and cutaneously metastatic melanoma. J Surg Oncol. 2014; 109(4): 308–13. PubMed Abstract \| Publisher Full Text\n\nHoffmann R, Wenzel E, Huth A, et al.: Cytokine mRNA levels in Alopecia areata before and after treatment with the contact allergen diphenylcyclopropenone. J Invest Dermatol. 1994; 103(4): 530–3. PubMed Abstract \| Publisher Full Text\n\nGulati N, Suárez-Fariñas M, Fuentes-Duculan J, et al.: Molecular characterization of human skin response to diphencyprone at peak and resolution phases: therapeutic insights. J Invest Dermatol. 2014; 134(10): 2531–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTian S, Krueger JG, Li K, et al.: Meta-analysis derived (MAD) transcriptome of psoriasis defines the “core” pathogenesis of disease. PLoS One. 2012; 7(9): e44274. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSuárez-Fariñas M, Pellegrino M, Wittkowski KM, et al.: Harshlight: a “corrective make-up” program for microarray chips. BMC Bioinformatics. 2005; 6: 294. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnson WE, Li C, Rabinovic A: Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics. 2007; 8(1): 118–27. PubMed Abstract \| Publisher Full Text\n\nLowes MA, Suárez-Fariñas M, Krueger JG: Immunology of psoriasis. Annu Rev Immunol. 2014; 32: 227–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClark RA, Chong B, Mirchandani N, et al.: The vast majority of CLA+ T cells are resident in normal skin. J Immunol. 2006; 176(7): 4431–9. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9426", "date": "13 Jul 2015", "name": "Jörg Prinz", "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\nChronic immune-mediated inflammatory diseases represent major challenges in medicine. The reasons for the perpetuation of inflammation are only incompletely understood. Explanations for the chronicity of immune-mediated inflammatory diseases are essentially required to initiate causal therapeutic strategies.The authors of the manuscript have chosen a novel approach to obtain insights into the molecular mechanisms associated with chronically perpetuating inflammation.For this purpose, they focus on two clearly defined T-cell mediated disorders serving as model diseases for inflammation: delayed-type i.e. T-cell mediated hypersensitivity reaction (DTH) in response to an obligatory hapten antigen, diphencyprone (DPCP), and psoriasis vulgaris, a chronic T-cell mediated inflammatory skin disease. Both diseases are particularly suited for this approach. DPCP-specific DTH reactions are immune mediated, resolve naturally and have well-defined self-limited kinetics. This allows for the accurate definition of time points for the analysis of inflammatory peak reactions and resolving inflammatory activity. Psoriasis instead constitutes a persistent T-cell mediated skin inflammation.The approach to obtain insights into the mechanisms perpetuating inflammation consists of the analysis and comparison of gene expression profiles in lesional and non-lesional skin probes of both diseases. Regarding psoriasis the authors refer to meta-analysis derived transcriptome data that cover the gene expression profiles from 193 lesional and non-lesional biopsy pairs as recently published. For DPCP-treated patients 11 lesional/non-lesional sample pairs were included for different time points during inflammation. Like the psoriasis transcriptome analyses the data acquisition in DCPC samples has been peer peer-reviewed in a high ranking scientific journal.To achieve meaningful results regarding the posed question, the authors define two groups of genes for comparison. One comprehensive gene list includes positive regulatory or inflammatory genes that enhance inflammation, the other genes particularly related to negative regulatory or immunosuppressive functions. The composition of the gene lists is based on scientifically verified Gene ontology terms. In addition to the microarray data, for select genes expression levels are verified by quantitative RT-PCR and by immunohistochemistry analyses of protein expression in skin samples.As a result the authors report, that at the time of the peak inflammatory response DPCP-induced DTH reactions contain significantly more genes related to suppression of inflammation than psoriasis lesions. This results in an altered balance between positive vs. negative regulatory transcripts in psoriasis compared to DTH reactions, which is maintained throughout DTH healing.The authors conclude that in psoriasis, but also beyond this disease the lack of negative immune regulatory genes may be related to inflammatory disease chronicity. They furthermore propose that DPCP reactions may serve to examine regulatory immune pathways.The study addresses a highly relevant question. The scientific approach is novel, original and represents a truly innovative solution to obtain novel insights into immune regulation. The comparison of psoriasis and DCPC reactions is reasonable because in both of them inflammatory reactions result from T-cell driven mechanisms. The methods are sound and appropriate for the analysis. The underlying data are extensive and due to the large sample size create a reliable and credible basis for the study. Accordingly, results and conclusions are well supported. The abstract provides an adequate summary of the study. The results are clearly presented. The discussion is short, draws a clear and valid conclusion and avoids unnecessary speculations. Indeed, the insights from the study provide a novel approach to understand the chronicity of inflammation. My personal assessment is that of an intriguing approach, which by itself is revealing and at the same time opens novel access paths into the investigation of chronic immune-mediated inflammation.I support the publication of this article in the present form without any reservation. There are no changes necessary.", "responses": [] }, { "id": "9246", "date": "16 Jul 2015", "name": "Wilson Liao", "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\nIn this welcome addition to the field of psoriasis transcriptomics, Gulati and colleagues sought to test whether expression of genes involved in immune regulation in transient condition, diphencyprone (DPCP) induced delayed-type hypersensitivity (DTH), was significantly different from expression of genes involved in immune regulation of a chronic T cell-mediated disease, psoriasis vulgaris. Employing microarray and qRT-PCR based approaches to profile the transcriptomes from biopsied samples, the authors found that negative immune regulatory genes were significantly overexpressed in the DPCP vs. healthy controls comparison, while such a pattern of overexpression was not observed in the psoriasis vs healthy controls comparison. This lack of overexpression of key negative regulatory genes was found in both the microarray and the qRT-PCR analyses. The authors conclude that the relative dearth of negative immune regulatory genes in psoriasis may be associated with the persistent inflammation found in psoriasis and that strategies to restore negative regulators may have therapeutic implications. This paper is technically sound, with well accepted protocol employed in all wet experiments. The statistical analysis was also performed with well accepted and sufficiently rigorous methodology. The evidence supports the authors’ central claims. P-values for differential expression analysis were adjusted for multiple hypotheses by the Benjamini-Hochberg method to control for the FDR. We believe that this paper will be of interest to researchers in the field of psoriasis transcriptomics and genetics, as well as the broader fields of autoimmunity and dermatology. The article could be strengthened by addressing the following comments and questions:The authors note that “genes with low variation and low expression in most samples were filtered out prior to the analysis”. We think that the filtering threshold for variance and expression should be reported as well. Odds ratios are reported in the results but it is not clear how they were estimated in the methods. Furthermore, interpretation of the odds ratios is unclear as well. In Table 1, we have questions regarding the FCH and the p and FDR values. Is the FCH the absolute FC? Is it log2? We think that the labeling should be clearer. What do p and FDR values of “0E+00” indicate? Is it for p and FDR values that are high or low? Could the authors’ results be confounded by the fact that psoriatic skin has a larger proportion of keratinocytes relative to DTH skin? If the negative regulators were primarily expressed in non-keratinocyte cell types, then biopsy of psoriatic skin compared to DTH skin might show a lower proportion of negative regulators based simply on cell proportions rather than intrinsic immunologic differences. We feel that the biology of psoriasis and DTH (or contact dermatitis) can be expanded upon in the discussion section. For instance, a number of the genes listed in Table 1 are known T regulatory cell signature genes (including CTLA4 and IL2RA; please see Ferraro et al., 2014). Is the transcriptional profile of the observed DTH reaction consistent with the effects of a T regulatory cell effect, for example as described in Rosenblum et al. (2011)? An apparent difference between the DTH model and psoriasis is that in the DTH model, antigen was transiently given, whereas in psoriasis it is possible that self-antigens are continually expressed. Might this contribute to the observed transcriptional differences? Have the authors’ examined the transcriptional profile of a DTH model in which antigen is chronically given?", "responses": [] }, { "id": "9578", "date": "20 Jul 2015", "name": "Eugene William St. Clair", "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\nIn this paper, Gulati and colleagues have tested the hypothesis that psoriatic skin lesions have significantly lower expression of negative immune regulators compared to peak DTH reactions from diphencyprone (DPCP). This study was done using two previously generated data sets described in previous publications from this group: 1) a study in 11 healthy human volunteers of recall responses to DPCP at 3-day (peak) and 14-day (resolution) using immunohistochemical techniques to visualize the infiltrating immune cells and gene expression (Affymetrix HGU133 Plus 2.0 Arrays) to perform a molecular profiling of the day 3 and day 14 DPCP responses, and in 6 cases 4-8 months later (reference 5); and 2) a meta-analysis derived (MAD) transcriptome of psoriatic skin from 3 different studies using Affymetrix HGU133 Plus 2.0 arrays in which microarray raw data were available through public databases (reference 6), termed the MAD-3 transcriptome. The results show an imbalance in the expression of negative regulatory genes in DPCP day-3 peak reactions compared to psoriatic skin lesions, favoring the expression of more negative immune regulators in the former vs. the latter group (52 genes vs 7 genes). They generated odds ratios for the positive and negative immune regulators from a list of negative regulatory and positive regulatory genes derived from a GO search. The odds ratio for negative immune regulators was significantly higher in the DPCP day-3 (peak) group than the psoriatic skin group, while the odds ratios for the positive immune regulators were not statistically significantly different between the two groups. The enrichment for negative immune regulators held true also at day-14 (resolving) and day 120 (late). These findings were confirmed by qRT-PCR of several key genes and by immunohistochemistry. This study is technically sound and the paper clearly written, with acceptable figures and tables. The title and the abstract are appropriate for the research described in the article. The methods are described in sufficient detail for the experiments to be replicated, although it would be useful to provide more detail about the methods underlying the calculations of the odds ratios. The interpretation of the results are reasonable based on the data and the conclusions provide a novel conceptual framework for understanding the fundamental mechanisms of skin homeostasis in psoriasis. I had one question about the results and their interpretation. In the study from reference 5, the results suggested different patterns of immune responses after DPCP challenge, referred to as subgroups A and B, where subgroup B had higher levels of negative immune regulators at day-3 than subgroup A. Interestingly, subgroup B fit the typical kinetics of a DTH response in which T cell and DC infiltrates peaked at day-3 and were diminished at day-14; whereas, subgroup A showed an unexpected increase in the numbers of T cells and DC’s at day-14 compared to day-3. It might be useful to know if the higher expression of negative immune regulators in DPCP skin compared with psoriatic skin was driven by the higher levels of these genes in subgroup B, or if it was generally true across both subgroups.", "responses": [] }, { "id": "9716", "date": "30 Jul 2015", "name": "Luis Puig", "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 paper under review provides interesting research data with immunopathogenic implications for both delayed-type hypersensitivity (DTH) reactions and psoriasis. Using both microarray and quantitative RT-PCR the authors found that peak DTH reactions induced by diphencyprone (DPCP) 3 days post-challenge has significantly higher expression of negative immune regulators than psoriasis lesional skin. These regulators include interleukin-10 (IL-10), cytotoxic T lymphocyte-associated 4 (CTLA4), programmed cell death 1 (PD1), programmed cell death 1 ligand 1 (PDL1), programmed cell death 1 ligand 2 (PDL2), and indoleamine 2,3-dioxygenase (IDO1), and their increased expression was confirmed at protein level by immunohistochemistry. According to the authors, these findings might explain at least in part the disease chronicity in psoriasis and might provide new therapeutic avenues.Of course, further studies are required to extricate the dynamics of immune activation in DTH vs psoriasis lesions; even data on DPCP reactions at 14 days can be found in another paper by the same authors (Gulati et al., 2014) detailed microarray and quantitative RT-PCR data on the time course of evolving lesions of psoriasis is lacking. Koebner phenomenon, guttate psoriasis and rebounds following clobetasol propionate suppression of plaque psoriasis might provide interesting models in this respect. The authors discuss the potential role of IL-10 regulation in DPCP immunotherapy of alopecia areata, and it might be interesting in this context to recall “Renbök phenomenon” or inverse Koebner phenomenon, namely the observation of normal hair growth in psoriatic plaques in patients with co-existing psoriasis and alopecia areata (Mirmirani, 2015).Pathogenic differences between DTH and psoriasis are not limited to the expression of genes involved in negative immune regulation, as illustrated by the lack of effect of blocking Th1/Th17 pathways with ustekinumab on the elicitation of ACD by patch testing (Nosbaum et al., 2012), even though IL-1 β activation of dendritic epidermal T cells and production of IL-17 seems to be important in experimental contact hypersensitivity (Nielsen et al., 2014). On the other hand, nickel-sensitized psoriasis patients develop a delayed but otherwise typical allergic contact dermatitis (ACD) even in close proximity to pre-existing psoriasis plaques, with a predominantly Th17 response in psoriasis and a stronger Th2 and cytotoxic immune response in ACD (Quaranta et al., 2014). Th9 cells are skin homing or skin resident cells which have been identified both in psoriasis and ACD, but IL-9 can have pro-inflammatory or regulatory activities depending on context (Liu et al., 2014). The innate immune responses to damage associated molecular patterns and reactive oxygen species are of foremost importance for the elicitation of ACD in vivo (Martin, 2015).Furthermore, the gene expression signatures and pathomechanisms of DTH reactions may differ according to the antigen (Dhingra et al., 2014) and duration of exposure. Nickel and other metal ions are able to elicitate ACD through the activation of toll-like receptor 4 (TLR4) (Schmidt et al., 2010), direct activation of the NLRP·inflammasome with production of IL-1β (Li et al., 2014) and other mechanisms (Vennegaard et al., 2014).Negative immune regulators include cytokines such as IL-10, IL-35 and transforming growth factor β (TGF β) and membrane proteins which are greatly interdependent and subject to cross-regulation, and several cell types, including Foxp3+ regulatory T cells, monocytes and B cells can be involved in down-regulation of inflammatory and autoimmune processes. Some of these cell types are characterized by their plasticity, so their phenotype might be subject to therapeutic modulation.In conclusion, the paper by Gulati and coworkers represents a seminal work in the study of the comparative dynamics of immune activation and downregulation in several skin disorders, with potential therapeutic implications.", "responses": [] } ]	1	https://f1000research.com/articles/4-149
https://f1000research.com/articles/4-21/v1	22 Jan 15	{ "type": "Research Article", "title": "Bias in peer review: a case study", "authors": [ "Richard Walker", "Beatriz Barros", "Ricardo Conejo", "Konrad Neumann", "Martin Telefont", "Beatriz Barros", "Ricardo Conejo", "Konrad Neumann", "Martin Telefont" ], "abstract": "Peer review is the \"gold standard\" for evaluating journal and conference papers, research proposals, on-going projects and university departments. However, it is widely believed that current systems are expensive, conservative and prone to various forms of bias. One form of bias identified in the literature is “social bias” linked to the personal attributes of authors and reviewers. To quantify the importance of this form of bias in modern peer review, we analyze three datasets providing information on the attributes of authors and reviewers and review outcomes: one from Frontiers - an open access publishing house with a novel interactive review process, and two from Spanish and international computer science conferences, which use traditional peer review. We use a random intercept model in which review outcome is the dependent variable, author and reviewer attributes are the independent variables and bias is defined by the interaction between author and reviewer attributes. We find no evidence of bias in terms of gender, or the language or prestige of author and reviewer institutions in any of the three datasets, but some weak evidence of regional bias in all three. Reviewer gender and the language and prestige of reviewer institutions appear to have little effect on review outcomes, but author gender, and the characteristics of author institutions have large effects. The methodology used cannot determine whether these are due to objective differences in scientific merit or entrenched biases shared by all reviewers.", "keywords": [ "Peer review", "bias", "nationality", "gender", "language", "prestige", "random intercept model", "authors", "reviewers" ], "content": "Introduction\n\nPeer review is the “gold standard” for the evaluation of journal and conference papers, research proposals, on-going projects and university departments and there is a strong consensus in the scientific community that it improves the quality of scientific publications1,2. As reported by Armstrong, “journal peer review is commonly believed to reduce the number of errors in published work, to serve readers as a signal of quality and to provide a fair way to allocate journal space”3. Surveys of authors and expert reviewers4–6 show that this view is widely held. However, many members of the scientific community also believe that peer review is expensive, conservative and prone to bias2,7–18. Critics point to the major delays it introduces into the publication process17,19, to biases against particular categories of papers (e.g. studies challenging conventional wisdom20; replication studies21,22 and studies reporting negative results12,23), to the unreliability of the review process23–25), to its inability to detect errors and fraud26, and to unethical practices by editors and reviewers27,28.\n\nAnother common criticism of peer review is that it is prone to “social bias”29: certain categories of reviewer may have conscious or unconscious biases for or against particular categories of author (e.g. authors of a particular gender, authors coming from institutions in a particular geographical area, authors from institutions in English-speaking or non-English speaking countries, authors belonging to low or high prestige institutions). Attempts to measure these biases have given contrasting results. For example, an experimental study by Lloyd30 shows that manuscripts with female author names have a far higher acceptance rate when they are reviewed by female rather than male reviewers (62% vs. 21%). Along the same lines, a widely quoted study of grant awards in Sweden31 suggests that proposals from male candidates receive systematically higher evaluations than those from female candidates with similar academic records, a result confirmed by a recent follow-up study15. Similarly, a study of the introduction of double blind review in the journal Behavioral Ecology reports that blinding reviewers to author gender led by an increase in papers with female first authors, absent in other very similar journals32. However, the interpretation of this study has been contested33, and some ex post analyses of publication patterns indicate that differences in the way male and female reviewers review papers do not affect review outcomes34. Other studies have found robust effects of author gender on peer review results but do not determine whether these are the result of bias or of differences in the scientific merit of articles submitted by male and female authors35.\n\nOther forms of social bias (e.g. bias for or against authors from particular geographical areas, language bias, bias in favor of authors from high prestige institutions) have been studied less frequently than gender bias but have also produced contrasting results, comprehensively reviewed by Lee and colleagues29. Several studies have shown, for instance, that journals favor authors located in the same country as the publisher36,37. Thus, a retrospective analysis by Link38 suggests that American reviewers are significantly more likely to accept a paper by another American author than a paper by an author of a different nationality. However, other studies indicate that American reviewers are actually more lenient to non-American than to American authors39. A study by Tregenza and colleagues40 shows that papers by authors from institutions in countries whose native language is English are more successful than papers by authors from institutions in other countries, an opinion shared by many commentators (for example41). However, the authors explicitly state that this is not necessarily a sign of bias and other studies have not found the same effect (e.g.42). Peters and Ceci43 report a quasi-experiment demonstrating that papers by authors from high-prestige institutions have a significantly higher chance of acceptance than similar papers by authors with less prestigious affiliations and participants in surveys of authors are reported to believe in such an effect44. However, to the knowledge of the authors, there is little observational evidence for this effect.\n\nAttempts to remedy the weaknesses of traditional peer review have led to a diversification of peer review practices, for instance through the use of author-blind and non-selective review, the removal of traditional reviewer anonymity, and the introduction of various forms of community review. To date, however, there have been few attempts to measure their effectiveness. Furthermore, many past studies of bias in peer review are relatively old and it is not clear whether or how far biases detected in the past have been affected by changes in social attitudes. In response to these concerns, we analyze data for the peer review systems used in Frontiers (an open access publishing house which uses a novel interactive review process), three computer science conferences (CAEPIA2003, JITEL2007, and SINTICE07) held in Spain between 2003 and 2007 and four international computer science conferences (AH2002, AIED2003, ICALP2002 and UMAP2011), held between 2002 and 2011.\n\nFrontiers is a large open access scientific publisher, which published its first journal in 2007. In January 2015, Frontiers had a portfolio of 49 open-access journals with over 50,000 researchers serving on its editorial boards and more than 380 academic specialty sections. Papers are published within these sections. Each paper is assigned to a scientist acting as the editor who coordinates the review process, and is responsible for publication/rejection decisions. Reviewers are selected automatically, based on the match between their individual specialties, and key words in articles submitted for review (or can also be assigned manually by the editor). During the review process, reviewers remain anonymous. However, accepted papers carry the names of their reviewers. This gives reviewers a strong incentive not to accept papers before they have reached a good level of quality.\n\nThe Frontier’s review system is designed not so much to select papers as to improve the quality of papers in a collaborative, interactive dialog between authors and reviewers. At the beginning of the review process, reviewers are asked to answer a series of open questions concerning different aspects of the paper (see Table 1). The precise set of questions depends on the nature of the paper (original research, review paper, etc.). Authors answer reviewer questions through an interactive forum. This possibility reduces misunderstandings and can significantly accelerate the submission process.\n\nIn addition to replying to open questions, reviewers can express their overall evaluation of the paper on a range of numerical scales (see Figure 1). However, the use of these scales is not mandatory. Nowhere in the process do papers receive an aggregate numerical score. The final decision to accept or reject a paper is taken by the journal editor, based on the overall results of the interactive review process. Acceptance rates are high. Of the papers in the Frontiers database that had reached a final publication/rejection decision on the date when we extracted the data, 91.5% were published and only 8.5% were rejected.\n\nThe conferences in these two datasets all used WebConf (http://WebConf.iaia.lcc.uma.es), a computerized system for managing the submission and review of conference papers. WebConf was developed by a team from Malaga University led by one of the authors (RC). WebConf implements a classical review process similar to the processes used by Springer, Elsevier and other large commercial publishers. Conference contributions are usually reviewed by three independent reviewers, occasionally by two or four. Reviewers are chosen by the conference program chair who draws on a database of potential experts called the program committee. In general, the program committee is made up of authors who have previously submitted papers in a particular area of research and have expressed their willingness to act as reviewers. The WebConf system suggests a list of potential reviewers based on the degree of matching between paper topics and reviewers field of expertise. The final selection is based on the judgment of the program chair.\n\nReviewers express their opinion of a paper in a conference-specific review form in which they assign scores to the paper on a number of separate scales, covering key areas of evaluation (typically including soundness, originality, clarity etc.) and textual comments. Scores on individual scales are usually expressed in terms of categories (typically: poor, fair, good, excellent). The final publication decision depends on the program chair. If one of the reviewers expresses a strongly negative view of a paper, it will typically be rejected. In cases where there is a very significant difference in reviewers’ opinions, the program chair can ask for an additional review. Acceptance rates vary between a minimum of 29.3% and a maximum of 87%. The combined acceptance rate for the seven conferences in the WebConf database was 57.9%.\n\n\nMethods\n\nFrontiers. The Frontiers database includes details of all authors and reviewers for all scientific papers submitted to Frontiers (N=8,565) between June 25, 2007 and March 19, 2012, the name of the journal to which the paper was submitted, the article type (review, original research etc.), the name and institutional affiliations of the authors and reviewers of specific papers, individual reviewer scores for the summary scales shown in Figure 1, and the overall review result (accepted/rejected). At the time of the analysis, 2,926 papers had not completed the review process and were excluded. In another 1,089 cases, reviewers had not assigned numerical scores to the paper, which could not therefore be considered. Our final analysis used 9,618 reviews, for 4,549 papers. Most of the papers in the database come from the life sciences. The majority of authors and reviewers come from Western Europe and Northern America. However, the database contains a substantial number of authors and reviewers from other parts of the world.\n\nSpanish computer science conferences (IEEE-Spain). This dataset includes details of 1,131 reviews referring to 411 papers submitted to three IEEE conferences (CAEPIA2003, SINTICE2007 and JTEL2007). The majority of authors and reviewers for these papers come from institutions in Spain and Portugal. The data provided include the name of the conference to which the contribution was submitted, the type of contribution (poster, short paper, full paper etc.) the name, gender and institutional affiliations of the authors and reviewers of specific contributions, individual reviewers scores and the final decision (accepted/rejected). All the papers in the database are in the area of computer science.\n\nInternational computer science conferences (IEEE-International). This dataset provides data for 2,194 reviews, referring to 793 papers submitted to four IEEE conferences (AH2002, AIED2003, ICALP2002 and UMAP2011), managed using WebConf and involving authors and reviewers from all over the world. This dataset provided the same data collected for IEEE-Spain.\n\nNames of authors and reviewers were canonized: accent and symbols were removed; double spaces replaced by single spaces, and upper-case characters replaced with lower-case characters. Names were rewritten in the normalized form <first name, last name>. Intermediate names were omitted.\n\nNames of institutions (universities, research institutions, companies) were canonized as above. After normalization, the name of the institution was recoded using the first three words in the full name.\n\nNeither the Frontiers nor the WebConf databases included data for author and reviewer gender. In the Frontiers case, gender was inferred semi-automatically in a multistep process. First we matched the first names contained in our database to an open source dictionary providing genders for more than 40,000 first names used in different countries (gender-1.0.0.tgz, downloadable from http://pecl.php.net/package/gender). We then used volunteers of different nationalities (Chinese, Egyptian, Indian, Japanese, Korean, Turkish) to assign genders to first names not contained in the dictionary. Additional names were assigned by manually searching for specific authors and reviewers on Google and Facebook. At the end of this process, we were able to assign genders to 9,472 out of 11,729 authors and reviewers. The majority of unassigned names were Asian (mainly Chinese). This is a potential source of bias in our analysis. Genders for WebConf authors and reviewers were inferred using a similar procedure.\n\nTo ascertain possible country and regional biases, we assigned each author and reviewer to the country of the institution to which they were affiliated. Actors with multiple affiliations were assigned to the country of the first affiliation listed. The USA, UK, Scotland, N. Ireland, Ireland, Australia, Canada, and New Zealand were classified as English-speaking countries. All other countries were classed as non-English speaking.\n\nAuthors’ and reviewers’ affiliated institutions were classified in terms of their position in the 2012 Shanghai academic ranking of world universities for the life sciences (http://www.shanghairanking.com/FieldLIFE2012.html) (Frontiers) and for computer sciences (http://www.shanghairanking.com/SubjectCS2012.html) (WEBCONF). Rank was coded as a categorical variable with values 1 (rank 1–50), 2 (rank 51–200), 3: unclassified. Given the process used to normalize institution names, it is possible that a small number of institutions were misclassified.\n\nFrontiers. The Frontiers review process produces a very low rejection rate. This means that the database used for our study contained relatively few rejected papers (N=478). To create a more informative indicator of reviewers’ evaluations, we computed for each paper the average of the scores expressed by each individual reviewer for the summary scales shown in Figure 1. A comparison between the distributions of scores for rejected and published papers (see Figure 2) clearly demonstrates the validity of the indicator. However, it should be noted that the indicator is a construction of the authors and played no role in the review process.\n\nWebConf. The WebConf system asks each reviewer to assign an overall score to the paper he/she has just reviewed. Scores are expressed on a scale of 0 to 10.\n\nFor the purposes of the study, we define bias as the interaction terms ∂ij in the random intercept model:\n\ny = b + µij + βijAi + γijRj + δijAiRj + ε (1)\n\nwhere y denotes the score given in the review, b denotes the random intercept, i indexes properties of authors, j indexes properties of reviewers, and ε is the error term. This method is similar but not identical to the method proposed in45.\n\nGiven a factor F, such as region, the variables Ai and Rj are indicator (dummy) variables indicating that the first author and reviewer belong to categories i and j of factor F, respectively, that is:\n\nIF one author belongs to category i of F, Ai=1, ELSE Ai=0 (2)\n\nAND\n\nIF one reviewer belongs to category j of F, Rj=1, ELSE Rj=0 (3)\n\nThus βij is the fixed effect of author category, γij is the fixed effect of reviewer category and δij is the fixed effect of the interaction between author and reviewer category. Since the expected value of the random intercept b is 0, the fixed effects allow us to estimate the following mean scores:\n\nGij = µij + βij + γij + ∂ij author in category i, reviewer in category j\n\nGijˇ = µij + βij author in category i, reviewer not in category j\n\nGiˇj = µiˇj + γij author category i, reviewer in category j\n\nGiˇjˇ = µij author not in category i, reviewer not in category j.\n\nWe define bias Bij of reviewers from category j of factor F towards authors from category i by the expression:\n\nBij = (Gij – Gij ˇ) – (Giˇj – Giˇjˇ) = ∂ij (4)\n\nSince the intercept and main effects cancel out, bias is the interaction term ∂ij and does not depend on the main effects βij and γij. In other words, it is independent of any general tendency of authors in category i to write better papers than other authors, or of any tendency of reviewers in category j to give generally higher scores. In this setting, reviewers from category j are biased in favor of authors from category i, if Bij >0 and are biased against authors from category i if Bij <0. Bias is significant at a level α if we can reject the null hypothesis:\n\nH0 : Bij = 0\n\nOtherwise we assume absence of bias.\n\nThe majority of papers in our databases had multiple authors. In preliminary studies, we explored statistical models that used this data in different ways: (i) the model used only the properties of the first author, (ii) the model used only the properties of the last author, (iii) the model considered the properties of all the authors. The three approaches yielded similar results (data not shown). In what follows, we apply the first method, unless otherwise stated.\n\nTo illustrate the concept of bias defined in (4), consider a factor with two levels such as gender. Let i and j denote female (F), and iˇ and jˇ male (M). Then the terms Gij = GFF, Gijˇ = GFM, Giˇj = GMF, and Giˇjˇ = GMM, have the following meanings:\n\nGFF : mean score when first author and reviewer are female\n\nGFM : mean score when first author is female and reviewer is male\n\nGMF : mean score when first author is male and reviewer is female\n\nGMM : mean score when first author and reviewer are male\n\nIf we assumed that papers by female authors have the same quality as papers by male authors,\n\nGFF – GFM > 0 (5)\n\nwould imply that female reviewers are biased in favor of female authors and\n\nGMF – GMM > 0 (6)\n\nwould imply that female reviewers are biased in favor of male authors.\n\nHowever, we cannot make this assumption. We therefore conclude that female reviewers are biased for or against female authors (BFF ≠ 0) only if (5) and (6), yield different results. If both were equal and positive, we conclude that females give higher scores than men regardless of the gender of the author. If female reviewers have a positive bias, this always implies a negative bias on the part of male reviewers and vice versa. By construction, this method cannot detect biases shared by all reviewers (e.g. a bias against female authors, or authors from a particular geographical region, shared by all reviewers, regardless of gender or geographical origin).\n\nIt should also be noted that, in our example, if B = (GFF – GFM) – (GMF – GMM) > 0 the following statements are equivalent:\n\nFemale reviewers are biased in favor of female authors\n\nFemale reviewers are biased against male authors\n\nMale reviewers are biased against female authors\n\nMale reviewers are biased in favor of male authors\n\nSimilarly, if B > 0, we can draw the following equivalent conclusions:\n\nFemale reviewers are biased against female authors\n\nFemale reviewers are biased in favor of male authors\n\nMale reviewers are biased in favor of female authors\n\nMale reviewers are biased against male authors\n\n\nResults\n\nThe automatic gender assignment program assigned genders to first authors and reviewers for 8,114 reviews from Frontiers, 1,131 from IEEE (Spain) and 2,194 from IEEE (International). The relative proportions of male first authors and reviewers (male authors: 70.0%–73.9%; male reviewers: 75.2–79.3%) were similar in all three datasets. In the Frontiers and IEEE (International) datasets, means scores for male first authors were significantly higher than those for female first authors (Frontiers: difference=0.07, p=0.034; IEEE International: difference=0.28, p=0.001). The IEEE (Spain) dataset showed the same pattern but the difference was not significant (difference=0.39, p=0.40). Reviewer gender had no significant effect on review scores in any of the datasets. In none of the data sets did the interaction between author and reviewer gender have a significant effect. In brief, none of the datasets showed evidence of gender bias. The significance of these results will be examined in the discussion. Complete data for the analysis can be found in Data Files 1–4.\n\nThe analysis examined the role of the region of first author and reviewer institutions in determining review scores and tested for possible regional bias. Authors and reviewers were grouped into 11 geographical regions (Africa, Australia/New Zeeland, Central America/Caribbean, Central Asia, Eastern Asia, Eastern Europe, Middle East/North Africa, South America, Southern Asia, Southern Europe, and Western Europe) according to the location of their respective institutions. To avoid problems with the convergence of the mixed model algorithm and to guarantee the statistical power of the analysis, pairs of first author/reviewer regions with less than 30 reviews were discarded. Distributions of author and reviewer regions differed significantly among the three datasets. In the Frontiers and International IEEE datasets, the majority of authors and reviewers came from institutions in North America and Western Europe, while the majority of authors and reviewers in the Spanish IEEE dataset, came from institutions in Southern Europe.\n\nAll three datasets showed large differences in the scores of first authors from different regions (see Table 2), which were statistically significant even after correction for multiple hypothesis testing. In the Frontiers dataset, authors from N. America scored significantly higher than authors from all other regions, whereas authors from E. Asia, E. Europe, S. Asia and Southern Europe scored significantly lower. In the IEEE (Spain) dataset, authors from Southern Europe scored higher than authors from other regions, whereas authors from North America scored lower. In the IEEE (international dataset) authors from N. America again scored significantly higher than authors from other regions, while authors from Africa and Central Asia scored lower.\n\nIn the Frontiers and IEEE (Spain) datasets, we detected no significant differences in scores from reviewers from different regions (not shown). In the IEEE (International) dataset reviewers from Australia/NZ and from Southern Asia gave scores that were significantly higher than the average for reviewers from other regions and reviewers from Western Europe gave scores that were significantly lower (not shown).\n\nTo test for bias, we applied the random intercept model to all author/review region pairs with more than 30 reviews (see Table 3). After application of the Bonferroni correction for multiple hypothesis testing, the Frontiers dataset showed no evidence of interaction between author and reviewer region and the other two datasets showed only very limited evidence (IEEE – Spain: bias of reviewers from S. Europe in favor of authors from E. Asia; IEEE – International: bias of North American reviewers in favor of authors from Eastern Asia). In none of the datasets did we find any evidence for regional biases previously reported in the literature (e.g. bias of North American reviewers in favor of North American authors). We conclude that regional bias has only a limited effect on review scores and that those biases that do exist are idiosyncratic to particular review systems. Full data for the analysis can be found in Data Files 5–8.\n\nWe hypothesized that reviewers could be biased against papers written by authors who were not native English speakers. We, therefore, analyzed potential reviewer bias for and against papers, whose first authors came or did not come from institutions in English-speaking countries. As a further test, we analyzed potential bias for and against papers, which had, or did not have, at least one author belonging to an institution in an English speaking country.\n\nThe Frontiers and the IEEE (International) datasets, both included large numbers of authors and reviewers, from institutions in native English-speaking and from non-English-speaking countries. In contrast, approximately 97% of the authors and reviewers in the IEEE (Spain) dataset came from Spain. Since none of the papers with an English-language first author, and only one paper with at least one English author, were reviewed by an English-language reviewer, it was not possible to measure bias using the random intercept model. This dataset was therefore discarded from the subsequent analysis.\n\nIn the remaining datasets, papers with a first author from an institution in a non-English-speaking country scored significantly lower that papers with first authors from institutions in an English speaking country, regardless of whether they were reviewed by native English-speaking or non-native English speaking reviewers (Frontiers: difference=-0.22, p<0.001; IEEE International: difference=-0.54, p<0.001). Reviewer language had no significant effect on score (Frontiers: difference=0.04, p=0.040; IEEE International: difference=0.01; p=0.80). In neither case did we find a significant interaction between author and reviewer language (not shown). Results for papers with at least one author from an institution in an English-speaking country were similar. Details of the analysis are shown in Data Files 9–11.\n\nReviewers from institutions with high academic prestige could be biased in favor of authors from other high prestige institutions and against authors from lower ranking institutions. To test this possibility, we classified all authors and reviewers in the three datasets by the position of their institutions in the Shanghai classifications, as described earlier.\n\nThe Frontiers and the IEEE International datasets both contained significant numbers of authors and reviewers from universities in all three categories. However, nearly all the authors and reviewers in the IEEE Spain dataset came from universities in category 3. Given the lack of data for authors and reviewers from higher-ranking institutions, this dataset was excluded from the subsequent analysis.\n\nIn the Frontiers dataset, authors from universities in category 1 scored significantly higher than authors from category 2 (difference=0.16, p=0.016) and from category 3 (difference=0.21, p<0.001), regardless of the origin of the reviewer. No significant difference was observed between the scores of authors in category 2 and 3 or between scores given by reviewers from institutions in different categories. In the IEEE International dataset, authors from universities in categories 1 and 2 both scored significantly higher than authors in category 3 (category 1: difference in scores=0.97, p=<0.001; category 2: difference in scores=0.89, p=<0.001) but there was no significant difference in their own respective scores (not shown). There was no significant difference between the scores given by reviewers in category 1 and the scores given by reviewers in categories 2 and 3. However, reviewers in category 2 gave significantly higher scores than reviewers in category 3 (difference in scores=0.28, p<0.001). Neither dataset showed a significant interaction between the prestige of author and reviewer institutions. In the IEEE (International) dataset, reviewers from institutions in category 1 gave higher scores to authors in category 3 than to authors in category 1, but the difference was not statistically significant (difference=0.48, p=0.059). Full details of the analysis are given in Data Files 12–14.\n\n\nDiscussion\n\nThe results of the study (see Table 4) show that the scores received by papers in peer review depend strongly on the characteristics of the first author (gender, geographical location, language and prestige of the author’s institution). In summary, male authors receive higher scores than female authors, authors from some geographical regions receive higher scores than authors from others; authors from institutions in English-speaking countries receive higher scores than authors in non-English-speaking countries; authors from high prestige institutions receive higher scores than authors from lower-prestige institutions. In contrast, we find no evidence that scores are affected by the personal characteristics of reviewers, no significant interactions between author and reviewer gender, language, and institutional prestige and only limited evidence of an interaction between author and reviewer region. In brief, the study provides little evidence for bias, at least in the sense in which bias is defined in our study (see below).\n\nThe review systems considered in the study are very different. The majority of papers in the Frontiers dataset came from the life sciences; all the papers in the IEEE datasets were from specialized areas of computer science. Frontiers adopts a novel interactive review process; the conferences in the IEEE (Spain) and the IEEE (International) datasets used a traditional approach. Authors and reviewers in the Frontiers and the IEEE International datasets come from all over world. The IEEE (Spain) dataset is dominated by authors and reviewers from Southern Europe. Despite these differences, analysis of the three datasets gave very similar results. This suggests that the findings of this study could be valid for a broad range of peer review systems. The large size of the datasets used in the analysis (in total 12,943 reviews of 5,743 papers) provides additional evidence of robustness. The main differences between the datasets were in their patterns of regional bias, which are different in each dataset. Unfortunately the many differences between the Frontiers and the IEEE systems make it impossible to untangle the roles of different contributory factors.\n\nThe finding that author characteristics have a significant effect on review scores is compatible with two distinct explanatory hypotheses. The first is that papers submitted by authors with a particular characteristic (e.g. authors from institutions in a particular region) are, on average, of higher scientific merit than papers by authors with different characteristics (e.g. authors from institutions in other regions). The second is that reviewers share a general bias against authors with particular characteristics, regardless of their own characteristics (e.g. reviewers from institutions in English and non-English speaking countries share a bias against authors from non-English speaking countries). The methodology of the study cannot distinguish between these hypotheses; in fact, the only way to detect generalized bias would be through experimental studies, comparing review scores when reviewers are blinded to particular characteristics of the author to the scores given when they are not (for example30,43). Such studies are extremely valuable. However, their experimental nature means that they are unable to demonstrate the existence or absence of bias in operational review systems. This suggests that observational and experimental methods are complementary, and that it will not be possible to gain a complete picture of bias in peer review without using both methods.\n\nThe study found no evidence for social bias in terms of author gender, and the language and prestige of author institutions and only weak evidence of regional bias. The findings for gender match results from a previous study which used the interaction between author and reviewer characteristics as a measure of bias46. However, it apparently contradicts results from previous studies showing significant bias, with respect to gender (e.g.30,43) as well as for region (e.g.38), and institutional prestige (e.g.43). Given that the majority of studies showing bias are relatively old, it is possible that changes in social attitudes have reduced or eliminated some of the biases they observed. In some cases, (e.g.31) studies measured effects that were independent of reviewer characteristics, which, as explained earlier, are invisible to the methodology used in this study. In this case, the results of our study would complement rather than contradict previous results. We suggest, nonetheless, that, at least in the case of gender, it is implausible that modern female reviewers are biased against female authors. The most parsimonious explanation of our results is that, with the possible exception of regional biases, social bias plays at most a minor role in determining review outcomes.\n\nOur study also found no significant differences between the scoring patterns of different categories of reviewer. These results, which were valid for all three datasets, contrast with previous findings showing significant differences in scoring practices between male and female reviewers34,47 and between reviewers from different countries47. However, there have been relatively few studies on this issue, and even these do not show a major impact of reviewer characteristics on the final outcomes of the review process34. Taken together, these results suggest that editors should not be over-concerned with the gender, language or institutional affiliation of the reviewers they choose for particular papers, though it could be useful to ensure a good regional balance among reviewers.\n\nOur study does not evaluate the full set of potential biases described in the peer review literature. For instance, we do not consider confirmation bias or alleged reviewer biases in favor of positive results, sophisticated experimental and statistical methodology, or against interdisciplinary studies, replication studies, etc. These are important limitations. A second limitation is that the study methodology has so far been tested with just three peer review systems, all applying to scientific papers. It is possible that other forms of peer review, such as peer review of grant applications, are subject to different forms of bias.\n\nIn conclusion, our study shows the important role of authors’ personal characteristics in determining the scores received in peer review, but finds little evidence for bias due to interactions between author and reviewer characteristics. These findings do not rule out generalized bias against authors with specific characteristics or forms of bias not considered in the study.\n\n\nData availability\n\nTo protect the identities of authors and reviewers, the source data used for this study have not been made public.", "appendix": "Author contributions\n\n\n\nRW, RC, BB and KN conceived the study; RW wrote the paper with contributions from KN and RC; MT prepared the Frontiers data for use in the analysis and implemented ad hoc software for this purpose; BB and RC prepared the Spanish and the International Computer Science Conference datasets; KN designed and implemented the statistical analysis. All authors contributed critical comments and agreed to the final content of the article.\n\n\nCompeting interests\n\n\n\nFrontiers was a partner in the SISOB project and contributed to the research described in this paper. The first author (RW) is a part-time employee of Frontiers.\n\n\nGrant information\n\nThe research described in this paper was supported by a grant from the European Union’s Seventh Framework Programme (FP7/2007-2013) under Grant Agreement 7PM- 266588 (SISOB).\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors gratefully acknowledge the technical support and constructive criticism received from members of the Frontiers team, in particular Fred Fenter, Kamila Markram, and Costanza Zucca. We further acknowledge the contributions of participants in SISOB meetings, where we presented preliminary versions of this study, in particular those from Ulrich Hoppe and Sandor Soos. A preliminary version of this paper was incorporated in “SISOB Deliverable 9.3: Study of Enhanced Evaluation”.\n\n\nReferences\n\nNature: Pros and cons of open peer review. Nat Neurosci. 1999; 2(3): 197–198. PubMed Abstract \| Publisher Full Text\n\nWare M: Peer review: benefits, perceptions and alternatives. Publishing Research Consortium. 2008. Reference Source\n\nArmstong JS: Management Science: What Does It Have to do With Management or Science? Mark Bull. 1998; 9: 1–13. Reference Source\n\nBradley JV: Pernicious publication practices. Bull Psychon Soc. 1981; 18(1): 31–34. Publisher Full Text\n\nFletcher RH, Fletcher SW: Evidence for the effectiveness of peer review. Sci Eng Ethics. 1997; 3(1): 35–43. Publisher Full Text\n\nMacNealy MS, Speck BW, Clements N: Publishing in technical communication journals from the successful author’s point of view. Tech Commun. 1994; 41(2): 240–259. Reference Source\n\nNature: Can peer review police fraud? Nat Neurosci. 2006; 9(2): 149. PubMed Abstract \| Publisher Full Text\n\nJefferson T, Rudin M, Brodney Folse S, et al.: Editorial peer review for improving the quality of reports of biomedical studies. Cochrane Database Syst Rev. 2007; (2): MR000016. PubMed Abstract \| Publisher Full Text\n\nBohannon J: Who’s afraid of peer review? Science. 2013; 342(6154): 60–65. PubMed Abstract \| Publisher Full Text\n\nBornmann L: Scientific peer review. Annu Rev Inf Sci Technol. 2011; 45(1): 197–245. Publisher Full Text\n\nEllison G: Is peer review in decline? Econ Inq. 2011; 49(3): 635–657. Publisher Full Text\n\nEmerson GB, Warme WJ, Wolf FM, et al.: Testing for the presence of positive-outcome bias in peer review: a randomized controlled trial. Arch Intern Med. 2010; 170(21): 1934–9. PubMed Abstract \| Publisher Full Text\n\nSmith R: Classical peer review: an empty gun. Breast Cancer Res. 2010; 12(Suppl 4): S13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSuls J, Martin R: The air we breathe: A critical look at practices and alternatives in the peer-review process. Perspect Psychol Sci. 2009; 4(1): 40–50. Publisher Full Text\n\nSandström U, Hällsten M: Persistent nepotism in peer-review. Scientometrics. 2008; 74(2): 175–189. Publisher Full Text\n\nAlberts B, Hanson B, Kelner KL: Reviewing peer review. Science. 2008; 321(5885): 15. PubMed Abstract \| Publisher Full Text\n\nBenos DJ, Bashari E, Chaves JM, et al.: The ups and downs of peer review. Adv Physiol Educ. 2007; 31(2): 145–152. PubMed Abstract \| Publisher Full Text\n\nSmith R: Peer review: a flawed process at the heart of science and journals. J R Soc Med. 2006; 99(4): 178–182. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArmstrong JS: Peer Review for Journals: Evidence on Quality Control, Fairness, and Innovation. Sci Eng Ethics. 1997; 3(1): 63–84. Publisher Full Text\n\nCampanario JM: Rejecting and resisting Nobel class discoveries: accounts by Nobel Laureates. Scientometrics. 2009; 81(2): 549–565. Publisher Full Text\n\nKerr S, Tolliver J, Petree D: Manuscript Characteristics Which Influence Acceptance for Management and Social Science Journals. Acad Manage J. 1977; 20(1): 132–141. Publisher Full Text\n\nCampanario JM: Peer review for journals as it stands today-Part 1. Sci Commun. 1998; 19(3): 181–211. Publisher Full Text\n\nMahoney MJ: Publication prejudices: An experimental study of confirmatory bias in the peer review system. Cognit Ther Res. 1977; 1(2): 161–175. Publisher Full Text\n\nKravitz RL, Franks P, Feldman MD, et al.: Editorial peer reviewers’ recommendations at a general medical journal: are they reliable and do editors care? PLoS One. 2010; 5(4): e10072. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerron DM: Is expert peer review obsolete? A model suggests that post-publication reader review may exceed the accuracy of traditional peer review. Surg Endosc. 2012; 26(8): 2275–2280. PubMed Abstract \| Publisher Full Text\n\nSchroter S, Black N, Evans S, et al.: Effects of training on quality of peer review: randomised controlled trial. BMJ. 2004; 328(7441): 673. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTravis GDL, Collins HM: New Light on Old Boys: Cognitive and Institutional Particularism in the Peer Review System. Sci Technol Hum Values. 1991; 16(3): 322–341. Publisher Full Text\n\nSieber JE: Improving the peer-review process relies on understanding its context and culture. Nature. 2012. Publisher Full Text\n\nLee CJ, Sugimoto CR, Zhang G, et al.: Bias in peer review. J Am Soc Inform Sci Technol. 2013; 64(1): 2–17. Publisher Full Text\n\nLloyd ME: Gender factors in reviewer recommendations for manuscript publication. J Appl Behav Anal. 1990; 23(4): 539–543. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWenneras C, Wold A: Nepotism and sexism in peer-review. Nature. 1997; 387(6631): 341–343. PubMed Abstract \| Publisher Full Text\n\nBudden AE, Tregenza T, Aarssen LW, et al.: Double-blind review favours increased representation of female authors. Trends Ecol Evol. 2008; 23(1): 4–6. PubMed Abstract \| Publisher Full Text\n\nWebb TJ, O’Hara B, Freckleton RP: Does double-blind review benefit female authors? Trends Ecol Evol. 2008; 23(7): 351–353. PubMed Abstract \| Publisher Full Text\n\nGilbert JR, Williams ES, Lundberg GD: Is there gender bias in JAMA’s peer review process? JAMA. 1994; 272(2): 139–142. PubMed Abstract \| Publisher Full Text\n\nBornmann L, Mutz R, Daniel HD: Gender differences in grant peer review: A meta-analysis. J Informetrics. 2007; 1(3): 226–238. Publisher Full Text\n\nDaniel HD, Russey WE: Guardians of science: Fairness and reliability of peer review: VCH Weinheim. 1993. Publisher Full Text\n\nErnst E, Kienbacher T: Chauvinism. Nature. 1991; 352: 560. Publisher Full Text\n\nLink AM: US and non-US submissions: an analysis of reviewer bias. JAMA. 1998; 280(3): 246–247. PubMed Abstract \| Publisher Full Text\n\nMarsh HW, Jayasinghe UW, Bond NW: Improving the peer-review process for grant applications: reliability, validity, bias, and generalizability. Am Psychol. 2008; 63(3): 160–8. PubMed Abstract \| Publisher Full Text\n\nTregenza T: Gender bias in the refereeing process? Trends Ecol Evol. 2002; 17(8): 349–350. Publisher Full Text\n\nHerrera AJ: Language bias discredits the peer-review system. Nature. 1999; 397(6719): 467. PubMed Abstract \| Publisher Full Text\n\nLoonen MP, Hage JJ, Kon M: Who benefits from peer review? An analysis of the outcome of 100 requests for review by Plastic and Reconstructive Surgery. Plast Reconstr Surg. 2005; 116(5): 1461–1472. PubMed Abstract \| Publisher Full Text\n\nPeters DP, Ceci SJ: Peer-review practices of psychology journals: The fate of published articles, submitted again. Behav Brain Sci. 1982; 5(2): 187–195. Publisher Full Text\n\nGillespie GW, Chubin DE, Kurzon GM: Experience with NIH peer review: researchers’ cynicism and desire for change. Sci Technol Hum Val. 1985; 10(3): 44–54. Publisher Full Text\n\nBornmann L, Mutz R, Daniel HD: How to detect indications of potential sources of bias in peer review: A generalized latent variable modeling approach exemplified by a gender study. J Informetrics. 2008; 2(4): 280–287. Publisher Full Text\n\nJayasinghe UW, Marsh HW, Bond N: A multilevel cross-classified modelling approach to peer review of grant proposals: the effects of assessor and researcher attributes on assessor ratings. J R Stat Soc Ser A Stat Soc. 2003; 166(3): 279–300. Publisher Full Text\n\nBorsuk RM, Aarssen LW, Budden AE, et al.: To name or not to name: The effect of changing author gender on peer review. BioScience. 2009; 59(11): 985–989. Publisher Full Text" }	[ { "id": "7425", "date": "02 Feb 2015", "name": "Lutz Bornmann", "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\nPeer review is the most important instrument for assessing scientific research. However, the instrument is not without drawbacks. As the most important weaknesses, a missing reliability, fairness and predictive validity have been seen1. The study of Walker, Barros, Conejo, Neumann and Telefont (2015) deals with the fairness of the peer review process: They investigated “social biases” in the processes of Frontiers - an open access publishing house with a novel interactive peer review process, and two peer review processes from Spanish and international computer science conferences.The study is very interesting. I recommend that the authors revise the manuscript according to the following points:On page 5, Walker, et al. describe the process of normalizing authors’ and reviewers’ names. Here, Walker, et al. should ensure that the names are completely cleaned: the same author and the same institutional unit should receive the same name. A general problem of this kind of data from peer review processes is that they are not cleaned and contain several name variants. Walker, et al. present their results with the reporting of statistical significance information. I recommend that not only the statistical, but also the practical significance of the results (effect sizes) should be reported.2,3,4 In the Discussion section the following major limitation of the study should be mentioned: The quality of the papers was not controlled. For example, Bornmann and Daniel (2009)5 and Bornmann and Daniel (2010)6 investigated the peer review processes of the Angewandte Chemie – International Edition and Atmospheric Chemistry and Physics. They considered citations for the single papers as a proxy for quality. Although citations measure only one part of quality (namely impact), it is more favorable to consider them than doing not. When examining the association of bias variables and peer review outcomes without controlling quality it is impossible to establish unambiguously whether a particular group of papers receives more favorable recommendations or decisions due to these variables, or if the more favorable recommendations and decisions are simply a consequence of the papers’ scientific quality.", "responses": [ { "c_id": "1218", "date": "12 Feb 2015", "name": "Richard Walker", "role": "Author Response", "response": "I think these comments are very relevant and we will take them into account in the next version of our paper. As concerns the specific points you raise:In our study, authors' and reviewers' names are relevant for gender assignment. Institution names are critical for assignment of language, region, and institutional prestige to author and reviewer institutions. Informal checks on our cleaning process suggest that it does not introduce substantial errors into our analysis. In the next version of our paper we will introduce more formal checks. We agree with the reviewer on this point. The next version of our paper will include estimates of the practical significance of our results. We completely agree with the reviewer that some differences in scores may be due to differences in scientific quality and tried to make this point in our text. In the next version, we will attempt to clarify this issue." }, { "c_id": "1396", "date": "10 Jun 2015", "name": "Richard Walker", "role": "Author Response", "response": "Thank you for your comments, which we found extremely useful and constructive. Particularly useful was your suggestion that we should include effect sizes in our analysis. This suggestion has been implemented in this revised version of our paper. We would also like to thank you for your suggestions regarding quality control, complementing similar suggestions from Jigisha Patel. Thanks to these suggestions, we found a number of problems with the data, which had previously passed unnoticed. All these changes have been incorporated in the new version of the paper which we have just submitted.On page 5, Walker, et al. describe the process of normalizing authors’ and reviewers’ names. Here, Walker, et al. should ensure that the names are completely cleaned: the same author and the same institutional unit should receive the same name. A general problem of this kind of data from peer review processes is that they are not cleaned and contain several name variants.Thank you this comment, which, together with comments in a similar vein from Jigisha Patel led us to conduct a thorough review of our data. The review found that errors in the normalization of author, reviewer and institution names and missing values led to down-stream errors in automated gender assignment and in the assignment of university rankings. In the case of the university rankings we were able to correct a number of errors. In the case of the gender assignment, this would have been too onerous to be possible. Instead, as suggested by Jigisha Patel, we used sampling to check the error rates in our data, which we report in our text, together with a discussion of their significance. Walker, et al. present their results with the reporting of statistical significance information. I recommend that not only the statistical, but also the practical significance of the results (effect sizes) should be reported.2,3,4Thank you for this suggestion, which we have implemented in the revised version of our manuscript. We began by calculating Cohen's d for all our data. In the results and the discussion sections, we analyse the practical significance of the effects observed both in terms of the number of reviews concerned (which was always small) and the effects on publication decisions. In the cases we study, this was not large. However, in more selective review systems it could be larger. In the Discussion section the following major limitation of the study should be mentioned: The quality of the papers was not controlled. For example, Bornmann and Daniel (2009)5 and Bornmann and Daniel (2010)6 investigated the peer review processes of the Angewandte Chemie – International Edition and Atmospheric Chemistry and Physics. They considered citations for the single papers as a proxy for quality. Although citations measure only one part of quality (namely impact), it is more favourable to consider them than doing not. When examining the association of bias variables and peer review outcomes without controlling quality it is impossible to establish unambiguously whether a particular group of papers receives more favourable recommendations or decisions due to these variables, or if the more favourable recommendations and decisions are simply a consequence of the papers’ scientific qualityWe now make explicit reference to this limitation in our text. We go on to explain that most of the papers in our datasets have very few citations. This means, that in our case, we are unable to use citations as a proxy for quality." } ] }, { "id": "7428", "date": "13 Feb 2015", "name": "Jigisha Patel", "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 analyses the relationship between the attributes of authors and reviewers and reviewer outcomes in three datasets, one an innovative peer-review model and two which use traditional peer review. They find no evidence of gender or institutional bias and limited interaction between author and reviewer region.Although social bias in peer review has been investigated in the past, I think this study is timely and the question of whether this previously found bias still exists with changing social norms and values is interesting. However, there is a need for some clarification of the methods before the significance of the authors’ findings can be determined.Major revisionsThe authors are very clear that their study investigates ‘bias’ as defined by the interaction between author and reviewer attributes, or bias as a function of reviewer characteristics (ref 29). They make the point that their methodology cannot determine biases that are shared by all reviewers regardless of reviewer characteristics. In the Introduction, the authors switch between describing previous research on reviewer characteristics and previous research on author characteristics. For example, paragraph 2 is predominantly about studies of reviewer characteristics, but ends on author characteristics. Paragraph 3 appears to begins on author characteristics, but then cites studies on reviewer characteristics. It would be much easier for the reader to understand what contribution this study makes to the literature if the authors made a clearer distinction between current evidence on reviewer characteristics (including those studies which looked at both reviewer and author characteristics) and other research on bias, which would include research focused only on author characteristics. As part of the peer review process for Frontiers reviewers can, if they wish, complete the ratings questionnaire shown in Fig 1 and it was these ratings that were used in the study analysis. I think it would be useful for the authors to clarify the following:Do Frontiers reviewers complete the rating independently of each other before the collaborative process? If this is the case, it is not clear to me why individual scores for each paper were averaged for the Frontiers dataset, but apparently not for the other two datasets. If the aim of this study is to investigate the interaction between reviewer attributes and those of authors, wouldn’t averaging the reviewer scores in this way confound this aim? Alternatively, if the ratings form is completed by reviewers after the collaborative process, how have the authors accounted for the potential confounding effect of the collaboration? Also, if reviewers complete the rating independently of each other, the characteristics of peer review for the Frontiers dataset used in this study are the same as that for the other datasets, i.e. it is single blind peer review. In all three datasets the reviewers are made aware of the authors names, but the authors do not know the reviewers’ – is that correct? In the discussion the authors state that the findings of this study could be valid for a broad range of peer review systems. However, this study did not include the interactive component of Frontiers peer review process, or if it did, it is not clear how. All three datasets appear to have used the single blind system of peer review. This statement in the discussion should be rephrased.MinorReference 31 is for a commentary. Can authors should provide the reference for the original Swedish study?Can the authors provide data on the error rate for their gender assignment process?I think the authors could provide a more informative title, for example, 'Bias in peer review: the interaction between reviewer and author characteristics.'Please note, I do not have the expertise to comment on the model and statistical analysis used in this study.", "responses": [ { "c_id": "1395", "date": "10 Jun 2015", "name": "Richard Walker", "role": "Author Response", "response": "We have now submitted a new version of our paper, which we have revised in the light of your comments, which we found extremely useful and constructive, though they caused us significant extra work. We are particularly grateful for your suggestions concerning the organization of the introduction to the paper, which we have attempted to take on board, and for your points on quality control (which were in the same spirit as comments from Lutz Bornmann. Thanks to these suggestions, we found a number of problems with the data, which had previously passed unnoticed.Comment: The authors are very clear that their study investigates ‘bias’ as defined by the interaction between author and reviewer attributes, or bias as a function of reviewer characteristics (ref 29). They make the point that their methodology cannot determine biases that are shared by all reviewers regardless of reviewer characteristics. In the Introduction, the authors switch between describing previous research on reviewer characteristics and previous research on author characteristics. For example, paragraph 2 is predominantly about studies of reviewer characteristics, but ends on author characteristics. Paragraph 3 appears to begins on author characteristics, but then cites studies on reviewer characteristics. It would be much easier for the reader to understand what contribution this study makes to the literature if the authors made a clearer distinction between current evidence on reviewer characteristics (including those studies which looked at both reviewer and author characteristics) and other research on bias, which would include research focused only on author characteristics.This was an extremely useful suggestion. We have now reorganized our introduction to talk first about effects regarding author characteristics, then to interactions between author and reviewer characteristics and finally to the characteristics of reviewers. We believe the paper gains significantly in clarity from this reorganization. As part of the peer review process for Frontiers reviewers can, if they wish, complete the ratings questionnaire shown in Fig 1 and it was these ratings that were used in the study analysis. Do Frontiers reviewers complete the rating independently of each other before the collaborative process? If this is the case, it is not clear to me why individual scores for each paper were averaged for the Frontiers dataset, but apparently not for the other two datasets. If the aim of this study is to investigate the interaction between reviewer attributes and those of authors, wouldn’t averaging the reviewer scores in this way confound this aim?The ratings questionnaire is filled in in the initial non-interactive part of the review process. We have clarified this in the text. Also, if reviewers complete the rating independently of each other, the characteristics of peer review for the Frontiers dataset used in this study are the same as that for the other datasets, i.e. it is single blind peer review. In all three datasets the reviewers are made aware of the authors names, but the authors do not know the reviewers’ – is that correct?Yes this is correct. We have clarified our description of the Frontiers process to show that it is single-blind. In the discussion the authors state that the findings of this study could be valid for a broad range of peer review systems. However, this study did not include the interactive component of Frontiers peer review process, or if it did, it is not clear how. All three datasets appear to have used the single blind system of peer review. This statement in the discussion should be rephrasedThe differences between the datasets concern not just the way the review is organized (interactive vs. non-interactive) but also the disciplines covered, and the geographical distribution of authors and reviewers. In the text, we clarify that all three review systems in the study are single blind. However, we hypothesize that reviewers preparing for an interactive process may behave differently from reviewers in a traditional review process. Reference 31 is for a commentary. Can authors should provide the reference for the original Swedish study?As you correctly note, the reference (now reference 30) was published in the form of a commentary. De facto, however, the article represents the first publication of results from an original study, which, to our knowledge was not published elsewhere, prior to the date on which the commentary appeared. Can the authors provide data on the error rate for their gender assignment process?On the basis of random sampling, we estimate errors rates of 7.5%, 0.0% and 5.2% for the Frontiers, IEEE (Spain) and IEEE (International) datasets respectively. We give these figures and discuss their significance in the text. I think the authors could provide a more informative title, for example, 'Bias in peer review: the interaction between reviewer and author characteristics.'We agree with this suggestion and have revised our title accordingly. It now reads, \"Personal attributes of authors and reviewers, social bias and the outcomes of peer review: a case study\"." } ] } ]	1	https://f1000research.com/articles/4-21
https://f1000research.com/articles/4-146/v1	09 Jun 15	{ "type": "Observation Article", "title": "Absence of kdr resistance alleles in the Union of the Comoros, East Africa", "authors": [ "Yoosook Lee", "Natalie Olson", "Youki Yamasaki", "Allison Chang", "Clare Marsden", "Ahmed Ouledi", "Gregory Lanzaro", "Anthony J. Cornel", "Natalie Olson", "Youki Yamasaki", "Allison Chang", "Clare Marsden", "Ahmed Ouledi", "Gregory Lanzaro", "Anthony J. Cornel" ], "abstract": "Knockdown resistance (kdr) and CYP9K1 genotypes were detected by a MOLDI-TOF based SNP genotyping assay (Sequenom iPLEX) in samples of Anopheles gambiae collected at 13 sites throughout the Union of the Comoros and Dar es Salaam, Tanzania during February and March 2011. All A. gambiae specimens collected in the Comoros were homozygous for the susceptible kdr alleles (+/+) while 96% of A. gambiae from Dar es Salaam were homozygous for the East African kdr resistant genotype (E/E). In contrast, all specimens from Dar es Salaam and the Comoros were homozygous for the cyp3 allele (c3/c3) at the CYP9K1 locus; the locus has been implicated in metabolic resistance against pyrethroid insecticides in West Africa. All specimens had typical A. gambiae genotypes for SNPs within the divergence Islands on all three chromosomes. Although further spatial and temporal studies are needed, the distribution of kdr genotypes between the Comoros and Tanzania further supports isolation of the Comoros populations from A. gambiae populations on mainland Africa.", "keywords": [ "DIS", "Dar es Salaam", "CYP9K1", "kncokdown resistence", "insecticide", "resistence" ], "content": "Introduction\n\nA majority of the human population residing in the Union of the Comoros (=94%) live in high malaria transmission zones1. Anopheles gambiae and Anopheles funestus (Giles) are the major malaria vectors in the Comoros2. Vector control efforts have concentrated on the adult stage using insecticide-treated bednets (ITNs) and indoor residual spraying (IRS) with DDT1. ITN distribution was initiated in the Comoros in 2005 and by 2014 roughly 40% of the population has access to ITNs1.\n\nLimited insecticide resistance surveillance has been conducted on malaria vectors in Union of Comoros, with to date, published records stemming only from investigations in Mayotte (an island administered by France), where A. gambiae were susceptible to multiple insecticides except for a larvicide, temephos3. Insecticide susceptibility studies have been conducted in neighboring East African countries such as in western Kenya (Chen et al. 2008. JME, Mathias et al. 2011, Malaria J, Ochomo et al. 2012 MVE), but little information is available on the coastal regions of Kenya. In Tanzania, information, based on small sample sizes, is available on the kdr allele frequency distribution in coastal districts of Muheza and Ilula (Dar es Salaam)4 where about one third from Dar es Salaam were homozygous for the kdr-East (L1014S) mutation.\n\nHere we present much needed data on kdr allele frequencies and include frequency data for a recently described pyrethroid metabolic resistance gene, CYP9K1. Allele frequencies for Anopheles gambiae collected at 13 sites in the Union of the Comoros, plus Dar es Salaam, Tanzania are presented (Table 1).\n\nNumbers (#) indicate site locations on the map in Figure 1.\n\n\nMethods\n\nA total of 362 indoor resting adults and larvae were collected from 13 locations from the three islands (Figure 1) making up the Union of the Comoros between February and March, 2011. Larvae were individually rinsed twice in bottled mineral water and placed in 80% ethanol for downstream genomic DNA extraction. A collection of A. gambiae sensu lato from Furvela, Mozambique were collected using light traps inside houses. Mosquitoes from Dar es Salaam were obtained from Dr. Kija Ngh’abi at Ifakara Health Institute.\n\nSite numbers corresponds to index number provided in Table 1.\n\nSamples were transported to the UC Davis Vector Genetics Laboratory for further genetic assay. DNA was extracted using a DNeasy extraction kit (Qiagen, Valencia, CA). Species were determined based on the combination of species diagnostic assays5,6 and a divergence island SNP (DIS) genotyping assay7.\n\nFor DIS, kdr and CYP9K1 genotyping, we used the Sequenom iPLEX Gold Genotyping Reagent Set (Catalog number: Sequenom 10158) on a MassArray (Sequenom) mass spectrometer at the UC Davis Veterinary Genetics Laboratory. This assay was slightly modified from the original DIS assay7 by adding the kdr and CYP9K1 markers, as described in Supplemental Document S1.\n\n\nResults & discussion\n\nA. gambiae from Dar es Salaam, Tanzania, had the kdr-East (L1014S) genotype at a frequency of 96%, which is higher than the frequency previously reported from Dar es Salaam by Kabula et al.7 where respectively, 1/3 and 2/3 of their samples were homozygous and susceptible for kdr-East (L1014S). In contrast, all A. gambiae from the Comoros were homozygous for the susceptible kdr alleles. All A. gambiae from both Tanzania and the Comoros were homozygous for the cyp3 allele for the CYP9K1 gene. All specimens from Furvela, Mozambique were A. merus (30/35) or A. arabiensis (5/35) and were excluded from further analysis.\n\nSignificant pressure to select for resistance to pyrethroid insecticides in A. gambiae and other indoor biting and resting malaria vectors likely occurs throughout sub-Saharan Africa because of intense IRS and ITN usage. A recent study in Mali noted an adaptive introgression of kdr resistant alleles from A. gambiae stably incorporated into the A. coluzzii genome under high ITN coverage environments8. A similar genomic signature of adaptive introgression was also observed in Ghana9. A. gambiae populations in the Comoros have had the opportunity, via transport by boat or air, to acquire resistant A. gambiae genotypes from neighboring countries such as Tanzania where high levels of insecticide resistance have been reported10. The failure of the Comoros population to acquire insecticide resistance alleles despite long term exposure to insecticide pressure1 may potentially be due to several factors or combination of factor including: (1) ITN coverage (<25% compared to >60% Mali) is not high enough to drive selection for resistance, (2) these populations are very isolated from mainland populations, requiring them to develop resistance de novo rather than from gene flow from neighboring populations, and/or (3) A. gambiae on the Comoros may be exophilic.\n\nOur study provides much needed information regarding the genetics of insecticide resistance in A. gambiae populations in the Comoros Islands. Although the malaria vectors in Comoros appear to be genetically predisposed to insecticide susceptibility, it is possible that these mosquitoes have developed phenotypic resistance via alternative mechanisms such as metabolic resistance other than CYP9K1 or behavior resistance (e.g. exophily). Further studies are needed to establish levels of phenotypic resistance against insecticides, as well as bionomics of the malaria vectors in this region to understand the impact of insecticide-based malaria control measures in the Comoros.", "appendix": "Author contributions\n\n\n\nYL conceived the study, designed experiments, performed data analysis and wrote manuscript. YY, AC, NO conducted experiments. CDM conducted field collections and conducted experiments. AO, GCL and AJC conducted field collections and wrote manuscript. All authors were involved in drafting this manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors also acknowledge financial support from NIH grants: 5R21AI062929.\n\nI confirm that the 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 Catelyn C. Nieman for assistance in DNA extraction and species diagnostic assay. We also thank Julia Malvick at the Veterinary Genetics Laboratory of UC Davis School of Veterinary Medicine for assistance in processing iPLEX SNP genotyping assay.\n\n\nSupplementary materials\n\nSupplemental Document S1.\n\nModification of the original DIS assay in 7.\n\nClick here to access the data.\n\n\nReferences\n\nWHO: World Malaria Report 2014. Switzerland: World Health Organization 2014. 2014. Reference Source\n\nAyala D, Goff GL, Robert V, et al.: Population structure of the malaria vector Anopheles funestus (Diptera: Culicidae) in Madagascar and Comoros. Acta Trop. 2006; 97(3): 292–300. PubMed Abstract \| Publisher Full Text\n\nPocquet N, Darriet F, Zumbo B, et al.: Insecticide resistance in disease vectors from Mayotte: an opportunity for integrated vector management. Parasit Vectors. 2014; 7: 299. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKabula B, Kisinza W, Tungu P, et al.: Co-occurrence and distribution of East (L1014S) and West (L1014F) African knock-down resistance in Anopheles gambiae sensu lato population of Tanzania. Trop Med Int Health. 2014; 19(3): 331–341. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScott JA, Brogdon WG, Collins FH: Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg. 1993; 49(4): 520–529. PubMed Abstract\n\nFavia G, Lanfrancotti A, Spanos L, et al.: Molecular characterization of ribosomal DNA polymorphisms discriminating among chromosomal forms of Anopheles gambiae s.s. Insect Mol Biol. 2001; 10(1): 19–23. PubMed Abstract \| Publisher Full Text\n\nLee Y, Marsden CD, Nieman C, et al.: A new multiplex SNP genotyping assay for detecting hybridization and introgression between the M and S molecular forms of Anopheles gambiae. Mol Ecol Resour. 2014; 14(2): 297–305. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNorris LC, Main BJ, Lee Y, et al.: Adaptive introgression in an African malaria mosquito coincident with the increased usage of insecticide-treated bed nets. Proc Natl Acad Sci U S A. 2015; 112(3): 815–820. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClarkson CS, Weetman D, Essandoh J, et al.: Adaptive introgression between Anopheles sibling species eliminates a major genomic island but not reproductive isolation. Nat Commun. 2014; 5: 4248. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKabula B, Tungu P, Malima R, et al.: Distribution and spread of pyrethroid and DDT resistance among the Anopheles gambiae complex in Tanzania. Med Vet Entomol. 2014; 28(3): 244–252. PubMed Abstract \| Publisher Full Text" }	[ { "id": "8978", "date": "22 Jul 2015", "name": "Beniamino Caputo", "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\nTitlePlease specify which kdr has been genotyped. Abstract The abstract provides an adequate summary of the article.Please insert that you collected just “1” site in Tanzania at Dar es Salaam. Moreover, you have 14 sampling sites in the Comoros according to Table 1.Please specify that you refer to Anopheles gambiae (Coetzee et al., 2013). Introduction Please explain your interest on CYP9K1 gene and add citations.Add references on insecticide resistance surveillance in Mozambique if any are available.Add the objective of the study (gene-flow of continental versus island).Minor comments:In the 1st paragraph, the last sentence change “has” to “had”.In the second paragraph substitute citation with numbers. MethodsStudy design and methods are quite well explained.Please specify how many larvae and adults have been collected from 13 locations from the three islands, and give more details in the results section.Modify Figure 1 according to Table 1 (Figure 1: image on the right side: you have missed out the number 14 and the number 15 is in a different location to that stated in Table 1. Therefore the number 15 in Figure 1 should be changed to a 14 and number 15 should be added at Dar es Salaam, Tanzania).In the table you list 15 collection sites but here you cite only 14. ResultsPlease specify the statement “malaria vectors in Comoros appear to be genetically predisposed to insecticide susceptibility”and add more details and references.Correct the citation Kabula et al. 7 to Kabula et al.4The samples in Mozambique seem not related to the study since only A. merus or A. arabiensis have been found.The discussions are balanced and justified based on the results obtained, even if continental sample (Tanzania) is very small.", "responses": [ { "c_id": "1478", "date": "23 Jul 2015", "name": "Yoosook Lee", "role": "Reader Comment", "response": "Thank you very much for your review.The kdr we genotyped is also known as L2014F. We will revise our title accordingly in the upcoming revision.Clarification on the species A. gambiae will be also made. The formal paper describing how we came across CYP9K1 gene and its evolutionary history among A. gambiae and A. coluzzii is under review and we should be able to add the proper citation in the next revision.We have looked for the peer-reviewed articles on insecticide resistance surveillance in Mozambique but we have not found one thus far. We welcome suggestions if you came across such publications.Other suggestion will be incorporated in the upcoming revision. Once we made revisions, we will post the detailed response to your review.Thank you very much again for your constructive comments!" } ] }, { "id": "8980", "date": "03 Aug 2015", "name": "Frederic Tripet", "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 an interesting short report describing the prevalence of two loci incriminated in metabolic and target-site resistance to pyrethroid pesticides in populations of the malaria mosquito A. gambiae s.s. on the Comoros Islands. The cyp3 allele of the CYP9K1 gene was found in all island specimens and also in a single population sampled from mainland Tanzania. In contrast the East African kdr (L1014S) allele was not found on the islands suggesting that these populations are fully geographically isolated from mainland populations, or alternatively that pesticide selection pressures are not very high on the islands due to limited vector control programmes. I just have a few general suggestions that may help make the current report a little bit more informative in key areas and thus make it more relevant to a broader readership. The current introduction and discussion are very succinct...Are these 2 loci the only major loci involved in resistance to pyrethroids and why were they chosen in particular? Additional background with references would be useful in the intro. As noted by the authors, there is limited information on pesticide resistance in general on the Comoros. The authors also state: 'Our study provides much needed information regarding the genetics of insecticide resistance in A. gambiae populations in the Comoros Islands.' It looks to me that what is first and foremost really needed are detailed bioassay surveys of pesticide resistance for the island populations. These would have provided a better context for the current study and would have made the interpretation of the distribution on the two kdr resistance loci easier. Is the cyp3 allele here an ancestral allele or has it swept through due to selection? Given the above, the main selling point of this study lies, in my view, in the use of the kdr locus as a marker for introgression as done previously in studies of West African A. gambiae populations. For many years, kdr was the only marker suggesting an absence of gene flow between A. coluzzii and A. gambiae s.s. in Mali. Hence, I would suggest expanding that part of the discussion a little bit. The current section omits some interesting parallels and details. In the same line of thought, discussing the new findings with those generated using neutral markers in a previous study for the same islands (Marsden et al., 2013) would be useful, again because of the similarities with approaches (neutral versus non-neutral markers) used in past studies of reproductive isolation. The authors mention possible gene flow from the mainland via boats. It may be worth mentioning that the main boat connections are from Mayotte, Zanzibar and Madagascar whose A. gambiae populations are not included in this and previous study. Why is that? What is the status of those populations? If the Comoros islands were to be used for possible mosquito release programmes, re-colonization from neighbouring islands would possibly also be a concern. On the other hand, gene flow from the Mozambique coast can only be dismissed if additional sampling was made from that region. This is clearly an area of the world whose vector populations are greatly understudied. Given that these islands represent some of the best locations for testing mosquito release programmes, this study, albeit a small contribution, represents an important step in the right direction.", "responses": [] }, { "id": "9761", "date": "03 Aug 2015", "name": "Frédéric Simard", "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 paper by Lee et al. provides strong evidence for the absence of mutant alleles at the well-characterized kdr locus in populations of the major malaria mosquito Anopheles gambiae from three islands of the Comoros archipelago. The lack of kdr mutants in these populations contrasts with the high frequency of the East-African kdr mutation (L1014S) in a continental population from Dar es Salam, Tanzania. These findings are consistent with results of work carried out by our research group in the neighbouring island of Mayotte, showing no evidence for phenotypic insecticide resistance in An. gambiae, as well as the absence of target site mutations at the kdr locus on this island. The authors conclude that, altogether, these results suggest restricted gene flow between continental and island populations of An. gambiae in this area. The paper is concise and clear. The title and abstract are appropriate, and they reflect adequately the content of the paper. There are, however, a few minor shortcomings to be addressed:The number and position of the sampling sites are not consistent in the text, Table and Figure: 13 sampling sites are mentioned in the text and abstract, 15 are shown in the Table and Figure; there are also inconsistencies with the identification codes between the table and the figure, and the caption of Table 1 indicates these codes with the symbol ‘#’ whereas the corresponding column name in the body of the Table is ‘idx’. In the first sentence of the ‘Results & discussion’ section, the authors state that “A. gambiae from Dar es Salaam, Tanzania, had the kdr-East (L1014S) genotype at a frequency of 96%,…”. The sentence should either state that 96% of the specimens were homozygous for the kdr allele (as in the abstract) or that the kdr mutation (instead of “genotype”—it is more appropriate as this is a single nucleotide polymorphism) was found at a frequency of 98% (as shown in Table 1). It is mentioned in the Introduction that “population access to ITNs” in 2014 in the Comoros was 40%, whereas it is reported that “ITN coverage” is <25% when discussing the results. Please explain the difference between these figures. Table S1 in the Supplementary Materials still uses the non-Linnean nomenclature of molecular forms instead of An. gambiae and An. coluzzii; we think this should now be superseded by the Linnean taxonomic nomenclature. There is little information in the paper about the CYP9K1 locus; for example, does it confer cross-resistance to DDT? Is the cyp3 allele wild-type or ‘resistant’? What is the phenotype of ‘resistant’ alleles? One or more references would be useful in this respect. In which year were the mosquitoes sampled in Dar es Salam? Moreover, the authors should:Include the references ‘Chen et al. 2008. JME, Mathias et al. 2011, MalariaJ and Ochomo et a l. 2012 MVE’ which are cited in the text (‘Introduction’) in the reference list; Specify in the ‘Methods’ (or perhaps in Table 1) which of the 362 specimens were collected as resting adults and which ones were collected as larvae; also, which steps were taken to avoid sampling individuals coming from the same mosquito progeny in larval samples, as this could have an impact on observed genetic diversity. Correct the reference number for the Kabula et al. paper cited in the ‘Results & discussion’ section. Use the abbreviation “An.” rather than “A.” for Anopheles, in agreement with taxonomic conventions and community usage. Italicize the adverb “sensu lato” given that it is from the Latin.Finally, it is in our view inappropriate to infer about gene flow between the Comoros and the African continent based on these results. (i) The locus is under selection, which is not ideal for gene flow inference, and, as discussed by the authors, the nature, strength, and geographic distribution of selective pressures may differ between the Comoros and the single continental population that was sampled. (ii) The level of resistance to pyrethroids and other insecticide compounds due to the L1014S mutation is probably low in An. gambiae (Ranson et al., 2000), and may differ according to genetic background. (iii) All other genetic markers used in the study, including CYP9K1 argue for gene flow occurring between these populations.Because there are no phenotypic data for the level of resistance of the An. gambiae populations that were included in the study, and because no historical and accurate data on insecticide usage in the Archipelago are available, it is at this point rather speculative to explain the absence of kdr mutations in the Comoros populations. Accordingly, the selective pressures for insecticide resistance operating on these populations need to be assessed before inferring the dynamics of gene exchange between these island and continental populations. As acknowledged by the authors, vector behaviour (feeding preferences, feeding time, endo-exophagy, endo-exophily) also needs to be investigated before any conclusion can be put forward to explain the differences observed.", "responses": [] } ]	1	https://f1000research.com/articles/4-146
https://f1000research.com/articles/4-145/v1	08 Jun 15	{ "type": "Review", "title": "The complement system in lupus nephritis", "authors": [ "Lihua Bao", "Patrick N. Cunningham", "Richard J. Quigg", "Patrick N. Cunningham", "Richard J. Quigg" ], "abstract": "The complement is part of the innate immune system and can be activated through one of three pathways. To prevent injury of self-tissue, complement is tightly regulated by over 30 proteins. Complement plays dual roles in the pathogenesis of systemic lupus erythematosus (SLE). On one hand, hereditary homozygous deficiencies of classical pathway components, such as C1q and C4, are associated with an increased risk for SLE, suggesting that complement is protective. On the other hand, complement is systemically consumed in both experimental and human SLE, suggesting its pathogenic role. Studies in genetically altered mice have shown that lack of complement inhibitors, such as complement factor H (CFH) or decay-accelerating factor (DAF) accelerates the development of experimental lupus nephritis; while treatment with recombinant protein inhibitors such as CR1-related protein y (Crry)-Ig, CR2-Crry, CR2-DAF and CR2-CFH ameliorates the disease development. Complement-targeted drugs, including soluble complement receptor 1 (TP10), C1 esterase inhibitor (C1-INH), and a monoclonal anti-C5 antibody (Eculizumab) have been shown to inhibit complement safely, and are now being investigated in a variety of clinical conditions. These clinical developments support their therapeutic use in lupus nephritis.", "keywords": [ "animal model", "complement", "lupus nephritis", "systemic lupus erythematosus", "therapy" ], "content": "\n\nThe involvement of the complement system in the pathogenesis of a number of autoimmune diseases including SLE is well accepted, yet its exact role is still not clear. On one hand, hereditary deficiencies of early classical pathway complement components predispose patients to SLE. On the other hand, activation of complement by immune complexes (ICs) is certainly a prominent feature in SLE that promotes tissue injury. Therefore, an imbalance of the complement system in either direction and the respective roles of the classical and alternative pathways can have complex effects on the disease phenotype. The current review will discuss the dual roles of complement in SLE, and in particular in lupus nephritis (LN).\n\n\nThe complement system\n\nThe complement system is an important part of innate immunity which defends the host against infectious microorganisms, clears ICs and dead cells, and serves as a bridge between innate and adaptive immunity1. Complement can be activated through classical, alternative and mannose-binding lectin (MBL) pathways, each with different initiators (shown schematically in Figure 1).\n\nShown are the three activation pathways – classical (CP), MBL (LP) and alternative pathways (AP) and the common intermediates of activation, C3 and C5 convertases. Regulatory proteins are in boxes, matched by color to the sites of inhibition. Anti-C5 antibody (Eculizumab) is also depicted. Cn, Clusterin; FI, complement factor I.\n\nThe classical pathway is activated when the C1q subunit of C1 binds with high avidity to the Fc portion of immunoglobulin (Ig) M (CH3 domain) or IgG (CH2 domain) in ICs. Besides Igs, C1q also binds and facilitates the removal of apoptotic cells, which affords it an important role in immune tolerance. Loss of complement self-regulatory protein CD46 (also known as Membrane Cofactor Protein, MCP), on apoptotic cells (by shredding into microparticles or cluster into blebs) was suggested as the mechanism of driving C3 opsonization and promotion of efferent removal of apoptotic cells2. The binding of C1q to its target triggers its conformational change and self-activation of C1r, followed by the activation of C1s3. Activated C1 (as the multiprotein C1qr2s2 complex) cleaves both C4 and C2 to generate C4a/C4b and C2a/C2b. C4b2a acts as a C3 convertase which cleaves and activates C3.\n\nThe alternative pathway is spontaneously activated by a C3 conformation change, which occurs slowly but continuously to generate hydrolyzed C3 (C3(H2O)). C3(H2O) then binds complement factor B (CFB) in the fluid phase. Upon factor B’s cleavage by complement factor D (CFD, also known as adipsin), an initial C3 convertase C3(H2O)Bb is formed, which can be stabilized by properdin, and promotes its own production to form C3bBb, the alternative pathway C3 convertase1.\n\nThe binding of MBL to terminal carbohydrate groups on certain microbes leads to the activation of the MBL-associated serine proteases (MASPs). Activated MASP cleaves C4 to C4a and C4b. Immobilized C4b induces the binding of C2 which is also cleaved by MASP and generates the C4b2a C3 convertase4.\n\nIrrespective of the pathway of activation, cleavage of C3 and C5 ultimately occurs, with the generation of the C3a and C5a anaphylatoxins, C3b opsonins, and C5b to start the non-enzymatic assembly of the C5b-9 membrane attack complex, which can result in cellular death or activation after membrane insertion5, which also occurs in erythrocytes (E) in paroxysmal nocturnal hemoglobinuria (PNH)6.\n\nGiven the potency of the complement system, natural fluid-phase and cell membrane-bound regulatory proteins acting throughout the three cascades are essential to prevent activation and injury to host tissues (Figure 1)1. The regulators of complement activation (RCA) gene family on human chromosome 1q3.2 (and a comparable location in mouse chromosome 1) includes MCP, complement receptor 1 (CR1), decay accelerating factor (DAF, also known as CD55), C4b-binding protein (C4bp), and CFH7,8. These proteins inhibit complement activation through interactions of their conserved short consensus repeats (SCRs) with fragments of C3 and/or C49.\n\nMCP is a cell surface glycoprotein which serves as a cofactor for factor I-mediated cleavage of C3b and C4b, thereby inhibiting the formation of C3/C5 convertases10. More recently, it was found that besides acting as a complement inhibitor, MCP also regulates T cell subsets during an immune response by promoting the activation of TH1 cells and their IL-10 production11. Studying of MCP in rodents has its obstacles, due to its restricted expression predominantly in testicular germ cells12, while in humans MCP has been identified in all cell types except erythrocytes10. DAF is a glycosylphosphatidylinositol (GPI)-anchored membrane protein, which binds C3b and C4b and accelerates the decay of C3 (C4b2a in the classical pathway and C3bBb in the alternative pathway) and C5 (C4b2a3b in the classical pathway and C3bBb3b in the alternative pathway) convertases13. CR1 is a single chain transmembrane glycoprotein which has the combined functions of DAF and MCP: it accelerates the decay of C3 and C5 convertases like DAF and it also serves as a cofactor for complement factor I-mediated inactivation of C3b and C4b into iC3b and iC4b with a similar function to MCP14. The plasma proteins, C4bp and complement factor H (CFH) also have cofactor activity for factor I-mediated inactivation of plasma C4b and C3b, in the classical and alternative pathways, respectively. Specific to rodents is the 65-kDa rodent complement regulatory protein, more commonly referred to as CR1-related gene/protein y (Crry) which has combined decay-accelerating and factor I cofactor activity for C3b and C4b, similar to CR115. Lastly, at either “end” of the complement cascades are C1-inhibitor (C1-INH) or CD59, which inhibit C1 activation and C5b-9 formation, respectively16,17.\n\n\nThe dual roles of complement in human SLE\n\nSLE is an autoimmune disorder caused by loss of tolerance to self-antigens, the production of autoantibodies, and deposition of complement-fixing ICs in injured tissues. SLE is characterized by a wide range of clinical manifestations and targeted organs, with LN being one of the most serious complications18. The pathogenic roles of complement activation in human SLE were indicated from years of clinical observations: low total complement hemolytic activity (CH50) and decreased C3 and C4 levels have been found in about 75% of SLE patients with focal nephritis and 90% in patients with diffuse nephritis19. Additionally, the co-localization of Ig isotypes IgG, IgA, and IgM with C1q, C4 and C3 (and C5b-9) (the so called “full house” pattern) in the glomeruli is almost exclusively present in glomeruli of patients with lupus nephritis20. Finally, complement split products such as C3d and C5b-9 can also be detected in the urine of SLE patients21.\n\nWhile frequent mutations in CFH and MCP found in atypical hemolytic uremic syndrome (aHUS) are not associated with lupus nephritis, SLE patients carrying these mutations do have an earlier onset of lupus nephritis22. SLE patients with lymphopenia and neutropenia were found to have lower level of DAF and CD59 on neutrophils, and SLE patients with anemia had lower expression of CD59 and CR1 on erythrocytes, and the expression is correlated with disease activity23. In neutrophils from SLE patients, IC down-regulates the levels of CR1 transcripts, both directly, and indirectly via inhibiting interferon (IFN) -γ induced CR1 expression, an example of how lupus activity affects CR1 expression24. Nonetheless, higher CR1 expression on SLE patient leucocytes indicates a better prognosis25, while lower leucocyte MCP expression is associated with lupus remission26. These findings not only suggest the pivotal roles of the complement system in the pathogenesis of SLE but may also potentially provide noninvasive markers for the disease activity.\n\nImpaired IC handling plays an important role in the pathogenesis in LN. Since the complement system is required at all steps of normal IC metabolism, any number of alterations can lead to pathological glomerular IC accumulation, particularly in conditions of IC excess, as in SLE. Studies have shown the association of SLE with low levels of CR1 on erythrocytes, a key site of binding and transferring IC to the mononuclear phagocyte system27, suggesting that a defective erythrocyte/IC-clearing system may be related to SLE pathogenesis.\n\nIn contrast to the widespread belief that generation of complement activation products in kidney and other disease sites is proinflammatory, patients with homozygous deficiencies of the C1 proteins (C1q or C1r/s) or C4 have a high prevalence (> 80%) of autoantibodies and SLE-like disease28. Successful treatment of C1q-deficient SLE patients with fresh frozen plasma or hematopoietic stem cell transplantation was also reported29,30. It is also noteworthy that in SLE patients, anti-C1q antibodies are associated with proliferative lupus nephritis, and anti-C1q antibody levels may indicate renal disease activity31. More recently, it was reported that in the presence of sera from individuals deficient in C1q, C4, C2 or C3, phagocytosis of apoptotic cells was decreased compared with studies using normal sera, indicating important roles of all these complement classical pathway components in clearance of apoptotic and necrotic cells32. Circulating cell-derived microparticles in patients with SLE carry more IgG, IgM, and C1q, and IgG-bearing microparticles are associated with autoantibodies and complement activation33. Interestingly, in the very rare cases of homozygous deficiency of C3, the most critical protein that affects all three pathways of complement activation, there is no association with SLE34.\n\nComplement also plays important roles in thrombotic complications associated with SLE. The risk of thrombosis is particularly high in SLE patients with antiphospholipid (aPL) autoantibodies. One explanation is that aPL-containing ICs bind to platelets which can subsequently activate the classical complement pathway, as supported by the fact that SLE patients have increased levels of C1q, C3d and C4d on their platelets, especially in patients with history of venous thrombosis. Sera from SLE patients with aPL autoantibodies has a higher capacity to activate the classical pathway on heterologous platelets35. Though C1q and C4d deposition on platelets is not specific for SLE, this is associated with venous thrombosis36. One study also suggests that small dense HDL particles may activate complement system and is related to subclinical atherosclerosis in SLE patients37.\n\n\nStudies of complement in experimental lupus models\n\nTwo of the best studied murine models that spontaneously develop lupus-like syndromes are the F1 cross between New Zealand Black and New Zealand White mice (NZB/W) and the MRL/MpJ-Tnfrsf6lpr/lpr/J (MRL/lpr) strain38. Similar to the female predominance in humans, only female NZB/W mice develop SLE. Both models have B cell hyperactivity, autoantibodies, hypocomplementemia, circulating and glomerular-bound ICs, and severe nephritis. As in humans, there is plenty of circumstantial evidence that complement activation is actively involved in the pathogenesis of glomerular disease in these mice. In early stages (4 and 5 months of age in MRL/lpr and NZB/W mice, respectively), granular deposition of mouse IgG, IgA, IgM and C3 are present largely in the mesangium, coincident with histopathology showing mesangial proliferation. As the disease progresses, there are glomerular capillary wall IC deposits, proliferation of intrinsic endothelial and mesangial cells, and infiltration with inflammatory cells. Eventually, crescent formation (more often in MRL/lpr mice) and glomerulosclerosis (more often in NZB/W mice) occurs, and mice die of renal failure38.\n\nThe manipulation of individual complement proteins through genetic techniques in lupus mouse strains and functional inhibition through the use of specific antibodies or antagonists have provided considerable insight into how complement is involved in this disease. Given that C3 is the common point connecting all three pathways in complement activation and is tightly regulated naturally, many of the studies in lupus mice have concentrated on activators and regulators of C3.\n\nDAF is a ubiquitously expressed GPI-anchored membrane protein that inhibits C3 activation through all pathways by inhibiting formation and accelerating decay of the C3/C5 convertase13. The relevance of DAF to lupus is suggested by the fact that DAF-deficient MRL/lpr mice had exacerbated lymphoproliferation, anti-chromatin autoantibody production and dermatitis, particularly evident in females, while nephritis appeared to be unaffected39. Furthermore, DAF-deficient MRL/lpr mice also deficient in C3 developed comparable lymphadenopathy, splenomegaly and anti-chromatin autoantibodies to what is seen in the complement sufficient mice, suggesting the protective effect of DAF in MRL/lpr autoimmunity is complement-independent. DAF-sufficient MRL/lpr chimeras with DAF-deficient MRL/lpr bone marrow developed significantly attenuated dermatitis compared with that in DAF-deficient MRL/lpr chimeras with DAF-sufficient MRL/lpr bone marrow, indicating that the protective effect of DAF on dermatitis is attributable to local expression40. The exacerbated dermatitis can be explained by the fact that DAF is strongly expressed in the skin while Crry is not.\n\nLike human CR1, Crry is an intrinsic membrane complement inhibitor that inhibits C3 convertases of all pathways, combining activities of human DAF and MCP15. Given that the Crry-deficient mice generated by Molina et al. have complete embryonic lethality from unrestricted maternal complement activation41, which makes generation of a straightforward Crry-deficient murine lupus model impossible. The critical role of Crry in regulating complement in the kidney came from our study of transplantation of Crry−/−C3−/− kidneys into wildtype hosts. Due to lack of Crry, there was marked inflammation in the tubulointerstitium which led to complete failure of the transplanted kidney within weeks (while the appropriate controls, including wild type kidneys transplanted into wild type or Crry−/−C3−/− mice remained normal)42. On the other hand, transgenic mice developed by our group that overexpressed a soluble form of Crry had less severe lupus nephritis as determined by blood urea nitrogen (BUN) and albuminuria measurements when crossed into the MRL/lpr strain. Since the spontaneous mortality in lupus mice is largely due to kidney disease, this translated into prolonged survival, while the underlying abnormal autoimmunity was not affected43. To make this complement inhibition more applicable to human SLE treatment, a recombinant soluble form of Crry (Crry-Ig) was also used in MRL/lpr mice. In chronic usage from early in the autoimmune disease until the end-stage, inhibited complement activation by Crry-Ig ameliorated lupus nephritis44. Interestingly, transcript profiling experiments showed that excessive matrix components such as collagens I, III, IV and VI were overexpressed in control MRL/lpr mice, which could be suppressed by complement inhibition with Crry-Ig. Potential explanations for these phenomena include Crry-Ig-mediated reductions in connective tissue growth factor and TGF-β1 expression, suggesting these profibrotic mediators are downstream of complement-induced injury and contribute to the progressive glomerulosclerosis in MRL/lpr mice45.\n\nIn both human and experimental SLE, CR1/CR2 expression decreases over time, suggesting that this plays a role in disease46,47. Yet, MRL/lpr mice deficient in CR1/CR2 had significantly lower levels of total IgG3 and specific IgG3 rheumatoid factor, supporting the role of CR1/CR2 in production of IgG3 in response to autoantigens. Nonetheless, this decrease of IgG3 autoantibodies did not lead to a reduction in features of lupus nephritis48.\n\nC4bp is a major soluble complement inhibitor of the classical pathway. Elevated serum levels of C4bp in SLE patients suggested its involvement in this disease49. More detailed studies addressing this possibility were performed by the Braun group, using a C4bp deficient MRL/lpr mouse model. While serum C4 levels are lower in MRL/lpr mice compared with normal C57BL/6 mice, starting as early as 3 weeks of age (i.e., prior to autoimmune disease onset), surprisingly deficiency of C4bp affected neither serum C4 levels nor the classical pathway hemolytic activity. With an unaffected complement classical pathway, C4bp deficiency did not lead to a significantly different outcome in these mice, which had similar amounts of glomerular C3 and IgG deposition, glomerulonephritis, tubulointerstitial inflammation, proteinuria and renal function. Moreover, the systemic immune response was not affected by the absence of C4bp in these mice50. One explanation for the negligible role of C4bp in lupus development in MRL/lpr mice is that the extensively activated classical pathway in this model simply overwhelms the absence of C4bp. In this situation, “a gain of function” of C4bp may reveal more information, as purified human C4bp successfully inhibited the classical complement pathway, leading to delayed or reduced disease development in these models51.\n\nIn physiological situations, there is spontaneous continuous low-level alternative pathway activation that is restrained by effective complement regulation. While the traditional thinking is that SLE is induced through IC-directed classical pathway activation, the involvement of the alternative pathway has also been suggested in many studies. Gilkeson et al. demonstrated that CFB- or CFD-deficient MRL/lpr mice had reduced glomerular C3 deposition associated with less severe glomerular histopathology52,53. These results imply that IC-directed classical pathway activation can recruit the potent alternative pathway to further amplify generation of C3 and C5 activation products. Complement factor H-related (CFHR) protein C is absent in the circulation of MRL/lpr and NZB/W mice before and after the onset of lupus, implying that polymorphic variation may contribute to the development of SLE54. Our group demonstrated CFH-deficient MRL/lpr mice developed a severe inflammatory diffuse lupus nephritis by 12 weeks, characterized by glomerular hyalinosis (“wire-loops”), and messagial, endocapillary and extracapillary proliferation, that resulted in the death of 50% of animals by 13 weeks of age55. These findings indicate that in contrast with the spontaneous disease in older CFH-deficient mice on mixed backgrounds56, loss of CFH accelerates the development of lupus nephritis and recapitulates the functional and structural features of human disease in MRL/lpr mice.\n\nC5b-9 deposition was found in diseased lupus kidneys more than two decades ago. Yet, compared with extensive studies focusing on C3 activation and its regulation, fewer studies have been done to investigate the downstream events following C3 activation in the pathogenesis of SLE. Wang et al. used a specific monoclonal antibody to inhibit C5 function in NZB/W mice. Six months of continuous therapy led to significantly delayed onset of proteinuria, improved renal pathological changes and prolonged survival, implicating a role of products of the terminal complement pathway, C5a and C5b-9 in lupus nephritis57. On the other hand, CD59a-deficient MRL/lpr mice had exacerbated skin disease and lymphoproliferation, through a complement-independent, autoimmunity regulatory mechanism, which is conferred by CD59a’s expression on both bone marrow-derived cell and peripheral tissues58.\n\nThe C3a and C5a anaphylatoxins are generated through complement activation when C3 and C5 are activated and cleaved. Increased expression of C3aR expression was found in glomeruli in human lupus nephritis59. C3aR and C5aR expression were significantly up-regulated at both the mRNA and protein levels and accompanied by a wider cellular distribution in MRL/lpr mouse kidneys60,61. This upregulated expression starts before the onset of kidney disease, supporting the idea that they may be involved in the development of disease, rather than simply being a consequence. Chronic administration of a specific C3aR antagonist (SB290157) led to significantly reduced kidney disease and prolonged survival in MRL/lpr mice61. Similarly, when C5a signaling was blocked in our studies with a specific antagonist60 or in those by Braun et al. through gene targeting62, MRL/lpr mice animals displayed attenuated renal disease and prolonged viability. The effects of blocking C3aR and C5aR in lupus mice had certain features in common, including less renal neutrophil and macrophage infiltration, apoptosis, and IL-1β expression60,61. Effects on chemokine expression were distinct, with C3aR- and C5aR-inhibited MRL/lpr mice having reduced CCL5 (RANTES) and CXCL2 (MIP-2) expression, respectively. C3aR-inhibited mice also had increased phosphorylation of protein kinase B (Akt), which we considered suggestive that C3aR signals promote renal cell apoptosis through an Akt pathway61. In C5aR-deficient MRL/lpr mice, there was a reduction in CD4+ T cell renal infiltration, lower titers of anti-double stranded DNA antibodies, and inhibition of interleukin (IL)-12 p20 and IFN-γ production, suggesting that Th1 responses are important to link C5a signaling in lupus nephritis62. In contrast, C3aR-deficient MRL/lpr mice had elevated autoantibody titers, more glomerular crescents and more severe intrarenal vasculitis, though it did not affect long-term renal injury or survival63. The mechanisms of different outcomes between short term (by using an antagonist) and long-term (by using gene-targeting) blockade of C3aR signaling in this lupus model still need to be clarified. This may be due to the fact that the particular C3aR antagonist (SB290157) used in this study also has partial agonist activity64.\n\nSince C3 is the converging point for all three complement pathways, alterations in C3 activation through manipulating its regulators, such as Crry or DAF, or blockade of the effects of C3 activation with inhibitors of C3aR, C5aR, or C5 have shown C3 activation is an important factor in the development of SLE. Surprisingly, C3 deficiency does not affect the development of nephritis in MRL/lpr mice, while glomerular IgG deposition is significantly increased65. This study is consistent with the important role of complement, and in particular C3, in the clearance of ICs66. The finding that homozygous deficiency of early components of the classical pathway other than C3 predispose to SLE suggests that physiological function of these molecules are protective in SLE. While the exact mechanism is still unknown, studies found that mice with generated C1q and C4 deficiencies had impaired ability to clear apoptotic debris67, leading to the accumulation of potentially immunogenic autoantigens and initiation of an autoimmune reaction in the right genetic setting. C1q-deficient mice had increased mortality and higher titers of autoantibodies, with 25% of the mice developing glomerulonephritis, characterized by glomerular IC deposits and apoptotic debris68. Further work suggests that C1q and DNase1 are important in the degradation of chromatin derived from necrotic cells69. It was also found that the IgG autoantibodies were responsible for the inhibition of macrophage removal of apoptotic cells through C3 deposition in MRL/lpr and NZB/W mice70. On the other hand, C3 deposition on nucleic acids is significantly reduced when dsDNA-specific IgG autoantibodies are present in the serum, which increase as disease progresses in MRL/lpr mice71.\n\nIn both human and experimental models of lupus nephritis, the complement system has a dual role. The classic pathway contributes to its protective role in the clearance of apoptotic material and circulating IC32,72. In theory chronic systemic complement inhibition may increase circulating IC and exacerbate disease44. To selectively target the desired complement regulator to the site of tissue injury, Tomlinson et al. used CR2 as the “guide” of complement inhibitors, given that sites of complement activation are “marked” by the presence of C3d73, and CR2 has natural affinity for C3d. Thus, low doses of chimeric CR2-DAF and CR2-CD59 efficiently protected target cells from complement attack74. CR2-DAF was targeted to the glomerulus in lupus nephritis while soluble DAF failed74. Long term (8 week) treatment of diseased MRL/lpr mice with low doses of CR2-Crry and CR2-CFH provided significant complement inhibition locally in the lupus glomerulus, which conferred significant reduction of proteinuria, as well as prolonged survival in these mice75,76. CR2-CFH treatment also significantly reduced glomerulonephritis76. The fact that selective alternative pathway inhibition by CR2-CFH provided significant reduction in glomerulonephritis while total complement inhibition with CR2-Crry did not, support the hypothesis that the alternative and classical pathways of complement have distinct roles in the pathogenesis of lupus nephritis76,77.\n\nInvestigations of complement activation in lupus nephritis did not only suggest potential therapies, but also diagnostic and monitoring tools. Using CR2-targeted MRI, glomerular C3b/iC3b/C3d deposition in MRL/lpr mice was found to increase with disease activity, suggesting its usage as a marker of onset and severity of lupus glomerulonephritis, which could be used to monitor the disease course and its response to therapy, repeatedly and noninvasively78.\n\n\nComplement-targeted therapy\n\nStrategies to manipulate the complement system in different human diseases have followed from successful animal studies, including those using recombinant intrinsic complement regulators and blocking antibodies. In addition to the treatment approaches indicated in studies using experimental models discussed above, we will focus on several promising therapeutic approaches that have been used in the treatment of human diseases, and may potentially extend to the treatment of human SLE and lupus nephritis.\n\nSoluble CR1 (sCR1) was first developed in 1990, and has both decay accelerator and cofactor activity in classical and alternative complement pathways. A current therapeutic form of soluble CR1 designated as TP10 (Avant Immunotherapeutics, Nedham, MA) has been used in clinical trials in several human diseases, including acute respiratory distress syndrome79, ischemia-reperfusion injury in patients undergoing lung transplantation80 or cardiac surgery81. These studies showed that sCR1 was well tolerated, with low development of anti-sCR1 antibodies81. More recently, soluble CR1 was found to prevent dysregulation of the alternative pathway C3 convertases in dense deposit disease (DDD) and C3 glomerulonephritis (C3GN), supporting clinical trial of soluble CR1 as a treatment for DDD and C3GN82.\n\nC1 esterase inhibitor (C1-INH) has been marketed as replacement therapy for hereditary angioedema (HAE) and ischemia-reperfusion injury. Plasma derived C1-INH has shown to be effective in prophylaxis, reducing the frequency of attacks of HAE, and shortening the duration of the acute attacks83. Recombinant human C1-INH also proved to be effective in the treatment of HAE patients during acute attacks84. C1-INH was also reported to successfully ameliorate both complement deposition on red blood cells (RBCs) and hemolysis in autoimmune hemolytic anemia85.\n\nThe most extensively used antibody targeting the complement system is a humanized monoclonal antibody (Eculizumab, Alexion Pharmaceuticals, Inc., Cheshire, CT) that directly binds human C5 and prevents its cleavage to C5a and C5b. In March 2007, Eculizumab (Soliris) is the first and the only US Food and Drug Administration (FDA) approved complement-inhibitor used in the treatment of human diseases. It was approved by the FDA for the treatment of Paroxysmal Nocturnal Hemoglobinuria (PNH), an acquired disorder of GPI-linked proteins including DAF and CD59 characterized by spontaneous complement activation and C5b-9-mediated hemolysis. It was subsequently approved for the treatment of atypical hemolytic uremic syndrome (aHUS) in 2011. Different phase 3 clinical trials showed that Eculizumab treatment showed better stabilization of hemoglobin, and reduced intravascular hemolysis with significant clinical improvements86, even with long term treatment87. Eculizumab increases aHUS patients’ platelet count, with 80% of patients achieving thrombotic microangiopathic event-free status88. With this beneficial effect of Eculizumab in aHUS, it is encouraging that a phase I trial has shown it is safe, with a prolonged interval of complement inhibition in SLE patients. Moving to the kidney, a multi-center phase II trial in the United States in which 122 patients with idiopathic membranous nephropathy were enrolled in a randomized placebo-controlled study of Eculizumab has been completed. Unfortunately, there was no difference comparing treatment with placebo in the primary outcome variable of urinary protein excretion. Because of the short-term treatment strategy (16 weeks) in a long-term disease, the study design may have been insufficient to uncover a true therapeutic effect. This is supported by the finding of an apparent benefit in patients enrolled in an open-label extension. Based on the impressive effects of long-term treatment with anti-C5 effects in lupus nephritis in NZB/W lupus mice57, Furie et al. conducted a phase I study with eculizumab in 24 patients with SLE. In this single center, randomized, placebo-controlled, double-blind, dose-ranging trial, patients were given a single intravenous dose of Eculizumab or placebo and followed for two months. Only mild adverse events were reported, and complement inhibition for 10 days in the 8 mg/kg group was observed89. We also designed a multi-center phase II trial using Eculizumab in proliferative lupus nephritis supported by the United States’ National Institutes of Health. Unfortunately, after enrollment of our first patient, this study encountered logistical delays and ultimately came to a complete halt. While this reflects the difficulties in clinical trials for a disorder such as lupus nephritis, based on what is known about the pathophysiology we remain interested in the potential efficacy of this drug in the treatment of lupus nephritis.", "appendix": "Author contributions\n\n\n\nAll authors contributed to the collaborative writing of this review. All authors have seen and agreed to the final content of this manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declared no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nWalport MJ: Complement. First of two parts. N Engl J Med. 2001; 344(14): 1058–1066. PubMed Abstract \| Publisher Full Text\n\nElward K, Griffiths M, Mizuno M, et al.: CD46 plays a key role in tailoring innate immune recognition of apoptotic and necrotic cells. J Biol Chem. 2005; 280(43): 36342–36354. PubMed Abstract \| Publisher Full Text\n\nGaboriaud C, Thielens NM, Gregory LA, et al.: Structure and activation of the C1 complex of complement: unraveling the puzzle. Trends Immunol. 2004; 25(7): 368–373. PubMed Abstract \| Publisher Full Text\n\nWallis R, Dodds AW, Mitchell DA, et al.: Molecular interactions between MASP-2, C4, and C2 and their activation fragments leading to complement activation via the lectin pathway. J Biol Chem. 2007; 282(11): 7844–7851. PubMed Abstract \| Publisher Full Text\n\nNicholson-Weller A, Halperin JA: Membrane signaling by complement C5b-9, the membrane attack complex. Immunol Res. 1993; 12(3): 244–257. PubMed Abstract \| Publisher Full Text\n\nJohnson RJ, Hillmen P: Paroxysmal nocturnal haemoglobinuria: nature's gene therapy? Mol Pathol. 2002; 55(3): 145–152. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarroll MC, Alicot EM, Katzman PJ, et al.: Organization of the genes encoding complement receptors type 1 and 2, decay-accelerating factor, and C4-binding protein in the RCA locus on human chromosome 1. J Exp Med. 1988; 167(4): 1271–1280. PubMed Abstract \| Free Full Text\n\nKingsmore SF, Vik DP, Kurtz CB, et al.: Genetic organization of complement receptor-related genes in the mouse. J Exp Med. 1989; 169(4): 1479–1484. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHourcade D, Holers VM, Atkinson JP: The regulators of complement activation (RCA) gene cluster. Adv Immunol. 1989; 45: 381–416. PubMed Abstract \| Publisher Full Text\n\nLiszewski MK, Post TW, Atkinson JP: Membrane cofactor protein (MCP or CD46): newest member of the regulators of complement activation gene cluster. Annu Rev Immunol. 1991; 9: 431–455. PubMed Abstract \| Publisher Full Text\n\nCardone J, Le Friec G, Vantourout P, et al.: Complement regulator CD46 temporally regulates cytokine production by conventional and unconventional T cells. Nat Immunol. 2010; 11(9): 862–871. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiwa T, Nonaka M, Okada N, et al.: Molecular cloning of rat and mouse membrane cofactor protein (MCP, CD46): preferential expression in testis and close linkage between the mouse Mcp and Cr2 genes on distal chromosome 1. Immunogenetics. 1998; 48(6): 363–371. PubMed Abstract \| Publisher Full Text\n\nLublin DM, Atkinson JP: Decay-accelerating factor: biochemistry, molecular biology, and function. Annu Rev Immunol. 1989; 7: 35–58. PubMed Abstract \| Publisher Full Text\n\nAhearn JM, Fearon DT: Structure and function of the complement receptors, CR1 (CD35) and CR2 (CD21). Adv Immunol. 1989; 46: 183–219. PubMed Abstract \| Publisher Full Text\n\nKim YU, Kinoshita T, Molina H, et al.: Mouse complement regulatory protein Crry/p65 uses the specific mechanisms of both human decay-accelerating factor and membrane cofactor protein. J Exp Med. 1995; 181(1): 151–159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorgan P, Harris C: Regulation in the activation pathways. Complement regulatory proteins. San Diego, CA, Academic Press, 1999; 41–136. Publisher Full Text\n\nMorgan P, Harris C: Regulation in the terminal pathway. Complement regulatory proteins. San Diego, CA, Academic Press, 1999; 137–170. Publisher Full Text\n\nde Zubiria Salgado A, Herrera-Diaz C: Lupus nephritis: an overview of recent findings. Autoimmune Dis. 2012; 2012: 849684. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nValentijn RM, van Overhagen H, Hazevoet HM, et al.: The value of complement and immune complex determinations in monitoring disease activity in patients with systemic lupus erythematosus. Arthritis Rheum. 1985; 28(8): 904–913. PubMed Abstract \| Publisher Full Text\n\nDalmasso AP: Complement in the pathophysiology and diagnosis of human diseases. Crit Rev Clin Lab Sci. 1986; 24(2): 123–183. PubMed Abstract\n\nManzi S, Rairie JE, Carpenter AB, et al.: Sensitivity and specificity of plasma and urine complement split products as indicators of lupus disease activity. Arthritis Rheum. 1996; 39(7): 1178–1188. PubMed Abstract \| Publisher Full Text\n\nJönsen A, Nilsson SC, Ahlqvist E, et al.: Mutations in genes encoding complement inhibitors CD46 and CFH affect the age at nephritis onset in patients with systemic lupus erythematosus. Arthritis Res Ther. 2011; 13(6): R206. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAlegretti AP, Schneider L, Piccoli AK, et al.: Diminished expression of complement regulatory proteins on peripheral blood cells from systemic lupus erythematosus patients. Clin Dev Immunol. 2012; 2012: 725684. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArora V, Mondal AM, Grover R, et al.: Modulation of CR1 transcript in systemic lupus erythematosus (SLE) by IFN-gamma and immune complex. Mol Immunol. 2007; 44(7): 1722–1728. PubMed Abstract \| Publisher Full Text\n\nArora V, Grover R, Kumar A, et al.: Relationship of leukocyte CR1 transcript and protein with the pathophysiology and prognosis of systemic lupus erythematosus: a follow-up study. Lupus. 2011; 20(10): 1010–1018. PubMed Abstract \| Publisher Full Text\n\nBiswas B, Kumar U, Das N: Expression and significance of leukocyte membrane cofactor protein transcript in systemic lupus erythematosus. Lupus. 2012; 21(5): 517–525. PubMed Abstract \| Publisher Full Text\n\nRoss GD, Yount WJ, Walport MJ, et al.: Disease-associated loss of erythrocyte complement receptors (CR1, C3b receptors) in patients with systemic lupus erythematosus and other diseases involving autoantibodies and/or complement activation. J Immunol. 1985; 135(3): 2005–2014. PubMed Abstract\n\nManderson AP, Botto M, Walport MJ: The role of complement in the development of systemic lupus erythematosus. Annu Rev Immunol. 2004; 22: 431–456. PubMed Abstract \| Publisher Full Text\n\nArkwright PD, Riley P, Hughes SM, et al.: Successful cure of C1q deficiency in human subjects treated with hematopoietic stem cell transplantation. J Allergy Clin Immunol. 2014; 133(1): 265–267. PubMed Abstract \| Publisher Full Text\n\nMehta P, Norsworthy PJ, Hall AE, et al.: SLE with C1q deficiency treated with fresh frozen plasma: a 10-year experience. Rheumatology (Oxford). 2010; 49(4): 823–824. PubMed Abstract \| Publisher Full Text\n\nYin Y, Wu X, Shan G, et al.: Diagnostic value of serum anti-C1q antibodies in patients with lupus nephritis: a meta-analysis. Lupus. 2012; 21(10): 1088–1097. PubMed Abstract \| Publisher Full Text\n\nGullstrand B, Mårtensson U, Sturfelt G, et al.: Complement classical pathway components are all important in clearance of apoptotic and secondary necrotic cells. Clin Exp Immunol. 2009; 156(2): 303–311. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNielsen CT, Østergaard O, Stener L, et al.: Increased IgG on cell-derived plasma microparticles in systemic lupus erythematosus is associated with autoantibodies and complement activation. Arthritis Rheum. 2012; 64(4): 1227–1236. PubMed Abstract \| Publisher Full Text\n\nDegn SE, Jensenius JC, Thiel S: Disease-causing mutations in genes of the complement system. Am J Hum Genet. 2011; 88(6): 689–705. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLood C, Eriksson S, Gullstrand B, et al.: Increased C1q, C4 and C3 deposition on platelets in patients with systemic lupus erythematosus--a possible link to venous thrombosis? Lupus. 2012; 21(13): 1423–1432. PubMed Abstract \| Publisher Full Text\n\nLood C, Tydén H, Gullstrand B, et al.: Platelet activation and anti-phospholipid antibodies collaborate in the activation of the complement system on platelets in systemic lupus erythematosus. PLoS One. 2014; 9(6): e99386. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParra S, Vives G, Ferré R, et al.: Complement system and small HDL particles are associated with subclinical atherosclerosis in SLE patients. Atherosclerosis. 2012; 225(1): 224–230. PubMed Abstract \| Publisher Full Text\n\nHahn BH, Kono DH: Animal models of SLE. in DJ W, BH H (eds): Dubois’ lupus erythematosus and related syndromes: Expert consult. Philadelphia, PA Elsevier saunders, 2013; 190–236. Publisher Full Text\n\nMiwa T, Maldonado MA, Zhou L, et al.: Deletion of decay-accelerating factor (CD55) exacerbates autoimmune disease development in MRL/lpr mice. Am J Pathol. 2002; 161(3): 1077–1086. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiwa T, Maldonado MA, Zhou L, et al.: Decay-accelerating factor ameliorates systemic autoimmune disease in MRL/lpr mice via both complement-dependent and -independent mechanisms. Am J Pathol. 2007; 170(4): 1258–1266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXu C, Mao D, Holers VM, et al.: A critical role for murine complement regulator Crry in fetomaternal tolerance. Science. 2000; 287(5452): 498–501. PubMed Abstract \| Publisher Full Text\n\nBao L, Wang Y, Chang A, et al.: Unrestricted C3 activation occurs in Crry-deficient kidneys and rapidly leads to chronic renal failure. J Am Soc Nephrol. 2007; 18(3): 811–822. PubMed Abstract \| Publisher Full Text\n\nBao L, Haas M, Boackle SA, et al.: Transgenic expression of a soluble complement inhibitor protects against renal disease and promotes survival in MRL/lpr mice. J Immunol. 2002; 168(7): 3601–3607. PubMed Abstract \| Publisher Full Text\n\nBao L, Haas M, Kraus DM, et al.: Administration of a soluble recombinant complement C3 inhibitor protects against renal disease in MRL/lpr mice. J Am Soc Nephrol. 2003; 14(3): 670–679. PubMed Abstract \| Publisher Full Text\n\nBao L, Zhou J, Holers VM, et al.: Excessive matrix accumulation in the kidneys of MRL/lpr lupus mice is dependent on complement activation. J Am Soc Nephrol. 2003; 14(10): 2516–2525. PubMed Abstract \| Publisher Full Text\n\nTakahashi K, Kozono Y, Waldschmidt TJ, et al.: Mouse complement receptors type 1 (CR1;CD35) and type 2 (CR2;CD21): expression on normal B cell subpopulations and decreased levels during the development of autoimmunity in MRL/lpr mice. J Immunol. 1997; 159(3): 1557–1569. PubMed Abstract\n\nWilson JG, Ratnoff WD, Schur PH, et al.: Decreased expression of the C3b/C4b receptor (CR1) and the C3d receptor (CR2) on B lymphocytes and of CR1 on neutrophils of patients with systemic lupus erythematosus. Arthritis Rheum. 1986; 29(6): 739–747. PubMed Abstract \| Publisher Full Text\n\nBoackle SA, Culhane KK, Brown JM, et al.: CR1/CR2 deficiency alters IgG3 autoantibody production and IgA glomerular deposition in the MRL/lpr model of SLE. Autoimmunity. 2004; 37(2): 111–123. PubMed Abstract \| Publisher Full Text\n\nBarnum SR, Dahlbäck B: C4b-binding protein, a regulatory component of the classical pathway of complement, is an acute-phase protein and is elevated in systemic lupus erythematosus. Complement Inflamm. 1990; 7(2): 71–77. PubMed Abstract\n\nWenderfer SE, Soimo K, Wetsel RA, et al.: Analysis of C4 and the C4 binding protein in the MRL/lpr mouse. Arthritis Res Ther. 2007; 9(5): R114. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlom AM, Nandakumar KS, Holmdahl R: C4b-binding protein (C4BP) inhibits development of experimental arthritis in mice. Ann Rheum Dis. 2009; 68(1): 136–142. PubMed Abstract \| Publisher Full Text\n\nElliott MK, Jarmi T, Ruiz P, et al.: Effects of complement factor D deficiency on the renal disease of MRL/lpr mice. Kidney Int. 2004; 65(1): 129–138. PubMed Abstract \| Publisher Full Text\n\nWatanabe H, Garnier G, Circolo A, et al.: Modulation of renal disease in MRL/lpr mice genetically deficient in the alternative complement pathway factor B. J Immunol. 2000; 164(2): 786–794. PubMed Abstract \| Publisher Full Text\n\nMehta G, Ferreira VP, Skerka C, et al.: New insights into disease-specific absence of complement factor H related protein C in mouse models of spontaneous autoimmune diseases. Mol Immunol. 2014; 62(1): 235–248. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBao L, Haas M, Quigg RJ: Complement factor H deficiency accelerates development of lupus nephritis. J Am Soc Nephrol. 2011; 22(2): 285–295. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPickering MC, Cook HT, Warren J, et al.: Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H. Nat Genet. 2002; 31(4): 424–428. PubMed Abstract \| Publisher Full Text\n\nWang Y, Hu Q, Madri JA, et al.: Amelioration of lupus-like autoimmune disease in NZB/WF1 mice after treatment with a blocking monoclonal antibody specific for complement component C5. Proc Natl Acad Sci U S A. 1996; 93(16): 8563–8568. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiwa T, Zhou L, Maldonado MA, et al.: Absence of CD59 exacerbates systemic autoimmunity in MRL/lpr mice. J Immunol. 2012; 189(11): 5434–5441. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMizuno M, Blanchin S, Gasque P, et al.: High levels of complement C3a receptor in the glomeruli in lupus nephritis. Am J Kidney Dis. 2007; 49(5): 598–606. PubMed Abstract \| Publisher Full Text\n\nBao L, Osawe I, Puri T, et al.: C5a promotes development of experimental lupus nephritis which can be blocked with a specific receptor antagonist. Eur J Immunol. 2005; 35(8): 2496–2506. PubMed Abstract \| Publisher Full Text\n\nBao L, Osawe I, Haas M, et al.: Signaling through up-regulated C3a receptor is key to the development of experimental lupus nephritis. J Immunol. 2005; 175(3): 1947–1955. PubMed Abstract \| Publisher Full Text\n\nWenderfer SE, Ke B, Hollmann TJ, et al.: C5a receptor deficiency attenuates T cell function and renal disease in MRLlpr mice. J Am Soc Nephrol. 2005; 16(12): 3572–3582. PubMed Abstract \| Publisher Full Text\n\nWenderfer SE, Wang H, Ke B, et al.: C3a receptor deficiency accelerates the onset of renal injury in the MRL/lpr mouse. Mol Immunol. 2009; 46(7): 1397–1404. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMathieu MC, Sawyer N, Greig GM, et al.: The C3a receptor antagonist SB 290157 has agonist activity. Immunol Lett. 2005; 100(2): 139–145. PubMed Abstract \| Publisher Full Text\n\nSekine H, Reilly CM, Molano ID, et al.: Complement component C3 is not required for full expression of immune complex glomerulonephritis in MRL/lpr mice. J Immunol. 2001; 166(10): 6444–6451. PubMed Abstract \| Publisher Full Text\n\nSheerin NS, Springall T, Carroll M, et al.: Altered distribution of intraglomerular immune complexes in C3-deficient mice. Immunology. 1999; 97(3): 393–399. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaylor PR, Carugati A, Fadok VA, et al.: A hierarchical role for classical pathway complement proteins in the clearance of apoptotic cells in vivo. J Exp Med. 2000; 192(3): 359–366. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBotto M, Dell'Agnola C, Bygrave AE, et al.: Homozygous C1q deficiency causes glomerulonephritis associated with multiple apoptotic bodies. Nat Genet. 1998; 19(1): 56–59. PubMed Abstract \| Publisher Full Text\n\nGaipl US, Beyer TD, Heyder P, et al.: Cooperation between C1q and DNase I in the clearance of necrotic cell-derived chromatin. Arthritis Rheum. 2004; 50(2): 640–649. PubMed Abstract \| Publisher Full Text\n\nKenyon KD, Cole C, Crawford F, et al.: IgG autoantibodies against deposited C3 inhibit macrophage-mediated apoptotic cell engulfment in systemic autoimmunity. J Immunol. 2011; 187(5): 2101–2111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPapp K, Végh P, Tchorbanov A, et al.: Progression of lupus-like disease drives the appearance of complement-activating IgG antibodies in MRL/lpr mice. Rheumatology (Oxford). 2010; 49(12): 2273–2280. PubMed Abstract \| Publisher Full Text\n\nSchifferli JA, Ng YC, Peters DK: The role of complement and its receptor in the elimination of immune complexes. N Engl J Med. 1986; 315(8): 488–495. PubMed Abstract \| Publisher Full Text\n\nThurman JM, Kulik L, Orth H, et al.: Detection of complement activation using monoclonal antibodies against C3d. J Clin Invest. 2013; 123(5): 2218–2230. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong H, He C, Knaak C, et al.: Complement receptor 2-mediated targeting of complement inhibitors to sites of complement activation. J Clin Invest. 2003; 111(12): 1875–1885. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAtkinson C, Qiao F, Song H, et al.: Low-dose targeted complement inhibition protects against renal disease and other manifestations of autoimmune disease in MRL/lpr mice. J Immunol. 2008; 180(2): 1231–1238. PubMed Abstract \| Publisher Full Text\n\nSekine H, Kinser TT, Qiao F, et al.: The benefit of targeted and selective inhibition of the alternative complement pathway for modulating autoimmunity and renal disease in MRL/lpr mice. Arthritis Rheum. 2011; 63(4): 1076–1085. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSekine H, Ruiz P, Gilkeson GS, et al.: The dual role of complement in the progression of renal disease in NZB/W F1 mice and alternative pathway inhibition. Mol Immunol. 2011; 49(1–2): 317–323. PubMed Abstract \| Publisher Full Text\n\nSargsyan SA, Serkova NJ, Renner B, et al.: Detection of glomerular complement C3 fragments by magnetic resonance imaging in murine lupus nephritis. Kidney Int. 2012; 81(2): 152–159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZimmerman JL, Dellinger RP, Straube RC, et al.: Phase I trial of the recombinant soluble complement receptor 1 in acute lung injury and acute respiratory distress syndrome. Crit Care Med. 2000; 28(9): 3149–3154. PubMed Abstract \| Publisher Full Text\n\nKeshavjee S, Davis RD, Zamora MR, et al.: A randomized, placebo-controlled trial of complement inhibition in ischemia-reperfusion injury after lung transplantation in human beings. J Thorac Cardiovasc Surg. 2005; 129(2): 423–428. PubMed Abstract \| Publisher Full Text\n\nLazar HL, Bokesch PM, van Lenta F, et al.: Soluble human complement receptor 1 limits ischemic damage in cardiac surgery patients at high risk requiring cardiopulmonary bypass. Circulation. 2004; 110(11 Suppl 1): II274–279. PubMed Abstract \| Publisher Full Text\n\nZhang Y, Nester CM, Holanda DG, et al.: Soluble CR1 therapy improves complement regulation in C3 glomerulopathy. J Am Soc Nephrol. 2013; 24(11): 1820–1829. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCraig TJ, Levy RJ, Wasserman RL, et al.: Efficacy of human C1 esterase inhibitor concentrate compared with placebo in acute hereditary angioedema attacks. J Allergy Clin Immunol. 2009; 124(4): 801–808. PubMed Abstract \| Publisher Full Text\n\nRiedl MA, Levy RJ, Suez D, et al.: Efficacy and safety of recombinant C1 inhibitor for the treatment of hereditary angioedema attacks: a North American open-label study. Ann Allergy Asthma Immunol. 2013; 110(4): 295–299. PubMed Abstract \| Publisher Full Text\n\nWouters D, Stephan F, Strengers P, et al.: C1-esterase inhibitor concentrate rescues erythrocytes from complement-mediated destruction in autoimmune hemolytic anemia. Blood. 2013; 121(7): 1242–1244. PubMed Abstract \| Publisher Full Text\n\nBrodsky RA, Young NS, Antonioli E, et al.: Multicenter phase 3 study of the complement inhibitor eculizumab for the treatment of patients with paroxysmal nocturnal hemoglobinuria. Blood. 2008; 111(4): 1840–1847. PubMed Abstract \| Publisher Full Text\n\nHillmen P, Muus P, Röth A, et al.: Long-term safety and efficacy of sustained eculizumab treatment in patients with paroxysmal nocturnal haemoglobinuria. Br J Haematol. 2013; 162(1): 62–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKeating GM: Eculizumab: a review of its use in atypical haemolytic uraemic syndrome. Drugs. 2013; 73(18): 2053–2066. PubMed Abstract \| Publisher Full Text\n\nBarilla-Labarca ML, Toder K, Furie R: Targeting the complement system in systemic lupus erythematosus and other diseases. Clin Immunol. 2013; 148(3): 313–321. PubMed Abstract \| Publisher Full Text\n\nKemper C, Pangburn MK, Fishelson Z: Complement nomenclature 2014. Mol Immunol. 2014; 61(2): 56–58. PubMed Abstract \| Publisher Full Text" }	[ { "id": "9817", "date": "10 Aug 2015", "name": "Jin-Xiong She", "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\nBao et al. review the complement system in Lupus Nephritis and provided updated literature survey on the field. The review is well written over all. Some things which needs to be addressed are as follows:Title of the review is very generic, which does not elicit interest in the reader. A large part of the introduction was dedicated to explaining complement system. Description of the main steps would do a better job. Individual lupus models should be grouped under three animal models to provide a clearer view to the readers about these three models. The review ends abruptly with description on efforts and interests of authors regarding conducting a failed clinical trial. Lack of a take home message and valid conclusions.", "responses": [ { "c_id": "1521", "date": "18 Aug 2015", "name": "Lihua Bao", "role": "Author Response", "response": "On behalf of co-authors Dr. Quigg and Dr. Cunningham, we appreciate your positive comment of \"The review is well written over all\". We also seriously considered the five comments and will address each of these in our new version of the article. Once again thank you very much for you comments." } ] }, { "id": "9822", "date": "17 Aug 2015", "name": "Luc Teyton", "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 current article is a very thorough review of the complement system and is nicely put in the context of lupus. Besides the known functions of complement factors in opsonization and immune complex clearance, it would be nice to add a short paragraph on less conventional mechanisms of action of some complement factors such as lipid binding, a topic of high relevance when one discusses apoptotic cell clearance and the relationship between the complement and coagulation systems. The intracellular functions of C3 would also be worth discussing.", "responses": [] } ]	1	https://f1000research.com/articles/4-145
https://f1000research.com/articles/4-141/v1	04 Jun 15	{ "type": "Review", "title": "Targeting metastatic breast cancer: problems and potential", "authors": [ "Sarah Deasy", "Karol Szczepanek", "Kent W. Hunter", "Sarah Deasy", "Karol Szczepanek" ], "abstract": "Breast cancer is one of the leading causes of cancer-related mortality of women in the United States. Since the majority of cancer deaths are due to metastases rather than the primary tumor, a better understanding of the biological mechanisms that lead to metastatic disease is critical to reduce breast cancer associated mortality. Current adjuvant therapies use the same broadly cytotoxic and targeted strategies against metastases as are used against the primary tumor. However, resistance to chemotherapy due to the cellular dormancy, high genotypic and phenotypic heterogeneity between primary tumor and metastases as well as among individual metastases, and the limitations in detection of disseminated tumor cells and micrometastases significantly hinder the efficiency of currently available therapies. While it is crucial to directly address the issue of metastatic dormancy and evaluate for anti-metastatic therapy the relevance of molecular targets chosen based on primary tumor profiling, it is also imperative to address metastasis-specific mechanisms of growth and survival that are likely to be distinct from those of the primary tumor. We believe that a three-pronged approach to therapy will be necessary to deal with progressive disease: blocking of further dissemination after diagnosis; eradication of disseminated tumor cells and prevention of the dormant-to-proliferative switch of those remaining; and elimination of established metastatic tumors. The implementation of this strategy requires a greater depth of knowledge of metastasis driver and maintenance genes and suggests the need for a “Metastasis Genome Atlas” project to complement the current investigations into cancer genomic landscapes.", "keywords": [ "breast cancer", "metastatic", "metastases", "tumor" ], "content": "Introduction\n\nBreast cancer remains one of the leading causes of cancer-related mortality among women in the US1. Since the primary tumor is usually resected upon detection, the majority of mortality is due to metastases. While 5-year relative survival for patients diagnosed with localized breast cancer is 98.5%, it plummets to a dismal 25% for patients with distant overt metastases1. Moreover, an estimated 20–50% of breast cancer patients diagnosed at an early stage are expected to develop metastatic disease, which can occur years or even decades after surgical removal of the primary tumor2. These statistics indicate that the establishment of effective therapies that target and prevent metastasis is of critical importance. Although the past decade has seen significant advances in the development of methods for detection and treatments of primary breast cancer, these treatments are generally ineffective at eliminating metastatic disease. This suggests either an inherent biological difference between primary tumors and distant metastases, a role of the microenvironment at the secondary site in inducing therapeutic resistance, or likely both.\n\nRecent progress in breast cancer treatment has been characterized by a shift from cytotoxic drugs, which broadly target highly proliferating cells, to more targeted therapies designed to attack a specific molecule or pathway through a known mechanism of action. This approach has been possible due to enormous technological leaps in genome sequencing, molecular biology, cancer genetics, genomics, and bioinformatics that allow for the identification of individual patient mutations and clinically actionable targets from tumor biopsy samples. Sequence information and patient-derived xenograft methods have shown promise in identifying individualized therapeutic strategies that are effective for specific patients, rather than the traditional population-based clinical strategies. However, while various sequencing projects have characterized the genomic landscape of breast cancer3,4, relatively little is known about the somatic diversity of metastatic tumors. Expression profiling has revealed multiple subtypes with different prognostic outcomes5 and led to identification of gene signatures that can stratify patients into groups with low- or high-risk for developing distant metastatic disease6. Still, targeted therapies against metastases have proven to be more challenging to develop compared to those targeting the primary tumor, since the critical driver molecules have not yet been firmly established.\n\nIn this review, we discuss the major challenges facing the development of efficient therapies against metastasis. We also present a strategy that we believe will improve the targeting of metastatic disease: blocking of further dissemination after diagnosis, eradication of disseminated tumor cells and prevention of the dormant-to-proliferative switch of those cells not eradicated, and elimination of established metastatic tumors. This strategy requires the elucidation of mechanisms that drive the establishment and maintenance of metastases.\n\n\nCurrent anti-metastatic treatments and therapies in development\n\nCurrent therapies against metastatic breast cancer address different stages of tumor progression7. Broad-spectrum therapies, such as cytotoxic chemotherapy, aim to kill actively dividing cancer cells. According to an extensive meta-analysis, combination chemotherapy can increase the 15-year survival rate of patients with early breast cancer8. However, a substantial number of these patients develop metastases or recurrent disease years after cessation of treatment, indicating the presence of dormant cells at the secondary site that are resistant to standard therapies. Therefore, the pathways that allow cancer cells to remain viable in a dormant state would provide ideal targets for eliminating minimal residual disease and preventing metastatic outgrowth after the cessation of treatment. Among them are the survival pathways involving Src and p389–12. Recent findings have shown that pharmacological inhibition of Src family kinase (SFK) signaling inhibited the proliferative outgrowth of dormant disseminated cells and the development of macrometastatic lesions12. Furthermore, dormant cell proliferation required ERK1/2 activation, and the treatment of cells undergoing the dormant-to-proliferative switch with the combination of Src inhibitor (AZD0530, saracatinib, AstraZeneca) and MEK1/2 inhibitor (AZD6244, selumetinib, AstraZeneca) resulted in apoptosis of dormant cells; neither of the inhibitors alone achieved this effect. Since MEK1/2 is a well-described upstream activator of ERK1/2, these observations complement previous reports showing that an ERKlow/p38high signaling ratio promotes tumor cell quiescence through a combination of signaling pathways promoting adaptive and basal survival and G0-G1 quiescence11. These reports indicate that therapies targeting signaling involved in the dormant-to-proliferative switch might represent promising clinical interventions against metastatic disease.\n\nMajor efforts of the pharmaceutical industry are now focused on the development of novel targeted therapies that function in a molecule-specific manner7. Examples of targeted therapies that are presently in clinical use for breast cancer are aromatase/estrogen synthase inhibitors (anastrozole, letrozole, and exemestane)13 and trastuzumab (Herceptin®), a monoclonal antibody against epidermal growth factor receptor 2 (ERBB2/HER2/neu)14. Initially investigated as a monotherapy and later in combination with chemotherapy, trastuzumab significantly improved patient response rate, time to progression, and overall survival compared to chemotherapy alone14. However, its use is limited only to HER2-positive breast cancers, the expression of which, as explained in the next section of this review, may not be constant between primary tumors and metastases, reducing the effectiveness of this strategy.\n\nBased on their specific role in the metastatic cascade, potential molecular targets for therapeutic intervention can be categorized as involved in either metastasis initiation or progression15,16. Due to the early dissemination of tumor cells, which is discussed in the next section, molecules important in metastasis initiation, such as regulation of cellular motility, angiogenesis, and invasion of local extracellular matrix (ECM), may not be efficient targets for preventing metastatic disease since patients at risk already have disseminated cells by the time of primary tumor diagnosis. However, if those molecules also play key roles in the later stages (progression) of metastasis, such as continued trafficking after initial escape from the primary tumor17, their targeting might be of clinical importance. In particular, molecules that promote survival at the distant sites, the dormancy-to-proliferation switch, and colonization of distant organs are very attractive targets for drug development.\n\nThe molecules identified and currently being investigated comprise both the tumor cell autologous factors, such as those involved in signal transduction, adhesion, motility, growth and survival, and the microenvironmental factors, including resident stromal cells, components of the immune system, chemokines, and promoters of angiogenesis. A growing catalog of potential molecular targets identified for inhibition of metastatic progression has been extensively reviewed7,18,19, including a recent overview of drug candidates in current pharmaceutical pipelines that target the mechanisms of tumor cell migration20. Effective anti-metastatic therapies need to address metastasis-specific characteristics that allow for the active targeting of metastatic cells. In the next section, we will discuss in more detail the challenges facing the treatment of metastasis from the perspective of drug development, specifically heterogeneity within and between metastases, as well as between metastases and primary tumors, constraints in detection and analysis, and the timing of intervention.\n\n\nChallenges in the treatment of metastatic disease\n\nThe logical approach to preventing metastatic disease is to inhibit the initial dissemination step. Dissemination used to be considered a late-stage event in the linear model of malignant disease evolution, due to the time necessary to accumulate essential somatic mutations required by tumor cells for motility and survival beyond the primary site. However, evidence from mouse models of spontaneous mammary tumorigenesis has shown that tumor cells are able to disseminate to the bone marrow as early as four weeks of age, when mammary tissue only appears in atypical ductal hyperplasia (ADH) or ductal carcinoma in situ (DCIS) stages21. Lung micrometastases in the BALB-NeuT model were found at 20 weeks, a full 3 weeks before tumors became visibly invasive. Additionally, early disseminated tumor cells (DTCs)—cells that left the primary tumor and spread to another tissue via the circulation—could be released from growth arrest and had the capability to develop into metastases. The clinical relevance of these findings was confirmed using bone marrow samples from 607 breast cancer patients at various stages of disease, concluding that small tumors produced similar numbers of DTCs to late-stage tumors. Other evidence, including metastasis without invasive cancer such as DCIS22, cancers of unknown primary origin23, and accidental transfer of cancer from clinically disease-free organ donors24 all supported the authors’ observations of early dissemination. These reports suggest that breast cancer patients likely have DTCs even before their primary cancer is diagnosed, making the targeting of the origin of metastatic disease a less promising therapy option.\n\nOne of the major challenges facing oncologists, particularly in the case of breast cancer, concerns DTCs that remain in tissues after adjuvant therapy in a sub-clinical, currently undetectable state, known as dormancy. Upon dissemination, cancer cells can enter dormancy either as individual non-proliferative, quiescent cells or as undetectable micrometastases held in a reduced or balanced proliferative/apoptotic state by lack of adequate resources (angiogenic dormancy) or an active immune system (immunogenic dormancy)25,26. These cells or micrometastases can eventually re-emerge and become proliferative years after a patient is thought to be cancer-free. Since the mechanisms of action of many cytotoxic therapies disrupt aspects of the mitotic process, these agents are most likely ineffective against the low or non-proliferative dormant cells27. There is also accumulating evidence from experimental models that activated survival mechanisms may be playing a role in protecting DTCs25. Regardless of the exact mechanism, DTCs are resistant to the broad-spectrum therapeutic strategies currently in use. Targeted therapy may provide additional benefit, but therapeutic regimens based on the primary tumor may not be effective in DTCs.\n\nAccumulating evidence indicates that metastases can be genetically and phenotypically different from the primary tumor28. This phenomenon has been reported for overt metastases that exhibit the opposite estrogen receptor, progesterone receptor, or HER-2 expression profile of the primary tumor. A meta-analysis examining estrogen receptor, progesterone receptor, and/or HER-2 expression in patients’ primary breast tumors and matched recurrent metastases included 48 studies published between 1983 and 201129 showing 20%, 33%, and 8% discordance for estrogen receptor, progesterone receptor, and HER-2 status, respectively, with a higher prevalence of positive to negative conversion. The authors suggest that this observation may be the result of individual DTCs reflecting heterogeneity of receptor expression within the primary tumor. Following endocrine therapy, the remaining resistant subpopulations of primary tumor cells are selectively able to disseminate to and grow at secondary sites, producing metastases with a different receptor expression profile than the majority of the original primary tumor. Since treatments are often determined based on the characteristics of the primary tumor, discordance between the primary tumor and metastatic phenotypes can render the treatment ineffective against DTCs. Conversely, some patients with receptor-negative primary tumors but positive metastases may benefit from available targeted therapies but currently are unlikely to receive these therapeutic options due to the clinical characteristics of the primary tumor.\n\nIn addition to differences between primary and metastatic tumors, there is also heterogeneity among individual metastases. More than one million cells per gram of tumor can be shed from the primary mass every day30 and these cells can come from any of the diverse subpopulations that compose the tumor at any point during tumor evolution. Therefore, the DTCs are a genetically and phenotypically heterogeneous group. However, not all of these cells have the ability to grow into clinically relevant distant lesions, otherwise every patient would have a high metastatic burden. In reality, the metastatic process is incredibly inefficient31,32, suggesting a great deal of selective pressure is experienced by DTCs and only those with the proper combination of characteristics are able to successfully colonize a secondary site. Though the successfully metastatic cells probably share many characteristics, the genetic changes that underlie those characteristics may not be identical, resulting in heterogeneity between metastases in addition to differences from the primary tumor33. Similar to the evaluation of HER-2 expression between primary tumor and metastases discussed earlier, an investigation into HER-2 expression revealed that 18% of metastases within the same patient showed discordant expression34. In another examination of single primary and metastatic breast cancer cells using immunofluorescence in situ hybridization (iFISH), cancer subtype markers, and statistical analysis, investigators measured genetic and phenotypic heterogeneity between distant metastases35. These results indicated significant diversity between distant metastases within the same patient. Furthermore, the authors observed substantial differences between metastases in the proportions of cells displaying subtype markers, indicating both genetic and phenotypic differences were prevalent between distant metastases from the same patient.\n\nAn additional challenge to successful treatment of metastatic disease is the ability to deliver the therapy to the metastatic tumor cells. Metastases are often present in multiple sites within a patient and drug accessibility between tissues, for example bone versus lung, may be significantly different. Certain tissues may act as “pharmacological sanctuaries” that protect cancer cells from available therapies due to barriers that prevent access to those cells. Evidence suggests that at least some brain metastases are shielded from pharmacologic intervention by the blood-brain barrier36. To cure patients of metastatic disease, effective therapies will have to be deliverable to all of the potential metastatic sites within the body. Since microenvironments differ in various secondary sites, combination anti-metastasis therapies may be necessary to deal with both intra-tumoral genetic heterogeneity as well as microenvironmentally-induced heterogeneity.\n\nIn summary, current adjuvant therapies are expected to control metastasis by targeting residual disease remaining after surgery. However, the use of adjuvant therapy is estimated to reduce the relative risk of relapse by only 19–37% depending on the age of the patient37. Thus, neither the current cytotoxic nor targeted adjuvant approaches appear to be as effective at preventing metastases as they are at targeting the primary tumor. Finding new molecular targets for preventing the growth of metastatic tumors should be based on the realization of how different this process is from early tumorigenesis.\n\n\nTake-home lessons for optimal design of future anti-metastatic therapies\n\nBased on our current knowledge of metastasis biology, we envision a three-pronged strategy for clinical intervention for progressive disease: prevention, stasis, and destruction. The prevention arm would be primarily applied as a neo-adjuvant therapy intended to reduce the potential metastatic capacity of tumor cells naturally shed from the primary tumor between diagnosis and surgical removal. As discussed above, this might require combination therapy for maximal efficiency since it has been demonstrated that tumor cells use multiple motility mechanisms (amoeboid, epithelial-to-mesenchymal transition, collective migration38) to initiate dissemination. This arm might also be used to help suppress potential metastatic cells displaced during surgical resection itself. Since breast cancers undergo early dissemination, the prevention strategy would not necessarily be effective against cells that disseminated to distant sites prior to diagnosis. For these cells, cytostatic or cytotoxic therapies need to be developed, specific to the biology of disseminated cells, to either hold the disseminated cells in a sub-clinical state for the remaining natural lifetime of the patient, or to eliminate the metastatic “seed” before it can establish clinically relevant lesions. Finally, for those patients unfortunate enough to develop metastases, therapeutics specifically developed and targeted to metastatic biology need to be developed to reduce morbidity and mortality associated with these secondary tumors.\n\nFor these strategies to work, appropriate targets that drive the establishment and maintenance of metastasis have to be identified. Regrettably, despite significant advances in our understanding of the cellular biology of metastatic disease, our knowledge of metastasis “driver” genes remains limited. This is due partially to the difficulty of studying the process, since much of the important biology occurs in a small fraction of disseminated single cells or micrometastases. At present, it is not possible to easily identify those cells that will evolve into macroscopic lesions for in depth analysis, or exactly when that critical event occurs. In addition, metastatic samples are difficult to obtain from human patients since they are not usually resected, but instead are subject to systemic therapy. Thus, although the common mutations that drive primary breast cancer are now known thanks to large sequencing projects, much less is known about the mutational spectrum in metastatic disease. Finally, there may be many genes that are associated with metastatic progression. Analysis of gene expression-based prognostic signatures suggests that there may be hundreds or even thousands of genes whose expression is correlated with the disease outcome39. Many of these genes are most likely expressed in the tumor microenvironment and therefore investigations solely in tumor cells have the potential to miss important molecules targetable for clinical intervention.\n\nContinued investigations based on the biology of primary tumors for the development of novel therapeutics are clearly important, and undoubtedly will provide additional benefit to patients. However, this strategy is potentially fraught with dangers, since therapies that can shrink primary tumors have been shown to have little effect, or even potentially adverse effects, on metastatic disease40,41. Furthermore, work from many laboratories has identified metastasis-associated genes that are not frequently mutated during cancer progression but are epigenetically silenced or exert their metastatic influence through alterations at the transcriptional level. Our own laboratory uses a meiotic genetics approach to identify somatically inherited polymorphisms that affect patients’ susceptibility to metastasis. This approach has generated a growing list of metastasis susceptibility genes with tumor-autonomous or stromal effects, including some with the potential to be actionable clinical targets42–48. The results of these studies suggest that a more comprehensive examination of metastasis biology, incorporating both genomic and transcriptional landscape mapping, may provide an entirely different set of genes that might be more advantageously targeted to specifically reduce metastatic burden, without necessarily impacting the primary tumor.\n\nNow that The Cancer Genome Atlas (TCGA) and other genomics projects have provided detailed information regarding the etiology of primary tumors, it is time to initiate similar projects for metastatic disease. Since metastatic tumors represent highly selected cells from a subset of the primary tumor localized in completely different microenvironments, these lesions should be recognized as related (but distinct) tumors with different biology. Thus, treatment strategies that are developed based on specific features of metastasis biology, used in parallel with those designed against the primary tumor, will provide a more effective method for combatting the final lethal stages of neoplastic disease than those strategies developed solely against the primary tumor. A Metastasis Genome Atlas Project would, however, have additional challenges compared to the TCGA. The major challenge would be access to the metastatic tumor samples since surgical resection of metastases is usually not performed due to increased risk without perceived patient benefit. In addition, metastatic tissue collected postmortem may be compromised or selected by therapies that potentially induce genomic changes, complicating analysis and interpretation of the results. Furthermore, it will be important to screen multiple metastases from different secondary sites within individual patients to help identify tissue-specific differences that might expose therapeutic vulnerabilities or resistance. Regardless of the challenges facing the Metastasis Genome Atlas Project concept, improved therapeutic strategies to reduce breast cancer morbidity and mortality would significantly benefit from a better understanding of the biology underlying the primary cause of patient distress, disseminated metastatic disease.", "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\nReferences\n\nHowlader N, Noone AM, Krapcho M, et al.: SEER Cancer Stat Rev, 1975–2011. 2014. Reference Source\n\nLu J, Steeg PS, Price JE, et al.: Breast cancer metastasis: challenges and opportunities. Cancer Res. 2009; 69(12): 4951–3. PubMed Abstract \| Publisher Full Text\n\nWatson IR, Takahashi K, Futreal PA, et al.: Emerging patterns of somatic mutations in cancer. Nat Rev Genet. 2013; 14(10): 703–18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGreaves M, Maley CC: Clonal evolution in cancer. Nature. 2012; 481(7381): 306–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerou CM, Sørlie T, Eisen MB, et al.: Molecular portraits of human breast tumours. Nature. 2000; 406(6797): 747–52. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\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 \| F1000 Recommendation\n\nEckhardt BL, Francis PA, Parker BS, et al.: Strategies for the discovery and development of therapies for metastatic breast cancer. Nat Rev Drug Discov. 2012; 11(6): 479–97. PubMed Abstract \| Publisher Full Text\n\nEarly Breast Cancer Trialists' Collaborative Group. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet. 2005; 365(9472): 1687–717. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nZhang XH, Wang Q, Gerald W, et al.: Latent bone metastasis in breast cancer tied to Src-dependent survival signals. Cancer cell. 2009; 16(1): 67–78. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nMarshall JC, Collins JW, Nakayama J, et al.: Effect of inhibition of the lysophosphatidic acid receptor 1 on metastasis and metastatic dormancy in breast cancer. J Natl Cancer Inst. 2012; 104(17): 1306–19. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nSosa MS, Avivar-Valderas A, Bragado P, et al.: ERK1/2 and p38α/β signaling in tumor cell quiescence: opportunities to control dormant residual disease. Clin Cancer Res. 2011; 17(18): 5850–7. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nEl Touny LH, Vieira A, Mendoza A, et al.: Combined SFK/MEK inhibition prevents metastatic outgrowth of dormant tumor cells. J Clin Invest. 2014; 124(1): 156–68. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nFerretti G, Bria E, Giannarelli D, et al.: Second- and third-generation aromatase inhibitors as first-line endocrine therapy in postmenopausal metastatic breast cancer patients: a pooled analysis of the randomised trials. Br J Cancer. 2006; 94(12): 1789–96. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nRimawi MF, Schiff R, Osborne CK: Targeting HER2 for the Treatment of Breast Cancer. Annu Rev Med. 2015; 66: 111–28. PubMed Abstract \| Publisher Full Text\n\nNguyen DX, Bos PD, Massagué J: Metastasis: from dissemination to organ-specific colonization. Nat Rev Cancer. 2009; 9(4): 274–84. PubMed Abstract \| Publisher Full Text\n\nMina LA, Sledge GW Jr: Rethinking the metastatic cascade as a therapeutic target. Nat Rev Clin Oncol. 2011; 8(6): 325–32. PubMed Abstract \| Publisher Full Text\n\nKim MY, Oskarsson T, Acharyya S, et al.: Tumor self-seeding by circulating cancer cells. Cell. 2009; 139(7): 1315–26. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nWang SH, Lin SY: Tumor dormancy: potential therapeutic target in tumor recurrence and metastasis prevention. Exp Hematol Oncol. 2013; 2(1): 29. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nRomero I, Garrido F, Garcia-Lora AM: Metastases in immune-mediated dormancy: a new opportunity for targeting cancer. Cancer Res. 2014; 74(23): 6750–7. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nSchaap-Nutt A, Thyssen DA, Drell DW, et al.: Anti-metastatics: an overview of drug candidates in current pipelines. Curr Pharm Des. 2014; 20(42): 6522–8. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nHüsemann Y, Geigl JB, Schubert F, et al.: Systemic spread is an early step in breast cancer. Cancer cell. 2008; 13(1): 58–68. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nCutuli B, Cohen-Solal-Le Nir C, De Lafontan B, et al.: Ductal carcinoma in situ of the breast results of conservative and radical treatments in 716 patients. Eur J Cancer. 2001; 37(18): 2365–72. PubMed Abstract \| F1000 Recommendation\n\nvan de Wouw AJ, Janssen-Heijnen ML, Coebergh JW, et al.: Epidemiology of unknown primary tumours; incidence and population-based survival of 1285 patients in Southeast Netherlands, 1984–1992. Eur J Cancer. 2002; 38(3): 409–13. PubMed Abstract \| F1000 Recommendation\n\nRiethmüller G, Klein CA: Early cancer cell dissemination and late metastatic relapse: clinical reflections and biological approaches to the dormancy problem in patients. Semin Cancer Biol. 2001; 11(4): 307–11. PubMed Abstract \| Publisher Full Text\n\nAguirre-Ghiso JA: Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer. 2007; 7(11): 834–46. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nSosa MS, Bragado P, Aguirre-Ghiso JA: Mechanisms of disseminated cancer cell dormancy: an awakening field. Nat Rev Cancer. 2014; 14(9): 611–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNaumov GN, Townson JL, MacDonald IC, et al.: Ineffectiveness of doxorubicin treatment on solitary dormant mammary carcinoma cells or late-developing metastases. Breast Cancer Res Treat. 2003; 82(3): 199–206. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nWelch DR: Microarrays bring new insights into understanding of breast cancer metastasis to bone. Breast Cancer Res. 2004; 6(2): 61–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAurilio G, Disalvatore D, Pruneri G, et al.: A meta-analysis of oestrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 discordance between primary breast cancer and metastases. Eur J Cancer. 2014; 50(2): 277–89. PubMed Abstract \| F1000 Recommendation\n\nButler TP, Gullino PM: Quantitation of cell shedding into efferent blood of mammary adenocarcinoma. Cancer Res. 1975; 35(3): 512–6. PubMed Abstract\n\nLuzzi KJ, MacDonald IC, Schmidt EE, et al.: Multistep nature of metastatic inefficiency: dormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am J Pathol. 1998; 153(3): 865–73. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nValastyan S, Weinberg RA: Tumor metastasis: molecular insights and evolving paradigms. Cell. 2011; 147(2): 275–92. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKlein CA: Parallel progression of primary tumours and metastases. Nat Rev Cancer. 2009; 9(4): 302–12. PubMed Abstract \| Publisher Full Text\n\nGancberg D, Di Leo A, Cardoso F, et al.: Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites. Ann Oncol. 2002; 13(7): 1036–43. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nAlmendro V, Kim HJ, Cheng YK, et al.: Genetic and phenotypic diversity in breast tumor metastases. Cancer Res. 2014; 74(5): 1338–48. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nLin NU, Amiri-Kordestani L, Palmieri D, et al.: CNS metastases in breast cancer: old challenge, new frontiers. Clin Cancer Res. 2013; 19(23): 6404–18. PubMed Abstract \| Publisher Full Text\n\nSingh AK, Loscalzo J; Brigham and Women's Hospital: The Brigham intensive review of internal medicine. New York: Oxford University Press; 2012. Reference Source\n\nFriedl P, Alexander S: Cancer invasion and the microenvironment: plasticity and reciprocity. Cell. 2011; 147(5): 992–1009. PubMed Abstract \| Publisher Full Text\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): 171–8. PubMed Abstract \| Publisher Full Text \| F1000 Recommendation\n\nBiswas S, Guix M, Rinehart C, et al.: Inhibition of TGF-beta with neutralizing antibodies prevents radiation-induced acceleration of metastatic cancer progression. J Clin Invest. 2007; 117(5): 1305–13. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nAlsarraj J, Faraji F, Geiger TR, et al.: BRD4 short isoform interacts with RRP1B, SIPA1 and components of the LINC complex at the inner face of the nuclear membrane. PloS One. 2013; 8(11): e80746. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nPark YG, Zhao X, Lesueur F, et al.: Sipa1 is a candidate for underlying the metastasis efficiency modifier locus Mtes1. Nat Genet. 2005; 37(10): 1055–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrawford NP, Alsarraj J, Lukes L, et al.: Bromodomain 4 activation predicts breast cancer survival. Proc Natl Acad Sci U S A. 2008; 105(17): 6380–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrawford NP, Qian X, Ziogas A, et al.: Rrp1b, a new candidate susceptibility gene for breast cancer progression and metastasis. PLoS Genet. 2007; 3(11): e214. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrawford NP, Walker RC, Lukes L, et al.: The Diasporin Pathway: a tumor progression-related transcriptional network that predicts breast cancer survival. Clin Exp Metastasis. 2008; 25(4): 357–69. PubMed Abstract \| Publisher Full Text \| Free Full Text \| F1000 Recommendation\n\nFaraji F, Hu Y, Wu G, et al.: An integrated systems genetics screen reveals the transcriptional structure of inherited predisposition to metastatic disease. Genome Res. 2014; 24(2): 227–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWinter SF, Lukes L, Walker RC, et al.: Allelic variation and differential expression of the mSIN3A histone deacetylase complex gene Arid4b promote mammary tumor growth and metastasis. PLoS Genet. 2012; 8(5): e1002735. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaraji F, Pang Y, Walker RC, et al.: Cadm1 is a metastasis susceptibility gene that suppresses metastasis by modifying tumor interaction with the cell-mediated immunity. PLoS Genet. 2012; 8(9): e1002926. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8896", "date": "04 Jun 2015", "name": "Patricia Steeg", "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", "responses": [] }, { "id": "8897", "date": "04 Jun 2015", "name": "Ann Chambers", "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 thoughtful and well-written short review of a clinically very important topic. This review presents a synthesis of approaches to deal clinically with metastatic breast cancer, and will stimulate thought in the field.", "responses": [] }, { "id": "8898", "date": "04 Jun 2015", "name": "Suzanne A Eccles", "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\nDr Hunter and colleagues have written an excellent, very timely and thought-provoking paper outlining the key issues in tackling metastatic breast cancer. It is almost unique in its ability to appear years or even decades after successful treatment of the primary tumor. They accurately pinpoint our failures so far to control metastatic disease and suggests that a ‘3-pronged attack’ will be required to improve survival: prevention of further spread, eradication of (or maintenance of dormancy in) disseminated cells and, most important of all, destruction of established micrometastases.The latter is certainly the main challenge: most primary breast cancers are successfully and completely removed by surgery, it is the already-disseminated tumor cells (DTC) that are the potential seeds of destruction of a human life. Much attention has been given to the early stages of metastasis: release of cells, motility, invasion, adhesion to endothelia, extravasation – perhaps because these steps are the easiest to manipulate experimentally. The onus is now on scientists to address whether any of these processes and their molecular mediators are still required once DTC are established at secondary sites. We need to know if they represent the ‘starter motor’ (essential initially but redundant once the process is underway) or the ‘engine’ (powering the continued progress) of metastases. Even if we can prevent further spread, it is important to define whether existing micrometastases (probably generated very early in the course of the disease) would prove fatal before any later waves of circulating cells take hold – possibly a case of shutting the stable door after the horse has bolted.Metastases break one of the fundamental ‘taboos’ of multicellular organisms – that of cells being able to survive and thrive in quite disparate, alien tissue environments – they no longer ‘know their place’. If we could define how this is achieved we may discover some common denominator actionable molecular targets for intervention that could be effective against widespread metastatic disease. Such studies, with the precision tools of genetic manipulation could perhaps be achieved in lower organisms as a starting point.So, key additional questions raised by this interesting piece are: which if any of the molecular mechanisms required to initiate and maintain primary tumors are also essential for establishing metastases and for how long do any such dependencies persist? Or are the determinants of successful secondary colonisation quite distinct? Do different organs have similar or unique requirements that enable cancer cells to form overt metastases? The paper touches on these considerations, but much more research is required in this challenging area. It is therefore essential, as Hunter rightly states, to facilitate better access to clinical material, especially relapsed and drug resistant secondary disease (which may only be practical at autopsy).Dr Hunter’s paper will certainly recalibrate the way we think about translational research in metastasis. It should inspire new generations of basic scientists and clinicians to work together to tackle more logically and effectively this last hurdle in our efforts to improve cure rates in breast and other cancers.", "responses": [] }, { "id": "8899", "date": "04 Jun 2015", "name": "Danny Welch", "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 perspective that deals with the challenges and opportunities of treating/preventing breast cancer metastasis. The manuscript is fair, balanced and covers the majority of issues relevant to the topic.", "responses": [] }, { "id": "8900", "date": "04 Jun 2015", "name": "Robin L. Anderson", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-141
https://f1000research.com/articles/3-242/v1	13 Oct 14	{ "type": "Research Note", "title": "Immunohistochemical visualization of mouse interneuron subtypes", "authors": [ "Simon Molgaard", "Maj Ulrichsen", "Simon Boggild", "Marie-Louise Holm", "Christian Vaegter", "Jens Nyengaard", "Simon Glerup", "Simon Molgaard", "Maj Ulrichsen", "Simon Boggild", "Marie-Louise Holm", "Christian Vaegter", "Jens Nyengaard" ], "abstract": "The activity of excitatory neurons is controlled by a small, but highly diverse population of inhibitory interneurons. These cells show a high level of physiological, morphological and neurochemical heterogeneity, and play highly specific roles in neuronal circuits. In the mammalian hippocampus, these are divided into 21 different subtypes of GABAergic interneurons based on their expression of different markers, morphology and their electrophysiological properties. Ideally, all can be marked using an antibody directed against the inhibitory neurotransmitter GABA, but parvalbumin, calbindin, somatostatin, and calretinin are also commonly used as markers to narrow down the specific interneuron subtype. Here, we describe a journey to find the necessary immunological reagents for studying GABAergic interneurons of the mouse hippocampus. Based on web searches there are several hundreds of different antibodies on the market directed against these four markers. Searches in the literature databases allowed us to narrow it down to a subset of antibodies most commonly used in publications. However, in our hands the most cited ones did not work for immunofluorescence stainings of formaldehyde fixed tissue sections and cultured hippocampal neurons, and we had to immunostain our way through thirteen different commercial antibodies before finally finding a suitable antibody for each of the four markers. The antibodies were evaluated based on signal-to-noise ratios as well as if positive cells were found in layers of the hippocampus where they have previously been described. Additionally, the antibodies were also tested on sections from mouse spinal cord with similar criteria for specificity of the antibodies. Using the antibodies with a high rating on pAbmAbs, stainings with high signal-to-noise ratios and location of the immunostained cells in accordance with the literature could be obtained, making these antibodies suitable choices for studying the GABAergic system.", "keywords": [ "antibody validation", "immunohistochemistry", "hippocampus", "spinal cord" ], "content": "Introduction\n\nHippocampal networks are composed of a large portion of excitatory principal cells and a smaller cohort of inhibitory interneurons1. Inhibitory interneurons release γ-aminobutyric acid (GABA), which is the major inhibitory neurotransmitter in the brain. Its principal action is mediated through ubiquitous fast ionotropic GABAA receptors by increasing the membrane permeability to Cl- ions2. This inhibitory mechanism regulates the excitability of both principal cells and GABAergic interneurons. In this way, GABA is able to efficiently control the rhythms of cortical networks3, which is believed to be of critical importance for information processing4. Breakdown of these brain rhythms in neuropsychiatric disorders, such as schizophrenia, depression and bipolar disorder, is thought to involve a defective GABA system5.\n\nInhibitory interneurons of the dentate gyrus is a highly diverse population and early studies identified up to 21 different subtypes in this region alone6. Immunostaining against GABA have shown discrepancy when compared to in-situ hybridization against glutamate decarboxylase, indicating that some cells may express very low levels of GABA leaving this as an insufficient choice for immunostaining7–9. These 21 subtypes can be distinguished based on axonal distribution, synaptic targets, neuropeptide or calcium-binding protein content and physiological characteristics10. In order to fully characterize a subtype, all parameters must be taken into account. When immunostaining against neuropeptides or calcium-binding proteins, this is not possible, and immunostaining therefore only allows characterization of subgroups.\n\nOne such subgroup is the parvalbumin expressing interneurons. Parvalbumin-labelled cell bodies are found primarily near the granule cell layer and are most prominent at the base of the granule cell layer. However, few are also found near the border of the granule cell and molecular layers and some in the hilus as well10. Although this is considered the largest group of the subgroups in the hippocampus, in the dentate gyrus these only represent around 20% of the total number of GABAergic interneurons as compared to around 40% in CA1 and CA311.\n\nSeveral distinct populations are found that express the calcium-binding protein calretinin. Most notably, calretinin is also found in mossy cells of the hilus12, and such mossy cells are particular numerous in the ventral hilus. Calretinin is also found in axon terminals of mossy cells which creates a dense band of labelling in the inner third of the molecular layer13.\n\nDespite labelling of mossy cells in the hilus, some GABAergic interneurons can also be found in the hilus near the granule layer14. These can often be distinguished by the more intense labelling when staining for calretinin compared to that of mossy cells.\n\nAnother subgroup is the somatostatin expressing interneurons. This subgroup comprises the largest group of GABAergic interneurons in the dentate gyrus and these are almost exclusively found within the hilus where they comprise approximately 55% of the total number of GABAergic interneurons with a slight increase from the dorsal to the ventral part of hippocampus15. As almost all somatostatin positive interneurons are found within the hilus, little labelling is found within the granule cell layer, except from a large number of axons from hilar somatostatin interneurons that project through this layer15,16.\n\nCalbindin has been found to be present in both inhibitory and excitatory neurons with a rather strong staining of granule cells in the dentate gyrus. Misplaced granule cells found in the stratum radiatum of the CA3 subfield are often mistaken for GABAergic interneurons but these are not positive for GABA1. All other cells in the dentate gyrus should be considered GABAergic interneurons and generally stain for GABA1. A precise percentage of calbindin interneurons is not available, but around 10–12% of total number of GABAergic interneurons is considered a close estimate17. Very few calbindin positive interneurons are found in the dentate gyrus compared to the CA-regions and these are difficult to detect due to the strong staining of granule cells, but calbindin positive interneurons can be found in the stratum moleculare and hilus1.\n\nImportantly, markers of hippocampal GABAergic interneurons do not readily apply to other regions such as the spinal cord GABAergic interneurons. The inhibitory interneurons of the spinal dorsal horn use primarily GABA and/or glycine. GABAergic interneurons are primarily located in laminae I, II and III of the dorsal horn and constitute approximately 25%, 30% and 40% of rat laminae I, II and III neurons, respectively18,19. The inhibitory effect of glycine is facilitated by activation of ionotropic ligand-gated glycine receptors that mediate an influx of chloride ions20 and within lamina I-III glycine immunostaining is largely restricted to GABAergic neurons18,19.\n\nGABAergic interneurons of the spinal dorsal horn can be identified by immunostaining against, for instance, parvalbumin and the neuronal form of nitric oxide synthase (n-NOS) besides GABA and glycine. Parvalbumin is expressed by a subpopulation of spinal cord dorsal horn interneurons that co-express GABA and glycine21–23. Conversely, calretinin, somatostatin and calbindin do not co-localize with GABA in interneurons of the dorsal horn, for which reason they are thought to co-localize to excitatory interneurons21,23–25. Thus, care should be taken when extrapolating interneuron markers from one region of the CNS to another. In the present study, we have evaluated a number of different antibodies (Table 2) against GABAergic markers using both cultured neurons and tissue sections. All tested antibodies have previously been reported to recognize GABAergic interneurons both in peer-reviewed publications and by the manufacturers.\n\n\nMaterials and methods\n\nAll experiments were approved by the Danish Animal Experiments Inspectorate under the Ministry of Justice (Permit 2011/561-119) and carried out according to institutional and national guidelines.\n\nFor a full list of reagents and chemicals, please see Table 1.\n\nHippocampal sections. Adult C57BL/6j Bomtac (wild type (wt)) mice (Taconic), aged 8 weeks were deeply anesthetized by intraperitoneal injection of 5 mg/ml pentobarbital and perfused transcardially with cold 4% (w/v) formaldehyde (pH 7,4, Hounisen) for five minutes. The brains were hereafter removed and post-fixed in 4% (w/v) formaldehyde overnight at 4°C. The next day the brains were moved to 30% (w/v) sucrose (Merck Millipore) for cryoprotection and left at 4°C for 48 hours, moulded in Tissue-Tek® (Sakura) and stored at -20°C. Coronal hippocampal slices (10 µm) were cut at -20°C using a Leica CM1900 cryostat (using low-profile disposable blades 819 from Leica Biosystems) and the sections were afterwards stored at -20°C until use.\n\nImmunostaining of tissue. Antigen epitopes shielded by formaldehyde cross-linked lysine side chains were retrieved in a heat-mediated antigen retrieval step using Target Retrieval Solution (Dako), according to manufacturers’ protocol. Hereafter, the slices were washed three times in Tris-buffered saline (TBS; pH 7.4) of ten minutes intervals, and incubated in a solution of TBS containing 0.3% Triton X-100 (Applichem) and 1% bovine serum albumin (BSA; Sigma) for thirty minutes. Following a ten minute washing step in TBS, the slices were incubated with primary antibody (Table 2) in a 50 mM Tris-based (TB) buffer solution (pH 7.4) containing 1% BSA (Sigma) at 4°C in a moisturized chamber overnight. The next day, the slices were left at room temperature (RT) for one hour, and subsequently washed three times in TBS. Slices were then incubated with secondary antibody (Table 3) in a 50 mM TB buffer solution containing 1% BSA (Sigma) at RT for four hours. Finally, the coverslips were washed three times five minutes in TBS, with Hoechst (5µg/µl, Sigma-Aldrich) being included in the last wash. The slices were hereafter mounted using Fluorescence Mounting Medium (Dako) and stored at 4°C. As negative controls of the immunostaining, simultaneous stainings were done using a similar protocol, except primary antibody was omitted.\n\nSpinal cord sections. Adult C57BL/6j Bomtac (wt) mice aged 16 weeks were deeply anaesthetized using 4% isoflurane (IsoFlo® vet, Abbott) prior to decapitation and hydraulic spinal cord extrusion26 using ice-cold phosphate-buffered saline (PBS; pH 7.4) as the extrusion liquid. Spinal cords were fixed in 4% (w/v) paraformaldehyde (PFA; Sigma) in PBS (pH 7.4) overnight at 4°C. The spinal cords were then cryoprotected overnight by immersion in 25% (w/v) sucrose in PBS (pH 7.4) at 4°C. Lumbar sections 2–4 of the spinal cords were isolated and embedded in TissueTek® (Sakura) prior to freezing, which was performed by lowering the tissue into dry-ice cold iso-pentane (VWR BDH Prolabo®). The tissues were stored at -80°C until further use. Transverse sections of 20 μm thickness were cut at -20°C using the CryoJane® Tape-Transfer System (Leica Microsystems) on a Leica CM1900 cryostat (using low-profile disposable blades 819 from Leica Biosystems) and the sections were stored at -20°C.\n\nImmunostaining of tissue. This step was done similar to previously described for immunostaining of hippocampal tissue.\n\nCulture of primary hippocampal neurons. Postnatal day 0 (P0) C57BL/6j Bomtac (wt) mice pups were sacrificed by decapitation, brains removed and hippocampi dissected into ice cold PBS. The tissue was dissociated for thirty minutes in 20 U/mL activated papain (Worthington Biochemical Corporation). After dissociation, the tissue was washed once in DMEM (Lonza) containing 0.01 mg/mL DNaseI (Sigma) before being triturated in DMEM (Lonza) containing 0.01 mg/mL DNaseI (Sigma). After this, Neurobasal-A medium (Gibco) containing B-27 Supplement (Gibco), 2 mM GlutaMAX (Gibco), 100 μg/mL Primocin (Invivogen) and 20 μM floxuridine + 20 μM uridine (Sigma) was added to the cells and the cells were seeded on poly-D-lysine (Sigma-Aldrich) and laminin (Invitrogen) pre-coated coverslips at a density of 100.000 cells per coverslip and left for fourteen days at 37°C and 5% CO2, with medium change every second day, before being fixed in PBS containing 4% PFA.\n\nImmunostaining of cultured hippocampal neurons. Neurons fixed in 4% PFA was briefly washed in PBS prior to three consecutive washes in PBS containing 0.1% Triton X-100 of ten minute intervals. Hereafter, the cells were washed once in PBS before being incubated in PBS containing 10% FBS (Gibco) for thirty minutes at RT. After this, the cells were incubated with primary antibody (Table 2) overnight at 4°C. The next day, the immunostaining were left at RT for one hour before continuing the immunostaining protocol. Hereafter, the cells were washed three times five minutes in PBS containing 0.1% Triton-X 100. Subsequently, the cells were incubated with secondary antibodies (Table 3) for four hours at RT. The coverslips were then washed two times five minutes in PBS followed by a five minute wash in PBS containing Hoechst (5µg/µl, Sigma-Aldrich) before being mounted using Fluorescence Mounting Medium (Dako) and stored at 4°C. As negative controls of the immunostaining, simultaneous stainings were done using a similar protocol, except primary antibody was omitted.\n\nConfocal microscopy. The samples were analysed on a Zeiss confocal LSM 780 microscope (Carl Zeiss) using 20X/0.8 M27 and 63X/1.20 W Korr (Water immersion correction ring) objectives. Appropriate filters were used upon excitation of the different fluorophores to match their maximum fluorescence emission. The channels used were H258 and A568 and they were configured to obtain the best signal during image acquisition of the samples in order to prevent bleed through between the different probes. The range indicator was used to adjust gain and offset so acquired images were optimally held within the dynamic range of the detector. Frame size was selected to be “optimal” and an averaging of 16 was selected upon image acquisition in order to acquire an appropriate number of pixels and to achieve a maximum of signal-to-noise-ratio, respectively. Image acquisition was performed with foci adjusted with respect to the 568 nm fluorophores, as they were used to visualize the markers of interneurons; parvalbumin, calretinin, calbindin and somatostatin. Processing of the acquired images were performed in Zen 2011 (Carl Zeiss) Image Processing. All images presented were subjected to similar brightness and contrast adjustments.\n\nThe use of each chemical can be found in the materials and methods section. The products are listed in alphabetic order.\n\nThe pAbmAbs rating reflects the average rating of the antibodies as of October 2014.\n\n\nResults and discussion\n\nInitially, we screened the antibody specificity by staining of cultured hippocampal neurons, evaluating antibodies based on their ability to mark a distinct subset of neurons. Hereafter, when staining hippocampal slices, the antibodies were rated based on the expected localization and abundance of interneurons positive for the specific staining.\n\nThe localization of parvalbumin interneurons within the dentate gyrus is very well described so cells staining positive in layers where parvalbumin interneurons are not expected were considered as unspecific immunostaining. For several of the immunostainings, very little, if any, signal was obtained. However, the anti-parvalbumin ab11427 antibody from Abcam gave a clear and intense staining of parvalbumin interneurons both in culture and in hippocampal tissue sections (Figure 1 and Table 2). As the positive neurons were found in layers of the dentate gyrus, where parvalbumin positive interneurons have previously been described to be located, at an expected frequency, this was considered a specific staining and was therefore rated with 5 out of 5 stars on pAbmAbs (www.pAbmAbs.com).\n\nA) shows specific and unspecific staining against parvalbumin in cultured hippocampal neurons. B) is a staining of hippocampal tissue, also showing a specific and an unspecific staining. Scale bar represents 20 µm.\n\nUnlike parvalbumin, calretinin is found not only in interneurons but also in mossy cells within the dentate gyrus. These can often be distinguished based on the intensity of the labelling. When rating these antibodies, the correct localization of positive neurons was therefore considered not only in relation to interneurons but also to mossy cells. Both antibodies from Millipore showed high specificity against calretinin, and especially the anti-calretinin ab5054 antibody gave a very specific staining with a high signal-to-noise ratio and was therefore given 5 out of 5 stars on pAbmAbs (Figure 2 and Table 2).\n\nA) shows specific and unspecific staining against calretinin in cultured hippocampal neurons. B) is a staining of hippocampal tissue, also showing a specific and an unspecific staining. Scale bar represents 20 µm.\n\nSimilarly, antibodies against somatostatin were evaluated based on signal-to-noise and localization of neurons positive for somatostatin. In most cases, staining against somatostatin gave a high background with very low signal. However, using the anti-somatostatin mab364 antibody from Millipore we observed a clear staining with a good signal-to-noise ratio (Figure 3 and Table 2) and therefore it received a rating of 5 out of 5 stars. The neurons positive for somatostatin were, as expected, found in the hilus of the dentate gyrus.\n\nA) shows specific and unspecific staining against calretinin in cultured hippocampal neurons. B) is a staining of hippocampal tissue, also showing a specific and an unspecific staining. Scale bar represents 20 µm.\n\nLike calretinin, calbindin is also expressed by non-inhibitory cells. When looking at the dentate gyrus, expression of calbindin by principal cells within the granule cell layer gives a weak immunostaining which might seem like unspecific binding, however that is not the case. Interneurons positive for calbindin can be recognized based on the location as well as increased intensity of the immunostaining. Due to the very low number of calbindin-interneurons in the hilus, this immunostaining can be hard to detect. Many of the antibodies we tested showed very little if any difference in staining intensity between interneurons and granule cells. However, using the anti-calbindin ab1778 antibody from Millipore we were able to distinguish between interneurons and granule cells (Figure 4 and Table 2). Since this antibody also shows very little background staining it was rated 5 stars on pAbmAbs.\n\nA) shows specific and unspecific staining against calbindin in cultured hippocampal neurons. B) is a staining of hippocampal tissue, also showing a specific and an unspecific staining. Scale bar represents 20 µm.\n\nParvalbumin positive cells of the spinal cord dorsal horn also represent a subgroup of GABAergic interneurons and immunostaining against parvalbumin can accordingly be used as a marker of GABAergic interneurons. When staining against parvalbumin with the anti-parvalbumin ab11427 antibody from Abcam they appeared to be largely restricted to laminae II-III of the dorsal horn, which is in accordance with previous findings27. The parvalbumin positive cells of laminae II-III were rather small and showed intense immunoreactivity in the nucleus and in the soma, as previously described22, making it easy to distinguish them from background staining. This antibody also appeared to stain neuronal processes of the dorsal horn and columns as well as the nuclei of ventral horn motor neurons, as previously described27–29. Although this antibody can be used to identify intense immunoreactive parvalbumin positive cells and function as a great marker of the parvalbumin positive subpopulation of GABAergic neurons of the spinal dorsal horn in locations previously described, it showed some background staining of spinal cord cryo-sections and was rated 4 out of 5 stars on pAbmAbs.\n\nUnlike interneurons of the hippocampus, calretinin can only be used as a marker of interneurons that do not contain GABA in the spinal cord24. The anti-calretinin AB5054 antibody from Merck Millipore works well for IHC of spinal cord cryo-sections (data not shown) and was rated 5 out of 5 stars on pAbmAbs, as it showed very low background staining and intense staining of a dense well-defined band of small calretinin immunoreactive cells in the superficial laminae of the dorsal horn and of large cells in lamina V-VI. These observations correlates with previous description of calretinin immunoreactivity in the spinal cord24, and indicates high specificity of the antibody.\n\nIn contrast to IHC of hippocampal sections with the anti-somatostatin MAB354 antibody from Millipore, this antibody gave a low signal when staining against somatostatin on spinal cord sections. Using this antibody, it was difficult to identify somatostatin positive cells in the spinal dorsal horn that otherwise previously have been described to be located in a dense band in lamina II of rat25 and mouse21 spinal dorsal horn. Therefore, the antibody was rated 2 out of 5 stars on pAbmAbs. This antibody was rated 5 out of 5 for the hippocampal staining, leading to an average rating of 3.5 on pAbmAbs.\n\nLike calretinin and somatostatin, calbindin can be used as a marker of spinal dorsal horn interneurons that do not contain GABA23. A dense band of calbindin immunoreactivity has previously been shown in lamina II and a more sparse band in lamina I, III and IV of the rat spinal dorsal horn23. This localization of calbindin immunoreactivity is also seen when using the anti-calbindin AB1778 antibody from Merck Millipore (data not shown). Also, the cells that constitute the central channel and motor neurons of the ventral horn also show calbindin immunoreactivity when staining with this antibody, which is in accordance with previously findings28,30. The antibody showed very intense staining of cytoplasm and nuclei, as well as processes of the outer lamina of the dorsal horn and showed low background staining. On the basis of these observations the antibody was rated 5 out of 5 stars on pAbmAbs.\n\n\nConclusion\n\nIn conclusion, staining against interneurons can be a very tedious task and great consideration is needed to ensure that it is actually only interneurons that are being stained. Optimizing protocols for immunostaining can be a, not only time consuming, but also an expensive task in a market full of different antibody options. By creating an information-sharing platform, pAbmAbs allows for a fast and cost-free screening of the current antibodies out there and thereby ensures that only the best antibodies are used. In the current study, we tested antibodies against parvalbumin, calretinin, calbindin and somatostatin, all markers of hippocampal GABAergic interneurons, both in culture and on hippocampal and spinal cord tissue. These antibodies were rated on specificity, and signal-to-noise ratio, for both tissue and culture. When immunostaining tissue, we also looked at the localization of positive cells within the tissue to ensure that only cells in the expected layers of the tissue stained positive for the GABAergic markers. When staining against parvalbumin we found that out of four different antibodies, the anti-parvalbumin ab11427 antibody from Abcam got a high score as it stained cells specifically with a high signal-to-noise ratio with the expected localization within the tissue. When staining against calretinin, the anti-calretinin ab5054 antibody from Millipore obtained the highest score on pAbmAbs. This antibody gave a very nice signal-to-noise ratio compared to the other antibodies tested. The anti-somatostatin mab354 antibody from Millipore was found to be the best antibody for stainings against somatostatin. Similar to the other antibodies with high pAbmAbs ratings, this also had a high signal-to-noise ratio compared to other antibodies tested. Finally, the anti-calbindin ab1178 antibody from Millipore got the highest rating out of the antibodies tested against this GABAergic subgroup. Overall, the antibody tested gave varying results when using our protocols. The specificities of the antibodies are therefore reflected on pAbmAbs which, by serving as a database, will help fast and cost-free evaluation of antibodies.\n\n\nData availability\n\nF1000Research: Dataset 1. Interneurons of hippocampus and spinal cord, 10.5256/f1000research.5349.d3668231", "appendix": "Author contributions\n\n\n\nSM and SG conceived this study. SM designed the experiments. SM, MU and SBH carried out the research. SG, JRN and CV contributed to the design of the experiment and expertise in immunohistochemistry. SM, SG and MU contributed to the preparation of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was funded by the Lundbeck Foundation, Danish Medical Research Council, Fonden til forskning af sindslidelser and Agnes og Poul Friis Fond.\n\n\nAcknowledgements\n\nWe thank Helene Andersen, Anja Aagaard and Benedicte Vestergaard for excellent technical assistance.\n\n\nReferences\n\nFreund TF, Buzsaki G: Interneurons of the hippocampus. Hippocampus. 1996; 6(4): 347–470. PubMed Abstract \| Publisher Full Text\n\nFarrant M, Nusser Z: Variations on an inhibitory theme: phasic and tonic activation of GABA(A) receptors. Nat Rev Neurosci. 2005; 6(3): 215–29. PubMed Abstract \| Publisher Full Text\n\nWang XJ, Buzsaki G: Gamma oscillation by synaptic inhibition in a hippocampal interneuronal network model. J Neurosci. 1996; 16(20): 6402–13. PubMed Abstract\n\nBuzsaki G, Draguhn A: Neuronal oscillations in cortical networks. Science. 2004; 304(5679): 1926–1929. PubMed Abstract \| Publisher Full Text\n\nBenes FM, Berretta S: GABAergic interneurons: implications for understanding schizophrenia and bipolar disorder. Neuropsychopharmacology. 2001; 25(1): 1–27. PubMed Abstract \| Publisher Full Text\n\nAmaral DG: A Golgi study of cell types in the hilar region of the hippocampus in the rat. J Comp Neurol. 1978; 182(4 Pt 2): 851–914. PubMed Abstract \| Publisher Full Text\n\nErlander MG, Tillakaratne NJ, Feldblum S, et al.: Two genes encode distinct glutamate decarboxylases. Neuron. 1991; 7(1): 91–100. PubMed Abstract \| Publisher Full Text\n\nRibak CE, Vaughn JE, Saito K: Immunocytochemical localization of glutamic acid decarboxylase in neuronal somata following colchicine inhibition of axonal transport. Brain Res. 1978; 140(2): 315–32. PubMed Abstract \| Publisher Full Text\n\nJinno S, Aika Y, Fukuda T, et al.: Quantitative analysis of GABAergic neurons in the mouse hippocampus, with optical disector using confocal laser scanning microscope. Brain Res. 1998; 814(1–2): 55–70. PubMed Abstract \| Publisher Full Text\n\nSomogyi P, Klausberger T: Defined types of cortical interneurone structure space and spike timing in the hippocampus. J Physiol. 2005; 562(Pt 1): 9–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRibak CE, Nitsch R, Seress L: Proportion of parvalbumin-positive basket cells in the GABAergic innervation of pyramidal and granule cells of the rat hippocampal formation. J Comp Neurol. 1990; 300(4): 449–61. PubMed Abstract \| Publisher Full Text\n\nBlasco-Ibanez JM, Freund TF: Distribution, ultrastructure, and connectivity of calretinin-immunoreactive mossy cells of the mouse dentate gyrus. Hippocampus. 1997; 7(3): 307–20. PubMed Abstract \| Publisher Full Text\n\nFujise N, Liu Y, Hori N, et al.: Distribution of calretinin immunoreactivity in the mouse dentate gyrus: II. Mossy cells, with special reference to their dorsoventral difference in calretinin immunoreactivity. Neuroscience. 1998; 82(1): 181–200. PubMed Abstract \| Publisher Full Text\n\nGulyas AI, Hajos N, Freund TF: Interneurons containing calretinin are specialized to control other interneurons in the rat hippocampus. J Neurosci. 1996; 16(10): 3397–411. PubMed Abstract\n\nJinno S, Kosaka T: Patterns of expression of neuropeptides in GABAergic nonprincipal neurons in the mouse hippocampus: Quantitative analysis with optical disector. J Comp Neurol. 2003; 461(3): 333–49. PubMed Abstract \| Publisher Full Text\n\nKatona I, Acsady L, Freund TF: Postsynaptic targets of somatostatin-immunoreactive interneurons in the rat hippocampus. Neuroscience. 1999; 88(1): 37–55. PubMed Abstract \| Publisher Full Text\n\nGulyas AI, Toth K, Danos P, et al.: Subpopulations of GABAergic neurons containing parvalbumin, calbindin D28k, and cholecystokinin in the rat hippocampus. J Comp Neurol. 1991; 312(3): 371–8. PubMed Abstract \| Publisher Full Text\n\nTodd AJ, Sullivan AC: Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivities in the spinal cord of the rat. J Comp Neurol. 1990; 296(3): 496–505. PubMed Abstract \| Publisher Full Text\n\nPolgar E, Hughes DI, Riddell JS, et al.: Selective loss of spinal GABAergic or glycinergic neurons is not necessary for development of thermal hyperalgesia in the chronic constriction injury model of neuropathic pain. Pain. 2003; 104(1–2): 229–39. PubMed Abstract \| Publisher Full Text\n\nZeilhofer HU: The glycinergic control of spinal pain processing. Cell Mol Life Sci. 2005; 62(18): 2027–35. PubMed Abstract \| Publisher Full Text\n\nHeinke B, Ruscheweyh R, Forsthuber L, et al.: Physiological, neurochemical and morphological properties of a subgroup of GABAergic spinal lamina II neurones identified by expression of green fluorescent protein in mice. J Physiol. 2004; 560(Pt 1): 249–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaing I, Todd AJ, Heizmann CW, et al.: Subpopulations of GABAergic neurons in laminae I–III of rat spinal dorsal horn defined by coexistence with classical transmitters, peptides, nitric oxide synthase or parvalbumin. Neuroscience. 1994; 61(1): 123–32. PubMed Abstract \| Publisher Full Text\n\nAntal M, Polgár E, Chalmers J, et al.: Different populations of parvalbumin- and calbindin-D28k-immunoreactive neurons contain GABA and accumulate 3H-D-aspartate in the dorsal horn of the rat spinal cord. J Comp Neurol. 1991; 314(1): 114–24. PubMed Abstract \| Publisher Full Text\n\nRen K, Ruda MA: A comparative study of the calcium-binding proteins calbindin-D28K, calretinin, calmodulin and parvalbumin in the rat spinal cord. Brain Res Brain Res Rev. 1994; 19(2): 163–79. PubMed Abstract \| Publisher Full Text\n\nProudlock F, Spike RC, Todd AJ: Immunocytochemical study of somatostatin, neurotensin, GABA, and glycine in rat spinal dorsal horn. J Comp Neurol. 1993; 327(2): 289–97. PubMed Abstract \| Publisher Full Text\n\nMeikle AD, Martin AH: A rapid method for removal of the spinal cord. Stain Technol. 1981; 56(4): 235–7. PubMed Abstract \| Publisher Full Text\n\nHughes DI, Sikander S, Kinnon CM, et al.: Morphological, neurochemical and electrophysiological features of parvalbumin-expressing cells: a likely source of axo-axonic inputs in the mouse spinal dorsal horn. J Physiol. 2012; 590(Pt 16): 3927–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInce P, Stout N, Shaw P, et al.: Parvalbumin and calbindin D-28k in the human motor system and in motor neuron disease. Neuropathol Appl Neurobiol. 1993; 19(4): 291–9. PubMed Abstract \| Publisher Full Text\n\nAntal M, Freund TF, Polgar E: Calcium-binding proteins, parvalbumin- and calbindin-D 28k-immunoreactive neurons in the rat spinal cord and dorsal root ganglia: a light and electron microscopic study. J Comp Neurol. 1990; 295(3): 467–84. PubMed Abstract \| Publisher Full Text\n\nHolm MM, Nieto-Gonzalez JL, Vardya I, et al.: Mature BDNF, but not proBDNF, reduces excitability of fast-spiking interneurons in mouse dentate gyrus. J Neurosci. 2009; 29(40): 12412–8. PubMed Abstract \| Publisher Full Text\n\nMolgaard S, Ulrichsen M, Boggild S, et al.: Dataset 1. Interneurons of hippocampus and spinal cord. F1000Research. 2014. Data Source" }	[ { "id": "6401", "date": "14 Oct 2014", "name": "Sally Lowell", "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 authors of this paper set out to identify antibodies that can be used to identify particular subtypes of GABAergic neurons within the mouse hippocampus. They explain that many of the antibodies that they have tested for this purpose do not seem to work, and they present data for those antibodies that they did find to be useful for this purpose. This article could therefore save readers a lot of wasted effort and money in identifying useful antibodies for their own studies in this area.The article starts with a nice overview of the different subtypes of GABAergic neurons and the markers that are commonly used to characterise them. I am not an expert in this area so I cannot review the accuracy of the information, but I did find it to be concise and useful introduction. There are a few issues that could be addressed to improve the results section of the article.Figure legends should explain what \"unspecific staining\" refers to. Does this mean secondary only control? If so then this form of labelling could be misleading as non-specific binding of the primary antibody will not be picked up by a secondary-antibody-only control. Why are there no Hoechst positive cells in some of the 'unspecific staining' panels? The evidence that these antibodies are specific to particular subtypes of neurons comes mainly from the observation that their staining pattern is restricted to the expected regions of the hippocampus. However the figures as presented do not make this clear. Is it possible to provide a clearer demonstration of these restricted expression domains on hippocampus sections? Perhaps it would be helpful to see an accompanying diagram showing the expected distribution of each marker on these sections? In Fig 3B, the 'specific staining' of antibody mAB354 is barely visible so it is difficult to assess whether there is a real difference between signal and background. The article reports results from immunostaining of the spinal cord, but no figures are presented to support the authors’ findings. It would be helpful to see these images.General comments: The article is written clearly and concisely, but would benefit from being proof-read for minor grammatical errors.", "responses": [ { "c_id": "1075", "date": "12 Nov 2014", "name": "Simon Molgaard", "role": "Author Response", "response": "\"Figure legends should explain what \"unspecific staining\" refers to. Does this mean secondary only control? If so then this form of labelling could be misleading as non-specific binding of the primary antibody will not be picked up by a secondary-antibody-only control.\" Reply: The figure legends have been changed to make this more clear. \"Why are there no Hoechst positive cells in some of the 'unspecific staining' panels?\" Reply: We apologize for the lack of clarity regarding this issue. The “unspecific” images with Hoechst are shown when we observed no signal at all. The “unspecific” images without Hoechst are shown when staining was present but distributed in an unspecific manner. \"The evidence that these antibodies are specific to particular subtypes of neurons comes mainly from the observation that their staining pattern is restricted to the expected regions of the hippocampus. However the figures as presented do not make this clear. Is it possible to provide a clearer demonstration of these restricted expression domains on hippocampus sections? Perhaps it would be helpful to see an accompanying diagram showing the expected distribution of each marker on these sections?\" Reply: The interneuron subtype distribution is presently used in a qualitative manner to validate whether positive staining localizes in the expected and with an expected frequency in the dentate gyrus of the hippocampus. Quantitative assessment of interneuron number using for example stereology is not feasible in the present study. \"In Fig 3B, the 'specific staining' of antibody mAB354 is barely visible so it is difficult to assess whether there is a real difference between signal and background.\" Reply: We agree that the staining is weaker; however it is still distinct and specific. \"The article reports results from immunostaining of the spinal cord, but no figures are presented to support the authors’ findings. It would be helpful to see these images.\" Reply: The images are already included in the data availability section." } ] }, { "id": "6403", "date": "17 Oct 2014", "name": "Tomi PJ Rantamäki", "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\nMolgaard et al. have investigated the suitability of various commercially available antibodies for the identification of GABAergic interneurons in mice. Among 13 tested antibodies against calbindin (3), calretinin (4), parvalbumin (4) and somatostatin (2), the authors found 1-2 antibodies per each marker that produced high quality, sensitive and specific staining in mouse brain sections and cultured hippocampal neurons obtained from P0 mouse pups. Their findings remind us about the tediousness of immunostainings in general and challenges about the identification of proper antibody and antibody conditions suitable for high quality research. Although the antibodies have not been investigated in all possible experimental conditions (e.g. fixation protocols, dilutions), the study provides very valuable reference information for researchers aiming to investigate GABAergic markers in mice.\n\nIn general the manuscript has been written and constructed well. Abstract nicely describes the summary of the study. The introduction provides very good background knowledge for the reader (incl. relevant citations). The methods section is described in a manner that allows scientific reproduction efficiently. Tables are clear and useful. Overall the representative figures are good but the paper would benefit with more comprehensive set of immunostainings (as main figures).\n\nI have few minor comments/questions:GABAergic interneurons are considered as small neuronal population in the text. In respect to glutamatergic neurons this is indeed the case, but overall I consider 20% quite a significant fraction (cf. monoaminergic neurons).The authors could have clarified what is “pAbmAbs” in the abstract.In the Introduction the authors state that breakdown of cortical network rhythms underlie neuropsychiatric disorders. I would rather say that alterations in cortical network rhythms in specific brain networks may underlie neuropsychiatric disorders.The authors could have explained that glutamate decarboxylase is expressed in GABAergic neurons and it synthesizes GABA (Introduction, first and second paragraph).The authors could have described the gender and amount of adult mice used for the study. Moreover, where all the antibodies tested in specimens derived from same conditions (e.g. same animal).I would have used “sections” rather than “slices” throughout the paper.Table 2. would be even more clear if the antibodies against the four different markers would have been divided from each other more clearly (e.g. using different background colors).Why there appears no Hoechst staining in some of the unspecific stainings?Why the authors choose not to show the representative figures from spinal cord?The authors could have emphasized that the quality of polyclonal antibodies is significantly determined by the lot/batch. This should be kept in mind when reproducing the findings.Figure legends should have been clearer. In optimal case, the reader understands the figures thoroughly without the main text (e.g. age of cultures).It would have been very useful to test, at least selected, antibodies in rats as well (brain sections, culture)", "responses": [ { "c_id": "1074", "date": "12 Nov 2014", "name": "Simon Molgaard", "role": "Author Response", "response": "\"GABAergic interneurons are considered as small neuronal population in the text. In respect to glutamatergic neurons this is indeed the case, but overall I consider 20% quite a significant fraction (cf. monoaminergic neurons).\" Reply: Comment well taken. The word “small” has now been removed from the abstract. \"The authors could have clarified what is “pAbmAbs” in the abstract.\" Reply: This has been added to abstract. \"In the Introduction the authors state that breakdown of cortical network rhythms underlie neuropsychiatric disorders. I would rather say that alterations in cortical network rhythms in specific brain networks may underlie neuropsychiatric disorders.\"Reply: We agree. The text is now changed accordingly. \"The authors could have explained that glutamate decarboxylase is expressed in GABAergic neurons and it synthesizes GABA (Introduction, first and second paragraph).\" Reply: This has now been clarified in the introduction. \"The authors could have described the gender and amount of adult mice used for the study. Moreover, where all the antibodies tested in specimens derived from same conditions (e.g. same animal).\" Reply: A paragraph has been added to the Methods clarifying these issues. \"I would have used “sections” rather than “slices” throughout the paper.\" Reply: This has now been corrected. \"Table 2. would be even more clear if the antibodies against the four different markers would have been divided from each other more clearly (e.g. using different background colors).\" Reply: Although we agree that a color-code would be desirable, we feel that the current table is sufficiently clear. \"Why there appears no Hoechst staining in some of the unspecific stainings?\" Reply: We apologize for the lack of clarity regarding this issue. The “unspecific” images with Hoechst are shown when we observed no signal at all. The “unspecific” images without Hoechst are shown when staining was present but distributed in an unspecific manner. \"Why the authors choose not to show the representative figures from spinal cord?\" Reply: The images are already included in the data availability section. \"The authors could have emphasized that the quality of polyclonal antibodies is significantly determined by the lot/batch. This should be kept in mind when reproducing the findings.\" Reply: Good point. In this regard we will refer to the article in the antibody validation collection by Dr Jan Voskuil. \"Figure legends should have been clearer. In optimal case, the reader understands the figures thoroughly without the main text (e.g. age of cultures).\" Reply: The figure legends have now been improved \"It would have been very useful to test, at least selected, antibodies in rats as well (brain sections, culture) \" Reply: We agree but as stated in the title we have only used mouse." } ] }, { "id": "6402", "date": "21 Oct 2014", "name": "Mei Yee Leung", "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 an informative, concise article with clear aims that highlights the potential difficulties in selecting the right antibodies for specific cell types and research applications. In this study, the authors systematically tested commercial antibodies against calbindin, calretinin, parvalbumin and somatostatin - markers of GABAergic interneuron subtypes. Of the 13 antibodies tested, only 4 were deemed reliable and useful for characterizing these subtypes in mouse brain.The authors’ rigorous approach in selecting the right antibodies is commendable, a fact which often goes unnoticed in publications. The ranking of antibody performance in pAbmAbs, a review-based platform is a much needed resource for scientists whose research depends on the validity of the antibodies.As someone not in this research area, I found the manuscript scientifically well construed and the rationale easy to follow. It is noteworthy that this study by Molgaard et al. adds to the understanding of GABAergic subtypes in the mouse hippocampus and spinal cord, information which is sparse in the literature. In addition, by demonstrating specific staining of these markers, they have confirmed previously reported localization of these cells. A few comments/suggestionsWould a change in the title be more appropriate? e.g. Immunofluorescent instead of immunohistochemicalIs there any reason why the age of the mouse used for hippocampal staining and that for spinal cord staining is different?It would add value to this paper is images of spinal cord staining was also shownAlthough the without primary images are very clean, using these antibodies on tissues not known to express these targets would be a more superior negative controlCould the authors offer an explanation for why the polyclonal antibodies appear to perform better than the monoclonals?To obtain better idea of reproducibility, it would be good to give an indication of how many times the experiment was performed and how many sections were stained per experiment", "responses": [ { "c_id": "1073", "date": "12 Nov 2014", "name": "Simon Molgaard", "role": "Author Response", "response": "\"Would a change in the title be more appropriate? e.g. Immunofluorescent instead of immunohistochemical\" Reply: OK \"Is there any reason why the age of the mouse used for hippocampal staining and that for spinal cord staining is different?\" Reply: No particular reason. The expression of GABAergic markers in the hippocampus and spinal cord of adult animals is expected to be fairly constant throughout adulthood. \"It would add value to this paper is images of spinal cord staining was also shown\" Reply: Done \"Although the without primary images are very clean, using these antibodies on tissues not known to express these targets would be a more superior negative control\" Reply: This point may rely on a possible misunderstanding. The images in the paper denoted “Unspecific staining” refers to stainings using primary antibodies for which no specific signal was observed. We have changed the legends to make this more clear. \"Could the authors offer an explanation for why the polyclonal antibodies appear to perform better than the monoclonals?\" Reply: We have no clear answer to this but it may be that polyclonal antibodies generally give a higher signal compared monoclonal antibodies due to the presence of multiple epitopes. \"To obtain better idea of reproducibility, it would be good to give an indication of how many times the experiment was performed and how many sections were stained per experiment\" Reply: This is now stated in the Methods section." } ] } ]	1	https://f1000research.com/articles/3-242
https://f1000research.com/articles/3-244/v1	15 Oct 14	{ "type": "Research Note", "title": "Application of ARID1A to murine formalin-fixed paraffin embedded tissue using immunohistochemistry", "authors": [ "Will Howat", "Jodi Miller", "Ioannis Gounaris", "Jodi Miller", "Ioannis Gounaris" ], "abstract": "ARID1A is a known suppressor of tumour formation and the Human Protein Atlas antibody HPA005456 has been demonstrated in previous literature to stain tumour tissue by immunohistochemistry (IHC) in formalin-fixed paraffin embedded human tissue and human cell lines. This article details the validation of this antibody for immunohistochemistry of formalin-fixed paraffin embedded murine tissue using a Leica BondMax immunostainer. Using Western blot and IHC on murine wild-type and knockout tissue we have demonstrated that this antibody to ARID1A correctly stains murine tissue by immunohistochemistry.", "keywords": [ "immunohistochemistry", "antibody", "ARID1A", "tissue" ], "content": "Introduction\n\nARID1A (AT-rich interactive domain 1a) is a member of the SWI/SNF family and its loss has been implicated as a factor in multiple premalignant and malignant conditions, including Barrett’s oesophagus and oesophageal carcinoma as well as endometrial and clear cell ovarian carcinomas and their precursor endometriotic lesions1–4. The ARID1A antibody from Human Protein Atlas is a rabbit antibody generated against a PrEST (Protein Epitope Signature Tag) fragment of the ARID1A gene and affinity purified against the same fragment5. It is thus designated as being “mono-specific” in that the affinity purification removes any non-specific or low affinity binders to the peptide. Through the Human Protein Atlas, the antibody has been tested on a wide variety of human tissue types and human malignancies, as well as for expression in immunofluorescence on U-2 OS, A-431 and U-251 MG cell lines. This demonstrates a nuclear expression in all cell lines and in the majority of tissue types6. However, the Western blot data were not supportive and did not produce staining corresponding to the expected size, although the data from a protein array did confirm a peak at the expected size. The antibody has been used to stain human colorectal cancers7, clear cell carcinomas8 and on a variety of clear cell cancer cell lines9 by immunohistochemistry.\n\nTo our knowledge, whilst the sequence homology between mouse and human ARID1A is 95%, this antibody has not been qualified using knockout tissue and has not been tested or published on murine tissue and this work represents the first data to do so.\n\n\nMaterials and methods\n\nDetails of all reagents with reference to the immunohistochemical staining procedure can be found in Table 1.\n\nAnti-ARID1A is a monospecific rabbit polyclonal generated to a PrEST sequence – PGLGNVAMGPRQHYPYGGPYDRVRTEPGIGPEGNMSTGAPQPNLMPSNPDSGMYSPSRYPPQQQQQQQQRHDSYGNQFSTQGTPSGSPFPSQQTTMYQQQQQNYK (Table 2). The lot number used for all validations was A40072 and for subsequent staining was D81856. A concentration of 1 µg/ml was used for initial validations and 0.5 µg/ml for final runs.\n\nDonkey anti-rabbit biotin (Jackson Immunoresearch, Table 2) is specific for Rabbit IgG (Heavy and Light chains) and was affinity purified to remove cross-reactions to Bovine, Chicken, Goat, Guinea Pig, Syrian Hamster, Horse, Human, Mouse, Rat and Sheep. All slides were stained with a concentration of 4.8 µg/ml.\n\nAnti-GAPDH was used as a loading control for Western blots and was a rabbit monoclonal (Cell Signaling, Table 2). Detection antibody for the Western blot for ARID1a was Goat anti-rabbit IR Dye 680LT (Li-Cor Biosciences, Table 2) used at a concentration of 0.1 µg/ml and detection of GAPDH was Goat anti-rabbit IR Dye 800CW (Li-Cor Biosciences, Table 2).\n\nAll tissues and cell pellets used during the validation were fixed in Neutral Buffered Formaldehyde as specified (Table 3) before being transferred directly to 70% ethanol for no longer than 3 days. Tissue processing was conducted on a Leica ASP300 through graded ethanols before clearing in xylene and impregnation in molten paraffin wax (Fisher). All tissue sections were cut on a Leica rotary microtome at 3 µm.\n\nWestern blot. Protein was extracted from the two clear cell carcinoma cell lines using a Tris-EDTA lysis buffer and run on a non-denaturing 3–8% Tris-acetate gel (Life Technologies). Following electrophoresis, the transfer membrane was probed with 0.2 µg/ml of anti-rabbit ARID1A (HPA005456) at 4°C overnight and 0.1 µg/ml anti-GAPDH (14C10) for the same length of time. Detection of the anti-rabbit ARID1A was with Goat anti-rabbit IRDye 680LT (Li-Cor Biosciences) and the GAPDH was with Goat anti-rabbit IRDye 800CW (Li-Cor Biosciences) both at 0.1 µg/ml.\n\nImmunohistochemistry. Slides were deparaffinised and rehydrated on a Leica ST5020 using Xylene (Sigma) for 2 × 10 mins and ethanol (Fisher), 2 × 100% ethanol followed by 1 × 70% ethanol for 5 mins each. Following staining, all slides were dehydrated, cleared and mounted and coverslipped in DPX (Fisher).\n\nThe antibody was validated on a Leica BondMax instrument using a Leica Intense R kit to a standardised in-house protocol. All reagents were from Leica as part of the Intense R kit and were conducted at room temperature, unless otherwise specified. All staining steps included individual washes in Leica Bond Wash after each step, as part of the protocol (Table 4). A full protocol for the validated conditions can be found in the supplementary material. In this protocol, the step named “primary” refers to the anti-ARID1a primary antibody.\n\nA slide using the same conditions and retrieval but omitting the primary antibody was used to control for any background staining due to the retrieval and detection steps.\n\nImaging. All slides were digitised using a Leica Scanscope AT2 at 0.5 µm/pixel resolution. Datasets can be viewed by downloading the Leica Imagescope free viewer at http://www.leicabiosystems.com/pathology-imaging/aperio-epathology/integrate/imagescope/.\n\n\nResults\n\nTo determine the correct cell line to utilise and to confirm the equivocal Western blot data from Human Protein Atlas, the antibody was used to stain a Western blot of two cell lines; ES-2 and RMG-II, both of which are cell lines derived from clear cell carcinoma. It could be demonstrated that the HPA ARID1A antibody showed positive expression in ES-2 cell lines at the expected size of 270 kDa and no staining for RMG-II. The loading control of GAPDH showed that there were no loading issues (Figure 1). Thus, these cell lines were chosen to be grown, formalin fixed and processed into paraffin wax for immunohistochemical validation.\n\nARID1A (red band) can be seen to be present at approximately 270kD in ES2 cell line only. GAPDH at 37kD (green band) represents loading control.\n\nFor immunohistochemical validation, ES-2 and RMG-II cell lines were stained using three antigen retrieval conditions; ER1 (Sodium Citrate, pH6), ER2 (Tris/EDTA, pH9) and Enzyme 1 (Proteinase K, 100 µg/ml) at a fixed antibody concentration of 1 µg/ml. The enzyme retrieval demonstrated no nuclear signal for either ES2 or RMG-II cell pellets and was discarded for future work (Figure 2; Dataset a). The ER2 condition did demonstrate significant nuclear staining in the ES2 cell pellet with minimal background staining in the RMG-II cell pellet (Figure 2; Dataset b). However, the staining in the ER1 condition was determined to give the best signal:noise ratio with no background cytoplasmic staining and crisp nuclear staining for the cell pellet (Figure 2; Dataset c). Control slides, omitting the primary antibody, were negative except for the ER2 condition in the RMG-II cell pellet where a weak cytoplasmic background could be seen (Figure 2; Dataset d). Thus there was minimal background inherent in the staining procedure. It was therefore determined that the antibody showed specificity for formalin-fixed paraffin embedded tissues and could be run on murine tissue.\n\nNPA denotes No Primary antibody control and represents the ER2 condition. Bar = 200 µm.\n\nMurine uterine tissue was used as positive control tissue samples, given the literature data on cell lines and endometrial tissue. The ER1 condition at 1 µg/ml demonstrated clean nuclear staining in the uterine epithelial compartment as well as nuclear staining of stromal cells. However, the nuclear staining in the stroma was not universal and distinct negative nuclei could be seen (Figure 3; Dataset e). There was no cytoplasmic or extracellular stromal background staining present and the antibody titrated successfully losing the intensity of staining, as expected (Dataset e). Following this, a concentration of 0.5 µg/ml was used for future preparations which provided clear and consistent staining in repeated batches using a different antibody lot (Dataset f).\n\nMurine uterine tissue stained by immunohistochemistry with anti-ARID1A using the ER1 condition and a concentration of 1 µg/ml demonstrating clear nuclear staining of the epithelial compartment (E) and negative nuclei in the stromal compartment (S). Bar = 100 µm.\n\nFinally, when applied to a genetically engineered, tamoxifen-induced, Arid1a knockout mouse model, the staining in the uterine epithelium could be completely abrogated (Arrow, Figure 4b) when compared to the same area in a wild-type animal (Arrow, Figure 4a), while not affecting the staining in the stromal compartment.\n\nMurine uterine tissue stained by immunohistochemistry with anti-ARID1A demonstrating nuclear staining in wild-type mice (A) but loss of epithelial staining after ARID1A knock-out in Arid1afl/fl mice (B). Bar = 100 μm.\n\n\nConclusions\n\nIt is clear from the use of ES2 and RMG-II cell lines that the Atlas Antibodies ARID1A antibody is specific for ARID1A in both Western blots and formalin-fixed paraffin embedded preparations of human origin and, coupled with the literature evidence, that it is validated in human tissue.\n\nThe staining pattern when applied to murine uterus showing a clear nuclear pattern, is again consistent with the literature on this protein. Crucially, however, when stained on an ARID1a knockout mouse model, the staining could be almost completely abrogated in the epithelial compartment and thus when taken in combination, it is clear that the anti-human ARID1a antibody is cross-reactive with murine tissue and can be used for this purpose.\n\n\nData availability\n\nF1000Research: Dataset 1. Whole slide images from antibody validation of HPA005456 for immunohistochemistry, 10.5256/f1000research.5514.d3676110", "appendix": "Author contributions\n\n\n\nWH wrote and conceived of the article, IG performed the Western blots and mouse experiments and requested validation of ARID1A in murine tissue, JM performed all immunohistochemical staining.\n\n\nCompeting interests\n\n\n\nThe authors do not declare any competing interests.\n\n\nGrant information\n\nWH and JM are funded by Cancer Research UK, IG is funded by an MRC Clinical Research Training Fellowship, grant number G1001957.\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\nThe authors would like to acknowledge the University of Cambridge, Cancer Research UK and Hutchison Whampoa Ltd. The authors would also like to acknowledge the staff at the Histopathology/ISH core facility at the Cancer Research UK Cambridge Institute for their assistance in preparing materials for this publication.\n\n\nReferences\n\nStreppel MM, Lata S, DelaBastide M, et al.: Next-generation sequencing of endoscopic biopsies identifies ARID1A as a tumor-suppressor gene in Barrett's esophagus. Oncogene. 2014; 33(3): 347–357. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWiegand KC, Hennessy BT, Leung S, et al.: A functional proteogenomic analysis of endometrioid and clear cell carcinomas using reverse phase protein array and mutation analysis: protein expression is histotype-specific and loss of ARID1A/BAF250a is associated with AKT phosphorylation. BMC Cancer. 2014; 14: 120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWiegand KC, Shah SP, Al-Agha OM, et al.: ARID1A mutations in endometriosis-associated ovarian carcinomas. N Engl J Med. 2010; 363(16): 1532–1543. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDulak AM, Stojanov P, Peng S, et al.: Exome and whole-genome sequencing of esophageal adenocarcinoma identifies recurrent driver events and mutational complexity. Nat Genet. 2013; 45(5): 478–486. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNilsson P, Paavilainen L, Larsson K, et al.: Towards a human proteome atlas: high-throughput generation of mono-specific antibodies for tissue profiling. Proteomics. 2005; 5(17): 4327–4337. PubMed Abstract \| Publisher Full Text\n\nHuman Protein Atlas. Reference Source\n\nWang K, Kan J, Yuen ST, et al.: Exome sequencing identifies frequent mutation of ARID1A in molecular subtypes of gastric cancer. Nat Genet. 2011; 43(12): 1219–1223. PubMed Abstract \| Publisher Full Text\n\nYamamoto S, Tsuda H, Takano M, et al.: Loss of ARID1A protein expression occurs as an early event in ovarian clear-cell carcinoma development and frequently coexists with PIK3CA mutations. Mod pathol. 2012; 25(4): 615–624. PubMed Abstract \| Publisher Full Text\n\nAnglesio MS, Wiegand KC, Melnyk N, et al.: Type-specific cell line models for type-specific ovarian cancer research. PLoS One. 2013; 8(9): e72162. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHowat WJ, Miller JL, Gounaris I: Whole slide images from antibody validation of HPA005456 for immunohistochemistry. F1000Research. 2014. Data Source\n\n\nSupplementary material\n\n\n\n\n\n\n\n\n\n" }	[ { "id": "6651", "date": "21 Nov 2014", "name": "Andrew D. Chalmers", "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 by Will Howat and colleagues validates an anti-ARID1A antibody and provides a good example of an antibody validation paper. In particular the use of negative cell lines and knockout mouse tissue demonstrate the extent of specificity of the antibody. The data is mostly well presented and clearly explained in the text.The manuscript would be suitable for indexing if some, mostly minor, issues can be addressed.I think the title should mention “antibody” to make clear it’s an antibody that is being validated. The authors say there is 95% sequence homology between human and mouse, it would be interesting to see how conserved the sequence the antibody was raised against is. The Materials and Methods give a good overview of the reagents and methods used for the IHC, but there is little information about the western blotting. The reagents and methods used should be added. It would be good to cite the source of the cells used and the mouse tissue. I was not clear if the lack of staining in RMG-II cells was expected or an unexpected but useful result? This should be explained. Figure 1 legend. Should read “represents the loading control” In figure 4B the KO epithelium is clearly negative, except one region in the top left which looks positive, I wondered what the authors felt about this? It is interesting that the stromal staining appears to be non-specific, I wonder if the authors ever did a no primary control? Does the stroma still stain? Also it would be worth mentioning this staining in the conclusions. Author contributions. Should it read “conceived the idea behind the article” or similar wording ?", "responses": [ { "c_id": "1154", "date": "06 Jan 2015", "name": "Will Howat", "role": "Author Response", "response": "Dear Dr Chalmers,Many thanks for your comments which we agree with and have taken on board. Specifically:We have modified the title to read \"Application of anti-ARID1a antibody...\" We have BLAST searched the protein sequence and it confirms 95% homology for the sequence used for immunisations. This has been added to the methods. We accept that the methods detailing the Western blotting are more limited than the IHC methods, as is the case of the methods detailing the production of the knockout mouse model. However, we feel that the article was designed as an example of antibody validation for immunohistochemistry and that while western blotting provides important additional data, it is not the focus of the article. We thus feel that there is sufficient information in the details behind the western blotting to repeat the experiment, without clouding the article with the full methods. The sources of the cells and mouse tissue have now been cited. The lack of staining of RMG-II cells was consistent with the literature (Anglesio et al.) and this information has been added to the results section. The figure legend has been modified. We believe that the focal staining is a result of incomplete recombination floxed Arid1a allele and Cre/ERT2. The resulting knockout is almost never 100% complete as in some cells recombination will not be induced due to issues such as low ligand (tamoxifen) penetration or failure of the ligand to induce recombination. Similarly, we believe that the staining in the stromal compartment is specific but represents a failure of delivery of Tamoxifen or of recombination. The data from no primary antibody (NPA) control which are all completely clean, now included in Fig 3, helps to demonstrate this. We have included these points in the conclusion. The author contributions have been modified.Kind Regards,Will Howat" } ] }, { "id": "6848", "date": "02 Dec 2014", "name": "Stephen McQuaid", "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 article, by Howatt et al. validates an antibody to ARID1A in murine FFPE samples, is a good example of the correct validations procedures that must be met when validating an antibody for use in tissue sections. Issues are relatively minor:What was the rationale for a starting concentration of 1μg/ml for initial validations? There is a focal region of the KO epithelium which clearly appears to be positive. The authors would need to explain this. Can the authors please add more discussion of the nature of the staining of the stromal cells. Do they consider this to be non-specific? Is such staining of stroma present in human tissue sections? Control slides, omitting primary antibody, were carried out during the cell line phase of the experiments. Were such slides run on the murine tissue sections and can the results be included in the data. I note that the stromal staining was still present at selected concentration of 0.5μg/ml (dataset f)", "responses": [ { "c_id": "1153", "date": "06 Jan 2015", "name": "Will Howat", "role": "Author Response", "response": "Dear Dr McQuaid,Thank you for your comments which we agree with and have taken on board. Specifically:All of the validations conducted at the Histopathology/ISH facility start with a dilution of 1:100 from the manufacturers provided stock concentration and we perform 3 retrieval methods; Sodium Citrate pH6, Tris EDTA pH9 & Proteinase K. We titrate as appropriate following establishment of the optimal retrieval method. The manufacturers original stock concentration was 100ug/ml, hence the 1ug/ml starting concentration. They have subsequently changed their stock concentration to 200ug/ml. We believe that the knockout model recombining a floxed Arid1a with CRE-ERT2 is not 100% complete as in some cells recombination will not be induced due to issues such as low ligand (tamoxifen) penetration or failure of the ligand to induce recombination. Thus explaining the area of focal staining in the epithelium. Similarly, although we have not investigated this further, we believe that the same problem underlies the stromal staining and is likely due to penetrance of the tamoxifen after being delivered intraperitoneally. Thus, given the wealth of data and the no primary antibody controls, now included in Fig 3, we feel that the stromal staining is not non-specific. Stromal staining does indeed occur in human tissue.All of the above have now been included in the conclusion section of the article.Kind Regards,Will Howat" } ] } ]	1	https://f1000research.com/articles/3-244
https://f1000research.com/articles/3-104/v1	09 May 14	{ "type": "Research Article", "title": "Simple, biologically-constrained CA1 pyramidal cell models using an intact, whole hippocampus context", "authors": [ "Katie A. Ferguson", "Carey Y. L. Huh", "Benedicte Amilhon", "Sylvain Williams", "Frances K. Skinner", "Katie A. Ferguson", "Carey Y. L. Huh", "Benedicte Amilhon", "Sylvain Williams" ], "abstract": "The hippocampus is a heavily studied brain structure due to its involvement in learning and memory. Detailed models of excitatory, pyramidal cells in hippocampus have been developed using a range of experimental data. These models have been used to help us understand, for example, the effects of synaptic integration and voltage gated channel densities and distributions on cellular responses. However, these cellular outputs need to be considered from the perspective of the networks in which they are embedded. Using modeling approaches, if cellular representations are too detailed, it quickly becomes computationally unwieldy to explore large network simulations. Thus, simple models are preferable, but at the same time they need to have a clear, experimental basis so as to allow physiologically based understandings to emerge. In this article, we describe the development of simple models of CA1 pyramidal cells, as derived in a well-defined experimental context of an intact, whole hippocampus preparation expressing population oscillations. These models are based on the intrinsic properties and frequency-current profiles of CA1 pyramidal cells, and can be used to build, fully examine, and analyze large networks.", "keywords": [ "Capacitance", "fluoresence", "hippocampus", "neurons" ], "content": "Introduction\n\nNetworks of excitatory and inhibitory neurons are essential components constituting the functional structures of our brains. Dysfunction is thought to occur when inappropriate excitation-inhibition balances occur1. From a modeling perspective, these balances are determined by the choice of parameters in the equations representing neurons and networks. Mathematical models of neurons and networks are developed so that they can be used to determine the mechanisms underlying brain functions. However, it is difficult to assess whether the mechanisms determined from mathematical models actually occur biologically2,3. Furthermore, it is well known that cellular models used in building network models affect and can dictate the network output4,5. To address this recognized difficulty we are developing models that are based on well-defined experimental contexts in which both the cellular and the network aspects of the model can be considered simultaneously6,7. Using such models, we aim to help determine, predict and test biologically based mechanisms.\n\nHere we use a well-defined experimental context of an in vitro intact, whole hippocampus preparation in which spontaneous population theta and theta-gamma rhythms are expressed8,9. Access to many cellular details is possible in vitro, and with a physiologically relevant output of theta and theta-gamma rhythms, a reasonable and functional scenario is also present. We have previously developed models of CA1 fast-spiking inhibitory cells in the same experimental context, and used them to show that model inhibitory networks of these fast-spiking cells exhibit sharp transitions between random and coherent firings at high frequencies (>90 Hz) when connectivity constraints were imposed6. In this article, we present the result of CA1 pyramidal cell models developed in the same context of this in vitro whole hippocampus preparation. Similar to our previous cellular model developments6, the models presented here are biologically-based at the cellular level, but do not have a biophysical representation in terms of conductance-based model representations. These CA1 models use a simple two variable mathematical formulation based on Izhikevich10 and they include rebound firing and adaptation characteristics. Using these simple, yet biologically-based models, we can build and examine several large network models that are aligned with the biology. Subsequent analyses of these large network models could determine the mechanisms by which particular cellular characteristics critically contribute to the population activities observed in our experimental context.\n\n\nMethods\n\nAnimals. Three mice (two female, post-natal day 20–29) were used. We used transgenic mice that expressed a fluorescent protein, tdTomato, under the control of the PV promoter. PV-Cre homozygote mice (strain name: B6;129P2-Pvalbtm1(cre)Arbr/J, stock number: 008069, Jackson Laboratory) were mated with a reporter line, Ai9 homozygote mice (strain name: B6;129S6-Gt(ROSA)26Sortm9(CAG-tdTomato)Hze/J, stock number: 007905, Jackson Laboratory) to generate PV-tdTomato mice. The mouse lines are on genetic backgrounds that are mixtures of C57BL/6 and a type of 129 mice; PV-Cre mice are C57BL/6;129P2 and Ai9 mice are C57BL/6;129S6. Mice were bred in-house at the Douglas animal facility and kept in standard laboratory cages with standard bedding and environmental enrichment. They were housed in a temperature-controlled room with a 12:12 hours dark/light cycle with food and water provided ad libitum. All animals were treated according to protocols and guidelines approved by McGill University and the Canadian Council of Animal Care (CCAC). Ethical approval was obtained to conduct this study (approval number: 2010-5827). The authors note that the use of scissors to decapitate mice at that age without administering anesthesia (which could have altered synaptic transmission) was approved by the CCAC.\n\nIntact hippocampal preparation. The acute preparation containing the whole hippocampus was obtained from PV-tdTomato mice according to a previously described protocol8. Briefly, after decapitation using scissors, the brain was quickly removed from the skull and placed in ice-cold high-sucrose solution, containing (in mM) 252 sucrose, 24 NaHCO3, 10 glucose, 3 KCl, 2 MgCl2, 1.25 NaH2PO4 and 1 CaCl2 (pH 7.3, oxygenated with 95% O2-5% CO2). From a hemisected brain, the septum and hippocampus along with the interconnecting fibers were carefully and rapidly dissected out using a microspatula. The preparation was trimmed in ice-cold high-sucrose solution (same contents as the high-sucrose solution listed above) using fine scissors to remove any remaining cortical tissue and the septum. Then, the surface of the preparation was cut at a ~45° angle to expose the pyramidal layer of CA1. The cut enabled visually guided patch-clamp recordings of pyramidal cells which yielded a higher success rate for whole-cell recordings compared to the blind-patch technique used previously8. The visual approach allowed identification of CA1 pyramidal cells by their soma location, morphology and the lack of PV-tdTomato fluorescence in the soma. The hippocampal preparation was then left to equilibrate in oxygenated room-temperature high-sucrose solution for 30 min - 1 h before recording. The preparation from only one hemisphere was used for recording from each mouse, and the preparation from either the left or the right hemisphere was chosen randomly for each experiment. Three mice were used in total; we used three intact hippocampal preparations from these mice (one from each mouse) and one pyramidal cell was recorded from each preparation, except for one preparation from which two cells were recorded.\n\nAll electrophysiological recordings were performed at 30 ± 2ºC, using artificial cerebrospinal fluid (aCSF) containing (in mM) 126 NaCl, 24 NaHCO3, 10 glucose, 4.5 KCl, 2 MgSO4, 1.25 NaH2PO4, 0.4 ascorbic acid and 2 CaCl2 (pH 7.3, oxygenated with 95% O2-5% CO2). The hippocampal preparation was placed in a custom-made submerged recording chamber lined with a nylon mesh, and firmly stabilized by carefully placing several lead weights at both septal and temporal poles of the hippocampal preparation. We placed the hippocampal preparation in the recording chamber, such that the CA1 was the most superficial and accessible sub-region for visualization and whole-cell recordings. Recordings were restricted to neurons located within the middle portion of the hippocampus (intermediate between septal and temporal poles of the preparation). The preparation’s stability in the recording chamber was extremely important as aCSF was perfused at the rate of 20–25 ml/min during recordings. Since the tissue is several millimeters thick, such a fast perfusion rate is necessary to ensure sufficient oxygenation. This fast perfusion rate is also required to generate intrinsic theta oscillations from intact hippocampal preparations8. In order to achieve temperature stability, aCSF was pre-heated using an electric skillet and further regulated via an automatic temperature controller (Warner Instruments, Hamden, CT). The electrophysiology setup was equipped with an upright BX51W1 Olympus microscope, a 20× water-immersion objective, Nomarsky optics, an infrared camera (Cohu, San Diego, CA), a variable-wavelength fluorescence system (PTI, Monmouth Junction, NJ) and a monochrome digital camera for fluorescence imaging (DAGE-MTI, Michigan City, IN). Patch pipettes were pulled from borosilicate glass capillaries (2.5–4 MΩ) and filled with the intra-pipette solution containing (in mM) 144 K-gluconate, 10 HEPES, 3 MgCl2, 2 Na2ATP, 0.3 GTP, 0.2 EGTA, adjusted to pH 7.2 with KOH. The patch electrode was controlled using a motorized micromanipulator (Sutter Instruments, Novato, CA). An Axopatch-1C amplifier (Axon Instruments, Foster City, CA) and pClamp9 software (Molecular Devices, Sunnyvale, CA) were used for recording. The junction potential was estimated at -15.2 mV and was corrected for. All drugs were obtained from Sigma-Aldrich (St. Louis, MO), unless otherwise noted.\n\nIntrinsic property characterization. Intrinsic properties were characterized in current-clamp mode following published protocols with minor modifications11. A mixture of synaptic blockers was used to inhibit synaptic activity: 5 μM 6,7-dinitroquinoxaline-2,3-dione disodium salt (DNQX), 5 μM bicuculline and 25 μM DL-2-amino-5-phosphonopentanoic acid sodium salt (DL-AP5) (Abcam, Toronto, Canada). Data analysis was done using custom Matlab software (MathWorks, Natick, MA). Once the whole-cell configuration was achieved, the cell’s resting membrane potential was noted and its spontaneous firing, if any, was recorded for 30 s. Access resistance and resting membrane potential were checked at regular intervals (every ~5 min) throughout the recording of the cell.\n\nThe frequency-current (f-I) profiles of the pyramidal cells are important to characterize, as we aim for our single cell model to respond to a variety of synaptic input strengths with frequencies similar to that observed experimentally. These f-I curves were determined by applying depolarizing current steps of 1 s duration to cells held in current clamp. Amplitudes were increased incrementally with step sizes of 25 pA for one of four cells, and 10 pA for three of four cells. The initial firing frequency was determined based on the inverse of the first inter-spike interval, and the final frequency was based on the inverse of the last inter-spike interval in the 1-second depolarizing step. For each cell, the approximate linear slope of the f-I curve above 5 Hz was determined using a least squares method. These values were chosen since above 5 Hz the slope was well-approximated by linearization. In addition, the minimum amount of current required to initiate a spike, the rheobase current (Irheo in pA), was determined. The action potential threshold was set to be the first voltage point such that the slope of the membrane potential exceeded 20 mV/ms12. The spike width was determined at the threshold value. In addition, the spike height from threshold and the minimum membrane potential reached following the spike were measured. Recordings were kept for analysis only if the neuron remained stationary; spikes overshot 0 mV (-15 mV junction potential corrected) and access resistance < 30 MΩ.\n\nWe built a pyramidal cell model based on Izhikevich’s10 simple spiking model structure. We chose this model as it captures the cell’s ability to produce rebound spiking, the approximate spike shape, and the frequency-current profile of the cell, including spike-adaptation. Thus, this model is relatively simple, but allows one to capture important biophysical properties of pyramidal cells.\n\nThe model has a fast variable representing the membrane potential, V (mV), and a slow “recovery” current given by the variable u (pA). In order to capture the spike width at threshold, we slightly modified the Izhikevich model by using a different “k” parameter above and below the spike threshold (khigh and klow respectively). The model is given by:\n\nCmV˙ = k(V–vr) (V–vt) – u + Iapplied\n\nu˙ = a[b(V – vr) – u] (1)\n\nif V ≥ vpeak, then V ← c, u ← u + d\n\nWhere k = klow if V ≤ vt ; k = khigh if V > vt\n\nThe parameters are as follows:\n\nCm (pF) is the membrane capacitance.\n\nvr (mV) is the resting membrane potential.\n\nvt (mV) is the instantaneous threshold potential.\n\nvpeak (mV) is the spike cut-off value.\n\nIapplied (pA) is the applied current, and represents the applied input into the cell.\n\na (ms-1) is the recovery time constant of the adaptation current.\n\nb (nS) describes the sensitivity of the adaptation current to subthreshold fluctuations. Greater values couple V and u more strongly resulting in possible subthreshold oscillations and low-threshold spiking dynamics. Further, the sign of b determines whether the effect of u is amplifying (b < 0) or resonant (b > 0).\n\nc (mV) is the voltage reset value.\n\nd (pA) is the total amount of outward minus inward currents activated during the spike and affecting the after-spike behaviour.\n\nk (nS/mV) represents a scaling factor. khigh is used to adjust the spike width after the threshold.\n\nAs with the experimental frequency-current (f-I) curve frequencies, we exhibited the “initial” and “final” f-I curves for each model, where the initial curves were based on the inverse of the first inter-spike interval (ISI) due to a one second current step, and the final curves were based on the inverse of the last ISI. A linear fit based on the least squares method (on frequencies over 10 Hz) was done for each curve (using 5 Hz gave essentially the same results). We then chose parameters in which our models exhibited similar initial and final f-I curves to those of the experimental pyramidal cells. To do so, we first set the resting membrane potential at the rheobase, the spike threshold, the minimum potential reached by the spike after-hyperpolarization, the spike peak, and the spike width at threshold (providing us with values for vr, vt, c, vpeak, and khigh respectively) based on our experimentally determined values (see Results). The threshold is defined in “Intrinsic Property Characterization”. Since our pyramidal cells exhibited resonant properties, we considered b values such that b > 0. We initially held b and klow constant (at 0.2 nS and 0.05 nS/mV), and varied a between 0 and 1 with an initial step size of 0.01, and d between 0 and 20 with a step size of 1. Choosing our a and d parameters that returned the best fits to our initial and final slopes, we then varied b between 0.1 and 10, and klow between 0 and 20 (both with a step size of 0.1). Noting that we required more adaptation, we then returned to vary a between 0 and 0.1 with an initial step size of 0.0001, and d again between 0 and 20 with a step size of 1.\n\n\nResults\n\nUsing the described whole hippocampal preparation and recordings from four CA1 pyramidal cells (from three mice), we developed our simple biologically-based cellular models. Our goal is to obtain representative CA1 pyramidal cell models that capture the essence of the experimental data in the described experimental context. By biologically-based, we mean that passive properties and spike characteristics are captured. Specifically, resting potential, spike threshold, spike width, spike peak and after-hyperpolarization potentials were incorporated into our models. In addition, the adaptation characteristics determined from the frequency-current (f-I) curves (see Methods) are captured. Although not directly characterized with experimental recordings here, our models express rebound firing. Rebound firing (from inhibition or hyperpolarization) has been shown in several experimental studies13,14 and is an important consideration in population theta activities.\n\nAnalysis of the experimental data yielded the following: at rheobase, the pyramidal cell membrane potential rested at -61.8 ± 2.9 mV (mean ± SEM), the spike threshold was -57.0 ± 2.2 mV, the after hyperpolarizing potential reached a minimum of -65.8 ± 3.83 mV, and the spike reached a peak at 22.6 ± 19.9 mV (n=4 cells). The spike width at threshold, was 3.6 ± 0.48 ms. Thus, we set vr = -61.8 mV, vt = -57.0 mV, c = -65.8 mV, vpeak = 22.6 mV, and khigh = 3.3 nS/mV in our models (see Methods). The remaining model parameters were chosen such that the rheobase and f-I curves of the cell model were similar to those of the cell recordings. To do so, we first considered the f-I curves of the cell recordings (Figure 1). To demonstrate the amount of adaptation that the cell exhibited, we created two f-I curves for each cell: one based on the first inter-spike interval (ISI) of the cell’s spiking during a one second current step (denoted initial curve, data points shown as asterisks in Figure 1), and one based on the last (final curve, data points shown as squares in Figure 1). If the cell only had one spike in the 1s trace, a frequency of 1 Hz was given. We note that two of our recordings (Pyramidal cell 1 and Pyramidal cell 2, shown in red and blue in Figure 1) exhibited stronger adaptation than the other two (Pyramidal cell 3 and 4, shown in green and black in Figure 1). Thus, we created pyramidal cell models based on the strongly adapting cells and the weakly adapting cells separately. While adaptation clearly exists (e.g., here,15,16), biological variability, precise protocols and contexts need to be considered. As such, we did not aim to exactly capture adaptation characteristics, but rather to capture strongly and weakly adapting cells as represented by the data here. Also, since our model has a simple, mathematical structure, it is limited in the extent of biologically-based characteristics that it can encompass.\n\n10 pA depolarizing steps were taken for all cells except Pyramidal cell 2 (25 pA steps). The initial (final) frequencies are shown by asterisks (squares), and the lines interpolate between the data points. Pyramidal cells 1 and 2 (shown in blue and red) have higher initial frequencies, and exhibit more adaptation than Pyramidal cells 3 and 4 (shown in light green and black).\n\nIt is important that our model f-I curve captures two important properties of the experimental data: the rheobase current (i.e. the minimum amount of current required to initiate a spike), and the approximate slope of the curve. If these properties are captured, then the model will spike with similar frequencies as the physiological cell given the same amount of synaptic input. We found that for the two strongly adapting pyramidal cells (1 and 2), the slope of the linear least squares approximation of the f-I curves above 5 Hz were 0.376 and 0.385 for the initial curves, and 0.030 and 0.040 for the final curves. Again, the initial frequencies are based on the inverse of the first inter-spike interval due to a one second current step, and the final frequencies are based on the last inter-spike interval. A series of depolarizing (25 and 10 pA) steps were used to determine the rheobase currents, which were 1.2 pA and 38.7 pA for the two strongly adapting cells (f-I curves, red and blue curves in Figure 1 or data in Figure 2).\n\nWe kept our previously determined parameters constant (i.e. vr = -61.8 mV, vt = -57.0 mV, c = -65.8 mV, vpeak = 22.6 mV, and khigh = 3.3 nS/mV), and set our membrane capacitance to Cm = 115 pA. We then varied the parameters a, b, d, and klow to produce multiple models. We determined the rheobase current and the slope of the initial and final f-I curve over 10 Hz (using a least squares approach) for each model in order to settle upon a final model in which our initial and final f-I slopes and rheobase approximated that which we determined biologically. We determined that a = 0.0012 ms-1, b = 3 nS, klow = 0.1 nS/mV, and d = 10 pA. This gave us a model f-I initial slope of 0.432, a final slope of 0.099, and a rheobase of ~0 pA (see Figure 2). As shown in Figure 3, the model shows strongly adapting firing (Figure 3A) and rebound firing when hyperpolarized (Figure 3B).\n\nThe initial data points are shown as asterisks and the final points shown as squares. The initial model curve is shown as a solid line (initial slope: 0.432), and the final curve shown as a dashed line (final slope: 0.099). The model rheobase (~0 pA) and slope approximate those determined experimentally.\n\nA: An example intracellular recording of Pyramidal cell 1 during current clamp with applied current of 188 pA (top, blue) is compared with the firing of our strongly adapting pyramidal cell model, also with an applied current of 188 pA (bottom, dark green). The firing rates and amount of adaptation of the model are similar to those of the experiment. B: Two examples of rebound firing in our strongly adapting model. In each case, a one-second step of hyperpolarizing input is applied (shown as dashed lines; top: 20 pA step, bottom: 50 pA step). In each case, the strongly adapting model produces rebound spiking (solid line), where more spiking occurs for larger amounts of hyperpolarizing input.\n\nFollowing a similar methodology, we created two separate models for the weakly adapting pyramidal cells: the first better captures the cell’s weak adaptation, especially for larger currents, but has a steep final slope, whereas the second captures the more gradual slope of the final f-I curve, but doesn’t exhibit the cell’s weak level of adaptation.\n\nWe found that for the two weakly adapting pyramidal cells, the slope of the linear least squares approximation of the f-I curves were 0.119 and 0.138 for the initial curves, and 0.013 and 0.044 for the final curves (values relating to experimental, data green and black in Figure 1 and Figure 4). A series of depolarizing (10 pA) steps were used to precisely determine the rheobase currents, which were 62.0 pA and -12.1 pA for the two weakly adapting cells. We kept our previously determined parameters constant (i.e. vr = -61.8 mV, vt = -57.0 mV, c = -65.8 mV, vpeak = 22.6 mV, and khigh = 3.3 nS/mV), and set our membrane capacitance to Cm = 300 pA, which allowed us to obtain the gradual f-I slope. We then varied the parameters a, b, d, and klow to produce multiple models, and again determined the rheobase current and the slope of the initial and final f-I curve over 10 Hz (using a least squares approach) for each model. In addition, to obtain an appropriate rheobase current, we included a shift in the applied current (Iapplied → Iapplied + IShift). For our first model, we determined that a = 0.001 ms-1, b = 3 nS, klow = 0.5 nS/mV, d = 5 pA, and Ishift = –45 pA. This gave us a model f-I initial slope of 0.136, a final slope of 0.089, and a rheobase of 5 pA (see purple solid and dashed lines in Figure 4). The second weakly adapting model is identical to the first, except that we explored smaller a parameter values in order to capture the gradual slope of the final f-I curve. For this model, a = 0.00008 ms-1, which gave an initial f-I slope of 0.136, a final slope of 0.048, and a rheobase of 5 pA (see purple solid and magenta dashed lines in Figure 4). An example of the weak adaptation in this case is shown in Figure 5A, and rebound firing for model 1 is shown in Figure 5B.\n\nThe initial data points are shown as asterisks and the final points shown as squares. The initial model curve is shown as a solid line (initial slope: 0.136), and the final curves are shown as dashed lines. Model 1 exhibits less adaptation (final slope for model 1: 0.089; final slope for model 2: 0.048), but higher final frequencies than model 2 (compare purple and magenta dashed lines). The model rheobase (5 pA) and slopes approximate those determined experimentally.\n\nA: An example intracellular recording of Pyramidal cell 3 during current clamp with applied current of 154 pA (top, light green) is compared with the firing of our weakly adapting pyramidal cell models, also with an applied current of 154 pA (model 1: middle, purple; model 2: bottom, magenta). The firing rates and amount of adaptation of the model are similar to those of the experiment. B: An example of rebound firing in our weakly adapting model 1. A one-second step of 1000 pA hyperpolarizing input is applied (dashed line). The weakly adapting model produces rebound spiking (solid line), but requires a reasonably large amount of applied input. The weakly adapting model 2 does not produce rebound spiking following steps of hyperpolarizing input in the physiological range.\n\n\nDiscussion and conclusion\n\nWe have developed simple, biologically-based cellular models of pyramidal cells from CA1 hippocampus. These models capture the frequency-current profiles of both strongly and weakly adapting cells. Importantly, we have used pyramidal cell recordings from a well-defined experimental context as a basis for our model development. We note that even though other simple pyramidal cell models exist with a conductance-based biophysical representation17, there is not a clear link with adaptation or rebound firing characteristics as exists in pyramidal cells. We note that while modifications have been done to such simple models (e.g., Stark et al.18 used the Olfusen et al. model19 which incorporated sodium currents, potassium current, voltage-dependent M currents, and a leak current, and added an h-current based on20 to incorporate rebound capabilities), the resulting cellular model outputs constraints with respect to experimental data are unclear, and as such, we consider them not to be biologically-based per se.\n\nOur models do not include the multiple voltage-gated channels known to be present in pyramidal cells, and are not spatially extended by being multi-compartment in nature. Multi-compartment, biophysical models of CA1 pyramidal cells exist21 as well as single-compartment biophysical models22 with various voltage-gated channels, but the goals in developing those models are different. Here, we are interested in capturing cellular characteristics in a well-defined experimental context (for subsequent large network explorations that take advantage of the experimental context), whereas these other studies have considered, for example, conductance balances to understand how they contribute to cellular output. We further note that the richness of detail in CA1 pyramidal cells is expanding (e.g., see16 in which countermodulation by metabotropic receptors in bursting or regular spiking pyramidal cells was shown). In essence, it is always the case that the mathematical models are a limited representation of the biology.\n\nIn previous work, we used adaptation characteristics from the literature to develop simple models of CA3 pyramidal cell models, and showed that population bursting could occur in excitatory networks if the adaptation characteristics were in line with the experimental data23. In this work we had a full set of experimental recordings and so could capture appropriate cellular characteristics more directly. Although there should not be large differences in some cellular characteristics (e.g., spike widths etc.), there could be differences in characteristics such as rheobase and adaptation amounts due to varying experimental contexts (e.g., solutions, recording setup details and so on – see13–16). Coupled with biological variability, it would be additionally challenging to be clear about model limitations in subsequent model usage. Here, with our simple mathematical model representation and knowledge of the biological variability in hand, one is easily aware of any changes in the parameters of the model that would result in large deviations from the experimental data.\n\nWe note that robust fitting strategies of experimental data to simple, mathematical representations of neurons are being developed24. However, there are some differences in the modeling goals. In the paper by Hertäg et al.24, the goal was to use the developed models based on in vitro recordings to predict spiking in an in vivo context. Here, we have developed our cellular models in the experimental context for which we build the network models to determine physiologically-based mechanisms. One can consider using our models in other contexts, keeping in mind the limitations associated with the models as developed. In essence, when incorporating various cellular characteristics, one should express the choices, rationale and reasoning behind the model development, which naturally stem from the modeling goals.\n\nIn conclusion, with our simple, developed models as presented here, we can proceed to considering very large networks that include these models in this experimental context. Furthermore, our simple model representations will also allow us to take advantage of developed theoretical analyses25,26.\n\n\nData availability\n\nZENODO: Data set of CA1 pyramidal cell models using an intact whole hippocampus preparation, DOI: 10.5281/zenodo.874727", "appendix": "Author contributions\n\n\n\nKAF and FKS conceived the study. CYLH, BA, SW designed the experiments. CYLH, BA carried out the experiments. KAF carried out the modeling and data analysis. FKS and KAF prepared the first draft of the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work has been supported by the Canadian Institutes of Health Research (CIHR Grant number MOP-102573) (SW), Natural Sciences and Engineering Research Canada (NSERC Grant Number: RGPIN-203700 (FKS), and an Ontario Graduate Scholarship (KAF).\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\nMarín O: Interneuron dysfunction in psychiatric disorders. Nat Rev Neurosci. 2012; 13(2): 107–20. PubMed Abstract \| Publisher Full Text\n\nSompolinsky H: Computational neuroscience: beyond the local circuit. Curr Opin Neurobiol. 2014; 25C: xiii–xviii. PubMed Abstract \| Publisher Full Text\n\nTiesinga P, Sejnowski TJ: Cortical enlightenment: are attentional gamma oscillations driven by ING or PING? Neuron. 2009; 63(6): 727–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSkinner FK, Ferguson KA: Modeling oscillatory dynamics in brain microcircuits as a way to help uncover neurological disease mechanisms: A proposal. Chaos. 2013; 23(4): 046108. PubMed Abstract \| Publisher Full Text\n\nWang XJ: Neurophysiological and computational principles of cortical rhythms in cognition. Physiol Rev. 2010; 90(3): 1195–1268. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFerguson KA, Huh CY, Amilhon B, et al.: Experimentally constrained CA1 fast-firing parvalbumin-positive interneuron network models exhibit sharp transitions into coherent high frequency rhythms. Front Comput Neurosci. 2013; 7: 144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHo EC, Strüber M, Bartos M, et al.: Inhibitory networks of fast-spiking interneurons generate slow population activities due to excitatory fluctuations and network multistability. J Neurosci. 2012; 32(29): 9931–46. PubMed Abstract \| Publisher Full Text\n\nGoutagny R, Jackson J, Williams S: Self-generated theta oscillations in the hippocampus. Nat Neurosci. 2009; 12(12): 1491–1493. PubMed Abstract \| Publisher Full Text\n\nJackson J, Goutagny R, Williams S: Fast and slow gamma rhythms are intrinsically and independently generated in the subiculum. J Neurosci. 2011; 31(34): 12104–17. PubMed Abstract \| Publisher Full Text\n\nIzhikevich EM: Simple model of spiking neurons. IEEE Trans Neural Netw. 2003; 14(6): 1569–72. PubMed Abstract \| Publisher Full Text\n\nHuh CYL, Goutagny R, Williams S: Glutamatergic neurons of the mouse medial septum and diagonal band of Broca synaptically drive hippocampal pyramidal cells: relevance for hippocampal theta rhythm. J Neurosci. 2010; 30(47): 15951–61. PubMed Abstract \| Publisher Full Text\n\nBekkers JM, Delaney AJ: Modulation of excitability by alpha-dendrotoxin-sensitive potassium channels in neocortical pyramidal neurons. J Neurosci. 2001; 21(17): 6553–60. PubMed Abstract\n\nChapman CA, Lacaille JC: Cholinergic induction of theta-frequency oscillations in hippocampal inhibitory interneurons and pacing of pyramidal cell firing. J Neurosci. 1999; 19(19): 8637–45. PubMed Abstract\n\nSun MK, Zhao WQ, Nelson TJ, et al.: Theta rhythm of hippocampal CA1 neuron activity: gating by GABAergic synaptic depolarization. J Neurophysiol. 2001; 85(1): 269–79. PubMed Abstract\n\nFernandez FR, White JA: Gain control in CA1 pyramidal cells using changes in somatic conductance. J Neurosci. 2010; 30(1): 230–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGraves AR, Moore SJ, Bloss EB, et al.: Hippocampal pyramidal neurons comprise two distinct cell types that are countermodulated by metabotropic receptors. Neuron. 2012; 76(4): 776–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nErmentrout GB, Kopell N: Fine structure of neural spiking and synchronization in the presence of conduction delays. Proc Natl Acad Sci U S A. 1998; 95(3): 1259–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStark E, Eichler R, Roux L, et al.: Inhibition-induced theta resonance in cortical circuits. Neuron. 2013; 80(5): 1263–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOlfusen MS, Whittington MA, Camperi M, et al.: New roles for the gamma rhythm: population tuning and preprocessing for the Beta rhythm. J Comput Neurosci. 2003; 14(1): 33–54. PubMed Abstract \| Publisher Full Text\n\nZemankovics R, Káli S, Paulsen O, et al.: Differences in subthreshold resonance of hippocampal pyramidal cells and interneurons: the role of h-current and passive membrane characteristics. J Physiol. 2010; 588(Pt 12): 2109–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoirazi P, Pissadaki EK: The making of a detailed CA1 pyramidal neuron model. In: Cutsuridis V, Graham B, Cobb S, Vida I (eds) Hippocampal microcircuits. Springer, New York, 2010; 317–352. Publisher Full Text\n\nNowacki J, Osinga HM, Brown JT, et al.: A unified model of CA1/3 pyramidal cells: An investigation into excitability. Prog Biophys Mol Biol. 2011; 105(1–2): 34–48. PubMed Abstract \| Publisher Full Text\n\nDur-E-Ahmad M, Nicola W, Campbell SA, et al.: Network bursting using experimentally constrained single compartment CA3 hippocampal neuron models with adaptation. J Comput Neurosci. 2011; 33(1): 21–40. PubMed Abstract \| Publisher Full Text\n\nHertäg L, Hass J, Golovko T, et al.: An approximation to the adaptive exponential integrate-and-fire neuron model allows fast and predictive fitting to physiological data. Front Comput Neurosci. 2012; 6: 62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNicola W, Campbell SA: Mean-field models for heterogeneous networks of two-dimensional integrate and fire neurons. Front Comput Neurosci. 2013; 7: 184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNicola W, Campbell SA: Bifurcations of large networks of two-dimensional integrate and fire neurons. J Comput Neurosci. 2013; 35(1): 87–108. PubMed Abstract \| Publisher Full Text\n\nFerguson KA, Huh CY, Amilhon B, et al.: Dataset of CA1 pyramidal cell models using an intact whole hippocampus preparation. 2014. Data Source" }	[ { "id": "4728", "date": "05 Jun 2014", "name": "Farzan Nadim", "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\nA simple mathematical yet biological model of hippocampal CA1 pyramidal cell has been developed to capture the cellular characteristics (f-I curve, rheobase current, firing rate adaptation and rebound spiking) of those neurons in an intact preparation of hippocampus. Although this study does not examine any biological question or hypothesis directly, the simplified models of CA1 pyramidal cells, in conjunction with a similar model of the fast spiking interneurons previously developed by this lab, can be used in future modeling studies that explore the network properties in the hippocampus. Comments:First paragraph of the Introduction:\"To address this recognized difficulty we are developing models that are based on well-defined experimental contexts in which both the cellular and the network aspects of the models can be considered simultaneously.\" Please specify more clearly about the network aspects, you are referring to. First paragraph of Introduction:\"Using such models, we aim to help determine, predict and test biologically based mechanisms.\"This is a bit misleading. In this work, the model only replicates biological behavior of neurons. There is no clear prediction of any biological phenomenon. End of the second paragraph of Introduction:\"we can build and examine several large network models that are aligned with the biology. Subsequent analyses of these large network models could determine the mechanisms by which particular cellular characteristics critically contribute to the population activities observed in our experimental context.\" It would be more exciting to construct network models and make biological predictions. Construction of simple cellular models to capture biological characteristics of individual neurons is not novel. In fact, the authors have used similar modeling approach to develop models in references 6 and 23. Second paragraph of Results - Stronger adaptation vs weaker adaptation - Has it been shown before? If yes, please include some references here. If not, I am wondering if it would be worthwhile to check the adaptation by comparing steady-state ISIs. Maybe different neurons have different time course of adaptation but at the end they show same level of adaptation. If we run the model for longer duration, how does the behavior change, does it continue to adapt or it reaches a steady-state? Second paragraph of Strongly adapting model:\"If these properties (f-I curve, rheobase current and adaptation) are captured, then the model will spike with similar frequencies as the physiological cell given the same amount of synaptic input.\" Is the applied protocol (continuous injections of currents for 1s) sufficient to characterize the response of these neurons in the presence of synaptic inputs? Second paragraph of Weakly adapting models:\"In addition, to obtain an appropriate rheobase current, we included a short in the applied current.\" I do not understand why Ishift term is included in Iapplied to adjust rheobase? Third paragraph of Discussion and conclusion:\"In previous work, we used adaptation characteristics from the literature to develop simple models of CA3 pyramidal cell models, and showed that population characteristics were in line with the experimental data.\" It would be interesting and more insightful to see how this work can be extended to network models as has been done previously (in references 6 and 23) by the authors. Third paragraph of Discussion and conclusion: \"Here, with our simple mathematical model representation and knowledge of the biological variability in hand, one is easily aware of any changes in the parameters of the model that would result in large deviations from the experimental data.\" This is very important aspect. But that would only be possible if the effects of parameters can be characterized. It would be interesting to see any such generalizations. Fourth paragraph of Discussion and conclusion:\"Here, we have developed out cellular models in the experimental context for which we build the network models to determine physiologically-based mechanisms.\" Building network models incorporating CA1 pyramidal cells and determining physiologically-based mechanisms is clearly not presented in this paper and should be indicated as future work. In Figure 3, please also show rebound firing of recorded cells, if available. In Figure 2, the label pointing to green dotted line is misspelled. It should be final curve instead of \"finial curve\".", "responses": [ { "c_id": "878", "date": "24 Jun 2014", "name": "Frances Skinner", "role": "Author Response", "response": "We would like to thank the reviewers for their helpful comments. We will respond more specifically and revise our article accordingly in the near future. We would like to point out that we focused on the single cell model and did not try to also include network simulations in this article as it was originally submitted as a 'Short Research Article' type. However, F1000Research recently replaced the 'Short Research Article' category with a shorter 'Research Notes' article type. As this article did not fit the criteria of a 'Research Note', it appears as a 'Research Article' instead." }, { "c_id": "1366", "date": "20 May 2015", "name": "Frances Skinner", "role": "Author Response", "response": "1-2: We have revised manuscript to make these points clear. 3: We agree. As mentioned in our more immediate response, this ‘short report’ was to present the cellular model. Network simulations using these models have been done (presented at CNS and SfN meetings), and the manuscripts are in preparation. These abstract references have been added. 4: Although it is clear that spike frequency adaptation can occur in CA1 pyramidal cells (as in additional figure and references), to the best of our knowledge, a specific distinction between strongly and weakly adapting has not been described in the literature for CA1 pyramidal cells, although such distinctions have been reported for CA3 pyramidal cells (Hemond et al., 2008). This is a challenging issue to be clear and distinct about due to both experimental and modeling limitations since the biophysical basis of spike frequency adaptation itself is not completely clear. For example, “… In most studies, however, the initial, steady-state, or mean firing frequency were not controlled…” (from Discussion in Fernandez and White, 2010). They were referring to IM current, but this points to the potential importance of details in different experimental contexts when comparing particular features. So, examination of the biophysical basis underlying adaptation have been explored (e.g., see Discussion in Fernandez and White 2010 ref as cited from above), and also, Ascoli et al. (2010) has shown that rebound is rarely observed in physiological conditions, and it can be unmasked by blocking A-type currents.Our intention here is not to say that there are two distinct types of adapting CA1 pyramidal models per se. For our models here, we aimed to be clear about how they were ‘biologically-based’ using our given context based on f-I curves, resting potential etc., but do not have a biophysical basis in terms of voltage-gated channels etc. Strongly and weakly adapting characteristics were observed (and modeled), but we think that it would be premature to suggest that this is a distinct (biological) separation without trying to dissect out the biophysical basis, doing more recordings of longer durations and so on. For our work here, the goal was to develop models based on a clear experimental context.We agree that this whole aspect needs to be explored fully and in more detail in future work, but we also think that it is important that this be done without separation of experiment and model details and context, and goals. 5: The model has reached a steady-state by the end of the first second, and thus this steady-state is reflected in our model final f-I curve. 6: We agree that the applied protocol used is not sufficient to characterize response to synaptic input. We have revised the particular sentence so that it is clear that we simply mean ‘to a first approximation’ (and would only mean ‘tonic’ input of the level of synaptic input). 7: Due to the simplicity of the model, it was difficult to capture the rheobase as well as the initial and final frequency current profiles in the weakly adapting model, and thus a term was needed to shift the curve laterally (Ishift). However, your comment brought to light that this shift would be more appropriately represented by an independent term directly affecting equation (1), rather than influencing the Iapplied current. This change in the representation is now reflected in the equation and description. 8: Yes, we agree. See point 3 above. 9: Yes, we agree and have added a couple of sentences to the Discussion in this regard. 10: We have revised that sentence accordingly and refer to abstracts of ongoing work. 11: Experimental data showing rebound firing is now included. 12: Fixed." } ] }, { "id": "4729", "date": "24 Jun 2014", "name": "Alla Borisyuk", "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 article presents minimal models for CA1 pyramidal cells, based on fitting the frequency-current curves that had been experimentally recorded in whole hippocampus preparation.There are two types of models, differing in some parameter values, to account for “strongly” and “weakly” adapting cell varieties. The goal of the present work is to present the models and their basic fit to single cell data, in preparation for future large network studies.Comments:I think this work would be much stronger if testable predictions were generated from the models. For example, what would be the responses to time varying inputs and how the responses would differ between the strongly and weakly adapting cells? It is not clear to me that obtaining one model for each of variety of cells is sufficient. For example, two strongly adapting cells have very different rheobase current values. If you fit the model to each cell individually, I expect that you will end up with two very different parameter sets. What would be the consequences for responses to time varying inputs, for example? Considerations like these will be very important when you decide how much variability and in which parameters to include in the network model. This is also related to the fact that a single model was not able to capture all desired properties of the weakly adapting model. I would like to see more discussion on the consequences of the spectrum of models for future network work.", "responses": [ { "c_id": "879", "date": "24 Jun 2014", "name": "Frances Skinner", "role": "Author Response", "response": "We would like to thank the reviewer for their helpful comments, which we will be taking into consideration. We would like to point out that we focused on the single cell model and did not try to also include network simulations in this article as it was originally submitted as a 'Short Research Article' type. However, F1000Research recently replaced the 'Short Research Article' category with a shorter 'Research Notes' article type. As this article did not fit the criteria of a 'Research Note', it appears as a 'Research Article' instead." }, { "c_id": "1365", "date": "20 May 2015", "name": "Frances Skinner", "role": "Author Response", "response": "We agree that it is may not sufficient to have one model for each of the variety of cells, and that it would be good to consider consequences of time varying input.However, as mentioned in our more immediate response, this work was to present the cellular model together with the experimental data. We were not aiming for testable predictions from the cellular models on their own, but we are aiming to use them in network models to help determine mechanisms in the biological system (see added abstract references), by focusing our models on the experimental context so that there can be a back and forth examination between model and experiment, given that the model will always be limited and the experimental details can vary between different contexts. Since we have a clear sense of the rationale and limitations of the cellular model, model network mechanistic insights would be viewed through those lens and help determine what modeling and experimental work should be further examined. For the time varying input aspect, this was raised (in a different way) by the other reviewers (see their point 6), and we have revised the paper accordingly. In essence, we see the fits as trying to match the experimental data to a first approximation.For the issue of different weakly and strongly adapting characteristics, please see our response to the other reviewers (point 4)." } ] } ]	1	https://f1000research.com/articles/3-104
https://f1000research.com/articles/4-139/v1	01 Jun 15	{ "type": "Research Article", "title": "A resistant-starch enriched yogurt: fermentability, sensory characteristics, and a pilot study in children", "authors": [ "Kayanush Aryana", "Frank Greenway", "Nikhil Dhurandhar", "Richard Tulley", "John Finley", "Michael Keenan", "Roy Martin", "Christine Pelkman", "Douglas Olson", "Jolene Zheng", "Kayanush Aryana", "Nikhil Dhurandhar", "Richard Tulley", "John Finley", "Michael Keenan", "Roy Martin", "Christine Pelkman", "Douglas Olson" ], "abstract": "The rising prevalence of obesity and the vulnerability of the pediatric age group have highlighted the critical need for a careful consideration of effective, safe, remedial and preventive dietary interventions. Amylose starch (RS2) from high-amylose maize (HAM) ferments in the gut and affects body weight.\n\nOne hundred and ten children, of 7-8 (n=91) or 13-14 (n=19) years of age scored the sensory qualities of a yogurt supplemented with either HAM-RS2 or an amylopectin starch. The amylopectin starch yogurt was preferred to the HAM-RS2-enriched yogurt by 7-8 year old panelists (P<0.0001). Appearance, taste, and sandiness scores given by 13- to 14-year-old panelists were more favorable for the amylopectin starch yogurt than for HAM-RS2-enriched yogurt (P<0.05). HAM-RS2 supplementation resulted in acceptable (≥6 on a 1-9 scale) sensory and hedonic ratings of the yogurt in 74% of subjects. Four children consumed a HAM-RS2-enriched yogurt for four weeks to test its fermentability in a clinical trial. Three adolescents, but not the single pre-pubertal child, had reduced stool pH (P=0.1) and increased stool short-chain fatty acids (SCFAs) (P<0.05) including increased fecal acetate (P=0.02), and butyrate (P=0.089) from resistant starch (RS) fermentation and isobutyrate (P=0.01) from protein fermentation post-treatment suggesting a favorable change to the gut microbiota. HAM-RS2 was not modified by pasteurization of the yogurt, and may be a palatable way to increase fiber intake and stimulate colonic fermentation in adolescents. Future studies are planned to determine the concentration of HAM-RS2 that offers the optimal safe and effective strategy to prevent excessive fat gain in children.", "keywords": [ "fatty acids", "fiber", "home nutrition support" ], "content": "Clinical relevancy statement\n\nResistant starch (RS) is a type of dietary fiber that people cannot digest, diluting caloric density, but is fermented by bacteria in the intestines into short chain fatty acids that have been shown in other studies to stimulate the production of appetite reducing hormones (see the text). We incorporated resistant starch into a yogurt that was generally accepted by children, increased their dietary fiber consumption and increased colonic fermentation in adolescents. This pilot data suggest the need for a study testing the ability of this yogurt to treat childhood obesity, a vulnerable group where non-food solutions are limited.\n\n\nIntroduction\n\nThe rapidly-growing prevalence of obesity in adults and children requires urgent remedial measures to avert individual and societal health care crises1. Few adult treatment strategies exist2, treating children is more challenging, yet childhood obesity is a growing health concern3. Pediatric vulnerability severely limits the use of pharmacological or surgical interventions. Even dietary treatment with energy and nutrient restriction for weight reduction may be detrimental to growth. Dietary resistant starch (RS) supplementation in food may offer a therapeutic opportunity to attenuate excessive fat gain in infants and children by reducing the caloric density while improving dietary quality1,2.\n\nRS are dietary carbohydrates that resist cooking processes and enzymatic digestion in the small intestine, are fermented by colonic microbiota and modify the gut flora4,5. The amount of RS in the human diet has progressively decreased with modern milling and food preparation methods. RS intake in medieval Europe was 50–100g/day6, it is estimated at 30–40g/day7 in developing countries, and has dropped to 3–8g/day in developed countries7–9. It is unlikely that modern human society will return to a diet of coarsely ground grains and legumes high in RS. However, RS is now available as an ingredient that can be incorporated into breads, cereal products and baked goods that are acceptable to the US population.\n\nMicrobiota-derived enzymes are needed to digest complex plant polysaccharides10 and RS-enriched diets increase butyrate-producing Clostridia in rodent feces11. A natural, granular, type 2 RS from high-amylose maize (HAM-RS2) decreases plasma cholesterol and triglycerides, increases satiety, increases insulin sensitivity12–22, and is anti-adipogenic in adult populations23–30. HAM-RS2 fermentation in the colon of rodents produces short chain fatty acids (SCFAs) such as acetate, propionate and butyrate that are absorbed through colonocytes, and change colonic microbiota composition25,31,32. Butyrate treatment increases gene expression of peptide tyrosine tyrosine (PYY) and proglucagon in ileal, primary colon and cecal epithelial cells of rats; elevates plasma Glucagon-like peptide-1 and -2 (GLP-1, GLP-2), and raises gene expression and protein production of Glucose transporter 2 (GLUT2)27,31,33–35. Clinical studies show that SCFAs increase in response to consumption of HAM-RS2 or RS from potatoes36–38, that the microbiota of rodents were modified39, and that butyrate was increased in rodents after dietary introduction of human feces40. Since yogurt can deliver dietary fibers to treat constipation in children41, the aim of the current work is to develop a palatable yogurt delivery vehicle for HAM-RS2 that will withstand pasteurization and demonstrate an increase in fecal pH and SCFAs in children and adolescents.\n\n\nMethods\n\nYogurt mixes were made by incorporating the starches individually into skim milk. The yogurt mixes were pasteurized at 65.5°C for 30 min, cooled to 40°C, inoculated with freshly thawed Streptococcus thermophilus (ST-M5) (3.1E+10 cfu/g, 1ml) and Lactobacillus bulgaricus (LB-12) (3E+10 cfu/g, 1ml) (Chr. Hansen Inc., Milwaukee, WI) per 3.785L (1 gallon), then incubated at 40°C until they reached a pH 4.5, and held at 4°C overnight. Blueberry puree (20% w/w) was incorporated into the yogurt the following day and amylopectin starch (15 g, control, AMIOCA® corn starch, Ingredion Incorporated, Bridgewater, NJ) or HAM-RS2 (15g, HI-MAIZE® 260 resistant starch, Ingredion Incorporated, Bridgewater, NJ) per 237ml serving was added to the yogurt (Creamery, College of Agriculture, LSU). A high performance liquid chromatography (HPLC) peak was detected in our HAM-RS2 sample and RS accounted for 38.2% of the sample.\n\nHAM-RS2 30g/237ml yogurt was used for in vitro testing. Six samples were prepared, coded, and tested blindly with half subjected to pasteurization. A modified Englyst method was used to quantify glucose release4. Intact granular structure of the starch was evaluated using birefringence light microscopy.\n\nThe Institutional Review Board (IRB) granted an exemption #HE13-1 (January 16, 2013) from continued oversight for the sensory study conducted in two groups of children evaluating the two yogurts. Ratings of satisfaction with the appearance, color, aroma, taste, thickness, sandiness, and palatability of each type of yogurt were scored by 110 children without communication. Ninety-one children were 7–8 years old (younger) and 19 were 13–14 years old (elder). The younger children were more willing to volunteer for the sensory study than the elder.\n\nSubjects with no dairy or starch-related allergies were recruited from The Louisiana State University Laboratory School and parental consent to participate was obtained along with the children’s assent. Participants were given yogurt samples in 85g cups with a snap-on lid. Cups were coded with a random three-digit number. Disposable plastic spoons and napkins were provided to prevent contamination between samples. Prior to the sensory evaluation, the children were provided with a “warm-up” yogurt sample to avoid the “first sample effect” due to possible previous consumption of other food items, and a cup of drinking water was provided to rinse their palate between samplings. Two evaluation forms were used, one with a face scale for the younger panelists and the other with a preference rating form for the older panelists, and clearly explained to each age group. The younger panelists indicated their yogurt preference by circling “smiling face (☺) as yes”, scored as 3, “neutral face (😐) as neither like nor dislike”, scored as 2, or “sad face (☹) as no”, scored as 1. The elder panelists evaluated the yogurt on a 1–9 scale (1-dislike extremely, 2-dislike very much, 3-dislike moderately, 4-dislike slightly, 5-neither like nor dislike, 6-like slightly, 7-like moderately, 8-like very much, 9-like extremely) for appearance, color, aroma, taste, thickness, and sandiness. The elder panelists evaluated the yogurt thickness by checking 1-too thin, 2-just about right, or 3-too thick; and the sandiness as 1-not grainy, 2-just about right, or 3-too grainy. Elder panelists answered the question “Is this product acceptable?” with 2-“yes” or 1-“no” answer.\n\nThe four-week pilot clinical trial was approved by the PBRC Institutional Review Board (IRB28012) and registered (http://clinicaltrials.gov/, NCT01338571) to determine the effects of consuming HAM-RS2-enriched yogurt on fecal pH and fecal SCFAs, pre- and post-consumption, in a healthy child and three healthy adolescents. The subjects (a 6-year-old female, two 10-year-old African-American females, and a 14-year-old Caucasian male) were recruited through the PBRC recruiting department. Parents signed a consent form and subjects signed an assent form. Subjects with gastrointestinal disease, on medications with the potential to alter the intestinal bacterial microbiota such as antibiotics and subjects with allergies to corn were specifically excluded.\n\nSubjects were weighed in the morning on an electronic scale in light street clothing without shoes or outer clothing and with pockets emptied. The electronic scale (Model 450, GSE Inc., Livonia, MI, USA) was calibrated daily using standardized weights and quarterly by an external service. Parents were given stool-collecting kits and instructed to collect a stool specimen from their child for 3 consecutive days at baseline and after 4 weeks of yogurt consumption. They were provided with ice packs, coolers and were instructed to return stool samples in the coolers to the research site on the day they were collected so they could be stored at -70°C until analysis.\n\nChildren were given HAM-RS2 10g plus 1g per year of age daily42 which was 16, 20, or 24g for the four subjects. A fresh supply of yogurt was given to the parents weekly and the daily yogurt was divided into servings at breakfast and dinner.\n\nMeasurements of fecal SCFAs and pH were previously published elsewhere18. Briefly, the frozen fecal specimens were thawed, homogenized and further diluted to wet sample in distilled water (0.5g/5ml). The pH was measured using a combination electrode. Samples were then acidified with metaphosphoric acid (250g/L, 1ml) containing ethyl-butyric acid (2g/L) as an internal standard. The mixture was vortexed, and centrifuged at 4°C for 10 minutes at 8,000 rpm to remove solids in the homogenized samples and syringe-filtered (33mm, Millipore, Billerica, MA). The filtrate was put into a gas chromatograph (GC) auto-sampler vial and capped. SCFAs in the effluent were analyzed using gas-liquid chromatography. The GC conditions (115°C for 0.1 min) were increased to 150°C for 0.1 min in increments of 10°C, then to 170°C for 2 min at increments of 11°C. The injector temperature was 250°C. Helium was the carrier gas with a flow rate of 60 ml/min and splitless injection was 60 ml/min. Single SCFAs were determined by retention time based on standards and the relative concentrations calculated based on the ratio of the peak areas of the sample to the internal standard.\n\nThe RS content differences of yogurt samples prepared with or without pasteurization were determined using the Student t-test (SAS 9.3, SAS Institute Inc., Cary, NC). The sensory data obtained from children were analyzed with a Randomized Block Design using panelists as blocks (GLM, SAS 9.1), and the paired t-test for HAM-RS2 score minus amylopectin score was performed. Differences between the types of yogurt were determined by differences of least squares mean ± SEM. The clinical data analysis of feeding yogurt to the four subjects was performed with the Student t-test (weight) and paired t-test (change in weight) (SAS 9.1). Alpha was set at 0.05.\n\nAll procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000.\n\n\nResults\n\nThe glucose release was detected by a modified Englyst method. A light microscope (200x, Leitz Wetzlar, Ortholux II, Ernst Leitz GmbH, Wetzlar, Germany) revealed morphologies of the starch granules in the yogurts as having an equal presence of birefringence indicating an intact granular structure (Figure 1). The RS content of the six yogurt samples varied minimally (from 45% to 51% on a dry weight basis with values of 51%, 45%, or 48% for the unpasteurized samples, and 45%, 45%, or 46% for the pasteurized samples (P>0.05).\n\nBar=30μm.\n\nGroup 1: Ninety-one 7- and 8-year-old panelists. The average scores were 1.538 for the HAM-RS2-yogurt and 2.143 for the amylopectin starch yogurt. The difference was -0.604±0.1 (t=6.05, P<0.0001) which indicated that the amylopectin starch yogurt was preferred over the HAM-RS2-yogurt.\n\nGroup 2: Nineteen 13- and 14-year-old panelists. No score differences were detected in color (6.95±0.36 vs. 7.05±0.28) and aroma (7.84±0.24 vs. 7.47±0.32) for the amylopectin starch compared to the HAM-RS2-yogurt, respectively (Table 1, P>0.05). However, appearance (6.84±0.34 vs. 4.58±0.38), taste (6.95±0.32 vs. 4.84±0.49), thickness (6.74±0.48 vs. 4.47±0.37), and sandiness (6.26±0.37 vs. 3.05±0.36) scores for the amylopectin starch yogurt were higher than for the HAM-RS2-yogurt (P<0.005).\n\nabMeans without a common superscript are significantly (P<0.05) different from each other.\n\nOn a 1–3 scale (1-too thin, 2-just about right, or 3-too thick), the amylopectin starch yogurt (2.26±0.13) was judged slightly thicker than just about right while the HAM-RS2-yogurt (1.16±0.09) was judged as too thin (P<0.0001, Table 2).\n\nabMeans without a common superscript are significantly (P<0.05) different from each other.\n\nUsing a 1–3 scale (1-not grainy, 2-just about right, and 3-too grainy), the HAM-RS2-yogurt (2.84±0.12) was judged as too grainy but was acceptable to 74% of the children of 13–14 years of age, while the amylopectin starch yogurt (1.95±0.12) was judged as just about right (Table 2, P<0.0001).\n\nThe amylopectin starch yogurt was always judged as acceptable (Table 2) and its acceptability on a 1–2 scale (1-not acceptable or 2-acceptable) was significantly higher than for the HAM-RS2-yogurt (P<0.05). The sensory study indicated that children preferred the amylopectin starch yogurt more than the HAM-RS2 added yogurt.\n\nAll adolescent participants finished the HAM-RS2-yogurt and returned the empty containers during the weekly clinic visits with no complaints regarding taste or compliance issues related to consumption of the yogurt.\n\nOne 10-year-old had a BMI of 19.8 kg/m2 and the other had a BMI of 27.1 kg/m2. The 14-year-old had a BMI of 31.5 kg/m2. All were otherwise healthy. The pre-pubertal child gained 1.9kg (39.5 to 41.4 kg). One of the adolescent females gained 3.2kg (49.6 to 52.8kg), the other one gained 0.4kg (69.5 to 69.9kg) and the adolescent male gained 1.7kg (89.9 to 91.6 kg) (P>0.05).\n\nSCFAs (μg/g wet stool weight) from carbohydrate fermentation were increased in the adolescent participants; in ascending order, butyrate (23%, 2,410±691 to 3,144±1,509µg, P=0.09), acetate (26%, 5,078±492 to 6,870±515µg, P=0.02), but not propionate (2,387±645 to 1,889±120µg, P>0.05). The isobutyrate from protein fermentation increased (39%, 285±31 to 471±58μg, P=0.01) (Figure 2). The stool pH of the adolescents was mildly reduced at the end of the fourth week with a trend toward a lower pH (2.8%, from 7.2±0.4 to 7.0±0.35, P=0.1, Figure 3).\n\nThe pre-pubertal child was not included.\n\nThe pH was increased in the pre-pubertal child.\n\nThe pre-pubertal participant responded to HAM-RS2-enriched-yogurt differently than the three adolescent children with an increase in stool pH (from 6.89 to 7.62). The stool SCFAs were decreased; in ascending order, isobutyrate (35%, from 526 to 186µg), butyrate (39%, from 4,028 to 1,571µg), acetate (52%, from 8,328 to 4,336µg), and propionate (65%, from 2,870 to 1,877µg) over the 4-week study.\n\n\nDiscussion\n\nThe prevalence of childhood obesity has increased 2- to 3-fold in just the last 25 years globally (see review 2). Childhood obesity is associated with co-morbidities similar to adults including hypertension, dysglycemia, dyslipidemia, inflammation and endothelial dysfunction2. Supplementing other treatment approaches with behavioral interventions may increase long term participation and is felt to be more important in the pediatric population than for adults2, but meta-analyses show prevention or treatment strategies to be ineffective. Currently, no truly effective pharmacological options are available for weight management, and surgery is restricted to a highly selected subgroup of very obese adolescent individuals. Medication and surgery have safety concerns in growing children and their efficacy is uncertain in the pediatric age group43. Novel treatments for childhood obesity offering safety, efficacy and acceptability are urgently needed44. Desirable attributes of an intervention for pediatric obesity include a preventive measure that attenuates excess fat accumulation while allowing for normal growth. RS is a natural food ingredient with a low risk profile that attenuates body fat accretion in experimental animal models, and is an excellent candidate to effectively combat childhood obesity. This feasibility study suggests that HAM-RS2-enriched foods likely alter microbiota composition, and this is supported by the increase in fecal SCFA content and lower pH. Yogurt was a generally acceptable vehicle for providing HAM-RS2. The yogurt cultures fermented lactose (milk sugar) and the RS granules in the final yogurt product were not damaged.\n\nEnriching the diet with RS which has been refined out of the US diet will improve dietary quality and may help to ease the severity of pediatric obesity. Although the amylopectin starch yogurt was preferred, our studies confirmed the general acceptability of incorporating HAM-RS2 into yogurt through taste and sensory testing in 91 7- to 8-year-olds and 19 13- to 14-year-old volunteers. The four subjects in our pilot study that consumed the HAM-RS2-enriched yogurt twice a day for weeks established the feasibility of feeding the HAM-RS2-enriched yogurt to children. We demonstrated a trend toward a reduction of pH and documented a significant increase in the SCFA content of the stools of the adolescent children. This agrees with previous studies that have found that adding HAM-RS2 to rodent diets reduced abdominal fat in association with increased fermentation20,21,37,44. Supplementing the diet with RS will need to be acceptable and palatable or children are likely to reject it in favor of low-fiber alternatives. Overall, the HAM-RS2-yogurt in the taste testing was acceptable to 74% of the children in the 13- to 14-year-old group, but 24% less acceptable in younger children. The knowledge that RS is healthy may increase the adoption of RS fortified foods, such as yogurt.\n\nPeople eat for volume and consume fewer calories when food has a lower caloric density38. RS and other dietary fibers reduce the caloric density of food12,13,44. RS is present in many different sources, which offers the opportunity to choose the RS with the greatest success in reducing or controlling body weight17. We have previously shown that the HAM-RS2 supplementation produces a 30% reduction in intestinal fat deposition in wild type C. elegans 28 and in rodents the same also reduced body fat25,44. Longer-term controlled studies are needed to determine if the reduced adiposity seen in animal models will occur in human populations. In a human pilot study, a HAM-RS2 (15g/day) supplemented diet enhanced insulin sensitivity by 56.5% in men over two to three months, which suggests that lower amounts of HAM-RS2 may also be efficacious. Beneficial changes in adiposity may occur over longer treatment periods12,13,21,22, and lower amounts of RS may further improve palatability – an important factor for long-term consumption.\n\nChildren maintain weight loss better than adults11. Although it is not clear why the pre-pubertal child in this study did not respond in the same way as the adolescents, it could represent differences in her intestinal microbiota or her pre-treatment diet which was not controlled nor queried. Further research will be necessary to explore the differential role of diet and the intestinal microbiota on the fermentation of RS before puberty. Weight gain in all of the children during the 4-week study may reflect the fact that they were growing.\n\n\nConclusion\n\nThe current study showed the acceptability and feasibility of using yogurt to deliver RS to adolescents which caused a change in SCFA and probably changed the gut microbiota. These preliminary data suggest the need to evaluate differences that may exist in the microbiota before and after puberty to determine whether the non-response of the pre-pubertal child represented an outlier or a real effect in pre-pubertal children. These preliminary results will need confirmation in a controlled trial so that the effects of growth can be taken into account in evaluating weight changes in longer-term studies using yogurt as a vehicle to deliver the functional food component HAM-RS2 in a range of doses into everyday foods that consumers enjoy. Our data encourage controlled studies in children and adolescents testing insulin sensitivity, effects on body weight, and potential differences between pre-pubertal and adolescent children in their microbiota response to RS. Hopefully, increased consumption of reduced-calorie foods in combination with increased physical activity12 will reduce weight gain, help to maintain a healthier weight, and lead to future improvements in public health for adults and adolescents alike.\n\n\nData availability\n\nF1000Research: Dataset 1. Sensory raw data from study participants, 10.5256/f1000research.6451.d4791845\n\nF1000Research: Dataset 2. Raw data for clinical study, 10.5256/f1000research.6451.d4800446\n\n\nConsent\n\nInformed consent was obtained from all patients being included in the study.", "appendix": "Author contributions\n\n\n\nEach individual has contributed to this manuscript as a qualified author and meets ALL of the requirements following the American Medical Association (AMA) manual guidelines. KJ Aryana, C Pelkman, and D Olson contributed to the conception and design of sensory study, acquisition of data, analysis and interpretation of the data, and drafting or editing the manuscript. R Tulley performed stool test and the SCFA analysis. FL Greenway and NV Dhurandhar designed the clinical study, contributed to its conception, acquisition of data, analysis and interpretation of the data, and drafting the manuscript. JW Finley, MJ Keenan, RJ Martin, and J Zheng contributed to designing of the sensory study, data analysis and interpretation, and drafting or editing the manuscript. All authors have approved the final version of the manuscript.\n\n\nCompeting interests\n\n\n\nC. Pelkman is an employee of Ingredion Incorporated. M. Keenan and R. Martin received grant support from Ingredion Incorporated. No competing interests were disclosed for other authors.\n\n\nGrant information\n\nThis project was funded in part by Louisiana State University Agricultural Center (LAB 93724) and the Pennington Biomedical Research Center. This work was partially supported by a NORC Center Grant (# 2P30DK072476) entitled “Nutritional Programming: Environmental and Molecular Interactions” sponsored by NIDDK. This work was supported in part by P50AT002776 from the National Center for Complementary and Alternative Medicine (NCCAM) and the Office of Dietary Supplements (ODS) which funds the Botanical Research Center of Pennington Biomedical Research Center and the Department of Plant Biology and Pathology in the School of Environmental and Biological Sciences (SEBS) of Rutgers University. This work was supported in part by 1 U54 GM104940 from the National Institute of General Medical Sciences of the National Institutes of Health which funds the Louisiana Clinical and Translational Science Center.\n\n\nAcknowledgements\n\nThis paper was approved for publication by the Director of Louisiana Agricultural Experiment Station as manuscript No. 2012-237-7031. High-amylose maize (HI-MAIZE 260 resistant starch) and AMIOCA corn starch were donated by Ingredion Incorporated. (Bridgeport, NJ). The authors would like to thank Dr. Danuta Janik for her assistance on the light microscopy.\n\n\nReferences\n\nAnderson JW, Patterson K: Snack foods: comparing nutrition values of excellent choices and \"junk foods\". J Am Coll Nutr. 2005; 24(3): 155–156. PubMed Abstract \| Publisher Full Text\n\nEbbeling CB, Pawlak DB, Ludwig DS: Childhood obesity: public-health crisis, common sense cure. Lancet. 2002; 360(9331): 473–482. PubMed Abstract \| Publisher Full Text\n\nDurant N, Baskin ML, Thomas O, et al.: School-based obesity treatment and prevention programs: all in all, just another brick in the wall? Int J Obes (Lond). 2008; 32(12): 1747–1751. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEnglyst HN, Kingman SM, Cummings JH: Classification and measurement of nutritionally important starch fractions. Eur J Clin Nutr. 1992; 46(Suppl 2): S33–S50. PubMed Abstract\n\nMahadevamma S, Shamala TR, Tharanathan RN: Resistant starch derived from processed legumes: in vitro and in vivo fermentation characteristics. Int J Food Sci Nutr. 2004; 55(5): 399–405. PubMed Abstract \| Publisher Full Text\n\nBirket A, Marie A: Dietary resistant starch interacts with non-starch polysaccharide and protein to influence colonic protein fermentation, with possible implication for colon cancer risk in humans. [Ph.D. thesis], Deakin University, Australia. 1997. Reference Source\n\nBaghurst KI, Record SJ, Baghurst PA: Dietary fiber, non-starch polysaccharide and resistant starch intakes in Australia. In: Spiller G ed. CRC Handbook of dietary Fiber in Human Health. Boca Raton, Florida.: CRC Press LLC. 2001; 583–591. Publisher Full Text\n\nDyssler PH: Estimation of resistant starch intake in Europe. In: van Amelsvoort JMH, ed. Wagneningen, The Netherlands.: EURESTA, N-G Asp.; 1994; 84–86.\n\nBrighenti F, Casiraghi MC, Baggio C: Resistant starch in the Italian diet. Br J Nutr. 1998; 80(4): 333–341. PubMed Abstract\n\nFlint HJ, Bayer EA, Rincon MT, et al.: Polysaccharide utilization by gut bacteria: potential for new insights from genomic analysis. Nat Rev Microbiol. 2008; 6(2): 121–131. PubMed Abstract \| Publisher Full Text\n\nSharp R, Macfarlane GT: Chemostat enrichments of human feces with resistant starch are selective for adherent butyrate-producing clostridia at high dilution rates. Appl Environ Microbiol. 2000; 66(10): 4212–4221. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobertson MD, Bickerton AS, Dennis AL, et al.: Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am J Clin Nutr. 2005; 82(3): 559–567. PubMed Abstract\n\nRobertson MD, Currie JM, Morgan LM, et al.: Prior short-term consumption of resistant starch enhances postprandial insulin sensitivity in healthy subjects. Diabetologia. 2003; 46(5): 659–665. PubMed Abstract \| Publisher Full Text\n\nBrown IL: Applications and uses of resistant starch. J AOAC Int. 2004; 87(3): 727–732. PubMed Abstract\n\nCassidy A, Bingham SA, Cummings JH: Starch intake and colorectal cancer risk: an international comparison. Br J Cancer. 1994; 69(5): 937–942. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Deckere EA, Kloots WJ, van Amelsvoort JM: Resistant starch decreases serum total cholesterol and triacylglycerol concentrations in rats. J Nutr. 1993; 123(12): 2142–2151. PubMed Abstract\n\nHiggins JA: Resistant starch: metabolic effects and potential health benefits. J AOAC Int. 2004; 87(3): 761–768. PubMed Abstract\n\nKendall CW, Esfahani A, Hoffman AJ, et al.: Effect of novel maize-based dietary fibers on postprandial glycemia and insulinemia. J Am Coll Nutr. 2008; 27(6): 711–718. PubMed Abstract \| Publisher Full Text\n\nAnderson GH, Cho CE, Akhavan T, et al.: Relation between estimates of cornstarch digestibility by the Englyst in vitro method and glycemic response, subjective appetite, and short-term food intake in young men. Am J Clin Nutr. 2010; 91(4): 932–939. PubMed Abstract \| Publisher Full Text\n\nWillis HJ, Eldridge AL, Beiseigel J, et al.: Greater satiety response with resistant starch and corn bran in human subjects. Nutr Res. 2009; 29(2): 100–105. PubMed Abstract \| Publisher Full Text\n\nJohnston KL, Thomas EL, Bell JD, et al.: Resistant starch improves insulin sensitivity in metabolic syndrome. Diabet Med. 2010; 27(4): 391–397. PubMed Abstract \| Publisher Full Text\n\nMaki KC, Pelkman CL, Finocchiaro ET, et al.: Resistant starch from high-amylose maize increases insulin sensitivity in overweight and obese men. J Nutr. 2012; 142(4): 717–723. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBray GA, Greenway FL: Current and potential drugs for treatment of obesity. Endocr Rev. 1999; 20(6): 805–875. PubMed Abstract \| Publisher Full Text\n\nZhou J, Martin RJ, Tulley RT, et al.: Dietary resistant starch upregulates total GLP-1 and PYY in a sustained day-long manner through fermentation in rodents. Am J Physiol Endocrinol Metab. 2008; 295(5): E1160–1166. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKeenan MJ, Zhou J, McCutcheon KL, et al.: Effects of resistant starch, a non-digestible fermentable fiber, on reducing body fat. Obesity (Silver Spring). 2006; 14(9): 1523–1534. PubMed Abstract \| Publisher Full Text\n\nGreenway F, O'Neil CE, Stewart L, et al.: Fourteen weeks of treatment with Viscofiber® increased fasting levels of glucagon-like peptide-1 and peptide-YY. J Med Food. 2007; 10(4): 720–724. PubMed Abstract \| Publisher Full Text\n\nZhou J, Hegsted M, McCutcheon KL, et al.: Peptide YY and proglucagon mRNA expression patterns and regulation in the gut. Obesity (Silver Spring). 2006; 14(4): 683–689. PubMed Abstract \| Publisher Full Text\n\nZheng J, Enright F, Keenan M, et al.: Resistant starch, fermented resistant starch, and short-chain fatty acids reduce intestinal fat deposition in Caenorhabditis elegans. J Agric Food Chem. 2010; 58(8): 4744–4748. PubMed Abstract \| Publisher Full Text\n\nVenema K, van den Abbeele P: Experimental models of the gut microbiome. Best Pract Res Clin Gastroenterol. 2013; 27(1): 115–126. PubMed Abstract \| Publisher Full Text\n\nden Besten G, van Eunen K, Groen AK, et al.: The role of short-chain fatty acids in the interplay between diet, gut microbiota, and host energy metabolism. J Lipid Res. 2013; 54(9): 2325–2340. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMangian HF, Tappenden KA: Butyrate increases GLUT2 mRNA abundance by initiating transcription in Caco2-BBe cells. JPEN J Parenter Enteral Nutr. 2009; 33(6): 607–617; discussion 617. PubMed Abstract \| Publisher Full Text\n\nTappenden KA, McBurney MI: Systemic short-chain fatty acids rapidly alter gastrointestinal structure, function, and expression of early response genes. Dig Dis Sci. 1998; 43(7): 1526–1536. PubMed Abstract \| Publisher Full Text\n\nZhou J, Keenan MJ, Fernandez-Kim SO, et al.: Dietary resistant starch improves selected brain and behavioral functions in adult and aged rodents. Mol Nutr Food Res. 2013; 57(11): 2071–2074. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTachon S, Zhou J, Keenan M, et al.: The intestinal microbiota in aged mice is modulated by dietary resistant starch and correlated with improvements in host responses. FEMS Microbiol Ecol. 2013; 83(2): 299–309. PubMed Abstract \| Publisher Full Text\n\nKeenan MJ, Janes M, Robert J, et al.: Resistant starch from high amylose maize (HAM-RS2) reduces body fat and increases gut bacteria in ovariectomized (OVX) rats. Obesity (Silver Spring). 2013; 21(5): 981–984. PubMed Abstract \| Publisher Full Text\n\nWang X, Brown IL, Khaled D, et al.: Manipulation of colonic bacteria and volatile fatty acid production by dietary high amylose maize (amylomaize) starch granules. J Appl Microbiol. 2002; 93(3): 390–397. PubMed Abstract \| Publisher Full Text\n\nLe Blay G, Michel C, Blottiere HM, et al.: Enhancement of butyrate production in the rat caecocolonic tract by long-term ingestion of resistant potato starch. Br J Nutr. 1999; 82(5): 419–426. PubMed Abstract\n\nCheng HH, Lai MH: Fermentation of resistant rice starch produces propionate reducing serum and hepatic cholesterol in rats. J Nutr. 2000; 130(8): 1991–1995. PubMed Abstract\n\nTappenden KA, Deutsch AS: The physiological relevance of the intestinal microbiota--contributions to human health. J Am Coll Nutr. 2007; 26(6): 679S-683S. PubMed Abstract \| Publisher Full Text\n\nSilvi S, Rumney CJ, Cresci A, et al.: Resistant starch modifies gut microflora and microbial metabolism in human flora-associated rats inoculated with faeces from Italian and UK donors. J Appl Microbiol. 1999; 86(3): 521–530. PubMed Abstract \| Publisher Full Text\n\nKokke FT, Scholtens PA, Alles MS, et al.: A dietary fiber mixture versus lactulose in the treatment of childhood constipation: a double-blind randomized controlled trial. J Pediatr Gastroenterol Nutr. 2008; 47(5): 592–597. PubMed Abstract \| Publisher Full Text\n\nTsiros MD, Sinn N, Coates AM, et al.: Treatment of adolescent overweight and obesity. Eur J Pediatr. 2008; 167(1): 9–16. PubMed Abstract \| Publisher Full Text\n\nMacLean PS, Higgins JA, Jackman MR, et al.: Peripheral metabolic responses to prolonged weight reduction that promote rapid, efficient regain in obesity-prone rats. Am J Physiol Regul Integr Comp Physiol. 2006; 290(6): R1577–R1588. PubMed Abstract \| Publisher Full Text\n\nBelobrajdic DP, King RA, Christophersen CT, et al.: Dietary resistant starch dose-dependently reduces adiposity in obesity-prone and obesity-resistant male rats. Nutr Metab (Lond). 2012; 9(1): 93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAryana K, Greenway F, Dhurandhar N, et al.: Dataset 1 in: A resistant-starch enriched yogurt: fermentability, sensory characteristics, and a pilot study in children. F1000Research. 2015. Data Source\n\nAryana K, Greenway F, Dhurandhar N, et al.: Dataset 2 in: A resistant-starch enriched yogurt: fermentability, sensory characteristics, and a pilot study in children. F1000Research. 2015. Data Source" }	[ { "id": "8858", "date": "30 Jun 2015", "name": "Patrick O’Neil", "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\nThere is much current attention to the role of the gut, especially gut microbiota, in the regulation of body weight. At the same time the US and other countries face escalating rates of childhood and adolescent obesity, suggesting the importance of obesity prevention efforts, of which there have been few shown to have significant impacts. This study examines one innovative potential approach to preventing obesity targeting food components, specifically the incorporation of dietary resistant starch into a common food item, yogurt. The study focused on the acceptability of the sensory characteristics of yogurt enhanced with one such resistant starch, compared to yogurt enhanced with a control amylopectin starch, among pre-adolescent and adolescent subjects. A small pilot 4-week trial also looked at the effects of the resistant starch when consumed by 4 subjects. The sensory evaluation study was well-designed with tight controls and a sizable N. Assessment of various sensory attributes was rather comprehensive among the adolescent sample, which unfortunately was substantially smaller than the sample of 7- to 8-year olds who could not be expected to make the finer sensory distinctions on which the adolescents were queried. The actual resistant starch content of the various yogurt samples was verified by analyses. The sensory evaluation study showed that the subjects consistently preferred the control yogurt to that containing the resistant starch. The adolescents found the resistant starch yogurts to be inferior to the control yogurts on appearance, taste, thickness, sandiness and acceptability. About 25% of the adolescent subjects rated the resistant starch as unacceptable. Among the younger children, whose only ratings were a smiling face, neutral face or sad/frowning face, the resistant starch yogurt was more strongly disliked, with nearly two-thirds of subjects awarding it a frowning face. The discussion does not adequately address these results, which show consistently that the sensory characteristics of the resistant starch yogurt preparation are not well received by children and adolescents. The conclusion that “…our studies confirmed the general acceptability of incorporating HAM-RS2 into yogurt…” seems to fly in the face of these results. The very small pilot clinical trial provided some support for the hypothesis that the resistant starch would produce changes in gut microbiota, findings worth following up with a larger study. It also showed that despite the limited acceptability ratings, the products could be consumed over a 4-week period. However, the weight gain seen among all four subjects is disturbing. Although the discussion dismisses this finding by saying that it may reflect growth, the amounts of weight gain seen over 4 weeks were 0.4, 1.7, 1.9 and 3.2 kg. Certainly the latter three gains, if continued over a year, would be quite excessive. Given that the point of this dietary intervention is to prevent excessive weight gain, these preliminary findings are cause for considerable concern. The authors have conducted a well-designed study of the feasibility of this innovative dietary intervention which has the purpose of obesity prevention. The findings show that it is possible to get children and adolescents to consume yogurt containing resistant starches, but more development is necessary to produce a food product with adequate acceptability. Aside from that, however, the assumption that this product, if consumed regularly, might avert excess weight gain needs further study, as the limited results here suggest its effect may be just the opposite of that which was expected and desired.", "responses": [] }, { "id": "10019", "date": "26 Aug 2015", "name": "Joanne Lupton", "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 are addressing an important issue -- the development of foods for children that will not impede their growth, but also not contribute to adiposity, that children will like to eat, and that will withstand pasteurization. In their case they have chosen to add a specific resistant starch (HAM-RS2)- to yogurt. There was a good rationale for them to select this resistant starch as previous studies done in rats, and some done in humans, have shown beneficial results. They performed a series of sensory tests comparing the yogurt with resistant starch to their control yogurt (containing amylopectin). Except for the color of the product all other sensory attributes were rated better for the control than the yogurt containing resistant starch.The statistical analyses appear to be acceptable except there was no stated allowance for multiple T tests. Despite this, it is clear that the children liked the control better than the one containing resistant starch. Instead of calling this product a \"bust\" the authors have chosen to call this a success as a certain percentage still liked the experimental product sufficiently to eat it. One wonders how the abstract would have read if the resistant starch containing product scores were reversed with the controls. One would imagine that they would have been delighted if not ecstatic.The \"clinical trial\" in children was important to do, as it provided information as to the acceptability of the resistant starch over time. However, there were only 4 children in this trial and one, who was pre-pubescent (age 6), was different from the others (2) 10 y and (1) 14, and the results from that individual were also different from the other 3. In one way this is really a preliminary test of the resistant starch, not a trial, as there was no control. All 4 children ate the resistant-starch yogurt, in amounts based, in part, on their ages. Again, the results of this trial were disappointing if weight gain was a primary outcome measure. All gained weight, and for some this was substantial. The resistant starch appeared to be fermented to SCFA, but the values for pH and SCFA were reported in a previous paper. If weight gain, or gain of muscle mass and not fat stores was an important endpoint and these children were still growing, the study would have benefitted by a control group of the amylopectin containing yogurt. The title of the paper seems incomplete as it doesn't mention the aim of the pilot study. I've tried to come up with the purpose of the clinical trial, but the fact that I'm unsure what the goal was is impeding me. Is it testing the fermentability of the resistant starch product? If so, it should provide pH and SCFA data. Is it testing whether or not children will eat this product for a month, then it should say that. If it's about weight gain/loss then it needed a control. In summary, I think this idea of accepting a paper and then asking experts in the field to comment is excellent, and I'm glad to be part of the process. These authors are experts in what they are doing, and they have found some important information about a natural, granular, type 2RS from high-amylose maize (HAM-RS2). Others working in this field can benefit from the fact that they presented their data in what could be characterized as a preliminary study. However, there are two points that bother me about the interpretation of their data. First, they put a very positive spin on the HAM-RS2 intervention, when by almost all accounts this was not a positive outcome. This really should be toned down. Second, the authors make statements about the benefits of fermentability that come across as \"facts\" when rather this is an open and unresolved issue. For example, when discussing HAM-RS2 they say \"decreases plasma cholesterol and triglycerides, increases satiety, increases insulin sensitivity and is anti-adipogenic in adult populations.\" Although they show several citations that have shown this, they have not reviewed the entire literature, and they make the statement as if it is fact. This needs to be qualified.\n\nIn a different place they say \"People eat for volume and consume fewer calories when food has a lower caloric density.\"\n\nThis is Barbara Rolls's hypothesis, but not everyone would agree with this statement. In fact, Rolls herself says it mostly applies to men, not women. So, please go back through your comments that are stated as FACTS and perhaps modify them.", "responses": [] } ]	1	https://f1000research.com/articles/4-139
https://f1000research.com/articles/4-138/v1	01 Jun 15	{ "type": "Research Note", "title": "Understanding carbon regulation in aquatic systems - Bacteriophages as a model", "authors": [ "Swapnil Sanmukh", "Krishna Khairnar", "Waman Paunikar", "Satish Lokhande", "Swapnil Sanmukh", "Krishna Khairnar", "Satish Lokhande" ], "abstract": "The bacteria and their phages are the most abundant constituents of the aquatic environment, and so represent an ideal model for studying carbon regulation in an aquatic system. The microbe-mediated interconversion of bioavailable organic carbon (OC) into dissolved organic carbon (DOC) by the microbial carbon pump (MCP) has been suggested to have the potential to revolutionize our view of carbon sequestration. It is estimated that DOC is the largest pool of organic matter in the ocean and, though a major component of the global carbon cycle, its source is not yet well understood. A key element of the carbon cycle is the microbial conversion of DOC into inedible forms. The primary aim of this study is to understand the phage conversion from organic to inorganic carbon during phage-host interactions.Time studies of phage-host interactions under controlled conditions reveal their impact on the total carbon content of the samples and their interconversion of organic and inorganic carbon compared to control samples. A total organic carbon (TOC) analysis showed an increase in inorganic carbon content by 15-25 percent in samples with bacteria and phage compared to samples with bacteria alone. Compared to control samples, the increase in inorganic carbon content was 60-70-fold in samples with bacteria and phage, and 50-55-fold for samples with bacteria alone. This study indicates the potential impact of phages in regulating the carbon cycle of aquatic systems.", "keywords": [ "interconversion", "microbial carbon pump", "carbon sequestration", "refractory carbon", "global carbon cycle" ], "content": "Introduction\n\nThe regulation of carbon in aquatic systems is a major biogeochemical process. The oceans’ surface takes up about 2% more CO2 gas than they release, a proportion of which dissolves into the water, forming carbonic acid. The increase in CO2 levels in oceans decreases the pH, resulting in acidification which affects the oceanic ecosystem1. Carbon also enters the seas through the food web via photosynthesis, but does not last for long periods and is either released into the atmosphere as CO2 or sinks to the ocean depths as dead organic matter. However, a significant amount of carbon is present in the water in the form of DOC2,4,5. The roles that ocean viruses play are very important in shaping microbial population sizes as well as in regenerating carbon and other nutrients6–8. It is estimated that every second, approximately 1023 viral infections occur in the ocean. Therefore, it should not be surprising that viruses are major influential forces behind biogeochemical cycles5–8.\n\nA key element of the carbon cycle is the microbial conversion of dissolved organic carbon into inedible forms. Microbes play a dominant role in “pumping” bioavailable carbon into a pool of relatively inert compounds. The microbial carbon pump (MCP) “may act as one of the conveyor belts that transports and stores carbon in oceans.” The MCP also appears to function in deep waters, where bacteria adapted to the high-pressure environment may be able to degrade refractory DOC. Hiroshi Ogawa et al., showed that marine microbes are able to convert bioavailable DOC to refractory DOC2,4,5.\n\nThe present communication represents time studies of phage-host interactions under controlled conditions, in order to analyze their impact on the total carbon content of the source (nutrient broth) and their interconversion between organic and inorganic forms of carbon with respect to control samples. The control sample is just the nutrient broth without the inoculation of bacterium and their respective phage.\n\n\nMaterials and methods\n\nThe experiment was designed to measure the inorganic carbon levels in three conditions: control (nutrient broth only), bacteria alone and bacteria with their specific phage. The bacterium used during our study was E. coli (ATCC, strain 13706) and the bacteriophage used was phi X174 (ATCC, strain 13706 B1). They represent a good model for carbon conversion and interconversion through phage-host interactions and their interaction can be easily determined by the instruments like TOC analyzer3,6,7.\n\nAll three experimental conditions were conducted in 1L of sterilized nutrient broth each as to have a defined composition of the nutrients available for our study (HiMedia Pvt. Ltd.). For the bacteria without phage condition, sterilized nutrient broth media was inoculated with 100 cfu/ml of E. coli (ATCC 13706) previously enriched and incubated at 37°C; for the bacteria with phage condition approximately 1 ml of 1000 pfu/ml of phage were added. All flasks were sealed and incubated at 37°C for 18 hours. For control condition, sterile uninoculated nutrient broth was kept at 4°C throughout the experiment.\n\nThe initial reading were analyzed by a total organic carbon (TOC) analyzer (Shimadzu, Japan Model: TOC-Vcph) after 18 hours of incubation for all three sets of samples were recorded as “0” hours reading and before inoculation of bacteria and phages (see Table 1 and Table 2). TOC analysis was further carried out after every 2 hours until a stationary state was achieved. The stationary phase for inorganic carbon was defined by no further increase or decrease in the reading of inorganic carbon.\n\nPlease refer Figure 1 and Figure 2 for understanding the principle of TOC analysis and different types of carbon compounds. The overall experiment was repeated for 10 times and their averages are represented in the Table 1 and Table 2.\n\n\nResults\n\nThe average results of the three sets are represented in Table 1 and Table 2, which show that the inorganic carbon content of the samples increased over time (except control) in both sets. The sample set with host-phage inoculation showed a increased reading of inorganic carbon levels compared to bacteria-only. There was an average 15–25 percent increase in inorganic carbon composition of sample set with host-phage inoculation. The result indicates that the phages may have role in regulation of carbon in aquatic systems through carbon sequestration or conversion in different biologically unavailable forms and can elevate inorganic carbon content levels in aqueous environments.\n\n\nDiscussion\n\nThe increase in inorganic carbon content may be due to lysis of the host cell releasing its refractory carbon compounds and respiration produced CO2 during utilization of carbon constituent for phage assembly and development. These controlled experiment mimics the continuous viral infections occurring in the different aquatic environments2,4,5. The consistent rise in the inorganic content is an indicator that, viruses somehow, seems to regulate carbon cycle to a greater extent as observed from the increase in IC level. The analytical results as indicated from the TOC analyzer are sole representation of phage lyses event and are worth analyzing further. If we are able to understand the biochemical mechanism and the byproducts generated during this whole process we may be able to determine the carbon sequestration in a better way. Considerable research activity needs to be initiated involving different environments conditions, parameters, sources, etc to facilitate better understanding of viral life cycle involving carbon cycle as an important area of future research. It can be proposed that carbon conversation during these studies gives us the clear ideas of the possible fate of carbon cycle and the role of phages. Similarly, we can also try to elucidate the role of phages (viruses) influencing other biogeochemical cycles including Nitrogen and Sulphur by using CHNS analyzer for better understanding of this process. It is also known that the infection of microbes also alters host metabolism significantly. Carbon sequestering algae like cyanobacteria are infected by cyanophages, which complicates our understanding further and demanding further in-depth studies. Lysogenic condition established by viruses under nutrient depleted condition or harsh environment can regulate the carbon utilization processes differently. Hence, the effect of viral infection on host metabolism remains unknown5–8.\n\nFuture work is essential for understanding the cellular processes especially infected (Lysogenic) host species. It will also prove helpful in deciphering the role of phages in regulating the carbon flow in the aquatic systems like oceans where their concentration outnumbered other species.", "appendix": "Author contributions\n\n\n\nAll authors have contributed equally to this work. All authors have seen and agreed to the content of the final manuscript.\n\n\nCompeting interests\n\n\n\nNo conflicts of interest were declared.\n\n\nGrant information\n\nThe work was carried out as an in-house activity and was supported by Council of Scientific and Industrial Research (CSIR), Government of India, New Delhi.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgment\n\nWe would like to thank Council of Scientific and Industrial Research (CSIR) for providing support and Center for Science and Environment, New Delhi for invitation to present our work during Second National Research Conference on Climate Change held on 5–6 November 2011 at New Delhi.\n\n\nReferences\n\nSuttle CA: Marine viruses--major players in the global ecosystem. Nat Rev Microbiol. 2007; 5(10): 801–812. PubMed Abstract \| Publisher Full Text\n\nOgawa H, Amagai Y, Koike I, et al.: Production of refractory dissolved organic matter by bacteria. Science. 2001; 292(5518): 917–920. PubMed Abstract \| Publisher Full Text\n\nClescerl LS, Greenberg AE, Eaton AD: Standard Methods for the Examination of Water and Wastewater. (20th ed.) Washington, DC: American Public Health Association. 1999. Reference Source\n\nStone R: Marine biogeochemistry. The invisible hand behind a vast carbon reservoir. Science. 2010; 328(5985): 1476-7. PubMed Abstract \| Publisher Full Text\n\nKolber ZS, Plumley FG, Lang AS, et al.: Contribution of aerobic photoheterotrophic bacteria to the carbon cycle in the ocean. Science. 2001; 292(5526): 2492-5. PubMed Abstract \| Publisher Full Text\n\nSanmukh S, Paunikar WN, Swaminathan S, et al.: Bacteriophages as a model for studying carbon regulation in aquatic system. Nature Precedings. 2012. Reference Source\n\nSanmukh SG, Paunikar WN, Meshram DB, et al.: The phage-host interaction as a model for studying carbon regulation in aquatic system. Presented at CSE-IIT Delhi Second National Research Conference on Climate Change on 5–6 November, 2011. Reference Source\n\nWeitz JS, Wilhelm SW: Ocean viruses and their effects on microbial communities and biogeochemical cycles. F1000 Biol Rep. 2012; 4: 17. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "8840", "date": "22 Jun 2015", "name": "Balendu Shekher Giri", "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 paper deals with an understanding of the use of Bacteriophages as a model for carbon regulation in aquatic systems The results show the increase in inorganic carbon content by 15-25 percent in samples with bacteria and phage compared to samples with bacteria alone with comparing to control samples, the increase in inorganic carbon content was 60-70-fold in samples with bacteria and phage, and 50-55-fold for samples with bacteria alone being reported by the authors. The biogeochemical process of carbon in aquatic environment is well discussed in the manuscript. The manuscript has been written in good English and journal guidelines have been followed strictly. The paper should be indexed only after revision by the authors.Abstract: This section is written very well with all the results and basic concepts. This section is also explaining the present work done by the authors. Introduction, Page 2, Line 3: \"The oceans’ surface takes up about 2% more CO2 gas than they release, a proportion of which dissolves into the water, forming carbonic acid” requires a reference. Introduction, Page 2, Line 6: “Carbon also enters the “seas” through the food web via photosynthesis, but does not last for long periods and is either released into the atmosphere as CO2 or sinks to the ocean depths as dead organic matter”. The “seas” should be “sea” and a reference is required. Introduction, Page 2, Line 23: Reference “Hiroshi Ogawa et al.,” should come with the year of publication. Material and methods, this section is written very well and you can understand the experiments conducted by the authors. I have only some queries and suggestions for this section:Page 2, Line 8: Kindly provide the manufacturer information for the TOC analyzer in the first usage. Result section is very short (even shorter than abstract) but it’s very informative and well written. Discussion section is also very well written with comparative study of this present work. But I think if you provide 2 or 3 more references then it will be better for this publication. Conclusion; I think authors have been forgotten to provide the conclusion part of this work and it is very important. Authors should be adding this section. Table 1 and 2 are presented very well and it’s looking like whole results are explained by them. Figure 3 and 4 are not cited in the text and it should be cited. These figures are the main results and could be explained in the results and discussion section. Figure 1 and 2 are presented very well and for figure 3 and 4 authors may modify with removing “in hours” by “h” and “in ppm” by “ppm”. Figure 3 and 4: I am unable to understand what differences between both figures are because all the things are same. Should be for table 1 and 2 (ppm 1 and ppm 2 should be come in the figure titles)?", "responses": [ { "c_id": "1429", "date": "23 Jun 2015", "name": "Swapnil Sanmukh", "role": "Author Response", "response": "Dear Sir,First of all thank you for approving our manuscript.Regarding your queries:-I think most of them are not quite critical but we will cite figure 3 and figure 4 as they are not cited in the article.-The figures are represented as per journals guidelines.-We will provide the make of TOC analyser when it is represented firstly in the article.-We have done the experiments in duplicate so that we can maintain homogeneity and have reproducible results minimizing manual errors. -The conclusion part is not included because the article is a research note, not a research article and it is not mandatory for this type of article.We would like to thanks for your time and suggestions for improving the quality of our manuscript." } ] }, { "id": "9219", "date": "15 Jul 2015", "name": "Mayur Bharat Kurade", "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 paper deals with Bacteriophages as a model for carbon regulation in aquatic systems. The increase in inorganic carbon content was 60-70-fold in samples with bacteria and phage, and 50-55-fold for samples with bacteria alone being reported by the authors. The authors presented their experimental results quite well enough, that can be easily understandable to the readers. This manuscripts meets the necessary standards for this journal. The authors should pay attention on few of the concerns enlisted below. This manuscript should be acceptable after the changes suggested herewith. The figures (3 and 4) are not mentioned in the manuscript text, The discussion part may be elaborated to have much clear idea of the present work and its impact. Please consider to use abbreviations instead of full forms, wherever necessary, e.g. '2 h' instead of '2 hours' in Materials and methods section, and '%' instead of 'percent' in \"Results\" section. Please follow the reference styles with the journal guidelines. e.g. Hiroshi Ogawa et al., ....should be corrected as Ogawa et al., followed by reference number in superscript. Reference number 3 is missing in introduction. References should be arranged in proper sequence. Please check all other references. Authors should elaborate the 'Results section', as results of Fig. 1 to 4 are not discussed. I suggest author to overcome short details in this particular section.", "responses": [ { "c_id": "1464", "date": "15 Jul 2015", "name": "Swapnil Sanmukh", "role": "Author Response", "response": "Dear Sir,First of all thank you for approving our manuscript.Regarding your queries:-We will change the abbreviations for the units used in the figures and results.-We will modify the reference style not properly cited for Hiroshi Ogawa et al as Ogawa et al.-We will elaborate the discussion as well as result section and cite the figures which are not mentioned or not discussed. We would like to thanks for your time and suggestions for improving the quality of our manuscript." } ] } ]	1	https://f1000research.com/articles/4-138
https://f1000research.com/articles/4-137/v1	01 Jun 15	{ "type": "Research Article", "title": "Combining complexity measures of EEG data: multiplying measures reveal previously hidden information", "authors": [ "Thomas Burns", "Ramesh Rajan", "Ramesh Rajan" ], "abstract": "Many studies have noted significant differences among human electroencephalograph (EEG) results when participants or patients are exposed to different stimuli, undertaking different tasks, or being affected by conditions such as epilepsy or Alzheimer's disease. Such studies often use only one or two measures of complexity and do not regularly justify their choice of measure beyond the fact that it has been used in previous studies. If more measures were added to such studies, however, more complete information might be found about these reported differences. Such information might be useful in confirming the existence or extent of such differences, or in understanding their physiological bases. In this study we analysed publically-available EEG data using a range of complexity measures to determine how well the measures correlated with one another. The complexity measures did not all significantly correlate, suggesting that different measures were measuring unique features of the EEG signals and thus revealing information which other measures were unable to detect. Therefore, the results from this analysis suggests that combinations of complexity measures reveal unique information which is in addition to the information captured by other measures of complexity in EEG data. For this reason, researchers using individual complexity measures for EEG data should consider using combinations of measures to more completely account for any differences they observe and to ensure the robustness of any relationships identified.", "keywords": [ "electroencephalograph", "complexity", "complexity measure", "sample entropy", "permutation entropy", "Lemel-Ziv complexity", "fractal dimension", "Weiner entropy" ], "content": "Introduction\n\nElectroencephalography (EEG) is a common, relatively non-invasive research and diagnostic tool. Its one-dimensional signals from localised peripheral regions on the head make it attractive for its simplistic fidelity and has allowed high clinical and basic research throughput. When it comes to interpreting EEG data, investigators have a wide range of analytical tools at their disposal (Dauwels et al., 2010; Delorme & Makeig, 2004) and in recent years have explored a number of novel relationships between measures of complexity (Cao & Slobounov, 2011; Dauwels et al., 2011; Jing et al., 2014; Sitt et al., 2014; Susmáková & Krakovská, 2008; Weiss et al., 2011). Studies which have included complexity measures, however, do not regularly include more than one or two such measures. For example, Dauwels et al. (2011) include the Lempel-Ziv (LZ) complexity measure (Lempel & Ziv, 1976) - an algorithmic-based measure - and regularity measures, but ignore potential chaotic and fractal measures. This is not to suggest that the LZ complexity measure or that regularity measures are meaningless, nor that chaotic and fractal measures are more or less important than other measures of complexity, but that all may be measuring different features. Thus, for a more complete and robust picture of any relationships found for one complexity measure in EEG data, it might be useful for investigators to include other measures in their analyses.\n\nThis study therefore aims to determine whether different measures of complexity of EEG signals correlate, and (if so) to what degree. To do this, a small battery of complexity measures were computed for publicly-available normative data and subsequently analysed for correlations. If some measures were found not to significantly correlate or correlate fully, this would suggest that these measures are detecting unique information which might otherwise have remained hidden to investigators who were computing only a single complexity measure from their data.\n\n\nMethods\n\nOne thousand, one hundred EEG recordings of 1-second duration from 13 healthy control subjects undergoing an object recognition psychophysics task were obtained from a publicly-available database created by Begleiter (1996) of the Neurodynamics Laboratory, State University of New York Health Center, Brooklyn, United States. The control subjects were selected so as to avoid disease-specific influences. While our sample size was limited by the database, prior studies which used this database reached significance (thus, independent power calculations were not performed). Detailed demographic, subject, recording, and task information can be found in the original study by Zhang et al. (1995). The following complexity measures were calculated in MATLAB for each recording: LZ algorithmic complexity (Lempel & Ziv, 1976), fractal dimension estimation (FD) (Higuchi, 1988), permutation entropy (PE) (Bandt & Pompe, 2002), Wiener entropy (WE) (Wiener, 1954), and spectral structure variability (SSV) (Singh, 2011). These measures were chosen on the basis of their broad representation of different conceptions of ‘complexity’, including informational theoretic, chaotic/fractal, and computational informatic approaches; details of how these measures are calculated and what they measure are well-described by their respective original proposers (Bandt & Pompe, 2002; Higuchi, 1988; Lempel & Ziv, 1976; Singh, 2011; Wiener, 1954) and so will not be repeated here (see Data Availability for code details). Many more measures exist than these, however as the principle aim of this paper was to determine if differences exist at all, any differences detected in this small cross-section of measures would sufficiently illustrate this. Results from the complexity measures were analysed by linear regression and significance (considered as p<0.05) for relationships between pairs of measures was calculated using Pearson product-moment correlation coefficients. For relationships which appeared to have non-linear components when viewing its scatterplot, binomial regression was attempted. Graphs and statistics were generated using MATLAB R2012a (7.14.0.739) and Microsoft Excel 2007.\n\n\nResults\n\nOf the ten pairs of measures, eight pairs exhibited highly significant (p<0.0001) correlations while two pairs - (i) PE and FD, (ii) WE and LZ - did not significantly correlate (Table 1 and Table 2). High degrees of spread were noted among all correlations.\n\nThese relationships were visualised using scatter plots (Figure 1 and Figure 2) to help determine if any of these relationships may be non-linear. Two such relationships - (i) LZ and FD, (ii) SSV and FD - appeared to follow a binomial trend (Figure 3), and binomial regression improved these relationships greatly.\n\nEight pairs of complexity measures of the EEG signals had a significant (p<0.0001) correlation. Although the relationships are significant, high degrees of spread are noticeable and some of the relationships may have non-linear components. EEG = electroencephalogram; LZ = Lempel-Ziv algorithmic complexity; FD = fractal dimension estimate (Higuchi method); PE = permutation entropy; SSV = spectral structure variability; WE = Wiener entropy (also known as spectral flatness).\n\nTwo pairs of complexity measures of the EEG signals were insignificant and uncorrelated - PE & FD (r=-0.0255, p=0.3990) and WE & LZ (r=-0.0472, p=0.1174). There appears to be no non-linear components or any evidence of a clear relationship between these pairs of measures. EEG = electroencephalogram; LZ = Lempel-Ziv algorithmic complexity; FD = fractal dimension estimate (Higuchi method); PE = permutation entropy; WE = Wiener entropy (also known as spectral flatness).\n\nTwo pairs of complexity measures of the EEG signals appeared to have noticeable non-linear relationships: (i) LZ and FD; and (ii) SSV and FD. Although these binomial relationships were - like their linear relationships - significant (p<0.0001), the binomial regressions produced less spread and appear to be truer representations of the relationships. EEG = electroencephalogram; LZ = Lempel-Ziv algorithmic complexity; FD = fractal dimension estimate (Higuchi method); SSV = spectral structure variability.\n\n\nDiscussion and conclusions\n\nSome - but not all - measures of complexity of EEG signals correlate, and to varying degrees of significance, e.g. we found no significant relationship between PE and FD but did find a significant relation between PE and LZ. To the best of our knowledge, this study represents the first report of such complexity measure differences in EEG signals. Of the many complexity measures available to researchers investigating EEG data, overreliance or overconfidence in any single measure therefore seems misplaced. As research groups who have attempted to classify or predict sleep stages or conscious states from EEG data have implicitly noted (Susmáková & Krakovská, 2008; Sitt et al., 2014; Weiss et al., 2011), no individual measure can reliably predict all possibly relevant physiology. Instead, combinations of measures are needed. In the same way, no individual measurement of complexity can reliably predict all possibly relevant complexity.\n\nIn part, the results from this short study reflect on a more generalised ambiguity of the concept of ‘complexity’. Who is to say, after all, that more is revealed about ‘complexity’ by FD than LZ? It seems that it cannot be said that either elucidate more or less about ‘complexity’, since both ultimately treat it in a different way on even a conceptual basis. This further reiterates the primary finding of the present study: by multiplying measures we can reveal information which was previously hidden or unknown to us. However, there are two caveats to this: (1) not all information may be physiologically or otherwise relevant all of the time (or ever); and (2) different datasets may, due to their differences in nature, show different levels of covariance between complexity measures.\n\nIt would be interesting for future studies to analyse previously-noted complexity differences - e.g., between patients with and without Alzheimer's disease (Dauwels et al., 2010) - to determine if these differences were measuring the same difference. Our results suggests they may not be. And if this is the case, more might be gleaned from the available data if more measures were applied in combination. It could even be possible that there exists entirely separate complexity dimensions, along which patients progress at different rates. Such information could therefore contain even more physiological, clinical, or other significance than previously thought.\n\n\nData availability\n\nA copy of MATLAB functions used in this study has been uploaded to GitHub and can be accessed here: https://github.com/tfburns/MATLAB-functions-for-complexity-measures-of-one-dimensional-signals.\n\nResults from these functions for the EEG data used can be found in the F1000Research repository (see below).\n\nF1000Research: Dataset 1. Calculated complexity measures for 1100 EEG recordings. The data are the results from MATLAB functions which calculated complexity measures for each EEG recording, 10.5256/f1000research.6590.d48983 (Burns & Rajan, 2015).", "appendix": "Author contributions\n\n\n\nTB conceived, wrote, and performed all analyses pertaining to the manuscript. RR assisted in the development of the conceptual and methodological components of the analyses. All authors have read and agreed to the final content of the manuscript.\n\n\nCompeting interests\n\n\n\nThe authors declared no competing interests.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThank you to MATLAB community members for their assistance with some of the functions used for calculating complexity measures in this paper.\n\n\nReferences\n\nBandt C, Pompe B: Permutation entropy: a natural complexity measure for time series. Phys Rev Lett. 2002; 88(17): 174102. PubMed Abstract \| Publisher Full Text\n\nBegleiter H: EEG data set of healthy and alcoholic adults completing psychophysics tasks [data-set]. Machine Learning Repository, University of California, Irvine [distributor]. 1996. Reference Source\n\nBurns T, Rajan R: Dataset 1: Combining complexity measures of EEG data: multiplying measures reveal previously hidden information. F1000Research. 2015. Data Source\n\nCao C, Slobounov S: Application of a novel measure of EEG non-stationarity as 'Shannon- entropy of the peak frequency shifting' for detecting residual abnormalities in concussed individuals. Clin Neurophysiol. 2011; 122(7): 1314–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDauwels J, Srinivasan K, Ramasubba Reddy M, et al.: Slowing and Loss of Complexity in Alzheimer’s EEG: Two Sides of the Same Coin? Int J Alzheimers Dis. 2011; 2011: 539621. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelorme A, Makeig S: EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004; 134(1): 9–21. PubMed Abstract \| Publisher Full Text\n\nDauwels J, Vialatte F, Cichocki A: Diagnosis of Alzheimer's disease from EEG Signals: where are we standing? Curr Alzheimer Res. 2010; 7(6): 487–505. PubMed Abstract \| Publisher Full Text\n\nHiguchi T: Approach to an irregular time series on the basis of the fractal theory. Physica D: Nonlinear Phenomena. 1988; 31(2): 277–83. Publisher Full Text\n\nJing L, Jiaqing Y, Xianzeng L, et al.: Using Permutation Entropy to Measure the Changes in EEG Signals During Absence Seizures. Entropy. 2014; 16(6): 3049–61. Publisher Full Text\n\nLempel A, Ziv J: On the complexity of finite sequences. IEEE Transactions on Information Theory. 1976; 22(1): 75–81. Reference Source\n\nSingh NC: Measuring the ‘complexity’ of sound. Pramana. 2011; 77(5): 811–6. Publisher Full Text\n\nSitt JD, King JR, El Karoui I, et al.: Large scale screening of neural signatures of consciousness in patients in a vegetative or minimally conscious state. Brain. 2014; 137(pt 8): 2258–70. PubMed Abstract \| Publisher Full Text\n\nSusmáková K, Krakovská A: Discrimination ability of individual measures used in sleep stages classification. Artif Intell Med. 2008; 44(3): 261–77. PubMed Abstract \| Publisher Full Text\n\nWeiss B, Clemens Z, Bódizs R, et al.: Comparison of fractal and power spectral EEG features: effects of topography and sleep stages. Brain Res Bull. 2011; 84(6): 359–75. PubMed Abstract \| Publisher Full Text\n\nWiener N: In The human use of human beings: Cybernetics and society. Boston Houghton Mifflin. 1954; 15–27. Reference Source\n\nZhang XL, Begleiter H, Porjesz B, et al.: Event related potentials during object recognition tasks. Brain Res Bull. 1995; 38(6): 531–8. PubMed Abstract \| Publisher Full Text" }	[ { "id": "8965", "date": "25 Jun 2015", "name": "Xiao Shifu", "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 an interesting article. The authors computed five complexity measures which were broadly representative, aiming to determine the correlations and differences between measures of complexity used in EEG signals study. Just as the findings of this study, more information can be revealed by multiplying measures instead of single complexity measures. Furthermore, the methods of combinations of complexity measures can be applied in future studies such as Alzheimer’s disease. With the coming of big data era, we need to adopt effective means to mine information hidden in many diseases. Perhaps, this paper gives us some reference.", "responses": [] }, { "id": "8846", "date": "24 Sep 2015", "name": "Alison Pack", "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 report addressing the utility of multiple measures used to evaluate EEG. EEG continues to be a valuable tool for research and clinical work. This paper nicely assesses through statistical analysis whether unique features of EEG can be helpful. The methods and analysis are well formulated. The conclusions are clearly articulated and are supported by research findings.", "responses": [] } ]	1	https://f1000research.com/articles/4-137
https://f1000research.com/articles/4-60/v1	09 Mar 15	{ "type": "Method Article", "title": "Prediction of multi-drug resistance transporters using a novel sequence analysis method", "authors": [ "Jason E. McDermott", "Paul Bruillard", "Christopher C. Overall", "Luke Gosink", "Stephen R. Lindemann", "Paul Bruillard", "Christopher C. Overall", "Luke Gosink", "Stephen R. Lindemann" ], "abstract": "There are many examples of groups of proteins that have similar function, but the determinants of functional specificity may be hidden by lack of sequence similarity, or by large groups of similar sequences with different functions. Transporters are one such protein group in that the general function, transport, can be easily inferred from the sequence, but the substrate specificity can be impossible to predict from sequence with current methods. In this paper we describe a linguistic-based approach to identify functional patterns from groups of unaligned protein sequences and its application to predict multi-drug resistance transporters (MDRs) from bacteria. We first show that our method can recreate known patterns from PROSITE for several motifs from unaligned sequences. We then show that the method, MDRpred, can predict MDRs with greater accuracy and positive predictive value than a collection of currently available family-based models from the Pfam database. Finally, we apply MDRpred to a large collection of protein sequences from an environmental microbiome study to make novel predictions about drug resistance in a potential environmental reservoir.", "keywords": [ "antibiotic resistance", "bacteria", "machine learning", "linguistics", "protein function", "multidrug resistance transporters", "microbiome" ], "content": "Introduction\n\nGram-negative bacteria are a major cause of many human diseases and, due to the emergence of antibiotic resistance, new means to combat them are a pressing international health issue. Recently the Center for Disease Control and Prevention (CDC) highlighted this problem, by stating that, “… new antibiotics will always be needed to keep up with resistant bacteria…” (CDC, 2013). Antibiotic resistance is mediated by several distinct mechanisms including enzymatic conversion of antibiotics and transporters that eliminate antibiotics from inside cells (Blair et al., 2015). Transporter superfamilies can be easily identified by standard sequence similarity but specific functional information (e.g. substrate specificity) can be more problematic.\n\nProtein function has traditionally been determined by costly and time-consuming experimental approaches. Tools to determine sequence similarity such as BLAST have enabled efficient annotation of novel proteins by transfer of function. Such methods have been very effective at delineating families of functionally similar proteins that have similar sequences. More flexible approaches using simple grammars like regular expressions and hidden Markov models have improved this process significantly (Bateman et al., 2000; Gough & Chothia, 2002). However, there remain many proteins that cannot be readily associated with known functions using these approaches, largely because they are unrelated by sequence. The field of linguistics is concerned with the structure of languages and studies morphology, syntax, and semantics. This task, which is grounded in mathematics, is directly analogous to the task of interpreting sequences of amino acids to predict function. To date, the application of linguistic-rooted approaches, such as generative grammars, to protein sequences and the use of rigorous and exhaustive approaches to optimize models has been extremely limited.\n\nGenerative grammars have a rich history in linguistic analysis with limited application to biological problems (Durbin et al., 1998). They can be classified in terms of the Chomsky hierarchy where grammars lower in the hierarchy (e.g., regular grammars) are simpler to understand, compute with, and parse; while grammars further up in the hierarchy are more complex but also have more descriptive power. Algorithms such as PROSITE (Hofmann et al., 1999) identify simple motifs in proteins using regular expressions, which are the simplest form of grammar (i.e. regular grammars). Hidden Markov models (HMM), a type of regular grammar, have also been applied to detect protein motifs and families. In addition to regular grammars, computational biologists have utilized stochastic context-free grammars for sequence modeling (Anderson et al., 2012; Dyrka et al., 2013). Such grammars are better at modeling palindromic sequences that are found in RNA structure. All three of these are limited, however, because they still require an underlying sequence alignment.\n\nGenerative grammars have a rich history in linguistic analysis with limited application to biological problems (Durbin et al., 1998). They can be classified in terms of the Chomsky hierarchy where grammars lower in the hierarchy (e.g., regular grammars) are simpler to understand, compute with, and parse; while grammars further up in the hierarchy are more complex but also have more descriptive power. Algorithms such as PROSITE (Hofmann et al., 1999) identify simple motifs in proteins using regular expressions, which are the simplest form of grammar (i.e. regular grammars). Hidden Markov models (HMM), a type of regular grammar, have also been applied to detect protein motifs and families. In addition to regular grammars, computational biologists have utilized stochastic context-free grammars for sequence modeling (Anderson et al., 2012; Dyrka et al., 2013). Such grammars are better at modeling palindromic sequences that are found in RNA structure. All three of these are limited, however, because they still require an underlying sequence alignment.\n\nThe regular expressions contained in the PROSITE database are identified using a manual process to first gather a set of examples of a functional class, perform a multiple sequence alignment on those examples, and finally generate a regular expression by looking at regions of the sequence that align and are generally functionally important, for example a phosphorylated residue or active site. A similar procedure is used to create hidden Markov models (HMMs) such as those found in the Pfam database, except that the process of determining a model is automated. Motif determination using these methods is practically limited to operation on families of related protein sequences that have been aligned and has been carried out manually for individual protein motifs (such as in the PROSITE database). Many proteins with the same function may not have significant sequence similarity to allow alignments to be easily or accurately performed. The dependence on multiple sequence alignments and manual construction of protein patterns limits the ability to provide insight into problematic protein motifs.\n\nPreviously we have described an effective approach to classification of problematic protein families such as bacterial type III secreted effectors that share little sequence similarity (McDermott et al., 2011; Samudrala et al., 2009). This method used a support vector machine to integrate different sequence-based features and did not use multiple sequence alignment; rather, because the secretion signal is located in the most N-terminal region of the proteins, it took advantage of this natural alignment of disparate sequences. For problematic protein families in which the discriminating motifs are located in different regions of the protein, methods are needed to be able to automatically identify motifs or features, even where the sequence background might be very noisy.\n\nIn this study we describe an application of the Proactive Intelligent Learning with Grammar (PILGram) method to protein sequences to develop patterns that can discriminate functional classes of proteins in an alignment-free manner. PILGram uses a genetic algorithm to automate feature selection and build regular expressions that discriminate between classes. We first show that PILGram is able to partially re-create PROSITE patterns for ser/thr phosphatase binding and for zinc fingers in an automated and alignment-free manner. We then apply PILGram to classify transporters involved in drug resistance from other transporter proteins and show that the resulting PILGram model performs better than existing HMM models at classifying proteins in this important functional class. Finally, we combine different PILGram models using a simple voting method to develop an effective classifier called MDRpred. The patterns identified by PILGram map to regions that are likely to be important for substrate specificity, highlighting regions that could be targeted for drug development. We show that PILGram can be a general tool for development of simple patterns for functional classification of protein sequences. As a demonstration we apply MDRpred to a metagenome from an environmental microbial community and highlight several high-confidence predictions of novel MDR transporter proteins.\n\n\nMethods\n\nTo examine the ability of PILGram to identify patterns from unaligned protein sequences we used sets of sequences used to define regular expressions for protein motifs from the PROSITE database. In this way we could compare the output of PILGram with the established PROSITE patterns that had been generated from the aligned set of protein examples. Proteins matching each indicated PROSITE pattern (positive examples) were obtained from the PROSITE website (http://prosite.expasy.org) as the “prosite.dat” file. UniProt identifiers were extracted from the “DR” fields and the matching sequences, obtained from the UniProt database, were listed as true positives “T”. Of the sequences in the UniProt database that did not match the positive examples, approximately 6000 were chosen at random (specific numbers given for each example) to serve as negative examples.\n\nTo construct a training set for multidrug resistance transporters we obtained the protein sequences of 6097 transporter proteins from the Transporter Classification Database [TCDB; (Saier et al., 2014)] along with family classifications. This database was searched for “drug resistance” giving 71 drug resistance (DR) transporters (See MDR_TCDB_positives.fasta and MDR_TCDB_negatives.fasta datasets). We then searched the protein sequence descriptions from the UniProt database and found an additional 89 sequences annotated with “[drug] resistance” that were not included in the TCDB annotations. We used the TCDB-annotated DR transporters as our positive examples because most are accompanied by references. The ‘candidate’ list of positive examples annotated by UniProt was held out of the training set so as not to interfere with classification. The remaining 5935 sequences were used as negative examples since they are annotated as transporters but not as DR transporters in either database.\n\nGenomic DNA was extracted from two unicyanobacterial consortia cultivated from a microbial mat inhabiting Hot Lake, WA (Lindemann et al., 2013) as previously described (Cole et al., 2014). Genome reconstructions were generated as reported by Nelson et al., (manuscript submitted). Briefly, paired-end reads were generated by the US Department of Energy (DOE) Joint Genome Institute (JGI; http://jgi.doe.gov) under CSP 701, quality trimmed using Trimmomatic (Bolger et al., 2014), and assembled using IDBA-UD (Peng et al., 2012) with a minimum contig size of 250 bp. Contigs longer than 2 Kb were binned using read coverage for each scaffold using Bowtie2 (Langmead & Salzberg, 2012) and samtools (Li et al., 2009). Gene models for the genome reconstructions were generated using Prodigal (Hyatt et al., 2010) and hand-curated in some instances. Additionally, axenic organisms isolated from the consortia were sequenced of 10 Kb libraries with PacBio and assembled by the JGI, also under CSP 701. The genomes of axenic organisms were shown to be identical to the corresponding genome reconstructions (Nelson et al., submitted), and replaced these reconstructions in the database, being more complete. For the axenic isolates, gene models were generated by IMG/ER (Markowitz et al., 2009).\n\nGenomic DNA was extracted from two unicyanobacterial consortia cultivated from a microbial mat inhabiting Hot Lake, WA (Lindemann et al., 2013) as previously described (Cole et al., 2014). Genome reconstructions were generated as reported by Nelson et al., (manuscript submitted). Briefly, paired-end reads were generated by the US Department of Energy (DOE) Joint Genome Institute (JGI; http://jgi.doe.gov) under CSP 701, quality trimmed using Trimmomatic (Bolger et al., 2014), and assembled using IDBA-UD (Peng et al., 2012) with a minimum contig size of 250 bp. Contigs longer than 2 Kb were binned using read coverage for each scaffold using Bowtie2 (Langmead & Salzberg, 2012) and samtools (Li et al., 2009). Gene models for the genome reconstructions were generated using Prodigal (Hyatt et al., 2010) and hand-curated in some instances. Additionally, axenic organisms isolated from the consortia were sequenced of 10 Kb libraries with PacBio and assembled by the JGI, also under CSP 701. The genomes of axenic organisms were shown to be identical to the corresponding genome reconstructions (Nelson et al., submitted), and replaced these reconstructions in the database, being more complete. For the axenic isolates, gene models were generated by IMG/ER (Markowitz et al., 2009).\n\nPhysiochemical properties (PPs) were calculated using the Python propy module (Cao et al., 2013). Properties were calculated using the 147 Composition, Transition, Distribution (CTD) descriptors in propy. The PP-protein regular expressions (PRE) were represented as a combination of regular expression from the standard PRE with one of the PPs. Score for a match was given by the mean of all PP scores for each sequence region matching the PRE.\n\nPILGram models were constructed using half the training data and performance was evaluated with the other half. PILGram optimization was based on accuracy:\n\n\n\nwhere TP, TN, FP, and FN are true positive, true negative, false positive, and false negative predictions, respectively. For final evaluation of models we also calculated positive predictive value:\n\n\n\nand area under the receiver operator characteristic (ROC) curve (AUC) (Salzberg, 1997).\n\nMachine learning methods, like SIEVE, take features as input to build a model. Features are the smallest elements derived from the examples (protein sequences) that can be categories (e.g. amino acid type) or values (e.g. solvent accessibility values). While the selection of salient features is critical for classification, most algorithms require their manual specification. PILGram (Proactive Intelligent Learning with Grammar) is an approach to automate the feature selection process and allows for the selection of irredundant features. PILGram does this by combining a genetic algorithm and a generative grammar, which is a formalized set of rules for combining features into different patterns in the form of parse trees. PILGram generates a large number of such trees and then applies a genetic algorithm, which iteratively recombines these trees to determine an optimal model for classification of the positive and negative examples. In this way PILGram specifies an absorbing Markov chain on the space of features, and given sufficient time, will always converge to a collection of optimal non-redundant features. The mathematical foundations of and explicit algorithm for PILGram are currently pending review, but the algorithm is perhaps best understood by example.\n\nConsider the following toy example: height, weight, and age data are gathered from a population and each person is labeled as obese or not. One might like to automate the determination of obesity using only height, weight, and age. It is known that the body mass index (BMI) is a good indicator of obesity and is given by (weight/(height × height)). In order to determine this quantity, PILGram might make use of the following grammar.\n\n〈expr〉::=(〈expr〉〈op〉〈expr〉)〈attr〉\n\n〈op〉::=+- × /\n\n〈attr〉::=height weight age\n\nIn this grammar the ‘’ symbol is to be read as ‘or’ and ‘::=’ can be read as ‘replace by.’ So the second line tells us that ‘〈op〉’ can be replaced by ‘+’, ‘-‘, ‘×’, or ‘/’. The symbols to the left of ::= are called non-terminal symbols. This grammar can be used to generate features as follows.\n\n1. Write down 〈expr〉.\n\n2. Locate any non-terminal symbol in your expression.\n\n3. Replace the chosen non-terminal according to the grammar.\n\n1. If there is a non-terminal symbol in your expression, then return to step 2.\n\nThis process can be viewed as a parse tree. That is, at step 1 one writes 〈expr〉. Then every time a non-terminal symbol is replaced one writes the replacement below the non-terminal symbol and connects each symbol in the replacement with the initial symbol with a line. A vertical line is placed below each non-terminal symbol that is not replaced. The resulting expression is then read from left to right along the ‘leaves’ of the resulting tree. For instance, BMI might be produced from the procedure as follows:\n\n〈expr〉→ (〈expr〉〈op〉〈expr〉) → (〈attr〉〈op〉〈expr〉) → (weight 〈op〉〈expr〉) → (weight/〈expr〉) → (weight/(〈expr〉〈op〉〈expr〉)) → (weight/(〈expr〉×〈expr〉)) → (weight/(〈attr〉×〈expr〉)) → (weight/(〈attr〉×〈attr〉)) → (weight/(height ×〈attr〉)) → (weight/(height × height)).\n\nThis is more succinctly expressed by the parse tree:\n\nWhile one might get lucky and generate this expression by random application of the above grammar, it is highly unlikely. However, one might generate (weight-height) and (age+height × height). While neither of these expressions are BMI, BMI can be produced by mutating and crossing these feature.\n\nMutation is a process by which a node in the parse tree is randomly selected and then replaced with another value such that the tree remains consistent with the generative rules of the grammar. In some cases, one might opt to re-build the tree below the replaced node thereby giving the algorithm greater flexibility. For instance, the first expression can be represented as a parse tree and mutated as follows:\n\nThe resulting feature, (weight/height), is more similar to BMI than the initial feature, and in fact performs better at classifying obesity. To arrive at BMI we could apply the crossing procedure to (weight/height) and (age+height × height).\n\nCrossing is a process by which two features are expressed as parse trees and two of their subtrees are exchanged so that the resulting parse trees are consistent with the grammar. For instance, BMI can be found by crossing (weight/height) and (age+height × height) as follows:\n\nNot all crossings and mutations will produce better features, and not all features should be considered for crossing or mutation. To handle this, PILGram behaves stochastically and preferentially selects features for mutation and crossing according to how well they perform. The guiding principle is that features which perform better should be closer to the optimal feature than those that do not. The entire PILGram algorithm can be outlined as follows:\n\n1. Select a grammar for feature generation and a fitness function to evaluate the features against.\n\n2. Randomly generate a population of features and determine the fitness of each feature.\n\n3. Randomly subsample the population where a feature is selected with probability proportional to its fitness.\n\n4. For each feature selected in step 3, copy the feature and randomly change the value of a random node in its parse tree in a manner consistent with the grammar. Return the initial feature and the result of the mutation to the population.\n\n5. Randomly subsample the population for pairs of features with each feature selected with probability proportional to its fitness.\n\n6. For each pair selected in step 5 produce a copy of their parse trees. Randomly select a subtree in each feature’s parse tree and exchange these subtrees ensuring that the exchange produces features which are consistent with the grammar. Return the two initial features and the two new features to the population.\n\n7. Compute the fitness of all features in the population and remove the least fit features until the population returns to its initial size.\n\n8. If the fittest feature has converged, then terminate the algorithm, otherwise return to step 3.\n\nA common variation of the algorithm is to randomly generate new features at the start of step 7 and add them to the population before reducing the population size. Another common modification is to iteratively apply the algorithm such that the fitness function is updated between iterations to account for the fittest feature. This allows one to generate a list of irredundant features. Unsurprisingly, the choice of generative grammar strongly influences the quality of the resulting features. Below we will make use of Perl’s regular expression grammar to produce motifs in an alignment free fashion (Supplemental Figure 1).\n\nPILGram has been applied in areas ranging from text analysis, which uses a combination of atomic features based on letter frequency based atomic features and regular expressions, and to image analysis, which uses more complex image-based atomic features. In both of these cases PILGram not only provided features that were optimal for classification, but that were also easily interpreted by a user (unpublished results). In addition to these application spaces, precursor technology has been applied to loop unrolling in the realm of compiler optimization (Leather et al., 2009) where it was found that learned features resulted in an increase from 48% of the theoretical efficiency bound (using expert driven features) to 76% of the theoretical bound using features automatically identified by a PILGram-like algorithm. We note that PILGram does not train a classifier, rather it selects features which means that any improvements are not the result of overfitting but instead, are a consequence of carefully chosen features.\n\n\nResults\n\n\n\n\nAlignment-free identification of discriminatory protein patterns in PROSITE\n\nTo test the ability of PILGram to identify discriminatory regular expressions from unaligned sequences we focused on a well-defined group of proteins with a known discriminatory pattern. We first examined the serine-threonine phosphatase pattern (PROSITE PS00125) by obtaining 166 sequences listed as true positives from PROSITE (see Methods). For negative examples we randomly selected 5344 sequences from UniProt that are not included in the positive sequences.\n\nWe applied PILGram to this dataset using a standard regular expression grammar modified for protein sequences (see Supplemental Table 1). The algorithm was terminated after 276 iterations when the fitness (classification accuracy) did not change over 10 consecutive iterations. The resulting pattern (Table 1) had a very high accuracy and positive predictive value (PPV) at 99.9% and 92%, respectively. The pattern identified by PILGram contains the core of the existing PROSITE pattern and performs nearly as well. However, the PILGram pattern required no sequence alignment or manual determination of a conserved pattern.\n\nThe ser/thr phosphatase pattern is relatively simple and does not include any gaps of variable size. We were interested in determining if PILGram would also work on a more complicated pattern and so chose the zinc finger pattern (PS00028), which is a somewhat variable arrangement of conserved cysteine and histidine residues. We obtained the 1997 sequences used for the construction of the PROSITE pattern and additionally collected 5435 randomly selected protein sequences from the UniProt database to serve as negative examples for this test example. Because individual runs converged on different predictive patterns we ran PILGram 10 times on the dataset. In principle, PILGram will always eventually converge to the optimal pattern. However, in practice there may be ‘flat regions’ over which the fitness function does not significantly vary with feature modification or local extrema. In such situations, PILGram may take significant time to escape these regions and it is more economical to employ a weak convergence test, run PILGram several times, and aggregate the features.\n\nThe resulting patterns (Table 2) vary in composition and accuracy, with a maximum accuracy obtained of about 92%. All patterns fall short of the manually determined PROSITE pattern that has an accuracy of 99%. It is interesting to note that none of the identified patterns perfectly matches any portions of the manually determined PROSITE pattern, though there are some consistently identified features such as multiple cysteine residues.\n\nWe examined the possibility that the patterns identified by PILGram would be synergistic in their discriminatory ability. For each example protein (positive and negative) we counted how many of the individual PILGram patterns matched, then used this number as a discriminator. We found that using this simple voting procedure increased the accuracy from 92% to a maximum of 95.3% when six or more patterns match a sequence (Figure 1). While this performance still does not reach the level of the original PROSITE pattern (99%), we believe it demonstrates the utility of PILGram for identifying patterns from unaligned sequences.\n\nMatches to PILGram-generated regular expression patterns for the zinc finger domain (represented in PROSITE PS00028) were counted (X axis) and accuracy (Y axis) calculated based on the known positives and negative examples datasets (see text). Peak accuracy of the approach is attained at six pattern matches.\n\nWe were interested to know if PILGram was identifying regions of the sequence that overlap with the PROSITE pattern. We identified regions in all positive example sequences that match the ten PILGram patterns and calculated a score for each sequence based on the number of matches, per residue, that PILGram identified in the real zinc finger region. On average, 3.4 PILGram patterns match each residue of the known PS00028 pattern, whereas the number of patterns matching arbitrary residues in the sequence was 2.1. This shows that PILGram identifies more patterns overlapping the canonical zinc finger motif. However, it is clear that PILGram-derived motifs may not be canonical and further work needs to be done in this area.\n\n\nDrug resistance transporters\n\nA more difficult task for functional classification is to develop a model that will discriminate a group of functionally related proteins that cannot be aligned by traditional sequence alignment methods, or where the alignment does not allow discrimination between closely related sequences with different functions. To test its utility with these kinds of problematic proteins we applied PILGram to develop a classifier for antibiotic drug resistance transporters.\n\nThough transporter superfamily members can be identified fairly readily using standard sequence alignment approaches, previous studies have shown that sequence similarity has limited utility for classifying of transporters by substrate specificity (Barghash & Helms, 2013). The same authors also showed separately that integrating simple data (amino acid composition, dipeptide composition) could be used to classify some substrate families with good accuracy (Schaadt et al., 2010; Schaadt & Helms, 2012), but these models have little potential for providing biological insight. Additionally, it remains unclear if there are members of functional families that have yet to be discovered because of lack of strong sequence similarity. ATP-binding cassette transporters (ABC), resistance-nodulation-cell division (RND) superfamily, and major facilitator superfamily (MFS) transporters are common superfamilies of proteins involved in the transport of a wide variety of different compounds, such as sugars, ions, peptides, and more complex organic molecules. Multidrug resistance (MDR) transporters are found in each of these superfamilies and are primary mediators of antibiotic drug resistance (Nikaido, 2009; Nikaido & Pages, 2012). Though MDR transporters actually encompass a range of substrate specificities because there are many types of drugs they export, we hypothesized that there would be some unifying features of MDR transporters that could be captured using PILGram.\n\nWe gathered a set of 73 known MDR transporter sequences (positive examples) from the TCDB (Saier et al., 2014) and used the remainder of sequences classified in the TCDB as non-MDR transporters (negative examples; 5935 sequences). This dataset (Supplemental Data MDR_TCDB_positives.fasta and MDR_TCDB_negatives.fasta) was used to train and cross-validate MDRpred as described below.\n\nWe first evaluated how well previously generated HMM models from the Pfam database could discriminate between MDR and non-MDR transporters. We identified four Pfam models that seem to definitively identify drug resistance transporters (PF00893, PF08370, PF00873, and PF13536) and applied them to the set of sequences considering a ‘hit’ as a sequence matched by any of the models with high confidence (E value < 1e-100). The Pfam models provide very good accuracy (~97%), but only identify 10 of 73MDR transporters (14%), and these are likely hits to many of the sequences used to create the models in the first place.\n\nWe examined the ability of PILGram to find patterns capable of identifying MDR transporters from other transporter sequences. Because we believed that transmembrane regions (TMRs) would be important features in this classification task we modified our protein regular expression (PRE) grammar (Supplemental Figure 1) to bias the feature generation processes toward producing TMRs (TMR-PRE). Additionally, we included a large set of different types of protein physiochemical properties in our PILGram search (PP-PRE). PILGram included the 147 types of properties as features that could be chosen during the search. If a physiochemical property was used in a search the score (value for that particular property) was calculated for all matches of the accompanying regular expression on a sequence. If there were multiple matches to the protein then the scores were averaged.\n\nUsing a 2-fold cross-validation approach (see Methods) we used PILGram to generate 36 models (Supplemental Table 1), approximately 12 models from each of the three grammars (PRE, TMR-PRE, and PP-PRE). The models had individual accuracies ranging from 70–75%, underperforming the combination of HMM models that already exist. However, application of the simple voting approach used above in which the number of models that matched each sequence was counted, improved the results dramatically. The accuracy and PPV for increasing numbers of model matches is shown in Supplemental Figure 2, and have maximum values at the most stringent threshold (requiring all patterns be matched) of 99% and 28%, respectively.\n\nTo examine whether the individual scores could be combined to provide better prediction we employed logistic regression and found that this improved our results somewhat (Figure 2; Supplemental Figure 3). As a comparison for the same ~97% accuracy level provided by the traditional methods (Pfam family matches) our method, we call MDRpred, identifies 37 of the MDR transporters from our training set (50%) versus 10 for the traditional methods. It is clear that further development is needed to improve classification of this important group, but our approach provides the best method to date of identifying drug resistance transporters using sequence alone.\n\nThe accuracy (blue line), positive predictive value (black line), and percentage of total MDR transporters identified (coverage; red line) are shown as a function of the score threshold used (X axis). The score is derived from a logistic regression on the complete set of 36 models generated (see text).\n\nIn addition to classification of sequences a second goal of this work is to identify biologically relevant regions of proteins that are responsible for protein function. We showed that PILGram can identify regions known to be functionally important in zinc fingers. Here we apply a similar approach to identify regions that may be important for drug resistance in transporters.\n\nWe first examined the overlap in patterns by clustering models based on the training sequences that they matched (Supplemental Figure 4). The models were arranged using hierarchical clustering and then seven clusters of similar models were identified. We found that most of the clusters exhibited some similarity in patterns and model from each cluster with the highest independent accuracy listed in Table 3. We found that applying logistic regression to combine these seven models provided a similar performance as the voting method, but underperformed the logistic regression on the complete set of models. This indicates that the seven models represent a large portion of the information in the approach but that the additional models add significant value.\n\nEmrD is an MDR transporter with a solved crystal structure (Yin et al., 2006). We examined the overlap of the PILGram models on the EmrD sequence and found that the maximum overlap in matched expressions from our models occurred in H3 69-103 and the loop following H4 118-131. The latter region has been highlighted as the ‘selectivity filter’, a loop extending in to the cytoplasm and that abrogates selectivity when mutated (Yin et al., 2006) (Figure 3). This suggests that for this case where a selectivity region is known, our model can correctly identify it, though more examples would be necessary to fully demonstrate this.\n\nThe structure of the MDR transporter EmrD from E. coli (2GFP) is shown with the regions of maximum pattern overlap shown in red. This region has been shown to be the selectivity filter for substrates transported by the protein, showing that MDRpred predictions can highlight functionally important regions.\n\n\nIdentification of novel MDR transporter candidates from environmental microbiomes\n\nNew antibiotic resistance mechanisms are thought to be acquired from a very large natural reservoir of environmental bacteria, most of which have not yet been characterized (D'Costa et al., 2007; Forsberg et al., 2012; Li et al., 2014). This means that novel antibiotics may face emergence of antibiotic resistance in pathogenic bacteria by lateral gene transfer or other means (Aminov & Mackie, 2007; Forsberg et al., 2012). We were interested in determining if our models could be used to identify candidate MDRs from environmental samples. We therefore searched a species-resolved metagenomic dataset acquired from consortia (Cole et al., 2014) cultivated from a phototrophic microbial mat in Hot Lake, Washington (Lindemann et al., 2013). Though soil microbial communities have been examined for antibiotic resistance potential previously (D'Costa et al., 2007) communities living in extreme environments such as Hot Lake have not. We postulated that these kinds of communities might be rich sources of novel MDR transporters given the manifold interactions between community members (Martinez et al., 2009; Piddock, 2006).\n\nWe first searched the 69010 protein sequences from the Hot Lake consortial metagenomes (Nelson et al., submitted) for known MDR transporters using the Pfam families (PF00893, PF08370, PF00873, and PF13536) and identified 118 high-confidence (E value < 1e-100) matches. Interestingly, when we examined a set of clones gathered from 18 soil samples and selected for expression of multidrug resistance phenotypes (Forsberg et al., 2012) we found only 14 MDR transporters at the same stringency, though one caveat is that the efficiency of expression of transporters could be a limitation in this system. This suggests that the Hot Lake community has a relatively large number of MDR transporters.\n\nWe searched the Hot Lake consortium metagenome using all 36 models and then ranked the results by number of matched sequences. A histogram of number of matching models is shown in Supplemental Figure 5. Because MDRpred was trained only on transporter proteins it cannot discriminate transporter proteins from non-transporters. That is, there are a significant number of false positive predictions that match proteins unlikely to be transporters. Accordingly, we filtered candidates to only those proteins identified as transporters by Pfam and at the highest stringency we identified five candidate MDR sequences (Table 4). This step is included in the overall MDRpred process to allow accurate prediction in entire genomes or metagenomes.\n\n\nDiscussion and conclusions\n\nThe explosion in number of sequences available from a large number of sources has driven the need for better methods to capture patterns in distinct groups of functionally related sequences. Our method, based on linguistic approaches to pattern identification, has several advantages over existing methods. Not requiring a sequence alignment means that important and discriminatory sequence regions can be identified from functionally similar proteins that may be highly evolutionarily divergent or where the evolutionary relationships are unclear. Having a wide range of grammars that can be applied in the framework is a significant strength, allowing for flexible pattern discovery. In the current paper we use only variants of a protein regular expression grammar, but other grammars can easily be used depending on the application. For example, context-free grammars could be applied to better identify potential non-local interactions between different regions in the protein sequences.\n\nIn the current study we have shown that PILGram can be successfully applied to identify patterns in proteins sequences, first by application to known functional sequences from the PROSITE database, and then by application to a set of proteins related by function but where functional determinants of specificity are not well understood. From our initial work with PROSITE families we found that some kinds of patterns may be more amenable to identification using PILGram, but this was a limited proof-of-concept application that would merit further characterization. In the case of the zinc finger pattern, which has variable spacing between active cysteine and histidine measurements we found that very accurate models could be obtained by taking a simple voting approach between multiple independent PILGram models.\n\nApplication of our approach to the MDR sequences identified a set of over 30 individual PILGram models that, when combined, provided very good accuracy and positive predictive value, relative to a combination of existing HMM models in Pfam. To our knowledge this is the first attempt to develop a predictive model for MDR transporters across families. Similar to our results with PROSITE patterns we found that these models could identify regions known to be important for substrate specificity in MDRs. This represents a step forward in classification of this important group of transporters.\n\nThe vast number of uncharacterized and often unculturable bacteria in environmental communities represent a large amount of genetic potential given the ability of bacteria to share genetic information. As an example application, we ran our method on sequences identified from a moderately complex community derived from an extreme environment, in this case the Hot Lake unicyanobacterial consortia (Cole et al., 2014). We identified five candidates that were strongly predicted by the combination of our models to be MDRs. Given that the positive predictive value of the combined method is nearly 30% it is likely that one or two of these predictions is a true positive. Further research is needed to be able to predict specific drug substrate specificities for MDRs and other transporters.\n\nWe believe that the method we describe, MDRpred, will complement well the other commonly used sequence annotation methods and that it provides a unique set of predictions about potential novel MDRs. Furthermore, the PILGram approach to identification of functional patterns in unaligned sequences has applications in a large number of other problematic protein groups where function is conserved over sequence.\n\n\nData availability\n\nA publication describing the PILGram software is currently in preparation (Gosink & Bruillard, manuscript in preparation) but the software is available upon request from the authors.\n\nFigshare: Prediction of multi-drug resistance transporters dataset. Doi: http://dx.doi.org/10.6084/m9.figshare.1326181 (McDermott et al., 2015).", "appendix": "Author contributions\n\n\n\nJ.E.M conceived of the study, applied the methods, interpreted results, wrote the manuscript. P.B. developed the software, devised the grammars used, and wrote the manuscript. C.O. analyzed results and wrote the manuscript. L.G. integrated results to provide final rankings and advised on statistics. S.R.L. provided data for microbiome applications and guidance in interpretation of results.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the Signatures Discovery Initiative, a component of the Laboratory Directed Research and Development Program at Pacific Northwest National Laboratory (PNNL), a multiprogram national laboratory operated by Battelle for the U.S. Department of Energy under Contract DE-AC05-76RL01830. A portion of this research was supported by the Genomic Science Program (GSP), Office of Biological and Environmental Research (OBER), U.S. Department of Energy (DOE) and is a contribution of the PNNL Foundational Scientific Focus Area.\n\nI confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nA. Accuracy of combined patterns for classification. Matches to PILGram-generated regular expression patterns for the MDR transporters were counted (X axis) and accuracy (Y axis) calculated based on the known positives and negative examples datasets (see text). Peak accuracy is attained when all 36 patterns match the sequence, indicating that the diversity of MDR transporter sequences is likely to be high. Redundancy analysis (Table 3) shows that a similar accuracy can be obtained with seven patterns. B. Positive-predictive value of combined patterns for classification. Positive predictive value (the percentage of true positives in all positive predictions; Y axis) was calculated for each number of MDR transporter pattern matches (X axis). The maximum value is reached in sequences that match all patterns.\n\nThe receiver-operator characteristic curves (ROC) are shown for the simple vote combination of 36 models (black line), the logistic regression combination of 36 models (blue line), and the logistic regression combination of the selected seven models shown in Table 3 (red line). The area under the curve (AUC) for each method is indicated in the legend. The results show that using all models in a score derived by logistic regression provides the best performance, though the other methods also perform adequately.\n\nFor each labeled sequence (rows) we assessed the presence (red cell) or absence (white cell) of a match to any of the 36 MDRpred regular expression models (columns). Hierarchical clustering was used to highlight relationships between models based on their patterns of matches. Dendrograms for sequences and models are shown. The patterns highlight seven groups of models that share a large number of predictions. Representative members from each of these clusters are shown in Table 3.\n\nThe number of sequences (Y axis) with N matches (X axis) in the Hot Lake metagenome (69,010 sequences total) is shown as a histogram. This plot can be compared with similar plots in Supplemental Figure 2 showing accuracy and positive predictive value at each of these stringency thresholds.\n\n\nReferences\n\nAminov RI, Mackie RI: Evolution and ecology of antibiotic resistance genes. FEMS Microbiol Lett. 2007; 271(2): 147–161. PubMed Abstract \| Publisher Full Text\n\nAnderson JW, Tataru P, Staines J, et al.: Evolving stochastic context--free grammars for RNA secondary structure prediction. BMC Bioinformatics. 2012; 13(1): 78. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarghash A, Helms V: Transferring functional annotations of membrane transporters on the basis of sequence similarity and sequence motifs. BMC Bioinformatics. 2013; 14: 343. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBateman A, Birney E, Durbin R, et al.: The Pfam protein families database. Nucleic Acids Res. 2000; 28(1): 263–266. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlair JM, Webber MA, Baylay AJ, et al.: Molecular mechanisms of antibiotic resistance. Nat Rev Microbiol. 2015; 13(1): 42–51. PubMed Abstract \| Publisher Full Text\n\nBolger AM, Lohse M, Usadel B: Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014; 30(15): 2114–2120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCao DS, Xu QS, Liang YZ: propy: a tool to generate various modes of Chou's PseAAC. Bioinformatics. 2013; 29(7): 960–962. PubMed Abstract \| Publisher Full Text\n\nCDC. ANTIBIOTIC RESISTANCE THREATS in the United States. 2013. Reference Source\n\nCole JK, Hutchison JR, Renslow RS, et al.: Phototrophic biofilm assembly in microbial-mat-derived unicyanobacterial consortia: model systems for the study of autotroph-heterotroph interactions. Front Microbiol. 2014; 5: 109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nD'Costa VM, Griffiths E, Wright GD: Expanding the soil antibiotic resistome: exploring environmental diversity. Curr Opin Microbiol. 2007; 10(5): 481–489. PubMed Abstract \| Publisher Full Text\n\nDurbin R, Eddy SR, Krogh A, et al.: Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids. Cambridge University Press, 1998. Reference Source\n\nDyrka W, Nebel JC, Kotulska M: Probabilistic grammatical model for helix-helix contact site classification. Algorithms Mol Biol. 2013; 8(1): 31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForsberg KJ, Reyes A, Wang B, et al.: The shared antibiotic resistome of soil bacteria and human pathogens. Science. 2012; 337(6098): 1107–1111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGough J, Chothia C: SUPERFAMILY: HMMs representing all proteins of known structure. SCOP sequence searches, alignments and genome assignments. Nucleic Acids Res. 2002; 30(1): 268–272. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHofmann K, Bucher P, Falquet L, et al.: The PROSITE database, its status in 1999. Nucleic Acids Res. 1999; 27(1): 215–219. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHyatt D, Chen GL, Locascio PF, et al.: Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics. 2010; 11: 119. 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): 357–359. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeather H, Bonilla E, O’Boyle M: Automatic Feature Generation for Machine Learning Based Optimizing Compilation. International Symposium on Code Generation and Optimization. 2009. Publisher Full Text\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\nLi W, Sharma M, Kaur P: The DrrAB efflux system of Streptomyces peucetius is a multidrug transporter of broad substrate specificity. J Biol Chem. 2014; 289(18): 12633–12646. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindemann SR, Moran JJ, Stegen JC, et al.: The epsomitic phototrophic microbial mat of Hot Lake, Washington: community structural responses to seasonal cycling. Front Microbiol. 2013; 4: 323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarkowitz VM, Mavromatis K, Ivanova NN, et al.: IMG ER: a system for microbial genome annotation expert review and curation. Bioinformatics. 2009; 25(17): 2271–2278. PubMed Abstract \| Publisher Full Text\n\nMartinez JL, Sanchez MB, Martinez-Solano L, et al.: Functional role of bacterial multidrug efflux pumps in microbial natural ecosystems. FEMS Microbiol Rev. 2009; 33(2): 430–449. PubMed Abstract \| Publisher Full Text\n\nMcDermott JE, Corrigan A, Peterson E, et al.: Computational prediction of type III and IV secreted effectors in gram-negative bacteria. Infect Immun. 2011; 79(1): 23–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcDermott JE, Bruillard P, Overall CC, et al.: Prediction of multi-drug resistance transporters dataset. Figshare. 2015. Data Source\n\nNikaido H: Multidrug resistance in bacteria. Annu Rev Biochem. 2009; 78: 119–146. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNikaido H, Pages JM: Broad-specificity efflux pumps and their role in multidrug resistance of Gram-negative bacteria. FEMS Microbiol Rev. 2012; 36(2): 340–363. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeng Y, Leung HC, Yiu SM, et al.: IDBA-UD: a de novo assembler for single-cell and metagenomic sequencing data with highly uneven depth. Bioinformatics. 2012; 28(11): 1420–1428. PubMed Abstract \| Publisher Full Text\n\nPiddock LJ: Multidrug-resistance efflux pumps - not just for resistance. Nat Rev Microbiol. 2006; 4(8): 629–636. PubMed Abstract \| Publisher Full Text\n\nSaier MH Jr, Reddy VS, Tamang DG, et al.: The transporter classification database. Nucleic Acids Res. 2014; 42(Database issue): D251–258. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSalzberg SL: On comparing classifiers: Pitfalls to avoid and recommended approach. Data Min Knowl Discov. 1997; 1(3): 317–328. Publisher Full Text\n\nSamudrala R, Heffron F, McDermott JE: Accurate prediction of secreted substrates and identification of a conserved putative secretion signal for type III secretion systems. PLoS Pathog. 2009; 5(4): e1000375. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchaadt NS, Christoph J, Helms V: Classifying substrate specificities of membrane transporters from Arabidopsis thaliana. J Chem Inf Model. 2010; 50(10): 1899–1905. PubMed Abstract \| Publisher Full Text\n\nSchaadt NS, Helms V: Functional classification of membrane transporters and channels based on filtered TM/non-TM amino acid composition. Biopolymers. 2012; 97(7): 558–567. PubMed Abstract \| Publisher Full Text\n\nYin Y, He X, Szewczyk P, et al.: Structure of the multidrug transporter EmrD from Escherichia coli. Science. 2006; 312(5774): 741–744. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "7889", "date": "19 Mar 2015", "name": "Robert Flight", "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\nClaimsImplement a linguistic-based approach that allows the identification of functional patterns from groups of functionally related proteins that does not require alignment of the proteins The method uses regular-expressions that are generated using a parse-tree that is modified via a genetic algorithm, and fitness is scored by accuracy using training data. Able to find discriminative patterns for serine-threonine phosphatases, zinc fingers, and multi-drug resistance (MDR) transporters Predict MDR transporters in a bacterial community from \"Hot Lake\" as a potential pool for novel MDR transporters that could be transferred to current bacteria as novel source of antibiotic resistance PILGram is able to identify and separate based on the binding region responsible for substrate specificityPraisesFrom this version of the manuscript (v1), the claims are justified. Regular expressions are a linguistic construct, the authors are able to reproduce previously defined regexes without prior alignment using the PILGram method, and classify zinc fingers and MDRs by counting the number of regexes matching a particular sequence, resulting in the MDRPred method. This method, MDRPred was then applied to a newly sequenced bacterial community and possibly novel MDR transporters identified.In addition, from the text, the generation and validation of the regexes was done in a statistically rigorous way, with half of the data used for training and half of the data used for testing / validation / calculation of metrics. This is nice to see in this kind of paper, as it has become the exception rather than the rule.ReservationsAlthough I think the general claims can be justified from the text, there are some areas of concern that I think should be addressed in a subsequent version of the manuscript. These reservations fall under these major areas, ordered in what I consider most important to least important:data availability for reproducibilityactual MDRPred codeactual id's for positive and negative examples Weak \"substrate specificity\" claim lack of description in the text leading to either lack of clarity or possible misunderstandingsPILGram detailsPhysiochemical Properties and TMRElaboration of clusteringSupplemental Table 1Describing REGEXE's language implying other methods are not \"linguistic\"Each of these reservations are further detailed below.Data & Code AvailabilityNot all of the code / data necessary to reproduce the results are currently provided. While acknowledging that the primary algorithm (PILGram) is currently awaiting publication and that this is not the place to describe the particulars of that software, I think there are still steps to be taken to improve the reproducibililty of this publication by providing more of the data. It should be noted that when the PILGram algorithm is published, this publication should be updated with references and links to make it easier for others to find. That being said, this is a publication about a method, and although the particulars of the method are well described, there is no accompanying code, scripts, even psuedocode supplied so that the reader might make use of the method themselves, either on the provided supplementary FASTA files, or on their own sequences. I searched github for the term \"mdrpred\", and also for the lead authors' name and twitter username to no avail. The need for an actual script or executable (preferably open source) is increased after reading the description of including PP-PRE and TMR-PRE, and calculating their matches, as this section is a little unclear as to how exactly that calculation is performed with no example (see comment below).In addition to the code, other data that should be included are:UniProt entries for positive and negative examples for serine-threonine phosphatases and zinc fingersgenome accession and gene annotations from the metagenome analysedlist of metagenome accessions annotated as MDR's using MDRPredDate of download of PROSITE data. prosite.dat on Mar 17, 2015 shows 198 positive matches for PS00125, and I'm assuming 2018 (hard to tell from file) positive matches for PS00028, versus 166 and 1997 sequences mentioned in the text.Text files of the regexes generated by PILGram in each caseWeak \"substrate specificity\" claimThis is mentioned in the abstract, and 2 times in the introduction. The wording in the abstract implies that the method is able to delineate substrate specificity, i.e. that the method can generate regexes that are specific for different substrates. However, the one result implies rather that the regexes identify the region responsible for substrate specificity (which is really neat). These seem to be two different things in my mind, and I think either the claim in the abstract and introduction should be dropped or clarified, especially given that there is only one example provided. Finally, the claim is further weakened in the current text because the word substrate is missing from the paragraph discussing the evidence for substrate specificity (Results, Drug resistance transporters, Functional motifs identified, last paragraph in that section, no mention of \"substrate\", just specificity).More so than the \"substrate specificty\" claim, I think the authors would do well to place more emphasis on the fact that all of this work is done on sets of sequences without alignment first! It might just be me, but this was to me one of the most important things in the paper (and something I will probably make use of in my own research), that did not seem to be highlighted enough.Lack of DescriptionPILGram Algorithmic DetailsAgain I acknowledge that this is not the place to detail the full inner-workings of the PILGram algorithm, and the example in the text for BMI is useful. However, most genetic algorithms have a defined chromosome length defining the solution. I would have expected an analogous situation for PILGram, in that one would have to define the length of the regular expression. This does not appear to be the case here, given the variety of reg-ex's noted for Zinc fingers and MDRs. As far as I can tell, this is likely due to the way that individual trees can be recombined, but it is not clear from the text how different length regexes result. Clarification of how different length regexes result would be useful.Physiochemical Properties and TMRI think I understand why the physiochemical properties (PP) and transmembrane region (TMR) score were included for the MDRPred, however there is currently no discussion of their inclusion or justification in the text. From the current description of them, it is also difficult to imagine how something matches the PP-PRE and TMR-PRE, including the PP and TMR scores as part of the match. Therefore I recommend:Having a better description of the PP and TMR scores in MethodsJustification for the inclusion of PP-PRE and TMR-PRE in MDRPred. Currently the only justification is \"Because we believed ...\". I would hazard a guess that the accuracy drops precipitously without them, but there is nothing in the text currently describing why they are needed.Giving examples of how some PPs are different for different AAsExample of calculation of PP score and TMR score for a regex matchExample of full match for a derived PP-PRE or TMR-PREElaboration of clusteringIn the Results, \"Functional motifs identified\", a description of clustering the generated models is provided. The current description is ambiguous. I think what was done was a vector of length 71 (corresponding to the number of training sequences) was generated for each model, with a 1 indicating a match to the model, and 0 indicating no match to the model. These 36 vectors (one for each model) were subsequently clustered using hierarchical clustering. No description of what distance metric was used to calculate the distance between the model vectors, nor which hierachical clustering method was used is provided. In the R stats package, there are: two variations of Ward's minimum variance method, the complete linkage method, the single linkage method, median, and centroid. The software, version, and algorithm reference should be provided for completeness.Supplemental Figure 4 should have the clusters indicated on the figure (boxes or something).Supplemental Table 1I believe supplemental table 1 could benefit from including:a description of what each column is beyond the title (for example, what is the difference between RESmall and RE??)a description of the PP that are included (it appears there are only 11 that end up being used)an indication of which are PRE, PP-PRE, and TMR-PREDescribing REGEXE'sI use regular expressions regularly, but even still I found it difficult to follow the regular expressions listed in the text without looking to a reference. A short description, even in the supplemental materials of general features of the regexes would be useful. For example, the fact that [ABC] means one of either A or B or C at that position, that {3, 8} means either 3 or 8 letters between the previous and the next thing, and that [^ABC] means none of either A or B or C.Further, having examples of what portion of a sequence is matched, especially for the serine-threonine case where the sequence interval in general overlaps between the PROSITE pattern and the PILGram derived pattern. But also having examples for the Zinc finger showing the attributes shown, or describing what part of the regex encodes which features would help a lot.Language implying other methods are not linguisticIn its current form, the abstract reads:\"In this paper we describe a linguistic approach to identify ...\"This implies that regular expressions in PROSITE and hidden markov models are not \"linguistic\" approaches. However, in the text, describing regular expressions used by PROSITE as the simplest form of grammar (regular grammars), and Hidden Markov models as a type of regular grammar (Introduction, paragraph 3). If these are grammars, then that implies they are linguistic approaches. In fact, from the description in the manuscript, PILGram generates regular expressions that in some cases are very similar to those used by PROSITE. It seems currently unclear as to how generating regular expressions using PILGram is a \"linguistic\" approach, but aligning and finding common features (as in PROSITE or HMMs) is not. I understand that PILGram is able to generate discriminative regexes without alignment first, and that is very useful (as exemplified by this manuscript), but from the current description that does not make it \"linguistic\". I admit I may be missing something in reading the current text in this area, as I am not a linguist.Other Simple ImprovementsMethods: under PILGram, first sentence, a reference is missing to SIEVE.Methods: PILGram example, example grammar, spaces around symbols would greatly improve the readabilityIn Results: actually identify the core of the PROSITE reg-ex that PILGramis able to capture, noting that PILGram drops the first and last AA in the PROSITEone, and adds Q to the set of alternatives compared to PROSITE.Results: paragraph 2 says \"Supplemental Table 1\", but I believe this should be\"Supplemental Figure 1\".Results: \"We found that applying logistic regression to combine these seven models provided a similar performance as the voting method, but underperformed the logistic regression on the complete set of models.\" how much did it underperform, curious minds want to know??Methods: hot lake peptide data, the wording implies single bacterium genome, however, the Results makes it clear that a metagenome is being used. One of these two sections should be modified to clarify whether it was a single genome or a metagenome. In addition, if these sequences have been submitted, accession numbers should be provided.", "responses": [ { "c_id": "1341", "date": "29 May 2015", "name": "Jason McDermott", "role": "Author Response", "response": "We thank both the reviewers for their insightful and very helpful comments. We have revised the manuscript according to the reviewers’ suggestions and feel that it is substantially improved in terms of clarity and potential for reproducibility. Importantly, we have provided more complete data, results, and code that we employ in the paper.Dr. Flight had a number of points grouped by subject matter so I’ve addressed each of them below using the same organization. Data and code availabilityDr. Flight’s points are very good. We now include the requested datasets in the manuscript and reference them appropriately (see below for details). We have put a script and associated files on Github that represents the MDRpred algorithm as an open source project at: https://github.com/biodataganache/MDRpred The following data files have been added to the manuscript:UniProt ids and sequences have been included for both positive and negative examples for the PS00125 and PS00028 PROSITE patterns.Genome accessions and links to genome annotations for all sequenced genomes in the metagenome used have been provided. Sequence bins (that is, sequences that are specific to a species that hasn’t been sequenced as an axenic culture) are currently being deposited in GenBank and the manuscript will be updated when accession numbers are available.A full list of high-confidence MDR predictions from the metagenome and their annotations are provided.The original PROSITE data records used in our analysis are provided. These both have been updated in PROSITE since our analysis and the numbers of sequences changed then.The lists regular expressions associated with each problem (the two PROSITE patterns and the MDR task) are now provided as text files.As soon as the PILGram code is released we will update the manuscript with a link to the software and citation for the publication. Weak substrate specificity claimWe have updated the manuscript in several places (Introduction, Results, and Discussion) to clarify our claim of substrate specificity. MDRpred predicts substrate specificity at a broader class level, essentially drug-type compound or not. We have included text to explain this distinction and also updated our discussion of specificity in the Results section to make clear that we mean substrate specificity at this broad level. We have also added stronger language about the lack of need for sequence alignment for our method to work, which we agree is one of the major points in the paper. PILGram Algorithmic detailsIn the Methods section we include a paragraph describing how the genetic algorithm operates on parse trees. It is indeed the case that the length of the regular expression is not fixed because genetic algorithm recombinations occur on these trees. We believe that this, combined with the other clarifications of the method now included in the revision, should adequately resolve this confusion. Physiochemical Properties and TMRTo address Dr. Flight’s comments we have greatly expanded our description of how the physicochemical properties and TMR scoring and grammars work. Also we have examined the contribution of models arising from each of these grammars to the overall method performance. Interestingly, we found that models from each grammar displayed very similar performance independently and each contributed to the final combined performance. We now provide examples of matches and of how the scores are calculated for different sequences and for different physicochemical properties. This was a good suggestion and we believe that the manuscript is really strengthened with these revisions. Elaboration of clusteringThe details of the clustering approach are now described in a new Methods subsection, “Pattern clustering”. Supplemental Table 1Supplemental Table 1 now includes column descriptions, descriptions for the PPs that were included, and a column indicating the source of each pattern (PRE, TMR-PRE, or PP-PRE). Describing REGEXE’sWe have added a subsection to the Methods titled, “Regular Expressions”, which summarizes interpretation of the regular expressions found in the manuscript. We have also added examples for the PROSITE patterns showing which portions of sequences were matched by the PILGram-generated patterns as new Tables 2 and 4 the Results section. The full alignment files are now provided as supplemental data files. Language implying other methods are not linguisticDr. Flight’s point is well-taken. It was not our intent to imply that other approaches, like those we mention (HMMs and PROSITE), are not derived from linguistics. We have revised the text throughout to make clear that other currently used bioinformatics methods are also derived from linguistics. Simple improvementsAll addressed." } ] }, { "id": "7890", "date": "25 Mar 2015", "name": "David Anthony Baltrus", "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\nGiven the growing problem of antibiotic resistance across bacterial pathogens, Multi Drug Resistant (MDR) transporters are an intrinsically important group of bacterial proteins. However, unlike other resistance protein families where precise characterization is possible (i.e. B-lactamases), and while we can often \"see\" MDR transporters in bacterial genomes due to sequence similarity, it is nearly impossible at the present time to accurately annotate what antibiotic substrates these transporters act on. As genome sequences pile up, and whole genome sequences begin to be used to predict drug resistances/sensitivities in clinical settings, the it becomes increasingly important to accurately describe the roles of MDR transporter complexes.The article from McDermott et al. focuses on using grammar based alignment free approaches in order to predict and classify MDR protein complexes, and develops a program called PILGram for this purpose. The authors do a good job of describing the problem they are addressing throughout the introduction, and giving examples of the utility of grammar based approaches to an audience that is likely not well versed in these analyses. Realistically, the results are not exceptional. The model does a great job of predicting ser/thr phosphatase patterns, but current approaches using similarity based searches do a pretty good job as well. The model does slightly worse than conventional methods with the prediction of zinc-fingers, likely because of their unstructured nature, but taking the consensus using grammar based approaches is still on par with other widely used methods. The authors don't improve on predictions for these two classes with grammar based methods, but they provide a good demonstration that such models can work on par with conventional analyses. I think it's important to develop both sequence based and sequence independent approaches and that these go hand in hand rather than act in a mutually exclusive way.The rubber meets the road when the authors try to predict novel MDR classes, and the results are not great. While numerically, the data seems to show that PILGram is able to be trained to identify MDR transporters with levels of accuracy above randomness, it misses a lot. On the other hand, so do conventional analyses, which is what makes this an interesting problem to tackle. Moreover, the authors use a metagenomic dataset (nicely done by the way, I wasn't expecting that) to try and predict novel MDR transporters. The data do suggest that PILGram can pick up something of a signal within these metagenomes compared to a soil sample, which is encouraging. However, I'm left with a bit of an unenthusiastic taste in my mouth when I see the table of \"high confidence novel transporter proteins\" and 3 of the 5 are annotated as some kind of transporters, and the other two are FtsH and a related protein. The authors do point out that it's likely that at least one of these is a true positive (my guess, it's not either of the last two), but it would be good in the discussion if the authors could further flesh out what differentiates the data that PILGram is giving you from simply looking through the annotations for \"transporter\" proteins given that 3 of 5 are likely transporting something based on the JGI annotation. Said more plainly, it would be good if the authors could describe what PILGram is telling them about the first three genes in table 4 that the annotations don't. I think this would really wrap the story up better.My overall impression is that this is a solid paper, albeit without really exceptional results. However, utilization of these sequence alignment independent grammar models and pipelines and descriptions for how they behave on real world data is a step forward and therefore worthy of being published. The data is solid, and the authors do a good job of describing the limitations. We need anything and everything possible to be able to predict MDR proteins given the large amounts of genome data that are going to be piling up. PILGram will only get better with larger training sets.Slight side note...I'm wondering whether glc-1 from C.elegans should be included in the training set for the \"Prediction of multi-drug resistance transporters dataset\" table. Seems weird to me given that these are bacterial proteins.", "responses": [ { "c_id": "1340", "date": "29 May 2015", "name": "Jason McDermott", "role": "Author Response", "response": "We thank both the reviewers for their insightful and very helpful comments. We have revised the manuscript according to the reviewers’ suggestions and feel that it is substantially improved in terms of clarity and potential for reproducibility. Importantly, we have provided more complete data, results, and code that we employ in the paper.Dr. Baltrus points out that two of the high-confidence predictions made by the method are annotated as non-transporter proteins. This issue arises from the fact that a preliminary screen was done on the metagenome to identify transporters (in the ‘Identification of novel MDR transporter candidates from environmental microbiomes’) using Pfam families. As explained in the paper this is necessary because MDRpred was trained only on transporter proteins and so may give spurious results when applied to non-transporters. However, it looks like the FtsH sequence was erroneously included as a transporter- probably because it has an ABC-associated ATPase domain. The other protein, listed as a “Bacterial cell division membrane protein”, has a membrane domain associated with O-antigen, but this appears to be involved in synthesis of O-antigen and not transport. These Pfam families have both been removed from our list for transporters, which is now provided as a data file. Table 4 (now Table 6) has been updated by removing these two predictions and a complete set of higher-confidence predictions is now included as Supplemental Data file. Dr. Baltrus also asked for clarification of what MDRpred would be giving beyond examining annotations for “transporter” proteins. The value of MDRpred is that it will predict which transporter proteins are capable of transporting a specific class of substrates, antibiotic drug compounds. As we point out in the text this is a broad class of compounds and is often incompletely defined for individual well-studied transporters. However, out method is able to accurately classify MDR transporters relative to other transporters that do not transport drug compounds. Even in the case of the first and third predictions (annotated as arabinose and lipid transporters, respectively) the specific annotations are based on best matches by Pfam or BLAST, and may not accurately represent the substrates that are actually transported depending on how close the matches are. We now include an extended discussion of the interpretation of the list of proteins found in Table 4 (now Table 6) in the Results section. Glc-1, a glutamate gated transporter, can confer drug resistance in C. elegans, though it appears that it does not transport drugs itself. We’ve removed it from the training set." } ] } ]	1	https://f1000research.com/articles/4-60
https://f1000research.com/articles/4-132/v1	28 May 15	{ "type": "Review", "title": "Defining the functional states of Th17 cells", "authors": [ "Youjin Lee", "Vijay Kuchroo", "Youjin Lee" ], "abstract": "The molecular mechanisms governing T helper (Th) cell differentiation and function have revealed a complex network of transcriptional and protein regulators. Cytokines not only initiate the differentiation of CD4 Th cells into subsets but also influence the identity, plasticity and effector function of a T cell. Of the subsets, Th17 cells, named for producing interleukin 17 (IL-17) as their signature cytokine, secrete a cohort of other cytokines, including IL-22, IL-21, IL-10, IL-9, IFNγ, and GM-CSF. In recent years, Th17 cells have emerged as key players in host defense against both extracellular pathogens and fungal infections, but they have also been implicated as one of the main drivers in the pathogenesis of autoimmunity, likely mediated in part by the cytokines that they produce. Advances in high throughput genomic sequencing have revealed unexpected heterogeneity in Th17 cells and, as a consequence, may have tremendous impact on our understanding of their functional diversity. The assortment in gene expression may also identify different functional states of Th17 cells. This review aims to understand the interplay between the cytokine regulators that drive Th17 cell differentiation and functional states in Th17 cells.", "keywords": [ "Th17", "T helper cells", "inflammation", "cytokine signaling" ], "content": "T helper subsets and links to autoimmune inflammation\n\nCD4 T cells are essential architects of host immune defense against pathogens1,2. Collectively, their effector function is mediated in part by a compilation of cytokines that directs differentiation, migration, homeostasis, regulation, and inflammation. Initially, CD4 T helper (Th) cells were grouped into two distinct subsets defined by production of unique cytokines: type 1 helper T cells (Th1) produce IFNγ as their signature cytokine and mediate immune responses against intracellular pathogens, and type 2 helper T cells (Th2) secrete interleukin (IL)-4, IL-5 and IL-13 and drive immune responses against extracellular pathogens, like parasites3. In recent years, the number of unique subsets has grown to include IL-9-producing Th9, follicular T helper cells (Tfh) and IL-17-producing Th17, as well as three subsets of T cells that regulate immune responses, including Type 1 regulatory cells (Tr1), follicular T regulatory cells (TfR) and T regulatory cell (Tregs) (Figure 1)4–6. Each of the effector subsets is not only critical for orchestrating a proper immune response against specific pathogens but is also a major contributor in the pathogenesis of a number of autoimmune inflammatory diseases7.\n\nCD4 T helper subsets identified with differentiating conditions as well as the signature cytokines they are known to produce. Th17 cells are further subtyped based on cytokine conditions that define pathogenic versus non-pathogenic states.\n\nFor a number of years, IL-12-induced Th1 cells were thought to be the main drivers of autoimmunity, based on the findings that IFNγ-secreting CD4 T cells were frequently found at the site of inflammation and treatment with IFNγ led to exacerbated disease in multiple sclerosis patients8,9. IL-12 is a heterodimeric cytokine composed of two subunits, IL-12p35 and IL-12p40, and is a critical factor for the differentiation of Th1 cells10,11. CD4 T cells express a heterodimeric IL-12 receptor (IL-12R) composed of IL-12Rβ1 and IL-12Rβ2 subunits12,13. Upon exposure to IL-12, the master transcription factor Tbx21 is induced, which transactivates IFNγ and the cells differentiate into Th1 cells14,15. The importance of Th1 cells in autoimmune diseases was further supported by findings that protection from experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis, was observed upon neutralization with anti-IL-12p40 or in IL-12p40–/– mice16,17. However, it became clear that Th1 cells may not be the exclusive drivers for autoimmunity when it was discovered that mice lacking critical components of the Th1 differentiation pathway, such as IFNγ, IFNγR, IL-12Rβ2, and IL-12p35, were highly susceptible to EAE, suggesting that Th1 cells may even be protective in autoimmune diseases18–22.\n\n\nDiscovery of IL-23- and Th17-associated pathogenic inflammation\n\nIn the late 1990s, a novel cytokine called IL-23 that belongs to the IL-12 family of cytokines was discovered23. Interestingly, similar to the functional IL-12 cytokine, IL-23 had an IL-23 p19 subunit, which combined with the IL-12 p40 subunit of IL-12, to develop a functional heterodimeric cytokine24. Loss of either IL-23 p19 or IL-12 p40 chains made mice highly resistant to the development of EAE and other autoimmune diseases, suggesting that IL-23 is a cytokine critical for development of autoimmunity17,25,26. However, unlike IL-12, IL-23 did not induce IFNγ production from naïve CD4 T cells24,27, but it was suggested that IL-23 may be critical for the generation of IL-17-producing Th17 cells. A series of in vitro studies showed that IL-23 could not induce differentiation of naïve T cells into IL-17-producing Th17 cells. In fact, it was discovered that the receptor for IL-23 was not even expressed on naïve CD4 T cells, suggesting that other cytokines besides IL-23 may be critical for the generation of Th17 cells28–30. In fact, we31 and others32,33 showed that Th17 cells are differentiated in the presence of TGF-β1 and IL-6, which resulted in the induction of a unique master transcription factor called RORγt. While IL-23 was not required for the differentiation of Th17 cells, it was revealed to be a critical factor for stabilization of the Th17 phenotype and in evoking pathogenic phenotype in Th17 cells. With ensuing studies it became clear that IL-23, not IL-12, was the critical cytokine for driving autoimmune inflammation. IL-23p19–/–, IL-12p40–/– and IL-23R–/– mice17,25,26 were completely protected from developing a number of murine models of autoimmune diseases including EAE, psoriasis, and colitis. Consistently, Genome Wide Association Scans have reported a strong genetic linkage to single nucleotide polymorphisms (SNP) in IL-23 or IL-23R, with increased susceptibility to several human autoimmune diseases34–40. However, the clearest role of Th17 cells in human autoimmune diseases was supported by clinical studies where neutralization of IL-17 by an anti-IL-17 antibody (Secukimumab) resulted in clinically beneficial results in a number of human autoimmune diseases, including psoriasis, ankylosing spondylitis, and multiple sclerosis41–45.\n\n\nHeterogeneity within the Th17 subset\n\nAlthough Th17 cells have become synonymous with autoimmune tissue inflammation, it is now clear that not all Th17 cells are pathogenic or induce tissue inflammation46. In human inflammatory bowel diseases (IBDs), neutralization of IL-17 or blockade of IL-17 receptor (IL-17Ra) resulted in disease exacerbation, suggesting a possible protective role by Th17 cells47. IL-17-producing T cells that line the gut mucosa do not induce inflammation but have been shown to be necessary to maintain the barrier function of the gut48. Commensal bacteria in the gut may play a critical role in the generation of Th17 cells in the lamina propria and, indeed, there is an absence of IL-17-producing cells in the lamina propria of the small intestines in germ-free mice49,50. There is also evidence suggesting that IL-17 is required to prevent pathologic gut inflammation in a CD4 T cell-mediated transfer model of colitis, as cells lacking the capacity to produce IL-17, or the lack of IL-17R in recipient mice, resulted in exacerbated colitis51,52. These studies alluded to a rather novel concept: that Th17 cells are not uniform in function. In fact, we53 and others54,55 have shown that Th17 cells come in two flavors: one in which they cause pathogenic tissue inflammation and autoimmune disease and the other that is non-pathogenic, in that they fail to provoke autoimmunity, especially in murine T cell models of inflammatory disease (Figure 1)53–55. Th17 cells differentiated in the presence of TGF-β1 and IL-656,57 co-produce IL-17 with IL-10, do not induce tissue inflammation, and in fact may inhibit autoimmune inflammation, and thus are characterized as “non-pathogenic” Th17 cells55. However, upon exposure to IL-23, a “switch” occurs in the Th17 cell transcriptome, which not only allows for stabilization of the Th17 phenotype but also converts non-pathogenic Th17 cells to become pathogenic53,58. These IL-23 experienced Th17 cells have been shown to promote destructive inflammation in numerous T cell-dependent murine models of autoimmunity53,58. IL-23 inhibits IL-10 production and instead promotes secretion of IL-22 and GM-CSF, suggesting that IL-23 drives the development of Th17 cells with unique functional properties59–61. This raises an important question: how does IL-23 induce pathogenicity in Th17 cells? Our studies revealed that IL-23 mediates important changes in the transcriptome of differentiating Th17 cells53. Besides the induction of a number of unique transcription factors, IL-23 induces TGF-β3 production in developing Th17 cells53. We showed that TGF-β3 together with IL-6 in vitro induces differentiation of pathogenic Th17 cells, without any need for further exposure to IL-2353. Similarly, John O’Shea54 and Chen Dong’s62 groups showed that naïve T cells exposed to IL-1β, IL-6 and IL-23 could induce Th17 cells that were highly pathogenic. Thus, by varying the cytokine cocktails in vitro, both pathogenic and non-pathogenic Th17 cells can be generated. Based on these observations, we undertook a systematic transcriptome analysis of Th17 cells in order to develop a novel gene signature that functionally distinguishes Th17 subsets.\n\n\nTranscriptional gene signatures for pathogenic Th17 cells\n\nWhen we compared the gene expression profiles of all known possible in vitro differentiation combinations that induce pathogenic and non-pathogenic Th17 cells, we found 434 genes that were differentially expressed between these different Th17 subtypes53. Of the 434 genes, 233 genes were differentially expressed between highly pathogenic and non-pathogenic Th17 cells53. Based on the biological function, we identified a representative subset of 23 genes that was highly suggestive of driving pathogenicity53. Pathogenic Th17 cells induced expression of various effector molecules that have been shown to be pro-inflammatory, such as Cxcl3, Ccl4, Ccl5, Csf2 (GM-CSF), Il3 (associated with Csf2), Il22, Gzmb (Granzyme B) and, interestingly, transcription factors that are associated with the Th1 phenotype such as Tbx21 (Tbet) and Stat453. Conversely, non-pathogenic Th17 cells revealed a gene signature that included molecules associated with regulation, such as Il10 and transcription factors that regulate IL-10 production, such as Ahr and Maf in addition to Ikzf3 (Aiolos)53. In addition, non-pathogenic Th17 cells express Il1rn (IL-1R antagonist) which might antagonize functions of IL-1 in differentiating Th17 cells into a pathogenic phenotype53. Based on the comparative gene expression profiles between pathogenic and non-pathogenic Th17 cells, our group identified a gene signature that may confer pathogenic phenotype to Th17 cells53.\n\nThe dichotomous nature of Th17 cells may not be a mere in vitro cytokine artifact but may have occurred naturally as a consequence of evolutionary pressures to defend against different types of pathogens. Federica Sallusto’s group was first to show that human Th17 cells producing IL-10 in conjunction with IL-17 have specificity for Staphylococcus aureus infection63. Conversely, Th17 cells that do not produce IL-10, but instead produce IFNγ with IL-17, have specificity for Candida albicans infection, suggesting that, evolutionarily, Th17 cells may have diverged to acquire different cytokine profiles, to become more adept in defense against specific pathogens63. This is in line with clinical observations with immune-deficient patients, where the loss of transcription factor Stat3, which inhibits development of all Th17 cells, results in hyper IgE syndrome and the patients develop rampant C. albicans and S. aureus infections64. Thus, based on our study, we’ve uncovered an interesting overlap in the gene expression profiles of Th17 cells specific for C. albicans or S. aureus in humans with pathogenic versus non-pathogenic Th17 cells in mice. The gene expression profile revealing an IL-17/IFNγ signature which was specific for C. albicans in humans had similarities to more pathogenic pro-inflammatory murine Th17 cells which cause severe EAE. Conversely, the gene profile for IL-17/IL-10-producing Th17 cells specific for S. aureus were comparable to a more non-pathogenic, regulatory gene signature53,54. This was highly suggestive of how evolutionary pressures have fine-tuned different effector cells for clearing different types of pathogens and utilized the same cells for the induction of tissue inflammation or to mediate tissue protection, albeit with small changes in the transcriptome.\n\n\nChallenges in understanding the functional outcome of Th17 heterogeneity\n\nIt has become clear in recent years that Th17 cells may have divergent functions53. We are just beginning to understand the functional consequences of this extensive heterogeneity of Th17 cells65. Though gene expression profiling has endowed us with the ability to identify a signature that distinguishes pathogenic from non-pathogenic Th17 cells53, we do not know how these cells are naturally derived in vivo or what their function is in mediating tissue homeostasis, effector function, inflammation or cancer. For example, do these pathogenic or non-pathogenic Th17 cells develop simultaneously during differentiation in the lymphoid tissue or is there plasticity in the development of Th17 cells such that they can inter-convert based on the environmental cues they receive? Or perhaps there is a sequential development: do non-pathogenic Th17 cells convert into pathogenic Th17 cells during the course of maturation or differentiation? Also, given that the location of the IL-17 producing cells in the peripheral tissue is critical in dictating their function, this raises the issue of how much the peripheral tissue microenvironment alters the developmental programming of Th17 cells. Much remains to be understood in terms of how and why Th17 cells retain heterogeneity and how it influences their functional states.\n\nIn recent years, examination of heterogeneity at a single-cell resolution has become possible by high throughput single-cell RNA sequencing of whole genomes and transcriptomes66,67. Single-cell RNA sequencing allows for profiling and characterization of expression variability on a genomic scale, which provides us with the ability to correlate this genomic heterogeneity with functional differences in Th17 cells68,69. In fact, single-cell RNA sequencing of Th17 cells obtained from different tissues and lymphoid organs is allowing identification of novel regulators of functional states (pathogenic versus non-pathogenic) of Th17 cells (unpublished observation from our lab). Transcriptomic analysis of T cells in the secondary lymphoid organs following activation does provide valuable clues into the differentiation state acquired by the T cells, but it does not identify the functional state that may be attained by Th17 cells upon arrival into the tissue niche. The functional states (pathogenic/non-pathogenic) of the Th17 cells may be partly dependent on the cytokine milieu and tissue microenvironment to which the cells migrate in order to mediate effector functions. Utilizing the pathogenicity gene signature derived from our earlier studies53 as one of the definable parameters used to analyze single-cell sequence data, our lab has discovered that the functional states of Th17 cells may be in constant flux as the T cells mediate tissue inflammation (unpublished observation). Uncovering key regulators that control effector functions of Th17 cells may permit novel treatment approaches for therapeutically inhibiting inflammation without affecting the protective functions of Th17 cells.\n\nHowever, assigning function to these novel regulators will require genetic manipulation undertaken at a large scale. Unfortunately, the only way to confirm the function of a gene is through the use of knockout mice or genetic knockdowns in the cells and disease models65,70. The use of viral vectors or transfection-based si-RNA delivery was not effective in this endeavor, due to the changes in either the differentiation or cell viability induced by these manipulations71,72. Also, to generate a knockout mouse of every novel regulator identified at a single-cell level is an impossible undertaking. To bypass these obvious limitations, our lab in collaboration with Hongkun Park’s lab has developed a novel system of silicon nanowire perturbations where newly discovered candidate genes can be knocked-down at a large scale, which has improved the process of functional validation65. Silicon nanowire perturbation allows for the delivery of siRNA effectively and efficiently into native T cells without the burden of activation or differentiation65,73,74.\n\n\nThe future\n\nArmed with next-generation sequencing and silicon nanowire knockdowns, the pathogenic potential of subpopulations within Th17 cells can be revealed and novel regulators that may drive functional heterogeneity can be effectively established. Understanding the epigenetic and transcriptional controls of various functional states of Th17 cells will undoubtedly reveal new treatment paradigms for autoimmune diseases as well as give us deeper insight into the complex network that drives inflammatory versus tissue-protective functions of Th17 cells.", "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\nReferences\n\nKhader SA, Gopal R: IL-17 in protective immunity to intracellular pathogens. Virulence. 2010; 1(5): 423–427. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhader SA, Guglani L, Rangel-Moreno J, et al.: IL-23 is required for long-term control of Mycobacterium tuberculosis and B cell follicle formation in the infected lung. J Immunol. 2011; 187(10): 5402–5407. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMosmann TR, Coffman RL: TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol. 1989; 7: 145–173. PubMed Abstract \| Publisher Full Text\n\nKing C, Tangye SG, Mackay CR: T follicular helper (TFH) cells in normal and dysregulated immune responses. Annu Rev Immunol. 2008; 26: 741–766. PubMed Abstract \| Publisher Full Text\n\nZhu J, Paul WE: Heterogeneity and plasticity of T helper cells. Cell Res. 2010; 20(1): 4–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZhu J, Yamane H, Paul WE: Differentiation of effector CD4 T cell populations (*). Annu Rev Immunol. 2010; 28: 445–489. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCox CA, Shi G, Yin H, et al.: Both Th1 and Th17 are immunopathogenic but differ in other key biological activities. J Immunol. 2008; 180(11): 7141–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZamvil SS, Steinman L: The T lymphocyte in experimental allergic encephalomyelitis. Annu Rev Immunol. 1990; 8: 579–621. PubMed Abstract \| Publisher Full Text\n\nPanitch HS, Hirsch RL, Haley AS, et al.: Exacerbations of multiple sclerosis in patients treated with gamma interferon. Lancet. 1987; 1(8538): 893–895. PubMed Abstract \| Publisher Full Text\n\nWolf SF, Temple PA, Kobayashi M, et al.: Cloning of cDNA for natural killer cell stimulatory factor, a heterodimeric cytokine with multiple biologic effects on T and natural killer cells. J Immunol. 1991; 146(9): 3074–3081. PubMed Abstract\n\nStern AS, Magram J, Presky DH: Interleukin-12 an integral cytokine in the immune response. Life Sci. 1996; 58(8): 639–654. PubMed Abstract \| Publisher Full Text\n\nChua AO, Chizzonite R, Desai BB, et al.: Expression cloning of a human IL-12 receptor component. A new member of the cytokine receptor superfamily with strong homology to gp130. J Immunol. 1994; 153(1): 128–136. PubMed Abstract\n\nPresky DH, Yang H, Minetti LJ, et al.: A functional interleukin 12 receptor complex is composed of two beta-type cytokine receptor subunits. Proc Natl Acad Sci U S A. 1996; 26(24): 14002–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGately MK, Renzetti LM, Magram J, et al.: The interleukin-12/interleukin-12-receptor system: role in normal and pathologic immune responses. Annu Rev Immunol. 1998; 16: 495–521. PubMed Abstract \| Publisher Full Text\n\nSzabo SJ, Kim ST, Costa GL, et al.: A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell. 2000; 100(6): 655–669. PubMed Abstract \| Publisher Full Text\n\nBrok Herbert PM, van Meurs M, Blezer E, et al.: Prevention of experimental autoimmune encephalomyelitis in common marmosets using an anti-IL-12p40 monoclonal antibody. J Immunol. 2002; 169(11): 6554–6563. PubMed Abstract \| Publisher Full Text\n\nLangrish CL, Chen Y, Blumenschein WM, et al.: IL-23 drives a pathogenic T cell population that induces autoimmune inflammation. J Exp Med. 2005; 201(2): 233–240. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGran B, Zhang G, Yu S, et al.: IL-12p35-deficient mice are susceptible to experimental autoimmune encephalomyelitis: evidence for redundancy in the IL-12 system in the induction of central nervous system autoimmune demyelination. J Immunol. 2002; 169(12): 7104–7110. PubMed Abstract \| Publisher Full Text\n\nBecher B, Durell BG, Noelle RJ: Experimental autoimmune encephalitis and inflammation in the absence of interleukin-12. J Clin Invest. 2002; 110(4): 493–497. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nZhang G, Gran B, Yu S, et al.: Induction of experimental autoimmune encephalomyelitis in IL-12 receptor-beta 2-deficient mice: IL-12 responsiveness is not required in the pathogenesis of inflammatory demyelination in the central nervous system. J Immunol. 2003; 170(4): 2153–2160. PubMed Abstract \| Publisher Full Text\n\nWillenborg DO, Fordham S, Bernard CC, et al.: IFN-gamma plays a critical down-regulatory role in the induction and effector phase of myelin oligodendrocyte glycoprotein-induced autoimmune encephalomyelitis. J Immunol. 1996; 157(8): 3223–3227. PubMed Abstract \| Faculty Opinions Recommendation\n\nFerber IA, Brocke S, Taylor-Edwards C, et al.: Mice with a disrupted IFN-gamma gene are susceptible to the induction of experimental autoimmune encephalomyelitis (EAE). J Immunol. 1996; 156(1): 5–7. PubMed Abstract \| Faculty Opinions Recommendation\n\nOppmann B, Lesley R, Blom B, et al.: Novel p19 protein engages IL-12p40 to form a cytokine, IL-23, with biological activities similar as well as distinct from IL-12. Immunity. 2000; 13(5): 715–725. PubMed Abstract \| Publisher Full Text\n\nLangrish CL, McKenzie BS, Wilson NJ, et al.: IL-12 and IL-23: master regulators of innate and adaptive immunity. Immunol Rev. 2004; 202(1): 96–105. PubMed Abstract \| Publisher Full Text\n\nCua DJ, Sherlock J, Chen Y, et al.: Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature. 2003; 421(6924): 744–748. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nCroxford AL, Mair F, Becher B: IL-23: one cytokine in control of autoimmunity. Eur J Immunol. 2012; 42(9): 2263–2273. PubMed Abstract \| Publisher Full Text\n\nAggarwal S, Ghilardi N, Xie M, et al.: Interleukin-23 promotes a distinct CD4 T cell activation state characterized by the production of interleukin-17. J Biol Chem. 2003; 278(3): 1910–1914. PubMed Abstract \| Publisher Full Text\n\nDiveu C, McGeachy MJ, Cua DJ: Cytokines that regulate autoimmunity. Curr Opin Immunol. 2008; 20(6): 663–668. PubMed Abstract \| Publisher Full Text\n\nMurphy CA, Langrish CL, Chen Y, et al.: Divergent pro- and antiinflammatory roles for IL-23 and IL-12 in joint autoimmune inflammation. J Exp Med. 2003; 198(12): 1951–7. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nUhlig HH, McKenzie BS, Hue S, et al.: Differential activity of IL-12 and IL-23 in mucosal and systemic innate immune pathology. Immunity. 2006; 25(2): 309–318. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nBettelli E, Carrier Y, Gao W, et al.: Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature. 2006; 441(7090): 235–238. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMangan PR, Harrington LE, O’Quinn DB, et al.: Transforming growth factor-beta induces development of the TH17 lineage. Nature. 2006; 441(7090): 231–234. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nVeldhoen M, Hocking RJ, Atkins CJ, et al.: TGFbeta in the context of an inflammatory cytokine milieu supports de novo differentiation of IL-17-producing T cells. Immunity. 2006; 24(2): 179–189. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nCargill M, Schrodi SJ, Chang M, et al.: A large-scale genetic association study confirms IL12B and leads to the identification of IL23R as psoriasis-risk genes. Am J Hum Genet. 2007; 80(2): 273–90. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nRahman P, Inman RD, Gladman DD, et al.: Association of interleukin-23 receptor variants with ankylosing spondylitis. Arthritis Rheum. 2008; 58(4): 1020–1025. PubMed Abstract \| Publisher Full Text\n\nRahman P, Inman RD, Maksymowych WP, et al.: Association of interleukin 23 receptor variants with psoriatic arthritis. J Rheumatol. 2009; 36(1): 137–140. PubMed Abstract \| Publisher Full Text\n\nHuber AK, Jacobson EM, Jazdzewski K, et al.: Interleukin (IL)-23 receptor is a major susceptibility gene for Graves' ophthalmopathy: the IL-23/T-helper 17 axis extends to thyroid autoimmunity. J Clin Endocrinol Metab. 2008; 93(3): 1077–1081. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nAli S, Srivastava AK, Chopra R, et al.: IL12B SNPs and copy number variation in IL23R gene associated with susceptibility to leprosy. J Med Genet. 2013; 50(1): 34–42. PubMed Abstract \| Publisher Full Text\n\nNúñez C, Dema B, Cénit MC, et al.: IL23R: a susceptibility locus for celiac disease and multiple sclerosis? Genes Immun. 2008; 9(4): 289–293. PubMed Abstract \| Publisher Full Text\n\nEirís N, González-Lara L, Santos-Juanes J, et al.: Genetic variation at IL12B, IL23R and IL23A is associated with psoriasis severity, psoriatic arthritis and type 2 diabetes mellitus. J Dermatol Sci. 2014; 75(3): 167–172. PubMed Abstract \| Publisher Full Text\n\nHofstetter HH, Ibrahim SM, Koczan D, et al.: Therapeutic efficacy of IL-17 neutralization in murine experimental autoimmune encephalomyelitis. Cell Immunol. 2005; 237(2): 123–130. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nLuger D, Silver PB, Tang J, et al.: Either a Th17 or a Th1 effector response can drive autoimmunity: conditions of disease induction affect dominant effector category. J Exp Med. 2008; 205(4): 799–810. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGenovese MC, Van den Bosch F, Roberson SA, et al.: LY2439821, a humanized anti-interleukin-17 monoclonal antibody, in the treatment of patients with rheumatoid arthritis: A phase I randomized, double-blind, placebo-controlled, proof-of-concept study. Arthritis Rheum. 2010; 62(4): 929–939. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nHuh JR, Leung MW, Huang P, et al.: Digoxin and its derivatives suppress TH17 cell differentiation by antagonizing RORγt activity. Nature. 2011; 472(7344): 486–490. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nXiao S, Yosef N, Yang J, et al.: Small-molecule RORγt antagonists inhibit T helper 17 cell transcriptional network by divergent mechanisms. Immunity. 2014; 40(4): 477–489. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaddur MS, Miossec P, Kaveri SV, et al.: Th17 cells: biology, pathogenesis of autoimmune and inflammatory diseases, and therapeutic strategies. Am J Pathol. 2012; 181(1): 8–18. PubMed Abstract \| Publisher Full Text\n\nHueber W, Sands BE, Lewitzky S, et al.: Secukinumab, a human anti-IL-17A monoclonal antibody, for moderate to severe Crohn's disease: unexpected results of a randomised, double-blind placebo-controlled trial. Gut. 2012; 61(12): 1693–1700. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nKamada N, Seo S, Chen GY, et al.: Role of the gut microbiota in immunity and inflammatory disease. Nat Rev Immunol. 2013; 13(5): 321–335. PubMed Abstract \| Publisher Full Text\n\nAtarashi K, Nishimura J, Shima T, et al.: ATP drives lamina propria TH17 cell differentiation. Nature. 2008; 455(7214): 808–812. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nIvanov II, Frutos Rde L, Manel N, et al.: Specific microbiota direct the differentiation of IL-17-producing T-helper cells in the mucosa of the small intestine. Cell Host Microbe. 2008; 4(4): 337–349. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nO’Connor W Jr, Kamanaka M, Booth CJ, et al.: A protective function for interleukin 17A in T cell-mediated intestinal inflammation. Nat Immunol. 2009; 10(6): 603–609. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nO’Connor W Jr, Zenewicz LA, Flavell RA: The dual nature of TH17 cells: shifting the focus to function. Nat Immunol. 2010; 11(6): 471–476. PubMed Abstract \| Publisher Full Text\n\nLee Y, Awasthi A, Yosef N, et al.: Induction and molecular signature of pathogenic TH17 cells. Nat Immunol. 2012; 13(10): 991–999. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nGhoreschi K, Laurence A, Yang X, et al.: Generation of pathogenic TH17 cells in the absence of TGF-β signalling. Nature. 2010; 467(7318): 967–971. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nMcGeachy MJ, Bak-Jensen KS, Chen Y, et al.: TGF-beta and IL-6 drive the production of IL-17 and IL-10 by T cells and restrain TH-17 cell-mediated pathology. Nat Immunol. 2007; 8(12): 1390–1397. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMcGeachy MJ, Cua DJ: Th17 cell differentiation: the long and winding road. Immunity. 2008; 28(4): 445–453. PubMed Abstract \| Publisher Full Text\n\nKorn T, Bettelli E, Oukka M, et al.: IL-17 and Th17 Cells. Annu Rev Immunol. 2009; 27: 485–517. PubMed Abstract \| Publisher Full Text\n\nJäger A, Dardalhon V, Sobel RA, et al.: Th1, Th17, and Th9 effector cells induce experimental autoimmune encephalomyelitis with different pathological phenotypes. J Immunol. 2009; 183(11): 7169–7177. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCodarri L, Gyülvészi G, Tosevski V, et al.: RORγt drives production of the cytokine GM-CSF in helper T cells, which is essential for the effector phase of autoimmune neuroinflammation. Nat Immunol. 2011; 12(6): 560–567. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nDuvallet E, Semerano L, Assier E, et al.: Interleukin-23: a key cytokine in inflammatory diseases. Ann Med. 2011; 43(7): 503–511. PubMed Abstract \| Publisher Full Text\n\nEl-Behi M, Ciric B, Dai H, et al.: The encephalitogenicity of TH17 cells is dependent on IL-1-and IL-23-induced production of the cytokine GM-CSF. Nat Immunol. 2011; 12(6): 568–575. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nChung Y, Chang SH, Martinez GJ, et al.: Critical regulation of early Th17 cell differentiation by interleukin-1 signaling. Immunity. 2009; 30(4): 576–587. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nZielinski CE, Mele F, Aschenbrenner D, et al.: Pathogen-induced human TH17 cells produce IFN-γ or IL-10 and are regulated by IL-1β. Nature. 2012; 484(7395): 514–518. PubMed Abstract \| Publisher Full Text \| Faculty Opinions Recommendation\n\nMa CS, Chew GY, Simpson N, et al.: Deficiency of Th17 cells in hyper IgE syndrome due to mutations in STAT3. J Exp Med. 2008; 205(7): 1551–1557. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nYosef N, Shalek AK, Gaublomme JT, et al.: Dynamic regulatory network controlling TH17 cell differentiation. Nature. 2013; 496(7446): 461–468. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nKalisky T, Blainey P, Quake SR: Genomic analysis at the single-cell level. Annu Rev Genet. 2011; 45: 431–445. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlainey PC, Quake SR: Dissecting genomic diversity, one cell at a time. Nat Methods. 2014; 11(1): 19–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrombetta JJ, Gennert D, Lu D, et al.: Preparation of Single-Cell RNA-Seq Libraries for Next Generation Sequencing. Curr Protoc Mol Biol. 2014; 107: 4.22.1–4.22.17. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nShalek AK, Satija R, Adiconis X, et al.: Single-cell transcriptomics reveals bimodality in expression and splicing in immune cells. Nature. 2013; 498(7453): 236–240. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nYosef N, Zalckvar E, Rubinstein AD, et al.: ANAT: a tool for constructing and analyzing functional protein networks. Sci Signal. 2011; 4(196): pl1. PubMed Abstract \| Publisher Full Text\n\nDardalhon V, Herpers B, Noraz N, et al.: Lentivirus-mediated gene transfer in primary T cells is enhanced by a central DNA flap. Gene Ther. 2001; 8(3): 190–198. PubMed Abstract \| Publisher Full Text\n\nMcManus MT, Haines BB, Dillon CP, et al.: Small interfering RNA-mediated gene silencing in T lymphocytes. J Immunol. 2002; 169(10): 5754–5760. PubMed Abstract \| Publisher Full Text\n\nShalek AK, Robinson JT, Karp ES, et al.: Vertical silicon nanowires as a universal platform for delivering biomolecules into living cells. Proc Natl Acad Sci U S A. 2010; 107(5): 1870–1875. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation\n\nShalek AK, Gaublomme JT, Wang L, et al.: Nanowire-mediated delivery enables functional interrogation of primary immune cells: application to the analysis of chronic lymphocytic leukemia. Nano lett. 2012; 12(12): 6498–6504. PubMed Abstract \| Publisher Full Text \| Free Full Text \| Faculty Opinions Recommendation" }	[ { "id": "8799", "date": "28 May 2015", "name": "Richard Flavell", "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", "responses": [] }, { "id": "8800", "date": "28 May 2015", "name": "Jagadeesh Bayry", "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", "responses": [] }, { "id": "8801", "date": "28 May 2015", "name": "Dan R. Littman", "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", "responses": [] } ]	1	https://f1000research.com/articles/4-132

Subsets and Splits

No community queries yet

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