link
stringlengths 41
45
| date
stringlengths 9
9
| paper
dict | reviews
listlengths 1
6
| version
int64 1
5
| main
stringlengths 38
42
|
|---|---|---|---|---|---|
https://f1000research.com/articles/7-569/v1
|
10 May 18
|
{
"type": "Research Article",
"title": "Does gonorrhoea screening intensity play a role in the early selection of antimicrobial resistance in men who have sex with men (MSM)? A comparative study of Belgium and the United Kingdom",
"authors": [
"Chris R. Kenyon",
"Irith De Baetselier",
"Tania Crucitti",
"Irith De Baetselier",
"Tania Crucitti"
],
"abstract": "Background: It is unclear why antimicrobial resistance in Neisseria gonorrhoeae in the United Kingdom (UK) and the United States has tended to first appear in men who have sex with men (MSM). We hypothesize that increased exposure to antimicrobials from intensive STI screening programmes plays a role. Methods: We assess if there is a difference in the distribution of azithromycin, cefixime and ceftriaxone minimum inhibitory concentrations (MICs) between MSM and women in the United Kingdom (UK) where 70% of MSM report STI screening in the past year vs. Belgium where 9% report STI screening in the past year. Our hypothesis is that MICs of the MSM should be higher than those of the women in the UK but not Belgium. Data for the MICs were taken from the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) in the UK in 2010/2011 and 2014 and a similar national surveillance programme in Belgium in 2013/2014 (the first most complete available data). We used the Mann–Whitney test to compare the MIC distributions between MSM and women within each country Results: In the UK the MICs for all three antimicrobials were significantly higher in MSM than women at both time points (P all <0.0005). In Belgium only the MIC distribution for azithromycin was higher in MSM (P<0.0005). Conclusion: The findings for cefixime and ceftriaxone, but not azithromycin are compatible with our hypothesis that screening-intensity could contribute to the emergence of AMR. Numerous other interpretations of our results are discussed.",
"keywords": [
"Neisseria gonorrhoeae",
"antimicrobial resistance",
"screening"
],
"content": "Introduction\n\nA striking feature of the patterning of antimicrobial resistance (AMR) is how it has repeatedly emerged in core-groups, either sex workers or men who have sex with men (MSM) with high rates of partner change1. In the last two decades AMR in the United Kingdom (UK) and the United States (USA) has tended to first appear in MSM2–5. In the UK for example, the prevalence of cefixime resistance (following the switch to cefixime therapy for Neisseria gonorrhoeae (NG) in 2005) increased from 0% in 2005 to 33.1% in 2010 in MSM, whilst remaining under 7% in heterosexual men and women (Figure 1)2. In the USA, UK and the Netherlands, the prevalence of AMR to at least one of ciprofloxacin/cefixime/cefotaxime/azithromycin has been noted to be higher in MSM3,4,6. This association has not, however, been found in other countries. An analysis of gonococcal AMR in the 24 countries participating in European Gonococcal Antimicrobial Surveillance Programme (Euro-GASP) in 2015, for example, found that cefixime and ciprofloxacin resistance rates were not higher in MSM compared to heterosexual men7. Azithromycin (AZM) resistance prevalence was however higher in men (both MSM and heterosexuals) than women.\n\nWe hypothesize that these differences in the emergence of AMR may be in part explained by differences in the intensity of NG/CT (Chlamydia trachomatis) screening for MSM. The percent of MSM who report being screened for NG/CT varies considerably between countries. In the 38 countries in the European MSM Internet Survey, for example the proportion of MSM who reported anal screening for sexually transmitted infections (STIs) ranged from 4.4% in Serbia to 70.6% in Malta (median 16.0, IQR 13.5-28.4)8. A higher screening intensity would be expected to translate into greater antimicrobial exposure. A study that modelled the sexual network of a population of Belgian MSM, for example, found that increasing screening intensity from 3.5% to 50% of MSM annually would reduce NG prevalence marginally but at the expense of a 12-fold increase in antimicrobial exposure8,9.\n\nIn this preliminary study to test the hypothesis that screening intensity played a role in the selection of AMR in NG we contrast the difference in azithromycin, cefixime and ceftriaxone minimum inhibitory concentration (MIC) distributions between MSM and women in the UK (an intensive-screening country; 70% of MSM report annual STI screening8) with those in Belgium (a low-screening country; 9% of MSM report annual STI screening8) in the years 2010 to 2015. The overall consumption of these antimicrobials in these two countries was not too dissimilar. Between 2010 and 2015, Belgians consumed slightly more cephalosporins but fewer macrolides than the inhabitants of the UK (cephalosporins: 966 vs. 905 standard units per 1000/population/year; macrolides 1960 vs. 3063 standard units per 1000/population/year, respectively10).\n\nIn Belgium, guidelines changed from ciprofloxacin to ceftriaxone 125mg IM or spectinomycin 2g IM in 200811,12. In 2012 azithromycin was added for treatment of NG and ceftriaxone dosage was increased: ceftriaxone 500mg IM plus azithromycin 2g PO13,14. In the UK, cefixime 400mg PO took over from ciprofloxacin in 2005 as preferred therapy3. In 2011, this was switched to ceftriaxone 500mg IM plus azithromycin 1g PO3,15. Thus between 2008 and 2012 therapy in Belgium/the UK was mostly ceftriaxone/cefixime whereas from 2012 dual therapy was recommended in both countries.\n\n\nMethods\n\nBecause the sampling and susceptibility testing methodologies vary slightly between Belgium and the UK, we do not directly compare the MICs between the two countries. Rather we assess if there is a difference in the distribution of MICs between MSM and women in each country. The rationale we use is as follows. If intensive screening in MSM plays a role in generating AMR in MSM then in the intensive-screening country we would expect to find a shift in distribution towards higher MICs in MSM compared to women for the antimicrobials used as treatment in the screening programme. In the low-screening country there should be no difference in distribution between MSM and women. We compare MSM with women rather than heterosexual men to avoid the problem of misclassification of men who occasionally have sex with men but regard themselves as heterosexual16.\n\nAMR surveillance in Belgium: All laboratories in Belgium are requested to send NG isolates to the National Reference Centre for STIs (NRC-STI) at the Institute of Tropical Medicine. The agar dilution method was used to determine MICs according to the CLSI guidelines17.\n\nAMR surveillance in the United Kingdom: The Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) is a sentinel surveillance programme for AMR in NG in the UK. It incorporates a network of genitourinary medicine (GUM) clinics chosen to give regional representation across England and Wales. Isolates from approximately 10% of patients with gonorrhoea, collected over a 3-month period (July–September) each year, undergo susceptibility testing via MIC determination using the agar dilution method at the Public Health England’s sexually transmitted bacteria reference unit (PHE STBRU)18. Demographic and behavioural data are gathered retrospectively and then linked to laboratory data3.\n\nThe data for Belgium was taken directly from NRC-STI. The details regarding sexual orientation started to be reported in sufficient numbers from 2013 onwards. Because the absolute number of isolates from Belgium are low we present analyses from the combined data from 2013 and 2014.\n\nThe data for the UK was extracted from the GRASP annual reports2,5,18. This included digitalization of the percent distribution of MIC by sexual orientation/gender graphs using GetData Graph Digitizer 2.26. We analyze the data of 2010 for ceftriaxone and of 2011 for azithromycin and cefixime, as well as the data of 2014 for the three antimicrobials.\n\nFor the UK data ethics approval was obtained from local regional research committees and from the northwest multicentre research ethics committee3. In Belgium no additional ethical approval was necessary because only fully anonymized routine surveillance data were used.\n\nWe used the Mann–Whitney test to assess if there was a difference in the MIC distributions between MSM and women within each country. Stata 13 was used for all analyses.\n\n\nResults\n\nThe STI reference laboratory received 1224 NG isolates from 78 laboratories in 2013/2014. Of these, 1150 were successfully cultured and tested. 941 (81.8%) were men, 190 (16.5%) women and 19 unknown gender. 183 (19.5%) of the men reported being heterosexual, 201 (21.4%) MSM and data was missing in 557 (59.2%) men.\n\nThe distribution of the azithromycin MICs was significantly higher in MSM compared to women (Median MIC 0.25, [IQR 0.25-0.50] vs. 0.25 [0.125-0.25]; P<0.0005) but there were no differences in the MIC distributions for cefixime or ceftriaxone (Table 1; Figure 2). The MIC distribution for azithromycin was slightly right-shifted in MSM compared to women (Figure 2). The distribution of the MICs for cefixime in women appeared bimodal, as was the MIC distribution for ceftriaxone in women and to a lesser extent in men.\n\n*** P<0.0005 (P-values are from Man-Whitney tests comparing MICs distributions between MSM and women in each country); IQR – Interquartile range\n\nThe number of isolates provided by the GRASP surveys was as follows: 2010: MSM 600, women 306; 2011: MSM 665, women 312; 2014: MSM 1073, women 192. For further details please refer to the individual annual reports2,5,18.\n\n2010–2011: The MIC distributions for all three antimicrobials were statistically significantly higher in MSM than women (Azithromycin: 0.25, [IQR 0.125-0.50] vs. 0.06 [0.03-0.125], cefixime: 0.008, [IQR 0.004-0.03] vs. 0.002 [0.002-0.004] ceftriaxone: 0.008, [IQR 0.004-0.03] vs. 0.008 [0.004-0.008]; All P<0.0005). For all three antimicrobials the distribution was right-shifted in MSM compared to women (Figure 2). The distributions of the MICs for cefixime and ceftriaxone in MSM appeared bimodal.\n\n2014: The MIC distributions for all three antimicrobials were statistically significantly higher in MSM compared to women and were shifted to the right but less so than in 2010 or 2011 (Figure 2, Table 1).\n\nThe distributions of the MICs for cefixime in MSM appeared bimodal, but with a shift to the left of the second mode compared to 2011. The bimodal appearance of the MIC distribution for ceftriaxone in 2014 is less pronounced compared to 2010.\n\n\nDiscussion\n\nA better understanding of the factors underpinning the genesis of AMR in NG could assist with efforts to prevent the further development of AMR. In this study we find that the MIC distribution for azithromycin, ceftriaxone and cefixime (particularly in 2010) is right shifted in MSM compared to women in the UK. In Belgium only the distribution of azithromycin is right-shifted in this way. In addition, we find that the magnitude of this right-shift decreased in the UK between 2010/2011 and 2014. As a result, the proportion of MSM in the UK with higher ceftriaxone MICs and cefixime MICs has declined between 2010 and 2014. These findings are commensurate with UK and European surveillance data showing a decline in the proportion of third generation cephalosporin AMR2,7,15. A plausible reason for this decline has been the introduction of high dose ceftriaxone which has more favourable pharmacokinetic parameters than cefixime2,3,19. Dual therapy with azithromycin may also have played a role7,20.\n\nWhat explains the right-shifting of cefixime and ceftriaxone in MSM versus women in the UK but not Belgium? An important difference in the pharmacoecology experienced by NG in the two countries was the use of cefixime in the UK (until 2011) compared to ceftriaxone monotherapy in Belgium (until 2012). Ceftriaxone's longer half-life than cefixime may have played a role in preventing MIC drift in Belgium19. Both women and men were treated with cefixime in the UK and this would thus not explain why the right-shifting occurred predominantly/only in MSM. The higher NG screening (and therefore antibiotic exposure rates) in MSM in the UK compared to Belgium is one of many possible explanations. This explanation stems from the insight that the intensity of exposure to antimicrobials plays a crucial role in the genesis of AMR21. A range of studies have found close correlations at ecological levels between the intensity of exposure to a particular antimicrobial and AMR to that antimicrobial22–26.\n\nArguing against the screening-intensity explanation is the fact that the right shifting of AZM occurred in MSM in both countries. This finding suggests either that some other factor is responsible for the right shifting in MSM (such as total macrolide use for all indications) or that the MSM sexual pharmacoecology is more susceptible to the development of AMR for azithromycin than other antibiotics27. The higher proportion of time NG spends in the rectum in MSM compared to heterosexual sexual networks, for example, could lead to an enhanced selection pressure for/availability of mtrR-related and erm mutations27. Macrolides have also been shown to have a particularly long adverse effect on the resistome, with changes noted for up to 4 years post therapy28,29. These considerations may mean that relatively low azithromycin exposure may be sufficient to generate a right shift in MIC.\n\nWe also observed changes in the bimodal distribution of cetriaxone and cefixime in 2014 versus 2010–2011 in the UK. The shift to the left of the second mode and almost disappearance of the bimodal distribution is reassuring as it may indicate that the previous emergence of a less susceptible population is temporarily under control and regaining susceptibility towards cefixime and ceftriaxone.\n\nThere are a number of alternative explanations for why AMR may arise sooner in MSM than women. MSM may be more likely to travel abroad and acquire more resistant NG in this way4,30. MSM are more likely to be HIV-infected and may as a result use more antimicrobials4. Some studies have found that even after stratifying for HIV-infection status, MSM still report consuming more antimicrobials4. Both treatment of symptomatic and asymptomatic STIs may play a role here. Finally, the fact that NG spends proportionately more time in the oropharynx and rectum in MSM (compared to heterosexuals) may offer it more opportunities for acquisition of resistance genes and mutations4,27. It is however unlikely that these explanations can explain the differences between NG AMR in MSM vs. women in the UK compared to Belgium.\n\nThe numerous weaknesses of our study design preclude firm conclusions. These limitations include the fact that we only include two countries, and we have limited data on the full range of potential explanatory variables (such as general antimicrobial consumption, NG/CT screening rates in women). There were also important methodological differences in how the surveillance was conducted in the two countries (such as sampling methodology, sensitivity testing). Whereas the GRASP sentinel methodology has been shown to yield fairly representative samples for the UK31, an equivalent study has not been conducted in Belgium.\n\nAlthough we cannot, on the basis of this study, conclude that the intensity of NG screening plays a role in the genesis of AMR in NG we also cannot reject this hypothesis. Further studies that could test this hypothesis include: 1) assessing the correlation between NG screening intensity in MSM and the prevalence of AMR in MSM in a greater number of countries; 2) community level randomized controlled trials assessing the impact of NG/CT screening on AMR and NG prevalence and 3) more detailed longitudinal assessments of the effects of repeated antibiotic exposure on the resistome and microbiome of MSM cohorts with higher risk behaviour32.\n\n\nData availability\n\nDataset 1: Minimum inhibitory concentrations distributions for Neisseria gonorrhoeae isolates analyzed 10.5256/f1000research.14869.d20317333",
"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\nLewis DA: The role of core groups in the emergence and dissemination of antimicrobial-resistant N gonorrhoeae. Sex Transm Infect. 2013; 89 Suppl 4: iv47–51. PubMed Abstract | Publisher Full Text\n\nPublic Health England: Surveillance of antimicrobial resistance in Neisseria gonorrhoeae: Key findings from the ‘Gonococcal resistance to antimicrobials surveillance programme' (GRASP) and related surveillance data. Public Health England: London. 2011.\n\nIson CA, Town K, Obi C, et al.: Decreased susceptibility to cephalosporins among gonococci: data from the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) in England and Wales, 2007-2011. Lancet Infect Dis. 2013; 13(9): 762–8. PubMed Abstract | Publisher Full Text\n\nKirkcaldy RD, Zaidi A, Hook EW 3rd, et al.: Neisseria gonorrhoeae Antimicrobial Resistance Among Men Who Have Sex With Men and Men Who Have Sex Exclusively With Women: The Gonococcal Isolate Surveillance Project, 2005-2010. Ann Intern Med. 2013; 158(5 Pt 1): 321–8. PubMed Abstract | Publisher Full Text\n\nPublic Health England: Surveillance of antimicrobial resistance in Neisseria gonorrhoeae: Key findings from the ‘Gonococcal resistance to antimicrobials surveillance programme' (GRASP) and related surveillance data. Public Health England: London. 2010.\n\nde Vries HJ, Van der Helm JJ, Schim van der Loeff MF, et al.: Multidrug-resistant Neisseria gonorrhoeae with reduced cefotaxime susceptibility is increasingly common in men who have sex with men, Amsterdam, the Netherlands. Euro Surveill. 2009; 14(37): 3, pii: 19330. PubMed Abstract | Publisher Full Text\n\nCole MJ, Spiteri G, Jacobsson S, et al.: Overall Low Extended-Spectrum Cephalosporin Resistance but high Azithromycin Resistance in Neisseria gonorrhoeae in 24 European Countries, 2015. BMC Infect Dis. 2017; 17(1): 617. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThe EMIS Network: EMIS 2010: The European Men-Who-Have-Sex-With-Men Internet Survey. Findings from 38 countries. Stockholm: European Centre for Disease Prevention and Control, 2011. Reference Source\n\nBuyze J, Vanden Berghe W, Hens N, et al.: Current levels of gonorrhoea screening in MSM in Belgium may have little effect on prevalence: a modelling study. Epidemiol Infect. 2018; 146(3): 333–338. PubMed Abstract | Publisher Full Text\n\nThe Center For Disease Dynamics Economics & Policy: Antibiotic use. 2018. Reference Source\n\nBelgische gids voor anti-infectieuze behandeling in de ambulante praktijk. editie 2006 [Internet]. Reference Source\n\nBelgische gids voor anti-infectieuze behandeling in de ambulante praktijk. 2008. editie 2nd ed [Internet]. Reference Source\n\nBelgische gids voor anti-infectieuze behandeling in de ambulante praktijk. editie 2012. [Internet]. Reference Source\n\nSchweikardt C, Goderis G, Elli S, et al.: Prescription of Antibiotics to Treat Gonorrhoea in General Practice in Flanders 2009-2013: A Registry-Based Retrospective Cohort Study. J Sex Transm Dis. 2017; 2017: 1860542. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClifton S, Bolt H, Mohammed H, et al.: Prevalence of and factors associated with MDR Neisseria gonorrhoeae in England and Wales between 2004 and 2015: analysis of annual cross-sectional surveillance surveys. J Antimicrob Chemother. 2018; 73(4): 923–932. PubMed Abstract | Publisher Full Text\n\nHué S, Brown AE, Ragonnet-Cronin M, et al.: Phylogenetic analyses reveal HIV-1 infections between men misclassified as heterosexual transmissions. AIDS. 2014; 28(13): 1967–75. PubMed Abstract | Publisher Full Text\n\nClinical and Laboratory Standards Institute (CLSI): Performance standards for antimicrobial susceptibility testing; seventeenth informational supplement. CLSI document M100-S1. Wayne, PA, USA: CLSI, 2007. Reference Source\n\nPublic Health England: Surveillance of antimicrobial resistance in Neisseria gonorrhoeae: Key findings from the ‘Gonococcal resistance to antimicrobials surveillance programme' (GRASP) and related surveillance data. Public Health England: London. 2014.\n\nChisholm SA, Mouton JW, Lewis DA, et al.: Cephalosporin MIC creep among gonococci: time for a pharmacodynamic rethink? J Antimicrob Chemother. 2010; 65(10): 2141–8. PubMed Abstract | Publisher Full Text\n\nKirkcaldy RD, Bartoces MG, Soge OO, et al.: Antimicrobial Drug Prescription and Neisseria gonorrhoeae Susceptibility, United States, 2005-2013. Emerg Infect Dis. 2017; 23(10): 1657–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCantas L, Shah SQ, Cavaco LM, et al.: A brief multi-disciplinary review on antimicrobial resistance in medicine and its linkage to the global environmental microbiota. Front Microbiol. 2013; 4: 96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBruyndonckx R, Hens N, Aerts M, et al.: Exploring the association between resistance and outpatient antibiotic use expressed as DDDs or packages. J Antimicrob Chemother. 2015; 70(4): 1241–4. PubMed Abstract | Publisher Full Text\n\nGoossens H, Ferech M, Vander Stichele R, et al.: Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet. 2005; 365(9459): 579–87. PubMed Abstract\n\nRiedel S, Beekmann SE, Heilmann KP, et al.: Antimicrobial use in Europe and antimicrobial resistance in Streptococcus pneumoniae. Eur J Clin Microbiol Infect Dis. 2007; 26(7): 485–90. PubMed Abstract | Publisher Full Text\n\nBronzwaer SL, Cars O, Buchholz U, et al.: A European study on the relationship between antimicrobial use and antimicrobial resistance. Emerg Infect Dis. 2002; 8(3): 278–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHicks LA, Chien YW, Taylor TH Jr, et al.: . Outpatient antibiotic prescribing and nonsusceptible Streptococcus pneumoniae in the United States, 1996–2003. Clin Infect Dis. 2011; 53(7): 631–9. PubMed Abstract | Publisher Full Text\n\nKenyon C, Osbak K: Certain attributes of the sexual ecosystem of high-risk MSM have resulted in an altered microbiome with an enhanced propensity to generate and transmit antibiotic resistance. Med Hypotheses. 2014; 83(2): 196–202. PubMed Abstract | Publisher Full Text\n\nJakobsson HE, Jernberg C, Andersson AF, et al.: Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome. PLoS One. 2010; 5(3): e9836. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJernberg C, Lofmark S, Edlund C, et al.: Long-term ecological impacts of antibiotic administration on the human intestinal microbiota. ISME J. 2007; 1(1): 56–66. PubMed Abstract | Publisher Full Text\n\nMatteelli A, Carosi G: Sexually transmitted diseases in travelers. Clin Infect Dis. 2001; 32(7): 1063–7. PubMed Abstract | Publisher Full Text\n\nHughes G, Nichols T, Ison CA: Estimating the prevalence of gonococcal resistance to antimicrobials in England and Wales. Sex Transm Infect. 2011; 87(6): 526–31. PubMed Abstract | Publisher Full Text\n\nKenyon C: Risks of antimicrobial resistance in N. gonorrhoeae associated with intensive screening programs in PrEP programs. Clin Infect Dis. 2018. PubMed Abstract | Publisher Full Text\n\nKenyon C, De Baetselier I, Crucitti T: Dataset 1 in: Does gonorrhoea screening intensity play a role in the early selection of antimicrobial resistance in men who have sex with men (MSM)? A comparative study of Belgium and the United Kingdom. F1000Research. 2018. Data Source"
}
|
[
{
"id": "34584",
"date": "14 Jun 2018",
"name": "Henry J C de Vries",
"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 nicely written paper proposing a new and refreshing hypothesis that gonorrhoea screening in a larger proportion of a certain population and subsequent treatment could induce AMR. Although the authors admit that they cannot confirm their hypothesis, they do claim to see an association in support of their claim. There is a multitude of other explanations for the association found which aren't properly discussed and more sound evidence is needed to confirm their statement. It is of interest though to report on their findings.\n\nI have some comments to consider though.\nIn the method section it is stated that MSM were compared with women rather than heterosexual men to avoid the problem of misclassification of men who occasionally have sex with men but regard themselves as heterosexual. In the same fashion it is possible that heterosexual women might have sex with bisexual males and thus be exposed to the MSM pharmacoecology described here. Please consider the effect of this option in the light of choosing the control group.\n\nThe MIC right shift for all 3 antibiotics has decreased from 2010/11 to 2014 in the UK this is attributed to higher dosages of cephalosporins given and the addition of azi to the recommended therapy. This finding can be interpreted as an argument against the hypothesis of the authors; correct treatment of a confirmed infection does not lead to the induction of AMR since the strain is eradicated and cannot develop AMR.\n\nIt is stated also that ceftriaxone's longer half-life than cefixime may have played a role in preventing MIC drift in Belgium. This is counter intuitive. A longer antimicrobial half life is associated with the induction of AMR due to the prolonged exposure of bacteria to sub therapeutic concentrations of antibiotic during re-exposure. See also: Decreased Azithromycin Susceptibility of Neisseria gonorrhoeae Isolates in Patients Recently Treated with Azithromycin. Wind CM, de Vries E, Schim van der Loeff MF, van Rooijen MS, van Dam AP, Demczuk WHB, Martin I, de Vries HJC. Clin Infect Dis. 2017 Jul 1;65(1):37-45. [Ref-1]\n\nIn the second paragraph of the discussion it isn't mentioned that: “This explanation stems from the insight that the intensity of exposure to antimicrobials plays a crucial role in the genesis of AMR [ref 21]. Here, the reference is misquoted, Cantas et al specifically address the non-therapeutic and low-level dosage use of antimicrobials that lead to AMR induction. This is not the case in the UK setting where MSM are treated with therapeutic dosages, and only after infection has been confirmed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3743",
"date": "19 Jun 2018",
"name": "Chris Kenyon",
"role": "Author Response",
"response": "Dear Prof. de VriesThank you for your useful suggestions which we respond to below:1. This is a valid concern. Repeating the analyses using heterosexual men instead of women as the control group makes little difference to the MSM vs. control MIC frequency distribution curves. This is evident if one looks at the individual GRASP reports. We can share the figures for Belgium if there is interest in this.2. and 3. It is true that we do not know why the right shift has declined. It is also true that a long half of an antibiotic is frequently associated with the induction of resistance. Azithromycin is a good example of this as noted in the reference you refer to. We have argued elsewhere that a N. gonorrhoeae (Ng) infection during the long declining half life of azithromycin is a plausible risk factor for inducing AMR (ref 27). This is however quite different to what is being argued here. If an antibiotic does not attain a Ng requires a free time above MIC for cephalosporins for at least 10-20 hours. If this is not attained this will place a selection pressure to develop AMR. Because the half life of cefixime is shorter than that of ceftriaxone (3.4 hours vs. 8.45 hours) there is a higher risk of not attaining the required free time above MIC and thereby selecting for AMR (doi:10.1093/jac/dkq289). Other factors such as the ratio of the Mutant Prevention Concentration to MIC ratio may also play a role but ultimately we do not know with certainty the reasons why cefixime is more selective for resistance.What we do have good experimental evidence for however is that for a number of bug-drug combinations cefixime is more prone to AMR than other third generation cephalosporins. One of the most convincing studies of this was an in a in vitro differential selection study by Negri et al., who found that cefixime was the best selector of penicillin resistance in Streptococcus pneumoniae (compared to amoxicillin, cefuroxime and cefotaxime (PMID: 8141563). The mechanism underpinning this effect has not been clearly elucidated but a number of authors have speculated that it may be related at least in part to cefixime's shorter half life. We will add this discussion to the next version of the paper.4. In the next version of the paper we will include the references listed below to better back up this claim. The Cantas reference should however remain as it is a useful overview of the importance of considering total antimicrobial consumption in an ecosystem perspective. The main pathway from high antimicrobial consumption to AMR is not via subtherapeutic dosing but rather factors such as antimicrobial induced changes to the resistome and microbiome which can then be taken up by Ng via mechanisms such as transformation, plasmids. This and other mechanisms are outlined in the references below as well as refs 27-29 and 32 above.Hicks LA, Chien YW, Taylor TH, Jr., Haber M, Klugman KP, Active Bacterial Core Surveillance T. Outpatient antibiotic prescribing and nonsusceptible Streptococcus pneumoniae in the United States, 1996-2003. Clin Infect Dis. 2011;53(7):631-9. Epub 2011/09/06. doi: 10.1093/cid/cir443. PubMed PMID: 21890767.Bronzwaer SL, Cars O, Buchholz U, Molstad S, Goettsch W, Veldhuijzen IK, et al. A European study on the relationship between antimicrobial use and antimicrobial resistance. Emerg Infect Dis. 2002;8(3):278-82. Epub 2002/04/03. doi: 10.3201/eid0803.010192. PubMed PMID: 11927025; PubMed Central PMCID: PMCPMC2732471.Riedel S, Beekmann SE, Heilmann KP, Richter SS, Garcia-De-Lomas J, Ferech M, et al. Antimicrobial use in Europe and antimicrobial resistance in Streptococcus pneumoniae. European Journal of Clinical Microbiology & Infectious Diseases. 2007;26(7):485-90. doi: 10.1007/s10096-007-0321-5. PubMed PMID: WOS:000247930300007.Goossens H, Ferech M, Vander Stichele R, Elseviers M, Group EP. Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet. 2005;365(9459):579-87. Epub 2005/02/15. doi: 10.1016/S0140-6736(05)17907-0. PubMed PMID: 15708101.Bruyndonckx R, Hens N, Aerts M, Goossens H, Cortinas Abrahantes J, Coenen S. Exploring the association between resistance and outpatient antibiotic use expressed as DDDs or packages. J Antimicrob Chemother. 2015;70(4):1241-4. Epub 2015/01/15. doi: 10.1093/jac/dku525. PubMed PMID: 25585511.Klein EY, Van Boeckel TP, Martinez EM, Pant S, Gandra S, Levin SA, et al. Global increase and geographic convergence in antibiotic consumption between 2000 and 2015. Proc Natl Acad Sci U S A. 2018;115(15):E3463-E70. Epub 2018/03/28. doi: 10.1073/pnas.1717295115. PubMed PMID: 29581252; PubMed Central PMCID: PMCPMC5899442.van de Sande-Bruinsma N, Grundmann H, Verloo D, Tiemersma E, Monen J, Goossens H, et al. Antimicrobial drug use and resistance in Europe. Emerg Infect Dis. 2008;14(11):1722-30. Epub 2008/11/04. doi: 10.3201/eid1411.070467. PubMed PMID: 18976555; PubMed Central PMCID: PMCPMC2630720."
}
]
},
{
"id": "35298",
"date": "20 Jul 2018",
"name": "Michelle J Cole",
"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 article that tests the hypothesis that the emergence of gonococcal AMR in MSM is due to increased intensity of screening for gonorrhoea (GC). The authors’ findings are based primarily on an ecological analysis and, if these data are available in Belgium, we believe an analysis considering the GC testing history of each person whose MIC data are included would yield useful results. We would be keen to discuss the proposed hypothesis and interpretation of GRASP data with you, and have included comments for the authors’ consideration:\nThe evidence for different levels of screening intensity in MSM is available from EMIS; is there a comparable source of these data for heterosexual women? Also, did the recruitment strategy to participate in EMIS vary between the UK and Belgium? If so, this would meant that response to the question on the history of STI screening may be less comparable between these two countries. We believe a comparison of MSM with heterosexual men may be more useful, even with the limitations of underreporting of same-sex contact in heterosexually-identifying men. This is because the numbers of isolates are more comparable, women sometimes have different antimicrobial susceptibility profiles from heterosexual men and the cultures available from women are not representative of the circulating isolates in the community due to the difficulty in culturing from women.\n\nIn England, most gonorrhoea and chlamydia are diagnosed in people under the age of 25 years, with over one million chlamydia tests conducted annually though the National Chlamydia Screening Programme (NCSP) for 15 to 24 year olds. NCSP testing coverage is more than twice as high in women (28%) than men (11%) of that age-group (data here). Additionally, dual (CT/GC) NAAT platforms are commonly used for the NCSP and, while the positive predictive value of a gonorrhoea test in a community sample is very low, people with false positive results may be incorrectly prescribed antibiotics to treat gonorrhoea. These two aspects may work against the screening intensity hypothesis. There was no evidence of an association between azithromycin resistant NG and being diagnosed previously with chlamydia or gonorrhoea (as discussed in this paper, which the authors also cite: Clifton et al1); ths suggests that those who get tested for STIs more frequently do not have higher azithromycin MICs.\n\nThe authors only considered GRASP data from 2010, 2011 and 2014, but this analysis could be strengthened by including data from more years, including more recently published data: https://www.gov.uk/government/publications/gonococcal-resistance-to-antimicrobials-surveillance-programme-grasp-report. Alternatively, could the authors please specify these three years were selected, or why data from different countries aren’t compared within the same years? Also, could you please clarify why the MIC distributions for all three antibiotics were not analysed for the three time points? For example, the ceftriaxone data from the UK in 2011 could have been considered. In addition, cefixime resistance was widespread across Europe in 2010 due to the ST1407 clone. Cefixime resistance in Belgium was similar than the UK in 2010 according to the Euro-GASP data (2) and it would not be surprising if the burden of this resistance was in MSM in Belgium also. This analysis would be strengthened by including Belgium data from 2010/2011.\n\nInterestingly, Euro-GASP 2015 and 2016 data show much higher cefixime resistance in Belgium than the UK. According to the proposed hypothesis, should the opposite pattern have been observed? It would also be interesting to speculate the level of cefixime resistance in MSM in the UK if asymptomatic, multiple-site screening was not in place.\n\nIn the UK, gonorrhoea screening guidelines for MSM attending specialist sexual health services were updated in 2010, and this has led to an increase in gonorrhoea tests since then. This testing trend should be considered when interpreting these findings, as the level of testing prior to 2010 would be much lower than those of more recent years. Furthermore, the upturn in screening coincided with a rapid decline in cephalosporin resistance. Between 2010 (gonorrhoea screening guideline) and 2012 (change to dual therapy), the increased screening at sexual health services would have detected more cases and these were treated with cefixime which has been shown to be less effective, particularly at the pharynx (Barbee 2013). This would suggest that the increased prevalence of cefixime resistance in MSM may have been due to the usage of an antibiotic which is less effective than ceftriaxone, rather than the result of intensive screening.\n\nFigure 1 shows ciprofloxacin resistance was higher initially in heterosexual men, rather than MSM; therefore, resistance does not always emerge in MSM first. The labelling of the graphs in figure 2 is unclear – could this be clarified?\n\nIn the discussion and in reference to the link between MSM and travel – this may be the case, but we have found no data to support this: Town et al3. In addition one of the references cited mentions that heterosexual men reported more sex abroad than MSM - Matteelli et al4. As you have used publicly available GRASP data collected for routine surveillance purposes, the ethics statement should be amended accordingly. Similarly, references to the ‘STBRU’ should be to ‘PHE’.\n\nAs sexual orientation data was missing for 59% of the Belgium isolates, what are your views on how representative the sample of MSM whose data are considered in this analysis is? As many of the MIC medians were the same and some of MIC distributions that are declared as significantly different do not appear very different on inspection, particularly for 2014 UK cefixime and ceftriaxone, it would be useful if additional analysis were performed, such as linear regression with the geometric MIC means.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3855",
"date": "09 Aug 2018",
"name": "Chris Kenyon",
"role": "Author Response",
"response": "We would be keen to discuss the proposed hypothesis and interpretation of GRASP data with you, and have included comments for the authors’ consideration: Reply:Thank you for your interest and most useful comments. We would be most interested in collaborating on further studies along these lines. The evidence for different levels of screening intensity in MSM is available from EMIS; is there a comparable source of these data for heterosexual women? Also, did the recruitment strategy to participate in EMIS vary between the UK and Belgium? If so, this would meant that response to the question on the history of STI screening may be less comparable between these two countries. Reply:We looked and unfortunately could not find a comparable source for women. The recruitment strategies for EMIS were similar in the two countries. Because EMIS did not collect nationally representative samples we cannot exclude the possibility of a sampling bias that differed between the UK and Belgium. The large numbers recruited in EMIS in both countries and the fact that the results for the screening and other questions are commensurate with other data sources however argues against such a bias. We believe a comparison of MSM with heterosexual men may be more useful, even with the limitations of underreporting of same-sex contact in heterosexually-identifying men. This is because the numbers of isolates are more comparable, women sometimes have different antimicrobial susceptibility profiles from heterosexual men and the cultures available from women are not representative of the circulating isolates in the community due to the difficulty in culturing from women. Reply:Repeating the analyses with heterosexual men instead of women makes very little difference to the results. As we note above, visual comparisons of the MIC frequency distributions for azithromycin, cefixime and ceftriaxone between heterosexual men and women in the annual GRASP reports reveal little difference. We would be interested to repeat this analysis in using a larger number of countries/subpopulations as outlined in the conclusion. In England, most gonorrhoea and chlamydia are diagnosed in people under the age of 25 years, with over one million chlamydia tests conducted annually though the National Chlamydia Screening Programme (NCSP) for 15 to 24 year olds. NCSP testing coverage is more than twice as high in women (28%) than men (11%) of that age-group (data here). Additionally, dual (CT/GC) NAAT platforms are commonly used for the NCSP and, while the positive predictive value of a gonorrhoea test in a community sample is very low, people with false positive results may be incorrectly prescribed antibiotics to treat gonorrhoea. These two aspects may work against the screening intensity hypothesis. Reply:This is a useful observation. We have thought at some length about how screening intensity may produce antimicrobial resistance (AMR) in N. gonorrhoeae (Ng). This has lead to the pharmacoecological theory of AMR (connectivity AMR theory) which posits that it is the combination of dense sexual networks plus excess antimicrobial consumption (such as from intense screening) which is responsible. The dense sex network generates the high prevalence of Ng and the antimicrobial exposure then initially lowers prevalence but in the process generates a fitness advantage for resistant Ng [4]. If this theory is correct then intensive screening in the general heterosexual population in the UK (with its low connectivity network) would not have the same effect on selecting for resistance because the prevalence of Ng is low. Populations of sex workers would also be predicted to be at risk for the genesis of AMR, which has been observed. We have added a section in the discussion to make this clearer. These considerations are best understood by means of the following diagram (Figure 1) which is taken from our recent paper on the topic [4]: URL of Figure 1: https://wwwnc.cdc.gov/eid/article/24/7/17-2104-f2 Figure 1. High network connectivity combined with excess antimicrobial drug exposure from N. gonorrhoeae pre-exposure prophylaxis could produce antimicrobial resistance. A dense sexual network translates into a high equilibrium prevalence of N. gonorrhoeae (red squares) at time-point 1. Active N. gonorrhoeae screening of 50% of this population every 3 months results in 50% lower N. gonorrhoeae prevalence at time-point 2 (3 months later) but at the expense of an altered resistome (AScr; black squares represent 3 patients with N. gonorrhoeae cleared by screening/treatment. The unchanged underlying network connectivity results in a force that pushes N. gonorrhoeae back toward its equilibrium prevalence, placing recently cured patients at high risk for reinfection at a time when their resistomes are enriched with resistance genes. Early reinfecting N. gonorrhoeae take up these resistance genes by transformation. In the absence of screening and excess antimicrobial drug use (ANoScr) N. gonorrhoeae prevalence would not decline but there would be no pressure to select for antimicrobial resistance. Gray squares indicate uninfected persons; lines represent sexual relationships. There was no evidence of an association between azithromycin resistant NG and being diagnosed previously with chlamydia or gonorrhoea (as discussed in this paper, which the authors also cite: Clifton et al1); ths suggests that those who get tested for STIs more frequently do not have higher azithromycin MICs. Reply:We found the paper by Clifton et al. most interesting but are also aware of the paper by Wind et al., which found that receipt of azithromycin in the previous 30 days was associated with an increased MIC [5]. More important however is the ecological perspective. It is plausible that excess AMR is exerting its effect at the population level. Thus if population A has 30 fold higher macrolide consumption than population B we know from a range of studies that macrolide resistance in S. pneumonia and numerous other pathobionts is much more likely to emerge in population A than B. This has been clearly shown in ecological level studies [6-8]. The association may be harder to establish at an individual level but if one thinks from a pharmacoecologic perspective (and includes considerations of how Ng can acquire AMR from Neisseriaceae and other commensals as it transits through a population) it is easy to see why this is the case [4]. The deleterious effect of macrolides on the resistome at an individual level have been long established [7]. These considerations are of considerable importance given the threat of untreatable Ng. We have recently calculated the macrolide and cephalosporin exposure that 3 site, 3 monthly screening for gonorrhoea and chlamydia places on PrEP cohort (Unpublished results). We did this via conducting a literature review of the incidence of gonorrhoea and chlamydia in PrEP studies that conducted 3 site, 3 monthly screening. We found that screening results in macrolide consumption rates that considerably exceed those in high macrolide consumption populations where consumption has been strongly associated with macrolide resistance. The authors only considered GRASP data from 2010, 2011 and 2014, but this analysis could be strengthened by including data from more years, including more recently published data: https://www.gov.uk/government/publications/gonococcal-resistance-to-antimicrobials-surveillance-programme-grasp-report. Alternatively, could the authors please specify these three years were selected, or why data from different countries aren’t compared within the same years? Also, could you please clarify why the MIC distributions for all three antibiotics were not analysed for the three time points? For example, the ceftriaxone data from the UK in 2011 could have been considered. In addition, cefixime resistance was widespread across Europe in 2010 due to the ST1407 clone. Cefixime resistance in Belgium was similar than the UK in 2010 according to the Euro-GASP data (2) and it would not be surprising if the burden of this resistance was in MSM in Belgium also. This analysis would be strengthened by including Belgian data from 2010/2011. Reply:We agree it is frustrating not to have data from Belgium in 2010/11. As we point out in the methods section we could not use this data as we do not have data as to sexual orientation in sufficient number from this period. In the methods we state:The details regarding sexual orientation started to be reported in sufficient numbers from 2013 onwards For the UK data we agree it would be interesting to look at the data from all the available years. We chose the first year at or after 2010 when antimicrobial MIC frequency distributions were reported by gender/sexual orientation. This was 2010 for ceftriaxone and 2011 for azithromycin and cefixime. We then looked at all 3 antimicrobials in 2014 so as to use the same time period for the comparison with Belgium. We considered that this analysis was sufficient for our purposes but acknowledge the reviewers point that there would be a multiplicity of other ways of doing this analysis. Looking at the MIC distributions from other years we consider it likely that this would not substantially change the results. Our analysis involved comparing MIC distributions between MSM and women in the two countries and not comparing MSM or women between the two countries. Interestingly, Euro-GASP 2015 and 2016 data show much higher cefixime resistance in Belgium than the UK. According to the proposed hypothesis, should the opposite pattern have been observed? It would also be interesting to speculate the level of cefixime resistance in MSM in the UK if asymptomatic, multiple-site screening was not in place. Reply:This is an excellent point. The cefixime resistance figures in Belgium 2015/2016 are remarkably high according to the Euro GRASP figures but it must be remembered that these figures are based on a small sample of all national samples for this time period. The full results for 2016 are 597 isolates tested of which 36 (6%°/1(0,1%) had decreased sensitivity to cefixime according to EUCAST/CLSI breakpoints. The results in 2015 were similar: 630 isolates tested with 52 (8,3%)/1(0,1%) classified as decreased sensitivity. These lower rates of resistance are a more accurate representation than the Euro GRASP figures [3]. < >In the UK, gonorrhoea screening guidelines for MSM attending specialist sexual health services were updated in 2010, and this has led to an increase in gonorrhoea tests since then. This testing trend should be considered when interpreting these findings, as the level of testing prior to 2010 would be much lower than those of more recent years. Furthermore, the upturn in screening coincided with a rapid decline in cephalosporin resistance. Between 2010 (gonorrhoea screening guideline) and 2012 (change to dual therapy), the increased screening at sexual health services would have detected more cases and these were treated with cefixime which has been shown to be less effective, particularly at the pharynx (Barbee 2013). This would suggest that the increased prevalence of cefixime resistance in MSM may have been due to the usage of an antibiotic which is less effective than ceftriaxone, rather than the result of intensive screening. Figure 1 shows ciprofloxacin resistance was higher initially in heterosexual men, rather than MSM; therefore, resistance does not always emerge in MSM first. The labelling of the graphs in figure 2 is unclear – could this be clarified? In the discussion and in reference to the link between MSM and travel – this may be the case, but we have found no data to support this: Town et al3. In addition one of the references cited mentions that heterosexual men reported more sex abroad than MSM - Matteelli et al4. As you have used publicly available GRASP data collected for routine surveillance purposes, the ethics statement should be amended accordingly. Similarly, references to the ‘STBRU’ should be to ‘PHE’. As sexual orientation data was missing for 59% of the Belgium isolates, what are your views on how representative the sample of MSM whose data are considered in this analysis is? As many of the MIC medians were the same and some of MIC distributions that are declared as significantly different do not appear very different on inspection, particularly for 2014 UK cefixime and ceftriaxone, it would be useful if additional analysis were performed, such as linear regression with the geometric MIC means.https://doi.org/10.3201/eid2407.172104.Reply:Thanks for this information which is indeed relevant and has been included as a major caveat in the discussion. Between 2010 (gonorrhoea screening guideline) and 2012 (change to dual therapy), the increased screening at sexual health services would have detected more cases and these were treated with cefixime which has been shown to be less effective, particularly at the pharynx (Barbee 2013). This would suggest that the increased prevalence of cefixime resistance in MSM may have been due to the usage of an antibiotic which is less effective than ceftriaxone, rather than the result of intensive screening. Reply:We agree this is a possibility and have expanded the discussion section to reflect his point. Figure 1 shows ciprofloxacin resistance was higher initially in heterosexual men, rather than MSM; therefore, resistance does not always emerge in MSM first. Reply:This is true as is the repeated emergence of AMR in sex workers and their contacts. The labelling of the graphs in figure 2 is unclear – could this be clarified? Reply:We have asked the publisher to increase the size of the figure to a single page so as to make it easier to read the labeling. In the discussion and in reference to the link between MSM and travel – this may be the case, but we have found no data to support this: Town et al3. In addition one of the references cited mentions that heterosexual men reported more sex abroad than MSM - Matteelli et al4. Reply:Thanks. We have added a sentence pointing this out in the discussion. As you have used publicly available GRASP data collected for routine surveillance purposes, the ethics statement should be amended accordingly. Similarly, references to the ‘STBRU’ should be to ‘PHE’. Reply:Both these changes have been made. As sexual orientation data was missing for 59% of the Belgium isolates, what are your views on how representative the sample of MSM whose data are considered in this analysis is? Reply:This is a major limitation and as we note in the limitations section means we need to be cautious in any conclusions we draw from this study. As many of the MIC medians were the same and some of MIC distributions that are declared as significantly different do not appear very different on inspection, particularly for 2014 UK cefixime and ceftriaxone, it would be useful if additional analysis were performed, such as linear regression with the geometric MIC means. Reply:Numerous other analyses could be done but this is a very simple and limited analysis. We believe that given the major limitations of the study described above and more fully in the paper, it would be inappropriate to conduct further complicated statistical analyses. The data suggest some differences in the relationship between MIC distributions between the 2 countries. The next step should be to find better and more extensive datasets to evaluate further test the hypothesis. We conclude the paper with a description of what we think these analyses could be.References:1. Chisholm SA, Mouton JW, Lewis DA, Nichols T, Ison CA, Livermore DM. Cephalosporin MIC creep among gonococci: time for a pharmacodynamic rethink? J Antimicrob Chemother. 2010;65(10):2141-8. Epub 2010/08/10. doi: 10.1093/jac/dkq289. PubMed PMID: 20693173.2. Negri MC, Morosini MI, Loza E, Baquero F. In vitro selective antibiotic concentrations of beta-lactams for penicillin-resistant Streptococcus pneumoniae populations. Antimicrob Agents Chemother. 1994;38(1):122-5. Epub 1994/01/01. PubMed PMID: 8141563; PubMed Central PMCID: PMCPMC284406.3. Cole MJ, Spiteri G, Jacobsson S, Woodford N, Tripodo F, Amato-Gauci AJ, et al. Overall Low Extended-Spectrum Cephalosporin Resistance but high Azithromycin Resistance in Neisseria gonorrhoeae in 24 European Countries, 2015. BMC Infect Dis. 2017;17(1):617. Epub 2017/09/13. doi: 10.1186/s12879-017-2707-z. PubMed PMID: 28893203; PubMed Central PMCID: PMCPMC5594611.4. Kenyon C, Schwartz IS. A combination of high sexual network connectivity and excess antimicrobial usage induce the emergence of antimicrobial resistance in Neisseria gonorrhoeae Emerg Infect Dis. 2018;24(7). doi: https://doi.org/10.3201/eid2407.172104.5. Wind CM, de Vries E, Schim van der Loeff MF, van Rooijen MS, van Dam AP, Demczuk WH, et al. Decreased azithromycin susceptibility of Neisseria gonorrhoeae isolates in patients recently treated with azithromycin. Clin Infect Dis. 2017. Epub 2017/04/04. doi: 10.1093/clinid/cix249. PubMed PMID: 28369420.6. Goossens H, Ferech M, Vander Stichele R, Elseviers M, Group EP. Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet. 2005;365(9459):579-87. Epub 2005/02/15. doi: 10.1016/S0140-6736(05)17907-0. PubMed PMID: 15708101.7. Malhotra-Kumar S, Lammens C, Coenen S, Van Herck K, Goossens H. Effect of azithromycin and clarithromycin therapy on pharyngeal carriage of macrolide-resistant streptococci in healthy volunteers: a randomised, double-blind, placebo-controlled study. The Lancet. 2007;369(9560):482-90.8. Garcia-Rey C, Aguilar L, Baquero F, Casal J, Dal-Re R. Importance of local variations in antibiotic consumption and geographical differences of erythromycin and penicillin resistance in Streptococcus pneumoniae. J Clin Microbiol. 2002;40(1):159-64. Epub 2002/01/05. PubMed PMID: 11773111; PubMed Central PMCID: PMCPMC120130."
}
]
}
] | 1
|
https://f1000research.com/articles/7-569
|
https://f1000research.com/articles/7-1210/v1
|
07 Aug 18
|
{
"type": "Research Article",
"title": "Exploring machine learning: A bibliometric general approach using SciMAT",
"authors": [
"Juan Rincon-Patino",
"Gustavo Ramirez-Gonzalez",
"Juan Carlos Corrales",
"Juan Rincon-Patino",
"Juan Carlos Corrales"
],
"abstract": "Background: Machine learning is becoming increasingly important for companies and the scientific community. In this study, we perform a bibliometric analysis on machine learning research, in order to provide an overview of the scientific work during the period 2007-2017 in this area and to show trends that could be the basis for future developments in the field. Methods: This study is carried out using the SciMAT tool based on results extracted from Scopus. This analysis shows the strategic diagrams of evolution and a set of thematic networks. The results provide information on broad tendencies of machine learning. Results: The results show that SciMAT is a useful tool to carry out a science mapping analysis, and emphasizes the premise that machine learning has boundless applications and will continue to be an interesting research field in the future. Conclusions: Some of the conclusions exposed in this study show that classification algorithms have been widely studied and represent a relevant tool for generating different machine learning applications. Nonetheless, regression algorithms are becoming increasingly important in the scientific community, allowing the generation of solutions to predict diseases, sales, and yields, for example.",
"keywords": [
"machine learning",
"science mapping",
"bibliometrics",
"topic analysis",
"SciMAT"
],
"content": "Introduction\n\nThe machine learning field researches different human learning processes, the theoretical analysis of possible learning algorithms and methods for several application domains1. Studies based on machine learning have allowed scientists and companies to predict mass mortality events2, the quality of water3, segment clients in private banking4, automatically classify text5 or for the production of crops, such as cocoa6. Considering the growing interest of the scientific community in machine learning research and its challenges, it is interesting and necessary to analyze the field. A good approach for that purpose is science mapping analysis because it is a different way of visualizing information that allows a new researcher to become familiar with a field. An example of science mapping analysis providing an overview of the conceptual evolution of a field in medicine is proposed in 7. In this study, we perform a science mapping analysis to explore machine learning research. The objective of the study is to allow new data analysts to know the current knowledge base about machine learning and to have an initial point to explore applications in this field.\n\nThis article has the following structure: In the Methods section, we describe the research methodology, the dataset used, the tool configuration, and how the analysis was performed. The Results section presents the results of the science mapping analysis. The conclusions are at the end of the article.\n\n\nMethods\n\nWe used Scopus as the bibliographic source. We looked, in the third quarter of 2017, for references of articles and conferences about machine learning, using this concept as the search keyword (‘machine AND learning’), with results ranging from 2007 to 2017(Q2). The concept was searched for in the article title, abstract and keywords of articles found. These articles were sorted by date (newest first). All the articles, between 2007 and 2017, were taken in to account for performing the analysis with the aim of obtaining a general vision of the field.\n\nWe obtained 67,475 records that were saved using RIS format in different files sorted by year. Figure 1 shows a summary of the records, and shows that research in the field of machine learning has been growing steadily. It is important to observe that the results for 2017 are not in the figure since the records are only to the second half Q2 of year and database is permanently updated. However, these results were used for the analysis because they show trends, which is the primary objective of the present study.\n\nWe used the records from Scopus for executing the analysis with SciMAT version 1.1.04. In this tool, the unit of analysis was Words. Primarily, we did a deduplication process, grouping similar words (by plurals) and looking for synonyms or duplicates in words with the highest number of documents and repetitions, always trying to avoid bias to include the largest number of terms. After that, we divided the time interval (2007 – 2017) into six smaller periods: 2007–2009, 2010–2012, 2013–2014, 2015, 2016, 2017(Q2 – published papers up the second quarter of the year). We distributed the gaps this way in order to have a comparable number of articles in each one of them. Finally, we carried out the analysis with the following configuration: all the periods, author’s words as the unit of analysis, a minimum frequency for data reduction of 100 (excepting 2017 (Q2) with 50) and co-occurrence as a kind of network. Other configurations were: network reduction equal to one, association strength as a normalization measure, Simple centers algorithm with a maximum net size of seven and a minimum of five, core mapper for the document mapper, h-index and Sum citations as the quality measures and, lastly, association strength for the evolution and overlapping map.\n\nSciMAT shows a spatial representation of the way disciplines, fields, specialties and documents or authors related to one another8–10. For this purpose, the tool implements a longitudinal framework, which takes as its base a co-word analysis and the h-index. Co-word analysis firstly provides information on the themes of a research field and, secondly, enables to analyze and to track the evolution of a study field throughout consecutive periods of time11. The h-index is used to measure the impact of the various identified thematic areas8.\n\nIn this tool, we follow four steps as mentioned by 10: researching cluster detection, drawing strategic diagrams, plotting of thematic areas and carrying out a performance analysis. For the first one, the tool creates a network of keywords co-occurrence based on 12 and 13 and makes a clustering of keywords to topics, using the Simple center's algorithm. For the second step, according to 13, the cluster centrality and density rank values are relevant. The centrality measures the intensity of the interaction of a group with the others; if the cluster is boldly related to the field of research, then the link will be stronger. The density measures the intensity of internal links inside the group. According to 8 and 10, the themes of the clusters can be classified into four groups using these two measures: (1) Motor themes, which are both well developed and central to the research field; (2) basic and transversal themes, which are not sufficiently developed topics but significant for the area of investigation; (3) emerging or declining themes, which are weakly studied and marginal; (4) highly developed and isolated themes, which have well-developed internal links, but reduced external links and have only minimal importance to the field. The third step is the plotting of thematic areas, and the last one is to conduct the performance analysis. In this study, we analyzed quantitative measures, such as the number of documents and authors.\n\n\nResults\n\nTo analyze the most relevant topics of investigation in different years, SciMAT uses strategic diagrams. We generated a diagram for each period of the study. The charts are divided into four quadrants: the upper-right quadrant alludes to the motor themes, the upper-left quadrant to the highly developed and isolated themes, the lower-right one to the basic and transversal themes, and the lower-left one to the emerging or declining topics9.\n\nFigure 2 shows the strategic diagram of the period 2007–2009, which has 27 themes. During this time span, OPTIMIZATION, CONTROL-THEORY, and IMAGE-CLASSIFICATION are some of the emerging topics. SEQUENCE-ANALYSIS-PROTEIN appears as the motor theme with the highest density (0.83) and centrality (3.21) values, followed by DATABASES-PROTEIN, with centrality equal to 2.97 and density equal to 0.43. This suggests that machine learning was widely used to study the protein molecule in that period, covering, for example, studies to predict the structure or function of that molecule.\n\nDue to the great number of themes that appear in the strategic diagram, we decided to choose three themes, giving priority to those that may have a high impact application. Figure 3A shows the net for the theme DATABASES-PROTEIN, which has a density of 0.43, a centrality of 2.97 and a document count of 322. In this net, we can observe that an important topic is PROTEIN-STRUCTURE. Figure 3B presents the network for the theme DATASETS, which has a density of 0.06, a centrality of 1.37, a document count of 273 and strongly related topics such as CLASSIFIERS, SEMI-SUPERVISED-LEARNING and RANDOM-FOREST. Figure 3C shows the network for the theme IMAGE-CLASSIFICATION, which has a density of 0.11, a centrality of 1.3, a document count of 273 and relevant concepts, such as NEURAL-NETWORKS, SUPPORT-VECTOR-MACHINE (SVM) and IMAGE-PROCESSING.\n\n(A) DATABASES-PROTEIN; (B) DATASETS; (C) IMAGE-CLASSIFICATION.\n\nFigure 4 shows the strategic diagram of the period 2010–2012. The diagram has 35 themes. During this time span, VIRTUAL-REALITY, ROBOTS, and METADATA are some of the emerging themes, while CLASSIFICATION-ALGORITHM is one of the basic and transversal topics. AGED and PROTEIN-ANALYSIS appear as the motor subjects with the highest density (0.72 and 0.58, respectively) and centrality (1.91 and 2.18, respectively) values. Other important themes are CHEMISTRY AND GENETICS. This is a sign that during this period, topics on biology and health began to become relevant in applied machine learning research.\n\nFrom the second period, we selected three thematic networks. Figure 5A shows the net for the theme PROTEIN-ANALYSIS, which has a density of 0.58, a centrality of 2.18 and a document count of 229. This shows us that topics about proteins continue to be important in this period. Figure 5B presents the network for the theme CHEMISTRY, which is an emergent theme and has a density of 0.32, a centrality of 2.24 and a document count of 346. Figure 5C shows the network for the subject VIRTUAL-REALITY, which has a density of 0.07, a centrality of 0.98, a document count of 65 and is another emerging theme for the period 2010–2012.\n\n(A) PROTEIN-ANALYSIS; (B) CHEMISTRY; (C) VIRTUAL-REALITY.\n\nThe strategic diagram of the period 2013–2014 is shown on Figure 6, which has 32 themes. During this time frame, SENSORS, FACE-RECOGNITION and COMMERCE are some of the emerging topics. The IMAGE-INTERPRETATION-COMPUTER-ASSISTED topic appears as the motor theme, with the highest density (0.58) and centrality (2.61) values. This suggests that in this period there were studies on machine learning applied to different topics, such as sensor data, costs, and gesture recognition.\n\nWe selected three thematic networks from the third period (2013–2014). Figure 7A shows the network for the theme AMINO-ACID-SEQUENCE, which has a density of 0.42, a centrality of 1.91 and a document count of 276. Figure 7B presents the network for the theme MOBILE-DEVICES, which has a density of 0.21, a centrality of 0.76, a document count of 159 and strongly related topics such as HUMAN-COMPUTER-INTERACTION, E-LEARNING, and UBIQUITOUS-COMPUTING. Figure 7C shows the network for the theme COMMERCE, which has a density of 0.07, a centrality of 0.88, a document count of 205 and relevant concepts, such as SOCIAL-NETWORKING and COSTS.\n\n(A) AMINO-ACID-SEQUENCE; (B) MOBILE-DEVICES; (C) COMMERCE.\n\nFigure 8 shows the strategic diagram of the period 2015. The diagram has 25 themes. During this time span, SMARTPHONES, FACE-RECOGNITION and FORESTRY (label generated for algorithms such as Random-Forest or Decision-Trees) are some of the emerging themes. NUCLEAR-MAGNETIC-RESONANCE-IMAGING appears as the motor subject with the highest density (0.57) and centrality (2.81) values. Other important themes are COMPUTATIONAL-BIOLOGY and MEDICAL-IMAGING. We found that during this period biology and health were once again relevant topics in machine learning research.\n\nFrom the fourth time frame, we selected three thematic networks. Figure 9A shows the network for the theme NUCLEAR-MAGNETIC-RESONANCE-IMAGING, which has a density of 0.57, a centrality of 2.81 and a document count of 251. Figure 9B presents the network for the theme COMPUTATIONAL-BIOLOGY, which is an emergent theme and has a density of 0.44, a centrality of 1.8 and a document count of 320. Figure 9C shows the network for the topic SMARTPHONES, which has a density of 0.07, a centrality of 0.13, a document count of 144 and is another emerging theme for the 2015 period.\n\n(A) NUCLEAR-MAGNETIC-RESONANCE-IMAGING; (B) COMPUTATIONAL-BIOLOGY; (C) SMARTPHONES.\n\nThe strategic diagram of the period 2016 is shown in Figure 10, which has 24 themes. During this time span, UBIQUITOUS-COMPUTING and COMMERCE are some of the emerging themes. The MIDDLE-AGED theme –which refers to applications developed for middle-aged people– appears as the motor topic with the highest density (0.76) and centrality (2.04) values.\n\nWe selected three thematic networks from the fifth period (2016). Figure 11A shows the network for the theme MEDICAL-IMAGING, which has a density of 0.26, a centrality of 1.45 and a document count of 312. Figure 11B presents the network for the theme INTRUSION-DETECTION, which has a density of 0.3, a centrality of 0.89 and a document count of 289. Figure 11C shows the network for the theme UBIQUITOUS-COMPUTING, which has a density of 0.07, a centrality of 1.07, a document count of 132 and relevant concepts, such as AUTOMATION, SMARTPHONES and the INTERNET.\n\n(A) MEDICAL-IMAGING; (B) INTRUSION-DETECTION; (C) UBIQUITOUS-COMPUTING.\n\nFigure 12 shows the strategic diagram of the period 2017(Q2). The diagram has 16 themes. During this time span, HUMAN is the only emerging theme, while FORESTRY is one of the basic and transversal topics. This shows us the importance of algorithms such as Random-Forest or Decision-Trees during the last decade in the research on machine learning. MEDICAL-IMAGING appears as the motor theme with the highest density (0.51) and centrality (2.05) values. Once again, topics on health are relevant in machine learning research.\n\nFrom 2017(Q2), we selected three thematic networks. Figure 13A shows the net for the theme MEDICAL-IMAGING, which has a document count of 137. Figure 13B presents the network for the topic FORESTRY (label generated for algorithms such as Random-Forest or Decision-Trees), which has a density of 0.17, a centrality of 1.84 and a document count of 220.\n\n(A) MEDICAL-IMAGING; (B) FORESTRY.\n\n\nConclusions\n\nExposing emerging trends in the field of machine learning allows researchers to increase their understanding of the changes and the evolution over time of this research field. One of the primary objectives of a science mapping analysis is to highlight trends and possible relationships between the relevant topics of a research field. SciMAT is a useful tool to carry out a study based on this approach, which offers fundamental themes, based on a cluster generation. The results of the present study show that machine learning is an important and widely studied scientific area. The tendencies indicate that machine learning applications will still be of interest to the scientific community. The use of machine learning to predict diseases such as cancer14 or Alzheimer’s disease15, and in fields such as biology16, rehabilitation system17, commerce18, smartphones19 and ubiquitous computing20, will be a trend in the near future.\n\n\nData availability\n\nDataset 1: Data obtained from Scopus and SciMat project file, to be opened in SciMat. DOI, 10.5256/f1000research.15620.d21242521",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors are grateful to the Telematics Engineering Group (GIT) of the University of Cauca for scientific support and Innovacción Cauca project for master's scholarship granted to J. Rincon-Patino.\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\nMichalski RS, Carbonell JG, Mitchell TM: Machine Learning: An Artificial Intelligence Approach. Springer Berlin Heidelberg, 2013. Publisher Full Text\n\nCrisci C, Ghattas B, Perera G: A review of supervised machine learning algorithms and their applications to ecological data. Ecol Modell. 2012; 240: 113–122. Publisher Full Text\n\nLópez ID, Figueroa A, Corrales JC: Adaptive Prediction of Water Quality Using Computational Intelligence Techniques. In Computational Science and Its Applications -- ICCSA 2017: 17th International Conference, Trieste, Italy, July 3-6, 2017, Proceedings, Part II, Cham: Springer International Publishing, 2017; 45–59. Publisher Full Text\n\nSmeureanu I, Ruxanda G, Badea LM: Customer segmentation in private banking sector using machine learning techniques. J Bus Econ Manag. 2013; 14(5): 923–939. Publisher Full Text\n\nSebastiani F: Machine Learning in Automated Text Categorization. ACM Comput Surv. 2002; 34(1): 1–47. Publisher Full Text\n\nPlazas JE, López ID, Corrales JC: A Tool for Classification of Cacao Production in Colombia Based on Multiple Classifier Systems. In Computational Science and Its Applications -- ICCSA 2017: 17th International Conference, Trieste, Italy, July 3-6, 2017, Proceedings, Part II, Cham: Springer International Publishing, 2017; 60–69. Publisher Full Text\n\nMoral-Muñoz JA, Cobo MJ, Peis E, et al.: Analyzing the research in Integrative & Complementary Medicine by means of science mapping. Complement Ther Med. 2014; 22(2): 409–418. PubMed Abstract | Publisher Full Text\n\nMartínez MA, Cobo MJ, Herrera M, et al.: Analyzing the Scientific Evolution of Social Work Using Science Mapping. Res Soc Work Pract. 2015; 25(2): 257–277. Publisher Full Text\n\nCobo MJ, López-Herrera AG, Herrera-Viedma E, et al.: SciMAT: A new science mapping analysis software tool. J Am Soc Inf Sci Technol. 2012; 63(8): 1609–1630. Publisher Full Text\n\nCobo MJ, López-Herrera AG, Herrera-Viedma E, et al.: An approach for detecting, quantifying, and visualizing the evolution of a research field: A practical application to the Fuzzy Sets Theory field. J Informetr. 2011; 5(1): 146–166. Publisher Full Text\n\nGarfield E: Scientography: Mapping the tracks of science. Curr Contents Soc Behav Sci. 1994; 7(45): 5–10. Reference Source\n\nCallon M, Courtial JP, Turner WA, et al.: From translations to problematic networks: An introduction to co-word analysis. Soc Sc Inform. 1983; 22(2): 191–235. Publisher Full Text\n\nCallon M, Courtial J, Laville F: Co-word analysis as a tool for describing the network of interactions between basic and technological research: The case of polymer chemsitry. Scientometrics. 1991; 22(1): 155–205. Publisher Full Text\n\nWang Y, Tetko IV, Hall MA, et al.: Gene selection from microarray data for cancer classification—a machine learning approach. Comput Biol Chem. 2005; 29(1): 37–46. PubMed Abstract | Publisher Full Text\n\nMoradi E, Pepe A, Gaser C, et al.: Machine learning framework for early MRI-based Alzheimer’s conversion prediction in MCI subjects. Neuroimage. 2015; 104: 398–412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSwan AL, Mobasheri A, Allaway D, et al.: Application of Machine Learning to Proteomics Data: Classification and Biomarker Identification in Postgenomics Biology. OMICS. 2013; 17(12): 595–610. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYeh SC, Huang MC, Wang PC, et al.: Machine learning-based assessment tool for imbalance and vestibular dysfunction with virtual reality rehabilitation system. Comput Methods Programs Biomed. 2014; 116(3): 311–318. PubMed Abstract | Publisher Full Text\n\nYan J, Zhang C, Zha H, et al.: On Machine Learning Towards Predictive Sales Pipeline Analytics. In Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence. 2015; 1945–1951. Reference Source\n\nFaragher RM, Harle RK: SmartSLAM - An Efficient Smartphone Indoor Positioning System Exploiting Machine Learning and Opportunistic Sensing. In Proceedings of the 26th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2013). 2013; 1006–1019. Reference Source\n\nVentura D, Casado-Mansilla D, López-de-Armentia J, et al.: ARIIMA: A Real IoT Implementation of a Machine-Learning Architecture for Reducing Energy Consumption. In Ubiquitous Computing and Ambient Intelligence. Personalisation and User Adapted Services: 8th International Conference, UCAmI 2014, Belfast, UK, Cham: Springer International Publishing, 2014; 444–451. Publisher Full Text\n\nRincon-Patino J, Ramirez-Gonzalez G, Corrales JC: Dataset 1 in: Exploring machine learning: A bibliometric general approach using SciMAT. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15620.d212425"
}
|
[
{
"id": "36927",
"date": "09 Aug 2018",
"name": "Rajesh Kumar Tiwari",
"expertise": [
"Reviewer Expertise Machine learning classification",
"prediction models using Artificial Intelligence",
"Statistical analysis",
"Computational Biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article highlights about the bibliometric analysis of scientific works in the area of machine learning, published in Scopus indexed Journals during the period of 2007-2017. The manuscript is well written. However, I have following suggestions to make:\n\nCorrection in title of Manuscript: Exploring machine learning: A bibliometric general approach using SciMAT Tool\n\nDefinition of Machine learning should be incorporated in the introduction Part. Authors can refer and cite following articles: PMID: 237403901 and PMID: 283020412\n\nApart from mass mortality events, the quality of water, segment clients in private banking, automatically classify text, production of crops, the machine learning applications are also widely used in the Drug designing in pharmaceutical industries, academia and research. Authors should mention the same in the introduction part of the article and can refer and cite the following articles with PMID: 223462303, PMID: 265268294, PMID: 283828575, PMID: 292563446\n\nMethods: As the data set is collected from Scopus, the same should be cited in the manuscript.\n\nThe original source of SciMAT Tool should be cited in the manuscript.\n\nThe authors must elaborate more about the basis of division of the data set. It is mentioned that the time interval divided into six smaller periods: 2007–2009, 2010–2012, 2013–2014, 2015, 2016, 2017. As the division is not uniform. Was the gap made such a way so as to have comparable number of articles in each one of them only?\n\nThe machine learning algorithms are widely used in robotics and pharmaceutical properties prediction. They can also be included in conclusion part.\n\nApart from SciMAT, are some other similar tools available? Can the comparative analysis of results using the same data set from different tools be done to know that how conclusive the reported results are? This can be included as future prospect of the current research.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "37068",
"date": "15 Aug 2018",
"name": "Mikhail G. Dozmorov",
"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 “Exploring machine learning: A bibliometric general approach using SciMAT” by Rincon-Patino et al. presents an overview of areas of machine learning applications. The study is conducted using science mapping analysis with the SciMAT tool. A bibliography containing “machine” and “learning” keywords over the 10-year period (2007-2017) was extracted from Scopus. This period was split into six smaller periods, and the state of machine learning in each period was illustrated by two figures, strategic diagram, and selected thematic network. The paper is well written. The following recommendations are intended to improve the readability and message of the paper.\nStrategic diagrams may not be familiar for users who aren’t familiar with the SciMAT tool. Thus, the description of strategic diagrams given at the beginning of the Results section is better to be placed to the legend of the first figure. Also, clarify the meaning of “the basic and transversal themes.” Furthermore, the description of the lower-left quadrant, “the emerging or declining topics,” is confusing, it should be one or the other.\n\nAlthough the study claims to present a longitudinal analysis of the development of machine learning areas, it is the analysis and the description of individual time periods, presented on individual figures. I am missing the compare-and-contrast view of the analysis, which would have been possible if the figures would be presented side-by-side. Given the good readability of the strategic diagrams, it is perfectly possible to combine them into one figure. The same can be done for the selected thematic networks, by making edges shorter, thicker, and fonts - larger.\n\nIt is unclear whether the manuscript is about the overview of machine learning fields, or the tool SciMAT and the science mapping analysis. Besides simply describing what the results are at each time period, it would be good to have some conclusions about the longitudinal trends of the field.\n\nAlthough the rationale for choosing the non-equal smaller periods in understandable, the periods themselves should be better defined. At the very least, state that the presented periods are inclusive, that is, 2007-2009 spans the 3-year period. Better, adopt (MM/DD/YYYY-MM/DD/YYYY] notation.\n\nFigure 1 shows the growing number of documents containing “machine” and “learning” keywords. However, it is known that the number of publications per year grows by itself. It would be more informative to show the proportion of published documents that contain machine\" and “learning” keywords. The usefulness of representing the proportion of publications can be illustrated with the following R code:\nlibrary(MDmisc) # devtools::install_github('mdozmorov/MDmisc') library(ggplot2) p <- get_pubmed_graph(\"machine learning\", yearstart = 2007, yearend = 2017, normalize = TRUE, xlab = \"Year\", ylab = \"Proportion of all publications\") p ggsave(\"pubmed_machine_learning.png\", p, device = \"png\", height = 4)\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1210
|
https://f1000research.com/articles/5-2748/v1
|
23 Nov 16
|
{
"type": "Software Tool Article",
"title": "BgeeDB, an R package for retrieval of curated expression datasets and for gene list expression localization enrichment tests",
"authors": [
"Andrea Komljenovic",
"Julien Roux",
"Marc Robinson-Rechavi",
"Frederic B. Bastian",
"Andrea Komljenovic",
"Julien Roux",
"Marc Robinson-Rechavi"
],
"abstract": "BgeeDB is a collection of functions to import into R re-annotated, quality-controlled and reprocessed expression data available in the Bgee database. This includes data from thousands of wild-type healthy samples of multiple animal species, generated with different gene expression technologies (RNA-seq, Affymetrix microarrays, expressed sequence tags, and in situ hybridizations). BgeeDB facilitates downstream analyses, such as gene expression analyses with other Bioconductor packages. Moreover, BgeeDB includes a new gene set enrichment test for preferred localization of expression of genes in anatomical structures (“TopAnat”). Along with the classical Gene Ontology enrichment test, this test provides a complementary way to interpret gene lists. Availability: http://www.bioconductor.org/packages/BgeeDB/",
"keywords": [
"Bioconductor",
"R Package",
"Collective Data Access",
"Gene expression",
"Gene Enrichment Analysis"
],
"content": "Introduction\n\nGene expression levels influence the behavior of cells, the functionality of tissues, and a wide range of processes from development and aging to physiology or behavior. It is of particular importance that researchers are able to take advantage of the vast amounts of publicly available gene expression datasets to reproduce and validate results, or to investigate new research questions1–3.\n\nTo that purpose, one should be able to easily query and import gene expression datasets generated using different technologies, and their associated metadata. The R environment4 has now become a standard for bioinformatics and statistical analysis of gene expression data, through the Bioconductor framework and its many open source packages5,6. It is thus desirable to provide access to gene expression datasets programmatically and directly in R. For example, the Bioconductor packages ArrayExpress7, GEOquery8 and SRAdb9 provide access to the reference databases ArrayExpress10, GEO11 and SRA12 respectively.\n\nHowever, such databases are primary archives aiming at comprehensiveness. They include gene expression datasets and other functional genomics data, generated from diverse experimental conditions, of diverse quality. The data provided are heterogeneous, with some datasets including only unprocessed raw data, and others including only data processed using specific analysis pipelines. For instance, over the 44,177 RNA array assay experiments stored in ArrayExpress with processed data available as of October 2016, 7,520 do not include the raw data. Metadata are often provided as free-text information that is difficult to query. For instance, the GEO database encourages submitters of high-throughput sequencing experiments to provide MINSEQE elements, but does not enforce this practice (see, e.g., GEO submission guidelines, and GEO Excel template for submissions). Unless the user needs to retrieve a specific known dataset from its accession number, it can be difficult to identify relevant available datasets. This can ultimately constitute an obstacle to data reuse.\n\nOne response to this diversity of primary archives is topical databases1. They can be useful for researchers of specialized fields, and even more so if they propose an R package for data access. For example, the BrainStars Bioconductor package allows access to microarray data of mouse brain regions samples from the BrainStars project13,14. The ImmuneSpaceR Bioconductor package allows access to the gene expression data generated by the Human Immunology Project Consortium15. Such efforts allow better control of the data and annotation quality, but by nature they include a limited number of conditions, which only fit the needs of specialized projects. Similarly, numerous “ExperimentData” packages are available on the Bioconductor repository, which each include a single curated and well-formatted expression dataset (see http://www.bioconductor.org/packages/release/BiocViews.html#___ExpressionData). But these packages are rarely updated and are mostly meant to be used as examples in software packages vignettes, for teaching, or as supplementary data for publications.\n\nFinally, added-value databases aim at filtering, annotating, and possibly reprocessing all or some of the datasets available from the primary archives1. For example, a Bioconductor package was recently released to access the Expression Atlas, which includes a selection of microarray and RNA-seq datasets from ArrayExpress that are re-annotated and reprocessed16,17. Similarly, the recount Bioconductor package provides access to a dataset of 2,040 reanalyzed human RNA-seq samples from SRA (see https://jhubiostatistics.shinyapps.io/recount/)18–20.\n\nThe Bgee database (http://bgee.org/)21 is another added-value database, which currently offers access to reprocessed gene expression datasets from 17 animal species. Bgee aims to compare gene expression patterns across tissues, developmental stages, ages and species. It provides manually curated annotations to ontology terms, describing precisely the experimental conditions used. It integrates expression data generated with multiple technologies: RNA-Seq, Affymetrix microarrays, in situ hybridization, and expressed sequence tags (ESTs). An important characteristic of Bgee is that all datasets are manually curated to retain only “normal” healthy wild-type samples, i.e., excluding gene knock-out, treatments or diseases. Finally, Bgee datasets are carefully checked for quality issues, and reprocessed to produce normalized expression level, calls of presence/absence of expression, and of differential expression. Bgee thus provides a reference of high-quality and reusable gene expression datasets that are relevant for biological insights into normal conditions of gene expression. Release 13 of Bgee includes 526 RNA-seq libraries, 12,736 Affymetrix chips, 349,613 results from 46,619 in situ hybridization experiments and 3,185 EST libraries. Release 14 of Bgee is in preparation and will notably include 5,746 RNA-seq libraries from 29 animal species, including 4,860 human libraries from the GTEx project22,23.\n\nUntil recently the Bgee database lacked programmatic access to data through an R package, a shortcoming that we have addressed with the release of the BgeeDB Bioconductor package, available at http://www.bioconductor.org/packages/BgeeDB/. The package provides functions for fast extraction of data and metadata. The data structures used in the package can be easily incorporated with other Bioconductor packages, offering a wide range of possibilities for downstream analyses.\n\nMoreover, in BgeeDB we introduce the possibility to run TopAnat analyses, i.e., anatomical expression enrichment tests on gene lists provided by the user. This functionality is based on the topGO package24,25, modified to use Bgee data (A. Alexa, personal communication). TopAnat is similar to the widely used Gene Ontology enrichment test26–28. But in our case, the enrichment test is applied to terms from an anatomical ontology, mapped to genes by expression patterns. As a result, TopAnat allows for discovery of tissues where a set of genes is preferentially expressed. This feature is available as a web-tool at http://bgee.org/?page=top_anat, but the R package offers more flexibility in the choice of input data and analysis parameters, and possibilities of inclusion within programs or pipelines.\n\nIn the following sections we provide some typical examples of usage of the BgeeDB package.\n\n\nMethods\n\nR >= 3.3\n\nBioconductor >= 3.4\n\nBgeeDB package version >= 2.0.0\n\nWorking internet connection\n\n\n\n\nUse cases\n\nThe first step of data retrieval is to initialize a new Bgee reference class object, for a targeted species and data type. Normalized expression levels are currently available in the BgeeDB package for two data types: Affymetrix microarrays and Illumina RNA-seq. The list of species available in the Bgee database for each data type, along with their NCBI taxonomy IDs and common names can be obtained with the listBgeeSpecies() function. By default, data will be downloaded from the latest Bgee release, but this can be changed with the release argument.\n\nNext, the functions getAnnotation(), getData(), and formatData() can be called to respectively download the annotations of datasets, download the actual expression data, and reformat the expression data for more convenient use. Of note, BgeeDB creates a directory to store the downloaded annotation files and datasets, by default in the user’s R working directory, but this can be changed with the pathToData argument. These versioned cached files make it faster for the user to return to previously used data and allow for offline work.\n\nMicroarray dataset retrieval. In the following example, we look for a microarray dataset in mouse (Mus musculus), spanning multiple early developmental stages, including zygote. At the time of publication the latest Bgee release is 13.2, so if one needs to strictly reproduce the output of the code below in the future, the release=\"13.2\" argument needs to be added when creating the Bgee object (see Supplementary file S1 and Supplementary file S2).\n\n\n\nThis creates a list of two data frames, one including the annotation of experiments, and one including the annotation of each individual sample, i.e., hybridized microarray chip. For mouse, there are 694 Affymetrix experiments and 6,077 samples available in Bgee release 13. Anatomical structures and developmental stages are annotated using the Uberon ontology29,30. Below, we are selecting the experiments for which at least one sample is annotated to the zygote stage (UBERON:0000106).\n\n\n\nThis yields three microarray experiments, with accessions GSE1749, E-MEXP-51 and GSE18290. Among these, the accession E-MEXP-51, submitted to ArrayExpress by Wang and colleagues31, includes samples from more developmental stages than the other two, so we use this in the next steps. For this experiment, raw data were available from ArrayExpress, so samples were fully normalized with gcRMA32 version 2.40.0 through the Bgee pipeline.\n\n\n\nThe experiment includes 35 samples that passed Bgee quality controls. They originate from 12 developmental stages: primary and secondary oocyte, zygote, early, mid and late 2-cells embryo, 4-cells embryo, 8-cells embryo, 16-cells embryo, early, mid and late blastocyst, although the developmental stages ontology used is not precise enough yet to differentiate some of these conditions: the early, mid and late 2-cells stages are annotated as Theiler stage 2 embryo, and the 4-cells and 8-cells stages are annotated as Theiler stage 3 embryo. All samples were hybridized to the Affymetrix GeneChip Murine Genome U74Av2 microarray. Let us download the normalized probesets intensities measured for all samples.\n\n\n\nThe resulting data frame lists for each sample (column “Chip.ID”), the 9,017 probesets on the microarray (column “Probeset.ID”), their mapping to Ensembl gene IDs33 (column “Gene.ID”), their logged normalized intensities (column “Log.of.normalized.signal.intensity”), and a presence/absence call and quality (columns “Detection.flag” and “Detection.quality”).\n\nAs this format might not be the most convenient for downstream processing of an expression dataset, we offer the formatData() function, which creates an ExpressionSet object including the expression data matrix, the probesets annotation to Ensembl genes and the samples' anatomical structure and stage annotation into (assayData, featureData and phenoData slots respectively). This object class is of standard use in numerous Bioconductor packages.\n\n\n\nThe callType option of the formatData() function could alternatively be set to present or present high quality to display only the intensities of probesets detected as actively expressed.\n\nThe result is a nicely formatted Bioconductor object including expression data and their annotations, ready to be used for downstream analysis with other Bioconductor packages.\n\nRNA-seq dataset retrieval. We now search Bgee for a RNA-seq dataset sampling brain and liver tissues (Uberon Ids UBERON:0000955 and UBERON:0002107 respectively) in macaque (Macaca mulatta), and including multiple biological replicates for each tissue.\n\n\n\nAccessions GSE4163734 and GSE3035235 both include biological replicates for brain and liver. We focus on GSE41637 for the next steps since it includes three replicates of each tissue, vs. only two for GSE30352. We download the dataset and reformat it to obtain an ExpressionSet including counts of mapped reads on each Ensembl gene for each sample.\n\n\n\nInstead of mapped read counts, it is also possible to fill the data matrix with expression levels in the RPKM unit (reads per kilobase per million reads), using the option stats=\"rpkm\". In the next Bgee release (release 14), it will be possible to obtain expression levels in the TPM unit (transcript per million)36,37 from pseudo-mapping of reads computed in Bgee using the Kallisto software38.\n\nPresence/absence calls retrieval. It is often difficult to compare expression levels across species39, and even within species, across datasets generated by different experimenters or laboratories40–42. Batch effects have indeed been shown to impact extensively gene expression levels, confounding biological signal differences.\n\nEncoding gene expression as present or absent in a sample allows a more robust comparison across such conditions. In addition to retrieving RNA-seq and Affymetrix quantitative expression levels, BgeeDB also allows to retrieve calls of presence or absence of expression computed in the Bgee database for each gene (RNA-seq) or probeset (Affymetrix), in the column “Detection.flag” of the data.E.MEXP.51 and data.GSE41637 objects created above. And interestingly, expression calls are also available in Bgee for ESTs and in situ hybridization data, as well as for the consensus of the four data types for each triplet “gene / tissue / developmental stage”.\n\nA powerful use of these expression calls is the anatomical expression enrichment test “TopAnat”. TopAnat uses a similar approach to Gene Ontology enrichment tests26, but genes are associated to the anatomical structures where they display expression, instead of to their functional classification. These tests allow detecting where a set of genes is preferentially expressed as compared to a background universe (Roux J., Seppey M., Sanjeev K., Rech de Laval V., Moret P., Artimo P., Duvaud S., Ioannidis V., Stockinger H., Robinson-Rechavi M., Bastian F.B.; unpublished report). We show an example of such an analysis in the section “Anatomical expression enrichment analysis” below.\n\nOf note, the expression calls imported from BgeeDB can also be used for other downstream analyses. For example, when studying protein-protein interaction datasets, it might be biologically relevant to retain only interactions for which both members are expressed in the same tissues43,44.\n\nClustering analysis. A variety of downstream analyses can be performed on the imported expression data. Below we detail an example of gene expression clustering analysis on the developmental time-series microarray experiment imported above. The analysis, performed with the Mfuzz package45,46 (version 2.34.0 for this paper), aims at uncovering genes with similar expression profiles across development. We can readily start with the ExpressionSet object previously created.\n\n\n\nThe resulting plot can be seen in Figure 1.\n\nThe x-axis displays sample names (column “Chip.ID” of the data.E.MEXP.51 object).\n\nDifferential expression analysis. Below, we detail a differential expression analysis, with the package edgeR47,48 (version 3.16.1 for this paper), on the previously imported RNA-seq dataset of macaque tissues. We aim at isolating genes differentially expressed between brain and liver.\n\n\n\nThe resulting plot can be seen in Figure 2.\n\nSignificantly differentially expressed genes (FDR < 1%) are highlighted in red.\n\nThe loadTopAnatData() function loads the names of anatomical structures, and relationships between them, from the Uberon anatomical ontology (based on parent-child “is_a” and “part_of” relationships). It also loads a mapping from genes to anatomical structures, based on the presence calls of the genes in the targeted species. These calls come from a consensus of all data types specified in the input Bgee class object. We recommend to use all available data types (RNA-seq, Affymetrix, EST and in situ hybridization) for both genomic coverage and anatomical precision, which is the default behavior if no dataType argument is specified when the Bgee class object is created.\n\nBy default, presence calls of both high and low quality are used, which can be changed with the confidence argument of the loadTopAnatData() function. Finally, it is possible to specify the developmental stage under consideration, with the stage argument. By default expression calls generated from samples of all developmental stages are used, which is equivalent to specifying stage=\"UBERON:0000104\" (“life cycle”, the root of the stage ontology). Data are stored in versioned tab-separated cached files that will be read again if a query with the exact same parameters is launched later, to save time and server resources, and to work offline.\n\nIn this example, we use expression calls for zebrafish genes using all sources of expression data.\n\n\n\nWe look at the expression localization of the genes with an annotated phenotype related to pectoral fin (i.e., genes which upon knock-out or knock-down led to abnormal phenotypes of pectoral fin or its components). Zebrafish phenotypic data are available from the ZFIN database49 and integrated into the Ensembl database50. We thus retrieve the targeted genes using the biomaRt51 Bioconductor package (version 2.30.0 for this paper).\n\n\n\nThis gives a list of 150 zebrafish genes implicated in the development and function of pectoral fin. The next step of the analysis relies on the topGO Bioconductor package. We prepare a modified topGOdata object allowing to handle the Uberon anatomical ontology instead of the Gene Ontology, and perform a GO-like enrichment test for anatomical terms. As for a classical topGO analysis, we need to prepare a vector including all background genes, and with values 0 or 1 depending if genes are part of the foreground or not. The choice of background is very important since the wrong background can lead to spurious results in enrichment tests52. Here we choose as background all zebrafish Ensembl genes with an annotated phenotype from ZFIN.\n\n\n\nAt this step, expression calls are propagated through the whole ontology (e.g., expression in the forebrain will also be counted as expression in the brain, the nervous system, etc). This can take some time, especially if the gene list is large.\n\nFinally, we launch an enrichment test for anatomical terms. The functions of the topGO package can directly be used at this step. See the vignette of this package for more details25. Here we use a Fisher test, coupled with the “weight” decorrelation algorithm.\n\n\n\nFinally, we implement a function to display results in a formatted table. By default anatomical structures are sorted by their test p-value, which is displayed along with the associated false discovery rate (FDR53) and the enrichment fold. Sorting on other columns of the table (e.g., on decreasing enrichment folds) is possible with the ordering argument. Of note, it is debated whether a FDR correction is relevant on such enrichment test results, since tests on different terms of the ontologies are not independent. An interesting discussion can be found in the vignette of the topGO package.\n\n\n\nThe 22 anatomical structures displaying a significant enrichment at a FDR threshold of 1% are show in Table 1. The first term is “paired limb/fin bud”, and the second “pectoral fin”. Other terms in the list, especially those with high enrichment folds, are clearly related to pectoral fins (e.g., “pectoral appendage cartilage tissue”), substructures of fins (e.g., “fin bone”), or located next to them (e.g., “ceratohyal cartilage”). This analysis shows that genes with phenotypic effects on pectoral fins are specifically expressed in or next to these structures. More generally, it proves the pertinence of TopAnat analysis for the characterization of lists of genes.\n\nThe “weight” algorithm of the topGO package was used to decorrelate the structure of the ontology.\n\n\nConclusion\n\nIn summary, the BgeeDB package serves as a bridge between data from the Bgee database and the R/Bioconductor environment, facilitating access to high-quality curated and re-analyzed gene expression datasets, and significantly reducing time for downstream analyses of the datasets. Moreover, it provides access to TopAnat, a new enrichment that makes sense of lists of genes by uncovering their preferential localization of expression in anatomical structures. The TopAnat workflow is straightforward; for users already using topGO in their analysis pipelines, performing a TopAnat analysis on the same gene list only requires 6 additional lines of code.\n\n\nSoftware availability\n\nSoftware available from: http://www.bioconductor.org/packages/BgeeDB/\n\nLatest source code: https://github.com/BgeeDB/BgeeDB_R\n\nArchived source code as at the time of publication: https://doi.org/10.5281/zenodo.16376854",
"appendix": "Author contributions\n\n\n\nAK and JR contributed equally to this work. AK developed the initial BgeeDB R package and made it available in Bioconductor. JR implemented the enrichment analyses, and refined the data download part. FBB developed the server-side responses. MRR and FBB tested and commented on the package development. AK and JR wrote the manuscript. All authors discussed the results and implications and commented on the manuscript at all stages.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by SIB Swiss Institute of Bioinformatics project Bgee, Swiss National Science Foundation grant 31003A_153341, SystemsX.ch project AgingX, and Etat de Vaud.\n\n\nSupplementary material\n\nR markdown file including code from the paper.\n\nClick here to access the data.\n\nPDF file including the results of execution of the code from File S1.\n\nClick here to access the data.\n\n\nReferences\n\nRung J, Brazma A: Reuse of public genome-wide gene expression data. Nat Rev Genet. 2013; 14(2): 89–99. PubMed Abstract | Publisher Full Text\n\nIoannidis JP, Allison DB, Ball CA, et al.: Repeatability of published microarray gene expression analyses. Nat Genet. 2009; 41(2): 149–55. PubMed Abstract | Publisher Full Text\n\nWan X, Pavlidis P: Sharing and reusing gene expression profiling data in neuroscience. Neuroinformatics. 2007; 5(3): 161–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Development Core Team: R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2007. Reference Source\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–21. 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\nKauffmann A, Rayner TF, Parkinson H, et al.: Importing ArrayExpress datasets into R/Bioconductor. Bioinformatics. 2009; 25(16): 2092–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis S, Meltzer PS: GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics. 2007; 23(14): 1846–7. PubMed Abstract | Publisher Full Text\n\nZhu Y, Stephens RM, Meltzer PS, et al.: SRAdb: query and use public next-generation sequencing data from within R. BMC Bioinformatics. 2013; 14(1): 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKolesnikov N, Hastings E, Keays M, et al.: ArrayExpress update--simplifying data submissions. Nucleic Acids Res. 2015; 43(Database issue): D1113–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarrett T, Wilhite SE, Ledoux P, et al.: NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res. 2013; 41(Database issue): D991–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeinonen R, Sugawara H, Shumway M, et al.: The sequence read archive. Nucleic Acids Res. 2011; 39(Database issue): D19–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrainStars Bioconductor package. http\n\nKasukawa T, Masumoto KH, Nikaido I, et al.: Quantitative expression profile of distinct functional regions in the adult mouse brain. PLoS One. 2011; 6(8): e23228. PubMed Abstract | Publisher Full Text | Free Full Text\n\nImmuneSpaceR Bioconductor package. http\n\nExpressionAtlas Bioconductor package. http\n\nPetryszak R, Keays M, Tang YA, et al.: Expression Atlas update--an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 2016; 44(D1): D746–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCollado-Torres L, Nellore A, Kammers K, et al.: recount: A large-scale resource of analysis-ready RNA-seq expression data. bioRxiv. 2016. Publisher Full Text\n\nFrazee AC, Langmead B, Leek JT: ReCount: A multi-experiment resource of analysis-ready RNA-seq gene count datasets. BMC Bioinformatics. 2011; 12(1): 449. PubMed Abstract | Publisher Full Text | Free Full Text\n\nrecount Bioconductor package. http\n\nBastian F, Parmentier G, Roux J, et al.: Bgee: Integrating and Comparing Heterogeneous Transcriptome Data Among Species. Data Integr Life Sci. 2008; 5109: 124–31. Publisher Full Text\n\nGTEx Consortium: Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: Multitissue gene regulation in humans. Science. 2015; 348(6235): 648–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelé M, Ferreira PG, Reverter F, et al.: Human genomics. The human transcriptome across tissues and individuals. Science. 2015; 348(6235): 660–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlexa A, Rahnenführer J, Lengauer T: Improved scoring of functional groups from gene expression data by decorrelating GO graph structure. Bioinformatics. 2006; 22(13): 1600–7. PubMed Abstract | Publisher Full Text\n\ntopGO Bioconductor package. http\n\nRhee YS, Wood V, Dolinski K, et al.: Use and misuse of the gene ontology annotations. Nat Rev Genet. 2008; 9(7): 509–15. 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\nDessimoz C, Škunca N editors: The Gene Ontology Handbook. Humana Press; 2017; XII, 305. Publisher Full Text\n\nMungall CJ, Torniai C, Gkoutos GV, et al.: Uberon, an integrative multi-species anatomy ontology. Genome Biol. 2012; 13(1): R5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaendel MA, Balhoff JP, Bastian FB, et al.: Unification of multi-species vertebrate anatomy ontologies for comparative biology in Uberon. J Biomed Semantics. 2014; 5(1): 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang QT, Piotrowska K, Ciemerych MA, et al.: A genome-wide study of gene activity reveals developmental signaling pathways in the preimplantation mouse embryo. Dev Cell. 2004; 6(1): 133–44. PubMed Abstract | Publisher Full Text\n\nWu Z, Irizarry RA, Gentleman R, et al.: A Model-Based Background Adjustment for Oligonucleotide Expression Arrays. J Am Stat Assoc. 2004; 99(468): 909–17. Publisher Full Text\n\nYates A, Akanni W, Amode MR, et al.: Ensembl 2016. Nucleic Acids Res. 2016; 44(D1): D710–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerkin J, Russell C, Chen P, et al.: Evolutionary dynamics of gene and isoform regulation in Mammalian tissues. Science. 2012; 338(6114): 1593–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrawand D, Soumillon M, Necsulea A, et al.: The evolution of gene expression levels in mammalian organs. Nature. 2011; 478(7369): 343–8. PubMed Abstract | Publisher 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): 281–5. PubMed Abstract | Publisher Full Text\n\nLi B, Dewey CN: RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011; 12: 323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBray NL, Pimentel H, Melsted P, et al.: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016; 34(5): 525–7. PubMed Abstract | Publisher Full Text\n\nRoux J, Rosikiewicz M, Robinson-Rechavi M: What to compare and how: Comparative transcriptomics for Evo-Devo. J Exp Zool B Mol Dev Evol. 2015; 324(4): 372–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGilad Y, Mizrahi-Man O: A reanalysis of mouse ENCODE comparative gene expression data [version 1; referees: 3 approved, 1 approved with reservations]. F1000Res. 2015; 4: 121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeek JT, Scharpf RB, Bravo HC, et al.: Tackling the widespread and critical impact of batch effects in high-throughput data. Nat Rev Genet. 2010; 11(10): 733–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkey JM, Biswas S, Leek JT, et al.: On the design and analysis of gene expression studies in human populations. Nat Genet. 2007; 39(7): 807–8. PubMed Abstract | Publisher Full Text\n\nDeane CM, Salwiński Ł, Xenarios I, et al.: Protein Interactions: Two Methods for Assessment of the Reliability of High Throughput Observations. Mol Cell Proteomics. 2002; 1(5): 349–56. PubMed Abstract | Publisher Full Text\n\nKotlyar M, Pastrello C, Sheahan N, et al.: Integrated interactions database: tissue-specific view of the human and model organism interactomes. Nucleic Acids Res. 2016; 44(D1): D536–D41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFutschik ME, Carlisle B: Noise-robust soft clustering of gene expression time-course data. J Bioinform Comput Biol. 2005; 3(4): 965–88. PubMed Abstract | Publisher Full Text\n\nMfuzz Bioconductor package. Reference Source\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\needgeR Bioconductor package. Reference Source\n\nHowe DG, Bradford YM, Conlin T, et al.: ZFIN, the Zebrafish Model Organism Database: increased support for mutants and transgenics. Nucleic Acids Res. 2013; 41(Database issue): D854–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpudich GM, Fernández-Suárez XM: Disease and Phenotype Data at Ensembl. Curr Protoc Hum Genet. 2011; Chapter 6: Unit 6.11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nbiomaRt Bioconductor package. Reference Source\n\nTimmons JA, Szkop KJ, Gallagher IJ: Multiple sources of bias confound functional enrichment analysis of global -omics data. Genome Biol. 2015; 16(1): 186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenjamini Y, Hochberg Y: Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Series B Stat Methodol. 1995; 57(1): 289–300. Reference Source\n\nKomljenovic A, Roux J, Robinson-Rechavi M, et al.: BgeeDB/BgeeDB_R: Bgee R package release 2.0.0. Zenodo. 2016. Data Source"
}
|
[
{
"id": "17925",
"date": "07 Dec 2016",
"name": "Virag Sharma",
"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 the manuscript, Komljenovic et al. present BgeeDB which is an R package for retrieval of expression datasets which have been curated. Additionally, they also provide a method (TopAnat) to determine tissue-specific enrichments for a given list of genes and species.\nThe former is a very useful resource because there is clearly a need for a database that provides gene expression datasets which are homogenous in nature and are of comparable quality. The BgeeDB database contains gene expression datasets from 17 species across different tissues and developmental stages, which is impressive. The fact that the database can be queried via a Bioconductor package should ensure that the database will be used by both - wet-lab biologists and computational scientists. Similarly, the TopAnat method also provides a useful functionality to determine anatomical expression enrichment on a user specified list.\nI have a few minor comments regarding the manuscript:\nThe authors should include some details about how they have reprocessed the gene expression datasets that are a part of BgeeDB. At the moment, it is rather unclear how this was achieved. I assume that the authors have an automated pipeline in place but it would be beneficial for readers to know how this was done.\n\nThe authors state that “TopAnat allows for discovery of tissues where a set of genes is preferentially expressed”. Is TopAnat the only tool that offers such a functionality? A brief background of similar tools that are currently available will be useful for the readers.\n\nI was not able to run the workflow that the authors have included in the Supplementary material:\nSee below:\n\nsource(\"https://bioconductor.org/biocLite.R\") biocLite(\"BgeeDB\") biocLite(c(\"edgeR\", \"Mfuzz\", \"biomaRt\")) library(BgeeDB) listBgeeSpecies()\nbgee_affymetrix <- Bgee$new(species=\"Mus_musculus\", dataType=\"affymetrix\", release=\"13.2\") Error in envRefSetField(.Object, field, classDef, selfEnv, elements[[field]]) : 'dataType' is not a field in class \"Bgee\"\n## Turns out that I need to use \"datatype\" instead of \"dataType\" bgee_affymetrix <- Bgee$new(species=\"Mus_musculus\", datatype=\"affymetrix\")\n\nbgee_affymetrix <- Bgee$new(species=\"Mus_musculus\", datatype=\"affymetrix\", release=\"13.2\") Error in envRefSetField(.Object, field, classDef, selfEnv, elements[[field]]) : 'release' is not a field in class \"Bgee\"\n####\nAt this moment, I did not try further. The authors need to clearly state what version of BgeeDB was used to create this workflow. If something has changed, then this needs to be appropriately addressed. I tried using “release=13.2” but it did not work.\n\nI did manage to run an enrichment test for anatomical terms though with some tweaking\n## Again an error message bgee_topanat <- loadTopAnatData(species=\"Danio_rerio\") Error in loadTopAnatData(species = \"Danio_rerio\") : Problem: the specified speciesId is not among the list of species in Bgee.\n## This works though myTopAnatData <- loadTopAnatData(species=\"7955\")\n####\nThe rest of the work-flow went smoothly and I was able to get a list of anatomical structures sorted by their p-value\nhead(tableOver)\n\norganId\n\norganName annotated significant 12 UBERON:0004357\n\npaired limb/fin bud 144\n\n41 2 UBERON:0000151\n\npectoral fin 420\n\n70 22 UBERON:2000040\n\nmedian fin fold 51\n\n18 9 UBERON:0003051\n\near vesicle 304\n\n41 15 UBERON:0005729\n\npectoral appendage field 16\n\n10 16 UBERON:0007390 pectoral appendage cartilage tissue 17\n\n9\n\nexpected foldEnrichment\n\npValue\n\nFDR 12\n\n7.15\n\n5.734266 1.622480e-22 1.445630e-19 2\n\n20.85\n\n3.357314 1.037552e-18 4.622296e-16 22\n\n2.53\n\n7.114625 7.171001e-12 2.129787e-09 9\n\n15.09\n\n2.717031 3.135769e-10 6.984926e-08 15\n\n0.79\n\n12.658228 4.004917e-10 7.136762e-08 16\n\n0.84\n\n10.714286 2.411891e-08 3.581659e-06\n\nIt would be useful if the authors could include a feature that allows the TopAnat method to print the 41 genes which represent the paired limb/fin bud. At some point, the users might want to revisit their gene lists and tag their genes based on the different anatomical structures.\nOther tools that perform Enrichment tests, for example Enrichr1, have this feature and this is extremely useful, in my opinion.",
"responses": [
{
"c_id": "3777",
"date": "07 Aug 2018",
"name": "Frederic Bastian",
"role": "Author Response",
"response": "The authors should include some details about how they have reprocessed the gene expression datasets that are a part of BgeeDB. At the moment, it is rather unclear how this was achieved. I assume that the authors have an automated pipeline in place but it would be beneficial for readers to know how this was done. There is now a complete and updated documentation for the Bgee pipeline: https://github.com/BgeeDB/bgee_pipeline We have included this information in the manuscript, as well as a brief outline of the analyses we perform, see \"Introduction\" section. --- The authors state that “TopAnat allows for discovery of tissues where a set of genes is preferentially expressed”. Is TopAnat the only tool that offers such a functionality? A brief background of similar tools that are currently available will be useful for the readers. We have added a paragraph describing similar tools, see end of the \"Introduction\" section. --- I was not able to run the workflow that the authors have included in the Supplementary material: [...] At this moment, I did not try further. The authors need to clearly state what version of BgeeDB was used to create this workflow. If something has changed, then this needs to be appropriately addressed. I tried using “release=13.2” but it did not work. We suspect that the reviewer did not use the latest version of the package (maybe the Bioconductor release itself needs to be updated first). The reviewer could maybe uninstall the BgeeDB package and rerun the following steps: source(\"https://bioconductor.org/biocLite.R\") biocLite(\"BgeeDB\") sessionInfo() With a package version >= 2.6.2, the errors should disappear. The R, Bioconductor, and BgeeDB package version requirements are listed at the beginning of the \"Methods\" section. If the problem persists, could the reviewer post the sessionInfo() results? Of note, the “release” argument is used to specify a particular Bgee release, but this is independent of the package version. --- I did manage to run an enrichment test for anatomical terms though with some tweaking ## Again an error message bgee_topanat <- loadTopAnatData(species=\"Danio_rerio\") Error in loadTopAnatData(species = \"Danio_rerio\") : Problem: the specified speciesId is not among the list of species in Bgee. ## This works though myTopAnatData <- loadTopAnatData(species=\"7955\") #### Again, this should be solved by updating to the last BgeeDB version --- The rest of the work-flow went smoothly and I was able to get a list of anatomical structures sorted by their p-value [...] It would be useful if the authors could include a feature that allows the TopAnat method to print the 41 genes which represent the paired limb/fin bud. At some point, the users might want to revisit their gene lists and tag their genes based on the different anatomical structures. Other tools that perform Enrichment tests, for example Enrichr, have this feature and this is extremely useful, in my opinion. This is a good point. It is possible to cross the geneList vector with the expression mapping present in the myTopAnatData object. Another approach is to use functions that are inherited from the topGO package. For the “paired limb/fin bud” term: myTerm <- \"UBERON:0004357\" termStat(myTopAnatObject, myTerm) # 198 genes mapped to this term for Bgee 14.0 and Ensembl 84 genesInTerm(myTopAnatObject, myTerm) # 48 significant genes mapped to this term for Bgee 14.0 and Ensembl 84 annotated <- genesInTerm(myTopAnatObject, myTerm)[[\"UBERON:0004357\"]] annotated[annotated %in% sigGenes(myTopAnatObject)] We have added this example at the end of the \"Anatomical expression enrichment analysis\" section."
}
]
},
{
"id": "18221",
"date": "14 Dec 2016",
"name": "Daniel S. Himmelstein",
"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 describes the BgeeDB R package, which provides a programmatic interface for accessing Bgee gene expression data. Bgee is a valuable resource because it integrates gene expression results across many experiments. Previously, I've used Bgee for its presence/absence of expression calls and its differential expression calls.\nIn my opinion, Bgee's ability to provide a genome-wide profile of expression for a given species, developmental stage, and anatomical structure is its most powerful capability. It was not clear to me whether BgeeDB provides this functionality. For example, can the user retrieve the normalized expression level across several experiments for the same species-stage-anatomy combination? In general, I think users will be more interested in this high-level functionality than the low level access BgeeDB currently provides. An example here would likely clear things up for me.\nIs it possible to integrate expression levels across Affymetrix and RNA-Seq experiments?\nThe Zenodo archive of the source code specifies GPLv3 as the license. This is great, but it's ideal to also add a LICENSE file to the GitHub.\nIt looks like there are at least two potential places where bug reports should be filed: on Bioconductor Support and GitHub Issues. It would be nice to clarify the preferred location for filing bug reports go and opening pull requests.\nCurrently, the GitHub repository BgeeDB/BgeeDB_R mentioned in the manuscript is forked from wirawara/BgeeDB. I expect this may cause some confusion, as BgeeDB/BgeeDB_R should be the upstream repository that users fork and contribute back to. If you make wirawara/BgeeDB private, this should break the relationship. @wirawara can then fork BgeeDB/BgeeDB_R to continue contributions if desired.\nFinally, I created some GitHub issues as part of this review:\nSample annotation variable names\n\nA less messy default download directory",
"responses": [
{
"c_id": "3776",
"date": "07 Aug 2018",
"name": "Frederic Bastian",
"role": "Author Response",
"response": "In my opinion, Bgee's ability to provide a genome-wide profile of expression for a given species, developmental stage, and anatomical structure is its most powerful capability. It was not clear to me whether BgeeDB provides this functionality. For example, can the user retrieve the normalized expression level across several experiments for the same species-stage-anatomy combination? In general, I think users will be more interested in this high-level functionality than the low level access BgeeDB currently provides. An example here would likely clear things up for me. Indeed there is currently no easy way to do this. As mentioned in https://github.com/BgeeDB/BgeeDB_R/issues/7, it would be nice to have a getDataByCondition function that would return all processed data for chips / libraries matching a queried organ/stage/(sex)/(strain). But it was hard to set priorities for the initial development (should the package complement the web interface, or be orthogonal to it?), and we will likely implement it in the near future. --- Is it possible to integrate expression levels across Affymetrix and RNA-Seq experiments? If the reviewer means to integrate present/absent expression calls, it is relatively easy to get all the genes expressed in one tissue and all sub-tissues from Affymetrix and RNA-Seq data, although a dedicated method could be added, for instance: library(BgeeDB) bgee_human <- Bgee$new(species='Homo_sapiens', dataType=c('rna_seq', 'affymetrix')) my_data <- loadTopAnatData(bgee_human) calls_by_tissue <- reverseSplit(my_data$gene2anatomy) # pick you favorite tissue, for example liver calls_by_tissue[[\"UBERON:0002107\"]] And this can be limited by stage too, for example: my_data <- loadTopAnatData(bgee_human, stage=\"UBERON:0000068\") We have noted in the issue 7 mentioned above to add a direct function to do this. If the reviewer means to integrate levels of expression, it is then not the aim of Bgee: Bgee integrate different data types and different experiments, processed and normalized independently (but in a consistent manner). --- The Zenodo archive of the source code specifies GPLv3 as the license. This is great, but it's ideal to also add a LICENSE file to the GitHub. We have added the LICENSE file to GitHub (GPL 3.0). --- It looks like there are at least two potential places where bug reports should be filed: on Bioconductor Support and GitHub Issues. It would be nice to clarify the preferred location for filing bug reports go and opening pull requests. We have added the preferred location for filing bug reports at the end of the \"Introduction\" section (GitHub), and in the DESCRIPTION file of the source code. --- Currently, the GitHub repository BgeeDB/BgeeDB_R mentioned in the manuscript is forked from wirawara/BgeeDB. I expect this may cause some confusion, as BgeeDB/BgeeDB_R should be the upstream repository that users fork and contribute back to. If you make wirawara/BgeeDB private, this should break the relationship. @wirawara can then fork BgeeDB/BgeeDB_R to continue contributions if desired. We thank the reviewer for the suggestion, we have now made wirawara/BgeeDB private. --- Finally, I created some GitHub issues as part of this review: Sample annotation variable names https://github.com/BgeeDB/BgeeDB_R/issues/5 We have replied on the issue. Our answer was that is a bit of a controversial topic. For example Google's R Style Guide (https://google.github.io/styleguide/Rguide.xml#identifiers) advise against the use of underscores (although they do not justify why, and we agree that the \"words separated with dots\" convention can be disturbing for python users). --- A less messy default download directory This point was discussed in https://github.com/BgeeDB/BgeeDB_R/issues/4. We notably mention that another directory can be specified by using the \"pathToData\" argument. This parameter is mentioned at the end of the section \"Data download and import of normalized expression levels\". We agree that a default directory should be used in future releases."
}
]
},
{
"id": "17980",
"date": "16 Dec 2016",
"name": "Leonardo Collado-Torres",
"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\nIn this manuscript the authors describe the BgeeDB Bioconductor package and show how to use it (as of Bioconductor 3.4) to interact with Bgee1 in order to get the data from Bgee into your R session. This allows users to then perform differential expression analyses and integrate Bgee with other data sets such as unpublished data. The manuscript includes code that shows how to use BgeeDB and showcases it's different features including their unique anatomical expression enrichment analysis method.\nI find interesting that you can use BgeeDB to get data from different platforms and from different organisms. Most of this can be done using other packages such as GEOquery, but BgeeDB makes it so the user doesn't have to do all the processing of the data and standarization over multiple projects.\nMy main concern with the manuscript in its current form and the BgeeDB package itself is the lack of clarity on how the data has been processed and how the anatomical expression test works. That is, it could potentially become a black box that produces interesting output but hides information that could be important.\nFor example, I'm sure some of the Affymetrix data could be downloaded with other packages and I do not know what would be the differences between the raw data and the data downloaded via BgeeDB. Is the data in BgeeDB normalized? If so, how? The help pages of BgeeDB, the package vignette, the original Bgee publication1 and http://bgee.org/?page=doc did not help me fully answer these questions (I might have missed the information). Maybe the functions in BgeeDB could print a message describing the main steps of how a given data set was processed or this could be added to the help pages. I currently ignore if all data sets were treated the same. For instance, is all the Affymetrix data normalized with the same method and same parameters? I assume that the answer is yes but I don't know. I suggest that the authors describe in more detail the data available in Bgee. The authors might want to consider making the processing code public at https://github.com/BgeeDB or citable via figshare.\nWith the anatomical expression test it's not clear to me how to interpret the results from BgeeDB::makeTable(). I understand that the authors will describe the details of how their anatomical test works in a future publication, which they did before with Bgee1 and Homolanto2. Ideally, the anatomical expression test would have been described first followed by BgeeDB. Without hindering the current plan, I believe that the authors could provide a summary of how TopAnat works. Then they can explain it fully in the planned future TopAnat publication. I am also curious on how users could use their own data to improve the TopAnat results, although that could be work for the TopAnat paper or future work.\nI think that the manuscript is overall well written and will be more appealing if the data and main features (TopAnat) are described in more detail. I hope that the authors are not discouraged by my report.\n\nBest, Leonardo\n\nMinor comments:\nI'm an author of recount3 which is incorrectly cited here. The pre-print version of https://jhubiostatistics.shinyapps.io/recount/ had data from 2040 different projects which together made up more than 60,000 RNA-seq samples. The current version has data from over 70,000 Illumina human RNA-seq samples from SRA, GTEx and TCGA.\n\nI don't think that it makes sense to include the str() calls in the paper. They do make sense in the supplementary material (the html and pdf rendered versions of the paper code) since those include the output. Also, while str() shows all the details of an object, it can encourage users to write code that depends on the internal structure of the object. You might want to consider adding accessor functions.\n\nIf you added indentation the code that runs over multiple lines would be easier to read. You can use the Bioconductor standard of using 4 spaces at the start of the line. Also make sure that object names don't get split into multiple lines. For example check the line after the \"list experiments including both brain and liver samples\" comment where \"Anatomical.entity.ID\" gets split into \"Anatomical.e\" and \"ntity.ID\" in the html version of the paper. Copy pasting works fine, but if someone prints the paper they might introduce errors can be avoided with better formatting. F1000Research's team should be able to tell you what is the character limit per line to use so that the PDF and HTML versions look great. The formatR package might be useful here.\n\nI would not use numerical indexes in the code since the results could change with time in such a way that the current code would not work in the future or worse, it might run without error but change the results in a way a new user would not notice. For example, change the code on the line after the \"order developmental stages\" comment which currently reads:\ndata.E.MEXP.51.formatted <- data.E.MEXP.51.formatted[, c(5,8,9,3,2,1,4,7,6)]\n\nThe comment that reads with \"retrieve anatomical structures enriched at a 1% FDR threshold\" is mixed with the code. That is, you are missing a new line character.\n\nReference 46 is incorrect. It's edgeR, not eedgeR.\n\nThe package's vignette is missing a title as currently shown at http://bioconductor.org/packages/release/bioc/html/BgeeDB.html.\n\nI recommend adding internal R links to your manual pages. For example, ?topAnat mentions loadTopAnatData(). Those links make it easier for a user to browse the help pages.\n\nI was able to run all the code without any edits (beyond that new line issue I already mentioned) using Bioconductor 3.4 (current Bioc-release) on R 3.3.1. Here are my session details:\n> options(width = 120) > devtools::session_info() Session info ----------------------------------------------------------------------------------------------------------- setting value\n\nversion R version 3.3.1 (2016-06-21) system\n\nx86_64, mingw32\n\nui\n\nRStudio (0.99.902)\n\nlanguage (EN)\n\ncollate English_United States.1252\n\ntz\n\nAmerica/Mexico_City\n\ndate\n\n2016-12-15\n\nPackages --------------------------------------------------------------------------------------------------------------- package\n\n* version date\n\nsource\n\nAnnotationDbi * 1.36.0\n\n2016-10-18 Bioconductor\n\nassertthat\n\n0.1\n\n2013-12-06 CRAN (R 3.3.1) BgeeDB\n\n* 2.0.0\n\n2016-10-18 Bioconductor\n\nBiobase\n\n* 2.34.0\n\n2016-10-18 Bioconductor\n\nBiocGenerics * 0.20.0\n\n2016-10-18 Bioconductor\n\nBiocInstaller * 1.24.0\n\n2016-10-18 Bioconductor\n\nbiomaRt\n\n* 2.30.0\n\n2016-10-18 Bioconductor\n\nbitops\n\n1.0-6\n\n2013-08-17 CRAN (R 3.3.1) class\n\n7.3-14\n\n2015-08-30 CRAN (R 3.3.1) data.table\n\n1.10.0\n\n2016-12-03 CRAN (R 3.3.2) DBI\n\n0.5-1\n\n2016-09-10 CRAN (R 3.3.1) devtools\n\n1.12.0\n\n2016-06-24 CRAN (R 3.3.1) digest\n\n0.6.10\n\n2016-08-02 CRAN (R 3.3.1) dplyr\n\n0.5.0\n\n2016-06-24 CRAN (R 3.3.1) DynDoc\n\n* 1.52.0\n\n2016-10-18 Bioconductor\n\ne1071\n\n* 1.6-7\n\n2015-08-05 CRAN (R 3.3.1) edgeR\n\n* 3.16.4\n\n2016-11-27 Bioconductor\n\nGO.db\n\n* 3.4.0\n\n2016-10-22 Bioconductor\n\ngraph\n\n* 1.52.0\n\n2016-10-18 Bioconductor\n\nIRanges\n\n* 2.8.1\n\n2016-11-08 Bioconductor\n\nlattice\n\n0.20-34 2016-09-06 CRAN (R 3.3.1) limma\n\n* 3.30.6\n\n2016-11-29 Bioconductor\n\nlocfit\n\n1.5-9.1 2013-04-20 CRAN (R 3.3.1) magrittr\n\n1.5\n\n2014-11-22 CRAN (R 3.3.1) matrixStats\n\n0.51.0\n\n2016-10-09 CRAN (R 3.3.1) memoise\n\n1.0.0\n\n2016-01-29 CRAN (R 3.3.1) Mfuzz\n\n* 2.34.0\n\n2016-10-18 Bioconductor\n\nR6\n\n2.2.0\n\n2016-10-05 CRAN (R 3.3.1) Rcpp\n\n0.12.8\n\n2016-11-17 CRAN (R 3.3.2) RCurl\n\n1.95-4.8 2016-03-01 CRAN (R 3.3.1) rsconnect\n\n0.6\n\n2016-11-21 CRAN (R 3.3.2) RSQLite\n\n1.1-1\n\n2016-12-10 CRAN (R 3.3.2) S4Vectors\n\n* 0.12.1\n\n2016-12-01 Bioconductor\n\nSparseM\n\n* 1.74\n\n2016-11-10 CRAN (R 3.3.2) tibble\n\n1.2\n\n2016-08-26 CRAN (R 3.3.1) tidyr\n\n* 0.6.0\n\n2016-08-12 CRAN (R 3.3.1) tkWidgets\n\n1.52.0\n\n2016-10-18 Bioconductor\n\ntopGO\n\n* 2.26.0\n\n2016-10-18 Bioconductor\n\nwidgetTools\n\n* 1.52.0\n\n2016-10-18 Bioconductor\n\nwithr\n\n1.0.2\n\n2016-06-20 CRAN (R 3.3.1) XML\n\n3.98-1.5 2016-11-10 CRAN (R 3.3.2)\n\nRegarding Virag Sharma's peer review report4, I assume that Virag was using an earlier R version (and thus an earlier Bioconductor version). The current development version of BgeeDB uses \"dataType\" and not \"datatype\", just like the release version. Check https://github.com/Bioconductor-mirror/BgeeDB/search?utf8=%E2%9C%93&q=datatype. Hopefully the authors won't change the spelling of arguments in the future since that's confusing for users, although that's certainly doable following the deprecated/defunct code cycle.\nRegarding Daniel S. Himmelstein's peer review report5, there is no need to add a license file when the license is specified in the DESCRIPTION file of an R package. See https://github.com/Bioconductor-mirror/BgeeDB/blob/master/DESCRIPTION#L14 where they state that the license is GPL-2. Although the authors should make sure that they correctly specify which license their software is released on: GPL-2 or GPLv3 as Daniel mentioned. Regarding where to place bug reports, the authors could resolve this by specifying the \"BugReports\" field in their DESCRIPTION file. For example see https://github.com/Bioconductor-mirror/recount/blob/master/DESCRIPTION#L63. I also agree with Daniel that currently BgeeDB has a bit of a messy download structure. I would prefer if the files were downloaded in a single directory (say \"bgee_downloads\") instead of the current working directory.",
"responses": [
{
"c_id": "3775",
"date": "07 Aug 2018",
"name": "Frederic Bastian",
"role": "Author Response",
"response": "My main concern with the manuscript in its current form and the BgeeDB package itself is the lack of clarity on how the data has been processed and how the anatomical expression test works. That is, it could potentially become a black box that produces interesting output but hides information that could be important. For example, I'm sure some of the Affymetrix data could be downloaded with other packages and I do not know what would be the differences between the raw data and the data downloaded via BgeeDB. Is the data in BgeeDB normalized? If so, how? The help pages of BgeeDB, the package vignette, the original Bgee publication1 and http://bgee.org/?page=doc did not help me fully answer these questions (I might have missed the information). We have made public the Bgee pipeline source code at https://github.com/BgeeDB/bgee_pipeline. We also have added a paragraph at the end of the \"Introduction\" section, pointing to the relevant part of the documentation for RNA-Seq and Affymetrix analyses, and describing them in brief. --- Maybe the functions in BgeeDB could print a message describing the main steps of how a given data set was processed or this could be added to the help pages. We have opened an issue on our tracker related to this point, see https://github.com/BgeeDB/BgeeDB_R/issues/22. We will add a function pointing to the relevant documentation in a future release. --- I currently ignore if all data sets were treated the same. For instance, is all the Affymetrix data normalized with the same method and same parameters? I assume that the answer is yes but I don't know. The Affymetrix data are not treated in the same way depending on whether the raw data were available, or only the data processed by using the MAS5 software. This is clarified at the end of the \"Introduction\" section. Also, in the package, this information about raw data availability can be retrieved in the annotation data frame. --- I suggest that the authors describe in more detail the data available in Bgee. The authors might want to consider making the processing code public at https://github.com/BgeeDB or citable via figshare. We have made public the Bgee pipeline source code at https://github.com/BgeeDB/bgee_pipeline. --- With the anatomical expression test it's not clear to me how to interpret the results from BgeeDB::makeTable(). I understand that the authors will describe the details of how their anatomical test works in a future publication, which they did before with Bgee and Homolanto. Ideally, the anatomical expression test would have been described first followed by BgeeDB. Without hindering the current plan, I believe that the authors could provide a summary of how TopAnat works. Then they can explain it fully in the planned future TopAnat publication. We have added a brief description of how TopAnat works in the \"Introduction\" section. --- I am also curious on how users could use their own data to improve the TopAnat results, although that could be work for the TopAnat paper or future work. This represents an advanced use of TopAnat that we don't find suitable for the paper. But users can override the association file, mapping genes to anatomical structures in the BgeeDB directory, to use their own data. Also, since the source code of the package is public, users can also modify the mapping files used by modifying the source code. --- I'm an author of recount which is incorrectly cited here. The pre-print version of https://jhubiostatistics.shinyapps.io/recount/ had data from 2040 different projects which together made up more than 60,000 RNA-seq samples. The current version has data from over 70,000 Illumina human RNA-seq samples from SRA, GTEx and TCGA. We have updated the number in our paper. We apologize for the mistake. --- I don't think that it makes sense to include the str() calls in the paper. They do make sense in the supplementary material (the html and pdf rendered versions of the paper code) since those include the output. Also, while str() shows all the details of an object, it can encourage users to write code that depends on the internal structure of the object. You might want to consider adding accessor functions. We have removed str() calls from the paper. For the future, we will think of adding accessor functions, although several are already available thanks to the topGO package. --- If you added indentation the code that runs over multiple lines would be easier to read. You can use the Bioconductor standard of using 4 spaces at the start of the line. Also make sure that object names don't get split into multiple lines. For example check the line after the \"list experiments including both brain and liver samples\" comment where \"Anatomical.entity.ID\" gets split into \"Anatomical.e\" and \"ntity.ID\" in the html version of the paper. Copy pasting works fine, but if someone prints the paper they might introduce errors can be avoided with better formatting. F1000Research's team should be able to tell you what is the character limit per line to use so that the PDF and HTML versions look great. The formatR package might be useful here. We didn't know about the formatR package and will have a look at it. In the meantime, we have split such offending lines, as identified by the reviewer. --- I would not use numerical indexes in the code since the results could change with time in such a way that the current code would not work in the future or worse, it might run without error but change the results in a way a new user would not notice. For example, change the code on the line after the \"order developmental stages\" comment which currently reads: data.E.MEXP.51.formatted <- data.E.MEXP.51.formatted[, c(5,8,9,3,2,1,4,7,6)] We have replaced all lines using numerical indexes, with use of column names. --- The comment that reads with \"retrieve anatomical structures enriched at a 1% FDR threshold\" is mixed with the code. That is, you are missing a new line character. This was fixed. --- Reference 46 is incorrect. It's edgeR, not eedgeR. This was fixed. --- The package's vignette is missing a title as currently shown at http://bioconductor.org/packages/release/bioc/html/BgeeDB.html. This was added. --- I recommend adding internal R links to your manual pages. For example, ?topAnat mentions loadTopAnatData(). Those links make it easier for a user to browse the help pages. We thank the reviewer for the suggestion, and we will implement this in a future release. --- I was able to run all the code without any edits (beyond that new line issue I already mentioned) using Bioconductor 3.4 (current Bioc-release) on R 3.3.1. Here are my session details: > options(width = 120) > devtools::session_info() [...] Regarding Virag Sharma's peer review report4, I assume that Virag was using an earlier R version (and thus an earlier Bioconductor version). The current development version of BgeeDB uses \"dataType\" and not \"datatype\", just like the release version. Check https://github.com/Bioconductor-mirror/BgeeDB/search?utf8=%E2%9C%93&q=datatype. Hopefully the authors won't change the spelling of arguments in the future since that's confusing for users, although that's certainly doable following the deprecated/defunct code cycle. This is indeed a change that we introduced in an earlier version, in an effort to name all our arguments in a consistent manner. We will try not to change this in the future. --- Regarding Daniel S. Himmelstein's peer review report5, there is no need to add a license file when the license is specified in the DESCRIPTION file of an R package. See https://github.com/Bioconductor-mirror/BgeeDB/blob/master/DESCRIPTION#L14 where they state that the license is GPL-2. Although the authors should make sure that they correctly specify which license their software is released on: GPL-2 or GPLv3 as Daniel mentioned. We have updated the DESCRIPTION file in the development branch of Bioconductor. The package is now released under the GPL-3.0 license. --- Regarding where to place bug reports, the authors could resolve this by specifying the \"BugReports\" field in their DESCRIPTION file. For example see https://github.com/Bioconductor-mirror/recount/blob/master/DESCRIPTION#L63. This was done. --- I also agree with Daniel that currently BgeeDB has a bit of a messy download structure. I would prefer if the files were downloaded in a single directory (say \"bgee_downloads\") instead of the current working directory. While another directory can be specified by using the \"pathToData\" argument, it is true that the solution proposed by the reviewer would be convenient, and we will try to update the package accordingly in the future."
}
]
}
] | 1
|
https://f1000research.com/articles/5-2748
|
https://f1000research.com/articles/6-1834/v1
|
13 Oct 17
|
{
"type": "Research Article",
"title": "Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico",
"authors": [
"Dhanya Haridas",
"Justin C. Biffinger",
"Thomas J. Boyd",
"Preston A. Fulmer",
"Leila J. Hamdan",
"Lisa A. Fitzgerald",
"Dhanya Haridas",
"Justin C. Biffinger",
"Thomas J. Boyd",
"Preston A. Fulmer",
"Leila J. Hamdan"
],
"abstract": "Background: Shipwrecks serve as a rich source for novel microbial populations that have largely remained undiscovered. Low temperatures, lack of sunlight, and the availability of substrates derived from the shipwreck’s hull and cargo may provide an environment in which microbes can develop unique metabolic adaptations.\n\nMethods: To test our hypothesis that shipwrecks could influence the microbial population involved in denitrification when a consortium is grown in the laboratory, we collected samples proximate to two steel shipwrecks in the northern Gulf of Mexico. Then under laboratory conditions, we grew two independent denitrifying microbial consortia. Each consortium was grown by using the BART assay system and analyzed based on growth kinetics, ion chromatography and 16S amplicon sequencing. Results: Both denitrifying consortia were different from each other based on varied growth profiles, rates of nitrate utilization and 16S amplicon sequencing. Conclusions: Our observations conclude that the laboratory grown water column microbial consortia from deep-sea shipwrecks in the Gulf of Mexico are able to undergo aggressive denitrification.",
"keywords": [
"shipwreck",
"denitrification",
"microbe",
"16S",
"Gulf of Mexico"
],
"content": "Introduction\n\nThe biogeochemical process that transforms dissolved inorganic nitrogen to nitrogen gas is known as denitrification (DN). This metabolic pathway impacts the nitrogen (N) cycle by returning elemental N to the atmosphere1,2. It can alternatively be defined as the reduction of more oxidized forms of nitrogen (NO3-, NO2-, NO and N2O) to N2 gas, where it can be linked to the oxidation of iron, sulfur and reduced carbon species3. It is primarily performed by facultative heterotrophic or chemolithoautotrophic bacteria under anoxic or very low-oxygen conditions3, where microorganisms utilize nitrate or nitrite as the terminal electron acceptor4. DN, alongside other biogeochemical processes (carbon and sulfur cycles), plays a key role in maintaining the nutrient balance in marine habitats5.\n\nIn recent years, shipwrecks have been identified as areas from which novel microbial species have been isolated, because of the introduction of foreign material to the area6. Thus, they would be an ideal location to discover unique microorganisms and metabolic activity, as these areas are known to be diverse habitats for macroorganisms in the marine environment7. The goal of this research was to prospect for novel DN microbial consortia near deep-sea shipwrecks in the Gulf of Mexico, culture the consortia under laboratory conditions and determine their DN activity. In this study, we collected water samples proximal to two steel shipwreck sites located in the northern part of the Gulf of Mexico, and analyzed the denitrifying and culturing potential of the microbial consortia obtained from the two sites.\n\n\nMethods\n\nWe obtained water samples ~600 m down current from two steel-hulled shipwrecks investigated as part of the Shipwreck Corrosion, Hydrocarbon Exposure, Microbiology and Archaeology (SCHEMA) study, which addresses the effect of the 2010 Deepwater Horizon spill on deep-sea shipwrecks in the northern Gulf of Mexico (http://www.boem.gov/GOM-SCHEMA/). Samples were collected onboard the R/V Pelican using a CTD-rosette during the PE15-22 expedition in May 2015. The shipwreck Halo, is a steel-hulled steam tanker, resting in ~140 m of water, and ~50 miles west of the Mississippi River’s Southwest Pass. The double steel-hulled German U-Boat U-166 shipwreck rests in ~1400 m of water within 10 km of the Macondo wellhead, the epicenter of the 2010 Deepwater Horizon spill. The water samples were stored in sterile plastic bottles at 4°C until further use.\n\nThe commercially available denitrifying Biological Activity Reaction Test (DN-BART) assay (HACH, Colorado, USA) was used to enrich for DN bacteria. Briefly, the lyophilized media in the DN-BART was solubilized with 15 mL water sample from either the Halo or U-166 shipwreck site. The assay was performed based on manufacturer’s instructions, with the exception that the assay was incubated for 30 days instead of the suggested 4 days. The enriched microbial consortium obtained from the DN-BART assay was used as the inoculum to perform the growth curve and Ion Chromatographic (IC) studies.\n\nNunc tubes (Chemglass, NJ, USA) containing 10 mL of modified Indole Nitrite medium (pancreatic digest of casein 20 g/L, disodium phosphate 2 g/L, dextrose 1 g/L, potassium nitrate 1 g/L) were used for all assays. Sterile nitrogen gas was bubbled through the media for 15 min prior to inoculation to de-gas and establish an anaerobic environment. The enriched Halo and U-166 DN consortium derived from the DN-BART assay was used as the inoculum to perform the growth curve studies. Each tube was inoculated with 100 µL of DN-BART consortium. The inoculated Nunc tubes were analyzed for a period of 24 h at 30°C (Excella E25, Fisher Scientific, MA, USA). The optical density of the samples was measured at 600 nm (OD600) every 2 h post-inoculation using the Spectronic 200 Spectrophotometer (Thermo Scientific, PA, USA) over a 24 h period after inoculation. All inoculated samples were done in triplicates.\n\nThe nitrate/nitrite concentrations were tracked by IC using a Dionex IC-3000 IC fitted with an IonPac AS16 column. The mobile phase was 9 mM Na2CO3 at 1.0 mL/min and a Dionex 7 anion standard mix was used for calibration before and half way through sample runs. The DN microbial consortium was cultured similarly to the growth curve assay (i.e. anaerobically to enable the denitrifying conditions to develop). The Halo and U-166 cultures were sampled (1 mL) using a sterile syringe and needle (BD, NJ, USA) every 2 h over a period of 24 h from the inoculated Nunc tubes. The sample was centrifuged at 12,000 × g for 3 min in a sterile 1.5 mL eppendorf tube (Eppendorf, NY, USA). The supernatant was additionally syringe filtered (0.2 µm filter Syringefilter, SC, USA) and stored in 1 mL IC vials (Thermo Scientific, Pittsburg, PA). All analyzes were performed in duplicate or triplicate. Nitrate concentrations were often above the calibration level (100 mg/L) and are annotated as estimated values (J). Standard errors for replicate measurements ranged from 0 to 2.26% with an average of 0.31% for the aggregate runs.\n\nTo determine the DN phylotypes present in the DN consortia, genomic DNA was isolated from the Halo and U-166 denitrifying microbial consortia after 24 h of growth and 16S amplicon sequencing of the V4 region was performed by Seqmatic (Fremont, CA).\n\n\nResults\n\nThe depth, temperature, salinity and dissolved oxygen (DO) of the water samples obtained from the region proximal to the Halo and U-166 shipwreck sites were obtained with the CTD (Table 1). The water sample obtained proximal to the U-166 shipwreck site had a higher DO content (6.6 mg/L) and lower temperature (4.3°C), as compared to the Halo shipwreck site that had 4.1 mg/L DO and 17.7°C. This indicates that both the water samples have varied hydrographic conditions.\n\nThe depth, temperature, salinity and dissolved oxygen (DO) for the water samples collected from Halo and U-166 are listed below.\n\nBoth water samples were also analyzed for the presence of denitrifying microbial consortia using the commercially available DN-BART assay. Upon performing the assay, it was observed that Halo and U-166 water samples did not produce foam or bubbles around the ball or in the tube after 4 days, the recommended duration for developing a positive reaction. Hence, due to the nature of the unique water samples, the assay was continued for 30 days. Following the 30 days, foam was detected around the ball, providing evidence for potential DN microbial consortia from the Halo and U-166 shipwreck sites. The DN consortia that were enriched using the DN-BART assay were further analyzed for microbial growth, nitrate/nitrite media concentrations and microbial composition over 24 h.\n\nTo determine the growth profile of the DN consortia, the growth curve assay was performed. It was observed that the DN consortia from Halo grew (OD600 = 0.980) much slower than the U-166 consortia (OD600 = 2.448) over the 24 h period (Figure 1; Dataset 1). Thus providing the first evidence that both the DN consortia are different from each other. All analyses were performed in triplicate (Figure 1).\n\nThe growth profiles of both DN consortia were analyzed for a period of 24 h at 0D600. The U-166 DN consortia grew at a much faster rate compared to the Halo DN consortia.\n\nIon chromatography (IC) studies were performed to identify the denitrifying potential of the isolated microbial consortia. The Halo microbial supernatants showed a steady decline in nitrate concentration (734 mg/L to 0.7 mg/L) as microbial growth entered into the logarithmic growth phase. As the nitrate concentration decreased, there was an increase in nitrite concentration from 1.4 mg/L, to a maximum of 130 mg/L and tapered down to 4.3 mg/L at 24 h. The U-166 microbial consortium rapidly converted nitrate into nitrite, as shown with a decrease in nitrate concentration (730 mg/L to 2.5 mg/L) followed by an increase in nitrite concentration (0 to 240 mg/L), which was later followed by a subsequent decrease in nitrite levels to 2.2 mg/L (Figure 2; Dataset 2).\n\nSamples were collected every 2 h for a period of 24 h and nitrate and nitrite levels were determined. Note: Nitrate values were above the calibration level (100 mg/L) and are thus estimates (but proportional).\n\nSince the growth curve and IC studies indicated that the DN consortia from Halo and U-166 are mutually exclusive, we wanted to determine the microbial composition of both Halo and U-166 DN consortia using 16S amplicon sequencing. The Halo DN consortium primarily consisted of the Pseudomonas genus (98.1%), while the U-166 DN consortium was dominated by the Citrobacter genus (72.6%). At the species level, P. tropicalis and P. aeruginosa for Halo, and C. werkmanii and C. freundii for U-166 were primarily detected (Figure 3; Dataset 3)), thus indicating that both DN consortia are mutually exclusive.\n\nGenomic DNA was isolated from the Halo and the U-166 DN consortium and the V4 region of the 16S was analyzed.\n\n\nDiscussion\n\nThe deep sea, identified with shelf depths greater than 200 m, has been documented to be the largest hypoxic and anoxic environment present on earth8. The varied living conditions mentioned earlier induces microbes to adopt unique metabolic adaptations. Hence, the marine dark biosphere has been recognized as a rich resource of unique microbial populations. Apart from the unique microbial life detected in the marine dark biosphere, shipwreck sites located in the deep sea also serve as a rich source of distinct flora and fauna6. Using two different shipwrecks at varying depth and material allows for the comparison of the metabolic activity of DN microbial consortia isolated from steel shipwreck sites.\n\nOne of the biggest challenges in characterizing new microbes from the deep-sea is the ability to successfully culture them in the laboratory. The initial approach to identifying a DN consortium was to assess growth using the commercially available DN-BART assay. The DN-BART assay provided the necessary nutrients in a modified nitrate medium and the presence of a potential DN microbial consortium from both Halo and U-166 shipwreck sites was confirmed. To further characterize the DN microbial consortia, the turbidity of the media was monitored and the nitrate/nitrite concentrations were examined every 2 h over a 24 h time period.\n\nThe water sample from the Halo shipwreck site was able to grow under the conditions set forth in this study, but at a much slower rate when compared to U-166. When the Halo microbial consortium began its logarithmic growth, there was a steady decline in the nitrate concentration and a subsequent increase in the nitrite concentration in the supernatant. The U-166 DN consortium also grew and the turbidity of the culture was greater as compared to Halo DN consortium. The IC studies further corroborated the decrease in nitrate levels and a concurrent increase in the nitrite concentration during the logarithmic phase of growth. Further, the U-166 consortia consumed nitrite, most likely as nitrate was completely consumed, at a rate 2 times slower than that observed in the Halo DN consortium (30 mg/L compared to 65 mg/L respectively). To determine the DN phylotypes present in the DN consortia, 16S amplicon sequencing was performed, and it was observed that at the species level P. tropicalis and P. aeruginosa for Halo and C. werkmanii and C. freundii for U-166 were the most dominant and are known denitrifiers9,10. It was also observed that the Citrobacter dominating the U-166 DN consortia consumed nitrate at a rate that was faster than other industrial microbial consortia containing Citrobacter adapted for denitrification11.\n\nThis study indicates that Halo and U-166 were good prospecting sites for novel microbial consortia related to denitrification. Each shipwreck site has a distinct DN consortium which can be grown under laboratory settings. The U-166 DN microbial consortium performs denitrification at a much faster rate than the Halo DN microbial consortium and most known industrial microbial consortia. This elevated DN activity could be the result of local hydrodynamic conditions or the proximity to the shipwreck, but additional studies are needed to identify the exact parameters. In conclusion, both DN consortia isolated from novel prospecting sites (shipwrecks) in the Gulf of Mexico can be cultured in the laboratory and can utilize a DN metabolic pathway for growth.\n\n\nData availability\n\nDataset 1: Growth curve assay: Growth curve studies were performed over a 24 h period for both the Halo and U-166 DN consortia. The optical density of the cultures were measured every 2 h at 600nm. DOI, 10.5256/f1000research.12713.d17976512\n\nDataset 2: Ion chromatography studies: Nitrate and nitrite levels of the DN consortia isolated from Halo and U-166 sites were determined every 2 h for a period of 24 h using ion chromatography. DOI, 10.5256/f1000research.12713.d17976613\n\nDataset 3: 16S amplicon sequencing: Genomic DNA was isolated from DN consortia after 24 h of growth and the V4 region was analyzed using the 16S metagenomics sequencing. DOI, 10.5256/f1000research.12713.d17976714",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nFunding was provided by the Office of Naval Research (ONR) through the Naval Research Laboratory (PE# 61153N), the Bureau of Ocean Energy Management (BOEM) No. M13PG00020, and the Navy Platform Support Program.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank the officers, crew and scientific party on board the R/V Pelican during the April-May 2015 expedition (PE15-22).\n\n\nReferences\n\nRivera-Monroy VH, Lenaker P, Twilley RR, et al.: Denitrification in coastal Louisiana: A spatial assessment and research needs. J Sea Res. 2010; 63(3–4): 157–172. Publisher Full Text\n\nSeitzinger S, Harrison JA, Böhlke JK, et al.: Denitrification across landscapes and waterscapes: a synthesis. Ecol Appl. 2006; 16(6): 2064–2090. PubMed Abstract | Publisher Full Text\n\nHulth S, Aller RC, Canfield DE, et al.: Nitrogen removal in marine environments: recent findings and future research challenges. Mar Chem. 2005; 94(1–4): 125–145. Publisher Full Text\n\nSeitzinger SP: Denitrification in freshwater and coastal marine ecosystems: Ecological and geochemical significance. Limnol Oceanogr. 1988; 33(4part2): 702–724. Publisher Full Text\n\nFalkowski PG: Evolution of the nitrogen cycle and its influence on the biological sequestration of CO2 in the ocean. Nature. 1997; 387: 272–275. Publisher Full Text\n\nSánchez-Porro C, Kaur B, Mann H, et al.: Halomonas titanicae sp. nov., a halophilic bacterium isolated from the RMS Titanic. Int J Syst Evol Microbiol. 2010; 60(Pt 12): 2768–2774. PubMed Abstract | Publisher Full Text\n\nPerkol-Finkel S, Shashar N, Benayahu Y: Can artificial reefs mimic natural reef communities? The roles of structural features and age. Mar Environ Res. 2006; 61(2): 121–135. PubMed Abstract | Publisher Full Text\n\nDanovaro R, Snelgrove PV, Tyler P: Challenging the paradigms of deep-sea ecology. Trends Ecol Evol. 2014; 29(8): 465–475. PubMed Abstract | Publisher Full Text\n\nHernandez D, Rowe JJ: Oxygen regulation of nitrate uptake in denitrifying Pseudomonas aeruginosa. Appl Environ Microbiol. 1987; 53(4): 745–750. PubMed Abstract | Free Full Text\n\nRehr B, Klemme JH: Formate dependent nitrate and nitrite reduction to ammonia by Citrobacter freundii and competition with denitrifying bacteria. Antonie Van Leeuwenhoek. 1989; 56(4): 311–321. PubMed Abstract | Publisher Full Text\n\nCyplik P, Juzwa W, Marecik R, et al.: Denitrification of industrial wastewater: Influence of glycerol addition on metabolic activity and community shifts in a microbial consortium. Chemosphere. 2013; 93(11): 2823–2831. PubMed Abstract | Publisher Full Text\n\nHaridas D, Biffinger J, Boyd T, et al.: Dataset 1 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. Data Source\n\nHaridas D, Biffinger J, Boyd T, et al.: Dataset 2 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. Data Source\n\nHaridas D, Biffinger J, Boyd T, et al.: Dataset 3 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. Data Source"
}
|
[
{
"id": "31967",
"date": "26 Mar 2018",
"name": "Maxim Rubin Blum",
"expertise": [
"Reviewer Expertise Marine Microbiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 aims to examine the cultivable denitrifying microbial consortia from marine environment near deep-sea shipwrecks in the Gulf of Mexico. The authors hypothesized that the shipwrecks could influence the microbial population involved in denitrification. The methods used in this study, however, have not adequately tested this hypothesis, as no control samples were obtained. On the other hand, the authors were able to report enrichment of two interesting bacterial consortia, one of which appeared to perform denitrification at a much faster rate than that measured for most known industrial microbial consortia, providing an opportunity to study these bacteria in more detail in the future.\n\nIf I understood correctly, only a single water sample from each location was used for the denitrifying Biological Activity Reaction Test, allowing to isolate a denitrifying consortium per sampling site. It is likely that these bacteria occur in a patchy manner in the environment, thus further sampling may reveal very different denitrifying consortia. Broadening the sampling effort is needed to test if either environmental factors or presence of shipwrecks affects the diversity of denitrifying bacteria.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "31968",
"date": "29 Mar 2018",
"name": "Dimitri Kalenitchenko",
"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 entitled “Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico” by Haridas et al. explores the potential of steel shipwrecks to host a specialized community with a focus on denitrifying bacteria. Overall the paper is well written and easy to follow, their results are relevant but before suggesting the acceptance of the manuscript, I suggest that authors resubmit a modified version of the manuscript that include a detailed material and method section.\n\n16S amplicon sequencing section :\nPlease provide the DNA extraction kit details Please provide the name of the sequencing platform and primers Please explain the bioinformatics treatment including the taxonomic affiliation used\n\nDenitrifying microbial consortia :\nWhy did the authors extend the incubation time up to 30 days, I am wondering if they checked for contamination from external source.\n16S amplicon sequencing :\nI am very surprised about the very low diversity they obtained, I agree that they enriched the community in DN bacteria but with only two tubes for two conditions I am not convinced they do not just select for a contaminant or a bacterium in each tube. How did the authors obtained these percentages, BLAST ?\n\nDiscussion\n\nI think the conclusion that shipwreck are good prospecting area for novel microbial consortia is highly speculative based on these data especially without a reference point before the water mass flow across the wreck.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3712",
"date": "15 Jun 2018",
"name": "Dhanya Haridas",
"role": "Author Response",
"response": "Dear Reviewer,Thank you for taking the time to review and suggest comments/edits to the manuscript. We have included information about the DNA isolation kit, the sequencing platform on which the 16S metagenomics studies were performed and the bioinformatics pipeline used. We hope the reviewer will find the responses satisfactory.You expressed your concern on the need for us to extend the incubation time from 4 days to 30 days for the BART assay. We appreciate the reviewer raising this concern and the reason we decided to extend the incubation time was due to the uncharacterized marine environmental samples used and it is known the concentration of viable microbes is less (as determined by a 3H-Leucine incorporation assay) than when grown in a rich medium. Hence extending the incubation time in this study resulted in an increased concentration of the viable microbial culture which helped determine if the assay was positive for microbes which could perform denitrification.You mentioned another concern about the study selecting for a contaminant. The raw data (excel file) containing all the sequencing tags obtained from the 16S metagenomics sequencing has been provided. The authors would like to mention that the microbial population obtained from the BART assay served as the starting inoculum for all the assays mentioned in the paper and therefore any potential aerobic or facultative aerobic bacteria cultured in the BART assay medium would not have propagated in the Nunc tube due to the anaerobic nature of the environment, thus further diminishing the diversity and potential for an aerobic contaminant. The percentages were part of the bioinformatics report submitted to us by Seqmatic. For example, in the Halo sample the total number of reads at the species level was 169,918 of which 59,117 reads were classified as Pseudomonas tropicalis (By Seqmatic) thus making it 34.72% of the total species population.We hope we addressed all your concerns in this response and thank you once again for reviewing the manuscript. Respectively,Dhanya Haridas"
}
]
}
] | 1
|
https://f1000research.com/articles/6-1834
|
https://f1000research.com/articles/6-2150/v1
|
19 Dec 17
|
{
"type": "Software Tool Article",
"title": "Software for web-based tic suppression training",
"authors": [
"Jonathan K. Black",
"Kevin J. Black",
"Jonathan K. Black"
],
"abstract": "Exposure and response prevention (ERP) is a first-line behavior therapy for obsessive-compulsive disorder, and has also been tested in Tourette syndrome (TS). However, ERP for tic disorders requires intentional tic suppression, which for some patients is difficult even for brief periods. Additionally, practical access to behavior therapy is difficult for many patients, especially those in rural areas. The authors present a simple, working web platform (TicTrainer) that implements a strategy called reward-enhanced exposure and response prevention (RE–ERP). This strategy sacrifices most expert therapist components of ERP, focusing only on increasing the duration of time for which the user can suppress tics through automated differential reinforcement of tic-free periods (DRO). RE–ERP requires an external tic monitor, such as a parent, during training sessions. The user sees increasing digital rewards for longer and longer periods of successful tic suppression, similar to a video game score. TicTrainer is designed with security in mind, storing no personally identifiable health information, and has features to facilitate research, including optional masked comparison of tics during DRO vs. noncontingent reward conditions. A working instance of TicTrainer is available from https://tictrainer.com/",
"keywords": [
"behavior therapy",
"software",
"tic disorders",
"Tourette syndrome",
"reward"
],
"content": "Introduction\n\nRecent years have seen increasing evidence for and acceptance of behavior therapies for tic disorders such as Tourette syndrome (Capriotti et al., 2014). Tic suppression plays a key role in these therapies (Specht et al., 2014). One of these is exposure and response prevention (ERP), a first-line treatment for obsessive-compulsive disorder. ERP also showed similar tic reduction efficacy to the most extensively studied behavior therapy for tics in a randomized, controlled trial (Verdellen et al., 2004). However, tic suppression, an essential component of ERP, is difficult or frustrating for some tic patients. Fortunately, we showed that even children with recent-onset tic disorders could suppress tics when brief tic-free periods were rewarded immediately by small tokens (Greene et al., 2015). Our experience with that study, and in an unrelated project, suggested to us that automating the process of immediate rewards for tic-free intervals might facilitate ERP, even in those who at first could suppress tics only for a few seconds at a time (Black et al., 2017; Miller et al., 2015).\n\nHere we present a simple, web-based tool to facilitate training intended to allow increasing periods of tic suppression. Design goals included using this program to record tics, to provide rewards for tic suppression in a video-game-like format that many children would be familiar with, to gradually increase rewards for increasing periods of tic suppression, to respect confidentiality, to gather anonymous information that can be used to assess use patterns and initial information about efficacy and safety, and to provide features that facilitate research use, i.e. creation of research subject accounts that can be assigned to different reward schedules at different points in time.\n\n\nMethods\n\nThe TicTrainer server is written in JavaScript, using the Node.js runtime.\n\nThe system is designed to ensure security of users’ personally identifying information. Instead of recording a name, each user is assigned a simple account ID used to log on. To track the collective age of users and to allow potential individualized training dependent on age, the user is asked his or her birth month and year. However, the system saves only a randomly chosen birthdate within 45 days of the 15th day of the month supplied.\n\nAdministrative accounts also can be created. Admins can flag certain users as research participants and assign them to receive rewards differently from regular users. They can also view system and session log files, or create another admin account.\n\nEach user and trainer account has its own text-based account data file on the server. This is a simple way to store modest amounts of data without using a database, at the possible expense of keeping numerous files open simultaneously if traffic to a single site becomes very heavy. System design may need to change if the number of users increases very substantially.\n\nA user’s ID, password, and other account information are sent back and forth as needed between client and server. When a user types his ID and password into the “Manage Account” sign in page, for example, the server sends back a unique web page that includes the credentials, as he/she provided them, and some information loaded from his/her account data. If he/she edits something in his/her account, those same credentials are sent back to the server for authentication along with the new data.\n\nAny page on the website that has information specific to a signed-in user is created dynamically by the server. These dynamic web pages are stored as .dynh files on the server. (“.dynh” is a custom extension which stands for DYNamic Html.) These files have designated locations for the server to inject the required unique fields (usually these include the user’s ID and password) before sending the edited page back to the client. Almost every page involved in a training session is stored and generated this way by the server, so as to preserve the user’s credentials as he/she typed them at the beginning of the session.\n\nDuring a session, the server mediates between the user and trainer via a session log file. When the trainer records a user’s tic, the trainer’s page sends an XMLHttpRequest asking the server to write a line to the log file. On the other side, the user’s page continually checks the session log file on the server for changes, so it can reset the rewards when the user tics and end the session when the trainer leaves. At the end of a session, the log file is archived with the end time in the filename.\n\nTicTrainer runs on a node.js server. Users need a web browser that supports javascript. The program may not function properly on browsers that do not support HTML5, and currently does not work with Microsoft browsers.\n\nA visitor to TicTrainer first registers an account. User and trainer accounts are created separately. Either one next goes to the “Manage Account” page and links to another account. Specifically, trainers specify the users they train, and users specify the trainers that can train them.\n\nDuring a session, the trainer is presented with two buttons: “Tic Detected,” and “End Session.” When “Tic Detected” is pressed, the server logs a tic. Trainers also see a 1-minute timer progressing continuously next to an “I’m Here” button, which restarts that timer (as does the “Tic Detected” button). If the trainer presses “End Session,” closes the page, or lets the “I’m Here” timer elapse, the session ends. The timer helps ensure that the trainer stays engaged in watching and recording the user’s tics.\n\nThe user’s session page displays a large counter for their current point total, followed by a superscript number indicating the current point accrual rate. Each time the user refrains from ticcing for a number of seconds equal to his/her level, their point total increases by the current rate, and the rate then increases by the square of the current level (rate is capped at 10 × levels2). Finally, a user “levels up” when his/her points exceed 1000 × levels2. He/she then also receives “coins” equal to the square of the previous level. The coins can be traded in for digital medals at an online store. Parents or clinicians may choose to provide tangible rewards for the digital coins or medals. The medals are displayed during training sessions and on the user’s “Manage Account” page. In total, this reward strategy is intended to provide users increasing incentives to suppress their tics for increasingly long intervals.\n\nFor potential use in controlled trials, research participant accounts can also be assigned to an alternative (control) reward strategy, noncontingent reward (Greene et al., 2015; Himle et al., 2008). In this case the admin user can set the initial mean frequency of rewards. This frequency may be set based on the participant’s previously recorded tic frequency, to better mask the treatment allocation. Thereafter the frequency of rewards increases automatically based on the subject’s achieved “level,” with the intention of approximately matching reward frequency with the two methods.\n\n\nUse cases\n\nFigure 1 shows a “user” window (for the person with tics) and a “trainer” window as they might appear during a session. Normally the two windows would appear on separate devices. This user is currently on level one, with 958 points, and is accruing 5 points for every second during which no tics are detected. This user has not yet earned any coins.\n\nLeft: the “user” window (person with tics). Right: the “trainer” window (clinician or other trained observer). Typically the two windows would appear on separate devices.\n\nSupplementary File 1 is a session log file for a test session (no human subjects were observed). Admin users can view or download these session log files, from which they can compute for each session any of the following:\n\nmeasures of tic frequency and tic suppression, e.g.\n\n◦ mean frequency of tics\n\n◦ longest tic-free interval\n\n◦ number of 10-second tic-free intervals\n\nnumber of rewards\n\nother metrics, e.g.\n\n◦ tests of whether inter-tic intervals fit a fractal pattern (Peterson & Leckman, 1998)\n\n◦ tests of the timing of tics vis-à-vis timing of rewards\n\n\nConclusions\n\nThis simple web-based platform is available at TicTrainer.com, and provides features that allow prospective trials including with different reward schedules. Features that could yet be implemented include adding self-report (and/or trainer-report) measures of inter-session tic severity or of other symptoms, or measures of premonitory sensations/urge intensity before, during or after sessions (Himle et al., 2007; Specht et al., 2014; Verdellen et al., 2008).\n\n\nSoftware availability\n\nTicTrainer available from: https://tictrainer.com/\n\nSource code available from https://github.com/jonkb/TicTrainer-node\n\nArchived source code as at time of publication: http://doi.org/10.5281/zenodo.1098270 (Black, 2017).\n\nLicense: MIT",
"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\nAcknowledgments\n\nA preprint of this work was posted on Authorea.com.\n\n\nSupplementary material\n\nSupplementary File 1: Sample regular log file. The log file from a sample training session. This test user account was not flagged for research, so this log file demonstrates the typical reward strategy (differential reinforcement of other behavior, or DRO).\n\nClick here to access the data.\n\n\nReferences\n\nBlack JK: TicTrainer V.2. Zenodo. 2017. Data Source\n\nBlack JK, Koller JM, Black KJ: TicTimer software for measuring tic suppression [version 1; referees: 1 approved, 1 approved with reservations]. F1000Research. 2017; 6: 1560. Publisher Full Text\n\nCapriotti MR, Himle MB, Woods DW: Behavioral Treatments for Tourette Syndrome. J Obsessive Compuls Relat Disord. Elsevier BV, 2014; 3(4): 415–20. Publisher Full Text\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\nHimle MB, Woods DW, Conelea CA, et al.: Investigating the effects of tic suppression on premonitory urge ratings in children and adolescents with Tourette's syndrome. Behav Res Ther. 2007; 45(12): 2964–76. PubMed Abstract | Publisher Full Text\n\nHimle MB, Woods DW, Bunaciu L: Evaluating the role of contingency in differentially reinforced tic suppression. J Appl Behav Anal. 2008; 41(2): 285–89. PubMed Abstract | Free Full Text\n\nPeterson BS, Leckman JF: The Temporal Dynamics of Tics in Gilles De La Tourette Syndrome. Biol Psychiatry. 1998; 44(12): 1337–48. PubMed Abstract | Publisher Full Text\n\nMiller B, Lim AN, Heidbreder AF, et al.: An Automated Motion Detection and Reward System for Animal Training. Cureus. 2015; 7(12): e397. 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(12): 297–318. PubMed Abstract | Publisher Full Text\n\nVerdellen CW, Keijsers GP, Cath DC, et al.: Exposure with Response Prevention versus Habit Reversal in Tourettes’s Syndrome: a Controlled Study. Behav Res Ther. 2004; 42(5): 501–11. PubMed Abstract | Publisher Full Text\n\nVerdellen CW, Hoogduin CA, Kato BS, et al.: Habituation of Premonitory Sensations during Exposure and Response Prevention Treatment in Tourette’s Syndrome. Behav Modif. 2008; 32(2): 215–27. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "30166",
"date": "02 Feb 2018",
"name": "Christine A. Conelea",
"expertise": [
"Reviewer Expertise Tic disorders",
"behavioral paradigms"
],
"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 TicTrainer, a newly created software program designed to track tic occurrences and deliver reward for successful suppression during live observation. Tic suppression contingent reward has reliably been shown to acutely enhance tic suppression in laboratory settings, but little work has been done to date to translate this finding into the clinical context. TicTrainer is an important step in this direction, and appears to be a promising tool for both research and clinical applications. To note, I was unable to evaluate the website itself due to “time out” errors, so my review is based only on the manuscript. Suggestions for enhancing the manuscript are as follows:\nIntroduction: In the sentence, “ERP also showed similar tic reduction..” clarify if the other treatment being referred to is Habit Reversal Therapy/Comprehensive Behavioral Intervention for Tics. It would be helpful to the reader to summarize the outcomes reported for ERP for tics in terms of response rate and degree of change in tic severity reported in prior research.\nA key feature of ERP is that it aims to train tic suppression in the presence of the premonitory urge so that one can habituate to those aversive sensations. This should be noted in the introduction, particularly since this program would train suppression in a manner that is de-linked from the urge.\n\nA brief summary of the literature demonstrating the robust, acute enhancement effect of contingent reward on tic suppression should be included in the introduction. This is a consistently replicated finding, which further supports the rationale for creating a more automated reward delivery program.\n\nThe introduction suggests an intent to utilize data input into the software for research purposes; however, at present the tictimer.com site does not appear to inform users of this intent. It is unclear to me if this is meant as a research only tool at this time, or if anyone can access the site. Clarify whether other possibility identifying information can be collected (e.g., IP address).\n\nIt seems that each “level” corresponds to a particular reinforcement schedule. Specific details about these schedules and how they change as one progresses through the program would be helpful. I assuming the program is intended to shape longer tic suppression durations but am not clear if the program aims to do this with an increasing or decreasing reward density. It also sounds like the reward schedule can be manipulated in a research context, so clarification of what the default settings vs. customizable settings would be useful.\n\nA section discussing limitations of the current iteration of the software would be useful, such as reliance on live observation by the trainer (as tic frequency can be impacted by observation, and real-time judgements about tic occurrences can be difficult). The program only codes tic occurrence dichotomously, which precludes analyses related to tic type (e.g., motor vs. vocal, simple vs. complex).\n\nIt would be beneficial to clarify whether this software has been tested in those with tics yet.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3868",
"date": "06 Aug 2018",
"name": "Kevin J Black",
"role": "Author Response F1000Research Advisory Board Member",
"response": "This paper describes TicTrainer, a newly created software program designed to track tic occurrences and deliver reward for successful suppression during live observation. Tic suppression contingent reward has reliably been shown to acutely enhance tic suppression in laboratory settings, but little work has been done to date to translate this finding into the clinical context. TicTrainer is an important step in this direction, and appears to be a promising tool for both research and clinical applications. To note, I was unable to evaluate the website itself due to “time out” errors, so my review is based only on the manuscript.The web site is working now. Steps for registering and starting the first session were originally quite complicated. We now provide careful, step-by-step instructions for both “users” and “trainers,” and we believe this also increases the likelihood of successful use. Suggestions for enhancing the manuscript are as follows:Introduction: In the sentence, “ERP also showed similar tic reduction..” clarify if the other treatment being referred to is Habit Reversal Therapy/Comprehensive Behavioral Intervention for Tics. It would be helpful to the reader to summarize the outcomes reported for ERP for tics in terms of response rate and degree of change in tic severity reported in prior research.Yes, the comparison was HRT/CBIT, which had more robust published evidence for efficacy. Since this is a software paper, the introduction was light on detail about ERP. Still, we now provide somewhat more information, as requested.A key feature of ERP is that it aims to train tic suppression in the presence of the premonitory urge so that one can habituate to those aversive sensations. This should be noted in the introduction, particularly since this program would train suppression in a manner that is de-linked from the urge.We now discuss this potentially important difference in the next-to-last paragraph of the introduction, and briefly mention it again at the end of the conclusion.A brief summary of the literature demonstrating the robust, acute enhancement effect of contingent reward on tic suppression should be included in the introduction. This is a consistently replicated finding, which further supports the rationale for creating a more automated reward delivery program.Done (next-to-last paragraph of introduction). The introduction suggests an intent to utilize data input into the software for research purposes; however, at present the tictimer.com site does not appear to inform users of this intent. It is unclear to me if this is meant as a research only tool at this time, or if anyone can access the site. Clarify whether other possibility identifying information can be collected (e.g., IP address).Anyone can access the site and use the software as is (or create their own instantiation on their own server to change its details and control its use). The program is not intended to track IP address, but servers can do so. It seems that each “level” corresponds to a particular reinforcement schedule. Specific details about these schedules and how they change as one progresses through the program would be helpful. I assuming the program is intended to shape longer tic suppression durations but am not clear if the program aims to do this with an increasing or decreasing reward density. It also sounds like the reward schedule can be manipulated in a research context, so clarification of what the default settings vs. customizable settings would be useful. Higher levels give higher reward density but reward frequency is slower. We provide the details in the revised ms. (“For each session, the reward point rate starts at zero, and the rate resets to zero after each tic. Each time the user refrains from ticcing for a number of seconds equal to his/her level, their point total increases by the current rate, and the rate then increases by the square of the current level (rate is capped at 10 × levels 2).”In other words, at level 1, the first second without a tic is rewarded with 0 points, the 2nd second without a tic is rewarded by 1 point, and so on until the eleventh (and beyond) consecutive second without a tic is rewarded with 10 points. Once a tic happens, the rate resets to 0 point per second. At level 2, two seconds must pass before the first reward, and the rate then increases to 4 points. After two additional seconds the rate increases to 8 points, and so on to a maximum of 40 points per 2 seconds. As for the question about research use, the current version of the software allows administrators to change a participant from the contingent reward schedule above to a noncontingent reward strategy in which rewards are given regardless of tics at a fixed initial rate that then increases with each subsequent level. A section discussing limitations of the current iteration of the software would be useful, such as reliance on live observation by the trainer (as tic frequency can be impacted by observation, and real-time judgements about tic occurrences can be difficult). The program only codes tic occurrence dichotomously, which precludes analyses related to tic type (e.g., motor vs. vocal, simple vs. complex).Done (last paragraph). It would be beneficial to clarify whether this software has been tested in those with tics yet.We’re just offering the software. We don’t yet know whether it will work or how well it is tolerated (though of course we want to find out). Added to last paragraph."
},
{
"c_id": "3880",
"date": "07 Aug 2018",
"name": "Kevin J Black",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Oops, that sentence should have read:\"Higher levels give higher reward magnitude, but reward frequency is slower.\""
}
]
},
{
"id": "30850",
"date": "05 Mar 2018",
"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\nThe concept of the TicTrainer is highly valuable. Unfortunately I was unable to review the current version of the TicTrainer website directly due to issues with registration (it won’t accept ID – if crucial for this peer review, please provide some guidance in a Supplementary file or in the reply). In general, the authors should comment whether there is a plan in place to overcome the inability of Microsoft browsers to support the program, as this may be a major limitation for diffusion.\n\nLike the other reviewer, I also think that the Introduction is too simplistic in comparing efficacy between ERP and CBIT, whereas in fact the quality of the evidence supporting these two behavioral treatments is different, and has been systematically reviewed and even meta-analysed. A short paragraph summarizing the evidence supporting the efficacy of ERP, also with respect to CBIT, would be useful.\n\nMoreover, the Introduction should help a reader who is not fully acquainted with ERP understanding the technique of this behavioral approach, particularly the way in which trainer and patient interact during treatment sessions, and how this type of interaction is suitable to be operationalized using a software tool. All this is a bit given for well established in the paper, whereas in fact the literature on alternative modes of administering ERP is scant.\n\nPerhaps I am missing some basic characteristic of the TicTrainer software, but from this manuscript it is not clear whether the “user” window and the “trainer” must appear on separate devices, or whether they could be opened on the same computer simultaneously. Can the authors comment on whether this is a specific choice of the authors of the software (and why), or a technical limitation that aims to be overcome in software upgrades?\n\nFinally, I think that the lack of premonitory urge measures is an important limitation. On an individual patient basis, the lack of efficacy of contingent reward application to ERP might depend on the intensity of the premonitory urges and by an insufficient habituation to the urge. If the urges are not measured somehow, it would be difficult to monitor treatment response. A recent paper by Brandt et al., Cortex 2006 proposes a real time, computer-based urge monitoring technique; it would be interesting to integrate/implement something similar in this software.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3869",
"date": "06 Aug 2018",
"name": "Kevin J Black",
"role": "Author Response F1000Research Advisory Board Member",
"response": "The concept of the TicTrainer is highly valuable. Unfortunately I was unable to review the current version of the TicTrainer website directly due to issues with registration (it won’t accept ID – if crucial for this peer review, please provide some guidance in a Supplementary file or in the reply).The web site is working now. Steps for registering and starting the first session were originally quite complicated. We now provide careful, step-by-step instructions for both “users” and “trainers,” and we believe this also increases the likelihood of successful use. In general, the authors should comment whether there is a plan in place to overcome the inability of Microsoft browsers to support the program, as this may be a major limitation for diffusion. This is a quite reasonable concern, but we have not yet identified the specific problem. However, since other free browsers for MS Windows, Apple OS, Linux, Android and Chrome OS are in very common use, we suspect this will not constitute a real barrier to its use.1. Like the other reviewer, I also think that the Introduction is too simplistic in comparing efficacy between ERP and CBIT, whereas in fact the quality of the evidence supporting these two behavioral treatments is different, and has been systematically reviewed and even meta-analysed. A short paragraph summarizing the evidence supporting the efficacy of ERP, also with respect to CBIT, would be useful. I agree that the evidence for efficacy of ERP is less solid than for CBIT, but ERP advocates reasonably point to the Verdellen, 2004, trial comparing it to HRT, and their own experience. In any case, a long discussion seems out of place in a short paper introducing software. We added a citation in the introduction to the ESSTS review of behavior therapies for TS. 2. Moreover, the Introduction should help a reader who is not fully acquainted with ERP understanding the technique of this behavioral approach, particularly the way in which trainer and patient interact during treatment sessions, and how this type of interaction is suitable to be operationalized using a software tool. All this is a bit given for well established in the paper, whereas in fact the literature on alternative modes of administering ERP is scant.The introduction now points out that exposure by focusing on premonitory urges is an important part of traditional ERP and we discuss in introduction and conclusion how this approach differs. 3. Perhaps I am missing some basic characteristic of the TicTrainer software, but from this manuscript it is not clear whether the “user” window and the “trainer” must appear on separate devices, or whether they could be opened on the same computer simultaneously. Can the authors comment on whether this is a specific choice of the authors of the software (and why), or a technical limitation that aims to be overcome in software upgrades?The original idea was for Johnny with tics to be seated at a computer or tablet and Mom to be using her smart phone, but it can be done on a single screen. We now specify this in “Use cases”: “Figure 1 shows a “user” window (for the person with tics) and a “trainer” window as they might appear during a session. In typical use, the two windows would appear on two separate devices (but they can be opened on the same device, as shown here).” 4. Finally, I think that the lack of premonitory urge measures is an important limitation. On an individual patient basis, the lack of efficacy of contingent reward application to ERP might depend on the intensity of the premonitory urges and by an insufficient habituation to the urge. If the urges are not measured somehow, it would be difficult to monitor treatment response. A recent paper by Brandt et al., Cortex 2006 proposes a real time, computer-based urge monitoring technique; it would be interesting to integrate/implement something similar in this software.We also are very interested in monitoring premonitory urge severity, though that likely would affect outcome. The conclusions now include the following: “Features that could yet be implemented include adding self-report (and/or trainer-report) measures of … premonitory sensations/urge intensity before, during or after sessions (Brandt et al., 2016; Himle et al., 2007; Specht et al., 2014; Verdellen et al., 2008).”"
}
]
},
{
"id": "30608",
"date": "15 Mar 2018",
"name": "Cara W.J Verdellen",
"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\nTicTrainer is developed to support exposure and response prevention (ERP) in the behavioral treatment of tics. ERP is considered a firstline intervention for tic disorders according to European Guidelines (Roessner et al; Verdellen et al, ESSTS, 2011). Patients are encouraged to suppress tics for prolonged periods of time allowing them to get used to unpleasant premonitory urges that often accompany tics. TicTrainer is a very valuable, nice and simple tool that supports the patient in practicing ERP at home, also in the absence of a therapist. It is also based on the finding that tic suppression is enhanced by reinforcement of tic-free periods. TicTrainer stores anonymous information to assess use patterns and information about efficacy and safety. This makes TicTrainer a promising software tool for both clinical and research ends.\n\nUnfortunately, I was unable to evaluate the website itself because after creating an account I did not manage to make it work (‘Input Error: improper ID’). Therefore, my review is based only on the manuscript. I recommend the authors to pay attention to these ‘’technical problems’’, and also to get it supported by Microsoft browsers, as this may be a major limitation for usage.\n\nTicTrainer can be used to enhance tic suppression ability (response prevention of tics), but in its present form, it does not include the element of exposure to premonitory urges. Since exposure is an important part of ERP (learning to tolerate the urge, not give in to it by performing tics), it is highly recommended to integrate this element, for example by adding a measure of sensations/urges.\n\nMore literature demonstrating the effect of direct reinforcement on tic suppression should be included in the introduction as it supports the rational for this tool in enhancing tic suppression ability.\n\nIt is described in this manuscript that ‘’preferably’’ two devices are needed; one for the patient and one for the trainer, who can be a parent/partner. It doesn’t say if it also works on one device (do you see 2 windows?). A disadvantage of two devices may be that a patient always needs another person to practice with, which is not always possible. It is suggested to consider integrating both tic detection and suppression/reinforcement in one tool/device (like the app BT-Coach, see bt-tics.com/bt-coach).\nFor what age is this tool applicable?\n\nIs it right that a patient doesn’t see his tic suppression (record) times on the window, only ‘’levels’’ and ‘’earned coins’’? It may be encouraging and reinforcing to also receive feedback on the exact tic suppression (record) times (as described in the protocol of ERP).\n\nSince the tool is based on receiving rewards for tic suppression, perhaps more details can be given about what kind of rewards can be used; an example of a rewarding system could be useful. Perhaps ‘’mini’’ games can be integrated in the tool that can be done directly after practicing, depending on the number of coins that are earned/levels achieved?\n\nOverall I think this is a very promising tic suppression tool!\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3870",
"date": "06 Aug 2018",
"name": "Kevin J Black",
"role": "Author Response F1000Research Advisory Board Member",
"response": "TicTrainer is developed to support exposure and response prevention (ERP) in the behavioral treatment of tics. ERP is considered a firstline intervention for tic disorders according to European Guidelines (Roessner et al; Verdellen et al, ESSTS, 2011). Patients are encouraged to suppress tics for prolonged periods of time allowing them to get used to unpleasant premonitory urges that often accompany tics. TicTrainer is a very valuable, nice and simple tool that supports the patient in practicing ERP at home, also in the absence of a therapist. It is also based on the finding that tic suppression is enhanced by reinforcement of tic-free periods. TicTrainer stores anonymous information to assess use patterns and information about efficacy and safety. This makes TicTrainer a promising software tool for both clinical and research ends. Unfortunately, I was unable to evaluate the website itself because after creating an account I did not manage to make it work (‘Input Error: improper ID’). Therefore, my review is based only on the manuscript. I recommend the authors to pay attention to these ‘’technical problems’’, and also to get it supported by Microsoft browsers, as this may be a major limitation for usage. The web site is working now. Steps for registering and starting the first session were originally quite complicated. We now provide careful, step-by-step instructions for both “users” and “trainers,” and we believe this also increases the likelihood of successful use. TicTrainer can be used to enhance tic suppression ability (response prevention of tics), but in its present form, it does not include the element of exposure to premonitory urges. Since exposure is an important part of ERP (learning to tolerate the urge, not give in to it by performing tics), it is highly recommended to integrate this element, for example by adding a measure of sensations/urges.In the edited introduction and conclusion, we now discuss this as a potential limitation, as a possible future improvement, and as an opportunity to test whether explicit exposure (focus on the urge) is essential to efficacy in ERP of tics (or whether by contrast tic suppression practice itself, if scheduled and repeated, can lead to benefit). This last possibility seems the more plausible given that habituation’s relevance to ERP efficacy is now in doubt. The relevant portions of the revised text include: “There is substantial evidence that contingent reward enhances tic suppression without producing a rebound effect, at least for periods of time up to 30-40 minutes (Himle et al., 2007; Himle et al., 2008; Woods et al., 2008; Specht et al., 2013; Brabson et al., 2016; Conelea et al., 2018). Within a single session, contingently reinforced tic suppression produced decreases in tic frequency whether or not the participants were directed to attend to premonitory urges (Specht et al., 2013). However, whether repeated practice of tic suppression leads to sustained, clinically relevant improvement is not known, because unlike typical ERP, tic suppression alone does not explicitly encourage attention to the premonitory urge (the “exposure” part of ERP). … The software itself does not encourage attention to premonitory urges, a focus of traditional ERP. One can provide instructions to do so outside of the software, or use the software without explicit exposure instructions in order to test whether ERP’s efficacy depends on this component.” More literature demonstrating the effect of direct reinforcement on tic suppression should be included in the introduction as it supports the rational for this tool in enhancing tic suppression ability.Done: “There is substantial evidence that contingent reward enhances tic suppression without producing a rebound effect, at least for periods of time up to 30-40 minutes (Himle et al., 2007; Himle et al., 2008; Woods et al., 2008; Specht et al., 2013; Brabson et al., 2016; Conelea et al., 2018). Within a single session, contingently reinforced tic suppression produced decreases in tic frequency whether or not the participants were directed to attend to premonitory urges (Specht et al., 2013).” It is described in this manuscript that ‘’preferably’’ two devices are needed; one for the patient and one for the trainer, who can be a parent/partner. It doesn’t say if it also works on one device (do you see 2 windows?). A disadvantage of two devices may be that a patient always needs another person to practice with, which is not always possible. It is suggested to consider integrating both tic detection and suppression/reinforcement in one tool/device (like the app BT-Coach, see bt-tics.com/bt-coach).We now address this question in “Use cases”: “Figure 1 shows a “user” window (for the person with tics) and a “trainer” window as they might appear during a session. In typical use, the two windows would appear on two separate devices (but they can be opened on the same device, as shown here).”We feel that BT-Coach is a wonderful (and more mature) application. However, the TicTrainer approach grew out of our recent study on children with recent onset of tics, who often are less likely to notice their own tics. We feel that tic monitoring by a third party may be a feature rather than a drawback in certain use cases.For what age is this tool applicable?We don’t know for certain but expect that anyone who could sit through a session of even 10 minutes or so may be able to use it. Is it right that a patient doesn’t see his tic suppression (record) times on the window, only ‘’levels’’ and ‘’earned coins’’? It may be encouraging and reinforcing to also receive feedback on the exact tic suppression (record) times (as described in the protocol of ERP).Correct, though the increasing reward rate with increasing duration of tic suppression (through 10 seconds initially, or 10 × level seconds later) is a weak proxy. We agree that the record times would be a potentially useful addition. We have added to conclusion the following: “Showing the user’s personal record times (maximum achieved tic-free duration) would be another feature of interest.” Since the tool is based on receiving rewards for tic suppression, perhaps more details can be given about what kind of rewards can be used; an example of a rewarding system could be useful. Perhaps ‘’mini’’ games can be integrated in the tool that can be done directly after practicing, depending on the number of coins that are earned/levels achieved?That is a great idea that at this time is beyond our skill. One benefit of our releasing this code is that others with time and skill beyond ours can add to it. Overall I think this is a very promising tic suppression tool!Thank you!"
}
]
}
] | 1
|
https://f1000research.com/articles/6-2150
|
https://f1000research.com/articles/7-725/v1
|
11 Jun 18
|
{
"type": "Software Tool Article",
"title": "Identifier Mapping in Cytoscape: idmapper",
"authors": [
"Adam Treister",
"Alexander R. Pico",
"Adam Treister"
],
"abstract": "Identifier Mapping, the association of terms across disparate taxonomies and databases, is a common hurdle in bioinformatics workflows. The idmapper app for Cytoscape simplifies identifier mapping for genes and proteins in the context of common biological networks. This app provides a unified interface to different identifier resources accessible through a right-click on the table's column header. It also provides an OSGi programming interface via Cytoscape Commands and CyREST that can be utilized for identifier mapping in scripts and other Cytoscape apps, and supports integrated Swagger documentation.",
"keywords": [
"Cytoscape",
"ID Mapping",
"Identifiers",
"BridgeDb"
],
"content": "Introduction\n\nCytoscape is an integrated network visualization tool and analysis platform1,2. Within its common workflows, identifier mapping remains a challenge when working with biological data from different sources. This problem has been addressed by the BridgeDB project3, which created clients and services to translate between various identifiers. The original BridgeDb app4 for Cytoscape was written to provide an exhaustive set of functions to match the full capabilities of BridgeDb. Though this provided the needed functionality, its basic usage was unnecessarily complex. The idmapper app is a useful alternative, providing a subset of critical features with a simplified interface bundled into Cytoscape. Now, without any installation or configuration, Cytoscape users can right-click on a table header to map that column’s data to a different namespace (Figure 1). Although, the breadth of coverage is smaller than the full-featured BridgeDb app, it still covers over a dozen identifier data sources, including Ensembl, EntrezGene, HGNC, KEGG, Uniprot and various specied-specific sources. Because idmapper supports Cytoscape’s new CyREST interface, identifier mapping can be included in scripted workflows, and driven from R or python programs.\n\nFour options are presented to the user when accessing idmapper from within the Cytoscape GUI, each with common default or inferred values to reduce the number of steps required of the user.\n\n\nImplementation\n\nFrom within Cytoscape, a user initiates an ID mapping operation by right-clicking on the header of a column containing identifiers in the Table Panel. In the most common cases the type of identifier can be guessed by idmapper based on the its format. Table 1 shows the supported data sources and example identifier formats. The app looks at the first ten entries and choose the source from the option that matches corresponding regular expressions. This number of identifiers iteratively sampled is set by a static variable called N_Iterations. The algorithm for inferring the data source is implemented in IdGuess.java.\n\nCurrently supported identifier databases, their BridgeDb system codes, their species specificity and an example identifier.\n\nThere are two different tasks supported by the idmapper app. ColumnMappingTask is activated by the right-click mouse event on a table header. It infers the current table and column from the information that comes from the mouse event. In order to support automation, we added MapColumnCommandTask as an analog that is exposed specifically for Commands and CyREST access. These tasks eventually result in the same algorithms being invoked.\n\n\nUse cases\n\nThe idmapper app provides the same basic functionality of the BridgeDb app with less fuss. Users do not have to install it, launch it, make configuration decisions or think about which database they are accessing. The app comes bundled with every Cytoscape release. As such it usage in Cytoscape via the interactive GUI (graphical user interface) is documented in the Cytoscape manual, http://manual.cytoscape.org/en/stable/Node_and_Edge_Column_Data.html#mapping-identifiers.\n\nTo map an identifier from one source to another, right click on the column header of your identifier. Select the option to Map Column to bring up the idmapper dialog (Figure 1).\n\nThe idmapper dialog presents a few choices the user can override before performing ID mapping. The default Species is determined by the previous selection made per network, providing a \"smart and sticky\" behavior. The available choices for the identifier data sources are determined by the species. The Map from data source is automatically selected based on an inspection of the first ten identifiers found in the column clicked on by the user. This can easily be overridden by the pull down menu. The To data source must be selected by the user; Ensembl is presented by default. Finally, the Force single checkbox offers to simplify the results of ID mapping by ignoring one-to-many cases and only keeping the first result. If the option is off, a list of results will appear in the column. This can easily be overridden by clicking the toggled checkbox.\n\nThe command interface does not use the same tasks as the GUI. In the GUI use case, the app knows the current context of where the command was activated, i.e., the network, table and column. This information must explicitly be provided as paramaters to the command interface to perform the same operation. Thus, in addition to species, mapFrom, mapTo and forceSingle, the command line operation of idmapper also requires networkName, table and columnName (see next section for more details).\n\nIn the scripting environment, idmapper provides all of its functionality in a single call (Figure 2). This means that identifier mapping can be incorporated into Cytoscape automation workflows with a single additional command.\n\nThe functionality of idmapper is contained in this singular function: map column.\n\nThe map column function takes the following parameters:\n\ncolumnName (string): Specifies the column name where the source identifiers are located\n\nforceSingle (string, optional): When multiple identifiers can be mapped from a single term, this forces a singular result\n\nmapFrom (string): Specifies the data source describing the existing identifiers\n\nmapTo (string): Specifies the data source identifiers to be returned as a result in a new column\n\nnetworkName (string, optional): Which network is used in the mapping.\n\nspecies (string): The common or latin name of the species to which the identifiers apply, e.g., Human, Homo sapiens, Mouse, Mus musculus, Rat, Rattus norvegicus, Frog, Xenopus tropicalis, Zebra fish, Danio rerio, Fruit fly, Drosophila melanogaster, Mosquito, Anopheles gambiae, Arabidopsis, Arabidopsis thaliana, Yeast, Saccharomyces cerevisiae, E. coli, Escherichia coli, Tuberculosis, Mycobacterium tuberculosis, Worm, Caenorhabditis elegans\n\ntable (string, optional): Which table is used as the source of the identifiers, e.g., \"node\" for the default node table\n\nWith Cytoscape running, the map column function can be called from any scripting environment or programming language that supports REST calls. In the case of R and Python scripts, there are dedicated packages to make this even easier. The RCy3 package wraps this command in an R function called mapTableColumn to conform to other table functions (https://www.bioconductor.org/packages/release/bioc/html/RCy3.html). The py2cytoscape library similarly provides this command as a python function, cyclient.idmapper.map_column (https://github.com/cytoscape/py2cytoscape).\n\nA sample script demonstrates how to map identifiers via RCy3, covering the most common use cases (https://github.com/cytoscape/RCy3/blob/master/vignettes/Identifier-mapping.Rmd).\n\nThe Yeast Perturbation sample network provided with Cytoscape can be loaded from the Starter Panel and provides gene identifiers of the form “YDL194W”. These are actually Ensembl-supported identifiers for Yeast, distinct from the typical “ENSXXXG00000123456” form as presented in Table 1. This presents a special case that users will need to be aware of when selecting species and source database or mapFrom in the GUI. In terms of automation, you could generate a new column of Entrez Gene IDs in this network with these calls:\n\n\n\nWhen working with protein interaction networks, for example those from the STRING database (see https://apps.cytoscape.org/apps/stringapp), you may want to translate to gene identifiers. The idmapper app supports this case as well, but one should be aware of the assumptions involved when making this translation. Since most genes encode for many proteins, you may have many-to-one mappings in your results. For all human networks imported from STRING using the StringApp5, the following commands will perform an ID mapping from Uniprot-TrEMBL (proteins) to Ensembl (genes):\n\n\n\n\nLimitations\n\nThe idmapper app provides easy access to a critical subset of ID mapping functionality originally covered by the BridgeDb app. When users run into the limitations of idmapper, they still have the option of installing and using the full-featured BridgeDb app from https://apps.cytoscape.org/apps/bridgedb. Examples of limitations include support for additional species or data sources. The BridgeDb app includes more of both as well as means to access custom data sources.\n\n\nData and software availability\n\n1. Software available from the Cytoscape App Store: https://apps.cytoscape.org/apps/idmapper\n\n2. Latest source code: https://github.com/cytoscape/idmapper\n\n3. Archived source code as at the time of publication: https://doi.org/10.5281/zenodo.12468146\n\n4. License: Apache License, Version 2.0",
"appendix": "Author contributions\n\n\n\nAT and ARP participated in the design of the described software. AT implemented the software. AT and ARP contributed to the writing of this article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe would like to acknowledge funding from National Institute of General Medical Sciences [P41GM103504 (ARP, AT)].\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\nJianjiong Gao and Chao Zhang for their work on the original BridgeDb app (https://f1000research.com/articles/3-148/v14). Nuno Nunes for his work on the BridgeDb web service.\n\n\nReferences\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\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\nvan Iersel MP, Pico AR, Kelder T, et al.: The BridgeDb framework: standardized access to gene, protein and metabolite identifier mapping services. BMC Bioinformatics. 2010; 11: 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGao J, Zhang C, van Iersel M, et al.: BridgeDb app: unifying identifier mapping services for Cytoscape [version 1; referees: 2 approved]. F1000Res. 2014; 3: 148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSzklarczyk D, Morris JH, Cook H, et al.: The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45(D1): D362–D368. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTreister A, Ono K, Zmasek C, et al.: cytoscape/idmapper: 3.6.3 (Version 3.6.3). Zenodo. 2018. Data Source"
}
|
[
{
"id": "34908",
"date": "12 Jun 2018",
"name": "Ruth Isserlin",
"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 entitled \"Identifier Mapping in Cytoscape: idmapper\" by Adam Treister and Alexander Pico presents a new implementation of id mapping available directly through Cytoscape with no additional configuration.\n\nThere is a lot of discussion throughout the paper about BridgeDB. It is not clear what the relationship between idmapper and BridgeDB is. It is understood that idmapper is a simplification and alternative of the aforementioned tool but does idmapper rely on BridgeDB? Do they share a codebase?\nOne thing that might be a good addition to the implementation section is a short discussion on the backend. Where are the mappings coming from? Is the app dependent on any online resources or are the mappings static and stored within the Cytoscape instance? This is related to the previous BridgeDB question as given that there is no discussion about the backend I thought maybe the info was coming from BridgeDB and that was the reason it was included.\nIn the implementation section, Table 1,\nit might be useful to have a column specifying the data source names that the app recognizes. (for example in the table one of the data sources listed is \"Uniprot TrEMBL\" and later in Case 2 in the command the map.from(RCy3) and source_selection(py2cytoscpae) the db is specified as \"Uniprot-TrEMBL\" (with a dash)). Given that this is the implementation section it might be good to list the regular expressions that each identifier recognizes (provided that they aren't too messy) and any exceptions they look for. Later on in the case 1 you mention an exception to the basic regular expression behaviour for yeast. Are there more exceptions that the app handles? It is also unclear why the \"Code\" column is required. Might be nice to separate the data sources that can handle any species and those that are species specific.\nIn the use cases section \"The default Species is determined by the previous selection made per network, providing a \"smart and sticky\" behavior. \" It is unclear but the previous selection was?\nIn the use cases section in the specific cases two example use cases are presented, species specific and protein to gene conversions. It would be helpful to list these and other common use cases at the start of the use case section as well. One of the most common use cases being going from non-descriptive identifiers (like entrez gene ids, and ensembl ids) to something more understandable such as species specific IDs (HGNC or MGD, Is it possible to map to proper gene symbols?)\nIn the use case section it is stated that if the ID maps to multiple identifiers there is an option, \"Force single\", that when selected the app selects the first result. How are the returned IDs sorted? Is the first match the \"best\" match, alphabetical, random?\nIn the Cytoscape automation section in the parameters section for the species option all the available mappings are listed but for the mapFrom and mapTo no options are listed. (if the recognized data source name is add to Table 1 you can just reference the table here or if the first column of Table 1 are the recognized names it would be good to reference it here). Also, the parameters listed for Cytoscape automation section are very different from the parameters used in the use cases which can be very confusing. Maybe adding an example using the RCy3 commandsGet option under RCy3 and py2cytoscape examples just showing how the user can use all the parameters as specified using the command directly.\n\nMinor comments/questions:\nIn the introduction \"Uniprot and various specied-specific sources\" should be \"Uniprot and various species-specific sources\" In the implementation section \"The app looks at the first ten entries and choose the source\" would be better as \"The app looks at the first ten entries and chooses the source\" In the implementation section - \"This number of identifiers iteratively sampled is set by a static variable called N_Iterations. The algorithm for inferring the data source is implemented in IdGuess.java.\" - This is a little confusing why this is needed. Is this a parameter the user can control or tweak? In Use cases section - \"As such it usage in Cytoscape via\" should be \"As such its usage in Cytoscape via\" In use case 2 \"you may want to translate to gene identifiers\" might be better as \"you may want to translate protein identifiers (for example: Uniprot-TrEMBL) to gene identifiers\" Can idmapper convert a list column? (in the example use case where the network is an enrichment map and each node contains a set of genes as opposed to each node being a gene) What is the resulting column name? What if a column with that name already exists?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3766",
"date": "26 Jun 2018",
"name": "Alexander Pico",
"role": "Author Response",
"response": "Ruth, thank you so much for your thorough review. These clarifications, fixes and additions have greatly improved the article. Version 2 should be released soon, addressing all the issues you raised. We decided not to include the regular expressions in Table 1, however. They are messy and are what you'd expect from the example identifiers provided, which we feel do a better job of communicating what to expect from each data source. We hope you'll have a chance to look over version 2 and find that it meets your expectations."
}
]
},
{
"id": "34906",
"date": "03 Jul 2018",
"name": "Augustin Luna",
"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 Cytoscape app provides functionality that is widely useful for Cytoscape users for converting network identifiers to different databases. Some points not completely clear:\nThe list of regular expressions used for inference, do they come from identifiers.org? What happens if the inference fails? Does the app try to pick the closest matching regular expression? For use of RCy3, is RCy3 a generic package to interact with any REST function in any Cytoscape package? Or did the developers of RCy3 specially include the mapTableColumn function to access the idmapper app?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3795",
"date": "04 Jul 2018",
"name": "Alexander Pico",
"role": "Author Response",
"response": "Augustin, thanks for the review.1. The regular expressions come from BridgeDb (which in turn gets them from identifiers.org). If there isn't a match to the regular expressions (or if more than one system is matched), then it just picks the first option in the list. A few of the system types aren't well specified. It's a simple matter to override this in the UI. We will add a sentence or two in the next version of the paper in response to all reviewers.2. Right. RCy3 supports both a generic function call and a specific mapTableColumn function call. Since this is a \"core\" app, the RCy3 package supports custom convenience functions to make this operation easier to use and better documented. We will add a more detailed description and contrasting example in the next version of the paper."
}
]
},
{
"id": "34907",
"date": "19 Jul 2018",
"name": "Nadezhda Doncheva",
"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 idmapper, a simple yet very useful app for converting node identifiers from one data source to another. It is provided as part of the widely used network analysis and visualization software Cytoscape and the mapping functionality can be used by users in several different ways, including the GUI, Cytoscape commands, and from R or Python scripts vie the Cytoscape REST interface. It supports common data sources such as UniProt and Ensembl. Both the app and the manuscript are in good quality and there are a few minor aspects to be addressed.\nIn the Implementation section, it is important to clarify which extent the idmapper app relies on the BridgeDB mappings/app and also say a few words about how these mappings are done in BridgeDB and what the versions of the used data sources are. Maybe it can be a subsection called Dependencies or Backend? In the use Cases section, it would be useful to provide the Cytoscape command call in addition to the R and python function calls. Figure 2 is not very informative as it is now. Maybe it would be better if it includes a screenshot of the “Example Value” code? In addition, the term “singular function” is used in mathematics and might be confusing for some readers.\nThe app works nicely both through the Cytoscape GUI and the command/swagger interface. However, it would be great if the documentation can be improved a little bit so it is more consistent and informative. In particular:\nCould a list of all possible species and all possible data sources names be provided both in the manuscript and in the documentation of the command in Cytoscape/Swagger? It could be included as part of the description of the map column function in the Cytoscape automation section. The species are most likely the ones listed in the species parameter description but it is clearly stated anywhere. Is there a comma missing in the sentence: ”The combined common or latin name”? It seems that one can use all three of those: the common name, the latin name or the combined one (although there is a warning if anything but the combined one is used). Are the data sources exactly and only the ones in Table 1? There is an example with “Uniprot” (GUI, swagger documentation and example), one with “Uniprot-TrEMBL” (Case 2 in the paper) and in the table it is written “Uniprot TrEMBL”. Are all of those the same or not and if not what are the differences? For the table, it says “node” in the command documentation and “default node” in the swagger example. Both of them work, but it would be good to have a list of possible values as it is the case for the species parameter. Make clear (maybe in table 1) that Ensembl refers to Ensmbl Gene identifiers and not Proteins. Would it be possible that the parameters of the python function have the same names as the parameters in all other functions? Only recommended if it does not break already existing scripts.\n\nThere were two minor issues while testing the app:\nWhen running the following command from within Cytoscape or the swagger documentation on the galFiltered network (or on a STRING network), a warning and an error message come up in the Task Manager. It seems to work nonetheless, so could you check what is going on there? idmapper map column columnName=\"name\" mapFrom=\"Ensembl\" mapTo=\"Entrez\" species=\"Yeast\" warning: value not contained in list of possible values possible items = [] and error: networkTable not found. The data source inferring works well for the identifiers, but in the case that the species is wrong (e.g. switching from a yeast to a human network), changing the species resets the Map from column to the first entry in the list.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3849",
"date": "06 Aug 2018",
"name": "Alexander Pico",
"role": "Author Response",
"response": "Nadya, thank you for the review. We have prepared a version 2 of the paper that addresses your main comments and corrections. A couple additional notes that are not reflected in the updated text: The R and Python library projects are independently developed, so coordinating on parameter names is impossible and in many cases not desired since consistency within each is much more important than across them, i.e., most users will pick one or the other and want maximum consistency with the conventions of that language, etc. Figure 2 is a requirement for all Cytoscape Automation papers. Even for cases like ours where there is just one operation. It's our fault for making such a simple app! Thanks for the bug reports. We will try to reproduce and fix these for future releases of the app."
}
]
}
] | 1
|
https://f1000research.com/articles/7-725
|
https://f1000research.com/articles/7-252/v1
|
01 Mar 18
|
{
"type": "Opinion Article",
"title": "The inflammation paradox: Why are Tsimane protected against Western diseases while Westerners are not?",
"authors": [
"Jens Freese",
"Rainer Johannes Klement",
"Helmut Lötzerich",
"Rainer Johannes Klement",
"Helmut Lötzerich"
],
"abstract": "Recently, observational studies in the Amazon region showed that the indigenous Tsimane in Bolivia appear protected against non-communicable diseases (NCDs) such as obesity, type 2 diabetes, and cardiovascular diseases despite increased inflammatory markers. These findings stand in contrast to Western societies, where an increasing body of evidence demonstrates that low-grade-inflammation is the driver of NCDs. In 2013 and 2014, we carried out two field studies (Eifel studies) with Westerners who returned to a simulated Palaeolithic lifestyle in a National park for 4 days and detected elevated inflammation markers, analogous to the conditions of the Tsimane. We here propose three hypotheses for this inflammatory paradox.",
"keywords": [
"Tsimane",
"Low-Grade Inflammation",
"Eifel studies",
"Non-communicable diseases"
],
"content": "Abbreviations\n\nCRP, C-reactive protein; NCD, non-communicable diseases; NK cells, natural killer cells; LGI, low-grade inflammation; LPS, lipopolysaccharides; NFκB, nuclear-factor-kappa-B\n\n\nIntroduction\n\nRecently, observational studies in the Amazon region showed that the indigenous Tsimane in Bolivia appear protected against non-communicable diseases (NCDs) such as obesity, type 2 diabetes, and cardiovascular diseases, despite increased inflammatory markers1. These findings stand in contrast to Western societies, where an increasing body of evidence demonstrates that low-grade-inflammation (LGI) is the driver of NCDs2–4.\n\n\nReport\n\nCompared to US reference values, Tsimane exhibit markedly high levels of eosinophilic and neutrophilic granulocytes, B lymphocytes and natural killer cells. The leukocyte counts of Tsimans (8,600–12,000 cells/μL) are 1.5 times, and lymphocytes 1.2 to 1.6 times higher than in the US population (6,700–7,900 cells/μL)5,6. Eosinophilic granulocytes, primarily indicative of parasitic infections, are 7-fold elevated. Consequently, the immunoglobulin E values are also significantly higher (150–200-fold). Important biomarkers for inflammation, such as neutrophil granulocytes (1.2 to 1.6-fold), blood sedimentation (30 mm/h to 15–20 mm/h) and C-reactive protein (CRP) values (higher from infant to adolescence), are also upregulated6. Also striking is the relatively high basal metabolism of the Tsimane compared to the US population. Gurven et al.7 argue that the elevated metabolic rate occurs to cover the energetic costs of an activated immune system in the tropical wilderness. This assumption indicates that 70% of this population is permanently infected with parasites. Despite elevated inflammation, this might be the reason why Tsimane are protected from NCD, because intestinal worms not only absorb fat that would then no longer be available to the host, but also increase the amount of type 2 anti-inflammatory T helper cells7.\n\nIn 2013 and 2014, we carried out two field studies (Eifel studies) with Westerners who returned to a simulated Palaeolithic lifestyle in a National park for 4 days8,9. Contrary to our expectations, in both studies, CRP, the main liver-derived biomarker that displays nonspecific inflammation, had increased significantly. The essential components of these interventions consisted of (i) the conversion to a paleo diet; (ii) the high range of locomotion (15 km/day in the Eifel study 2013, 16.4 km/day in the Eifel study 2014); (iii) a fasting period from 12 to 14 hours per day in conjunction with a low meal frequency resulting in undercaloric energy intake (1567 kcal in the Eifel study 2013, 1747 kcal in the Eifel study 2014). All mentioned factors have been shown to have anti-inflammatory effects10–14.\n\n\nDiscussion\n\nSince it cannot be assumed that civilized humans display major parasitic infections like the Tsimane, we provide the following hypothetical explanations for the stimulation of the immune system in the Eifel studies, which are likely to influence one another:\n\n\n\n1. Phyto-antibiotics (phytoncides), which plants release into the atmosphere to protect themselves against bacteria and insects, could have stimulated the innate immune system15. As studies from Japan and Korea have shown, so-called \"forest bathing\" (a multi-day hike through a forest) promotes the formation of high levels of natural killer cells (NK cells). This effect persists for up to 30 days after the intervention16,17. In addition, forest bathing also increases the activity of the cytolytic proteins perforin, granzyme A and granulysin in NK cells. Walks in the city, on the other hand, do not change the NK cell population or its activity17. These effects could have contributed to the increase in CRP levels in the Eifel studies, as most of the time participants spent in a forest area.\n\n2. The radical change from a near-sterile to a natural environment may have prompted the innate immune system to anticipate and prophylactically protect the organism against pathogens such as bacteria, parasites, fungi, and other microorganisms. Danger signals, called exogenous pathogen associated molecular patterns and endogenous danger associated molecular patterns, activate the innate immune system via Toll-like receptors, which can trigger a rapid antibacterial inflammatory response. This mechanism of action may have led to the development of an acute inflammation and resolution of a (chronic) LGI. In contrast to LGI, substances such as lipoxins, resolvins and protectins are formed in acute inflammation in order to end the inflammatory process18,19. Since no follow-up measurements were made in the Eifel studies, this hypothesis is currently only speculative.\n\n3. Despite the fact that the participants in the Eifel studies were in good mental and physical health, the level of physical stress due to the high workload combined with calorie restriction conditions could have increased cell depletion. It is well known that the destruction of cell structures, e.g. in burns, viral and bacterial infections or after high volume or intensity training, increases the endogenous load of lipopolysaccharides (LPS)20. LPS activate the innate immune system via Toll-like receptors and stimulate the activation of nuclear-factor-kappa-B (NFκB) intracellularly. NFκB on the one hand increases pro-inflammatory cytokine secretion and on the other hand inhibits the insulin signalling cascade. As a result, macrophages, and other immune cells switch from oxidative phosphorylation to anaerobic glycolysis. This metabolic reprogramming of M2 to M1 macrophages occurs to trigger a rapid antibacterial inflammatory response to pathogens21. A by-product of cell destruction is uric acid, which stimulates the release of CRP in the liver as part of the acute immune response22,23. In turn, CRP stimulates the production of antibodies from B lymphocytes to kill pathogens24. Due to the high range of locomotion in both Eifel studies, uric acid might have played the leading role in stimulating the immune system. Since uric acid has not been measured, future studies should include this marker to provide a possible confirmation of this hypothesis.\n\n\nOutlook\n\nThe fact that a chronic inflammatory situation in Tsimans protects against NCD, while it increases the incidence in LGI in Westerners, should be investigated in further studies with the hypotheses proposed here.",
"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\nKaplan H, Thompson RC, Trumble BC, et al.: Coronary atherosclerosis in indigenous South American Tsimane: a cross-sectional cohort study. Lancet. 2017; 389(10080): 1730–1739. PubMed Abstract | Publisher Full Text\n\nEgger G: Obesity, chronic disease, and economic growth: a case for \"big picture\" prevention. Adv Prev Med. 2011; 2011: 149158. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRuiz-Núñez B, Pruimboom L, Dijck-Brouwer J: Lifestyle and nutritional imbalances associated with Western diseases: causes and consequences of chronic systemic low-grade inflammation in an evolutionary context. J Nutr Biochem. 2013; 24(7):1183–201. PubMed Abstract | Publisher Full Text\n\nRodríguez-Hernández H, Simental-Mendía LE, Rodríguez-Ramírez G, et al.: Obesity and inflammation: epidemiology, risk factors, and markers of inflammation. Int J Endocrinol. 2013; 2013: 678159. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNational Health and Nutrition Examination Survey (n.d.): National health and nutrition examination survey. Reference Source\n\nBlackwell AD, Trumble BC, Maldonado Suarez I, et al.: Immune function in Amazonian horticulturalists. Ann Hum Biol. 2016; 43(4): 382–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGurven MD, Trumble BC, Stieglitz J, et al.: High resting metabolic rate among Amazonian forager-horticulturalists experiencing high pathogen burden. Am J Phys Anthropol. 2016; 161(13): 414–425. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFreese J, Ruiz-Núñez B, Heynck R, et al.: To restore health, “do we have to go back to the future?” The impact of a 4-day paleolithic lifestyle change on human metabolism – a pilot study. J Evol Health. 2016a; 1(1): 12. Publisher Full Text\n\nFreese J, Pardi DJ, Ruiz-Nunez B, et al.: Back to the future. Metabolic effects of a 4-day outdoor trip under simulated paleolithic conditions - new insights from the Eifel study. J Evol Health. 2016b; 1(1): 16. Publisher Full Text\n\nInoue H, Sakai M, Kaida Y, et al.: Blood lactoferrin release induced by running exercise in normal volunteers: Antibacterial activity. Clin Chim Acta. 2004; 341(1–2): 165–72. PubMed Abstract | Publisher Full Text\n\nKonner M, Eaton SB: Paleolithic nutrition: Twenty-five years later. Nutr Clin Pract. 2010; 25(6): 594–602. PubMed Abstract | Publisher Full Text\n\nLemke D, Klement RJ, Paul S, et al.: The Paleolithic Diet and its Significance for the Prevention and Treatment of Chronic Diseases. Aktuel Ernahrungsmed. 2016; 41(06): 437–449.\n\nLongo VD, Mattson MP: Fasting: molecular mechanisms and clinical applications. Cell Metab. 2014; 19(2): 181–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeinhold M, Shimabukuro-Vornhagen A, Franke A, et al.: Physical exercise modulates the homeostasis of human regulatory T cells. J Allergy Clin Immunol. 2016; 137(5): 1607–1610.e8. PubMed Abstract | Publisher Full Text\n\nLi Q, Nakadai A, Matsushima H, et al.: Phytoncides (wood essential oils) induce human natural killer cell activity. Immunopharmacol Immunotoxicol. 2006; 28(2): 319–33. PubMed Abstract | Publisher Full Text\n\nPark BJ, Tsunetsugu Y, Kasetani T, et al.: The physiological effects of shinrin-yoku (taking in the forest atmosphere or forest bathing): Evidence from field experiments in 24 forests across japan. Environ Health Prev Med. 2010; 15(1): 18–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q: Effect of forest bathing trips on human immune function. Environ Health Prev Med. 2010; 15(1): 9–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSerhan CN, Savill J: Resolution of inflammation: the beginning programs the end. Nat Immunol. 2005; 6(12): 1191–7. PubMed Abstract | Publisher Full Text\n\nSerhan CN: Novel lipid mediators and resolution mechanisms in acute inflammation: To resolve or not? Am J Pathol. 2010; 177(4): 1576–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSun Y, Shang D: Inhibitory Effects of Antimicrobial Peptides on Lipopolysaccharide-Induced Inflammation. Mediators Inflamm. 2015; 2015: 167572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Neill LA, Kishton RJ, Rathmell J: A guide to immunometabolism for immunologists. Nat Rev Immunol. 2016; 16(9): 553–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanellis J, Watanabe S, Li JH, et al.: Uric acid stimulates monocyte chemoattractant protein-1 production in vascular smooth muscle cells via mitogen-activated protein kinase and cyclooxygenase-2. Hypertension. 2003; 41(6): 1287–93. PubMed Abstract | Publisher Full Text\n\nChen CJ, Kono H, Golenbock D, et al.: Identification of a key pathway required for the sterile inflammatory response triggered by dying cells. Nat Med. 2007; 13(7): 851–856. PubMed Abstract | Publisher Full Text\n\nBehrens MD, Wagner WM, Krco CJ, et al.: The endogenous danger signal, crystalline uric acid, signals for enhanced antibody immunity. Blood. 2008; 111(3): 1472–1479. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "31343",
"date": "12 Mar 2018",
"name": "Michael D. Gurven",
"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\nWe appreciate that Freese and colleagues [FKL] attempt to understand what they refer to as the “inflammation paradox”, i.e. that people may experience high chronic levels of circulating inflammation (as measured by biomarkers such as high sensitivity C-reactive protein (CRP)), yet not succumb to the diseases such as heart disease and diabetes usually attributed to such high exposure in urban settings. We first documented this pattern among Tsimane Amerindians in Bolivia in 2009 when we found no cases of peripheral arterial disease among older adults despite high average levels of CRP, and no relationship between CRP and the continuous ankle-brachial index used to assess peripheral arterial disease risk1. More recently, we reported minimal coronary artery calcification that progressed 25 years more slowly than in the US,, and again, without any relationship with CRP2.\n\nFKL assert that the typically high inflammation the Tsimane experience may be protective against non-communicable diseases (NCDs), and that their healthy hearts may not be due simply to their having protective diets, high moderate-to-vigorous activity levels and other lifestyle factors. We interpret their position that high inflammation is protective based on their discussion of field Eifel studies which showed CRP elevations after an attempt to mimic a more traditional “Paleo” lifestyle on a 4 day forest trek with low-carb diet, caloric restriction and semi-fasting. They propose three reasons for why CRP might be elevated under these (and presumably Tsimane conditions), but yet have no relationship with NCDs and their inflammation-related risk factors like smoking and obesity: 1) forest bathing may induce innate immune activity from plants releasing phyto-antibiotics into the atmosphere; 2) trekking in the “natural environment” (but not the city) may induce anticipatory immune defenses; and 3) caloric restriction combined with vigorous activity may have led to “cell depletion” and increased innate immune activity, including CRP, via uric acid byproducts.\n\nFirst, we disagree with their conclusion that a “chronic inflammatory situation in Tsimans [sic] protects against NCD”. This has never been shown. Rather, we have just shown that Tsimane do not appear to suffer from NCDs like heart disease and diabetes, despite their elevated systemic inflammation. Elsewhere, we have proposed that high inflammation may not lead to increased heart disease risk under certain conditions that may be particular for traditional subsistence populations like the Tsimane3,4. High systemic inflammation might not lead to greater NCD risk when combined with: 1) low LDL, a physically demanding lifestyle and minimal obesity; 2) chronic helminthic infections that: a) modulate immune function toward more anti-inflammatory Th2 activity that helps prevents systemic inflammation from damaging arteries, b) increase basal metabolic rate and lower obesity risk; c) reduce blood lipids such as LDL cholesterol and triglycerides, and blood glucose. Also notable is their limited exposure to tobacco smoke.\n\nSecond, while the Eifel study results suggest that adopting a more “Paleo”-like lifestyle can improve cardiometabolic biomarkers, we disagree that they show that higher inflammation (or innate immune activity in general) typifies hunter-gatherers or other subsistence populations. While CRP increased by 67% after the four day forest trek of healthy adults, this increase was not statistically significant (see Table 4). Thus, given the small sample size (n=25), these speculations have limited basis and are premature. Moreover, we see other issues with generalizing from the Eifel study. Half of the study participants did not engage in regular physical activity (52% <3 hrs/week). Heavy physical activity could lead to acute increases in inflammation, whereas long-term effects of regular exercise are usually associated with reduced chronic low grade inflammation. It is also possible that during the experiment, some participants incurred minor injuries or acute infections, both of which could result in higher CRP. Psychosocial stress, as might occur under the Eifel study conditions of food restriction for example, can also increase inflammation, especially IL-6, IL-1B and CRP5.\nWhile Tsimane show evidence of chronic, elevated CRP across much of the life course6,7, it is yet to be determined how representative this pattern is among small-scale subsistence populations. Shuar forager-horticulturalists of Ecuador show elevated CRP only when infected, but otherwise have low baseline CRP levels8. A small sample of Hadza hunter-gatherers also showed reasonably low CRP levels (74% of 23 adults had CRP < 3 mg/L)9.\n\nAnticipatory stimulation of the innate immune system in the forest is an interesting idea; even visual exposure to photographs with disease cues (e.g. sneezing) can increase IL-6 by 24% compared to controls shown neutral stimuli10. However, while short-term changes in other biomarkers like blood lipids and blood glucose may be generalizable from the Eifel study to a long-term study, it is unclear whether any acute innate immune activity observed over a 4 day period would sustain in a longer-term study after habituation. As FKL argue, there are many other reasons CRP can be elevated that have little to do with inflammation, but might still be related to tissue stress or injury11.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? No",
"responses": [
{
"c_id": "3861",
"date": "06 Aug 2018",
"name": "Jens Freese",
"role": "Author Response",
"response": "We appreciate that Freese and colleagues [FKL] attempt to understand what they refer to as the “inflammation paradox”, i.e. that people may experience high chronic levels of circulating inflammation (as measured by biomarkers such as high sensitivity C-reactive protein (CRP)), yet not succumb to the diseases such as heart disease and diabetes usually attributed to such high exposure in urban settings. We first documented this pattern among Tsimane Amerindians in Bolivia in 2009 when we found no cases of peripheral arterial disease among older adults despite high average levels of CRP, and no relationship between CRP and the continuous ankle-brachial index used to assess peripheral arterial disease risk1. More recently, we reported minimal coronary artery calcification that progressed 25 years more slowly than in the US,, and again, without any relationship with CRP2. FKL assert that the typically high inflammation the Tsimane experience may be protective against non-communicable diseases (NCDs), and that their healthy hearts may not be due simply to their having protective diets, high moderate-to-vigorous activity levels and other lifestyle factors. We interpret their position that high inflammation is protective based on their discussion of field Eifel studies which showed CRP elevations after an attempt to mimic a more traditional “Paleo” lifestyle on a 4 day forest trek with low-carb diet, caloric restriction and semi-fasting. They propose three reasons for why CRP might be elevated under these (and presumably Tsimane conditions), but yet have no relationship with NCDs and their inflammation-related risk factors like smoking and obesity: 1) forest bathing may induce innate immune activity from plants releasing phyto-antibiotics into the atmosphere; 2) trekking in the “natural environment” (but not the city) may induce anticipatory immune defenses; and 3) caloric restriction combined with vigorous activity may have led to “cell depletion” and increased innate immune activity, including CRP, via uric acid byproducts. First, we disagree with their conclusion that a “chronic inflammatory situation in Tsimans [sic] protects against NCD”. This has never been shown. Rather, we have just shown that Tsimane do not appear to suffer from NCDs like heart disease and diabetes, despite their elevated systemic inflammation. Elsewhere, we have proposed that high inflammation may not lead to increased heart disease risk under certain conditions that may be particular for traditional subsistence populations like the Tsimane3,4. High systemic inflammation might not lead to greater NCD risk when combined with: 1) low LDL, a physically demanding lifestyle and minimal obesity; 2) chronic helminthic infections that: a) modulate immune function toward more anti-inflammatory Th2 activity that helps prevents systemic inflammation from damaging arteries, b) increase basal metabolic rate and lower obesity risk; c) reduce blood lipids such as LDL cholesterol and triglycerides, and blood glucose. Also notable is their limited exposure to tobacco smoke. Answer: We agree that there are no data showing that chronic inflammation in Tsimans protects them against NCD (correlation is not causation). We have erased this claim. Rather, we agree that the best explantion is probably the absence of typical NCD risk factors. Therefore, the Tsimane are noteworthy for being an example that chronic high inflammation is not sufficient for increasing NCD risk, given that other conditions are satisfied as you mentioned. This is now emphasized in the revision. Second, while the Eifel study results suggest that adopting a more “Paleo”-like lifestyle can improve cardiometabolic biomarkers, we disagree that they show that higher inflammation (or innate immune activity in general) typifies hunter-gatherers or other subsistence populations. While CRP increased by 67% after the four day forest trek of healthy adults, this increase was not statistically significant (see Table 4). Thus, given the small sample size (n=25), these speculations have limited basis and are premature. Moreover, we see other issues with generalizing from the Eifel study. Half of the study participants did not engage in regular physical activity (52% <3 hrs/week). Heavy physical activity could lead to acute increases in inflammation, whereas long-term effects of regular exercise are usually associated with reduced chronic low grade inflammation. It is also possible that during the experiment, some participants incurred minor injuries or acute infections, both of which could result in higher CRP. Psychosocial stress, as might occur under the Eifel study conditions of food restriction for example, can also increase inflammation, especially IL-6, IL-1B and CRP5. Answer: You are correct that high-grade inflammation is not a general hallmark of hunter-gatherer populations, and we apologize if we seem to have made such a statement. Our goal was simply to propose several possible hypotheses fort he observed acute elevations in CRP in the Eifel studies, partly in preparation for a new study that is going to start this summer. We also state in the revision that the data are not able to provide evidence for one of our hypotheses over any other, so that future studies should particularly put these hypotheses to test. While Tsimane show evidence of chronic, elevated CRP across much of the life course6,7, it is yet to be determined how representative this pattern is among small-scale subsistence populations. Shuar forager-horticulturalists of Ecuador show elevated CRP only when infected, but otherwise have low baseline CRP levels8. A small sample of Hadza hunter-gatherers also showed reasonably low CRP levels (74% of 23 adults had CRP < 3 mg/L)9. Answer: Thank you for pointing this out. We added the references to these studies that indicate that the Tsimane CRP levels are apparently higher than that of other indegenous people. Anticipatory stimulation of the innate immune system in the forest is an interesting idea; even visual exposure to photographs with disease cues (e.g. sneezing) can increase IL-6 by 24% compared to controls shown neutral stimuli10. However, while short-term changes in other biomarkers like blood lipids and blood glucose may be generalizable from the Eifel study to a long-term study, it is unclear whether any acute innate immune activity observed over a 4 day period would sustain in a longer-term study after habituation. As FKL argue, there are many other reasons CRP can be elevated that have little to do with inflammation, but might still be related to tissue stress or injury11. Answer: We agree that it is unclear if and when the elevated CRP levels found in the Eifel study participants would have decreased again. However, a similar study has found increased CRP levels after a 10-day trip through the wilderness. This is stated now in the last two sentences of the conclusions section: “In particular it remains to be determined, if, when and how the acute elevations in CRP level resolve if the participants would remain in Paleolithic living conditions. Interestingly, a study similar to the Eifel studies involving a10-day trip through the Pyrenees also reported a significant mild increase in CRP levels after the trip, showing that the acute phase repsonse could be sustained at least up to 10 days (29-30).“"
}
]
},
{
"id": "31345",
"date": "06 Apr 2018",
"name": "Colette Berbesque",
"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 statement is organised around two discoveries: the highly inflammatory immune profile in Tsimane associated with low prevalence of NCDs, and the increased inflammation in westerners whom participated in a 'Paleo-trek'1. From these observations, FKL reach the conclusion that chronic inflammation in the Tsimane protects against NCDs and propose explanations for why this might be the case. We will discuss this assumption in the following sections.\n\nFirst, we disagree with the assumption that Tsimane people stand as an adequate model for the study of the relationship between Low-grade inflammation (LGI) and non-communicable diseases (NCDs). LGI, associated with coronary diseases and other NCDs in westerners, is usually not triggered by specific stimuli such as infection or injury2, but strongly associated with other risk factors such as stressful situations3 and excess of adipose tissue4. Tismanes are less mobile than most hunter-gatherers, such as the Hadza - who demonstrate low CPR5. This increased sedentisation may be part of the reason they face a heavy parasitic load and bacterial infections6 likely responsible for their hightened inflammatory response7, along with low levels of cholesterol, LDL and HDL. It is also noteworthy that the major comparative study on Tsimane immune function failed to show that after childhood there was any significant increase in CRP, a major low-grade inflammatory marker7. The hypothesis that the parasitic burden triggering Th2 response may enhance anti-inflammatory cytokine levels and thus protect against NCDs is interesting. However, the absence of major NCDs risk factors, such as chronic stress, insulin-resistance and elevated fatty acid levels, brings into question the underlying assumption that the Tsimane may face common NCDs.\n\nFinally, we strongly reject the authors' conclusion that inflammation may protect against NCDs in Tsimane. Indeed, CRP levels were found to be a positive predictor of non-null coronary artery calcium scores8, suggesting CRP might actually be predictive of atherosclerosis, a major NCD, in Tsimane as well.\n\nAccording to the authors, the inflammatory state of the Tsimane could be partly explained by environmental factors also responsible for the pro-inflammatory profile observed in the participants of the Eifel experiment. We will review and discuss two of the hypotheses proposed by FKL to explain this result.\n\nFirst, a ‘forest bathing’ effect would have enhanced innate immune response and therefore result in an increase in CRP levels. Although we do not wish to undermine the potential role of the ‘forest bathing’ effect, to the best of our knowledge no evidence so far suggests that it could have an impact on CRP levels. Most studies focus on NK cells, cortisol levels and bio-markers of endothelial function. As CRP elevation in the Eifel participants failed to reach significance and NK cells were not measured, the ‘forest bathing’ effect remains speculative.\n\nSecond, as suggested by the authors, exposure to exogenous pathogen associated-molecules could have caused a shift towards Th1 response in Eifel participants. However, there is no evidence to support the assumption that this shift was 'prophylactic' as participants may have well indeed face acute infection from small wounds or respiratory pathogens in a forest environment. In order to strengthen their theoretical argument, the authors could provide references regarding the resolution of chronic inflammation following an episode of acute inflammation.\n\nFinally, we would like to suggest another potential origin for the pro-inflammatory immune profile in the Eifel participants. It seems participants exhibit the typical leukocyte count induced by stress and exercise9: neutrophilia, lymphopenia and eosinopenia. Such leukocyte pattern is supposedly the result of cortisol-induced cell-migration into peripheral tissue10 and does not fit with Tsimane leukocyte pattern7. Furthermore, exercise-induced response has been shown to gradually decrease in trained athletes. Long-term training is eventually followed by the release of anti-inflammatory cytokines11,12. Such difference between initial-shock immune response to intense exercise, as undertaken by untrained Eifel participants, and long-term immune adaptation to physical activity brings into question the reliability of the leukogram patterns measured on the fourth day of trek in the Eifel study. A shift towards a more anti-inflammatory pattern could therefore be expected in the Eifel participants after acclimatisation to this new lifestyle, regardless of environmental factors. It is therefore highly speculative that the Eifel participants’ immune profile would be a good model for the immune pattern of a Paleo-population.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? No",
"responses": [
{
"c_id": "3860",
"date": "06 Aug 2018",
"name": "Jens Freese",
"role": "Author Response",
"response": "The authors statement is organised around two discoveries: the highly inflammatory immune profile in Tsimane associated with low prevalence of NCDs, and the increased inflammation in westerners whom participated in a 'Paleo-trek'1. From these observations, FKL reach the conclusion that chronic inflammation in the Tsimane protects against NCDs and propose explanations for why this might be the case. We will discuss this assumption in the following sections. First, we disagree with the assumption that Tsimane people stand as an adequate model for the study of the relationship between Low-grade inflammation (LGI) and non-communicable diseases (NCDs). LGI, associated with coronary diseases and other NCDs in westerners, is usually not triggered by specific stimuli such as infection or injury2, but strongly associated with other risk factors such as stressful situations3 and excess of adipose tissue4. Tismanes are less mobile than most hunter-gatherers, such as the Hadza - who demonstrate low CPR5. This increased sedentisation may be part of the reason they face a heavy parasitic load and bacterial infections6 likely responsible for their hightened inflammatory response7, along with low levels of cholesterol, LDL and HDL. It is also noteworthy that the major comparative study on Tsimane immune function failed to show that after childhood there was any significant increase in CRP, a major low-grade inflammatory marker7. The hypothesis that the parasitic burden triggering Th2 response may enhance anti-inflammatory cytokine levels and thus protect against NCDs is interesting. However, the absence of major NCDs risk factors, such as chronic stress, insulin-resistance and elevated fatty acid levels, brings into question the underlying assumption that the Tsimane may face common NCDs. Answer: We apologize if we seemed to have suggested the Tsimane as an adequate model for studying the connection between LGI and NCD. This was not our aim. Rather, our aim was to point out the Tsimane as an interesting case of indigenous people who are apparently free from the NCD that plague the Western societies despite having chronically elevated CRP levels. This should make clear that they are not an adequate model, but rather an exceptional case worth further studying. In our revised manuscript, we strengthen the notion that Tsimane do not face the usual risk factors for NCDs that Westeners do, agreeing with your argument above. This is probably the best explanantion for their low incidence of NCDs. Finally, we strongly reject the authors' conclusion that inflammation may protect against NCDs in Tsimane. Indeed, CRP levels were found to be a positive predictor of non-null coronary artery calcium scores8, suggesting CRP might actually be predictive of atherosclerosis, a major NCD, in Tsimane as well. Answer: We are sorry for making causal claims where the data do not clearly support any. We have been more careful with our formulations in the revised manuscript, refraining from making any causal claims. According to the authors, the inflammatory state of the Tsimane could be partly explained by environmental factors also responsible for the pro-inflammatory profile observed in the participants of the Eifel experiment. We will review and discuss two of the hypotheses proposed by FKL to explain this result. First, a ‘forest bathing’ effect would have enhanced innate immune response and therefore result in an increase in CRP levels. Although we do not wish to undermine the potential role of the ‘forest bathing’ effect, to the best of our knowledge no evidence so far suggests that it could have an impact on CRP levels. Most studies focus on NK cells, cortisol levels and bio-markers of endothelial function. As CRP elevation in the Eifel participants failed to reach significance and NK cells were not measured, the ‘forest bathing’ effect remains speculative. Answer: We agree that this hypothesis is currently only speculative until further measurements have been made which are able to distinguish it from other hypotheses such as the acute phase response due to strenuous physical activity. We have re-phrased the Outlook section accordingly. The same is true for our second hypothesis. Nevertheless, we consider them possible explanations that should be considered in data interpretation. Our future research is planned to more specifically address the forest bathing hypothesis. Second, as suggested by the authors, exposure to exogenous pathogen associated-molecules could have caused a shift towards Th1 response in Eifel participants. However, there is no evidence to support the assumption that this shift was 'prophylactic' as participants may have well indeed face acute infection from small wounds or respiratory pathogens in a forest environment. In order to strengthen their theoretical argument, the authors could provide references regarding the resolution of chronic inflammation following an episode of acute inflammation. Answer: We had already noted that this hypothesis is currently only speculative. However, since we are mainly interested in explaining the acute rise of CRP levels, we have deleted any hypothesizing that this acute phase response could have led to the resolution of chronic LGI. An example for the resolution of chronic inflammation after acute stimulation of the immune system (although not neccessarily translatable to all NCDs) is the spontaneous remission of cancer after an acute episode of fever that is now more or less officially acknowledged (e.g. Jessy T (2011): Immunity over inability: The spontaneous regression of cancer. J Nat Sci Biol Med; 2(1): 43–49). Other examples from medical practice include treatment of chronic tendonitis through re-injury or using low-dose ionizing radiation to treat chronic inflammatory conditions. Finally, we would like to suggest another potential origin for the pro-inflammatory immune profile in the Eifel participants. It seems participants exhibit the typical leukocyte count induced by stress and exercise9: neutrophilia, lymphopenia and eosinopenia. Such leukocyte pattern is supposedly the result of cortisol-induced cell-migration into peripheral tissue10 and does not fit with Tsimane leukocyte pattern7. Furthermore, exercise-induced response has been shown to gradually decrease in trained athletes. Long-term training is eventually followed by the release of anti-inflammatory cytokines11,12. Such difference between initial-shock immune response to intense exercise, as undertaken by untrained Eifel participants, and long-term immune adaptation to physical activity brings into question the reliability of the leukogram patterns measured on the fourth day of trek in the Eifel study. A shift towards a more anti-inflammatory pattern could therefore be expected in the Eifel participants after acclimatisation to this new lifestyle, regardless of environmental factors. It is therefore highly speculative that the Eifel participants’ immune profile would be a good model for the immune pattern of a Paleo-population. Answer: Thank you for these thoughts which go along the lines of our third hypothesis, namely that the sudden physical activity increase would be responsible for the acute CRP increase. We have included this additional mechanism in our discussion and re-wrote a large part of the third hypothesis. We think that the revision now also distinguishes more clearly between the chronic high CRP levels of the Tsimane and the acute elevations observed in the Eifel studies. In no way did we intent to propose that the immune profile of the latter is a good model for the former."
}
]
}
] | 1
|
https://f1000research.com/articles/7-252
|
https://f1000research.com/articles/7-1200/v1
|
06 Aug 18
|
{
"type": "Research Article",
"title": "The antimicrobial activity of alcoholic extracts of leaves and bark from two varieties of ”Sangre de Drago” compared with the antimicrobial activity present in the latex of the same species",
"authors": [
"Cecilia Barba Guevara",
"Luis Montaluisa",
"María Elena Maldonado Rodriguez",
"Luis Montaluisa",
"María Elena Maldonado Rodriguez"
],
"abstract": "Background: This research was carried out in order to study the antimicrobial effectiveness of crude latex of two varieties of \"Sangre de Drago\": Croton lechleri Muller Arg. and Croton urucurana Baill and compare that effectiveness to the antimicrobial activity of the alcoholic extracts of its leaves and bark. Methods: The activity of the alcoholic extracts and latex were evaluated against bacterial strains of Staphylococcus epidermidis, Bacillus subtilis and Escherichia coli. The extraction of the alcoholic extracts (20% Tincture) of the leaves, bark and latex from the two Croton species was carried out by maceration using 70% alcohol as a menstruum, at room temperature, for 2 to 7 days, with shaking at least twice a day. A 20% tincture was obtained, from which the physical and chemical parameters were determined as indicated by the Ecuadorian Quality Control Standard for natural medicinal products. Results: It was found that both the alcoholic extracts of the plant material and the crude latex indicate antimicrobial activity for S. epidermidis, moderate antimicrobial activity for B. subtilis and no antimicrobial activity for E. coli. The moderate antimicrobial activity against B. subtilis, at doses of 125 p.p.m., is in line with the findings of previous studies by other authors. Conclusions: he antimicrobial activity of the latex of the two species against S. epidermidis is not registered in literature and, the negative antimicrobial activity for E. coli does not agree with what has been reported by previous studies.",
"keywords": [
"Sangre de Drago",
"Croton lechleri Muller Arg.",
"Croton urucurana Baill",
"antimicrobial activity"
],
"content": "Introduction\n\nIn Ecuador, there is a growing interest in medical plants from medical practitioners1,2. “Sangre de Drago“, the name by which the plant and its medicinal latex are known, are a group of Croton (Euphorbiaceae) tree species widely distributed in Latin America. In Ecuador, they are found in the Amazon region and, as in Bolivia, Colombia and Peru, the most commonly identified species of Sangre de Drago is Croton lechleri Muller Arg3. Croton urucurana Baill, a species found in the subtropics of South America, which is representative of the north of the country, islands and coasts of the Uruguay River, does not appear in the literature information for the Croton genus in Ecuador4.\n\nThe red latex and bark of Sangre de Drago have been used in popular South American indigenous medicine for different purposes, including wound healing5, diarrhoea, influenza, tonsillitis, intestinal disorders, herpes, fertility enhancement, tuberculosis6, hepatitis, cancer prevention, anti-inflammatory problems, acne, weight loss, coughs and colds7. In Ecuador, C. lechleri is used for treating haemorrhages, wounds and tuberculosis2. The chemical composition of latex has been widely studied, revealing the presence of different metabolites, amongst which are the alkaloid taspine and its salt, lignans, proanthocyanidins, terpenes and phenolic compounds. These metabolites are responsible for the different activities that this medicinal plant presents. Taspine and lignans have anti-inflammatory activity, taspine hydrochloride has wound healing activity, and phenolic compounds have antimicrobial and antiviral activity8. The antimicrobial activity of Sangre de Drago in Gram-positive bacteria has been demonstrated, such as Staphylococcus aureus ATCC 6538 and Staphylococcus epidermidis ATCC 12228; and with the Gram-negative Pseudomonas and Klebsiela FDA 6029. In this study, the antimicrobial activity of latex and of alcoholic extracts of leaves and bark of two species of Sangre de Drago was evaluated in vitro against specific bacterial strains in order to compare their antibacterial properties.\n\n\nMethods\n\nThe vegetal material used from the two species of Sangre de Drago was collected in the localities of Talag and Canelos belonging to the provinces of Napo and Pastaza, respectively, and located at 1°, 3’, 57” south latitude; 77°, 54’, 26” west longitude; 40° west longitude; and 1°, 35’, 22” south latitude; 77°, 44’, 48” west longitude; 40° west longitude, respectively. The sample collection was permitted by the Ministerio del Medio Ambiente, Ecuador (MAE), (registration number: 2015-RO. Nro. 449 and 905-RO.- suplemento 553). Aerial parts (leaves and bark) and latex of the trees aged 3 to 4 years and in a flowering-fructification stage were used. The crude latex for the analysis was obtained from indigenous merchants in the areas. One of the two latex samples corresponding to the Canelos sector was extracted from the same tree from which the plant material was collected. The collecting folders for each species were prepared and sent to the National Herbarium of Ecuador (QCNE) for taxonomic identification. The folder regarding the material collected in the Province of Napo was identified as C. lechleri Muller Arg. and the one corresponding to the province of Pastaza as C. urucurana Baill. The aerial parts (leaves and bark) of the two species were dried at 49°C, in a JP SELECTA series 0385081 stove with circulating air for 3 days. The plant material was removed at least once a day for uniform drying until a constant weight was achieved. Next, the parts were crushed in a hammer mill until a moderately thick powder was obtained. The batches of the plant material were identified for the corresponding analyses with the letters “l” and “b” for leaves and bark, respectively, accompanied by the letters T and C to identify provenance from Talag and Canelos, respectively. The latex samples were refrigerated at 4°C for 2 weeks, during which time their analysis began from their acquisition. They were coded with the letters “T” and “C” and the letter “C” was accompanied with numbers “1” and “2” for the latex provided by the indigenous merchants and for the latex obtained from the same tree as the vegetal material was collected, respectively.\n\nThe extraction of the alcoholic extracts (20 % Tincture) of the leaves, bark and latex from the two Croton species was carried out by maceration using 70% alcohol as a menstruum, at room temperature, for 2 to 7 days, with shaking at least twice a day. A 20% tincture was obtained, from which the physical and chemical parameters were determined as indicated by the Ecuadorian Quality Control Standard for natural medicinal products. Under the same conditions, the physical-chemical constants of the crude latex of Sangre de Drago samples were determined. For the determination of the refractive index, a Clean-Prison series 003021 ABBE refractometer was used. pH determination was carried out in a Metrohm potentiometer, type 1.744.0010.\n\nThe fractionation of the analysed alcoholic extracts was achieved by thin-layer chromatography, tapping 0.1 ml of the alcoholic extract by means of a glass capillary 1 cm from the bottom edge of 60GF254 silica gel plates. It was allowed to dry for a moment before each application and at the end of it. Then the plates were placed in saturated glass chambers for at least 30 minutes with steam from the components of each of the mobile phases used, which are listed below as ratios:\n\n• Butanol-acetic acid-water (BAW): 4-1-5\n\n• Chloroform-acetone-diethylamine: 7-3-1\n\n• Toluene-ethyl acetate: 70–30\n\nThe following were used as developers: 254 nm and 366 nm ultraviolet light, Dragendorff reagent, ammonia fumes. Each chromatography experiment was repeated twice.\n\nThe microbiological quality of the alcoholic extracts of leaves and barks, as with crude latex, was verified as outlined by the World Health Organization (WHO)10. The tally of bacteria, fungi and yeast was read after 5 days of incubation as indicated by the WHO. Regarding the reading for the determination of pathogenic bacteria, this was performed at 24 and 48 h of incubation with two repeats. The results of bacterial growth were compared with that described by Porter and Kaplan11 for each culture medium used, for the purpose of bacterial recognition.\n\nThe evaluation of the antimicrobial activity was carried out against six bacterial strains (Escherichia. coli, S. aureus, Pseudomonas aeruginosa, S. epidermidis, Bacillus subtilis and Streptococcus sp.) maintained in nutritious agar at 4°C according to Garcia and Uruburu12. The antifungal activity was not assessed since growth problems were encountered for the strain Candida albicans ATCC10231. The method used was described in the “CYTED” Research Techniques Manual9. The extracts were solubilized using dimethylsulfoxide (DMSO) to their respective dilutions (5000, 2500, 1250, 1000, 500, 250 and 125 p.p.m.). Gentamicin sulfate (0.1 mg/0.1 ml) was used as a standard antibiotic and as a growth inhibition solution. As a negative control, Muller Hilton Agar and Muller Hilton Agar boxes with DMSO (0.1 ml) were prepared.\n\n\nResults and discussion\n\nTable 1 shows the results of density, total solids, pH and refractive index of the alcoholic extracts of the bark, leaves and latex of the two Sangre de Drago varieties analyzed.\n\nAs can be seen, there is no marked difference between the values of density, total solids and pH of the different extracts analysed. The organoleptic characteristics, on the other hand, differ clearly between the extracts of the two Croton species. In the same way, the capillary analysis shows a clear differentiation between the extracts and basically between the latex extracts (Figure 1). The presence of a fluorescent blue colour on the fringe of the capillary analysis image was characteristic in all the analysed extracts. When the wet paper strip was exposed to ammonia fumes, revealed with 366 nm UV light and the environment left to dry, a weaker blue color was observed on the paper strips with 366 nm UV light. According to the consulted literature13, these results are characteristic of flavonoid-anthocyanidin type metabolites.\n\nTable 2 shows the values of the physical and chemical parameters of the crude latex of the two varieties of Sangre de Drago analysed. As can be seen, the organoleptic characteristics allow the two varieties to be clearly differentiated. In terms of density parameters, total solids and pH, the results vary for each species. The refractive index cannot be determined by the physical characteristics of the samples (colour and viscous consistency).\n\nð, density g/ml; ST, total solids.\n\nFigure 2 presents the results of the thin layer chromatography, RF of the spots and colours visualized with the developers: 366 nm UV light, 254 nm UV light, ammonia fumes, Drangendorf reagent, alcoholic extracts of leaves, bark and latex.\n\nAs can be seen, the separation of the respective fractions occurs with greater clarity in the plates run in the mobile phases of BAW and 70–30 toluene-ethyl acetate. The different fractions obtained are observed as yellow at the point of application and faint orange to the naked eye during the course of the shift.\n\nWhen revealed with 366 nm UV light, fluorescent orange spots are observed at different RF. When revealed with 254 nm UV light, faint grey spots are observed. The plates were passed through ammonia fumes and observed under 366nm UV light, where it was observed that the orange-coloured dots turned fluorescent blue. The phases were run in chloroform-acetone-diethylamine (7-2-1) mobile phase. Under natural light, no stain can be seen, but under the 366nm UV light, reddish spots are observed at a 0.48 RF and under 254 nm UV light, blue spots are observed. The RF of the calculated spots is closer to that reported in the literature for the alkaloid taspine (Theoretical RF of 0.50). When revealing the spots with Drangendorf’s reagent, a faint brown color was observed that disappeared after a few seconds. The presence of orange spots that turn blue in the presence of ammonia is characteristic of flavonoid-type phenolic compounds, and within these the anthocyanidins group, which are components present in Sangre de Drago14.\n\nThe results obtained from the count of bacteria, fungi and yeasts of the alcoholic extracts of leaves, bark and crude latex in colony forming units/ml of sample are found in Table 3.\n\nT, Talag; C, Canelos; C1, latex from unidentified tree; C2, latex from identified tree.\n\nAs can be observed, the values corresponding to the microbiological count of bacteria of the alcoholic extracts are within those established for crude drugs (see Table 4). The other two samples of latex did not present bacterial growth. The count of moulds and yeasts presents similar results to the bacterial count in the alcoholic extracts. The same cannot be said for the crude latex, because on the fifth day, the three analysed samples presented massive growth, which did not allow for the counting of the colony forming units. The presence of pathogenic batteries was negative for the alcoholic extracts and for the latex samples from “C1“ Talag and Canelos. This was not so for the “C2“ Canelos latex sample, which showed growth of pink colonies, characteristic of E. coli in red Violet Agar. The abundant bacterial growth with a metallic luster in Mac Conkey Agar of the planting in lines of characteristic colonies in Red Violet Agar confirmed the presence of E. coli. It was impossible to count the colony forming units, as they surpassed what was established by the WHO for this type of sample. All samples were negative for bacterial growth in Cetrimide Agar specific for P. aeruginosa and in Baird Parker Agar for S. aureus.\n\nSource: Quality Control Methods for Medicinal Plant Materials9\n\nFor the determination of the antibacterial activity by means of the method described in the CYTE Research Techniques Manual9, a battery of six bacteria was used (E. coli, S. aureus, P. aeruginosa, S. epidermidis, B. subtilis and Streptococcus sp.). The preliminary results of each of the bioassays performed are presented in Table 5, in which the moderate antimicrobial activity (±) for B. subtilis can be observed in the seven samples analysed (extracts and crude latex). For the remaining five bacteria, the antibacterial activity was negative (-) for the alcoholic extracts. The Sangre de Drago latex coded “T“ and “C1“ were active (+) for S. epidermidis at doses of 2500 p.p.m. and 5000 p.p.m. and the “C2“ latex is active (+) at a dose of 1250 p.p.m. For the four remaining bacteria, they did not exhibit antimicrobial activity. The negative activity (-) for E. coli does not agree with previous reports15–17. Regarding B. subtilis, it could be said that it is similar to results from other authors’ studies. The literature does not present information relating to activity on S. epidermidis. As for fungal activity, this test was not performed due to growth problems that arose with the C. albicans strain.\n\nLeyenda: Ec=Escherichia coli, Sa= Staphylococcus aureus, Pa= Pseudomona aeruginosa, Se= Staphylococcus epidermidis, Bs= Bacilus suptilis, Ssp= Streptococcus sp.\n\nR=resistant\n\nI= intermediate sensitivity\n\n\nConclusions\n\n1. The results of the physical-chemical constants of the alcoholic extracts do not allow one to differentiate the two Croton species studied. However, the organoleptic characteristics and the capillary analysis allow for differentiation between the alcoholic extracts and latex of C. lechleri and C. urucurana.\n\n2. The organoleptic characteristics of the latex of the two varieties allow one to differentiate them from each other. Regarding physical-chemical parameters, the values are very similar, which allows them to be used for their differentiation.\n\n3. The average pH value, of 3.7 for C. lechleri latex, 4.6 for C. urucuran C1 latex and 3.9 for C. urucurana C2 latex, are similar to that found in the literature (pH 4.4).\n\n4. The chromatographic analysis of the extracts of bark, leaves and latex of the two Sangre de Drago species allowed for the separation of this species’ characteristic components, alkaloids and phenols. These compounds make up the metabolites responsible for the biological properties of the Sangre de Drago.\n\n5. The antimicrobial activity of the alcoholic extracts of leaves, bark and latex was negative for S. epidermidis and E. coli, but not for B. subtilis, which was moderately positive.\n\n6. The three latex samples show antimicrobial activity for S. epidermidis. However, the sample of Canelos latex, obtained from the same tree that leaves and bark were collected from, presented antimicrobial activity at lower doses, 1250 p.p.m. Antimicrobial activity was not found in the reviewed literature. In relation to B. subtilis, moderate activity was observed in the three latex samples, an activity that was also evident in the alcoholic extracts of leaves and bark.\n\n7. The values of antimicrobial activity against E. coli and B. subtilis are found in the literature15–17 for C. Lechleri, but not for C. Urucurana, a species for which no information is available.\n\n\nData availability\n\nDataset 1. Images and raw data from all chromatography experiments under all development methods with RF values, in addition to complete microbiological counts and physical/chemical characteristics of extracts. Also included are images of the plates used in this study, although it should be noted that these images were captured after storage for 2 or more years. Chromatography data 1 contains data generated using a mobile phase of Butanol:acetic acid:water (4:1:5) visualized using ferric chloride. Chromatography data 2 contains data generated using a mobile phase of chloroform:acetone:water (7:3:1) visualized using ammonia vapor; Chromatography data 3 contains data using a mobile phase of chloroform:acetone:diethylamine (7:3:1) visualized using ammonia vapor. DOI: 10.5256/f1000research.14575.d21135818.",
"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\nTapia W, Garzón K, Granda N, et al.: Alexiteric activity of Costus pulverulentus C. Presl., Desmodium adscendens (Sw.) DC., Begonia glabra Aubl. and Equisetum bogotense on the poison of Bothrops asper (equis) [version 1; referees: 2 not approved]. F1000Research. 2018; 7: 136. Publisher Full Text\n\nRisco E, Vila R, Henriques A, et al.: Bases químicas y farmacològicas de utilizaciòn de sangre de drago. Revista de Fitoterapia. 2005; 5(2): 101–114.Reference Source\n\nRisco E, Iglesias J, Cañigueral S: Sangre de Drago. Interés Terapéutico del látex de Croton lechleri. 12eme Forum De Natura Rerum. Paris. 2001. Reference Source\n\nMoacir P: Ditermpenos clerodano de corteza de Croton urucurana Baillon. Recuperado el26 de October, 2015.\n\nPieters L: Aislamiento de un dihidrobenzofurano (Croton spp.) como inhibidor de la proliferaciòn celular. J Nat Prod. 1993; 899–900.\n\nTempesta MS: Proanthocyanidin polymers having antiviral activity and methods of obtaining same. 1993; 4, 212, 944. Reference Source\n\nUbillas R, Jolad SD, Bruening RC, et al.: SP-303, an antiviral oligomeric proanthocyanidin from the latex of Croton lechleri (Sangre de Drago). Phytomedicine. 1994; 1(2): 77–106. PubMed Abstract | Publisher Full Text\n\nZapata R: Actividad antimicrobiana in vitro de la droga comercializada como sangre de grado. Lima, Perú: UNMSM.\n\nWHO: Edition of Quality control methods for medicinal plant materials. 1998. Reference Source\n\n1. Plants, Medicinal. 2. Medicine, Herbal. 3. Medicine, Traditional. 4. Quality control. 5. Manuals. I. World Health. Organization. ISBN 978 92 4 150073 9 (NLM classification: QV 766), 2011. Reference Source\n\nPorter R, Kaplan J: Merck EM. Editorial Medica Panamericana, ISBN 9788498357530, 2014. Reference Source\n\nGarcía M, Uruburu F: La conservación de cepas microbianasColección microbiana de cultivos tipo (CECT) Universidad de Valencia, España. Temas de la actualidad.\n\nLock O: Investigación Fitoquímica. Métodos en el estudio de productos naturales. Pontificia Universidad Católica del Perú. Segunda Edición, 1994; 123–124. Reference Source\n\nRamírez G: Sangre de Drago (Croton lechleri Muell. Arg). Fitoterapia. Revisiones Monográficas. Natura Medicatrix. 2003; 21(4): 213–217. Reference Source\n\nZaidan MR, Noor Rain A, Badrul AR, et al.: In vitro screening of five local medicinal plants for antibacterial activity using disc diffusion method. Trop Biomed. 2005; 22(2): 165–170. PubMed Abstract\n\nYaya JH, Flores MF, Castro HL: Control microbiológico y evaluación de la actividad antibacteriana in Vitro de croton lechleri \"Sangre de grado\". [en línea], 2003. [revisado 2018]. Reference Source\n\nGallardo Vásquez GJ, Barboza Mejía L: Efecto cicatrizante del gel elaborado del látex de Croton lechleri \"Sangre de Drago\". Revista Científica Ciencia Médica. 2015; 18(1): 10–16. Reference Source\n\nBarba Guevara C, Montaluisa L, Maldonado Rodriguez ME: Dataset 1 in: The antimicrobial activity of alcoholic extracts of leaves and bark from two varieties of \"Sangre de Drago\" compared with the antimicrobial activity present in the latex of the same species. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.14575.d211358"
}
|
[
{
"id": "36857",
"date": "16 Aug 2018",
"name": "Dinithi C. Peiris",
"expertise": [
"Reviewer Expertise Cancer research",
"diabetes",
"antimicrobial",
"signalling pathways etc.."
],
"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\nGeuvara et al., describes a comparison of antimicrobial activities of leaves and bark of 2 varieties “Sandre de Dargo” to that of latex of the same species. However, several points should be corrected/clarified before acceptance.\n\nThe name of the botanist with complete affiliation after the scientific name “in Gram-positive bacteria has been demonstrated, such as Staphylococcus aureus ATCC 6538 and Staphylococcus epidermidis ATCC 12228” – Better to revise the sentence. Better to include the time and the period of collection. Did authors deposit a voucher specimen and allocated a herbarium number? Why did authors conducted alcoholic extraction? Was it based on previous studies? The introduction should include a survey of previous work of the plant the authors are investigating What are the gram-negative and gram-positive bacteria used? Please separate the strains accordingly. Better to expand antimicrobial testing experiment Did authors only count the colonies formed. You can go through the following paper for methodological details. Bandara K.V., Padumadasa C., Peiris L.D.C. (2018)1. Authors should report how they measure diameter of inhibition zone and MIC. Figure 1 – Please italic the scientific name and also better to indicate what the colors denote. What is the title of the table 1\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36855",
"date": "22 Aug 2018",
"name": "Gabriel Trueba",
"expertise": [
"Reviewer Expertise Microbial genetics"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper entitled \"The antimicrobial activity of alcoholic extracts of leaves and bark from two varieties of ”Sangre de Drago” compared with the antimicrobial activity present in the latex of the same species\" describes some chemical analysis and antimicrobial properties of the alcoholic extract and latex form a medicinal plant from Ecuador. The topic of the paper is interesting but there is little additional information compared with previously published reports. Additionally, the purification methods and physiochemical analysis are simple and insufficient to provide useful information. The paper should not have the results of standard microbiological parameters accepted by Ecuadorian institutions.\nMinor points\n\nBefore considering for indexing, the paper requires a lot of improvement in style. The authors should pay attention to the references. References 8 and 9 don't correspond with the text. Page 3: This should be eliminated \"Under the same conditions, the physical chemical constants of the crude latex of Sangre de Drago samples were determined. For the determination of the refractive index, a Clean-Prison series 003021 ABBE refractometer was used. pH determination was carried out in a Metrohm potentiometer, type 1.744.0010.\" Page 4, thin layer chromatography results are insufficient to identify taspine as indicated in the text Page 5. The methods utilized to detect E. coli are not adequate. Table 1 . legend requires a lot more information Figure 1. This figure does not provide any relevant information. Table 5. There is no explanation about what intermediate sensitivity means\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1200
|
https://f1000research.com/articles/7-1199/v1
|
06 Aug 18
|
{
"type": "Method Article",
"title": "Common ELIXIR Service for Researcher Authentication and Authorisation",
"authors": [
"Mikael Linden",
"Michal Procházka",
"Ilkka Lappalainen",
"Dominik Bucik",
"Pavel Vyskocil",
"Martin Kuba",
"Sami Silén",
"Peter Belmann",
"Alexander Sczyrba",
"Steven Newhouse",
"Ludek Matyska",
"Tommi Nyrönen",
"Michal Procházka",
"Ilkka Lappalainen",
"Dominik Bucik",
"Pavel Vyskocil",
"Martin Kuba",
"Sami Silén",
"Peter Belmann",
"Alexander Sczyrba",
"Steven Newhouse",
"Ludek Matyska",
"Tommi Nyrönen"
],
"abstract": "A common Authentication and Authorisation Infrastructure (AAI) that would allow single sign-on to services has been identified as a key enabler for European bioinformatics. ELIXIR AAI is an ELIXIR service portfolio for authenticating researchers to ELIXIR services and assisting these services on user privileges during research usage. It relieves the scientific service providers from managing the user identities and authorisation themselves, enables the researcher to have a single set of credentials to all ELIXIR services and supports meeting the requirements imposed by the data protection laws. ELIXIR AAI was launched in late 2016 and is part of the ELIXIR Compute platform portfolio. By the end of 2017 the number of users reached 1000, while the number of relying scientific services was 36.\nThis paper presents the requirements and design of the ELIXIR AAI and the policies related to its use, and how it can be used for serving some example services, such as document management, social media, data discovery, human data access, cloud compute and training services.",
"keywords": [
"authentication",
"authorisation",
"IAM",
"data access",
"GDPR",
"GA4GH"
],
"content": "1. Introduction\n\nReliable electronic authentication of users is a requirement to authorise and manage access to the key services and instruments, sensitive data and computing capacities targeted to life science use. Life Science services need to, for example,\n\n1. Identify and group users based on their home organisation (such as the employing institution) information reliably;\n\n2. Manage administrational costs as the number of users and their access rights increase;\n\n3. Prevent poor security practices, such as sharing of user credentials;\n\n4. Ensure access to the data and services is closed as justification, such as the user’s organisational affiliation or other status, changes;\n\n5. Provide separate views on services for users acting in different roles or status (like a bona fide researcher with scientific record of successful research on human data), and\n\n6. Provide an audit trail to comply with privacy laws, including a proper reaction to security incidents.\n\nIncreasing the availability of services and data that are subject to access control for scientific, technological and innovation opportunities is a widely-accepted goal in Europe. Within the life-science community this concept is highlighted in management of research consented human data. ELIXIR, the European research infrastructure for biological data, unites the Europe’s leading life-science organisations in managing and safeguarding the massive amounts of data being generated in publicly funded research. It coordinates, integrates, and sustains bioinformatics resources across its member states, and enables users in academia and industry to access vital data, tools, standards, computational and training services for their research. Research Infrastructures (RI), such as ELIXIR, must support federated access to data that needs to be access-controlled.\n\nTechnologies in access management are developing rapidly, and new concepts are needed globally to respond to new requirements, including the General Data Protection Regulation1, without creating unnecessary obstacles for research in e.g. human health. In the long term, cross-border access to data resources and services is suggested to lead to beneficial solutions between national and international organisations and their stakeholders.\n\nFor example, international collaboration in resource access enables a scalable data infrastructure. Data are going to be described using internationally recognised best practices and standards to enable its pooling. This is crucial for individual reference data set analysis, like the Finnish population genomics2, which is challenging to analyse without access to international reference data and a scalable processing platform.\n\nA number of life-science services must authenticate their users and enforce access and resource policies3. As an outcome, research services have deployed local access management solutions issuing credentials, typically usernames and passwords, for their customers. As a consequence, the users are overloaded by too many credentials they need to remember for login to different services, leading them to adopt poor practices like writing the passwords down or using the same passwords in different services.\n\nDeveloping and operating access management solutions is expensive, especially advanced services, such as, multi-factor authentication for improved security or user-friendly workflow systems for applying roles and permissions to access resources. These kind of services can be better deployed in a centralised setup where a single access management solution serves several research services (called relying services). Authentication and authorisation is not a core function for research and being able to off-load related functionality would enable the managers of the research services to focus on their core competencies.\n\nIn research and education, access management services have been subject of interest since the proliferation of the world-wide web during the latter part of 1990’s. The concept of an authentication and authorisation infrastructure (AAI) was introduced in Switzerland in 2001 as a service layer that provides AAI services for relying parties4. In the commercial sector, the concept of Identity and Access Management (IAM) has later emerged with similar thematic focus. Standards (such as SAML), open source products (such as Shibboleth and SimpleSAMLphp) and services (most importantly, research and education identity federations5) have all been established during the 2000’s, enabling relying services to make use of the researchers’ home university accounts for login.\n\nThe ELIXIR AAI development started in 2014 as one of the ELIXIR Compute platform key components based on a generic use cases that cover most ELIXIR services. ELIXIR AAI intends to serve the needs of the various ELIXIR communities’ services operated by the ELIXIR Nodes in different countries. Some of the communities (such as the plant and marine) have moderate security needs whereas others (such as the sensitive human data and the rare disease) have strict requirements for authentication and authorisation of the researchers.\n\nThis paper describes the key concepts of ELIXIR AAI and its uptake in the community during the first two years of operation. The work describes the requirements the authors have collected for the ELIXIR AAI based on use case analysis, introduces the ELIXIR AAI design to meet the requirements, provides an overview of the ELIXIR AAI policies to complement the design, and describes technology use by selected early adopters. ELIXIR AAI uptake is advancing rapidly. Therefore, our future work will focus on the sustainable expansion of the services into the broader life science service provider community.\n\n\n2. Requirements on ELIXIR AAI\n\nThis section presents the requirements that have driven the design of the ELIXIR AAI. The requirements were drawn from the analysis of the use cases collected in autumn 2014.\n\nResearchers within Europe and beyond can register an ELIXIR identity to access the relying services. The relying services identify each user by an ELIXIR identifier which is an opaque non-reassignable string. When a human-readable presentation of the ELIXIR identity is needed for instance in the user interface, a parallel ELIXIR username can be used instead. The ELIXIR identifier and the ELIXIR username are exemplified in Table 1 below.\n\nThe intention is that a person registers one ELIXIR identity and uses that throughout their career, just updating their affiliation information (see the next section) on an affiliation change. Although there is no practical means to prevent researchers from registering several ELIXIR identities, it is believed that having several ELIXIR identities would be confusing especially for the users themselves.\n\nThere is no password associated to an ELIXIR identity. Instead, the registration process requires the user to link to one or more of their existing academic or commercial accounts which are then used for login. Following external accounts can be associated to an ELIXIR identity:\n\nResearcher’s account in their home university or research institution provided it has an Identity Provider server available via the eduGAIN interfederation service. This approach is preferred because the home university is able to deliver also the fresh affiliation information for a researcher\n\nORCID identity as provided the ORCID community\n\nThe commercial identity providers Google and LinkedIn.\n\nFor accountability reasons, the ELIXIR identity represents a single natural person. ELIXIR AAI does not accept shared accounts or accounts representing a specific role (e.g. a position like “on-duty operations manager” that circulates among a pool of persons). This is excluded in the Acceptable Usage Policy that the users need to commit to when they register an ELIXIR identity and, when possible, by technical checks on the external accounts.\n\nELIXIR AAI enriches the ELIXIR identities with extra attributes to serve the authorisation needs of the relying services. This section describes the key attributes.\n\n2.2.1. User’s home organisation(s). In the ELIXIR research infrastructure, users are typically affiliated with an organisation (called a home organisation), such as, a university, research institution or a private company and the permissions to access services is often coupled to their continuing affiliation. The current affiliation attribute is a multivalued attribute that indicates the organisation(s) the end user is currently affiliated with and the type of the affiliation. A user can have several home organisations at the same time if they for example work or study in several organisations in parallel. The relying services who decide the user’s access rights based on their affiliation can monitor changes in the attribute.\n\nA user can have three types of affiliations with their home organisation as described in Table 2 below. Faculty affiliation type is believed to most accurately represent the person’s role as a researcher in the home organisation, although the exact definition leaves some institutional freedom. Member affiliation covers also other kind of Home organisation members, such as staff and students and affiliate the rest. The syntax and semantic of the attribute follows the eduPersonScopedAffiliation attribute defined in eduPerson schema6.\n\nThe preferred way to populate the affiliation attribute is to retrieve it programmatically from the home organisation using the eduGAIN service (see section 2.1). There are also other ways described in the table for those home organisations who do not support programmatic access. The validity of the attribute values is guaranteed by asking the user to refresh the value every 12 months using the procedure described above.\n\n2.2.2. Group membership. In many relying services, users’ access rights are coupled to their membership in a group associated with the service. For instance, if a project is assigned a resource quota in a computing cluster, any member of the project group should be able to log in and initiate computations. In that scenario, the resource quota would be assigned to a group and any member of the group can consume the shared resource allocation.\n\nELIXIR AAI has a service for managing ELIXIR users’ group memberships and roles in the groups they belong to. Management of groups is done using a web interface or an API.\n\nEach user can belong to one or multiple groups simultaneously. This is represented by the user having a “member” role in the group. A group member can have also arbitrary extra roles in the group, such as “secretary” or “chair”.\n\nEach group can have one or several owners (members with a special role “owner”). The creator of a group becomes the initial owner of the group. Owners are able to\n\ndelegate group ownership to other members or groups\n\nmanage the group’s properties (such as name)\n\ninvite group members (requires confirmation by the invited user)\n\nadd group members (no confirmation needed by the invited user)\n\nremove group members\n\nassign and delete additional attributes (roles) for users in the group\n\ndefine and manage registration form and process for the group\n\nThe Global Alliance for Genomics and Health is promoting a tiered access to sensitive human data, which consists of three access levels: public access for data with no access control needs, and registered as well as controlled access levels for data whose access is based on the users’ roles and entitlements.\n\nELIXIR AAI has services that support implementation of the registered and controlled access. These services support the dataset owner’s also in meeting their obligations originating from the General data protection regulation, including security by design, accountability and implementation of appropriate technical and organisational measures to secure the data in the registered and controlled access tier.\n\n2.3.1. Controlled access data. ELIXIR AAI has a service that researchers can use for applying for access rights to sensitive datasets and the dataset owners to approve or rejects the data access applications. The Principal Investigator of a project fills in a data access application on behalf of all project members and commits to the dataset’s license terms. The electronic application is then circulated to the Data Access Committee nominated by the dataset owner as defined in the workflow of the dataset. Once approved, the user’s entitlements to access the datasets are presented to the data archive or computing environment for access control enforcement. The dataset owners relying on the system can dynamically revoke and audit the users’ data access rights. Requirements of the service are presented in detail in 7.\n\n2.3.2. Registered access data. The process described above for the controlled access datasets is laboursome both for the applicant and the dataset owners and the application review process may introduce an unnecessary delay to the researcher’s data access.\n\nTo reduce the administrative burden, the Global Alliance for Genomics and Health has proposed that datasets whose privacy impact is limited could deploy a more light-weight schema where the researcher needs to just demonstrate they are a researcher in good standing and commit to the general terms of the registered access. Once done, the researchers could access any registered access datasets or services without further applications. An example of such services are the ELIXIR Beacons8. The requirements of registered access are further described in 9.\n\nELIXIR AAI supports registered access by providing a mechanism for the researcher to demonstrate their bona fide status and make the necessary attestations to confirm that they are committed to the terms of registered access data. As a result, the users’ ELIXIR identity is amended with an extra attribute declaring that they have a status as a bona fide researcher.\n\nSome of the relying services, especially those related to sensitive human data, expect not just fine-grained authorisation but also authentication which relies on multiple authentication factors (something you know, something you have, something you are).\n\nAuthentication to ELIXIR AAI is carried out by the external authentication providers the user has linked to their ELIXIR identities. Typically, those providers offer just password based authentication which is, despite its wide use, vulnerable to various attacks.\n\nELIXIR AAI offers a step-up authentication service for multi-factor authentication. The user is first authenticated by their external authentication provider and subsequently by a second authentication factor delivered by ELIXIR AAI.\n\n2.5.1. Authentication. Towards the relying services, ELIXIR AAI supports SAML 2.0 and OpenID Connect protocols for user authentication and delivery of the attributes described above. The relying service can also use a mechanism to request a step-up authentication.\n\nAlthough OpenID Connect and its underlying OAuth2 protocol can be extended also to non-web services, such as native apps, they are mostly limited to an environment where the user has a web browser to access or launch the service. Services such as data transfer may require authentication using X.509 certificates. ELIXIR AAI also supports credential translation where a user can receive a X.509 certificate after a successful authentication on ELIXIR AAI.\n\n2.5.2. Attribute push. In the basic scenario, the Relying party receives the necessary attributes from ELIXIR AAI using the SAML 2.0 or OpenID Connect protocols when the user logs in as described above. However, some services need a batch synchronisation of users’ attributes although they do not log in to the service. Examples of such services are mailing list services (becoming a group member subscribes them to an associated mailing list service) or IaaS cloud middleware (closing a project shuts down all associated virtual machines).\n\nELIXIR AAI serves those Relying services with batch based attribute synchronisation when needed.\n\n\n3. ELIXIR AAI design\n\nThis section describes the design of the ELIXIR AAI that covers the requirements presented in the previous section. All the components are implemented using open source software.\n\nFigure 1 below presents an overview of the ELIXIR AAI services. The components are described in detail in the next sections.\n\nIn the upper part of the figure are the services relying on the ELIXIR AAI for authentication and authorisation of their users. ELIXIR AAI can support simple tools and services such as the wikis or mailing lists with moderate authentication and authorisation requirements, as well as more complex relying services, such as archives of sensitive data, scientific workflow systems and IaaS clouds for processing the data.\n\nThe external authentication providers are down in the figure. eduGAIN, Google, ORCID and Linkedin are supported. A user can decide which one to use during login process.\n\nThe middle part of the figure presents the ELIXIR AAI service components. The ELIXIR AAI services for authentication are in the left hand side. The ELIXIR AAI services for decorating the authenticated identities with extra attributes for authorisation are in the right hand side.\n\nWhen a user has authenticated using an external authentication provider they first land on the ELIXIR proxy identity provider (IdP). The proxy IdP acts as a relying service towards the external authentication providers and as an Identity Provider towards the relying services and other AAI components. If the proxy IdP observes a new user it triggers automatically a user registration sequence during which the ELIXIR identifiers (see section 2.1) are assigned to them.\n\nThe proxy IdP together with the Perun system10 is responsible for account linking, which tries to avoid a user registering multiple ELIXIR identities by performing heuristics. When the system observes similarity between an existing and a new user account it proposes the user to link the external account to their existing ELIXIR identity, instead. The linking is done by asking the user to log in using the new account and, subsequently, using the existing account.\n\nThe proxy IdP is based on the SimpleSAMLphp product and is replicated geographically for high availability. The OpenID Connect integration to relying services is done using the MITREid product.\n\nIf the relying service requires multi-factor authentication, the user will be redirected to the step-up authentication service that performs a subsequent extra authentication using a second authentication factor.\n\nIn the current deployment, the step-up authentication relies on the Time-based One-Time Password (TOTP) standard11 and a smartphone app that the user needs to install in their smartphone and register to ELIXIR AAI. Once registered, the TOTP app delivers a 6 digit one-time password that is valid for one minute. The user needs to activate their smartphone app and type the one-time password the app gives to the Step-up authentication service’s web page before it expires.\n\nThe registration of the second authentication factor to the correct ELIXIR identity is based on delivering an SMS to the user’s registered cellphone number. ELIXIR is deploying a process to record the ELIXIR identity holder’s trusted cellphone number.\n\nThe service is based on a TOTP server side implementation made at ELIXIR-Finland12. In their smartphones the users can use any standard TOTP app available in the appstores, such as, FreeOTP or Google authenticator.\n\nThe credential translation component currently supports delivering X.509 certificates to the users. The component asks the user to log in using the Proxy IdP and issues them a certificate. The deployment is based on the CILogon software and RCAuth service pilot delivered by the AARC project and is currently operated by NIKHEF.\n\nThe X.509 issued credentials are currently used for gridFTP and access to some IaaS cloud middleware services.\n\nELIXIR directory is an abstract container for storing all ELIXIR identities and attributes. It serves as a common integration point for services that manage user, group and service attributes and authorisations in ELIXIR AAI. It is a service component internal to ELIXIR AAI, however it has API which can be used by relying services to obtain additional data online.\n\nAccess to all the data stored in ELIXIR directory is controlled. Majority of ELIXIR directory services is managed by Perun10 system.\n\nA user can self-manage some of their attributes in the ELIXIR directory, such as their e-mail address. ELIXIR AAI provides the users also a control panel where they can manage the external authentication providers linked to their ELIXIR identity and review all the data stored about them in ELIXIR directory.\n\nIn the bona fide role management service the user can register their bona fide status and make the attestations as described in section 2. The bona fide status is then stored as an extra attribute for the user’s ELIXIR identity and is available for the relying services.\n\nIn the current implementation, the user can demonstrate their researcher status by\n\n- Acquiring a “faculty” attribute from their home organisation (see section 2.2.1), or\n\n- Asking another ELIXIR user who has a “faculty” status in their home organisation to vouch for their bona fide status, or\n\n- matching their ELIXIR identity against the European PubMed archive to demonstrate they are an author of at least one publication.\n\nThe bona fide management service also asks the user to make the attestations9 that are necessary to receive the bona fide status. If several parallel attestations consumed by different relying services emerge, they will be supported, too.\n\nThe functionality is implemented using the REMS7 and Perun10 software.\n\nUsing the group management service of ELIXIR AAI a user can form new groups and manage their members. The groups can be hierarchical and the management of the subgroups can be delegated to other users or groups. Each group can have its own registration process defined. Managers of a group can then gather additional data from users or control how the users are invited into the group. The ELIXIR AAI can expose the group memberships to relying services via various channels like SAML attributes, OpenID Connect claims or directly push them to the service.\n\nELIXIR is currently using the group membership e.g. for managing membership in the projects and related mailing lists to build the ELIXIR services. The group memberships are also consumed by IaaS clouds integrated to ELIXIR. The group membership service is provided by the Perun10 software.\n\nRelying services can register their datasets to the dataset authorisation service and define application forms and related workflows for managing the dataset access rights. The ELIXIR user can then apply for access rights to a dataset and the application is circulated to the related Data Access Committee for approval. Once approved, the ELIXIR AAI uses OAuth2 protocol to expose the access rights to the relying service for access control enforcement.\n\nThe Dataset authorisation service is implemented using the REMS software7.\n\n\n4. ELIXIR AAI policies\n\nThe technical infrastructure introduced in the previous section is complemented by a set of policies relevant for use of the AAI. The policies described below cover the expectations on the behaviour of the end users, relying services and AAI operators.\n\nWhen a user registers an ELIXIR identity they must commit to ELIXIR AAI’s Acceptable Usage Policy (AUP)13. The AUP regulates how and for which purposes the relying services can be used and how the authentication credentials must be protected and used. The AUP also informs the user on the processing of their personal data, including the ELIXIR AAI log files.\n\nThe AUP intends to set the basic level of expectations on the behaviour of the users in the relying services and allows ELIXIR to exclude misbehaving users. Individual relying services can complement the AUP by their own terms of service, based on their specific needs.\n\nThe relying services of ELIXIR AAI must support life science research. This requirement allows also services beyond ELIXIR research infrastructure to integrate to the ELIXIR AAI, as long as the services benefit life science research. This is desirable because there are also e-infrastructures and other research infrastructures serving the ELIXIR community.\n\nIn order to integrate to ELIXIR AAI, the relying services need to respect the applicable laws, including the European data protection laws. If the service is established outside the European Economic Area and countries which guarantee adequate protection of personal data, the relying service must guarantee it takes appropriate safeguards to protect the ELIXIR identities it receives from ELIXIR AAI.\n\nELIXIR AAI is operated by the Czech and Finnish ELIXIR Nodes. The ELIXIR AAI policy obligates the AAI operators to take care of the information security and availability of the AAI services professionally. Normally the AAI services are expected to be available 24-by-7.\n\nThe AAI operators must use the personal data only for what is needed for the operations of the AAI. Log files must be used only for administrative, operational, accounting, monitoring and security purposes. The operators can provide statistics on the use of the AAI service but the statistics must not reveal individual users.\n\n\n5. Uptake and outreach\n\nELIXIR AAI became operational in November 2016. It is a service that belongs to the Compute platform of ELIXIR together with the cloud and data transfer services. In the end of 2017, ELIXIR AAI had 1097 users who belong to 124 groups and could log in to 36 relying services. Additionally, 71 relying services were testing their integration to the ELIXIR AAI, including major European e-Infrastructures EGI and EUDAT. An average of 3000 logins were observed in a month.\n\nEarly relying parties of the ELIXIR AAI include internal project management (e.g. Intranet), social media applications (e.g. Virtual coffee room), request tracking systems (e.g. RT or TOPdesk), data discovery (e.g. Beacon), data and sample access (e.g. THL biobank), and cloud computing service (e.g. EGI fedCloud, EMBL-EBI Embassy and de.NBI cloud). This section provides examples on the use of ELIXIR AAI in these relying services to manage users and access rights gives an overview of the possibilities.\n\nELIXIR Intranet is an internal document management service provided from the central ELIXIR Hub to manage documentation of the distributed ELIXIR infrastructure. The intranet relies on ELIXIR AAI for the user and group management. Experts belonging to a particular group like Heads of Nodes or Technical coordination have different access and visibility on the ELIXIR documentation.\n\nELIXIR Virtual Coffee Room is a service run by the Estonian ELIXIR Node to enable experts from ELIXIR communities to interact online. ELIXIR AAI is implemented in the user and group management. For instance, the training and technical experts from ELIXIR Nodes belong to an interest group with their own dedicated group spaces where they can virtually meet even though they are working across the Nodes.\n\nELIXIR Beacons8 are part of the global Beacon network and a GA4GH driver project influencing many of the emerging global standards that ensure interoperability for life-science researchers. The ELIXIR Beacons enable discovery of genomic datasets using a specialised query interface. Once a dataset has been discovered, ELIXIR AAI allows for the user to gain data access rights if they are recognised as a bona fide researcher to access registered access data. Eventually, the data discovery can lead the user to apply for controlled access rights from the data owners.\n\nNational Institute of Health and Welfare THL biobank is part of BBMRI and linked to ELIXIR Finland. THL uses ELIXIR AAI REMS to manage access application to biobank samples from the Finnish population-level cohorts, and datasets created from the samples. THL was the first sensitive data controller in Europe to federate data access authorisations electronically in collaboration with ELIXIR. Electronic data access entitlements coupled to the reliable identification users is part of the national strategy of Finland to comply with the General data protection regulation.\n\nELIXIR Germany’s de.NBI cloud and EMBL-EBI’s Embassy cloud trust ELIXIR AAI to manage user access to their cloud portal. The de.NBI cloud works as a technical integration point between compute and data resources for bioinformatics in Germany. The cloud access authorisation decisions are made by the local compute providers, and the resource allocation is recorded in ELIXIR AAI and pushed to the cloud centers. In the EMBL-EBI Embassy, user access attributes are released from ELIXIR AAI to the cloud provider when a user logs in to the cloud middleware.\n\nThe ELIXIR training portal TeSS relies on ELIXIR AAI, and the ELIXIR AAI team has organised three training events to help the potential relying services to integrate to ELIXIR, including providing ELIXIR AAI training material contents in TeSS. The hands on trainings have focused on helping the participants to install a SAML Service Provider or OpenID Connect Relying Party and configure it to rely on ELIXIR AAI for user authentication and consumption of user attributes in their services.\n\nFurther information and relying service integration instructions are available here. Statistics on the number of logins to relying services is available here.\n\n\n6. Future work\n\nELIXIR is one of the 13 life science research infrastructure recognised by ESFRI 2016 roadmap14. Although the infrastructures are different in their focus and security requirements, many of them share the need for authenticating and authorising their end users. Many of the researchers are also using services from several infrastructures which would be made simpler by a common AAI.\n\nSince 2016, ELIXIR has worked together with the other infrastructures in CORBEL, the ESFRI cluster project for life sciences. CORBEL has collected AAI use cases from the participating research infrastructures and compiled them to a requirements specification for a common Life Science AAI15. The requirements specification has been used as the basis for a Life Science AAI pilot in the context of AARC2 project. Funding has been applied for deploying the Life Science AAI into production.\n\nThe Global Alliance for Genomics and Health (GA4GH) is a policy-framing and technical standards-setting organization, seeking to enable responsible genomic data sharing within a human rights framework. Its data use & researcher identities (DURI) workstream facilitates and enables the harmonisation of researcher identities by defining who is a bona fide researcher and one or more identity providers that respects this definition and can provide a portable electronic identity. Compatibility with the DURI deliverables, such as the library card concept16, is a future work item for ELIXIR AAI.\n\n\n7. Conclusions\n\nELIXIR has developed a comprehensive AAI service portfolio for authentication of researchers and assisting the relying services to decide what the researchers are permitted to do in the service. The service belongs to the ELIXIR Compute platform and is deployed using open source components. The ELIXIR AAI is available for relying services that support life science research.\n\nThis paper described the requirements on the AAI service and how they have been compiled to a design of the ELIXIR AAI. The related policies were shortly described to complement the technical perspectives. The ELIXIR AAI has been in production since late 2016 and ELIXIR is inviting new services to integrate to it.\n\n\nData availability\n\nNo data are associated with this article\n\n\nSoftware availability\n\nThe ELIXIR AAI homepage with guides of how to connect to service and as well as policy documents is available here: https://www.elixir-europe.org/services/compute/aai\n\nCESNET/Perun\n\nSource code available from:https://github.com/CESNET/perun/tree/v3.1.0\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.130887417\n\nCESNET/Perun-mitreid\n\nSource code available from:https://github.com/CESNET/perun-mitreid/tree/v1.11.0\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.129981018\n\nCESNET/Perun-services\n\nSource code available from:https://github.com/CESNET/perun-services/tree/v3.1.0\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.130030019\n\nCESNET/Perun-simplesamlphp-module\n\nSource code available from:https://github.com/CESNET/perun-simplesamlphp-module/tree/elixir\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.130076920\n\nCESNET/Perunauthorize-simplesamlphp-module\n\nSource code available from: https://github.com/CESNET/perunauthorize-simplesamlphp-module/tree/v1.0.0\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.130076521\n\nCESNET/Proxystatistics-simplesamlphp-module\n\nSource code available from:https://github.com/CESNET/proxystatistics-simplesamlphp-module/tree/v1.1.0\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.130076122\n\nCESNET/Simplesamlphp\n\nSource code available from:https://github.com/CESNET/simplesamlphp/tree/v1.15.4\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.129840023\n\nCSC/Rems\n\nSource code available from: https://github.com/CSCfi/rems/tree/v2.1\n\nArchieved source code at time of publication: http://doi.org/10.5281/zenodo.129733624\n\nCESNET/Perun-mitreid is licenced under the Apache-2.0 licence, other components implemented by CESNET under the BSD-2-clause licence and the component by CSC under the MIT licence.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe projects receive funding from the European Union’s Horizon 2020 research and innovation programme [676559, 654248 and 730941]. CSC - IT Center Finland acknowledges Academy of Finland grant [292265] for ELIXIR Finland.\n\n\nAcknowledgments\n\nThe authors wish to acknowledge the ELIXIR implementation study 2017 and 2018 and the ELIXIR EXCELERATE project which are used for funding the development and operations of the ELIXIR AAI. The work was also inspired by the work in CORBEL and AARC2 projects.\n\n\nReferences\n\nRegulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation). Reference Source\n\nUniversity of Helsinki. Finngen project.\n\nAARC project: Recommandations on minimal assurance level relevant for low-risk research use cases. 2015. Reference Source\n\nDroz S, Graf C, Hassenstein G, et al.: Concept for an Electronic Academic Community in Switzerland and the creation of a Common Authentication and Authorization Infrastructure (AAI) for the Swiss Higher Education System. 2001. Reference Source\n\nREFEDS - The Voice of Research and Education Identity Federations. Reference Source\n\nInternet2: eduPerson Object Class Specification (201602). 2016. Reference Source\n\nLinden M, Nyrönen T, Lappalainen I: Resource Entitlement Management System. Selected papers of TNC2013 conference. Reference Source\n\nScollen S, Rambla J, Linden M, et al.: ELIXIR Beacon. 15th European Conference on Computational Biology (ECCB). 2016. Reference Source\n\nDyke SO, Kirby E, Shabani M, et al.: Registered access: a ‘Triple-A’ approach. Eur J Hum Genet. 2016; 24(12): 1676–1680. PubMed Abstract | Publisher Full Text | Free Full Text\n\nProchazka M, Licehammer S, Matyska L: Perun – Modern Approach for User and Service Management. IST-Africa Conference. 2014. Reference Source\n\nM'Raihi D, Machani S, Pei M, et al.: TOTP: Time-Based One-Time Password Algorithm. RFC 6238. 2011. Publisher Full Text\n\nTuomi A: Haka MFA. Reference Source\n\nELIXIR: Acceptable Usage Policy and Conditions of Use. 2016. Reference Source\n\nEuropean Strategy Forum on Research Infrastructures: Strategy Report on Research Infrastructures. Roadmap. 2016. Reference Source\n\nLinden M, Holub P, Lappalainen I, et al.: Common Authentication and Authorisation service for Life Science Research. TNC18 conference. Reference Source\n\nCabili MN, Carey K, Dyke SOM, et al.: Simplifying research access to genomics and health data with Library Cards. Sci Data. 2018; 5: 180039. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZlámal P, Stava M, Licehammer S, et al.: CESNET/perun: Release 3.1.0 (Version v3.1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1308874\n\nKuba M, Bučík DF: CESNET/perun-mitreid: Version 1.11 (Version v1.11.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1299810\n\nZlámal P, Stava M, Licehammer S, et al.: CESNET/perun-services: Release 3.1.0 (Version v3.1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1300300\n\nProchazka M, Vyskočil P: CESNET/perun-simplesamlphp-module: v1.0.0 (Version elixir). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1300769\n\nVyskočil P, Prochazka M: CESNET/perunauthorize-simplesamlphp-module: v1.0.0 (Version v1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1300765\n\nVyskočil P, Prochazka M: CESNET/proxystatistics-simplesamlphp-module: v1.1.0 (Version v1.1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1300761\n\nMorken O, Crespo JP, Solberg AÅ, et al.: CESNET/simplesamlphp: v1.15.4 (Version v1.15.4). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1298400\n\nJalkanen T, Rontu M, Kaasinen J, et al.: CSCfi/rems: Otaniementie (Version v2.1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1297336"
}
|
[
{
"id": "36859",
"date": "20 Aug 2018",
"name": "Bengt Persson",
"expertise": [
"Reviewer Expertise Bioinformatics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 article giving a clear presentation of the ELIXIR infrastructure for researcher authentication and authorisation (AAI). The AAI service enables for a life science researcher to obtain an ELIXIR identity that can be used for getting access to restricted ELIXIR services. The authors describe the AAI service systematically and thoroughly, including the rationale behind and the initial requirements, the AAI design and the different components of the system, AAI policies, and examples of applications. Links to software repositories and appropriate references are provided. In conclusion, this is an excellent contribution to the F1000Research.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "36861",
"date": "22 Aug 2018",
"name": "Dixie Baker",
"expertise": [
"Reviewer Expertise Information security technology",
"genomic research data protection"
],
"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 broad, multinational applicability and utility of the ELIXIR AAI, this article is of high value, and I recommend its acceptance with some modifications.\n\nRegarding explanation of the \"method,\" although Section 2 is titled \"Requirements for ELIXIR AAI,\" which begins \"This section presents the requirements that have driven the design of the ELIXIR AAI,\" in actuality, no \"requirements\" are articulated. Instead, the section describes the ELIXIR AAI system as implemented, so the reader is left to surmise what the \"requirements\" might be. Also, as described in section 2.1 and reinforced throughout, ELIXIR seems more akin to a broker of identities authenticated by multiple IdPs than an identity provider.\n\nWith respect to providing sufficient detail to enable replication, without clearly articulated requirements, it would be difficult for anyone to replicate this implementation. For example, Section 2.3 (within the \"Requirements\" section begins by explaining that the Global Alliance for Genomics and Health is \"promoting a tiered access to sensitive human data...\" Does this mean that anyone attempting to replicate this work must join with GA4GH in promoting this access model?\n\nRegarding whether the \"conclusions about the method and its performance adequately supported by the findings,\" this article does not present any conclusions or performance data.\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? No\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? No",
"responses": []
},
{
"id": "36860",
"date": "31 Aug 2018",
"name": "Nicolas Liampotis",
"expertise": [
"Reviewer Expertise federated identity and access management"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper describes the ELIXIR Authentication and Authorisation Infrastructure (AAI) which enables researchers to access ELIXIR services and resources using a single set of credentials.\n\nThe authors start by presenting the requirements extracted from the analysis of the use cases that were collected during the design phase of the ELIXIR AAI. The different requirements are described in detail and cover aspects related to the ELIXIR identity and its attributes, the authorised access to data, the need for stepping up the authentication strength via multi-factor authentication and the interfacing with service providers based on standard technologies such as SAML and OpenID Connect.\nThe authors then describe the key features of the ELIXIR AAI design that was driven from the analysed requirements. Following the presentation of the design, the paper discusses the set of policies pertaining to the end users, relying services and AAI operators.\nThe paper also provides information related to the uptake of the ELIXIR AAI during its first two years of operation in production.\nThe inclusion of links to the ELIXIR AAI documentation and the relevant software repositories is particularly useful for implementers/operators of AAI solutions for research collaboration.\nPlease consider the following edits:\nPage 4. \"Instead, the registration process requires the user to link to one\": remove \"to\" Page 4. ORCID identity as provided * the ORCID community: insert \"by\"\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "36862",
"date": "06 Sep 2018",
"name": "Marcus Hardt",
"expertise": [
"Reviewer Expertise Architectures in Federated Identity Management",
"Federated Research Infrastructures"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 an overview on the ELIXIR AAI. It is written on a very high level of abstraction.\n\nThe described system integrates services deployed in many different countries and organisations on one side with federated users on the other side. The requirements that lead to developing the system are outlined, all the important technical components are described and justified with respect to the requirements. The section about the policies that describes the duties of the involved parties and their legal / organisational interplay amongst each other is very short, and therefore incomplete. In the uptake section the authors prove their success with the number of registered users and the impressive number of available services.\nThe strength of this article is its high level of abstraction. This allows non-technical people to get an overview of the developed method. The downside of this is that many technically hard points that had to be solved were not able to be mentioned.\nFrom my perspective a few points deserve additional work:\nPolicies: You are following the AARC policy starter pack. I think the article would benefit, if you even if briefly - mentioned all the relevant policies in the field. Trust: You mention services that accept users authenticated elsewhere in the world. How was it possible to convince these services to trust other identity providers and/or group-managers in foreign countries that a given user is in fact the one he claims to be. Finally, a lessons learned section will help others to reproduce the presented method.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? No\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1199
|
https://f1000research.com/articles/6-618/v1
|
03 May 17
|
{
"type": "Research Article",
"title": "De novo whole-genome assembly of a wild type yeast isolate using nanopore sequencing",
"authors": [
"Hans J. Jansen",
"Ron P. Dirks",
"Michael Liem",
"Christiaan V. Henkel",
"G. Paul H. van Heusden",
"Richard J.L.F. Lemmers",
"Trifa Omer",
"Shuai Shao",
"Peter J. Punt",
"Herman P. Spaink",
"Hans J. Jansen",
"Ron P. Dirks",
"Christiaan V. Henkel",
"G. Paul H. van Heusden",
"Richard J.L.F. Lemmers",
"Trifa Omer",
"Shuai Shao",
"Peter J. Punt",
"Herman P. Spaink"
],
"abstract": "Background: The introduction of the MinIONTM sequencing device by Oxford Nanopore Technologies may greatly accelerate whole genome sequencing. It has been shown that the nanopore sequence data, in combination with other sequencing technologies, is highly useful for accurate annotation of all genes in the genome. However, it also offers great potential for de novo assembly of complex genomes without using other technologies. In this manuscript we used nanopore sequencing as a tool to classify yeast strains. Methods: We compared various technical and software developments for the nanopore sequencing protocol, showing that the R9 chemistry is, as predicted, higher in quality than R7.3 chemistry. The R9 chemistry is an essential improvement for assembly of the extremely AT-rich mitochondrial genome. Results: In this study, we used this new technology to sequence and de novo assemble the genome of a recently isolated ethanologenic yeast strain, and compared the results with those obtained by classical Illumina short read sequencing. This strain was originally named Candida vartiovaarae (Torulopsis vartiovaarae) based on ribosomal RNA sequencing. We show that the assembly using nanopore data is much more contiguous than the assembly using short read data. Conclusions: The mitochondrial and chromosomal genome sequences showed that our strain is clearly distinct from other yeast taxons and most closely related to published Cyberlindnera species. In conclusion, MinION-mediated long read sequencing can be used for high quality de novo assembly of new eukaryotic microbial genomes.",
"keywords": [
"Nanopore sequencing",
"de novo genome assembly",
"wild type yeasts",
"ethanologenic",
"Candida",
"Cyberlindera"
],
"content": "Introduction\n\nWith the development of robust second generation bioethanol processes, next to the use of highly engineered Saccharomyces cerevisiae strains1,2, non-classical ethanologenic yeasts are also being considered as production organisms3,4. In particular, aspects concerning the ability to use both C6 and C5 C-sources and feedstock derived inhibitor resistance have been identified as important for the industrial applicability of different production hosts3. In our previous studies we have identified a novel ethanologenic yeast, Wickerhamomyces anomala, as a potential candidate3. Based on this research, a further screen for alternative yeast species was initiated (Punt and Omer, unpublished study) Here we describe the isolation and genomic characterization of one of these new isolates, which was typed as Candida vartiovaarae based on ribosomal RNA analysis.\n\nWith the arrival of next generation sequencing and the assemblers that can use this type of sequencing data, whole genome shotgun sequencing of completely novel organisms has become affordable and accessible. As a result, a wealth of genomic information has become available to the scientific community leading to many important discoveries. While generating whole draft genomes has become accessible, these genomes are often fragmented due to the nature of these short read technologies5. Assembling the short read data into large contigs proved to be difficult because the short reads do not contain the information to span repeated structures in the genome. Approaches to sequence the ends of larger fragments partially mitigated this problem6.\n\nThe new long read platforms from Pacific Biosciences and Oxford Nanopore Technologies made it possible to obtain reads that span many kilobases7. Assemblies using this type of data are often more contiguous than assemblies based on short read data8,9.\n\nWe have employed the Oxford Nanopore Technologies MinIONTM device to sequence genomic DNA from the isolated Candida vartiovaarae strain. The same DNA was also used to prepare a paired end library for sequencing on the Illumina HiSeq2500. The sequence data were used in various assemblers to obtain the best assemblies.\n\n\nMaterials and methods\n\nIn our previous research3, a screening approach was developed to select for potential ethanologens using selective growth on industrial feedstock hydrolysates. Based on this approach, a previously identified microflora from grass silage was screened for growth on different hydrolysates from both woody and cereal residues. From this microflora, a strain was isolated (DDNA#1) after selection on a growth medium consisting of 10% acid-pretreated corn stover hydrolysate, which was shown to be most restrictive in growth due to the presence of relatively high amounts of furanic inhibitors.\n\nCells were grown at 30°C on plates with YNB (without amino acids) medium supplemented with 0.5% glucose. Cells were scraped from plates and resuspended in 5 ml TE. High MW chromosomal DNA was isolated from yeast isolate DNA#1 and Saccharomyces cerevisiae S288C using a Genomic-tip 100/G column, according to the manufacturer’s instructions (Qiagen).\n\nTo isolate intact chromosomal DNA from DDNA#1, a BioRad CHEF Genomic DNA Plug Kit was used. Briefly, yeast cells were treated with lyticase and the resulting spheroplasts were embedded in low melting point agarose. After incubation with RNase A and Proteinase K, the agarose plugs were thoroughly washed in TE. The DNA in the agarose plugs was separated on a 0.88% agarose gel in 1xTAE buffer on a Bio-Rad CHEF DRII system. The DNA was separated in four subsequent 12 hour runs at 3V/cm; run one and two used a constant switching time of 500 seconds, and in run three and four the switching time increased from 60 seconds to 120 seconds. The gel was afterwards stained with ethidium bromide and imaged.\n\nHigh molecular weight DNA from both DDNA#1 and Saccharomyces cerevisiae S288C was sheared using a nebulizer (Life Technologies). The sheared DNA was used to make genomic DNA libraries using the TruseqTM DNA sample preparation kit, according to the manufacturer’s instructions (Illumina Inc.). In the size selection step, a band of 330–350 bp was cut out of the gel to obtain an insert length of ~270 bp. From the resulting libraries, 4.5 million fragments were sequenced in paired end reads with a read length of 150 nt on an Illumina HiSeq2500, according to the manufacturer’s instructions. The HiSeq control software (HCS) and real time analysis (RTA) software, versions were 2.2.38 and 1.18.61, respectively, were used.\n\nThe genomic DNA was sequenced using nanopore sequencing technology. First the DNA was sequenced on R7.3 Flow Cells. Subsequently, multiple R9 and R9.4 Flow Cells were used to sequence the DNA. For R7.3 sequencing runs, we prepared the library using the SQK-MAP006 kit from Oxford Nanopore Technologies. In short, high molecular weight DNA was sheared with a g-TUBE (Covaris) to an average fragment length of 20 kbp. The sheared DNA was repaired using the FFPE Repair Mix, according to the manufacturer’s instructions (New England Biolabs). After cleaning the DNA with using an extraction process, using a ratio of 0.4:1 Ampure XP beads (Beckman Coulter) to DNA, the DNA ends were polished and an A overhang was added with the NEBNext End Prep Module (New England Biolabs). Then, prior to ligation, the DNA was again cleaned with an extraction using a ratio of 1:1 Ampure XP beads to DNA. The adaptor and hairpin adapter were ligated using Blunt/TA Ligase Master Mix (New England Biolabs). The final library was prepared by cleaning the ligation mix using MyOne C1 beads (Invitrogen).\n\nTo prepare 2D libraries for R9 sequencing runs, we used the SQK-NSK007 kit from Oxford Nanopore Technologies. The procedure to prepare a library with this kit is largely the same as with the SQK-MAP006 kit. 1D library preparation was done with the SQK-RAD001 kit from Oxford Nanopore Technologies. In short, high molecular weight DNA was tagmented with a transposase. The final library was prepared by ligation of the sequencing adapters to the tagmented fragments using the Blunt/TA Ligase Master Mix (New England Biolabs).\n\nThe prepared libraries were loaded on the MinION flow cell, which was docked on the MinION device. The MinKNOW software (version 0.50.2.15 for SQK-MAP006 libraries and version 1.0.5 for SQK-NSK007 and SQK-RAD001 libraries) was used to control the sequencing process and the read files were uploaded to the cloud based Metrichor EPI2ME platform for base calling. Base called reads were downloaded for further processing and assembly.\n\nThe sequence data from the Illumina platform was assembled using the Spades assembler (version 3.6.0), either alone or in combination with the nanopore data.\n\nFrom the base called read files produced by the Metrichor EPI2ME platform, a sequence file in fasta format was extracted using the R-package poRe v0.1710. For the assembly of the nanopore data, Canu v1.3 was used11. After assembly, the resulting contigs were polished with the short read data using PILON v1.1812. The sequencing data has been submitted to the European Nucleotide Archive and can be accessed at http://www.ebi.ac.uk/ena/data/view/PRJEB19912.\n\nA k-mer count analysis was done using Jellyfish (version 2.2.6)13 on the Illumina data. From the paired end reads, only the first read was truncated to 100 bp to avoid the lower quality part of the read. The second read was omitted from this analysis to avoid counting overlapping k-mers. Different k-mer sizes were used ranging from k=17 to 23. After converting the k-mer counts into a histogram format, this file was analyzed using the Genomescope tool, available at http://qb.cshl.edu/genomescope/ and https://github.com/schatzlab/genomescope.\n\nFrom 26S ribosomal RNA sequences available in the nucleotide database, Chen et al.14 have constructed a phylogenetic tree. The closest relative for which whole genome sequences are available is Cyberlindnera jadinii. To compare our draft genome assembly to this yeast species, we retrieved assemblies of two Cyberlindnera jadinii strains, namely NBRC 0988 (GenBank accession number, DG000077.1) and CBS1600 (GenBank accession number, CDQK00000000.1). We also used Saccharomyce cerevisiae S288C (GenBank accession number, GCA_000146045.2) in this comparison. We aligned those assemblies to the corrected draft assembly of our strain using MUMmer’s alignment generator NUCmer (version 3.1)15. NUCmer’s output was filtered with delta-filter, and the filtered results parsed to MUMmerplot, generating full-genome visualization between the pairs of different yeast species.\n\nReads generated on the Illumina platform were aligned to the published Candida vartiovaarae mitochondrial genome (Genbank accession number, KC993190.1) using Bowtie2 (version 2.2.5). Reads generated on the MinION platform were aligned using BWA-mem (version 0.7.15) with -x ont2d settings. Resulting bam files were sorted and viewed in IGV viewer (version 2.3).\n\n\nResults and discussion\n\nFrom a screen on 10% acid-pretreated corn stover hydrolysate, about 70 individual clones were obtained, only five of which were able to grow well on purely synthetic YNB-based medium. To determine the taxonomic status of these clones, chromosomal DNA was isolated and used for PCR amplification of the ribosomal ITS sequence using ITS specific primers (ITS1 and ITS416).\n\nBLAST analysis of these ITS sequences of all 5 isolates revealed a 100% identity to Candida vartiovaarae (Torulopsis vartiovaarae: NCBI accession number KY102493)\n\nAll five isolates were grown on different C-sources and showed growth on glucose, mannose, cellobiose, xylose and glycerol, while growth on L-arabinose was variable. No significant growth was found on galactose and rhamnose. Good growth (on glucose) occurred between 20–30°C, at pH3-7 (optimum 25°C, pH4-5). Based on the results, we concluded that all five isolates originated from a single source in the grass silage sample. Subsequent experiments were therefore carried out with a single isolate now named DDNA#1.\n\nWe took three approaches to assemble the genome of DDNA#1. The first approach used only short reads produced by the Illumina platform. After merging the paired end reads we obtained 1.08 Gbp of ~240 bp reads. The genome sequence that we obtained using the Spades assembler17 showed a very fragmented assembly that consisted of 14,764 contigs. The N50 of this assembly was only 2.2 kbp, possibly due to a high level of SNPs. We also assembled Saccharomyces cerevisiae S288C using a similar short read dataset that was made and sequenced in parallel. Here we obtained an assembly that consisted of 768 contigs with a longer N50 of 124 kbp. In the second approach, we used the Spades assembler to make a hybrid assembly by combining the short read data set and the corrected long reads that were produced by the Canu assembler11. From the original 2.05 Gbp nanopore sequence data with an average read length of 7.5 kbp, 389 Mbp was left after correction by Canu. This corrected dataset had an average read length of 7.9 kbp. This hybrid assembly consisted of 1904 contigs with an N50 of 255 kbp. As a third approach, we only used the long read data set and let the Canu assembler correct the longest reads with the shorter reads and then attempt an assembly. In this assembly we obtained 61 contigs with a N50 of 455 kbp (Table 1). It is clear from these results that using the long read data set alone produced the most contiguous assembly, as has been shown previously8,9.\n\nWe also used the nanopore datasets made with the R7.3 and R9 chemistry separately in the Canu assembler. The most notable difference between these assemblies is found in the mitochondrial genome. Only 16 kbp of this 33 kbp genome could be assembled with the R7.3 data, whereas the R9 assembly contained the complete mitochondrial genome (NCBI reference sequence, NC_022164.1). The mitochondrial genome has a very low GC content (21%) and in the extragenic regions more A and T homopolymers are found. Very few R7.3 reads mapped to this region, but in the R9 dataset there are many more reads that represent this region (Figure 1). It has been shown that the R7.3 data especially has a bias against A and T homopolymers. This bias is reduced in R9, but not completely absent18,19. Even after correction of the long reads and assembly in Canu the contig sequences still contain errors11. We have used PILON12 and the complementary Illumina data from this strain to correct the assembled contigs. This led to a minor increase in size of the assembly.\n\nReads from both the Illumina, and the nanopore platform were aligned to the Candida vartiovaarae mitochondrial genome (Genbank accession number, KC993190.1) to show the difference in coverage between the different platforms and chemistry versions.\n\nThe Illumina sequence data of our DDNA#1 isolate were submitted to the Genomescope13 software package to analyze the k-mer count distribution, using k-mer size = 19 at an average coverage of 28.0x (Figure 2). The ‘haploid’ genome is predicted to contribute to the most abundant fraction, which corresponds with the second peak (dotted line) in the plot (Figure 2A). The first peak corresponds to sequence occurring exactly half as frequently as the main peak, so these are plausibly haplotypes. Due to the nature of k-mer counting, this peak often appears higher than the main peak, because a single SNP will affect all k-mers overlapping that position. The first two peaks contain about 10 Mbp of sequence. Additional peaks at higher coverage indicate duplications and repetitive DNA that are quite abundant, but correspond with less sequence than the second peak. Genomescope estimated a haploid genome size of between 12.00 and 12.01 Mbp. Additionally, Genomescope revealed 3.6% variety across the entire genome indicating that the genome of C. vartiovaarae has strong heterozygous properties (Figure 2B). A likely possibility is that areas in the genome are replicated and slightly diverged in sequence. This could also explain why we see a large tail of repeated k-mers (Figure 2A). It could also explain why our assembly still remained fragmented despite the relatively large amount of nanopore data that was used in the assembly.\n\nGenomescope attempts to find k-mer count peaks, low and high coverage peaks indicating hetero- and homozygosity. (A) We find ~13× and ~28× coverage for hetero- and homozygous fractions in our dataset. Exact peak positions are determined with a log transformation. Evaluating the slope between coverage points reveals the peak positions indicating hetero- and homozygosity, for lower and higher coverage, respectively. (B) Table showing the most important metrics from this k-mer analysis.\n\nAs a further means to validate our assembled contigs and determine if they match the actual chromosome length, we have separated the chromosomes on an agarose gel using pulsed field gel electrophoresis. The gel image in Figure 3 shows five bands that represent the chromosomes of this yeast strain. The smallest band has a length that corresponds to the length of the mitochondrial genome (33 kbp). Additional fragments of 450, 1200, and 1500 kbp are also found. The intensity of the band that runs above the 2200 kbp marker band suggests that it actually contains more than one distinct fragment. To make the genome size fit to the estimate derived from the assembly and k-mer analysis (~12.5 mbp), three ~3 Mbp chromosomes should be postulated. The uncertainty in chromosome size estimate based on pulsed field electrophoresis gels is high because of the large chromosome size and the fact that it is difficult to determine if more than one fragment is present in the gel at a given position. Our conclusion that the top band represents three or more chromosomes is in agreement with the genome sequences of two related C. jadinii strains, namely CBS1600 and NBRC 0988.\n\nIn lane 1, the chromosomes of Saccharomyces cerevisiae were loaded as a marker. Sizes of the chromosomes in the marker lane are indicated. In lane 2, the chromosomes of Candida vartiovaarae DDNA#1 were loaded.\n\nWe have compared the assembled contigs of our C. vartiovaarae isolate DDNA#1 strain to yeast genome sequences that are already deposited in the nucleotide database. Comparison of our yeast strain with the well characterized S. cerevisiae assembly showed negligible genomic similarity (Figure 4A). From 26S ribosomal RNA sequences available in the nucleotide database, Chen et al.14 have constructed a phylogenetic tree. The closest relatives for which whole genome sequences are available are C. jadinii strains CBS1600 and NBRC 0988. An initial comparison between CBS1600 and NBRC 0988 revealed that these two strains show high homology (Figure 4B). The genomic similarity between our strain and C. jadinii strains CBS1600 and NBRC 0988 is much lower (Figures 4C and D). In conclusion, these data show that wild type yeast strains are very heterogeneous, despite a high similarity based on ribosomal RNA ITS sequences. Therefore, the data suggest that nanopore sequencing is an essential new tool to classify yeast strains. Of course, the nanopore sequence data in combination with other sequencing technologies is highly useful for accurate annotation of all genes in the genome.\n\nDashed lines indicate contigs (start and stop positions) and the area between dashed lines indicates the contig size. Blue and orange dots are hits in reverse and forward orientation, respectively. Diagonal lines indicate sequence and synteny conservation across species. (A) Comparison between Saccharomyces cerevisiae S288c (horizontal axis) and Candida vartiovaarae isolate DDNA#1 (vertical axis). (B) Comparison between Cyberlindnera jadinii strains CBS1600 (horizontal axis) and NBRC 0988 (vertical axis). (C) Comparison between Candida vartiovaarae isolate DDNA#1 (vertical axis) and Cyberlindnera jadinii strain CBS1600 (horizontal axis). (D) Comparison between Candida vartiovaarae isolate DDNA#1 (vertical axis) and Cyberlindnera jadinii strain NBRC 0988 (horizontal axis).",
"appendix": "Author contributions\n\n\n\nHPS conceived the study. PJP, HPS, HJJ, and RPD designed the experiments. HJJ, RJLFL, PvH, TO, and SS performed the experiments. HJJ, ML, and CVH contributed to the data analysis. HJJ, RPD, and HPS 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\nHJJ and CVH are members of the Nanopore Community, and have previously received flow cells free of charge, as well as travel expense reimbursements from Oxford Nanopore Technologies.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nZhang GC, Liu JJ, Kong II, et al.: Combining C6 and C5 sugar metabolism for enhancing microbial bioconversion. Curr Opin Chem Biol. 2015; 29: 49–57. PubMed Abstract | Publisher Full Text\n\nSànchez Nogué V, Karhumaa K: Xylose fermentation as a challenge for commercialization of lignocellulosic fuels and chemicals. Biotechnol Lett. 2015; 37(4): 761–772. PubMed Abstract | Publisher Full Text\n\nZha Y, Hossain AH, Tobola F, et al.: Pichia anomala 29X: a resistant strain for lignocellulosic biomass hydrolysate fermentation. FEMS Yeast Res. 2013; 13(7): 609–617. PubMed Abstract | Publisher Full Text\n\nHarner NK, Wen X, Bajwa PK, et al.: Genetic improvement of native xylose-fermenting yeasts for ethanol production. J Ind Microbiol Biotechnol. 2015; 42(1): 1–20. PubMed Abstract | Publisher Full Text\n\nSimpson JT, Pop M: The theory and practice of genome sequence assembly. Annu Rev Genomics Hum Genet. 2015; 16: 153–172. PubMed Abstract | Publisher Full Text\n\nKoren S, Phillippy AM: One chromosome, one contig: complete microbial genomes from long-read sequencing and assembly. Curr Opin Microbiol. 2015; 23: 110–120. PubMed Abstract | Publisher Full Text\n\nUrban JM, Bliss J, Lawrence CE, et al.: Sequencing ultra-long DNA molecules with the Oxford Nanopore MinION. BioRxiv. 2015. Publisher Full Text\n\nBerlin K, Koren S, Chin CS, et al.: Assembling large genomes with single-molecule sequencing and locality-sensitive hashing. Nat Biotechnol. 2015; 33(6): 623–630. PubMed Abstract | Publisher Full Text\n\nChakraborty M, Baldwin-Brown JG, Long AD, et al.: Contiguous and accurate de novo assembly of metazoan genomes with modest long read coverage. Nucleic Acids Res. 2016; 44(19): e147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWatson M, Thomson M, Risse J, et al.: poRe: an R package for the visualization and analysis of nanopore sequencing data. Bioinformatics. 2015; 31(1): 114–115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoren S, Walenz BP, Berlin K, et al.: Canu: scalable and accurate long-read assembly via adaptive k-mer weighting and repeat separation. BioRxiv. 2016. Publisher Full Text\n\nWalker BJ, Abeel T, Shea T, et al.: Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS One. 2014; 9(11): e112963. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarçais G, Kingsford CA: A Fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics. 2011; 27(6): 764–770. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen B, Huang X, Zheng JW, et al.: Candida mengyuniae sp. nov., a metsulfuron-methyl-resistant yeast. Int J Syst Evol Microbiol. 2009; 59(Pt 5): 1237–1241. PubMed Abstract | Publisher Full Text\n\nKurtz S, Phillippy A, Delcher AL, et al.: Versatile and open software for comparing large genomes. Genome Biol. 2004; 5(2): R12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu J: Fungal DNA barcoding. Genome. 2016; 59(11): 913–932. PubMed Abstract | Publisher Full Text\n\nBankevich A, Nurk S, Antipov D, et al.: SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012; 19(5): 455–477. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIp CL, Loose M, Tyson JR, et al.: MinION Analysis and Reference Consortium: Phase 1 data release and analysis [version 1; referees: 2 approved]. F1000Res. 2015; 4: 1075. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain M, et al.: MinION Analysis and Reference Consortium: Phase 2 data release and analysis of R9.0 chemistry. F1000Research. In Press."
}
|
[
{
"id": "23377",
"date": "27 Jun 2017",
"name": "Mile Šikić",
"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 presented de novo whole-genome assembly of a wild type yeast isolate using nanopore sequencing. They tried three different approaches to assemble the genome: using Illumina reads only, using both Illumina and nanopore reads in a hybrid approach, and using the only nanopore reads for assembling and Illumina reads for polishing. The third approach resulted in the most contiguous assembly. In they work they use nanopore datasets made with R7.3, R9 and R9.4 chemistries.\nAlthough they used a correct procedure for genome assembly it would be interesting to compare their results with the following methods in the third approach:\nUsing minimap+ miniasm assembler in combination with Racon consensus tool and PILON\n\nUsing Canu + racon + PILON\n\nTry to polish nanopore assembly using Nanopolish\n\nIn addition, it would be valuable if they make their data publicly available to enable others to reproduce their results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3802",
"date": "03 Aug 2018",
"name": "Michael Liem",
"role": "Author Response",
"response": "Using minimap+ miniasm assembler in combination with Racon consensus tool and PILON Thank you for your suggestion, this strategy is now included in our study. Using Canu + racon + PILON Since Canu contains an integrated self-correction procedure prior to assembly we have not corrected the Canu contigs with Racon, however the combination Canu – PILON correction is part of our study, thank you. Try to polish nanopore assembly using Nanopolish Thank you for your suggestion, however, since we have combined different data sets from different chemistries and different laboratories, at different times, including filtering of these data, it’s relatively complicated to polish such datasets with Nanopolish. To balance out the effort-result ratio we have performed a double iteration PILON correction which shows to be sufficient to identify the majority of genes stored in the Fungi 0db9 database used by BUSCO. In addition, it would be valuable if they make their data publicly available to enable others to reproduce their results. Data has status in process, should be publicly accessible very soon"
}
]
},
{
"id": "23808",
"date": "07 Jul 2017",
"name": "Jean-Marc Aury",
"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\nWe read the manuscript by Jansen et al. titled “De novo whole-genome assembly of a wild type yeast isolate using Nanopore sequencing” with great interest. Authors describe their strategy to sequence and assemble a yeast strain using different methodologies: a short read strategy with Illumina reads alone and two hybrid approaches, the first one combining both short and long reads for the assembly and the second using long reads for the assembly and short reads for the correction of the consensus. In general, we think that this is a well put together study that reflects the current standard approaches for assembling genomes with both short and long reads. However, we have some questions/remarks that we would like the authors to answer.\nIt seems that the high level of polymorphism complicate the de novo assembly. If some regions are heterozygous, it should lead to a higher than expected assembly size. We think the authors should describe in more details the Illumina-only assembly especially the cumulative size (add a column in Table 1). As the error rate is low, with a high level of SNPs, both (Is the DDNA#1 isolate is a diploid yeast?) haplotypes should be segregated. On the contrary, the assembly length of the nanopore-only assemblies seems to be near the expected size (12Mb), does it mean that the error rate prevent to distinguish haplotypes? We think the authors should discuss in more details how haplotypes are resolved in their different assemblies.\n\nThe whole dataset (reads + final assembly) should be submitted in public repository to ensure full reproducibility.\n\nParagraph Illumina and MinION de novo genome assembly, line 38. Contigs were polished using the Pilon tool but line 7 of the same paragraph, authors indicate that the Spades assembly that was generated from Illumina reads alone was highly fragmented possibly due to a high level of SNPs in the DDNA#1 isolate. I think that to verify if the Pilon correction didn’t do more harm than good, authors could run the Busco tool (http://busco.ezlab.org/) on the assemblies, or annotate genes, before and after correction to verify if it didn’t introduce errors in the consensus due to heterogeneous input reads.\n\nParagraph Illumina and MinION de novo genome assembly, lines 14-15 it is said that the cumulative size of reads that was given as input to Canu was 2.05 Gb and that the corrected reads cumulative output size was equal to 389 Mb. I think that by default Canu only corrects 30X of the input read set (controlled by the corOutCoverage parameter) and since it is relatively close to 30-fold coverage of a yeast genome, I was wondering if authors leaved this parameter as default or if they moved up the limit and it could only correct around 30X of coverage. If this parameter was changed, I think it would be a good idea to indicate it.\n\nAuthors should add a table that contains standard metrics about the sequencing data (nanopore and illumina): number of reads, cumulative size, coverage, average read length…\n\nParagraph Full genome comparison, lines 12-15 it is said that the Nucmer’s ouput was filtered with the delta-filter software; please add the parameters used to filter out alignments. Moreover, if the yeast genomes used for the comparison are highly variable the nucmer software is not the best suited; maybe lastz (https://github.com/lastz/lastz) should better perform.\n\nThe smartdenovo assembler has been successfully applied to yeast genomes (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5466710/), it would be interesting to compare their results with a smartdenovo assembly.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3801",
"date": "03 Aug 2018",
"name": "Michael Liem",
"role": "Author Response",
"response": "1. It seems that the high level of polymorphism complicate the de novo assembly. If some regions are heterozygous, it should lead to a higher than expected assembly size. We think the authors should describe in more details the Illumina-only assembly especially the cumulative size (add a column in Table 1). As the error rate is low, with a high level of SNPs, both (Is the DDNA#1 isolate is a diploid yeast?) haplotypes should be segregated. On the contrary, the assembly length of the nanopore-only assemblies seems to be near the expected size (12Mb), does it mean that the error rate prevent to distinguish haplotypes? We think the authors should discuss in more details how haplotypes are resolved in their different assemblies. Statistical information on the Illumina derived assembly is now added to Table 1. Indeed the majority of assemblies based exclusively on nanopore data are haploid genomes, this comes together with the notion that most of these assembler are designed to reconstruct bacterial genomes. However Canu should be able to differentiate diploid haplotypes, that is for high coverage datasets. It appears 17x coveragehigh quality long length read data is insufficient to resolve the (partial) diploid genome of DDNA#1. 2. The whole dataset (reads + final assembly) should be submitted in public repository to ensure fullreproducibility. These should be publicly available now 3. Paragraph Illumina and MinION de novo genome assembly, line 38. Contigs were polished using the Pilon tool but line 7 of the same paragraph, authors indicate that the Spades assembly that was generated from Illumina reads alone was highly fragmented possibly due to a high level of SNPs in the DDNA#1 isolate. I think that to verify if the Pilon correction didn’t do more harm than good, authors could run the Busco tool (http://busco.ezlab.org/) on the assemblies, or annotate genes, before and after correction to verify if it didn’t introduce errors in the consensus due to heterogeneous input reads. Thank you for your suggestion this is now incorporated into the manuscript under methods/ results and discussion – genome assembly assessment based on gene expectation using BUSCO 4. Paragraph Illumina and MinION de novo genome assembly, lines 14-15 it is said that the cumulative size of reads that was given as input to Canu was 2.05 Gb and that the corrected reads cumulative output size was equal to 389 Mb. I think that by default Canu only corrects 30X of the input read set (controlled by the corOutCoverage parameter) and since it is relatively close to 30-fold coverage of a yeast genome, I was wondering if authors leaved this parameter as default or if they moved up the limit and it could only correct around 30X of coverage. If this parameter was changed, I think it would be a good idea to indicate it. In our Canu version corOutCoverage is set to 40x coverage by default and has not been changed. 5. Authors should add a table that contains standard metrics about the sequencing data (nanoporeand illumina): number of reads, cumulative size, coverage, average read length… Table is now added. 6. Paragraph Full genome comparison, lines 12-15 it is said that the Nucmer’s ouput was filtered with the delta-filter software; please add the parameters used to filter out alignments. Moreover, if the yeast genomes used for the comparison are highly variable the nucmer software is not the best suited; maybe lastz (https://github.com/lastz/lastz) should better perform. Thank you for your suggestion, we have performed similar whole genome alignments with Lastz and mummer, however we didn’t observe a noticeable difference based on whole genome comparison alone. It appears the assembly algorithm and input data characteristics are the major factors that influenced the contiguity and fragmentation of our assemblies. 7. The smartdenovo assembler has been successfully applied to yeast genomes (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5466710/), it would be interesting to compare their results with a smartdenovo assembly. Thank you for your suggestion, Smartdenovo has now been added to the set of assemblers and results are denoted in our manuscript. Indeed Smartdenovo is an assembler that performs relatively well on the dataset of our yeast strain."
}
]
},
{
"id": "24005",
"date": "17 Jul 2017",
"name": "Christina A. Cuomo",
"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 by Jansen et al describes comparison of de novo assemblies generated using Illumina or Oxford Nanopore sequence for the yeast Candida varitovaarae. The sequenced isolate was collected from a screen for new ethanologenic yeast species. Genomic DNA was sequenced using both platforms and de novo assemblies compared for overall metrics and representation of the mitochondrial genome. The final assembly was compared to those of other related yeast species to view conservation of synteny.\nOverall this is an interesting study in showing the advantage of utilizing long Oxford nanopore reads for assembly of a genome that was difficult to assemble using Illumina data. This description would be more compelling if the authors could address a few issues with the presentation of this data.\n1. In addition to genome size, the major factors that can influence the outcome of a de novo assembly are the repetitive sequence content, GC content, and level of heterozygosity. The authors suggest that repetitive sequence could explain large number of contigs; this could be directly addressed by identifying repetitive sequences in the assembly and evaluating contig ends. However there is also the suggestion in the text of some level of heterozygosity, which could better account for the low contig N50 they report in the Illumina assemblies. Whether or not the species is diploid and if so the level of heterozygosity is important to address in evaluating the performance of the two sequencing approaches and documenting the genomes for which long reads are most useful. This could be addressed for example using the Illumina data to identify heterozygous variants across the assembly.\n2. The authors use Pilon to correct the assembled contigs with Illumina data and note that this led to a minor increase in size of the assembly, suggesting there were some misassembled regions in the original Canu assembly. As the other genomes compared using Nucmer are distantly related, with many rearrangements, this could not be used to validate the Canu assembly. It would be helpful if the authors could more fully describe the errors identified and fixed by Pilon.\n\n3. Along the same lines, which statistics are for the final, best version of the assembly? Table 1 compares different combinations of Oxford Chemistry, however the authors also describe an additional step of Pilon polishing. It would be useful to contrast metrics, including sequence coverage levels and GC content, to those from the 2 Spades assemblies, as well as note which assembly is the final version.\n4. In Figure 1, the top scale is too small to read. Plotting the GC as a separate track would be helpful to compare to the R7 coverage level.\n5. For the PFG in Figure 3, a longer run may help separate the bright high MW band into separate chromosomes.\n6. The data does not appear to be submitted to a public repository; both the raw sequence and the final best assembly should be submitted to NCBI or the ENA.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3800",
"date": "03 Aug 2018",
"name": "Michael Liem",
"role": "Author Response",
"response": "1. In addition to genome size, the major factors that can influence the outcome of a de novo assembly are the repetitive sequence content, GC content, and level of heterozygosity. The authors suggest that repetitive sequence could explain large number of contigs; this could be directly addressed by identifying repetitive sequences in the assembly and evaluating contig ends. However there is also the suggestion in the text of some level of heterozygosity, which could better account for the low contig N50 they report in the Illumina assemblies. Whether or not the species is diploid and if so the level of heterozygosity is important to address in evaluating the performance of the two sequencing approaches and documenting the genomes for which long reads are most useful. This could be addressed for example using the Illumina data to identify heterozygous variants across the assembly. The estimated genome size comparison between nanopore mediated assemblies and hybrid Spades assembly is a first indication of the polyploid genome of our strain. Together with the abundant double gene copy BUSCO gene identification analysis and Spades hybrid contigs alignment to TULIP contigs we hope to have shown the diploid characteristics of DDNA#1, at least to partial extend. 2. The authors use Pilon to correct the assembled contigs with Illumina data and note that this led to a minor increase in size of the assembly, suggesting there were some misassembled regions in the original Canu assembly. As the other genomes compared using Nucmer are distantly related, with many rearrangements, this could not be used to validate the Canu assembly. It would be helpful if the authors could more fully describe the errors identified and fixed by Pilon. Increased assembly length after PILON correction is mainly due to corrected homopolymer stretches that are often underrepresented due to sequencing complexities of low complexity regions. This explanation has been added to the manuscript under results and discussion – Illumina and MinION de novo genome assembly. 3. Along the same lines, which statistics are for the final, best version of the assembly? Table 1 compares different combinations of Oxford Chemistry, however the authors also describe an additional step of Pilon polishing. It would be useful to contrast metrics, including sequence coverage levels and GC content, to those from the 2 Spades assemblies, as well as note which assembly is the final version. The final assembly is now described under results and discussion – Illumina and MinION de novo genome assembly. We have added sequence data statistics such as coverage and total amount of data. And aimed to highlight the error correction effect using BUSCO gene identification analysis. 4. In Figure 1, the top scale is too small to read. Plotting the GC as a separate track would be helpful to compare to the R7 coverage level. GC-content is now added to this figure and numbers and text have been made more clear. 5. For the PFG in Figure 3, a longer run may help separate the bright high MW band into separate chromosomes. We have tried many different run conditions and failed to properly resolve the largest bands. This may be different on a different system but we do not have access to such a system. 6. The data does not appear to be submitted to a public repository; both the raw sequence and the final best assembly should be submitted to NCBI or the ENA. Data has status in process, should be publicly accessible very soon"
}
]
},
{
"id": "23807",
"date": "27 Jul 2017",
"name": "Hayan Lee",
"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\nJansen et al. used Oxford Nanopore Technology with other short read sequencing technology, HiSeq 2500, to perform high-quality de novo genome assembly and classify yeast strain isolates, Candida vartiovaarae DDNA#1 from Saccharomyces cerevisiae S288C and Cyberlindrena jadinii CBS1600/NBRC 0988. They also exploited two versions of Nanopore flowcell chemistry and related software. Especially AT-rich mitochondria assembly using R7.3 and R9 comparison is very interesting.\n\nUsing similar short read data, N50 of DDNA#1 is 2.2kbp and that of S277C was 124Kbp. Probably authors want to perform repeat analysis for both strains to further study what makes such a performance gap.\n\nFor assembly approach two and three, authors used Canu to correct Nanopore reads with short reads. So basically all three approaches adopted short reads for correction or assembly purpose. Since Canu can perform self-correction with only long reads, it would be very interesting to compare self-corrected Nanopore reads assembly contiguity vs. short reads corrected Nanopore reads assembly contiguity.\n\nAuthors used two error correction methods; Canu and PILON, It would be helpful to consistently compare the correction performance of two software.\n\nAlthough C. jadinii stains are proposed to be the closest strain, given Figure 4, S288C looks much closer to DDNA#1. Probably authors want to take a close look at this.\n\nAll sequencing data should be available online for reproducibility.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3799",
"date": "03 Aug 2018",
"name": "Michael Liem",
"role": "Author Response",
"response": "Using similar short read data, N50 of DDNA#1 is 2.2kbp and that of S277C was 124Kbp. Probably authors want to perform repeat analysis for both strains to further study what makes such a performance gap. We hope to have shown that the performance gap can be overcome using long reads. Using long reads in either hybrid or with data exclusively from nanopore decreased fragmentation and increased contiguity. Suggesting genomic complexity caused initial difficulties during assembly of this strain. For assembly approach two and three, authors used Canu to correct Nanopore reads with short reads. So basically all three approaches adopted short reads for correction or assembly purpose. Since Canu can perform self-correction with only long reads, it would be very interesting to compare self-corrected Nanopore reads assembly contiguity vs. short reads corrected Nanopore reads assembly contiguity. We have compared Canu (self-corrected) results to assemblies made with Miniasm, TULIP and Smartdenovo corrected with Racon. It appears the assembly strategy is a crucial difference to contiguity and fragmentation as opposed to self- or post-assembly correction. Authors used two error correction methods; Canu and PILON, It would be helpful to consistently compare the correction performance of two software. The comparison between different assemblers and correction procedures should be more consistent now that we have separated the two task more prominently. Although C. jadinii stains are proposed to be the closest strain, given Figure 4, S288C looks much closer to DDNA#1. Probably authors want to take a close look at this. Although alignment hits between C. jadinii and S288C are more targeted towards the diagonal in this figure the alignment length is very short and the number of alignment hits is significantly lower compared to the other two strains. This underlines the poor synteny conservation between C. jadinii and S288C as compared to CBS1600 and NBRC 0988. C. jadinii compared to these two strains show many more alignment hits hence these strains are taken to be more similar. All sequencing data should be available online for reproducibility. Data has status in process, should be publicly accessible very soon"
}
]
}
] | 1
|
https://f1000research.com/articles/6-618
|
https://f1000research.com/articles/7-1189/v1
|
03 Aug 18
|
{
"type": "Research Article",
"title": "Androgen-dependent alternative mRNA isoform expression in prostate cancer cells",
"authors": [
"Jennifer Munkley",
"Teresa M. Maia",
"Nekane Ibarluzea",
"Karen E. Livermore",
"Daniel Vodak",
"Ingrid Ehrmann",
"Katherine James",
"Prabhakar Rajan",
"Nuno L. Barbosa-Morais",
"David J. Elliott",
"Teresa M. Maia",
"Nekane Ibarluzea",
"Karen E. Livermore",
"Daniel Vodak",
"Ingrid Ehrmann",
"Katherine James",
"Prabhakar Rajan",
"Nuno L. Barbosa-Morais",
"David J. Elliott"
],
"abstract": "Background: Androgen steroid hormones are key drivers of prostate cancer. Previous work has shown that androgens can drive the expression of alternative mRNA isoforms as well as transcriptional changes in prostate cancer cells. Yet to what extent androgens control alternative mRNA isoforms and how these are expressed and differentially regulated in prostate tumours is unknown. Methods: Here we have used RNA-Seq data to globally identify alternative mRNA isoform expression under androgen control in prostate cancer cells, and profiled the expression of these mRNA isoforms in clinical tissue. Results: Our data indicate androgens primarily switch mRNA isoforms through alternative promoter selection. We detected 73 androgen regulated alternative transcription events, including utilisation of 56 androgen-dependent alternative promoters, 13 androgen-regulated alternative splicing events, and selection of 4 androgen-regulated alternative 3′ mRNA ends. 64 of these events are novel to this study, and 26 involve previously unannotated isoforms. We validated androgen dependent regulation of 17 alternative isoforms by quantitative PCR in an independent sample set. Some of the identified mRNA isoforms are in genes already implicated in prostate cancer (including LIG4, FDFT1 and RELAXIN), or in genes important in other cancers (e.g. NUP93 and MAT2A). Importantly, analysis of transcriptome data from 497 tumour samples in the TGCA prostate adenocarcinoma (PRAD) cohort identified 13 mRNA isoforms (including TPD52, TACC2 and NDUFV3) that are differentially regulated in localised prostate cancer relative to normal tissue, and 3 (OSBPL1A, CLK3 and TSC22D3) which change significantly with Gleason grade and tumour stage. Conclusions: Our findings dramatically increase the number of known androgen regulated isoforms in prostate cancer, and indicate a highly complex response to androgens in prostate cancer cells that could be clinically important.",
"keywords": [
"Androgens",
"AR",
"prostate cancer",
"alternative splicing",
"alternative promoters",
"alternative 3' ends",
"transcription",
"mRNA isoforms"
],
"content": "Introduction\n\nA single human gene can potentially yield a diverse array of alternative mRNA isoforms, thereby expanding both the repertoire of gene products and subsequently the number of alternative proteins produced. mRNAs with different exon combinations are transcribed from most (up to 90%) human genes, and can generate variants that differ in regulatory untranslated regions, or encode proteins with different sub-cellular localisations and functions1–5. Altered splicing patterns have been suggested as a new hallmark of cancer cells6–8, and in prostate cancer there is emerging evidence that expression of specific mRNA isoforms derived from cancer-relevant genes may contribute to disease progression9–11.\n\nAndrogen steroid hormones and the androgen receptor (AR) play a key role in the development and progression of prostate cancer, with alternative splicing enabling cancer cells to produce constitutively active ARs11–13. The AR belongs to the nuclear receptor superfamily of transcription factors, and is essential for prostate cancer cell survival, proliferation and invasion14–16. Classically, androgen binding promotes AR dimerization and its translocation to the nucleus, where it acts as either a transcriptional activator or a transcriptional repressor to dictate prostate specific gene expression patterns17–23. The major focus for prostate cancer therapeutics has been to reduce androgen levels through androgen deprivation therapy (ADT), either with inhibitors of androgen synthesis (for example, abiraterone) or with antagonists that prevent androgen binding to the AR (such as bicalutamide or enzalutamide)24. Although ADT is usually initially effective, most patients ultimately develop lethal castrate resistant disease for which there are limited treatment options11,12.\n\nAndrogens and other steroid hormones have also been associated with alternative splicing. Recent RNA-sequencing-based analysis of the androgen response of prostate cancer cells grown in vitro and within patients following ADT identified a set of 700 genes whose transcription is regulated by the AR in prostate cancer cells25. However, in addition to regulating transcriptional levels, steroid hormone receptors can control exon content of mRNA10,26–29. In prostate cancer androgens can modulate the expression of mRNA isoforms via pre-mRNA processing and promoter selection9,10,18,30. The AR can recruit the RNA binding proteins Sam68 and p68 as cofactors to influence alternative splicing of specific genes, and studies using minigenes driven from steroid responsive promoters indicate that the AR can affect both the transcriptional activity and alternative splicing of a subset of target genes11,31,32. Other steroid hormones also coordinate both transcription and splicing decisions29. The thyroid hormone receptor (TR) is known to play a role in coordinating the regulation of transcription and alternative splicing27, and the oestrogen receptor (ER) can both regulate alternative promoter selection and induce alternative splicing of specific gene sets that can influence breast cancer cell behaviour28,33–35.\n\nIn previous work we used exon level microarray analysis to identify 7 androgen dependent changes in mRNA isoform expression10. However, to what extent androgen-regulated mRNA isoforms are expressed in clinical prostate cancer is unclear. To address this, here we have used RNA-Sequencing data to globally profile alternative isoform expression in prostate cancer cells exposed to androgens, and correlated the results with transcriptomic data from clinical tissue. Our findings increase the number of known AR regulated mRNA isoforms by 10 fold and imply that pre-mRNA processing is an important mechanism through which androgens regulate gene expression in prostate cancer.\n\n\nMethods\n\nCell culture was as described previously25,36. All cells were grown at 37°C in 5% CO2. LNCaP cells (CRL-1740, ATCC) were maintained in RPMI-1640 with L-Glutamine (PAA Laboratories, R15-802) supplemented with 10% Fetal Bovine Serum (FBS) (PAA Laboratories, A15-101). For androgen treatment of cells, medium was supplemented with 10% dextran charcoal stripped FBS (PAA Laboratories, A15-119) to produce a steroid-deplete medium. Following culture for 72 hours, 10 nM synthetic androgen analogue methyltrienolone (R1881) (Perkin-Elmer, NLP005005MG) was either added (Androgen +) or absent (Steroid deplete) for the times indicated.\n\nRNA-seq transcript expression analysis of previously generated data25 was performed according to the Tuxedo protocol37. All reads were first mapped to human transcriptome/genome (build hg19) with TopHat38/Bowtie39, followed by per-sample transcript assembly with Cufflinks40. The mapped data was processed with Cuffmerge, Cuffdiff and Cuffcompare, followed by extraction of significantly differentially expressed genes/isoforms; expression changes between cells grown with androgen and cells grown without androgens were assessed. Reference files for the human genome (UCSC build hg19) were downloaded from the Cufflinks pages: (UCSC-hg19 package from June 2012 was used.). The software versions used for the analysis were: TopHat v1.4.1, SAM tools Version: 0.1.18 (r982:295), bowtie version 0.12.8 (64-bit) and cufflinks v1.3.0 (linked against Boost version 104000). The Tuxedo protocol37 was carried out as follows: For steps 1–5, no parameters (except for paths to input/output files) were altered. In step 15, additional switches -s, -R, and -C were used when running cuffcompare. Steps 16–18 (extraction of significant results) were performed on the command line.\n\nCells were harvested and total RNA extracted using TRIzol (Invitrogen, 15596-026) according to manufacturer's instructions. RNA was treated with DNase 1 (Ambion, AM2222) and cDNA was generated by reverse transcription of 500ng of total RNA using the Superscript VILO cDNA synthesis kit (Invitrogen, 11754-050). Alternative events were analysed by either reverse transcriptase PCR or real-time PCR. Exon profiles were monitored and quantified using the Qiaxcel capillary electrophoresis system (Qiagen) and percentage inclusion was calculated as described previously10. Real time PCR was performed in triplicate on cDNA using SYBR® Green PCR Master Mix (Invitrogen, 4309155) and the QuantStudio 7 Flex Real-Time PCR System (Thermo Fisher Scientific). Samples were normalised using the average of three reference genes, GAPDH, β -tubulin and actin. Ct values for each sample were calculated using SDS 2.4 software (Applied Biosystems) and relative mRNA expression was calculated using the 2-ΔΔCt method. All primer sequences are listed in Supplementary Table 1. Raw Ct values are given in Dataset 141.\n\nThe following commercial antibodies were used in the study: anti-RLN2 rabbit monoclonal (Abcam, ab183505 1:1000 dilution), anti-TACC2 rabbit polyclonal antibody (11407-1-AP, Proteintech 1:500 dilution), anti-NDUFV3 rabbit polyclonal antibody (13430-1-AP, Proteintech 1:500 dilution), anti-actin rabbit polyclonal (A2668, Sigma 1:2000 dilution), anti-α-Tubulin mouse monoclonal (Sigma, T5168 1:2000 dilution), normal rabbit IgG (711-035-152, Jackson labs 1:2000 dilution) and normal mouse IgG (715-036-150, Jackson labs 1:2000 dilution).\n\nGene ontology (GO) analysis of RNA-Seq data was carried out as described previously42. Enrichment of GO terms (with b500 annotations) was calculated using the goseq R package (version 1.18.0). Genes were considered significant at a p-value threshold of 0.05 after adjustment using the Benjamini-Hochberg false discovery rate.\n\nAvailable clinical and processed RNA-Seq data from The Cancer Genome Atlas (TCGA) prostate adenocarcinoma (PRAD) cohort, comprising 497 tumour samples from as many patients with different stages / Gleason grades and 52 matched samples taken from normal prostate tissue (were downloaded from the Broad Institute TCGA Genome Analysis Center (Firehose 16/01/28 run https://doi.org/10.7908/C11G0KM943). Transcriptome data from the TCGA PRAD cohort were analysed for alternative isoform expression, with transcript models relying on TCGA GAF2.1, corresponding to the University of California, Santa Cruz (UCSC) genome annotation from June 2011 (hg19 assembly). This annotation encompassed 42 of the 73 androgen-regulated alternative mRNA isoform pairs identified. These were studied using two types of analysis: 1) differential transcript expression between tumour and normal prostate tissue and 2) correlation between isoform expression in tumour samples and Gleason score or tumour stage.\n\nDifferential isoform and gene expression analysis was performed on estimated read counts using the limma software R package (version 3.7) following its RNA-Seq analysis workflow44. This workflow was also used for differential isoform ratio analysis, relying on logit-transformed ratio (see below). An FDR-adjusted p-value of 0.05 for the moderated t-statistics was used as threshold for significance of differential expression. Individual isoform expression was estimated in TPM (transcripts per million mapped reads). The expression ratio, henceforth called PSI (percent spliced-in), of each annotated androgen-regulated isoform pair in each TCGA sample was calculated as the ratio between the expression of isoform 1 and the total expression of isoforms 1 and 2 combined, i.e. the sum of their expressions. For each isoform pair, ΔPSI is the difference of median PSI between the tumour and the normal groups of samples.\n\nTwo-tailed Spearman’s rank correlation tests were used to study the association between isoform expression and both Gleason score and tumour stage (these were used herein as numeric variables). An FDR-adjusted p-value of 0.05 was used as threshold for significance. Isoform expression differences between tumour and normal samples were considered equivalent to those detected in LNCaP cells under androgen stimulation when there was a statistically significant consistent change in the levels of the expected induced or repressed isoform (1 or 2), concomitant with no contradictory change in the PSI. Isoform “switches” were considered equivalent when there was a minimum (ΔPSI > 2.5%) and statistically significant consistent change in the PSI. Equivalent criteria were used to evaluate the equivalence between androgen-dependence and the associations with Gleason score and tumour stage.\n\n\nStatistical analysis\n\nStatistical analyses were conducted using the GraphPad Prism software (version 5.04/d). PCR quantification of mRNA isoforms was assessed using the unpaired student’s t-test.\n\nData is presented as the mean of three independent samples ± standard error of the mean (SEM). Statistical significance is denoted as * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 and **** p ≤ 0.0001.\n\n\nResults\n\nWe analysed previously published RNAseq data from LNCaP cells25 to globally profile how frequently androgens drive production of alternative mRNA isoforms in prostate cancer cells. This analysis identified a group of 73 androgen regulated alternative mRNA isoforms, which could be validated by visualisation on the UCSC Genome Browser45 (Table 1). 64 AR regulated mRNA isoforms were novel to this study. Experimental validation in an independent RNA sample set using RT-PCR confirmed 17/17 of these alternative events at the mRNA level (Supplementary Figure 1). 73% of genes (53/73) with identified alternative androgen regulated mRNA isoforms also changed their overall expression levels in response to androgens (Table 2). Some of the androgen regulated alternative events are in genes are already implicated in in either prostate cancer or other cancer types (summarised in Table 3). However, Gene Ontology analysis of these 73 genes did not identify any significantly enriched biological processes.\n\nThe 73 identified mRNA isoforms were generated via androgen-regulated utilisation of 56 alternative promoters, 4 alternative 3′ ends and 13 alternative splicing events (Figure 1A). Of the 56 androgen regulated alternative promoters that were identified, 23 alternative promoters were induced by androgens (including LIG4, Figure 1B), 26 promoters were repressed by androgens, and for 7 genes there was a switch in usage from one promoter to another (Table 1). The alternative splicing events that were under androgen control included 12 alternative exons and one androgen-regulated intron retention (Table 1). 10 of these are novel to this study, including exclusion of an alternative exon in ZNF678 (Figure 1C). Of the alternative exons, six genes contained switches in previously unannotated protein-coding exons in response to androgen-exposure. We also identified four androgen regulated alternative mRNA 3' end isoform switches, including a switch in the 3’ end of the mRNA transcript for the MAT2A gene (Figure 1D).\n\n(A) Analysis of RNAseq data from LNCaP cells grown with (A+) or without androgens (R1881) (steroid deplete, SD) for 24 hours identified 73 androgen regulated alternative mRNA isoforms. The 73 alternative events were generated via androgen-regulated utilisation of 56 alternative promoters, 4 alternative 3' ends and 13 alternative splicing events. (B) Androgens drive a promoter switch in the LIG4 gene, which produces an mRNA isoform with an alternative 5’UTR. Visualisation of our LNCaP cell RNA-seq reads for the LIG4 gene on the UCSC genome browser identified a switch from promoter 1 to alternative promoter 2 in cells grown in the presence of androgens. Promoter 2 is predicted to produce a different 5’UTR without influencing the protein sequence (left panel). Quantitative PCR using primers specific to each promoter indicate that in response to androgens there is repression of promoter 1 and induction of promoter 2 (right panel). (C) Androgens drive alternative splicing of the ZNF678 gene. Visualisation of our LNCaP cell RNA-seq reads for the ZNF678 gene on the UCSC genome browser identified a switch to inclusion of a cassette exon in the presence of androgens. Inclusion of the alternative cassette exon in the ZNF678 gene is predicted to induce a switch to an alternative non-coding mRNA isoform (left panel). Quantitative PCR using primers in flanking exons confirmed increased inclusion of the alternative exon in LNCaP cells exposed to androgens (right panel). (D) Androgens promote selection of an alternative 3’ end for the MAT2A gene. Visualisation of our LNCaP cell RNA-seq reads for the MAT2A gene on the UCSC genome browser indicates a switch to reduced usage of an alternative 3’ end in the presence of androgens (left panel). Quantitative PCR using primers specific to each isoform confirmed down-regulation of an alternative 3’ end (p<0.01). Alternative 3’ ends for the MAT2A gene are predicted to produce proteins with different amino acid sequences and to influence a known Pfam domain (right panel).\n\n48/73 (66%) of the androgen regulated alternative events detected in response to androgen stimulation are predicted to change the amino acid sequence of the resulting protein (Table 1). Some of these are already known to have a well characterised role in prostate cancer progression, including an alternative promoter in the oncogene TPD52 that produces a protein isoform called PrLZ (Figure 2A)46–49. Others are not so well characterised. Using western blotting we could detect a novel shorter protein isoform corresponding to androgen-driven selection of an alternative promoter in the TACC2 gene (Figure 2B); and exclusion of a cassette exon in the NDUFV3 gene, which we show also produces a novel shorter protein isoform (Figure 2C). We also detected a switch in the 3' end of the mRNA transcript for the MAT2A gene, which is predicted to produce a protein isoform with a shorter C-terminal domain (Figure 1D); and induction of an alternative 3' isoform of CNNM2, which is predicted to be missing a conserved CBS domain (Table 1 and Supplementary Figure 1).\n\n(A) Androgens induce an alternative promoter in the oncogene TPD52 that produces an isoform called PrLZ. Visualisation of our LNCaP cell RNA-seq reads for the TPD52 gene on the UCSC genome browser identified a switch from promoter 1 to alternative promoter 2 in cells grown in the presence of androgens. Promoter 2 is known to produce an alternative protein isoform of TPD52 known as PrLZ (left panel). Quantitative PCR using primers specific to each promoter indicate an induction of the PrLZ isoform in response to androgens (middle panel). PrLZ has an alternative N-terminal amino acid sequence which results in an alternative protein isoform and disrupts a known Pfam domain (right panel). (B) Androgens induce an alternative promoter in the TACC2 gene that produces a novel alternative protein isoform. Visualisation of our LNCaP cell RNA-seq reads for the TACC2 gene on the UCSC genome browser identified a switch from promoter 1 to alternative promoter 2 in cells grown in the presence of androgens. Promoter 2 is predicted to produce an alternative shorter protein isoform of TACC2 (isoform 2) (left panel). Quantitative PCR using primers specific to each promoter indicate a switch from isoform 1 to isoform 2 in response to androgens (middle panel). Detection of TACC2 protein in LNCaP by western blotting (cells were grown with or without androgens for 24 or 48 hours). Tubulin was used as a loading control. Exposure to androgens for 48 hours induces expression of the alternative TACC2 protein isoform (right panel). (C) Androgens drive alternative splicing of the NDUFV3 gene. Visualisation of our LNCaP cell RNA-seq reads for the NDUFV3 gene on the UCSC genome browser identified a switch to exclusion of a cassette exon in the presence of androgens (left panel). Quantitative PCR using primers in flanking exons confirmed less inclusion of the alternative exon in LNCaP cells exposed to androgens (middle panel). Exclusion of the alternative cassette exon is predicted to produce an alternative protein isoform. Detection of NDUFV3 protein in LNCaP cells using western blotting (right panel). (D) Androgens suppress an alternative promoter in the RLN2 gene, which produces a shorter non-coding mRNA isoform. Visualisation of our LNCaP cell RNA-seq reads for the RLN2 gene on the UCSC genome browser identified a switch from promoter 1 to alternative promoter 2 in cells grown in the presence of androgens. Promoter 2 is predicted to produce an untranslated non-coding mRNA isoform (left panel). Quantitative PCR using primers specific to each promoter indicated a significant switch in promoter usage in response to androgens (middle panel). Detection of RLN2 protein in LNCaP by western blotting (cells were grown with or without androgens for 48 hours). Actin was used as a loading control. As seen previously55, androgens suppress RLN2 protein levels.\n\n11 of the remaining identified androgen-regulated alternative events change the expression of mRNAs from coding to non-coding or untranslated (not predicted to produce a protein) (Table 1). These included promoter switches for the RLN1 and RLN2 genes which encode peptide hormones that may be important in prostate cancer5,50–55. Androgens drive a promoter switch in both RLN1 and RLN2 to produce predicted non-coding or untranslated mRNA isoforms, reducing expression of protein-coding RLN1 and RLN2 mRNA isoforms. To test whether prostate cancer cells turn off gene expression by switching between utilisation of promoters that generate coding and noncoding mRNAs, we analysed RLN2 protein levels. Consistent with our hypothesis and a previous study55, RLN2 protein production was negatively regulated by androgens in parallel to the switch to the non-coding mRNA isoform (Figure 2D).\n\n14 of the identified androgen-dependent mRNA isoforms lead to/result in coding mRNAs with altered 5’ untranslated regions (5′ UTR) with no impact on the coding sequence. These include a promoter switch in the LIG4 gene (Figure 1B).\n\nTo investigate potential links between androgen-dependent mRNA isoforms and tumourigenesis, we analysed the expression of 41 androgen-regulated mRNA isoform pairs in clinical prostate adenocarcinoma and normal prostate tissues. This analysis utilised transcriptomic data from 497 tumour samples and 52 normal samples in the PRAD TCGA cohort104. The remaining isoform pairs identified within our dataset have not been previously annotated by UCSC, therefore it was not possible to include them in our comparison. A description of the cohort used is summarised in Table 4.\n\nAll tumours were hormone naive (not subject to ADT) at the time of sample collection\n\n33 of the 42 mRNA isoform pairs exhibited significant differences in the expression of at least one of the isoforms, or in the isoform expression ratio between tumour and normal tissues (Table 5). 13 of those tumour-specific alterations mimicked the effect of androgen stimulation in LNCaP cells: the changes were in form of alternative promoters for TACC2, TPD52, NUP93, PIK3R1, RDH13, ZFAND6, CDIP1, YIF1B, LIMK2, and FDFT1; an alternative 3´ end in CNNM2; and alternative exons in NDUFV3 and SS18 (Figure 3, Table 5 & Supplementary Figure 2). Two of the alternative promoters (ZFAND6 and CDIP1) are predicted to introduce a change in the 5′UTR, whereas all the others are predicted to alter the resulting protein isoform. A number of mRNA isoforms that were androgen responsive in LNCaP cells showed tumour specific alterations opposite to the effect of androgen stimulation. These were LIG4, MAPRE2, OSBPL1A, SEPT5, NR4A1, and RCAN1 (all predicted to alter the resulting protein isoform except LIG4). For the remaining 14 mRNA isoform pairs, the data was inconclusive according to the consistency conditions listed in the methods section (Table 5).\n\nViolin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio in PSI (right panel) in normal and tumour samples. The mean log2 fold-change (logFC) in expression between tumour and normal samples and the associated FDR-adjusted p-value for the moderated t-statistic of differential expression are shown for both isoforms (left and central panels). The mean difference in PSI (deltaPSI) between tumour and normal samples and the associated FDR-adjusted p-value for the Mann-Whitney U test of differential splicing are shown (right panel).\n\nWe next investigated whether the identified androgen-dependent mRNA isoforms are differentially expressed during prostate cancer progression by correlating the expression levels of each isoform with Gleason scores and prostate tumour grades within the PRAD TCGA cohort (Figure 4 & Figure 5, Table 6 & Table 7 and Supplementary Figure 3 & Supplementary Figure 4). For 6 of the alternative mRNA isoforms responsive to androgens (made from alternative promoters in LIG4, OSBPL1A, CLK3, TSC22D3 & ZNF32 and utilising an alternative exon in ZNF121), the expression changed significantly with Gleason score and showed specific alterations consistent with the effect of androgen stimulation. Conversely, 9 alternative isoforms (which were androgen responsive in LNCaP cells) showed tumour specific alterations opposite to the effect of androgen stimulation (including an alternative promoters in NUP93 and the alternative 3´end of MAT2A). 3 androgen regulated mRNA isoforms (OSBPL1A, CLK3 and TSC22D3) change significantly with both Gleason grade and tumour stage.\n\nViolin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio (right panel) by Gleason grade. Their respective Spearman’s correlation coefficient (Rho) with grade and associated FDR-adjusted p-value are shown.\n\nViolin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio (right panel) by tumour stage. Their respective Spearman’s correlation coefficient (Rho) with stage and associated FDR-adjusted p-value are shown.\n\n\nDiscussion\n\nThe main function of the androgen receptor (AR) is as a DNA binding transcription factor that regulates gene expression. Here we show the AR can couple hormone induced gene transcription to alternative mRNA isoform expression in prostate cancer. In response to androgens, the AR can induce the use of alternative promoters, induce the expression of alternatively spliced mRNA isoforms, regulate the expression of non-coding mRNA transcripts, and promote the transcription of mRNA isoforms encoding different protein isoforms. Importantly, we also find that some of these alternative mRNA isoforms are differentially regulated in prostate cancer versus normal tissue and also significantly change expression during tumour progression. Our data suggest that most androgen regulated alternative mRNA isoforms are generated through alternative promoter selection rather than through internal alternative exon splicing mechanisms. This suggests expression of alternative isoforms of specific genes can be a consequence of RNA polymerase being recruited to different promoters in response to androgen stimulation. Alternative promoter usage has been observed for many genes and is believed to play a significant role in the control of gene expression4,105,106. Alternative promoter use can also generate mRNA isoforms with distinct functional activities from the same gene, sometimes having opposing functions11.\n\nAndrogen exposure further drives a smaller number of alternative splicing events suggesting that the AR could contribute to altered patterns of splicing in prostate cancer cells. Tumour progression is believed to be associated with a coordinated change in splicing patterns which is regulated by several factors including signalling molecules7. We also identified 4 AR regulated alternative mRNA 3′ end isoform switches. This is the first time that regulation of 3′ mRNA end processing has been shown to be controlled by androgens. The selection of alternative 3′ ends can produce mRNA isoforms differing in the length of their 3′ UTRs (which can lead to the inclusion or exclusion of regulatory elements and influence gene expression), or in their C-terminal coding region (which can contribute to proteome diversity)107–114. Defective 3′ mRNA processing of numerous genes has been linked to an oncogenic phenotype115–119, and the 3′ mRNA end profiles of samples from multiple cancer types significantly differ from those of healthy tissue samples115,119–121.\n\nBased on the findings presented in this study, we propose that activated AR has the ability to coordinate both transcriptional activity and mRNA isoform decisions through the recruitment of co-regulators to specific promoters. The genomic action of the AR is influenced by a large number of collaborating transcription factors122–124. Specifically, Sam68 and p68 have been shown to modulate AR dependent alternative splicing of specific genes and are significantly overexpressed in prostate cancer31,32. In future work it will be important to define the role of specific AR co-regulators in AR mediated isoform selection.\n\nSome of the androgen dependent mRNA isoforms identified here are predicted to yield protein isoforms that may be clinically important, or to switch off protein production via generation of noncoding mRNA isoforms. Although the functional significance of the alternative mRNA isoforms identified in this study is yet largely unexplored, as is their role in the cellular response to androgens, the presented results emphasize the importance of analysing gene regulation and function at the mRNA isoform level.\n\n\nData availability\n\nThe RNASeq data from LNCaP cells has been published previously https://doi.org/10.1016/j.ebiom.2016.04.01825\n\nThe RNAseq custom tracks are available in Supplementary File 1. To view these files please load them onto the UCSC website using the ‘My data’ tab and ‘custom tracks’. Then ‘Paste URLs or data’. The data is aligned to Feb 2009 (GRCh37/hg19).\n\nProstate adenocarcinoma cohort RNA-Seq data was downloaded from the Broad Institute TCGA Genome Analysis Center: Firehose 16/01/28 run https://doi.org/10.7908/C11G0KM943\n\nDataset 1: Real-time PCR raw Ct values 10.5256/f1000research.15604.d21287341\n\nDataset 2: Raw unedited western blot images 10.5256/f1000research.15604.d212874125",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by Prostate Cancer UK [PG12-34, S13-020 and RIA16-ST2-011].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary Table 1: Details of primer sequences used.\n\nClick here to access the data.\n\nSupplementary File 1: RNA-Seq reads custom tracks for visualisation on UCSC genome browser\n\nClick here to access the data.\n\nSupplementary Figure 1: PCR validation of 17 androgen regulated alternative events.\n\nClick here to access the data.\n\nSupplementary Figure 2: Differential alternative mRNA isoform expression in theTGCA PRAD cohort. Normal vs. tumour (unpaired samples) analysis. Violin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio in PSI (right panel) in normal and tumour samples. The mean log2 fold-change (logFC) in expression between tumour and normal samples and the associated FDR-adjusted p-value for the moderated t-statistic of differential expression are shown for both isoforms (left and central panels). The mean difference in PSI (deltaPSI) between tumour and normal samples and the associated FDR-adjusted p-value for the Mann-Whitney U test of differential splicing are shown (right panel).\n\nClick here to access the data.\n\nSupplementary Figure 3: Differential alternative mRNA isoform expression in the TGCA PRAD cohort across different Gleason grades. Violin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio (right panel) by Gleason grade. Their respective Spearman’s correlation coefficient (Rho) with grade and associated FDR-adjusted p-value are shown.\n\nClick here to access the data.\n\nSupplementary Figure 4: Differential alternative mRNA isoform expression in the TGCA PRAD cohort across different tumour stages. Violin-boxplots of expression in transcripts per million mapped reads (TPM) of Isoforms 1 (left panel) and 2 (central panel), and of their expression ratio (right panel) by tumour stage. Their respective Spearman’s correlation coefficient (Rho) with stage and associated FDR-adjusted p-value are shown.\n\nClick here to access the data.\n\n\nReferences\n\nJohnson JM, Castle J, Garrett-Engele P, et al.: Genome-wide survey of human alternative pre-mRNA splicing with exon junction microarrays. Science. 2003; 302(5653): 2141–4. PubMed Abstract | Publisher Full Text\n\nPan Q, Shai O, Lee LJ, et al.: Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008; 40(12): 1413–5. PubMed Abstract | Publisher Full Text\n\nWang ET, Sandberg R, Luo S, et al.: Alternative isoform regulation in human tissue transcriptomes. Nature. 2008; 456(7221): 470–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavuluri RV, Suzuki Y, Sugano S, et al.: The functional consequences of alternative promoter use in mammalian genomes. Trends Genet. 2008; 24(4): 167–77. PubMed Abstract | Publisher Full Text\n\nTrapnell C, Williams BA, Pertea G, et al.: Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol. 2010; 28(5): 511–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHanahan D, Weinberg RA: Hallmarks of cancer: the next generation. Cell. 2011; 144(5): 646–74. PubMed Abstract | Publisher Full Text\n\nOltean S, Bates DO: Hallmarks of alternative splicing in cancer. Oncogene. 2014; 33(46): 5311–8. PubMed Abstract | Publisher Full Text\n\nLadomery M: Aberrant alternative splicing is another hallmark of cancer. Int J Cell Biol. 2013; 2013: 463786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRajan P, Elliott DJ, Robson CN, et al.: Alternative splicing and biological heterogeneity in prostate cancer. Nat Rev Urol. 2009; 6(8): 454–60. PubMed Abstract | Publisher Full Text\n\nRajan P, Dalgliesh C, Carling PJ, et al.: Identification of novel androgen-regulated pathways and mRNA isoforms through genome-wide exon-specific profiling of the LNCaP transcriptome. PLoS One. 2011; 6(12): e29088. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Livermore K, Rajan P, et al.: RNA splicing and splicing regulator changes in prostate cancer pathology. Hum Genet. 2017; 136(9): 1143–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLivermore KE, Munkley J, Elliott DJ: Androgen receptor and prostate cancer. AIMS Mol Sci. 2016; 3(2): 280–99. Publisher Full Text\n\nMills IG: Maintaining and reprogramming genomic androgen receptor activity in prostate cancer. Nat Rev Cancer. 2014; 14(3): 187–98. PubMed Abstract | Publisher Full Text\n\nSnoek R, Cheng H, Margiotti K, et al.: In vivo knockdown of the androgen receptor results in growth inhibition and regression of well-established, castration-resistant prostate tumors. Clin Cancer Res. 2009; 15(1): 39–47. PubMed Abstract | Publisher Full Text\n\nHara T, Miyazaki H, Lee A, et al.: Androgen receptor and invasion in prostate cancer. Cancer Res. 2008; 68(4): 1128–35. PubMed Abstract | Publisher Full Text\n\nHååg P, Bektic J, Bartsch G, et al.: Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. J Steroid Biochem Mol Biol. 2005; 96(3–4): 251–8. PubMed Abstract | Publisher Full Text\n\nMunkley J, Oltean S, Vodák D, et al.: The androgen receptor controls expression of the cancer-associated sTn antigen and cell adhesion through induction of ST6GalNAc1 in prostate cancer. Oncotarget. 2015; 6(33): 34358–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Rajan P, Lafferty NP, et al.: A novel androgen-regulated isoform of the TSC2 tumour suppressor gene increases cell proliferation. Oncotarget. 2014; 5(1): 131–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMassie CE, Lynch A, Ramos-Montoya A, et al.: The androgen receptor fuels prostate cancer by regulating central metabolism and biosynthesis. EMBO J. 2011; 30(13): 2719–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Lafferty NP, Kalna G, et al.: Androgen-regulation of the protein tyrosine phosphatase PTPRR activates ERK1/2 signalling in prostate cancer cells. BMC Cancer. 2015; 15: 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Livermore KE, McClurg UL, et al.: The PI3K regulatory subunit gene PIK3R1 is under direct control of androgens and repressed in prostate cancer cells. Oncoscience. 2015; 2(9): 755–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKarantanos T, Corn PG, Thompson TC: Prostate cancer progression after androgen deprivation therapy: mechanisms of castrate resistance and novel therapeutic approaches. Oncogene. 2013; 32(49): 5501–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCai C, He HH, Chen S, et al.: Androgen receptor gene expression in prostate cancer is directly suppressed by the androgen receptor through recruitment of lysine-specific demethylase 1. Cancer Cell. 2011; 20(4): 457–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYuan X, Cai C, Chen S, et al.: Androgen receptor functions in castration-resistant prostate cancer and mechanisms of resistance to new agents targeting the androgen axis. Oncogene. 2014; 33(22): 2815–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Vodak D, Livermore KE, et al.: Glycosylation is an Androgen-Regulated Process Essential for Prostate Cancer Cell Viability. EBioMedicine. 2016; 8: 103–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAuboeuf D, Hönig A, Berget SM, et al.: Coordinate regulation of transcription and splicing by steroid receptor coregulators. Science. 2002; 298(5592): 416–9. PubMed Abstract | Publisher Full Text\n\nSatoh T, Katano-Toki A, Tomaru T, et al.: Coordinated regulation of transcription and alternative splicing by the thyroid hormone receptor and its associating coregulators. Biochem Biophys Res Commun. 2014; 451(1): 24–9. PubMed Abstract | Publisher Full Text\n\nDutertre M, Gratadou L, Dardenne E, et al.: Estrogen regulation and physiopathologic significance of alternative promoters in breast cancer. Cancer Res. 2010; 70(9): 3760–70. PubMed Abstract | Publisher Full Text\n\nDowhan DH, Hong EP, Auboeuf D, et al.: Steroid hormone receptor coactivation and alternative RNA splicing by U2AF65-related proteins CAPERalpha and CAPERbeta. Mol Cell. 2005; 17(3): 429–39. PubMed Abstract | Publisher Full Text\n\nCochrane DR, Wang Z, Muramaki M, et al.: Differential regulation of clusterin and its isoforms by androgens in prostate cells. J Biol Chem. 2007; 282(4): 2278–87. PubMed Abstract | Publisher Full Text\n\nRajan P, Gaughan L, Dalgliesh C, et al.: The RNA-binding and adaptor protein Sam68 modulates signal-dependent splicing and transcriptional activity of the androgen receptor. J Pathol. 2008; 215(1): 67–77. PubMed Abstract | Publisher Full Text\n\nClark EL, Coulson A, Dalgliesh C, et al.: The RNA helicase p68 is a novel androgen receptor coactivator involved in splicing and is overexpressed in prostate cancer. Cancer Res. 2008; 68(19): 7938–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalama SA, Mohammad MA, Diaz-Arrastia CR, et al.: Estradiol-17β upregulates pyruvate kinase M2 expression to coactivate estrogen receptor-α and to integrate metabolic reprogramming with the mitogenic response in endometrial cells. J Clin Endocrinol Metab. 2014; 99(10): 3790–9. PubMed Abstract | Publisher Full Text\n\nLal S, Allan A, Markovic D, et al.: Estrogen alters the splicing of type 1 corticotropin-releasing hormone receptor in breast cancer cells. Sci Signal. 2013; 6(282): ra53. PubMed Abstract | Publisher Full Text\n\nBhat-Nakshatri P, Song EK, Collins NR, et al.: Interplay between estrogen receptor and AKT in estradiol-induced alternative splicing. BMC Med Genomics. 2013; 6: 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, McClurg UL, Livermore KE, et al.: The cancer-associated cell migration protein TSPAN1 is under control of androgens and its upregulation increases prostate cancer cell migration. Sci Rep. 2017; 7(1): 5249. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrapnell C, Roberts A, Goff L, et al.: Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012; 7(3): 562–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim D, Pertea G, Trapnell C, et al.: TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013; 14(4): R36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangmead B, Salzberg SL: Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012; 9(4): 357–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrapnell C, Hendrickson DG, Sauvageau M, et al.: Differential analysis of gene regulation at transcript resolution with RNA-seq. Nature Biotechnol. 2013; 31(1): 46–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Maia TM, Ibarluzea N, et al.: Dataset 1 in: Androgen-dependent alternative mRNA isoform expression in prostate cancer cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15604.d212873\n\nYoung MD, Wakefield MJ, Smyth GK, et al.: Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biol. 2010; 11(2): R14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCenter BITGDA: Analysis-ready standardized TCGA data from Broad GDAC Firehose 2016_01_28 run. Broad Institute of MIT and Harvard Dataset. 2016. Publisher Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTyner C, Barber GP, Casper J, et al.: The UCSC Genome Browser database: 2017 update. Nucleic Acids Res. 2017; 45(D1): D626–D34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoslan N, Bieche I, Bright RK, et al.: TPD52 represents a survival factor in ERBB2-amplified breast cancer cells. Mol Carcinog. 2014; 53(10): 807–19. PubMed Abstract | Publisher Full Text\n\nZhang H, Wang J, Pang B, et al.: PC-1/PrLZ contributes to malignant progression in prostate cancer. Cancer Res. 2007; 67(18): 8906–13. PubMed Abstract | Publisher Full Text\n\nZhang D, He D, Xue Y, et al.: PrLZ protects prostate cancer cells from apoptosis induced by androgen deprivation via the activation of Stat3/Bcl-2 pathway. Cancer Res. 2011; 71(6): 2193–202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu L, Shang ZF, Wang J, et al.: PC-1/PrLZ confers resistance to rapamycin in prostate cancer cells through increased 4E-BP1 stability. Oncotarget. 2015; 6(24): 20356–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeng S, Agoulnik IU, Bogatcheva NV, et al.: Relaxin promotes prostate cancer progression. Clin Cancer Res. 2007; 13(6): 1695–702. PubMed Abstract | Publisher Full Text\n\nFeng S, Agoulnik IU, Li Z, et al.: Relaxin/RXFP1 signaling in prostate cancer progression. Ann N Y Acad Sci. 2009; 1160: 379–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeschadim A, Summerlee AJ, Silvertown JD: Targeting the relaxin hormonal pathway in prostate cancer. Int J Cancer. 2015; 137(10): 2287–95. PubMed Abstract | Publisher Full Text\n\nSilvertown JD, Ng J, Sato T, et al.: H2 relaxin overexpression increases in vivo prostate xenograft tumor growth and angiogenesis. Int J Cancer. 2006; 118(1): 62–73. PubMed Abstract | Publisher Full Text\n\nSilvertown JD, Symes JC, Neschadim A, et al.: Analog of H2 relaxin exhibits antagonistic properties and impairs prostate tumor growth. FASEB J. 2007; 21(3): 754–65. PubMed Abstract | Publisher Full Text\n\nThompson VC, Morris TG, Cochrane DR, et al.: Relaxin becomes upregulated during prostate cancer progression to androgen independence and is negatively regulated by androgens. Prostate. 2006; 66(16): 1698–709. PubMed Abstract | Publisher Full Text\n\nOnodera Y, Takagi K, Miki Y, et al.: TACC2 (transforming acidic coiled-coil protein 2) in breast carcinoma as a potent prognostic predictor associated with cell proliferation. Cancer Med. 2016; 5(8): 1973–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTakayama K, Horie-Inoue K, Suzuki T, et al.: TACC2 is an androgen-responsive cell cycle regulator promoting androgen-mediated and castration-resistant growth of prostate cancer. Mol Endocrinol. 2012; 26(5): 748–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim DH, Oh SY, Kim SY, et al.: DNA ligase4 as a prognostic marker in nasopharyngeal cancer patients treated with radiotherapy. Asian Pac J Cancer Prev. 2014; 15(24): 10985–9. PubMed Abstract | Publisher Full Text\n\nJun S, Jung YS, Suh HN, et al.: LIG4 mediates Wnt signalling-induced radioresistance. Nat Commun. 2016; 7: 10994. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrupp K, Roettger L, Kluth M, et al.: Expression of DNA ligase IV is linked to poor prognosis and characterizes a subset of prostate cancers harboring TMPRSS2:ERG fusion and PTEN deletion. Oncol Rep. 2015; 34(3): 1211–20. PubMed Abstract | Publisher Full Text\n\nCao WH, Liu HM, Liu X, et al.: Relaxin enhances in-vitro invasiveness of breast cancer cell lines by upregulation of S100A4/MMPs signaling. Eur Rev Med Pharmacol Sci. 2013; 17(5): 609–17. PubMed Abstract\n\nBinder C, Hagemann T, Husen B, et al.: Relaxin enhances in-vitro invasiveness of breast cancer cell lines by up-regulation of matrix metalloproteases. Mol Hum Reprod. 2002; 8(9): 789–96. PubMed Abstract | Publisher Full Text\n\nMa J, Niu M, Yang W, et al.: Role of relaxin-2 in human primary osteosarcoma. Cancer Cell Int. 2013; 13(1): 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHombach-Klonisch S, Bialek J, Trojanowicz B, et al.: Relaxin enhances the oncogenic potential of human thyroid carcinoma cells. Am J Pathol. 2006; 169(2): 617–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRadestock Y, Willing C, Kehlen A, et al.: Relaxin enhances S100A4 and promotes growth of human thyroid carcinoma cell xenografts. Mol Cancer Res. 2010; 8(4): 494–506. PubMed Abstract | Publisher Full Text\n\nByrne JA, Frost S, Chen Y, et al.: Tumor protein D52 (TPD52) and cancer-oncogene understudy or understudied oncogene? Tumour Biol. 2014; 35(8): 7369–82. PubMed Abstract | Publisher Full Text\n\nByrne JA, Balleine RL, Schoenberg Fejzo M, et al.: Tumor protein D52 (TPD52) is overexpressed and a gene amplification target in ovarian cancer. Int J Cancer. 2005; 117(6): 1049–54. PubMed Abstract | Publisher Full Text\n\nZhao Z, Liu H, Hou J, et al.: Tumor Protein D52 (TPD52) Inhibits Growth and Metastasis in Renal Cell Carcinoma Cells Through the PI3K/Akt Signaling Pathway. Oncol Res. 2017; 25(5): 773–9. PubMed Abstract | Publisher Full Text\n\nLi J, Li Y, Liu H, et al.: The four-transmembrane protein MAL2 and tumor protein D52 (TPD52) are highly expressed in colorectal cancer and correlated with poor prognosis. PLoS One. 2017; 12(5): e0178515. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRubin MA, Varambally S, Beroukhim R, et al.: Overexpression, amplification, and androgen regulation of TPD52 in prostate cancer. Cancer Res. 2004; 64(11): 3814–22. PubMed Abstract | Publisher Full Text\n\nGoto Y, Nishikawa R, Kojima S, et al.: Tumour-suppressive microRNA-224 inhibits cancer cell migration and invasion via targeting oncogenic TPD52 in prostate cancer. FEBS Lett. 2014; 588(10): 1973–82. PubMed Abstract | Publisher Full Text\n\nMoritz T, Venz S, Junker H, et al.: Isoform 1 of TPD52 (PC-1) promotes neuroendocrine transdifferentiation in prostate cancer cells. Tumour Biol. 2016; 37(8): 10435–46. PubMed Abstract | Publisher Full Text\n\nShang ZF, Wei Q, Yu L, et al.: Suppression of PC-1/PrLZ sensitizes prostate cancer cells to ionizing radiation by attenuating DNA damage repair and inducing autophagic cell death. Oncotarget. 2016; 7(38): 62340–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi L, Xie H, Liang L, et al.: Increased PrLZ-mediated androgen receptor transactivation promotes prostate cancer growth at castration-resistant stage. Carcinogenesis. 2013; 34(2): 257–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang YF, Jan YH, Liu YP, et al.: Squalene synthase induces tumor necrosis factor receptor 1 enrichment in lipid rafts to promote lung cancer metastasis. Am J Respir Crit Care Med. 2014; 190(6): 675–87. PubMed Abstract | Publisher Full Text\n\nFukuma Y, Matsui H, Koike H, et al.: Role of squalene synthase in prostate cancer risk and the biological aggressiveness of human prostate cancer. Prostate Cancer Prostatic Dis. 2012; 15(4): 339–45. PubMed Abstract | Publisher Full Text\n\nSusini T, Berti B, Carriero C, et al.: Topoisomerase II alpha and TLE3 as predictive markers of response to anthracycline and taxane-containing regimens for neoadjuvant chemotherapy in breast cancer. Onco Targets Ther. 2014; 7: 2111–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSamimi G, Ring BZ, Ross DT, et al.: TLE3 expression is associated with sensitivity to taxane treatment in ovarian carcinoma. Cancer Epidemiol Biomarkers Prev. 2012; 21(2): 273–9. PubMed Abstract | Publisher Full Text\n\nYang RW, Zeng YY, Wei WT, et al.: TLE3 represses colorectal cancer proliferation by inhibiting MAPK and AKT signaling pathways. J Exp Clin Cancer Res. 2016; 35(1): 152. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakaya HI, Beckedorff FC, Baldini ML, et al.: Splice variants of TLE family genes and up-regulation of a TLE3 isoform in prostate tumors. Biochem Biophys Res Commun. 2007; 364(4): 918–23. PubMed Abstract | Publisher Full Text\n\nGuan H, Liu C, Fang F, et al.: MicroRNA-744 promotes prostate cancer progression through aberrantly activating Wnt/β-catenin signaling. Oncotarget. 2017; 8(9): 14693–707. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiménez-Mascarell P, Oyenarte I, Hardy S, et al.: Structural Basis of the Oncogenic Interaction of Phosphatase PRL-1 with the Magnesium Transporter CNNM2. J Biol Chem. 2017; 292(3): 786–801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee JH, Zhao XM, Yoon I, et al.: Integrative analysis of mutational and transcriptional profiles reveals driver mutations of metastatic breast cancers. Cell Discov. 2016; 2: 16025. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang W, Sviripa V, Chen X, et al.: Fluorinated N,N-dialkylaminostilbenes repress colon cancer by targeting methionine S-adenosyltransferase 2A. ACS Chem Biol. 2013; 8(4): 796–803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrau M, Feo F, Pascale RM: Pleiotropic effects of methionine adenosyltransferases deregulation as determinants of liver cancer progression and prognosis. J Hepatol. 2013; 59(4): 830–41. PubMed Abstract | Publisher Full Text\n\nWang X, Guo X, Yu W, et al.: Expression of methionine adenosyltransferase 2A in renal cell carcinomas and potential mechanism for kidney carcinogenesis. BMC Cancer. 2014; 14: 196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaldonado LY, Arsene D, Mato JM, et al.: Methionine adenosyltransferases in cancers: Mechanisms of dysregulation and implications for therapy. Exp Biol Med (Maywood). 2018; 243(2): 107–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTomasi ML, Cossu C, Spissu Y, et al.: S-adenosylmethionine and methylthioadenosine inhibit cancer metastasis by targeting microRNA 34a/b-methionine adenosyltransferase 2A/2B axis. Oncotarget. 2017; 8(45): 78851–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCizkova M, Vacher S, Meseure D, et al.: PIK3R1 underexpression is an independent prognostic marker in breast cancer. BMC Cancer. 2013; 13: 545. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCheung LW, Hennessy BT, Li J, et al.: High frequency of PIK3R1 and PIK3R2 mutations in endometrial cancer elucidates a novel mechanism for regulation of PTEN protein stability. Cancer Discov. 2011; 1(2): 170–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMa J, Hou X, Li M, et al.: Genome-wide methylation profiling reveals new biomarkers for prognosis prediction of glioblastoma. J Cancer Res Ther. 2015; 11 Suppl 2: C212–5. PubMed Abstract | Publisher Full Text\n\nThorsen K, Schepeler T, Øster B, et al.: Tumor-specific usage of alternative transcription start sites in colorectal cancer identified by genome-wide exon array analysis. BMC Genomics. 2011; 12: 505. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLuo A, Zhang X, Fu L, et al.: Zinc finger factor ZNF121 is a MYC-interacting protein functionally affecting MYC and cell proliferation in epithelial cells. J Genet Genomics. 2016; 43(12): 677–85. PubMed Abstract | Publisher Full Text\n\nLiu C, Zheng L, Wang H, et al.: The RCAN1 inhibits NF-κB and suppresses lymphoma growth in mice. Cell Death Dis. 2015; 6(10): e1929. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEspinosa AV, Shinohara M, Porchia LM, et al.: Regulator of calcineurin 1 modulates cancer cell migration in vitro. Clin Exp Metastasis. 2009; 26(6): 517–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHang X, Wu Z, Chu K, et al.: Low expression of DCXR protein indicates a poor prognosis for hepatocellular carcinoma patients. Tumour Biol. 2016; 37(11): 15079–85. PubMed Abstract | Publisher Full Text\n\nEbert B, Kisiela M, Maser E: Human DCXR - another 'moonlighting protein' involved in sugar metabolism, carbonyl detoxification, cell adhesion and male fertility?. Biol Rev Camb Philos Soc. 2015; 90(1): 254–78. PubMed Abstract | Publisher Full Text\n\nCho-Vega JH, Vega F, Schwartz MR, et al.: Expression of dicarbonyl/L-xylulose reductase (DCXR) in human skin and melanocytic lesions: morphological studies supporting cell adhesion function of DCXR. J Cutan Pathol. 2007; 34(7): 535–42. PubMed Abstract | Publisher Full Text\n\nCho-Vega JH, Tsavachidis S, Do KA, et al.: Dicarbonyl/L-xylulose reductase: a potential biomarker identified by laser-capture microdissection-micro serial analysis of gene expression of human prostate adenocarcinoma. Cancer Epidemiol Biomarkers Prev. 2007; 16(12): 2615–22. PubMed Abstract | Publisher Full Text\n\nAbiatari I, Gillen S, DeOliveira T, et al.: The microtubule-associated protein MAPRE2 is involved in perineural invasion of pancreatic cancer cells. Int J Oncol. 2009; 35(5): 1111–6. PubMed Abstract | Publisher Full Text\n\nZhang JH, Li AY, Wei N: Downregulation of long non-coding RNA LINC01133 is predictive of poor prognosis in colorectal cancer patients. Eur Rev Med Pharmacol Sci. 2017; 21(9): 2103–7. PubMed Abstract\n\nZhang J, Zhu N, Chen X: A novel long noncoding RNA LINC01133 is upregulated in lung squamous cell cancer and predicts survival. Tumour Biol. 2015; 36(10): 7465–71. PubMed Abstract | Publisher Full Text\n\nBrown-Endres L, Schoenfeld D, Tian F, et al.: Expression of the p53 target CDIP correlates with sensitivity to TNFα-induced apoptosis in cancer cells. Cancer Res. 2012; 72(9): 2373–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFraser M, Sabelnykova VY, Yamaguchi TN, et al.: Genomic hallmarks of localized, non-indolent prostate cancer. Nature. 2017; 541(7637): 359–64. PubMed Abstract | Publisher Full Text\n\nAyoubi TA, Van De Ven WJ: Regulation of gene expression by alternative promoters. FASEB J. 1996; 10(4): 453–60. PubMed Abstract | Publisher Full Text\n\nBaek D, Davis C, Ewing B, et al.: Characterization and predictive discovery of evolutionarily conserved mammalian alternative promoters. Genome Res. 2007; 17(2): 145–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMillevoi S, Vagner S: Molecular mechanisms of eukaryotic pre-mRNA 3' end processing regulation. Nucleic Acids Res. 2010; 38(9): 2757–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHollerer I, Grund K, Hentze MW, et al.: mRNA 3'end processing: A tale of the tail reaches the clinic. EMBO Mol Med. 2014; 6(1): 16–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTian B, Manley JL: Alternative cleavage and polyadenylation: the long and short of it. Trends Biochem Sci. 2013; 38(6): 312–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoore MJ: From birth to death: the complex lives of eukaryotic mRNAs. Science. 2005; 309(5740): 1514–8. PubMed Abstract | Publisher Full Text\n\nMiyamoto S, Chiorini JA, Urcelay E, et al.: Regulation of gene expression for translation initiation factor eIF-2 alpha: importance of the 3' untranslated region. Biochem J. 1996; 315(Pt 3): 791–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTakagaki Y, Seipelt RL, Peterson ML, et al.: The polyadenylation factor CstF-64 regulates alternative processing of IgM heavy chain pre-mRNA during B cell differentiation. Cell. 1996; 87(5): 941–52. PubMed Abstract | Publisher Full Text\n\nEdwalds-Gilbert G, Veraldi KL, Milcarek C: Alternative poly(A) site selection in complex transcription units: means to an end? Nucleic Acids Res. 1997; 25(13): 2547–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLutz CS: Alternative polyadenylation: a twist on mRNA 3' end formation. ACS Chem Biol. 2008; 3(10): 609–17. PubMed Abstract | Publisher Full Text\n\nMayr C, Bartel DP: Widespread shortening of 3'UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell. 2009; 138(4): 673–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEnciso-Mora V, Hosking FJ, Di Stefano AL, et al.: Low penetrance susceptibility to glioma is caused by the TP53 variant rs78378222. Br J Cancer. 2013; 108(10): 2178–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStacey SN, Sulem P, Jonasdottir A, et al.: A germline variant in the TP53 polyadenylation signal confers cancer susceptibility. Nat Genet. 2011; 43(11): 1098–103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiestner A, Tehrani M, Chiorazzi M, et al.: Point mutations and genomic deletions in CCND1 create stable truncated cyclin D1 mRNAs that are associated with increased proliferation rate and shorter survival. Blood. 2007; 109(11): 4599–606. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin Y, Li Z, Ozsolak F, et al.: An in-depth map of polyadenylation sites in cancer. Nucleic Acids Res. 2012; 40(17): 8460–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorris AR, Bos A, Diosdado B, et al.: Alternative cleavage and polyadenylation during colorectal cancer development. Clin Cancer Res. 2012; 18(19): 5256–66. PubMed Abstract | Publisher Full Text\n\nFu Y, Sun Y, Li Y, et al.: Differential genome-wide profiling of tandem 3' UTRs among human breast cancer and normal cells by high-throughput sequencing. Genome Res. 2011; 21(5): 741–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Q, Li W, Liu XS, et al.: A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol Cell. 2007; 27(3): 380–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndreu-Vieyra C, Lai J, Berman BP, et al.: Dynamic nucleosome-depleted regions at androgen receptor enhancers in the absence of ligand in prostate cancer cells. Mol Cell Biol. 2011; 31(23): 4648–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao JC, Fong KW, Jin HJ: FOXA1 acts upstream of GATA2 and AR in hormonal regulation of gene expression. Oncogene. 2016; 35(33): 4335–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunkley J, Maia TM, Ibarluzea N, et al.: Dataset 2 in: Androgen-dependent alternative mRNA isoform expression in prostate cancer cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15604.d212874"
}
|
[
{
"id": "36819",
"date": "05 Sep 2018",
"name": "Sebastian Oltean",
"expertise": [
"Reviewer Expertise alternative splicing",
"prostate cancer",
"diabetes (renal complications)"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nProstate cancer (PCa) is still a significant health problem in UK and across the world. Though a small minority of patients progress to aggressive forms, the absolute number is quite significant due to the high incidence of PCa among men. Therefore, investigation of molecular mechanisms of PCa progression is very important and will hopefully unravel novel therapeutic targets.\n\nAlternative splicing (AS) has been shown to occur in over 94% of genes in humans. It is therefore a crucial level of gene regulation and not surprisingly involved in virtually every physiological and pathological process. AS de-regulation has been implicated in many diseases, including cancer and in particular PCa, and interestingly, many times it has been shown to drive cancer pathology independently of transcription.\n\nSince androgens are main players in PCa, the idea of analysing global changes in AS in response to androgens is very welcome to the field. The authors found 10 times more AS isoforms regulated by androgens than previously reported in data from cell culture, most of them occurring through alternative promoter mechanism. They have confirmed and validated part of these changes. They have also analysed the isoforms changes between adenocarcinoma and normal tissues as well as during progression through the Gleason stages of PCa.\n\nThis is a very well thought and executed study, with may informative results. I have a suggestion for the discussion part:\n\none issue in global analysis of splice isoforms is which ones are causal (ie maintain and aggravate the phenotype) and which ones are just associated with the pathological progression; while a full answer to this would need experimental evidence on each individual splicing event, could the authors discuss 1-2 examples, if possible, where the changes at protein level (either sequence or expression level of a particular isoform) would hypothetically have a causal role\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "38286",
"date": "18 Sep 2018",
"name": "Cyril F. Bourgeois",
"expertise": [
"Reviewer Expertise Transcription and alternative splicing",
"transcriptomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 Munkley and colleagues identifies in a comprehensive manner novel alternative mRNA isoforms regulated by androgens. Interestingly most isoforms result from a choice between alternative promoters, suggesting that regulation takes place mostly at the transcriptional level, but they identified also a few alternative cassette exons and 3' ends. They show experimental validation for 17 isoforms. Beside increasing the number of identified genes in the context of androgen-treated prostate cancer LNCaP cells, the authors analysed the expression of those new isoforms in a large cohort of prostate tumours. They found the expression of some of the mRNA isoforms is positively correlated in the androgen-treated cell and in cancer versus normal samples, and find further correlation with the tumour grade and stage for 3 alternative isoforms.\nOverall this is an interesting work that clearly deserves to be published, as it reveals new potentially interesting target genes for prostate cancer. I have only a couple of comments/questions that may help to improve the strength of the manuscript.\nDid the authors try to experimentally validate the regulation of alternative isoforms for the 3 most interesting genes, i.e. OSBPL1A, CLK3 and TSC22D3, which is correlated to tumour stage ? As these new isoforms are predicted to alter the protein sequence, is it possible to discuss or predict what could be the impact of these modifications for these proteins, with regards to what is known about their function and/or in the context of prostate cancer?\nLooking at the RNA-seq profiles for the validated examples, it seems to me that in some cases, especially for RLN1 and RLN2, regulation of promoter choice correspond also to changes in the 3 end of the transcript (the peak seems to be shifted to the 3' end). Such examples may have escaped the in silico prediction, but can you make any comment on this ?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "37702",
"date": "04 Oct 2018",
"name": "Jennifer Byrne",
"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\nProstate cancer is a common cancer in men that is driven in part through deregulated androgen signalling. The importance of androgen inhibitors in prostate cancer therapy and the clinical challenges posed by the development of androgen-resistant disease both justify the detailed description of the effects of androgen treatment on gene transcription and alternative splicing in prostate cancer cells. In this sense, the analyses reported by Munkley and colleagues represent valuable additions to the literature. However, further explanation of the results presented would increase the reader’s ability to understand these results and their significance, and to identify which results should be prioritised for further research. I have therefore provided some specific suggestions to increase the accessibility of the data as they are currently presented. Specific comments\nThe genes shown in Tables 1, 2, 3 and 5 are not shown in alphabetical order. It is unclear how these genes are ranked and why they are shown in the orders displayed. It would be helpful for any groupings of genes to be more clearly displayed in these tables where this is relevant. It would be helpful to more clearly indicate which findings are novel and which are supported by the literature in Tables and/or Figures. In Table 1, a number of genes are shown in bold, but this is not explained. In Table 1, it would be helpful to annotate the isoform ID’s shown (columns towards the right side of Table). What does “comparable” mean here? It is challenging to show data for a large number of genes, most of which the authors will not be familiar with. However, in Figure 2, incorrect information is shown for the TPD52 gene (panel A). The PrLZ isoform is actually longer than the TPD52 isoform (through an extended N-terminal sequence specific to PrLZ), yet the sizes of these isoforms indicated in the diagram at the right have been switched (TPD52 is incorrectly shown to be the longer isoform). The authors should check whether this is an isolated error or whether other data for the TPD52 and PrLZ isoforms have been switched (for example in Figure 3). It would be helpful for Table 4 to include percentages as well as sample numbers so that readers can compare the composition of the TCGA PRAD cohort with other published cohorts. Analyses compared differential isoform expression in prostate cancer and normal tissue. The cohort included 497 prostate cancer patients, for which 52 had matched normal tissue (Table 5, Figure 3). I’ve assumed that these analyses compared transcript levels in the 497 prostate cancer cases with those in the 52 normal tissue cases. However, given that the 52 normal tissue cases had matched tumour samples available, were analyses conducted to compare the 52 matched cases? These analyses could be argued to be more robust through comparing matched samples, albeit in a smaller cohort. Table 5 should indicate the numbers of tumour and normal tissue samples compared. Some data in Tables 5, 6, and 6 are shown in bold, but this is not explained. I could not open Dataset 2. Could this be made available as a pdf file? All violin plots (Figures 3-5, also supplementary data) should specify the sample numbers compared, either below the X axis or in the figure legend if the same sample numbers apply to every plot shown.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1189
|
https://f1000research.com/articles/7-1188/v1
|
03 Aug 18
|
{
"type": "Research Article",
"title": "A pilot study on assessing the role of intra-operative Flow 800 vascular map model in predicting onset of vasospasm following micro vascular clipping of ruptured intracranial aneurysms",
"authors": [
"Sunil Munakomi",
"Deepak Poudel",
"Deepak Poudel"
],
"abstract": "Objective To ascertain the predictive value of intra-operative FLOW 800 vascular map model in predicting onset of post-operative clinical vasospasm and delayed cerebral ischemia among patients undergoing micro-vascular clipping of ruptured intracranial aneurysms. Material and methods A total of 40 patients were enrolled in the study and their variables such as age, World Federation of Neurological Surgeons (WFNS) grade at presentation, Computerized Tomography (CT) Fisher grading, location of the aneurysms, and Indocyanine Green (ICG) flow status were compared and statistically analyzed along with differences in Absorption Intensities (AI) and difference in time lag values obtained from the FLOW 800 vascular map studies for predicting onset of vasospasm. Results The Receiver Operating curve (ROC) of the model for predicting post-operative vasospasm was highest (.892) for difference in the AI followed by CT Fisher grading (.778), difference in time lag (.700) and WFNS grading (.699).Analysis of variance for different variables studied in our model for predicting vasospasm was significant for all except for age (.991) and the ICG flow through the parent vessel (.079).Multivariate analysis done for predicting the vasospasm was significant for all variables except for age (.869) and ICG main flow (.196) Conclusion\nOur study confirmed the role of FLOW 800 study model in predicting clinical vasospasm. Inclusion of this entity would therefore help in taking timely and correct therapeutics measures to ensure better patient outcomes.",
"keywords": [
"Aneurysm",
"vasospasm",
"ICG study",
"FLOW 800 study"
],
"content": "Introduction\n\nProgressive narrowing of vessels can occur in up to 70% of patients within 2 weeks of intra-cranial aneurysmal rupture, among which 30% develop delayed ischemic neurological deficits1,2. Although hypertension, hypervolemia and hemodilution (triple-H) therapy is commonly instituted to counteract this process, there is a paucity of evidence for recommending it for prophylactic purposes3. The American Heart Association/American Stroke Association guidelines also recommend induced hypertension targeted euvolemia for managing delayed cerebral ischemia (DCI)4. The World Federation of Neurological Surgeons (WFNS) grade of the patient at presentation and the location of epicenter of the subarachnoid bleed has shown to be positively correlated with the onset of vasospasm5,6. Since indocyanine green (ICG) flow studies are safe, easily applicable, readily reproducible and now routinely utilized technique during the microvascular clipping of aneurysms, addition of FLOW 800, an automated vascular map study generated by the microscope with the provision for quantitative study of flow velocities and time lag for appearance of the dye between relevant vessels, can help us segregate and outline groups at high-risk of developing post-operative vasospasm and form an evidence-based management algorithm for better therapeutic benefits and clinical outcomes.\n\n\nMethods\n\nA total of 40 patients who underwent microvascular clipping for ruptured intra-cranial aneurysms in the Department of Neurosurgery at Nobel Medical College and Teaching Hospital (Biratnagar, Nepal) between January 2017 and June 2018 were enrolled in the study. Those patients who refused to participate in the study or who left the hospital against medical advice during the course of the study were excluded from the study. Moreover, exclusions were also made in extreme circumstances wherein ICG study was not possible owing to intra-operative brain swelling or inability to visualize the relevant vessels owing to small operating field or close proximities between relevant vessels rendering difficulties in correctly specifying regions of interests (ROI).\n\nThe retrospective acquisition of data of these patients, with regards to their age, Glasgow coma scale (GCS) during initial presentation, WFNS grade, computerized tomography (CT) Fisher grading7, ICG flow status and FLOW 800 mapping (version 2.21) (extrapolated from a Pentero surgical microscope; Carl Zeiss Co., Germany) was conducted and the outcome in terms of occurrence of radiological and clinical vasospasm was analyzed.\n\nTo obtain ICG flow status, after the permanent clip was applied, the setting of the microscope was changed from the usual white light to the infrared mode. Next, ICG was injected and visualization was looked for within the desired vessels. This is a qualitative study, but we can record the time lag for appearance of dye. Following this, with the help of the FLOW 800 software, a vascular map was automatically generated from the IR 900 video data. There is also provisions for quantitative analysis of the flow dynamics in terms of average absorption intensity (AI) and time lag for appearance of the dye by selecting appropriate regions of interests (ROI) within the vessels.\n\nIn our study model of FLOW 800 mapping, the lower normal limit for normal difference in average absorption intensity (DfAI) between the parent and the branching vessel was taken at 50%. Similarly, the maximum upper limit for time-lag for appearance of the flow between the parent and the branching vessel was kept at 6 seconds (Figure 1–Figure 4). This time limit was extrapolated from the pooled results of ICG and FLOW 800 studies in patients who did not show any vasospasm in the post-operative period. The average intensity was chosen owing to its easy applicability, calculation and easy reproducibility, thereby minimizing scoring bias. Currently, there is a paucity in the literature with regards to normal values for AIs and time-lag for specific vessels following FLOW 800 study so as to form a specific reference standard. Values from these studies were compared to clinical findings and CT images, which served as the surrogate marker for vasospasm.\n\nTo minimize study bias and the post hoc effects, the study was blinded among the investigators who studied the clinical and radiological outcomes from those who analyzed the ICG and FLOW 800 vascular studies. All patients received an intravenous ICG bolus of 25 mg dissolved in 5 ml of 0.9% saline for the study. A Pentero surgical microscope (Carl Zeiss Co., Germany) was utilized for studying ICG as well as FLOW 800 vascular studies.\n\nPost-operatively, patients were given 60 mg (oral or via naso-gastric feeding tube) nimodipine every 4 hours and fluid management was titrated to achieve a central venous pressure of 8–12 mmHg, a hematocrit of 30–35% and a mean arterial pressure (MAP) of more than 20 mmHg greater than the preoperative MAP value. Clinical (presence of new onset neurological deficits in the post-operative period) and radiological (presence of features of ischemia or infarction) evidence of vasospasm were stringently assessed and recorded in the post-operative period.\n\nThe study was approved by the Institutional Review Committee (IRC) of Nobel Medical College and Teaching Hospital (NMCTH) (approval number 134/2018). Written informed consent was taken from the relatives or next of kin of the patients (owing to the poor neurological status of the patients and the emergent need for operative management) for their inclusions in the study and usage of their relevant clinical data for resource measures.\n\nData were recruited and analyzed using the SPSS version 16 software. Statistical analysis was done utilizing receiver operating curve (ROC) with area under curve (AUC) values, Analysis of variance (ANOVA) and multivariate logistic regression analysis along with logistic coefficient curve study among the considered variables applying vasospasm as the final outcome. No post hoc analysis was done beyond those factors pertaining to our study model.\n\n\nResults\n\nThe incidence of clinical vasospasm and delayed cerebral ischemia was 40% in this study. CT Fisher grade 3 was seen in 42.5% of cases and grade 4 in 37.5% of cases. The incidence of anterior communicating artery aneurysm was observed in 62.5% of cases.\n\nThe receiver operating curve (ROC) of the model for predicting post-operative vasospasm was highest (area under the curve (AUC)=0.892) for difference in the AI of FLOW 800 study followed by CT Fisher grading (AUC=0.778), difference in time lag in FLOW 800 (AUC=0.700) and WFNS grading (AUC=0.699) (Figure 5 and Table 1) thereby verifying the aim of our study for its routine inclusion as an intra-operative adjunct to ICG flow study.\n\nWFNS, World Federation of Neurological Surgeons; ICG, indocyanine green; AI, absorption intensity.\n\nANOVA for variables studied in our model for predicting vasospasm was significant for WFNS grade, CT Fisher grade, location of aneurysms, ICG flow through branching vessels, difference in flow velocities (DfAI) and time lag in dye appearance (DfTM) but not for age of the patients (P=0.991) and the ICG flow through the parent vessel (P=0.079) (Table 2). Multivariate analysis done for predicting the vasospasm was significant for all variables except for age (P=0.869) and ICG main flow (P=0.196) (Table 3).\n\ndf, degrees of freedom; WFNS, World Federation of Neurological Surgeons; ICG, indocyanine green; AI, absorption intensities.\n\ndf, degrees of freedom; WFNS, World Federation of Neurological Surgeons; ICG, indocyanine green; AI, absorption intensities.\n\nThe correlation and the coefficient curves between relevant variables and onset of post-operative vasospasm have been shown in the Figure 6 and Figure 7, which are also positively correlated with WFNS, CT Fisher grade, difference in velocities (DfAI) and time lag (DfTM) obtained from the FLOW 800 software in positively predicting the onset of post-operative vasospasm.\n\nThe specific variable is represented in in the ‘x’ axis and risk of vasospasm in the ‘y’ axis. DfAI, difference in average absorption intensity between parent and branching vessels; DfTM, difference in time lag for appearance of dye between parent and branching vessels.\n\nDfAI, difference in average absorption intensity between parent and branching vessels; DfTM, difference in time lag for appearance of dye between parent and branching vessels.\n\n\nDiscussion\n\nCerebral vasospasm has surpassed re-bleeding as the main cause of death and major disability among patients with ruptured intracranial aneurysms9. Inadvertent clipping of the parent vessels, its branches or the perforators causing compromised blood flow has been seen in 12–21% of cases, with subsequent occurrence of significant vasospasm in up to 10% of these cases10,11. Cerebral vasospasm still leads to mortality in 7% of patients, and to severe disability in another 7% of cases, even in experienced hands and the best neurosurgical care12–14. Visual inspection alone does not verify perfect placement of the clips and guarantee the patency of the relevant vessels10,15.\n\nIn this aspect, intraoperative angiography is still the gold standard in confirming the patency of the parent and their branching vessels along with their relevant perforators16,17. Intra-operative angiography has facilitated clip readjustment in almost 44% of cases. However, technical complexities, the risk of radiation hazards and major neurological complications (0.4–2.6%) preclude its frequent intra-operative application18. The application of intraoperative angiography also prolongs the operative time by almost 40 minutes19.\n\nMicroscope-integrated ICG angiography has shown to be a valuable alternative for assessing real-time intra-operative vascular mapping with minimal risks and hazards compared to intra-operative digital subtraction angiography20,21. ICG is a near-infrared (NIR) fluorescent dye; its absorption and emission peaks (805 and 835 nm, respectively) are ideally suited for usage in vascular studies since confounding absorption from other endogenous chromophores are minimal within these ranges22. The dye remains confined within the vascular compartment after binding to specific plasma proteins. The operating microscope can be integrated with a laser light source (wavelength within the ICG absorption band) and a camera that is capable of exciting, visualizing and subsequently transforming acquired ICG images into real-time vascular road maps21,23–25.\n\nFLOW 800 is microscope-integrated software capable of automatically reconstructing time-resolved quantitative analysis of ICG angiography studies. The resulting data can be displayed and stored as either time-to-arrival maps or time-intensity curves specific for selected regions of interests (ROI)26. The software reproduces a map graded by the averages of the AIs and the time lag for the same based on time to half maximal fluorescence within the selected ROIs27. ICG angiography has complication rate of less than 0.1%28,29. The discordance between ICG and intraoperative digital subtraction angiography (DSA) in previous series were reported to be in the range of 10–20%30–32. A recent study has shown to be correlating with post-operative DSA findings in almost 97% of cases23.\n\nThe main limitation of ICG angiography is that it can only study the vascular flow within the field of the operating microscope. Furthermore, blood clots, brain tissue and sometimes applied aneurysm clips can also preclude the proper visualization of the vessels, thereby requiring further adjustments of the microscope. The image quality can also be hindered by calcifications, atherosclerotic plaques or thromboses within the aneurysm33.\n\nThe limitations of having that microscope in the direct line-of-sight of the region of interest can be counteracted by simultaneous use of an endoscope25,34. A recent study suggested the use of intra-arterial ICG at a reduced dosage for better image quality with minimized time-lag between successive studies, owing to its rapid clearance, unlike intravenous studies wherein the ICG remains for around 10 minutes35. Micro-vascular Doppler study is a simple and readily applicable alternative for assessing the vascular patency36. A micro-Doppler study, however, lacks quantitative assessment and is also highly influenced on the insonation angles during its placement by the operator33,36. The sensitivity and specificity for determining the accurate flow are limited to 85–90%37,38. Intraoperative alterations, such as brain shift following retractor removal, probable induced late mechanical thrombosis and sometimes the Coanda effect induced by clips are not detectable by Doppler study19,39. Other more recent advancements involve somato-sensory evoked potentials (SSEPs)40,41. However, they have limitations in predicting ischemia outside the perimeter of the corticospinal tract25. The ‘ultimate, all-in-one’ diagnostic tool has not yet been designed42.\n\nProphylactic hypervolaemic therapy is unlikely to confer any additional benefit in minimizing vasospasm42. Treatment with triple-H therapy causes complications in 10–20% of patients, with pulmonary edema the most common adverse effect13,43. Moreover, there can be exacerbation of cerebral edema and an added risk of bleeding from unsecured and sometimes hemorrhagic infarctions in ischemic regions9.\n\nThe advantage of using FLOW 800 intra-operatively is that it facilitates the timely undertaking of corrective measures, such as readjustment of the clips. Calcium overloading, which triggers phosphorylation of the contractile proteins of the arterial smooth muscles thereby leading to vasospasm, can be minimized by using topical sodium nitroprusside (SNP)44,45. Moreover, complications such as pulmonary edema due to aggressive medical management can be minimized by placement of a swan gauge catheter to keep the pulmonary capillary wedge pressure below the colloid oncotic pressure (COP)46. A serial bedside transcranial Doppler (TCD) study can be utilized in these patients to assess changes in the flow velocities and accordingly modifying the treatment algorithm47. Moreover, relevant rescue interventions, such as hemodynamic augmentation or intra-arterial vascular manipulations can be timely initiated for prevention as well as management of refractory vasospasms48.\n\nThere are some limitations in our study. The results of our study were derived from sample size of only 40 patients, and therefore needs further confirmation from multi-centric randomized control trials with the inclusion of larger cohorts. There is also a prerequisite for an ICG-integrated operating microscope with added facilities for FLOW 800 vascular study. In cases of the repeated use of ICG, there may be bias in the extrapolated results of FLOW 800 owing to false fluorescence from the retained dye. There is also provision for inter-rater bias when selecting the appropriate ROIs among vessels in close proximities, thereby increased tendency for false results. However, this improves with practice since there is not a steep learning curve.\n\n\nConclusion\n\nThe addition of FLOW 800 quantitative mapping following a routinely performed ICG study can precisely help determine patients at high risk of post-operative vasospasm. Timely actions, such as readjusting clips, the local administration of drugs or aggressive medical or interventional management can be undertaken. Additional measures, such as the placement of a swan gauge catheter to minimize complications and TCD for continuous monitoring of these patients can be utilized for better clinical outcomes. Further studies are recommended for confirming the role of FLOW 800 software as a valuable adjunct to intra-operative ICG vascular studies.\n\n\nData availability\n\nDataset 1. Demographic information and the results of each diagnostic technique performed for each patient. https://doi.org/10.5256/f1000research.15627.d2128758.",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nDorsch NW, King MT: A review of cerebral vasospasm in aneurysmal subarachnoid haemorrhage part I: incidence and effects. J Clin Neurosci. 1994; 1(1): 19–26. PubMed Abstract | Publisher Full Text\n\nWeir B, Grace M, Hansen J, et al.: Time course of vasospasm in man. J Neurosurg. 1978; 48(2): 173–8. PubMed Abstract | Publisher Full Text\n\nAwad IA, Carter LP, Spetzler RF, et al.: Clinical vasospasm after subarachnoid hemorrhage: response to hypervolemic hemodilution and arterial hypertension. Stroke. 1987; 18(2): 365–72. PubMed Abstract\n\nConnolly ES Jr, Rabinstein AA, Carhuapoma JR, et al.: Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2012; 43(6): 1711–37. PubMed Abstract | Publisher Full Text\n\nCharpentier C, Audibert G, Guillemin F, et al.: Multivariate analysis of predictors of cerebral vasospasm occurrence after aneurysmal subarachnoid hemorrhage. Stroke. 1999; 30(7): 1402–1408. PubMed Abstract\n\nNomura Y, Kawaguchi M, Yoshitani K, et al.: Retrospective analysis of predictors of cerebral vasospasm after ruptured cerebral aneurysm surgery: influence of the location of subarachnoid blood. J Anesth. 2010; 24(1): 1–6. PubMed Abstract | Publisher Full Text\n\nFisher CM, Kistler JP, Davis JM: Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980; 6(1): 1–9. PubMed Abstract | Publisher Full Text\n\nMunakomi S, Poudel D: Dataset 1 in: A pilot study on assessing the role of intra-operative Flow 800 vascular map model in predicting onset of vasospasm following micro vascular clipping of ruptured intracranial aneurysms. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15627.d212875\n\nSen J, Belli A, Albon H, et al.: Triple-H therapy in the management of aneurysmal subarachnoid haemorrhage. Lancet Neurol. 2003; 2(10): 614–621. PubMed Abstract | Publisher Full Text\n\nMacdonald RL, Wallace MC, Kestle JR: Role of angiography following aneurysm surgery. J Neurosurg. 1993; 79(6): 826–32. PubMed Abstract | Publisher Full Text\n\nAlexander TD, Macdonald RL, Weir B, et al.: Intraoperative angiography in cerebral aneurysm surgery: a prospective study of 100 craniotomies. Neurosurgery. 1996; 39(1): 10–7; discussion 17-8. PubMed Abstract | Publisher Full Text\n\nKivisaari RP, Porras M, Ohman J, et al.: Routine cerebral angiography after surgery for saccular aneurysms: is it worth it? Neurosurgery. 2004; 55(5): 1015–24. PubMed Abstract | Publisher Full Text\n\nKassel NF, Torner JC, Haley EC Jr, et al.: The International Cooperative Study on the Timing of Aneurysm Surgery. Part 1: Overall management results. J Neurosurg. 1990; 73(1): 18–36. PubMed Abstract | Publisher Full Text\n\nPickard JD, Murray GD, Illingworth R, et al.: Effect of oral nimodipine on cerebral infarction and outcome after subarachnoid haemorrhage: British aneurysm nimodipine trial. BMJ. 1989; 298(6674): 636–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyer B, Urbach H, Schaller C, et al.: Immediate postoperative angiography after aneurysm clipping--implications for quality control and guidance of further management. Zentralbl Neurochir. 2004; 65(2): 49–56. PubMed Abstract | Publisher Full Text\n\nImizu S, Kato Y, Sangli A, et al.: Assessment of incomplete clipping of aneurysms intraoperatively by a near-infrared indocyanine green-video angiography (Niicg-Va) integrated microscope. Mini Inva Neurosurg. 2008; 51(4): 199–203. PubMed Abstract | Publisher Full Text\n\nMartin NA, Bentson J, Viñuela F, et al.: Intraoperative digital subtraction angiography and the surgical treatment of intracranial aneurysms and vascular malformations. J Neurosurg. 1990; 73(4): 526–33. PubMed Abstract | Publisher Full Text\n\nde Oliveira JG, Beck J, Seifert V, et al.: Assessment of flow in perforating arteries during intracranial aneurysm surgery using intraoperative near-infrared indocyanine green videoangiography. Neurosurgery. 2007; 61(3 Suppl): 63–72; discussion 72–3. PubMed Abstract | Publisher Full Text\n\nStendel R, Pietilä T, Al Hassan AA, et al.: Intraoperative microvascular Doppler ultrasonography in cerebral aneurysm surgery. J Neurol Neurosurg Psychiatry. 2000; 68: 29–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRaabe A, Beck J, Gerlach R, et al.: Near-infrared indocyanine green video angiography: a new method for intraoperative assessment of vascular flow. Neurosurgery. 2003; 52(1): 132–9; discussion 139. PubMed Abstract | Publisher Full Text\n\nRaabe A, Nakaji P, Beck J, et al.: Prospective evaluation of surgical microscope-integrated intraoperative near-infrared indocyanine green videoangiography during aneurysm surgery. J Neurosurg. 2005; 103(6): 982–9. PubMed Abstract | Publisher Full Text\n\nScerrati A, Della Pepa GM, Conforti G, et al.: Indocyanine green video-angiography in neurosurgery: a glance beyond vascular applications. Clin Neurol Neurosurg. 2014; 124: 106–13. PubMed Abstract | Publisher Full Text\n\nMa CY, Shi JX, Wang HD, et al.: Intraoperative indocyanine green angiography in intracranial aneurysm surgery: Microsurgical clipping and revascularization. Clin Neurol Neurosurg. Elsevier; 2009; 111(10): 840–846. PubMed Abstract | Publisher Full Text\n\nÖzgiray E, Aktüre E, Patel N, et al.: How reliable and accurate is indocyanine green video angiography in the evaluation of aneurysm obliteration? Clin Neurol Neurosurg. 2013; 115(7): 870–8. PubMed Abstract | Publisher Full Text\n\nMielke D, Malinova V, Rohde V: Comparison of intraoperative microscopic and endoscopic ICG angiography in aneurysm surgery. Neurosurgery. 2014; 10 Suppl 3: 418–25; discussion 425. PubMed Abstract | Publisher Full Text\n\nKamp MA, Slotty P, Turowski B, et al.: Microscope-integrated quantitative analysis of intraoperative indocyanine green fluorescence angiography for blood flow assessment: first experience in 30 patients. Neurosurgery. 2012; 70(1 Suppl Operative): 65–73; discussion 73–4. PubMed Abstract | Publisher Full Text\n\nAcerbi F, Cavallo C, Ferroli P: Letter: Intraoperative Assessment of Blood Flow with Quantitative Indocyanine Green Videoangiography: The Role for Diagnosis of Regional Cerebral Hypoperfusion. Neurosurgery. 2016; 78(2): E310–E312. PubMed Abstract | Publisher Full Text\n\nRegillo CD: The present role of indocyanine green angiography in ophthalmology. Curr Opin Ophthalmol. 1999; 10(3): 189–96. PubMed Abstract | Publisher Full Text\n\nWoitzik J, Horn P, Vajkoczy SP: Intraoperative control of extracranial-intracranial bypass patency by near-infrared indocyanine green videoangiography. J Neurosurg. 2005; 102(4): 692–8. PubMed Abstract | Publisher Full Text\n\nCaplan JM, Sankey E, Yang W, et al.: Impact of indocyanine green videoangiography on rate of clip adjustments following intraoperative angiography. Neurosurgery. 2014; 75(4): 437–43; disucssion 444. PubMed Abstract | Publisher Full Text\n\nWashington CW, Zipfel GJ, Chicoine MR, et al.: Comparing indocyanine green videoangiography to the gold standard of intraoperative digital subtraction angiography used in aneurysm surgery. J Neurosurg. 2013; 118(2): 420–7. PubMed Abstract | Publisher Full Text\n\nDoss VT, Goyal N, Humphries W, et al.: Comparison of intraoperative indocyanine green angiography and digital subtraction angiography for clipping of intracranial aneurysms. Interv Neurol. 2015; 3(3–4): 129–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCui H, Wang Y, Yin Y, et al.: Role of intraoperative microvascular Doppler in the microsurgical management of intracranial aneurysms. J Clin Ultrasound. 2011; 39(1): 27–31. PubMed Abstract | Publisher Full Text\n\nJoo SP, Kim TS: The Clinical Importance of Perforator Preservation in Intracranial Aneurysm Surgery: An Overview with a Review of the Literature. Chonnam Med J. 2017; 53(1): 47–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuroda K, Kinouchi H, Kanemaru K, et al.: Intra-arterial injection fluorescein videoangiography in aneurysm surgery. Neurosurgery. 2013; 72(2 Suppl Operative): ons141–150; discussion ons150. PubMed Abstract | Publisher Full Text\n\nMarchese E, Albanese A, Denaro L, et al.: Intraoperative microvascular Doppler in intracranial aneurysm surgery. Surg Neurol. 2005; 63(4): 336–342; discussion 342. PubMed Abstract | Publisher Full Text\n\nJanjua N, Mayer SA: Cerebral vasospasm after subarachnoid hemorrhage. Curr Opin Crit Care. 2003; 9(2): 113–19. PubMed Abstract\n\nLysakowski C, Walder B, Costaza MC: Transcranial Doppler versus angiography in patients with vasospasm due to a ruptured cerebral aneurysm: a systematic review. Stroke. 2001; 32(10): 2292–98. PubMed Abstract\n\nRobinson JL, Roberts A: Operative treatment of aneurysms and Coanda effect: a working hypothesis. J Neurol Neurosurg Psychiatry. 1972; 35(6): 804–809. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKato Y, Sano H, Nagahisa S: Endoscope-assisted microsurgery for cerebral aneurysms. Minim Invasive Neurosurg. 2000; 43(2): 91–7. PubMed Abstract | Publisher Full Text\n\nManninen PH, Patterson S, Lam AM: Evoked potential monitoring during posterior fossa aneurysm surgery: a comparison of two modalities. Can J Anaesth. 1994; 41(2): 92–7. PubMed Abstract | Publisher Full Text\n\nLennihan L, Mayer SA, Fink ME, et al.: Effect of hypervolemic therapy on cerebral blood flow after subarachnoid hemorrhage: a randomized controlled trial. Stroke. 2000; 31(2): 383–91. PubMed Abstract\n\nSolenski NJ, Haley EC Jr, Kassell NF: Medical complications of aneurysmal subarachnoid hemorrhage: a report of the multicenter, cooperative aneurysm study. Participants of the Multicenter Cooperative Aneurysm Study. Crit Care Med. 1995; 23(6): 1007–17. PubMed Abstract | Publisher Full Text\n\nThomas JE, Rosenwasser RH, Armonda RA: Safety of intrathecal sodium nitroprusside for the treatment and prevention of refractory cerebral vasospasm and ischemia in humans. Stroke. 1999; 30(7): 1409–16. PubMed Abstract | Publisher Full Text\n\nThomas JE, Nemirowski A, Zelman V, et al.: Rapid reversal of endothelin-1-induced cerebral vasoconstriction by intrathecal administration of nitric oxide donors. Neurosurgery. 1997; 40(6): 1245–9. PubMed Abstract | Publisher Full Text\n\nCorsten L, Raja A, Guppy K, et al.: Contemporary management of subarachnoid hemorrhage and vasospasm: The UIC experience. Surg Neurol. 2001; 56(3): 140–8; discussion 148–50. PubMed Abstract | Publisher Full Text\n\nKincaid MS, Souter MJ, Treggiari MM, et al.: Accuracy of transcranial Doppler ultrasonography and single-photon emission computed tomography in the diagnosis of angiographically demonstrated cerebral vasospasm. J Neurosurg. 2009; 110(1): 67–72. PubMed Abstract | Publisher Full Text\n\nDurrant JC, Hinson HE: Rescue therapy for refractory vasospasm after subarachnoid hemorrhage. Curr Neurol Neurosci Rep. 2015; 15(2): 521. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "37112",
"date": "29 Aug 2018",
"name": "Guru Dutta Satyarthee",
"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\nIntra-operative Neurosurgical Armamentarium to deal with vasospasm development following aneurysmal subarachnoid hemorrhage.\nAuthors reported utility of intra-operative FLOW 800 vascular map model with indigo cyanine [ICG] fluoroscopic angiographic images obtained with microscope with appropriate facility in predicting onset of clinical cerebral vasospasm during post-operative period and delayed cerebral ischemic deficit in a cohort of forty patients suffering with aneurysmal subarachnoid hemorrhage and managed microsurgical clipping. Factors predicting risk of vasospasm onset included parameters obtained with FLOW 800 vascular map model like the difference in the absorption intensities and time lag.\nHowever, major disadvantages of the vascular map model include the cost factor as advanced microscope along with appropriate software is a pre-requisite, in addition to the availability of dye, and part of vessel evaluated is very limited to microscope field with direct line of-sight of target vessels but affected by obstructed microscope field by presence of blood clots and aneurysm clips. The anatomical factor of vessel and aneurysm can also impair image quality of atherosclerotic plaques, aneurysmal sac calcifications or thrombosis. Authors rightly advocated vascular model with ICG angiography can act as one of adjunct for aneurysm surgery in neurosurgical armamentarium besides vascular Doppler, endoscope, with capability of predicting vasospasm and appropriate measure can be applied right from intraoperative period like drainage of hematoma, installation of papavarin, Omaya reservoir and installation of papavarin in the post-operative period besides stellate ganglion block, and hemodynamic interventions can definitely aid in improving outcome. However, the size of the cohort sample was modest.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36812",
"date": "04 Sep 2018",
"name": "Lekhjung 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\nI consider this paper as an excellent report and is an important addition to the existing information on the subject. I personally adored the concept of exploring better markers to identify patients of ruptured intracranial aneurysm at high risk of postoperative clinical vasospasm and delayed cerebral ischemia and the promising findings of components of ICG flow status as one of the important components and predictors as tested by various statistical methods. I hope the author intend to study the concept in larger number of patients with ruptured intracranial aneurysm.\nIn addition I would like to add few comments:\n\nAbstract:\n\nIn the methodology section, the authors can mention ….. (Last line) ….predicting onset of vasospasm and delayed cerebral ischemia, as per the objective of the study. In the conclusion section, it may be better to replace word “confirmed” in “our study confirmed…..” I feel this study has found/demonstrated the role FLOW 800…… rather than establishing it.\nIntroduction:\nThis section is well written with the emphasis on the potential role of information generated by FLOW 800 vascular mapping. However the last sentence appears to be long. The authors could break up the sentences with pertinent references so that it is easily understood.\n\n“Indocyanine green (ICG) flow studies are safe, easily applicable, readily reproducible and now routinely used during the microvascular clipping of aneurysms. [Reference?] A note on current status of ICG studies in predicting vasospasm would be better here.\nMethods:\nThe author mentions CLINICAL VASOSPASM in introduction and in methodology section “radiological vasospasm” is also mentioned. Kindly elaborate. It is wiser to mention why ANOVA was chosen when multivariate analysis is also done in this case. Why were these tests chosen and how would they help in interpretation in this test.\nDiscussion:\nI think the authors should discuss more about their findings with appropriate reference rather than reviewing about the machine and methods.\nConclusion:\nWith the limitation explained by the authors, this study obviously may not precisely identify patients at high risk of post-operative vasospasm, so it may be prudent to conclude by remaining confined to the objective and the the relevant result of the study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36816",
"date": "13 Sep 2018",
"name": "Amit Agrawal",
"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 present study the authors share there preliminary experience with automated vascular mapping to identify sub-group of patient who are high-risk of developing post-operative vasospasm following aneurysmal surgery for spontaneous subarachnoid hemorrhage. It will be beneficial if an evidence based algorithm can be generated and can be used to improve clinical outcomes. Although the article is well written, however the study includes a small number of patients; data were acquired retrospectively and does not include all consecutive cases. As authors have suggested transcranial Doppler study (TCD) at frequent interval can be used as a non-invasive tool to determine the velocity in the cerebral vessels and can be correlated with FLOW 800 studies. Additionally we need to determine the cost-effectiveness of the procedures particularly in settings where the resources are limited. To better understand the benefits of FLOW 800 we need a larger and long term experience.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1188
|
https://f1000research.com/articles/7-1187/v1
|
03 Aug 18
|
{
"type": "Research Article",
"title": "Clinical characteristics and outcomes of extremity soft tissue sarcomas at Thailand’s national tertiary referral center",
"authors": [
"Jomjit Chantharasamee",
"Nichanan Tanapathomsinchai",
"Suvimol Niyomnaitham",
"Tauangtham Anekpurithanang",
"Nichanan Tanapathomsinchai",
"Suvimol Niyomnaitham",
"Tauangtham Anekpurithanang"
],
"abstract": "Background: Due to the heterogeneity of soft tissue sarcomas, conflicting data have been reported regarding treatment outcomes. Moreover, factors such as histologic subtype, treatment compliance, and treatment tolerability may influence outcomes of treatment. The aim of this study was to investigate the clinical characteristics, prognostic factors, survival outcomes, and outcomes of treatment in patients with extremity soft tissue sarcoma who underwent complete tumor resection at Siriraj Hospital. Methods: Medical records of patients with extremity soft tissue sarcoma underwent total tumor resection at Siriraj Hospital during the January 2007 to December 2016 study period were included. Data collected included demographic data, tumor characteristics, type of surgery, and postsurgical intervention. Regimen, dose, cycle, and treatment compliance data were collected in patients who received adjuvant chemotherapy. Results: A total of 58 patients with a median age of 53.5 years were included. In total, 13 patients received adjuvant chemotherapy. Tumor grade 3 with size >10 cm was a baseline factor found to be a significant predictor of unfavorable disease-free survival (DFS) (p<0.001). Median DFS was 56.2 months in the chemotherapy group and 20.5 months in the non-chemotherapy group (p=0.29). Median OS was 77.2 months in the chemotherapy group and 66.6 months in the non-chemotherapy group (p=0.24). A total of 20% of patients in chemotherapy group developed grade 3-4 hematologic toxicity. Conclusion: Tumor grade 3 >10 cm was the only baseline characteristic found to be a significant predictor of unfavorable DFS in univariate analysis. No conclusive benefit of adjuvant chemotherapy in term of DFS and OS.",
"keywords": [
"extremity soft tissue sarcoma",
"disease-free survival",
"overall survival",
"chemotherapy",
"outcomes"
],
"content": "Background\n\nSoft tissue sarcoma is a rare tumor that represents less than 1% of all malignancies1. Complete surgical resection with adequate margin is the mainstay of treatment for patients with this condition. The prognosis of patients with soft tissue sarcoma varies according to histologic subtype, tumor size, tumor location, margin status, and tumor depth2–6. Given the heterogeneity of soft tissue sarcomas, it is difficult to determine the value of prognostic factors and the survival outcomes among the different histological subtypes. In term of adjuvant treatment, the impact of adjuvant chemotherapy treatment on survival in soft tissue sarcoma has been investigated in several phase 3 randomized controlled trials7–10 Although those studies found adjuvant chemotherapy effective for improving outcome in soft tissue sarcoma, the efficacy of this adjuvant treatment continues to be debated. Data from meta-analyses and large randomized trials reported by the European sarcoma group indicated postoperative chemotherapy improved relapse-free survival (RFS) in patients with extremity soft tissue sarcoma, but data regarding improvement in OS were conflicting7–14\n\nThe aim of this study was to investigate the clinical characteristics, prognostic factors, survival outcomes, and outcomes of treatment in extremity soft tissue sarcoma patients who underwent complete tumor resection at Siriraj Hospital.\n\n\nMethods\n\nMedical records of patients diagnosed with extremity soft tissue sarcoma and treated with total tumor resection at Siriraj Hospital (Bangkok, Thailand) during the January 2007 to December 2016 study period were included. Siriraj Hospital is Thailand’s largest university-based national tertiary referral center. The protocol for this study was approved by the Siriraj Institutional Review Board, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand (approval no. 835/2559).\n\nPatients aged ≥15 years who were diagnosed with non-metastatic extremity soft tissue sarcoma with tumor >5 cm or World Health Organization tumor grade 3 of any size15 were eligible for inclusion. Patients meeting any one of the following criteria were excluded: distant metastasis prior to resection date, soft tissue sarcoma at non-extremity site and Ewing sarcoma. Collected data included age, gender, ECOG16, tumor size, histological subtype, site, type of surgery, margin status, and postsurgical intervention. Chemotherapy agent/regimen, dose, cycle, and treatment compliance data were also collected in patients who received adjuvant chemotherapy. Toxicities were evaluated according to the Common Terminology Criteria for Adverse Events version 3.0. Regarding survival analysis, disease-free survival (DFS) was defined as the duration from curative surgery to the documented date of relapse. Overall survival (OS) was defined as the duration from diagnosis to either i) the date of death from any cause registered in the Civil Registration Database; or, ii) the censor date of December 31, 2016.\n\nPatient characteristics were described using descriptive statistics. Categorical variables are reported as frequency and percentage, and continuous variables are presented as median and range. Kaplan-Meier survival analysis was performed to estimate DFS and OS. The log-rank test was used for univariate analysis to evaluate the effect of baseline characteristics on survival. The Cox proportional hazard model was used for multivariate analysis. All statistical analyses were performed using SPSS Statistics version 21.0 (SPSS, Inc., Chicago, IL, USA). p<0.05 was regarded as being statistically significant.\n\n\nResults\n\nA total of 58 patients (25 males, 33 females) with a median age of 53.5 years were included. The histologic subtypes were undifferentiated pleomorphic sarcoma (UPS) (26/58, 44.8%), liposarcoma (8/58, 13.8%), synovial sarcoma (5/58, 8.6%), and malignant peripheral nerve sheath tumor (5/58, 8.6%). In total, 50 of 58 patients had an ECOG score of 0-1, but 8 patients in the non-chemotherapy group had ECOG of 2 or more. Overall median tumor size was 9 cm. Median tumor size was also 9 cm for both the chemotherapy and non-chemotherapy groups. Patient demographic and clinical characteristics are given in Table 1.\n\nUPS, undifferentiated pleomorphic sarcoma.\n\nA total of 68% of patients received adjuvant treatment after curative resection (81%, 47/58, were limb-sparing surgery and 19%, 11/58 were amputation). Adjuvant radiation therapy was given to 67% (37/58) of all patients. Among the 13 patients in the adjuvant chemotherapy group, 69% (9/13) were treated with doxorubicin and ifosfamide. Other regimens included cisplatin plus doxorubicin in two patients with UPS (in whom it could not be identified whether from bone or soft tissue in origin), and in one patient with leiomyosarcoma (for whom the rationale for this suboptimal usage), and cyclophosphamine–doxorubicin–vincristine regimen in one patient with rhabdomyosarcoma (Table 2). Primary prophylaxis by filgrastim was not routinely administered to patients in this study.\n\nA total of 92% (12/13) of patients in the adjuvant chemotherapy group completed the planned 4 or 6 cycles of chemotherapy. One of 13 patients received only 1 cycle of chemotherapy and was lost to follow-up. Median dose of doxorubicin and ifosfamide was 56 mg/m2 (range: 37.5–73) and 4.8 g/m2 (range: 3.6–5.3), respectively. The reason for chemotherapy dose reduction in most cases was hematological toxicity. A total of 53% (7/13) of patients in the chemotherapy group required dose reduction, and 30% (4/13) needed to delay chemotherapy.\n\nFrom 69 cycles of adjuvant chemotherapy, the incidence of treatment-related adverse events of any grade was 93% (66/69). No treatment-related death was observed in this study. A total of 23% (16/69) of patients developed grade 3–4 hematologic toxicity. Grade 3 mucositis developed in one patient. Febrile neutropenia occurred in two patients (15.4%). In total, two patients experienced non-hematologic toxicities that led to dose reduction or treatment discontinuation. No cardiac toxicity was reported in patients who received doxorubicin (Table 3).\n\nAt a median follow-up of 51 months, 46.7% (21/45) and 30.87% (4/13) of patients in the non-chemotherapy group and chemotherapy group had disease recurrence, respectively (p=0.38). Distant metastasis occurred in 80% (20/25) of all relapse patients. The most common site of metastasis was the lung (15/20) (Table 4). Median DFS in all patients was 33 months. Median DFS in the chemotherapy group was 56.2 months versus 20.5 months in the non-chemotherapy group (p=0.29) (Figure 1). Regarding OS, only 20% of patients had died by the censor date of December 21, 2016. The median OS in the entire 58 patient study population was 74.8 months, with a cumulative survival at 4 years of 84%. Median OS in the chemotherapy group was slightly shorter than in the non-chemotherapy group (66.6 vs 77.2 months, respectively; p=0.24) (Figure 2). Cumulative survival at 4 years in the chemotherapy group was 92% vs 81% in the non-chemotherapy group.\n\nThe treatments given after disease relapse are shown in Table 5. Among the 25 patients with disease relapse, 18 (72%) of them proceeded to post-relapse interventions. There were three patients with distant recurrence that exhibited no evidence of disease (NED) after subsequent treatment. Of these three patients, one underwent pulmonary resection, and the other two achieved complete remission after receiving palliative chemotherapy.\n\nKaplan-Meier survival analysis was performed to estimate DFS and OS, the log-rank test was used for univariate analysis to evaluate the effect of baseline characteristics on survival, and the Cox proportional hazard model was used for multivariate analysis (Table 6). Shorter duration from curative surgery to first date of chemotherapy administration was identified as a significant favorable prognostic factor for DFS (p=0.02). Median duration from surgery to first chemotherapy administration was close to being significantly shorter in the non-relapse group than in the relapse group (28 vs 45.5 days; p=0.059). Tumor grade 3 with size >10 cm was a baseline factor found to be a significant predictor of unfavorable DFS (p<0.001); median DFS in this group was 9.4 months versus 36.4 months in the others. No other independent factors, including histologic subtype, age, margin status, ECOG, or type of surgery, were significantly associated with DFS. In subgroup analysis of patients who received chemotherapy, no significant difference in DFS was observed relative to any baseline characteristics. Tumor grade 3 with size >10 cm showed a trend toward worse OS, but the association did not achieve statistical significance (hazard ratio: 2.8, 95% CI: 0.8-9.9; p=0.09) (Figure 3). No clinical characteristics were found to be significantly correlated with either DFS or OS in multivariate analysis.\n\n*p-value<0.05 indicates statistical significance **By linear-regression method. HR, hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group.\n\nKaplan-Meier curves of overall survival (OS) (A) and disease-free survival (DFS) (B), separated tumor grade 3/>10 cm size status.\n\n\nDiscussion\n\nThe aim of this study was to investigate the clinical characteristics, prognostic factors, survival outcomes, and outcomes of treatment in extremity soft tissue sarcoma patients who underwent complete tumor. We included only patients with high-risk extremity soft tissue sarcoma in order to minimize the heterogeneity of the study population.\n\nRegarding baseline characteristics, the median age of 53.5 years in our study was much higher than the median age of 46 years reported in the European Organization for Research and Treatment of Cancer (EORTC) and Italian Sarcoma Group (ISG) trials9,13, but similar to a study by Liu et al. from Taiwan18. The median tumor size of 9 cm found in this study was slightly smaller than the median tumor size of 10 cm reported from the ISG trial9. However, the median tumor size of 9 cm in both the chemotherapy and non-chemotherapy groups in this study were larger than the median tumor size of 8.6 cm and 7.5 cm in the chemotherapy and non-chemotherapy groups in the EORTC trial, respectively12. The predominant histological subtype in this study was UPS (26/58, 45%), which is higher than the 29% and 18% rates reported from the ISG trial9, EORTC trial13, respectively, but similar to 27% rate reported by Roger et al.19 Synovial sarcoma in this study was 8.9% (5/58), which is considerably lower than the 25% rate in the ISG and the 14% rate in the EORTC trial9,13. A total of 87% (50/58) of patients had negative surgical margin, which was higher than 37%, 38% and 37% reported in the EORTC, Taiwan and Korean studies, respectively13,18,20.\n\nFor tumor grading, subjects in our study had more tumor grade 1 than patients in the ISG trial (7/58; 12% vs 0%)9. In our analysis, we also attempted to identify baseline characteristics that portend an unfavorable prognosis. In previous studies, tumor size and grading were reported to be prognostic markers, especially if the tumor was larger than 10 cm or larger than 5 cm and of a high grade2–6. This study revealed that tumor grade 3 >10 cm was associated with significantly shorter DFS. Shorter duration from curative surgery to commencement of chemotherapy administration was found to be a significant favorable prognostic factor for DFS (p=0.02).\n\nMoreover, the median time from surgery to first chemotherapy administration in non-relapse patients was nearly significantly shorter than in relapse patients (28 vs. 45.5 days; p=0.059). This 4-week period is comparable to the reported 4-week time frame within which chemotherapy was commenced in several previous studies7–14. We found tumor grade 3 with size >10 cm showed a non-significant trend toward association with worse OS.\n\nFrom our survival analysis, the median follow-up period was 51 months, which is similar to durations of follow-up reported in previous studies7–14,18–20. Overall median DFS was 33 months in our study population. Median DFS in this study was 56 months in the chemotherapy group and 20.5 months in the non-chemotherapy group, which are longer than the 48-month and 16-month DFS durations, respectively, reported in the ISG trial9. One possible explanation for this difference between studies is that our study consisted of patients with lower risk of tumor grade 3 and size >5 cm for relapse, as compared with 100% of patients in the ISG trial9. In this study, 8.6% (5/58) of patients had tumors smaller than 5 cm, and 12% (6/50) had tumor grade 1. Only 24% (12/50) of patients in our study had tumor grade 3 that was larger than 10 cm, which is much lower than the 54% reported from the ISG trial9.\n\nWhen we compared the median DFS in this study with the 90-month and 78-month DFS of EORTC study in the chemotherapy and non-chemotherapy groups, respectively13, the shorter survival in our study was probably due to greater heterogeneity among patients in the EORTC study. This heterogeneity included 27% with small tumors, less cases with tumor grade 3 (46% in EORTC vs. 30/50; 60% in our study), various primary sites, and median tumor size (7.5 cm and 8.6 cm in the chemotherapy and non-chemotherapy groups, respectively) that was smaller than the median tumor size by group in our study. The 4-year cumulative OS in the chemotherapy and non-chemotherapy groups in this study was 92% and 81%, respectively, compared to 69% and 50%, respectively, in the ISG trial9.\n\nThe efficacy of adjuvant chemotherapy was also evaluated. We proceeded with our analysis despite potential limitations, which included those inherent to retrospective studies, and unbalanced baseline characteristics and sample sizes between the chemotherapy and non-chemotherapy groups, which could result in some degree of selection bias relative to several clinical characteristics. The median age of patients in the chemotherapy group was lower than that in the non-chemotherapy group (42 vs 55 years, respectively). Moreover, chemotherapy group patients had an ECOG performance status score of 0-1, while all ECOG 2-3 patients were put into the non-chemotherapy group. As such, both age and ECOG score may influence physician decision. A total of 13 patients (22%) in our study received adjuvant chemotherapy compared to 34% reported in Taiwan, which had similar baseline patient characteristics to our study18. On the contrary, a report from Korea demonstrated only 7.3% of patients received adjuvant chemotherapy20.\n\nFor survival analysis compared between patients who received and who did not receive adjuvant chemotherapy, we found no statistically significant difference for either DFS or OS between groups (56 vs 20.5 months, respectively; p=0.29). It should be noted that this study was a retrospective review. There are, therefore, many confounding factors involving the benefits of chemotherapy in this small sample size of patients that received chemotherapy (13/58; 22%). Secondly, chemotherapy may not improve survival benefit, even in high-risk population as established in previous studies, including a long-term (5-year) follow-up evaluated by an Italian Sarcoma Group study and a Cochrane Review of Sarcoma Meta-analysis Collaboration10,11,13.\n\nThe incidence of distant relapse in this study was 34.4% (20/58), as compared to the relapse rate of 47% (49/104) in patients studied in the ISG trial9. The 4-year cumulative incidence of distant metastasis in ISG study in the chemotherapy and non-chemotherapy groups was 44% and 45%, respectively, compared to 31% and 37% in our study. Post-relapse treatment was also evaluated. In this study, 18/25 (72%) patients were commenced on interventions that included surgery, radiation, and/or chemotherapy.\n\nCompliance with adjuvant chemotherapy was also evaluated and reported. Most patients were given doxorubicin and ifosfamide, except in specific subtypes (Table 2). The median dose of doxorubicin and ifosfamide was 56 mg/m2/cycle and 4.8 g/m2/cycle, respectively, which was lower than reported doses from previous studies7–14. We found that 53% (7/13) of patients needed dose reduction or delay in treatment due to hematologic toxicity. Most cases did not use prophylactic growth factor to prevent febrile neutropenia. This may explain the higher rate of dose reduction in our study compared to the dose reduction of 26% of patients in the ISG study9. All patients in the ISG trial received growth factor for secondary prophylaxis. Non-fatal non-hematologic toxicity was found in 62% (43/69) of patients in this study, which was similar to rates reported in previous studies7–14.\n\nThis study has some notable limitations. As previously described, the potential for selection bias, incomplete medical records, and disproportionate sample size between the chemotherapy and non-chemotherapy groups made it difficult to evaluate contributing factors. The patients enrolled in this study were also from a single center, which is located in a major urban metropolis. Our findings may, therefore, not be generalizable to other geographic settings within Thailand. Moreover, our center is Thailand’s largest tertiary referral hospital, which means that we are often referred patients with complicated and intransigent. Lastly, tumors with heterogeneous histology subtype with a wide range of intrinsic tumor behavior, such as mitotic rate, grade, and undetermined molecular subtype were included. Further study is warranted in a more specific population.\n\n\nConclusion\n\nTumor grade 3 >10 cm was the only baseline clinical characteristic found to be a significant predictor of unfavorable DFS in univariate analysis. No conclusive benefit of adjuvant chemotherapy was observed relative to DFS and OS.\n\n\nData availability\n\nDataset 1. Complete raw demographic and clinical information for each patient included in this study. DOI: https://doi.org/10.5256/f1000research.15793.d21294717.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declare that no grants were involved in supporting this work.\n\n\nReferences\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\nCoindre JM, Terrier P, Bui NB, et al.: Prognostic factors in adult patients with locally controlled soft tissue sarcoma. A study of 546 patients from the French Federation of Cancer Centers Sarcoma Group. J ClinOncol. 1996; 14(3); 869–877. PubMed Abstract | Publisher Full Text\n\nPisters PW, Leung DH, Woodruff J, et al.: Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J ClinOncol. 1996; 14(5): 1679–1689. PubMed Abstract | Publisher Full Text\n\nLevay J, O'Sullivan B, Catton C: Outcome and prognostic factors in soft tissue sarcoma in adults. Ann Surg. 1975; 63(7): 1437–1451.\n\nSinger S, Corson JM, Gonin R, et al.: Prognostic factors predictive of survival and local recurrence for extremity soft tissue sarcoma. Ann Surg. 1994; 219(2): 165–173. PubMed Abstract | Free Full Text\n\nUeda T, Aozasa K, Tsujimoto M: Multivariate analysis for clinical prognostic factors in 163 patients with soft tissue sarcoma. Cancer. 1988; 62(7): 1444–1450. PubMed Abstract | Publisher Full Text\n\nItaliano A, Delva F, Mathoulin-Pelissier S, et al.: Effect of adjuvant chemotherapy on survival in FNCLCC grade 3 soft tissue sarcomas: a multivariate analysis of the French Sarcoma Group Database. Ann Oncol. 2010; 21(12): 2436–2441. PubMed Abstract | Publisher Full Text\n\nBramwell V, Rouesse J, Steward W, et al.: Adjuvant CYVADIC chemotherapy for adult soft tissue sarcoma--reduced local recurrence but no improvement in survival: a study of the European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group. J Clin Oncol. 1994; 12(6): 1137–1149. PubMed Abstract | Publisher Full Text\n\nFrustaci S, Gherlinzoni F, De Paoli A, et al.: Adjuvant chemotherapy for adult soft tissue sarcomas of the extremities and girdles: results of the Italian randomized cooperative trial. J Clin Oncol. 2001; 19(5): 1238–1247. PubMed Abstract | Publisher Full Text\n\nFrustaci S, De Paoli A, Bidoli E, et al.: Ifosfamide in the adjuvant therapy of soft tissue sarcomas. Oncology. 2003; 65 Suppl 2: 80–84. PubMed Abstract | Publisher Full Text\n\nAdjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: meta-analysis of individual data. Sarcoma Meta-analysis Collaboration. Lancet. 1997; 350(9092): 1647–1654. PubMed Abstract | Publisher Full Text\n\nLe Cesne A, Ouali M, Leahy MG, et al.: Doxorubicin-based adjuvant chemotherapy in soft tissue sarcoma: pooled analysis of two STBSG-EORTC phase III clinical trials. Ann Oncol. 2014; 25(12): 2425–32. PubMed Abstract | Publisher Full Text\n\nWoll PJ, Reichardt P, Le Cesne A, et al.: Adjuvant chemotherapy with doxorubicin, ifosfamide, and lenograstim for resected soft-tissue sarcoma (EORTC 62931): a multicentre randomised controlled trial. Lancet Oncol. 2012; 13(10): 1045–1054. PubMed Abstract | Publisher Full Text\n\nGronchi A, Frustaci S, Mercuri M, et al.: Short, full-dose adjuvant chemotherapy in high-risk adult soft tissue sarcomas: a randomized clinical trial from the Italian Sarcoma Group and the Spanish Sarcoma Group. J ClinOncol. 2012; 30(8): 850–856. PubMed Abstract | Publisher Full Text\n\nFletcher CD, Hogendoorn P, Mertens F, et al.: WHO Classification of Tumours of Soft Tissue and Bone. 4th ed. Lyon, France: IARC Press; 2013. Reference Source\n\nOken MM, Creech RH, Tormey DC, et al.: Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982; 5(6): 649–655. PubMed Abstract | Publisher Full Text\n\nChantharasamee J, Tanapathomsinchai N, Niyomnaitham S, et al.: Dataset 1 in: Clinical characteristics and outcomes of extremity soft tissue sarcomas at Thailand’s national tertiary referral center. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15793.d212947\n\nLiu CY, Yen CC, Chen WM, et al.: Soft tissue sarcoma of extremities: the prognostic significance of adequate surgical margins in primary operation and reoperation after recurrence. Ann Surg Oncol. 2010; 17(8): 2102–11. PubMed Abstract | Publisher Full Text\n\nNgan R, Wang E, Porter D, et al.: Soft-tissue sarcomas in the Asia-Pacific region: a systematic review. Asian Pac J Cancer Prev. 2013; 14(11): 6821–32. PubMed Abstract | Publisher Full Text\n\nKim YB, Shin KH, Seong J, et al.: Clinical significance of margin status in postoperative radiotherapy for extremity and truncal soft-tissue sarcoma. Int J Radiat Oncol Biol Phys. 2008; 70(1): 139–44. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "37579",
"date": "10 Sep 2018",
"name": "Robert S. Benjamin",
"expertise": [
"Reviewer Expertise Sarcomas"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript reports the results of treatment of patients with high-risk primary soft-tissue sarcomas of the extremities treated at Thailands’s national tertiary referral center. The report covers 58 patients treated over 10 years, 13 of whom received adjuvant chemotherapy. Univariate and multivariate statistical analyses are presented with a major emphasis on the potential beneficial effects of chemotherapy. Unfortunately, there are too few patients treated with chemotherapy to make any real conclusions; however the authors’ conclusions, based solely on statistical significance, are highly misleading and emphasize the pitfalls of overinterpretation of small data sets. The authors list median DFS and survival figures, but it is not clear how these were derived. Customarily, one uses the median from the Kaplan-Meier curves; but the median DFS and OS in the chemotherapy group from Figures 1 and 2 show that the median DFS and OS for the chemotherapy group has not been reached. One could argue that the improvement in DFS and OS is compelling, regardless of the statistical significance. It may be due to selection bias; but I don’t know how to select which patient with high-risk primary soft-tissue sarcoma is likely to do well, and if I did, that is not the patient I would subject to the toxicities of adjuvant chemotherapy. The authors would do better to present their data without the statistical analysis on the effects of chemotherapy simply stating that the small number of patients treated precludes adequate statistical analysis.\n\nI have some additional specific suggestions for improvement:\n\nIn the abstract, the authors state, “Due to the heterogeneity of soft tissue sarcomas, conflicting data have been reported regarding treatment outcomes. Moreover, factors such as histologic subtype, treatment compliance, and treatment tolerability may influence outcomes of treatment.” I would eliminate the first part of the first sentence. First of all, it is not simply heterogeneity that accounts for conflicting data as the authors note in the second sentence. I would suggest, “Conflicting data have been reported regarding treatment outcomes of patients with primary soft-tissue sarcomas. Factors such as primary site, histologic subtype, treatment regimen, treatment compliance, and treatment tolerability may influence outcomes of treatment.”\n\nFigure 3 requires relabeling and better explanation. What is pattern 1 and 2? What are the authors trying to illustrate? The greater than sign did not show up in the PDF download.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1187
|
https://f1000research.com/articles/7-1186/v1
|
03 Aug 18
|
{
"type": "Case Report",
"title": "Case Report: Root resorption caused after pulp death of adjacent primary molar",
"authors": [
"Maha M. Azab",
"Dalia M. Moheb",
"Osama I. El Shahawy",
"Dalia M. Moheb",
"Osama I. El Shahawy"
],
"abstract": "Necrotic decayed primary molars with necrotic pulp tissues may show periapical involvement and root resorption. In this case report, a pediatric patient with a very common chief complain and clinical picture of necrotic badly decayed molar, introduced a very interesting case when radiographic investigation was performed, which showed that root resorption of the adjacent healthy molar occurred. The current report is, to the best of our knowledge, the first to report such finding in primary dentition.",
"keywords": [
"Root resorption",
"Necrotic tooth",
"Pulpectomy"
],
"content": "Introduction\n\nRoot resorption is the physiologic or pathologic loss of dentin and/or cementum and/or bone1.\n\nPrimary teeth can go through either type, but other than resorption during the shedding process resorption is considered pathologic. Inflammatory root resorption is not a rare finding in the pediatric community, with spread of infection from a carious tooth as a main cause2,3. In the present case, the interesting finding is that root resorption did not only occur in the carious, necrotic tooth but also occurred in the adjacent vital tooth.\n\n\nCase report\n\nA seven and half year-old boy visited the outpatient clinic of Pediatric Dentistry Department, Faculty of Dentistry, Cairo University in June 2015 with a chief complaint of pain on the lower right molar area. The patient’s mother stated that the pain was at times throbbing in nature, and child is not able to chew on this side.\n\nClinical examination showed a badly decayed, lower second primary molar with related localized intraoral abscess, where the lower first primary molar was intact. The patient had poor oral hygiene; he had not received any professional dental care, and was very apprehensive.\n\nRadiographic examination revealed root resorption and bone rarefaction related to lower second primary molar. The interesting finding was a considerable amount of root resorption of the distal root of the adjacent lower first primary molar (Figure 1A).\n\nThe case was managed by performing pulpectomy4 to the lower second primary molar, with root canals filled with calcium hydroxide paste with iodoform (Metapex, Meta Biomed, Republic of Korea). The tooth was then restored with high viscosity glass ionomer (GC Fuji IX GP capsule, GC corporation, Tokyo, Japan) (Figure 1B). The lower first primary molar was not touched and instead monitored. No antibiotics or analgesics was prescribed.\n\nUnfortunately, the patient’s mother did not want follow-up appointments in person, however, she was contacted on the phone, after 2 weeks, 3 months and 6 months, and she said everything was fine and there was no swelling or pain.\n\nAt about 8 months from the treatment appointment, the patient’s mother visited the outpatient clinic with the patient for other reasons, and decided to pass by the Pediatric Dentistry Department for patient follow-up. Clinical examination showed no signs or symptoms, occlusal restoration was intact, and radiographic examination revealed arrested root resorption, on both molars, and an increase in the density of bone although this was not at a normal level yet (Figure 1C).\n\nA) Pre-operative radiograph: Bone rarefaction and root resorption in first and second primary molars; B) Post-operative radiograph: Pulpectomy treatment in lower second primary molar; C) 8 months post-operative: cessation of root resorption.\n\nTable 1 shows the patient’s timeline of symptoms, treatment and follow-up.\n\n\nDiscussion\n\nCaries-related inflammatory root resorption is caused when bacteria from infected pulp stimulate resorptive cells, thus removal of infected pulp is necessary for cessation of the condition5.\n\nThe only previously reported similar case was a periapical lesion adjacent to a tooth with failing root canal therapy, where healing did not occur till extraction of the adjacent tooth6.\n\nIn the current case, the treatment choice for the lower second primary molar was obvious and clear. The problem with the adjacent tooth, which was intact but suffered from root resorption, is that it is not clear by signs, symptoms and investigation whether the root resorption is just caused (due to proximity) by resorptive cells stimulated from bacteria from the necrotic pulp chamber of lower second primary molar, or if bacteria or bacterial toxins have spread to the lower first primary molar, causing retrograde infection, which would have necessitated pulp therapy to the first primary molar as well.\n\nWe have chosen the more conservative treatment plan, which involved the pulpectomy of lower second primary molar and follow-up for the lower first primary molar, which turned out to be appropriate, where mother reported.\n\n\nPatient perspective\n\nThe patient’s mother was pleased with the more conservative treatment performed, as the child was very apprehensive, and she preferred the least clinical procedure possible. She and the child were satisfied with the results as clinical symptoms subsided after treatment.\n\n\nConclusion\n\nAlthough a rare finding, one should consider the possibility of root resorption caused by periapical infection of adjacent tooth, when no other symptoms are present, as the least invasive treatment and follow-up should be tried first.\n\n\nConsent\n\nWritten informed consent for publication of the clinical details and images was obtained from the patient's mother.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nNe RF, Witherspoon DE, Gutmann JL: Tooth resorption. Quintessence Int. 1999; 30(1): 9–25. PubMed Abstract\n\nSantos BZ, Bosco VL, Silva JYB, et al.: Physiological and pathological factors and mechanisms in the process of root resorption of deciduous teeth. RSBO (Online). 2010; 7(3): 332–9. Reference Source\n\nVieira-Andrade RG, Drumond CL, Alves LP, et al.: Inflammatory root resorption in primary molars: prevalence and associated factors. Braz Oral Res. 2012; 26(4): 335–40. PubMed Abstract | Publisher Full Text\n\nBharuka SB, Mandroli PS: Single- versus two-visit pulpectomy treatment in primary teeth with apical periodontitis: A double-blind, parallel group, randomized controlled trial. J Indian Soc Pedod Prev Dent. 2016; 34(4): 383–90. PubMed Abstract | Publisher Full Text\n\nFinucane D, Kinirons MJ: External inflammatory and replacement resorption of luxated, and avulsed replanted permanent incisors: a review and case presentation. Dent Traumatol. 2003; 19(3): 170–4. PubMed Abstract | Publisher Full Text\n\nFrank AL: Inflammatory resorption caused by an adjacent necrotic tooth. J Endod. 1990; 16(7): 339–41. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "36785",
"date": "08 Aug 2018",
"name": "Mariem O. Wassel",
"expertise": [
"Reviewer Expertise pediatric dentistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe title can be changed to \"primary molar root resorption after pulp death of an adjacent primary molar\" to clarify that root resorption occurred to an adjacent primary tooth not a permanent one. Start the abstract with decayed primary molars, delete \"necrotic\" Indicate in abstract the age of the child and that resorption occurred in an adjacent primary molar Replace root resorption by inflammatory root resorption in key words Give hint about child's medical history Write the details of pulpectomy (single or 2 visits), irrigation used Why is SSC not fitted? In ref 6 cited text (page 3, first paragraph), please specify whether this was a primary or permanent tooth In conclusion, please start a new sentence from \"when no other symptoms are present.........\" and delete \"as\" in the same sentence\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
},
{
"id": "37166",
"date": "31 Aug 2018",
"name": "Zafer C Cehreli",
"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 reports a very rare occurrence of the spreading of periapical infection to a healthy primary molar root. The non-vital tooth causing the infection was treated endodontically without any intervention to the vital neighboring primary molar. After 8 months, arrest of the resorption was evident along with improved periapical healing.\n\nMinor comments:\nPlease discuss the effect of coronal restoration on the prognosis of endodontic treatment.\n\nPlease discuss the fate of extruded root canal medicament.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1186
|
https://f1000research.com/articles/6-2172/v1
|
22 Dec 17
|
{
"type": "Research Article",
"title": "Collaboration between a human group and artificial intelligence can improve prediction of multiple sclerosis course: a proof-of-principle study",
"authors": [
"Andrea Tacchella",
"Silvia Romano",
"Michela Ferraldeschi",
"Marco Salvetti",
"Andrea Zaccaria",
"Andrea Crisanti",
"Francesca Grassi",
"Andrea Tacchella",
"Silvia Romano",
"Michela Ferraldeschi",
"Marco Salvetti",
"Andrea Zaccaria"
],
"abstract": "Background: Multiple sclerosis has an extremely variable natural course. In most patients, disease starts with a relapsing-remitting (RR) phase, which proceeds to a secondary progressive (SP) form. The duration of the RR phase is hard to predict, and to date predictions on the rate of disease progression remain suboptimal. This limits the opportunity to tailor therapy on an individual patient's prognosis, in spite of the choice of several therapeutic options. Approaches to improve clinical decisions, such as collective intelligence of human groups and machine learning algorithms are widely investigated. Methods: Medical students and a machine learning algorithm predicted the course of disease on the basis of randomly chosen clinical records of patients that attended at the Multiple Sclerosis service of Sant'Andrea hospital in Rome. Results: A significant improvement of predictive ability was obtained when predictions were combined with a weight that depends on the consistence of human (or algorithm) forecasts on a given clinical record. Conclusions: In this work we present proof-of-principle that human-machine hybrid predictions yield better prognoses than machine learning algorithms or groups of humans alone. To strengthen this preliminary result, we propose a crowdsourcing initiative to collect prognoses by physicians on an expanded set of patients.",
"keywords": [
"Multiple sclerosis",
"Machine learning",
"Random Forest",
"collective intelligence",
"Hybrid predictions",
"Crowdsourcing"
],
"content": "Introduction\n\nThe natural course of multiple sclerosis (MS) is extremely variable, ranging from extremely mild to very aggressive forms. Most patients experience an initial relapsing-remitting (RR) phase, in which symptoms appear and fade. Eventually, remissions fail and the disease proceeds to a secondary progressive (SP) form, leading to incremental disability. The palette of disease-modifying treatments is becoming relatively large, in principle opening the possibility to tailor the therapy to meet the specific needs of each patient. Unfortunately, the accuracy of parameters to predict the rate of disease progression remains suboptimal.\n\nBeing all the above therapies preventive, in the absence of exact prognostic indicators we have to accept that a proportion of patients is either under- or over-treated. This is a serious concern as the disease can be severely disabling, and some of the available therapies can lead to adverse events that can be worse than the disease itself. Thus, the possibility to formulate a prognosis as exact as possible is becoming increasingly appealing.\n\nIn the clinics, as in any other fields of human knowledge, innovative approaches based on machine learning and collective reasoning methods are used in an attempt to succeed where traditional methods of forecasting failed. Machine learning algorithms catch complex relations among existing data to an extent beyond standard regression models. Good performances have been obtained for the diagnosis of Parkinson's disease and the prognosis of disease progression in amyotrophic lateral sclerosis (Dinov et al., 2016; Küffner et al., 2015). For MS, machine learning algorithms can correctly classify disease course in about 70 % of cases of both clinically definite MS and of clinically isolated syndrome (Fiorini et al., 2015; Wottschel et al., 2014; Zhao et al., 2017), a good result that still requires improvement to become of clinical value. Through collective reasoning, or collective intelligence, groups of lay people may perform as well as experts. In principle, the larger the group, the higher the prediction accuracy (see for review Ponsonby & Mattingly, 2015), which led to the development of several crowdsourcing initiatives for diagnostic purposes (for instance, Candido dos Reis et al., 2015; Lau et al., 2016). However, when expert people are involved, even small groups can outperform the best among them, at least when a yes/no answer to well-defined diagnostic questions is requested based on radiographic/ histological images, (Kurvers et al., 2016; Sonabend et al., 2017; Wolf et al., 2015). Studies with medical students show that working in pairs ameliorates diagnostic ability, with further improvements when group size increases (Hautz et al., 2015; Kämmer et al., 2017), in line with the core idea of Collective intelligence.\n\nCombination of human and machine predictions into hybrid forecasts exploits human intuitive reasoning and computer classification capabilities, potentially boosting both. Indeed, at least in the case of predicting the course of actions in American football games within the frame of prediction markets, hybrid groups performed better than either humans or computers. (Nagar & Malone, 2011). In this paper, we report the promising results of a preliminary study on the combination of predictions made by humans with those of a machine learning algorithm on the progression of multiple sclerosis in a set of patients. Machine learning and collective intelligence performed almost equally well, but their combination yielded a small, yet statistically significant, improvement in the reliability of the forecasts on disease evolution over different time periods.\n\nThese results indicate that it is worth deepening the study of human and machine clinical predictions, as well as the potentiality of hybrid predictions, for which we propose a crowdsourcing approach on a platform specifically designed for this analysis (DiagnoShare).\n\n\nMethods and results\n\nOur dataset is composed by clinical records gathered during 527 visits of 84 outpatients followed at the Multiple Sclerosis service of Sant’Andrea hospital in Rome. Parameters evaluated during each visit are listed in Supplementary Table 1. All patients had clinically definite MS in the RR stage at the time of the visit(s) included in the database. Data potentially revealing the identity of the patients was removed from the shared database. For each visit, we noted if the patient was in RR or SP stage after 180, 360 and 720 days, so that predictions could be compared with the true progression of disease in each patient (Supplementary File: TrueOutcomes.xlsx).\n\nUse of database for research purposes was authorized by the Ethical committee of Sapienza University (Authorization n. 4254_2016, dated November 2, 2016).\n\nHaving a correctly labelled dataset, in which each entry is associated to the outcome, we used the Random forest supervised approach to classification (Breiman, 2001; Liaw & Weiner, 2002), using the Scikit-learn toolbox version 0.16.1.\n\nTo benchmark the performance of the trained models, we used a modified k-folding strategy. Since data was limited (a set of 527 records), and not independent, as it had been obtained from 84 patients, with a simple random k-folding the training set would be composed of many correlated same-patient data. Even worse, some of the data from patients present in the training set would be used to validate the model in the benchmarking stage. As a consequence, the model would overfit the training data, misleadingly showing very good performance. Being presented with many data from the same patient, the model optimizes its ability in recognizing patients themselves, through their highly correlated clinical variables.\n\nTo avoid these problems, we developed an alternative approach, training the algorithm with the following rules:\n\n1. We excluded all visits from one patient from the dataset\n\n2. We built 50 training sets, each composed by 83 records, one (randomly chosen) for every remaining patient\n\n3. We trained 50 Random Forest models, one for each training set.\n\n4. We computed the probability of the transition from RR to SP by averaging the predictions of the 50 models on all the visits of the excluded patient. Predictions consisted in scores from 0 (Extremely unlikely) to 1 (Highly probable).\n\nWe repeated the procedure for the 84 patients, obtaining an estimation of the probability of the RR to SP transition for each of the 527 clinical records. Three different prediction delays were considered, namely 180, 360 and 720 days. Results obtained are presented in Supplementary File: RF_Predictions.xlsx. The performance of the model was estimated by the Area Under the \"Receiver Operating Characteristic\" (ROC) Curve (AUC) computed on all the 527 examples. The AUC values obtained are shown in Table 1.\n\nFor each clinical record, the indicated agents evaluated the probability that disease evolved from the RR to the SP phase after 180, 360 or 720 days. Data represent the AUC values obtained for each method. *: P<0.001 when compared to Group or Random Forest values at the same time points.\n\nForty-two medical students in the final two years of their course (Sapienza University, Rome Italy, based within Sant'Andrea hospital), volunteered to participate in the task. All were familiar with clinical records in general, and were instructed on the meaning of each entry present in the medical records of MS patients. This part of the study was approved by the Ethical Committee of the Department of Physiology and Pharmacology, Sapienza University on July 13, 2017.\n\nFor adequate comparison with computer predictions, students evaluated 50 medical records, collected in a questionnaire, randomly extracted from the same dataset used for machine learning and estimated the probability that the patient would progress to the SP phase within 180, 360 and 720 days. Scores were from 0 (Extremely unlikely) to 5 (Highly probable). Predictions (see Supplementary file Student_Predictions.xlsx) were analysed, using the AUC.\n\nOn average, each clinical record was evaluated by 4 of the 42 students.\n\nPredictions were less accurate than those proposed by machine learning (Table 1). Standard deviation was larger for the 180 day time point, indicating that opinions on the long-term evolution of the disease are more widely shared, although they are not more precise. To evaluate the impact of collective intelligence, we measured the performance of Pairs, considering all visits evaluated by at least two individual students, randomly selecting only 2 scores when more were available. The prognoses were averaged before computing the AUC, which showed a marked increase (Table 1). Aggregation of all singles (Group) yielded a further small increase in the performance of the forecasting (Table 1), which almost equalled that of random forest algorithm.\n\nWe next integrated human and computer predictions into a hybrid prediction, which combines human clinical reasoning with the classification approach of machine learning algorithms. These different \"ways of reasoning\" possibly lead to quite divergent predictions on individual cases, a complementarity that should be exploited taking the difference into account when creating hybrid predictions.\n\nTo compare the two sets on equal grounds, predictions on each clinical record were ranked in order of consistence, for the two agents separately, that is agreement between students or decision trees in the random forest. Then, a normalized ranking was assigned, ranging from 1 for the most consistent predictions to 0 for the most scattered. The hybrid prediction score for each clinical record was then obtained by summing the two squared rankings, to emphasize the contribution of the most consistent agent.\n\nNote that a linear combination of rankings would result in a worse performance of hybrid predictions, as the information about the most consistent prediction between the two agents would be lost. A similarly degraded performance is observed when predictions are not ranked.\n\nSince our dataset is relatively small, as is the number of students that evaluated the clinical records, we used a bootstrap procedure to evaluate the statistical significance of the improvement. The bootstrap (Efron & Tibshirani, 1994; Felsenstein, 1985) consists in random sampling of the dataset that allows the estimation of confidence intervals.\n\nAs shown in Table 1 and Figure 1, hybrid predictions yielded a small but statistically significant (P<0.001) improvement in the prediction of disease course in time. Significance was evaluated from confidence limits using standard methods (Altman & Bland, 2011).\n\nThe box plot shows the distribution of the AUC obtained from the bootstrap. In particular, the colored boxes correspond to quartiles, while the lines show the full range of the generated AUCs.\n\n\nDiscussion\n\nA number of studies have investigated the possibility to increase the appropriateness of clinical decisions through collective intelligence of human groups (for instance, Kurvers et al., 2016; Sonabend et al., 2017; Wolf et al., 2015) or machine learning algorithms. The latter approach has been used in a great variety of tasks, and its value in the medical realm is possibly overstated (Chen & Asch, 2017). However, machine learning methods performed well for prognostic predictions (Küffner et al., 2015; Zhao et al., 2017). In particular, the Random forest approach provided good predictions on ALS course (Küffner et al., 2015).\n\nIn this work we present proof-of-principle that human-machine hybrid predictions attain prognostic ability above that of machine learning algorithms and groups of humans alone.\n\nThe duration of the RR phase before its shift into progression has always been difficult to predict, and possibly the random occurrence of relapses (Bordi et al., 2013) contributes to the lack of univocal indicators. No approach, no matter how good, can yield certainty when cause-effect relations are unknown. Thus, our aim has been to obtain predictions on the probability that MS patients in the RR phase will convert to a SP form within a certain time frame. Predictions on the course of real patients were provided by medical students and a random forest algorithm. A significant improvement of predictive ability was obtained when predictions were combined in a non-linear manner, with a weight that depends on the consistence of human (or algorithm) forecasts on a given clinical record.\n\nThis result can be considered in agreement with several studies on different medical issues showing that predictor's confidence correlates very well with the correctness of the prediction (Detsky et al., 2017; Hautz et al., 2015; Kämmer et al., 2017; Kurvers et al., 2016). Indeed, the concordance of different members of a given group (students or runs of the random forest model) can be taken as indicating that the agent is \"sure\" of the forecast.\n\nIn spite of the relatively basic machine learning technique used, the small number of students involved and their limited clinical knowledge, this work suggests that hybrid predictions can be useful to improve the prognosis of MS course. A deeper study is therefore of interest. To recruit more and more skilled humans, we propose a crowdsourcing initiative called DiagnoShare that is being advertised among physicians.\n\nA reliable tool to predict MS progression can be of aid to clinicians to tailor therapy to each patient, but also in clinical trials, to evaluate whether drugs modify the estimated outcome of each enrolled patient, as proposed for ALS (Küffner et al., 2015).\n\nIn the long run, it is possible that further developments in our ability to combine collective reasoning and machine predictions will have a profound impact also on the organization and management of medical care, particularly in hospital settings.\n\n\nData availability\n\nDataset 1: True outcome of patients, indexed as clinical records. More than one clinical record is pertinent to each patient. T_180, T_360, T_720: clinical conditions of the patient 180, 360 and 720 days after the visit in which clinical record was obtained. 0: still in RR phase; 1: transitioned to SP phase. DOI: 10.5256/f1000research.13114.d188355 (Tacchella et al., 2017a)\n\nDataset 2: Predictions on individual clinical records made by medical students. Each student worked on a questionnaire (lines labelled \"questionnaire\", column B.) listing 50 clinical reports (lines labelled \"Clinical report N\", columns B to AY) and made a prediction on the probability of RR –to–SP transition within 180, 360 and 720 days (lines labelled Prediction @ 180, 360, 720, columns B to AY)\n\nThe numbering of Clinical reports is the same used in Dataset 1. DOI: 10.5256/f1000research.13114.d188356 (Tacchella et al., 2017b)\n\nDataset 3: Predictions on individual clinical records made by a Random Forest algorithm. Score_180, Score_360, Score_720: Probability that the patient will transition to SP phase within180, 360 and 720 days after the visit in which clinical record was obtained. The numbering of Clinical reports is the same used in Dataset 1. DOI: 10.5256/f1000research.13114.d188357 (Tacchella et al., 2017c)",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nCENTERS is a special project of, and is supported by, Fondazione Italiana Sclerosi Multipla. AT and AZ acknowledge funding from the “CNR Progetto di Interesse CRISIS LAB”.\n\nAll 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 all the students that participated in the project.\n\n\nSupplementary material\n\nSupplementary Table 1: Parameters evaluated for each patient and included in clinical records.\n\nClick here to access the data.\n\n\nReferences\n\nAltman DG, Bland JM: How to obtain the P value from a confidence interval. BMJ. 2011; 343: d2304. PubMed Abstract | Publisher Full Text\n\nBordi I, Umeton R, Ricigliano VA, et al.: A mechanistic, stochastic model helps understand multiple sclerosis course and pathogenesis. Int J Genomics. 2013; 2013: 910321. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBreiman L: Random Forests. Mach Learn. 2001; 45(1): 5–32. Publisher Full Text\n\nCandido Dos Reis FJ, Lynn S, Ali HR, et al.: Crowdsourcing the General Public for Large Scale Molecular Pathology Studies in Cancer. EBioMedicine. 2015; 2(7): 681–689. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen JH, Asch SM: Machine Learning and Prediction in Medicine - Beyond the Peak of Inflated Expectations. N Engl J Med. 2017; 376(26): 2507–2509. PubMed Abstract | Publisher Full Text\n\nDetsky ME, Harhay MO, Bayard DF, et al.: Discriminative Accuracy of Physician and Nurse Predictions for Survival and Functional Outcomes 6 Months After an ICU Admission. JAMA. 2017; 317(21): 2187–2195. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDinov ID, Heavner B, Tang M, et al.: Predictive Big Data Analytics: A Study of Parkinson’s Disease Using Large, Complex, Heterogeneous, Incongruent, Multi-Source and Incomplete Observations. PLoS One. 2016; 11(8): e0157077. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEfron B, Tibshirani RJ: An introduction to the bootstrap. CRC press, 1994. Reference Source\n\nFelsenstein J: CONFIDENCE LIMITS ON PHYLOGENIES: AN APPROACH USING THE BOOTSTRAP. Evolution. 1985; 39(4): 783–791. PubMed Abstract | Publisher Full Text\n\nFiorini S, Verri A, Tacchino A, et al.: A machine learning pipeline for multiple sclerosis course detection from clinical scales and patient reported outcomes. Conf Proc IEEE Eng Med Biol Soc. 2015; 2015: 4443–6. PubMed Abstract | Publisher Full Text\n\nHautz WE, Kämmer JE, Schauber SK, et al.: Diagnostic performance by medical students working individually or in teams. JAMA. 2015; 313(3): 303–304. PubMed Abstract | Publisher Full Text\n\nKämmer JE, Hautz WE, Herzog SM, et al.: The Potential of Collective Intelligence in Emergency Medicine: Pooling Medical Students' Independent Decisions Improves Diagnostic Performance. Med Decis Making. 2017; 37(6): 715–724. PubMed Abstract | Publisher Full Text\n\nKurvers RH, Herzog SM, Hertwig R, et al.: Boosting medical diagnostics by pooling independent judgments. Proc Natl Acad Sci U S A. 2016; 113(31): 8777–8782. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKüffner R, Zach N, Norel R, et al.: Crowdsourced analysis of clinical trial data to predict amyotrophic lateral sclerosis progression. Nat Biotechnol. 2015; 33(1): 51–57. PubMed Abstract | Publisher Full Text\n\nLau WW, Sparks R, ; OMiCC Jamboree Working Group et al.: Meta-analysis of crowdsourced data compendia suggests pan-disease transcriptional signatures of autoimmunity [version 1; referees: 2 approved]. F1000Res. 2016; 5: 2884. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiaw A, Wiener M: Classification and regression by random Forest. R News. 2002; 2: 18–22. Reference Source\n\nNagar Y, Malone TW: Making Business Predictions by Combining Human and Machine Intelligence in Prediction Markets. Proceedings of the International Conference on Information Systems ICIS 2011. Shanghai, China. 2011. Reference Source\n\nPonsonby AL, Mattingly K: Evaluating New Ways of Working Collectively in Science with a Focus on Crowdsourcing. EBioMedicine. 2015; 2(7): 627–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSonabend AM, Zacharia BE, Cloney MB, et al.: Defining Glioblastoma Resectability Through the Wisdom of the Crowd: A Proof-of-Principle Study. Neurosurgery. 2017; 80(4): 590–601. PubMed Abstract | Publisher Full Text\n\nTacchella A, Romano S, Ferraldeschi M, et al.: Dataset 1 in: Collaboration between a human group and artificial intelligence can improve prediction of multiple sclerosis course: a proof-of-principle study. F1000Research. 2017a. Data Source\n\nTacchella A, Romano S, Ferraldeschi M, et al.: Dataset 2 in: Collaboration between a human group and artificial intelligence can improve prediction of multiple sclerosis course: a proof-of-principle study. F1000Research. 2017b. Data Source\n\nTacchella A, Romano S, Ferraldeschi M, et al.: Dataset 3 in: Collaboration between a human group and artificial intelligence can improve prediction of multiple sclerosis course: a proof-of-principle study. F1000Research. 2017c. Data Source\n\nWolf M, Krause J, Carney PA, et al.: Collective intelligence meets medical decision-making: the collective outperforms the best radiologist. PLoS One. 2015; 10(8): e0134269. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWottschel V, Alexander DC, Kwok PP, et al.: Predicting outcome in clinically isolated syndrome using machine learning. Neuroimage Clin. 2014; 7: 281–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao Y, Healy BC, Rotstein D, et al.: Exploration of machine learning techniques in predicting multiple sclerosis disease course. PLoS One. 2017; 12(4): e0174866. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "30349",
"date": "26 Feb 2018",
"name": "Bruno Bonetti",
"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 is interesting and intriguing, since it opens new possibilities in MS prognosis combining human expertise and \"artificial intelligence\". I do not understand why medical students have been chosen instead of (young) neurologists who may have additional skills in the specific task. Apart from this aspect, the manuscript is well written and easy to follow. Deserves publication.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3475",
"date": "01 Mar 2018",
"name": "Francesca Grassi",
"role": "Author Response",
"response": "Thank you very much for reviewing our paper.In this proof-of-concept study, we chose to work with medical students instead of neurologists because we wanted to test if even a group of relatively uneducated people can enhance the predictive ability of machine learning algorithms, which is now well established.We agree with you that the next step is to obtain predictions by neurologists and other medical doctors, and in fact we set up the platform DiagnoShare (http://www.phys.uniroma1.it/diagnoshare) to extend the study.Hopefully, we can soon extend this work with a final study"
}
]
},
{
"id": "31369",
"date": "21 Mar 2018",
"name": "Roger 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 is an interesting and clearly written article on using a machine learning method (random forests) and medical students to form \"hybrid\" predictions of disease progression in MS, specifically the conversion from RRMS to SPMS. The article claims that the results are a proof-of-principle that combining machine learning and human predictions is better than either approach alone.\n\nThe main strengths of the article are its clear writing, the reproducbility of the experiments, the clinical importance of the application, and topical nature of the subject, as machine learning for clinical prediction is such a hot topic that integration with the clinical workflow is a critical area of study.\nThe main limitations of the article are that only clinical parameters were used to perform the predictions, and the longitudinal nature of the data was not used to its full benefit. To realize the potential of machine learning for MS prediction, imaging parameters should be included (there is good literature on MS prediction using imaging), and examining changes over time is important for both machine (eg, using recurrent networks) and human raters (examining clinical changes over multiple time points). The article places some importance on having the computer and humans using the same set of input parameters, but I do not feel that this is warranted; the data should be selected to be most appropriate for each approach.\nGiven the above limitations, it is difficult to generalize the findings to say that hybrid predictions are better than either machine learning or humans. This could be true, and the article provides some support for that, but more work needs to be done to provide strong evidence.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3859",
"date": "01 Aug 2018",
"name": "Francesca Grassi",
"role": "Author Response",
"response": "First of all, thank you for taking the time to read our work and to give useful comments. We hope that our responses will clear your doubts. We list below the changes introduced in the new version, prompted by your observations. We hope that you agree with us that it is improved. The main limitations of the article are that only clinical parameters were used to perform the predictions, and the longitudinal nature of the data was not used to its full benefit. To realize the potential of machine learning for MS prediction, imaging parameters should be included (there is good literature on MS prediction using imaging), and examining changes over time is important for both machine (eg, using recurrent networks) and human raters (examining clinical changes over multiple time points). The article places some importance on having the computer and humans using the same set of input parameters, but I do not feel that this is warranted; the data should be selected to be most appropriate for each approach. ANSWER We agree with you that many other approaches could be used. As now stated in the text, we chose to explore predictions done using only clinical data, available to all neurologists, which have recently been independently demonstrated to have good predictive value (Goodin et al., 2018; reference added to the paper). Imaging data performed in real-world clinical settings do not have the standardization required for predictions either by experts or algorithms. However, future studies aimed at confirming this proof-of-principle, initial work will surely consider different options. Given the above limitations, it is difficult to generalize the findings to say that hybrid predictions are better than either machine learning or humans. This could be true, and the article provides some support for that, but more work needs to be done to provide strong evidence. ANSWER We completely agree with you that this is a preliminary, proof-of-concept work. We state it more clearly in the Discussion"
}
]
},
{
"id": "31371",
"date": "06 Apr 2018",
"name": "Bjoern Menze",
"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:\nI think exploring how to fuse multiple expert opinions is a very interesting line of research in computer aided diagnostics. Here, the authors test how to make use of lay persons, and I would agree that there are many tasks when a (briefly trained) lay person or non-expert can contribute significantly to an analytical task.\nIn the application here, I would be rather critical about this approach, though. For example, the authors write \"through collective reasoning, or collective intelligence, groups of lay people may perform as well as experts.\" I would not agree, by any means. How would a lay person without training be able to distinguish, for example, a stroke related white matter hyper-intensity from an MS lesion? Or even a large MR artifact? Averaging will reduce variance in prediction, but the individual prediction itself has to be unbiased. In other words: the layman predictor has to be correct on average. But how would they possibly be in case they have no idea about how to read these data? Moreover, the authors point out that \"studies with medical students show that working in pairs ameliorates diagnostic ability\". Is this because of a better discussion of the decision? With two subjects it cannot be the power of large numbers that this study relies on.\nInstead of exploring how to fuse layman's decisions, I would recommend the authors to explore how to fuse decisions of different algorithms, or from neurologists of different training/seniority level, or decisions based on different sources.\n\nTechnical:\nExperimental setup and evaluation: The authors describe a \"leave-one patient-out\" cross-validation as an innovation of their study. While this is a good approach, it is not new.\nAlgorithm and training: There are different classes - what is the distribution of those classes for the 84 patients? What is in the reports? Numbers? Free text? What features are input to the random forest algorithm? How many features at all? How did you train the algorithm (parameters \"mtry\", why 50 trees?) Without this information it is difficult to assess whether the performance of your random forest is bad (i.e., close to layman's predictions) because of an suboptimal training procedure, or because this is a hard problem indeed.\nFusion rule: (Described in the section \"To compare the two sets... of the most consistent agent.\") I don't understand what you do. How does summing a squared ranking lead to a prediction score? I assume you are using the normalized (and squared) ranking as a sort of weight? Why do you square the rankings? What happens when you use other non-linear transformations? Is there any way you illustrate the distributions so that we can follow your reasoning? How about presenting simple rules like averaging, or majority voting at least as a baseline we can compare against?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3858",
"date": "01 Aug 2018",
"name": "Francesca Grassi",
"role": "Author Response",
"response": "First of all, thank you for taking the time to read our work and to give useful comments. We hope that our responses will clear your doubts. We list below the changes introduced in the new version, prompted by your observations. We hope that you agree with us that it is improved In the application here, I would be rather critical about this approach, though. For example, the authors write \"through collective reasoning, or collective intelligence, groups of lay people may perform as well as experts.\" I would not agree, by any means. How would a lay person without training be able to distinguish, for example, a stroke related white matter hyper-intensity from an MS lesion? Or even a large MR artifact? Averaging will reduce variance in prediction, but the individual prediction itself has to be unbiased. In other words: the layman predictor has to be correct on average. But how would they possibly be in case they have no idea about how to read these data? ANSWER Although your point of view is quite understandable, there is a large body of literature on the topic of collective intelligence. In the hope to overcome your skepticism on this point, we added some more references to published work on diagnostic crowdsourcing initiatives. Moreover, the authors point out that \"studies with medical students show that working in pairs ameliorates diagnostic ability\". Is this because of a better discussion of the decision? With two subjects it cannot be the power of large numbers that this study relies on. ANSWER It is now better explained that the two quoted studies use different methods: real pairs in one, aggregated opinions in the other, yet both obtain better performances. Authors do not discuss the underlying processes, so we cannot indicate the real reason of a better performance. Instead of exploring how to fuse layman's decisions, I would recommend the authors to explore how to fuse decisions of different algorithms, or from neurologists of different training/seniority level, or decisions based on different sources. ANSWER Thank you for your suggestion. Understanding that we have to deepen out study (now repeatedly stated throughout the paper) we have developed DiagnoShare to obtain the predictions of clinicians of different expertise and we are investigating the performance of algorithm combinations. Experimental setup and evaluation: The authors describe a \"leave-one patient-out\" cross-validation as an innovation of their study. While this is a good approach, it is not new. ANSWER Thank you for your observation. Indeed, it is better to define our approach as a modified leave-one-out method. It is modified, because we not only left one patient out, we also included only one record for each of the remaining patients. Algorithm and training: There are different classes - what is the distribution of those classes for the 84 patients? What is in the reports? Numbers? Free text? What features are input to the random forest algorithm? How many features at all? How did you train the algorithm (parameters \"mtry\", why 50 trees?) Without this information it is difficult to assess whether the performance of your random forest is bad (i.e., close to layman's predictions) because of an suboptimal training procedure, or because this is a hard problem indeed ANSWER Thank you for pointing out that this part of the paper required clarification. As now better emphasized in the text, in this proof-of-concept work we considered only patients that actually transitioned to the SP phase, so there is a unique class of patients. Features input to the RF algorithm are listed in Supplementary Table 1. We added a statement to declare what types of numerical values we used in the work. The results presented show that the RF algorithm performs better than layman, as its performance is however better than that of individual medical students, that are not quite laymen, although not experts as well. In any case, the focus of the paper is not on the goodness of the algorithm, but on the value of combining different approaches to the prediction problem, which indeed has been resisting solution for many years of medical analysis. Fusion rule: (Described in the section \"To compare the two sets... of the most consistent agent.\") I don't understand what you do. How does summing a squared ranking lead to a prediction score? I assume you are using the normalized (and squared) ranking as a sort of weight? Why do you square the rankings? What happens when you use other non-linear transformations? Is there any way you illustrate the distributions so that we can follow your reasoning? How about presenting simple rules like averaging, or majority voting at least as a baseline we can compare against? ANSWER We agree with you that, indeed, this point is complex and we try a different explanation, hoping that it is clearer. First of all, ranking is inherent to building a ROC curve. Since we have only two agents (humans and RF algorithm), we cannot use a majority rule, we can only perform an average (linear or weighted) of the scores. For any clinical record, the final forecast is the average of \"unitary predictions\" by multiple individuals or decision trees. If \"unitary predictions\" of one agent are highly concordant, it means that the prediction is quite obvious for the agent, suggesting that is more probably correct than others. We ranked forecasts on clinical records in order of concordance of \"unitary predictions\" and emphasized the value of agreement by squaring the ranks. In line with other pieces of research, this weighted average performed better than linear averaging, as stated in the paper"
}
]
}
] | 1
|
https://f1000research.com/articles/6-2172
|
https://f1000research.com/articles/7-1167/v1
|
01 Aug 18
|
{
"type": "Research Article",
"title": "Utilization of full postnatal care services among rural Myanmar women and its determinants: a cross-sectional study",
"authors": [
"Aye Sandar Mon",
"Myo Kyi Phyu",
"Wilaiphorn Thinkhamrop",
"Bandit Thinkhamrop",
"Aye Sandar Mon",
"Myo Kyi Phyu"
],
"abstract": "Background: Mothers and their newborns are vulnerable to threats to their health and survival during the postnatal period. Full postnatal care (PNC) uptake decreases maternal deaths and is also essential for first 1,000 days of newborn’s life, but PNC usage is usually inadequate in rural areas. Little is known about the full PNC utilization among rural Myanmar women. This study, therefore, aimed to study the situation of the utilization of full PNC and examine its determinants. Methods: This community-based cross-sectional study was conducted in selected villages of the Magway Region, Myanmar. A total of 500 married women who had children aged under 2 years were selected using multistage cluster sampling and interviewed with semi-structured questionnaires. The determinants of full PNC usage were identified by generalized estimating equation (GEE) under a logistic regression framework. Results: Among 500 rural women, around a quarter (25.20%; 95% confidence interval (CI), 21.58-29.21%) utilized full PNC. Multivariable analysis revealed that factors associated with full PNC usage included mothers attaining educational level of secondary or higher (adjusted odds ratio (AOR), 2.16; 95% CI, 1.18-3.94), belonging to higher income level (AOR, 2.02; 95% CI, 1.11-3.68), having male involvement (AOR, 2.19; 95% CI, 1.02-4.69), being of low birth order (i.e. the first birth) (AOR, 3.26; 95% CI, 1.80-5.91), and having awareness of postnatal danger signs (AOR, 2.10; 95% CI, 1.15-3.83). Moreover, the presence of misconceptions on postnatal practice was identified as a strong barrier to adequate PNC usage (AOR, 0.12; 95% CI, 0.04-0.36). Conclusion: Most of the rural women practiced inadequate PNC in Myanmar. Maternal healthcare services at rural areas should be intensively promoted, particularly among women who had high birth order (greater number of births). Health education regarding perinatal misconceptions and danger signs, and benefits of full PNC services usage should be emphasized and urgently extended.",
"keywords": [
"postnatal care",
"full PNC utilization",
"rural women",
"Myanmar"
],
"content": "Introduction\n\nIn Myanmar, eliminating preventable maternal mortality remains one of the critical challenges to the health system, despite the fact that maternal and child health care has been prioritized. The maternal mortality ratio (MMR) was estimated as 282 per 100,000 live births (LBs) in 2014 Myanmar census report1. In South-East Asian region, Myanmar has a higher MMR than the regional average, which is 140 per 100,000 LBs1. The leading cause of maternal death was post-partum haemorrhage (PPH), and the second and third-leading causes were pregnancy-induced hypertension and abortion, respectively. Over three-quarters (77.4%) of maternal deaths in Myanmar occurred in women who resided in rural areas2. Even though rural women are likely to have higher birth rates, most of them have greater reluctance in seeking, reaching and receiving care from skilled providers3.\n\nIncreasing the quality and skilled postnatal care has recently been highlighted as a method of reducing preventable maternal mortality4. Moreover, effective and adequate PNC is also essential for the first 1,000 days of a child’s life. Improving the health of pregnant women and new mothers will not only reduce maternal morbidity and mortality, but also further reduce child mortality. It has been shown that motherless children have a higher chance of dying before their second birthday than those who have mothers alive5. The highest risk of maternal mortality is during delivery and in the immediate postnatal period, especially the first 24 hours6. Therefore, the World Health Organization (WHO) recommended the optimal timing of PNC should start as early as possible within 24 hours after birth, even if birth occurs at home. The recommended numbers of postnatal visits are at least three additional post natal contacts, in addition to the first contact within 24 hours of birth: on day 3, between days 7 and 14, and 6 weeks postpartum7.\n\nThe postpartum period is defined as the first six weeks (42 days) after birth. This period poses substantial risks and hazards for maternal and neonatal health, and a lack of quality health care during this time may result in mortality or disability, in addition to missed opportunities to promote healthy behaviors. The first hours and days after birth are the most crucial for both mother and neonate, despite the fact that those in the postnatal period are paid less attention by skilled care providers compared to those in the antenatal and intranatal periods7,8.\n\nA number of international studies have been conducted to determine the factors associated with postnatal care utilization in developing countries. Some have emphasized the timing of postnatal care visits9,10, but others have considered whether women received PNC at least once, regardless of the timing of the first visit or the number of visits11–15. In Myanmar, literature concerning postnatal care remains limited in spite of having many studies about antenatal and intranatal care. This study aims to explore the magnitude of rural women who received full PNC in addition to push and pull factors for full PNC utilization in Myanmar.\n\n\nMethods\n\nThis community-based cross-sectional survey was conducted at selected villages (anonymized for ethical reasons) in the Magway Region, which was chosen because its MMR, 343.6 per 100,000 LBs1, is higher than union average (282 per 100,000 LBs) and then 85% of residents in this region are from rural populations16. Data were collected between November 2016 and January 2017.\n\nThe required sample size of 500 participants was estimated based on the multiple logistic regression analysis, as described previously17. Married women aged 15-49 years who had children aged under 2 years of age and provided informed consent were eligible for this study. Woman who could not communicate properly due to physical or mental ill health were excluded from this study. The eligible samples were obtained by applying multistage cluster sampling method. Firstly, out of 26 townships, 4 were selected by simple random sampling using a lottery. From each township, a random selection of 5 villages (having not less than 18 women who had delivered 2 years prior to the survey) was done. As a result, 500 women fitting the eligibility criteria were recruited in person, with the assistance of local health authorities, from 21 villages by cluster sampling. The 2-year recall period was used to minimize recall bias.\n\nData were collected by face to face interview using semi-structured questionnaires (Supplementary File 1). Reliability of 0.86 was estimated by using Cronbach’s alpha. Validity was arranged by the three experts to obtain the finalized version of questionnaire. Preceding the interview, the researcher trained 10 enumerators for data collection and also explained the objectives and facts to follow while asking the questionnaires. Data were collected after participants had been informed about the purpose of the study, ensuring confidentiality to those taking part in the study.\n\nIn this study, the outcome variable was utilization of full PNC which was defined as the participants receiving at least four postnatal visits and the first visit within 24 hours of delivery. For analysis, the outcome responses were dichotomized into the women who reported less than four postnatal visits or postnatal care after 24 hours =0 and those who received four or more postnatal visits and the first visit within 24 hours =1. The independent variables measured were as followings: socio-demographic variables such as age of respondents, education level, average monthly per-capita income, male involvement, accessibility to PNC services. Moreover, knowledge of postnatal danger signs and perception on traditional birth attendants (TBAs) were also defined as explanatory variables. Finally, birth order (i.e. the order that the child was born to his/her family), number of AN visits and misconception regarding postnatal practices were considered as important independent variables in this study.\n\nSome independent variables are explained in detail as follows. Male involvement was considered if the woman was provided with transportation assistance for perinatal visits by her husband and the couple had mutual discussion for maternal healthcare usage. Accessibility to maternal care was defined as a combination of the time spent for travelling to the nearest health center and whether the mother could visit there during any season; that is, if the nearest health center was situated within less than 2 hours travelling distance and could be visited during any season, especially rainy season, this was counted as easy accessibility to nearest health center, otherwise, as difficulty in access. Regarding misconceptions, if a woman avoided certain foods, had behavioral restrictions or customs/practices that might threaten the health and survival of mothers and their babies within postnatal period, she belonged to the category of women having misconception on postnatal practices. The outcome variable and most of the independent variables were measured as categorical ones, except age, family income, birth order, numbers of antenatal visits and postnatal visits. However, for more simple analysis and better interpretation purposes, all numerical independent variables were categorized.\n\nThe statistical analysis was conducted using the STATA version 13.1. The socio-demographic and background characteristics of respondents were presented as frequencies and percentages for categorical variables and as summary statistics, such as mean ± standard deviation for continuous variables. The full PNC utilization rate with 95% confidence interval (CI) was also described. To explore the determinants on full PNC utilization, odds ratio with 95% CI was estimated using a generalized estimating equation (GEE) under multiple logistic regression framework. To take into account the correlation of an event occurring within the same village (i.e. those in the same village having similar access to a health clinic), for estimation of standard error, the GEE was applied18. The factors which were significant at p-value less than 0.25 in bivariate analysis were included in the GEE method. All statistical tests were two-sided and p-values less than 0.05 were considered as statistically significant.\n\nThe Khon Kaen University Ethics Committee for human research with reference number [HE592256] and the Ethical Committee of University of Public Health, Yangon, Myanmar [Ethical (6/2016)] approved this study. Permission to conduct this study was obtained from local responsible persons and health authorities (i.e. village administrative authorities and health authorities from Magway Regional Public Health Department, respectively). Participation in this study was entirely voluntary and informed consent was taken from all participants prior to interview. For participants younger than 18 years, consent was obtained from the individual’s guardian.\n\n\nResults\n\nOut of 500 respondents, nearly half of them (48.2%) were in the young adult age group of 25 to 35 years. The participants were aged between 17 and 47 years, with a mean age of 29.72±6.6 years (Table 1). Majority of the respondents and their spouses were in primary or below level of education, accounting 72.2% and 63.8% respectively. About 64% of the interviewee had no more than five family members. More than half of the respondents (60.8%) had low incomes (less than 50,000 Myanmar kyats (MMK)). Regarding accessibility, about half of respondents (44.8%) encountered difficulty in accessing their nearest health center (that is, they experienced more than 2 hours travel there or it was not easily accessed in the rainy season). In connection with male involvement, 46.8% of the participants were provided with assistance from their husband regarding maternal care usage, such as transportation assistance, and mutual discussion for seeking and receiving maternal healthcare services.\n\nThe average number of children that the respondents had during the study period was 2 (SD=1.4) and 34 respondents (6.8%) had 5 children and more. For just under half of the mothers (47.6%), the last child recently delivered was their first born (Table 2). Most of the mothers (76.2%) had low awareness of postnatal danger signs, including neonatal health risks. On the other hand, around a quarter (23.8%), classified as having a high level of awareness of postnatal danger signs, could name at least 3 out of 8 postpartum danger signs and 1 out of 6 neonatal danger signs. Nearly 50% perceived TBAs as skilled care givers. Only half of mothers received maternal healthcare (antenatal, intranatal and postnatal) from skilled healthcare providers, who included doctors, nurses, lady health visitors (skilled maternal care providers in rural areas) and midwives. Slightly under a quarter (23.6%) did not take antenatal care at all. Nearly two-thirds of the women in the study (64.4%) selected their home as their place of delivery. About one-third of mothers (32.6%) did not take postnatal care and just over a quarter (27.2%) received at least 4 visits (the WHO-recommended number of visits). Among the 337 respondents who took postnatal care, 83.68% received their first postnatal contact with skilled provider within 24 hours of delivery (the WHO-recommended timing of the first visit). The majority of these individuals (about 90%) received health services for both mother (84.8%) and newborns (97.3%). Regarding receipt of health education on breastfeeding and postnatal danger signs, around half of the mothers were provided with this information (breastfeeding, 48.1%; postnatal danger signs, 51.6%). Moreover, just under half of mothers could get knowledge about contraception methods (49.3%) although over three-quarters of them (75.4%) were provided with contraceptives. Almost all of them (98.8%) were given postnatal supplements, such as vitamin B1 and iron. Out of the 500 women, almost half of them (49.6%) had misconception regarding postnatal practices; these included food taboos such as avoiding the consumption of meat and some vegetables or behavioral restrictions such as avoiding going outside the delivery room within 7 days of the birth and massaging lower abdomen for the removal of impure blood.\n\n*Those that received more than one PNC service. TBA, traditional birth attendant.\n\nOf the 500 women in this study with children under 2 years age, 126 utilized full PNC, i.e. they received at least four postnatal visits and their first visit within 24 hours after childbirth (25.20% (95%CI, 21.58-29.21)) (Table 3). The results from bivariate analysis presented as the crude odds ratio (OR) along with its 95% CI, and P-value of each variable revealed that all of the factors in the Table 3 were statistically significant associated with full PNC: these were composed of age, education attainment of respondents and their husbands, income, accessing to health center, male involvement, birth order, awareness of postnatal danger signs, acceptance of TBA, types of health care provider, number of AN visits, place of delivery and misconception regarding postnatal practices.\n\nTBA, traditional birth attendant; ANC, antenatal care.\n\nAfter adjusting for covariates using multivariable analysis with multivariable logistic regression implemented with GEE, it was found out that the higher the degree of school education of the mother, the larger the odds of utilizing full PNC (adjusted odds ratio (AOR), 2.16; 95% CI, 1.18-3.94) (Table 4). The rural women earning higher incomes (≥50,000 MMK) were twice as likely to receive full PNC as their counterparts earning <50,000 MMK (AOR, 2.02; 95% CI, 1.11-3.68). The participants who received support from their spouses to receive PNC were 2.19 times more likely to utilize full PNC than those who did not receive male involvement (AOR, 2.19; 95% CI, 1.02-4.69). The respondents who were knowledgeable about postnatal danger signs were two times more likely to receive full PNC than those with low awareness (AOR, 2.10; 95% CI, 1.15-3.83). Delivery of the first child (AOR, 3.26; 95% CI, 1.8-5.91) was identified as a conclusive determinant of full PNC usage. The presence of misconceptions regarding postnatal practice had a strong negative impact on the utilization of full PNC, with an AOR of 0.12 (95% CI, 0.04-0.36).\n\n\nDiscussion\n\nThis community-based study was conducted to assess the extent of and determinants on full postnatal care utilization of rural Myanmar women. The present study highlighted the inadequate receipt of postnatal care among mothers in rural Myanmar. The prevalence of full PNC utilization was only 25.2%. A national survey focusing on the timing of postnatal visit revealed that the overall prevalence was 68%20. The variation in presenting this utilization rate might be due to different operational definitions for outcome variable in different studies. Moreover, comparing the proportions of complete ANC attendance and health facility delivery, that of full PNC usage is markedly lower among the participants of this study. The attainment of a higher level of education was significantly associated with the receipt of full PNC in the current study, which was consistent with other studies conducted in Bangladesh and Nepal9,21 and, in addition, also homogeneous with the findings of a national survey20. This might be due to the fact that mothers with higher education attainment are more likely to seek health information about safe motherhood, including newborn care, availability and accessibility to health care services from reliable sources of information. Studies undertaken in Indonesia, India and China indicated that the wealth of the mother was associated with the receipt of PNC11,13,22. Similarly, in our study, rural mothers with low per-capita income (less than 50,000 MMK; the amount below the international poverty line as determined by the World Bank), were less likely to use full PNC. The possible explanation might that low income resulted in financial hardship, leading to barriers for taking full PNC. This explanation was strongly supported by the notion that more than two-third of non-users in this study reported they didn’t receive PNC because of unaffordability in terms of time and money.\n\nIn the present study, male involvement in spousal discussion on receipt of maternal care services and accompanying the partner to health facility was observed to have a positive influence on full PNC utilization, fitting with data from a study from India in which male involvement and their knowledge about maternal health significantly related to the maternal healthcare utilization23. Regarding obstetric determinants, prior studies mentioned that factors such as birth order, knowledge about perinatal danger signs, antenatal attendance and place of delivery had association with PNC uptake9,13,14,21,22,24. This study also revealed that first birth order and high awareness of postnatal danger signs were very strong pull factors on full PNC utilization. However, unexpectedly, the frequency of antenatal attendance and place of delivery did not guarantee full PNC usage. The potential reason behind this might be the participants were not likely to be informed about the importance of PNC, its availability, recommended timing and targeted frequency of postnatal visits during antenatal visits and before discharge from health facility after delivery, leading to ignorance of PNC until mothers encountered any postnatal complication or abnormality. Moreover, in the current study, a significant proportion of rural women did not receive education and counseling relating to breastfeeding, postnatal danger signs and contraception. This indicated that there might be a weakness in delivering health messages from health care providers to rural mothers.\n\nConsistent with prior studies on postpartum belief and practice, misconceptions regarding postnatal practice were proved as barrier to PNC uptake by the evidence from the current research25,26. Rural women who had such misconceptions exhibited 88% lower usage of full PNC than those who did not. Based on Myanmar customs and traditional beliefs, food prohibition, behavioral restriction or both within the postpartum period were observed among nearly half of the participants (49.5%). Breastfeeding mothers who had postpartum food taboos perceived that meat consumption could make the newborn ill and that some vegetables, such as roselles, cause abdominal pain and flatulence for both mother and baby. Some mothers reported that they ate only fried fish, dried fish, dried prawns and soup during their postpartum period. This food avoidance practice might result in nutritional deficiency for both mothers and babies. Another common misconception perpetuated among rural women was that strict home confinement within 7 days after delivery; this behavioral restriction might bar to timely and adequate attendance of PNC.\n\nThis study has a number of strengths. This is the first study to reveal the prevalence and determinants of utilization of full PNC, based on the recommended timing and frequency of postnatal visits as per updated WHO postnatal guideline, among rural Myanmar women. Our data analysis, developed using our aforementioned sampling technique, is thereby more likely to provide valid estimates. In addition, the evidence obtained from the current research provides updated knowledge and assistance for the policy makers and healthcare providers to extend quality maternal healthcare package nationwide. Nonetheless, the present study has some limitations. The cross-sectional nature restricts the ability to draw cause-effect relationships between the potential predictors and full PNC utilization. Since the participants were reporting past experience and practice, there may have elicited recall bias. Nevertheless, a 2-year recall period was selected to minimize this bias.\n\n\nConclusion\n\nThe current study reported on the underutilization of postnatal care among rural Myanmar women. The key determinants on full PNC were education attainment, having higher income, male involvement, the first birth order, awareness of postnatal danger signs, and presence of postnatal misconception. On the basis of the evidence generated in this study, coverage of maternal healthcare emphasizing PNC should be intensified to reach out to less-educated mothers, those from low-income families and high-birth-order mothers. An awareness-raising program highlighting the importance and availability of postnatal care is essential to improve full PNC utilization; it is urgently needed to facilitate the health care providers for provision of essential and updated health information concerning safe motherhood and newborn care, in order to correct harmful misconceptions and upgrade knowledge regarding perinatal danger signs among rural women. Further study focusing on quality of PNC services and satisfaction on services the rural women received should be recommended.\n\n\nData availability\n\nDataset 1. Complete de-identified demographic information for each women taking part in the study, in addition to the answer provided to each question of the questionnaire. A dictionary of terms used in the dataset is also included. DOI: https://doi.org/10.5256/f1000research.15561.d21175019.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declare that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe health authorities and staff from Magway Regional Health Division are acknowledged for their kind support in field data collection.\n\n\nSupplementary material\n\nSupplementary File 1. The questionnaire given to each woman in this study, in the original Burmese and English.\n\nClick here to access the data.\n\n\nReferences\n\nDepartment of Population, Ministry of Labour, Immigration, and, Population: The 2014 Myanmar population and Housing Census: Thematic report on maternal mortality. 2016. Reference Source\n\nMOH: Maternal death review. Nay Pyi Taw, Myanmar: Ministry of Health; 2013.\n\nOo K, Win LL, Saw S, et al.: Challenges faced by skilled birth attendants in providing antenatal and intrapartum care in selected rural areas of Myanmar. WHO South East Asia J Public Health. 2012; 1(4): 467–76. PubMed Abstract | Publisher Full Text\n\nVictora CG, Requejo JH, Barros AJ, et al.: Countdown to 2015: a decade of tracking progress for maternal, newborn, and child survival. Lancet. 2016; 387(10032): 2049–59. PubMed Abstract | Publisher Full Text\n\nOwino B: The use of Maternal Health Care Services Socio-economic and demographic factors—Nyanza, Kenya. African French Research Institute. 2001; 21. Reference Source\n\nWorld Health Organization, Regional Office for South-East Asia: Compendium of maternal and neonatal health strategies in SEA Region. New Delhi: WHO Regional Office for South-East Asia; 2010. Reference Source\n\nWHO: WHO Recommendations on Postnatal care of mother and newborn. Geneva: WHO; 2013; 63. Reference Source\n\nWHO DoMPS: WHO Techanical Consultation of Postpartum and Postnatal Care. Geneva: World Health Organization; 2010; 65. Reference Source\n\nKhanal V, Adhikari M, Karkee R, et al.: Factors associated with the utilisation of postnatal care services among the mothers of Nepal: analysis of Nepal demographic and health survey 2011. BMC Womens Health. 2014; 14(1): 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRutaremwa G, Wandera SO, Jhamba T, et al.: Determinants of maternal health services utilization in Uganda. BMC Health Serv Res. 2015; 15: 271. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu X, Zhou X, Yan H, et al.: Use of maternal healthcare services in 10 provinces of rural western China. Int J Gynaecol Obstet. 2011; 114(3): 260–4. PubMed Abstract | Publisher Full Text\n\nDhaher E, Mikolajczyk RT, Maxwell AE, et al.: Factors associated with lack of postnatal care among Palestinian women: a cross-sectional study of three clinics in the West Bank. BMC Pregnancy Childbirth. 2008; 8(1): 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSingh PK, Rai RK, Alagarajan M, et al.: Determinants of maternity care services utilization among married adolescents in rural India. PLoS One. 2012; 7(2): e31666. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorku AG, Yalew AW, Afework MF: Factors affecting utilization of skilled maternal care in Northwest Ethiopia: a multilevel analysis. BMC Int Health Hum Rights. 2013; 13(1): 20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDarega B, Dida N, Tafese F, et al.: Institutional delivery and postnatal care services utilizations in Abuna Gindeberet District, West Shewa, Oromiya Region, Central Ethiopia: A Community-based cross sectional study. BMC Pregnancy Childbirth. 2016; 16: 149. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinistry of Immigration and Population DoP. The 2014 Myanmar Population and Housing Census The Union Report: Census Report Volume 2. Nay Pyi Taw, Myanmar: Department of Population, Ministry of Immigration and Population, 2015. Reference Source\n\nHsieh FY, Bloch DA, Larsen MD: A simple method of sample size calculation for linear and logistic regression. Stat Med. 1998; 17(14): 1623–34. PubMed Abstract | Publisher Full Text\n\nKleinbaum DG, Klein M: Logistic Regression for Correlated Data: GEE. Logistic Regression: A Self-Learning Text. Statistics for Biology and Health. Third ed. New York: Springer; 2010; 489–538. Reference Source\n\nMon AS, Phyu MK, Thinkhamrop W, et al.: Dataset 1 in: Utilization of full postnatal care services among rural Myanmar women and its determinants: a cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15561.d211750\n\nInternational MaI: Myanmar Demographic and Health Survey 2015-16: Key Indicators Report. 2016. Reference Source\n\nRahman MM, Haque SE, Zahan MS: Factors affecting the utilisation of postpartum care among young mothers in Bangladesh. Health Soc Care Community. 2011; 19(2): 138–47. PubMed Abstract | Publisher Full Text\n\nTitaley CR, Dibley MJ, Roberts CL: Factors associated with non-utilisation of postnatal care services in Indonesia. J Epidemiol Community Health. 2009; 63(10): 827–31. PubMed Abstract | Publisher Full Text\n\nSinha KC: Male involvement and utilization of maternal health services in India. Int J Sci Res Publ. 2016; 4(11): 1–13. Reference Source\n\nTesfahun F, Worku W, Mazengiya F, et al.: Knowledge, perception and utilization of postnatal care of mothers in Gondar Zuria District, Ethiopia: a cross-sectional study. Matern Child Health J. 2014; 18(10): 2341–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMrisho M, Obrist B, Schellenberg JA, et al.: The use of antenatal and postnatal care: perspectives and experiences of women and health care providers in rural southern Tanzania. BMC Pregnancy Childbirth. 2009; 9: 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiamond-Smith N, Thet MM, Khaing EE, et al.: Delivery and postpartum practices among new mothers in Laputta, Myanmar: intersecting traditional and modern practices and beliefs. Cult Health Sex. 2016; 18(9): 1054–66. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "36685",
"date": "06 Aug 2018",
"name": "Myo Myo Mon",
"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 should be accepted.\nIt’s a good article highlighting the determinant of utilization of PNC services among rural women in Myanmar. The work is clearly and accurately presented. The finding clearly fulfilled the objectives of the study. And it also includes the available current literature.\nThe study design is appropriate and the work is technically sound. Investigators used the sound study design, measured all necessary outcome and independent variables, and defining all the operational definitions. There are sufficient details of methods and analysis provided to allow replication by others. The statistical analysis and its interpretation are appropriate. Statistical methods are clearly described and used the correct method of analysis to fulfill the objectives. All the source data underlying the results are available to ensure full reproducibility. The conclusions drawn are adequately supported by the results.\nGeneral comments Please describe one or two sentence about pre-test if possible. If validity was ensured by obtaining the three experts’ comments, is it possible to show content validity index? If not, it’s OK. Please move the description about the categorization of level of awareness of postnatal danger signs under the findings section to include under the “assessment of variables”. Similar comment is for the definition of skilled care provider.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3884",
"date": "07 Aug 2018",
"name": "Aye Sandar Mon",
"role": "Author Response",
"response": "We would like to express our sincere gratitude and appreciation to Dr. Myo Myo Mon for your effort on this manuscript. Regarding the pretest, 25 reproductive-aged mothers who had under 2-years aged children from the selected rural area of Yangon Region were interviewed using same questionnaire that was constructed based on WHO Recommended Interventions for Improving Maternal and Newborn Health, WHO Guidelines Approved by the Guidelines Review Committee: Pregnancy, Childbirth, Postpartum and Newborn Care: A Guide for Essential Practice and Myanmar multiple indicator cluster survey. We are truly sorry that we cannot mention the content validity index. We agree your suggestion to add the operational definition and categorization of the variables, level of awareness on postnatal danger signs and type of maternal health care providers, under the subheading “Assessment of variables”."
}
]
},
{
"id": "36684",
"date": "08 Aug 2018",
"name": "Win Myint Oo",
"expertise": [
"Reviewer Expertise Epidemiology including research methodology and data analysis",
"Biostatistics",
"Public Health",
"NCDs",
"Infectious Diseases",
"Maternal and Child Health"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPage 3(Methods) You should describe the number of selected villages included in this study. Actually it was 21; isn’t it\n\nPage 3 (Study Participants) How did you get the sample size “500”; by using formula or soft ware? You better describe the sample size calculation in details. What do you mean that the sample size was estimated to be 500, based on the multiple logistic regression analysis, as described previously? Where and how you describe it? I cannot find it at anywhere! And you have selected four (4) from 26 townships and five (5) villages from each selected township. It made the number of villages included in your study to be (20). However, the actual number of villages included was (21). It is better to explain why it becomes 21, rather than 20.\n\nPage 3 (Assessment of Variables) Do we need to describe with/from whom the respondents received (or) sought PN care; either MW (mid-wife) or LHV (lady health visitor) in the operational definition of “utilization of full PNC? This is just my suggestion. You can decide not to describe! There are three types of determinants/factors associated with the outcome variable (independent variables such as age, education, average monthly per-capital income, male involvement and accessibility to PNC service, explanatory variables such as knowledge of post-natal danger signs and perception on TBAs, and important independent variables such as birth order, number of AN visit and misconception regarding post-natal practices) in your study! Why do you classify these into three categories? Based on what (or) Why? Is it possible to describe simply (for example, the variables considered as determinants of utilizing full PNC in this study were ------ [or] the present study considered -------- as independent variables)? Or main independent variables were ---- in this study. Those variables ------ were considered as confounders in the present study Please try to be consistent in utilizing the name of variables. (At first you use the name of independent variable as accessibility to PNC service but later you use its name as accessibility to the [nearest] health center. Accessibility to the nearest health center seems to be more relevant. Another one is concerned with perception on TBA. Although you use the term perception, later you changed it to acceptance. I think “acceptance” is more appropriate.\n\nPage 6 (Discussion) Please insert the in-text citation for the description of (less than 50,000 MMK; the amount below the international poverty line as determined by the World Bank).\n\nPage 8&9 (Conclusion) Based on the findings of your study, maternal education level/status was found to be significantly associated with utilization of full PNC. Therefore, you better add the statement/conclusion regarding with that variable into your conclusion! (For example, maternal education or education of women especially in rural area should be enhanced/improved in order to promote the utilization of full PNC or something like that).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36686",
"date": "13 Aug 2018",
"name": "Htin Zaw Soe",
"expertise": [
"Reviewer Expertise ‘Public Health’ focusing on vector-borne diseases especially malaria and dengue",
"neglected tropical diseases particularly schistosomiasis",
"nutrition and food safety",
"and cancer."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 is clearly and accurately presented with current literature. The study design is appropriate. The work is found to be reproducible. Statistical methods are suitable and interpretations are appropriate. Conclusions are supported by results.\n\nIn addition to these points, reviewer’s comments are provided as follow.\n\nMethods:\nStudy participants: Add ‘a’ in the sentence of ……….by applying a multistage cluster sampling method. Assessment of variables: Author uses interchangeably the words – ‘explanatory’ and ‘independent’, both of which have same meaning. To avoid reader’s confusion, the author should use only one word – either ‘explanatory’ or ‘independent’. Or use like ….independent (explanatory) variable and then continuously use ‘independent’ only throughout the text. Data analysis: In description of p values, in the text, small letter ‘p’ is used, and in the tables (3) and (4), capital letter ‘P’ is used. Author should use only one type either small or capital to have consistency throughout the text. The small one is better\nResults:\nAuthor uses the words – ‘spouse’ in the text and ‘husband’ in the table. Should use only one type. Table (1): Table construction should be like this. In Table (2) heading, ‘Factors related to’ is used and in the text ‘Factors relating to’ is used. Use only one type. Table (2): Table construction should be like this. Table (3) and (4): It is better if author constructs two columns separately for – ‘Variables’ - and ‘category’ as suggested above.\n\nDiscussion:\nThe prevalence of full PNC utilization should be described with 95% CI, rather than 25.2%. The word ‘our’ should not be used. Use ‘the present study’ instead of ‘our study’. Add ‘be’ in the sentence of …..The possible explanation might be that low income….. Add ‘were’ in the sentence of …..their knowledge about maternal health were significantly related…….. In the last sentence of Discussion, one related reference should be added to support the statement of ‘a 2 year recall period was selected to minimize this bias’.\nConclusion:\n\nThe determinants should not be described again because they have been already described before. In the last sentence of Conclusion, use ‘is’ instead of ‘should be’, reflecting an\n\nauthor’s strong suggestion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1167
|
https://f1000research.com/articles/7-164/v1
|
08 Feb 18
|
{
"type": "Research Article",
"title": "A retrospective study on atrial fibrillation after coronary artery bypass grafting surgery at The National Heart Institute, Kuala Lumpur",
"authors": [
"Ahmad Farouk Musa",
"Chou Zhao Quan",
"Low Zheng Xin",
"Trived Soni",
"Jeswant Dillon",
"Yuen Kah Hay",
"Rusli Bin Nordin",
"Chou Zhao Quan",
"Low Zheng Xin",
"Trived Soni",
"Jeswant Dillon",
"Yuen Kah Hay",
"Rusli Bin Nordin"
],
"abstract": "Background: Atrial fibrillation (AF) is common after cardiac surgery and has been associated with poor outcome and increased resource utilization. The main objective of this study is to determine the incidence of POAF in Malaysia and identify the predictors of developing POAF. The secondary outcome of this study would be to investigate the difference in mortality and morbidity rates and the duration of intensive care unit (ICU), high dependency unit (HDU) and hospital stay between the two. Methods: This is a retrospective single-center, cross sectional study conducted at the National Heart Institute, Malaysia. Medical records of 637 who underwent coronary artery bypass grafting (CABG) surgery in 2015 were accrued. Pre-operative, operative and post-operative information were subsequently collected on a pre-formulated data collection sheet. Data were then analyzed using IBM SPSS v23. Results: The incidence of POAF in our study stands at 28.7% with a mean onset of 45±33 hours post operatively. Variables with independent association with POAF include advancing age, Indian population, history of chronic kidney disease, left ventricular ejection fraction and beta-blocker treatment. The mortality rate is significantly higher statistically (p < 0.05), and similarly the incidence of stroke. The incidence of other post-operative complications was also significantly higher statistically. The duration of ICU, HDU and hospital stays were statistically longer (p < 0.001) with higher rates of ICU readmissions and reintubations seen. Conclusion: We conclude that the incidence of POAF in Malaysia is comparable to the figures in Western countries, making POAF one of the most commonly encountered condition after CABG with similar higher rates of mortality, poor outcomes and longer duration of stay, and therefore increased cost of care. Strategies to reduce the incidence of AF after cardiac surgery should favorably affect surgical outcomes and reduce utilization of resources and thus lower cost of care.",
"keywords": [
"atrial fibrillation",
"coronary artery bypass grafting",
"incidence",
"predictors",
"outcomes"
],
"content": "Introduction\n\nAtrial fibrillation (AF) is the most common arrhythmia after cardiac surgery with a reported incidence of 15–40%, a significant leap in incidence compared to the normal population1. Patients who underwent valve surgery or combined valve and coronary artery bypass grafting (CABG) have higher incidence of postoperative AF (POAF) than patients having CABG alone2. AF is especially common after mitral valve surgery, occurring in as many as 64% of patients3. POAF is self-limiting in most cases, but even when it is unlimited, it requires additional medical treatment and a prolonged hospital stay, and it consequently increases the cost of operative treatment4–6.\n\nPOAF is generally associated with an increased risk of morbid consequences and even mortality. The risk of developing cerebral infarction is more than doubled in patients with POAF, with a reported incidence of 5.3% compared to patients without POAF, which has a reported an incidence of 2.4%7. Stroke remains the main morbid event for POAF, as the recurrence rate for stroke stands at 20% per year regardless of whether it is chronic or intermittent AF. Other complications such as post-operative congestive heart failure, renal failure, infection and neurological changes all experienced significant increases in patients with POAF2,7.\n\nPOAF is also noted to increase the risk of mortality. A two-fold increase is reported in both the short-term mortality rate (1.9% to 3.6%, p<0.001) as well as the 1-year mortality rate (3.4% to 6.9%, p<0.001) in patients with POAF compared to their healthy counterparts. A similar pattern can also be seen in the 4-year mortality rate (8.8% to 15%) and the 8 year mortality rate (17% to 29%)8. In addition, patients with POAF are reported to be at a higher risk of dying from embolism9.\n\nThe duration of hospital stay of POAF patients is extended as well (10 days in POAF group vs 7 days in non-POAF group)9,10. Furthermore, patients with POAF are also more likely to be re-intubated and readmitted11.\n\nCo-morbid history. Several risk factors have been identified to increase a patient’s possibility of developing POAF12. Advanced age has consistently been related to a higher incidence of POAF. Aging has been associated with remodelling of the atria13.\n\nAnother pre-operative risk factor that has been consistently significant is cardiac remodelling in the form of left atrial dilatation and left ventricular hypertrophy, which may result in diastolic dysfunction14. These risk factors may be related to each other as advanced aged is associated with degenerative and inflammatory changes in the cardiac anatomy. These anatomic changes provide triggers for an episode of POAF15. On top of that, underlying cardiac disease, such as hypertension and valvular abnormalities, which are more common in the elderly, can cause cardiac remodelling as well. These two are also independent predictors of POAF16.\n\nMale gender has inconsistently been associated with POAF. Sex differences in ion-channel expression and hormonal effects on autonomic tone may explain this disparity17. Recently, metabolic syndrome (MS) as a risk factor to POAF has been brought to the attention of clinicians. Echahidi et al found a significant increase in the incidence of POAF in patients with mild as well as moderate-severe obesity18. On top of that, pre-operative low-density lipoprotein cholesterol (LDL-C) was found to be significantly higher in patients who developed POAF, enhancing the predictive value of a pre-operative lipid profile19.\n\nOther pre-operative risk factors include chronic renal failure, previous history of AF and a history of rheumatic heart disease20. Hypothyroidism has been linked to an increased incidence of POAF in a recent small retrospective review21. It remains to be seen whether this linkage is reproducible in future large cohort studies but this finding is exciting as hypothyroidism is an easily treatable condition.\n\nIntra-operative. Valve surgery is the most consistently linked risk factor to POAF after cardiac surgery. Several authors have expressed the possibility that the anatomical changes in the atrium secondary to valvular disease are more important than the procedure itself14.\n\nA longer aortic cross-clamp time is associated with an increase in the number of patients with POAF. Mathew et al reported a 6% increase in incidence of POAF with every fifteen-minute increase in cross-clamp time14. Surgical practices such as bicaval venous cannulation and pulmonary vein venting have been found to significantly increase the incidence of POAF. It is believed that these practices invade the already vulnerable atrium, which causes further discordance in electrical conduction14.\n\nPost-operative. Postoperative risk factors for POAF are less described but several risk factors are linked to POAF especially electrolyte imbalance such as hypomagnesemia and hypokalaemia22. It was noted that 69% of patients that eventually developed AF had hypokalaemia while only 24% of those who did not develop AF had hypokalaemia. Besides that, magnesium supplementation pre-operatively or early post-operative period in hypomagnesemia patients has shown to be preventative of AF. This effect is not apparent in those who have normal magnesium levels post-operatively22–25.\n\nIt is noted that patients with POAF also require higher amount of post-operative usage of inotropic agents26, though it is unclear whether post-operative usage of inotropic agents has a causative mechanism or it is simply used in patients who are relatively unstable and hence are already at a higher risk of developing POAF.\n\nA recent retrospective study also found a strong positive correlation can be seen between the time of POAF onset with the time of maximum blood sugar concentration27.\n\nSo far no data have been obtained on the incidence, predictors, and outcome of POAF in our population in Malaysia. This present retrospective study would form the baseline knowledge about this condition, and to compare it with Western figures, and would then form a basis for a possible prospective intervention in trying to reduce the occurrence of POAF.\n\n\nMethods\n\nThis is a single-centre, cross-sectional study of patients who underwent CABG at the National Heart Institute (IJN), Kuala Lumpur, Malaysia. Patient’s medical records were reviewed retrospectively to identify pre-operative factors that predict the development of atrial fibrillation in the post-operative period as well as their outcomes.\n\nEthical approval was obtained through the National Heart Institute Ethics Committee (IJNEC/16/15). No amendments were made throughout the duration of the study. This study was approved by the Monash University Human Research Ethics Committee (CF16/1984 – 2016001004). No patient consent was required by the ethics committees for this study, since there was no patient contact.\n\nThe sample size was calculated for this study using the Raosoft® sample size calculator28. The figure of 30% was expected for the incidence of POAF, consistent with previous studies9,10,12,13. With the total number of CABG done in Malaysia being unknown, an overestimated number of 20,000 cases a year was used for the purpose of this calculation. Using these figures, a sample size of 318 patients will be able to investigate the outcome with a 5% margin of error and a 95% confidence interval. We managed to obtain data of 637 patients, which further reduced the margin of error of our study to 3.5% with a 95% confidence interval.\n\nEvery patient who underwent CABG in IJN over the year of 2015 from January until June was included in this study. Patients who underwent CABG with an additional cardiac surgery in the same sitting were included in the study as well. Conversely, patients who underwent other cardiac surgery without CABG were not included in this study.\n\nThe exclusion criteria for this study is patients who had documented AF prior to CABG. Patients who were in sinus rhythm prior to CABG but had previous documented AF, including paroxysmal or intermittent episodes, were excluded from this study as well. This is to ensure that in any case a patient developed POAF, that episode of AF would be his/her first episode or what is known as recent-onset AF. Besides that, patient medical records would need to be complete, with more than half of the variables available for collection. A pre-formed data collection sheet was used to collect all variables (Supplementary File 1). A total of 680 patients fit the inclusion criteria. 43 patients were rejected in accordance to the exclusion criteria. The remaining 637 patients were included for analysis.\n\nStatistical analysis was carried out using IBM SPSS 23.0.\n\nThe prevalence of categorical variables was determined using frequency tables, while continuous variables were described in terms of mean and standard deviation. With these calculations, the general characteristics of the sample was established.\n\nNormality of continuous variables were tested using the degree of skewness and kurtosis. The patients were then grouped according to the onset of POAF and comparison of characteristics between groups were done. Categorical variables were compared using cross tabulations and Chi-square test or Fisher’s exact test was employed to assess significance, whichever more suitable. Student’s t-test was used to compare differences in mean of normal continuous variables between two groups. A p-value of less than 0.05 was considered statistically significant.\n\nThe secondary outcome of the study in terms of duration of ventilation, hospital stay, ICU stay and high dependency unit (HDU) stay did not follow the normal distribution, hence they were expressed with median value and interquartile range. We decided to establish a cut-off value to delineate between normal and prolonged duration. The cut-off value was determined with reference to hospital protocol.\n\nWe cross-tabulated the onset of POAF with the newly categorised groups and utilised Chi-square test or Fisher’s Exact test to assess significance. Additionally, the same analysis was used to compare the prevalence of post-operative complications between patient who developed POAF and patients who did not.\n\nIn the last part of the analysis, the group of patients who developed POAF were selected and split into two groups according to the duration of POAF and number of episodes they experienced. The groups were then cross-tabulated with the post-operative complications as well as prolonged ventilation and stay. Chi-square test or Fisher’s exact test were employed to assess significance.\n\nIn order to identify potential predictors to be included in the multivariate logistic regression modelling, univariate analysis was performed. The relationship between each pre-operative variable and the onset of POAF was explored. Variables with p value of less than 0.25 were included in the multivariate logistic regression. Besides that, variables that were shown to be significant predictors in previous studies were included in the multivariate logistic regression model as well. Variables included in the multivariate logistic regression model were assessed for collinearity.\n\n\nResults\n\nThe total number of patients included in this study was 637 patients. The patients included in the study sample are pre-dominantly elderly males, representing 82.8% of the cohort with a mean age of 60.66 ± 8.2 years old. The distribution of the patient’s age in our study follows a normal distribution with a minimum age of 30 and a maximum of 91. The majority of the patients (58.2%) were Malays. The Chinese population, standing at 17.6% of our study sample is slightly under represented, while the Indian population is slightly over represented at 22.1% of our sample.\n\nThere is a statistically significant difference between the mean age of the groups of patients without POAF and those with POAF (60.00 years vs 62.31 years, p=0.002). Besides that, the difference between the non-POAF and POAF group also achieve statistical significance in between the major populations in Malaysia (p=0.001). The Malay population has the highest incidence of POAF at 33.2%, followed by the Chinese population at 27.7%, while the Indian population has the lowest incidence of POAF at 16.3%.\n\n*Total n varies slightly for each item due to a small amount of missing data in each.\n\nIn total, 183 patients in our study developed AF in the post-operative period, representing 28.7% of the sample. The median time of development of POAF is 45 hours after surgery. The majority of POAF developed within 3-days after surgery with the 2nd day being the most common. Only 15% of the POAF in our study developed after more than 3-days post-operation. 95.6% of AF lasts less than 48 hours, while slightly more than one third of the patients had more than one episode of AF in their stay. However, only one patient was discharged with AF in our study, the rest were discharged with sinus rhythm. Of the 91.3% of patients who were treated for AF, amiodarone was the most common intervention, used in 82.5% of the patients. Other interventions used to treat AF were digoxin, beta-blocker and electrical cardioversion.\n\n*Total n varies slightly for each item due to a small amount of missing data in each.\n\nConsidering that Malaysia has one of the highest prevalence of obesity in the world, the median BMI of 26.29 kg/m2 was not unexpected. We noted that 63.7% of the study sample was categorised as overweight and 17.3% as obese, according to the Asian guidelines29. There was no significant difference between the two groups.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nIn terms of functional status, the patients in our study have a relatively normal functional status with 350 (56.7%) of the patients classified as New York Heart Functional Class (NYHA) I and 233 (37.8%) classified as NYHA II. By contrast, 34 (5.5%) patients were classified as NYHA III and none classified as NYHA IV. On top of that, the median left ventricular ejection fraction of our patients is 50.5%. There is a statistically significant difference in terms of ejection fraction between groups (51.0% in the non-POAF group vs. 49.0% in the POAF group, p=0.035).\n\nThere were 180 (28.5%) patients in total who has left atrial enlargement before operation. In total, 115 (25.6%) of them belong to the non-POAF group, while 65 (35.7%) were from the POAF group. The difference between the groups are statistically significant (p=0.010).\n\nAmong the co-morbid conditions that we investigated in our study, hypertension, diabetes mellitus and hypercholesterolemia are the three commonest diagnosed pre-operative co-morbidities in the patients in our study. We noticed 75.4% of our patients had previously been diagnosed with hypertension, 54.8% had diabetes mellitus and 69.5% had hypercholesterolemia. In total, 276 patients (43.3%) were previous or current smokers. Amongst all the investigated pre-operative medical conditions, chronic kidney disease (CKD) and end-stage renal failure with dialysis were the two conditions that observed a statistically significant difference between two groups. A total of 26 patients with previously diagnosed CKD also developed POAF (14.2%), compared to the 28 patients who did not develop POAF (6.2%). Similarly, 11 patients who underwent dialysis due to end stage renal failure before operation developed POAF, compared to the 6 who did not develop POAF.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nThe medications list tabulated below reflects the treatments given to our patient’s underlying medical condition. Anti-hypertensive, anti-lipid, anti-platelets and anti-angina agents were present in more than half of our patients’ medication list. The abundant prescription of these medications seen in our patients is not surprising as these are the staple medications used in cardiovascular diseases. Amongst the medications, beta-blocker is the only medication that has a statistical significant difference between the two groups. In total, 292 patients (64.3%) of the patients in the non-POAF group were prescribed beta blockers, while 134 (73.2%) of the patients in the POAF group were prescribed beta-blockers (p=0.031).\n\nAssociation between post-operative atrial fibrillation (POAF) and individual medication is presented.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nThe commonest performed surgery in our patients is CABG alone, performed in 586 patients (92.0%). Compound surgeries are relatively uncommon, only performed in 51 of the cases (8.0%). Amongst the compound surgeries, CABG and valve is the most commonly performed, representing 6.3% of the total surgeries performed. Mitral valve surgery is the most commonly performed valve surgery, representing 61.4% of the valve surgery performed with CABG in our patients.\n\nCABG, coronary bypass grafting.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nThe vast majority of the patients underwent on-pump CABG (608 patients, representing 95.4% of the total). Amongst them, 178 patients were in the POAF group, representing 97.3% of the POAF group; 430 patients were in the non-POAF group, representing 94.7% of the non-POAF group. However, the difference is not statistical significant. A total of 25 patients underwent off-pump CABG (3.9), while 4 patients underwent on pump beating heart CABG (0.6%).\n\nThe median bypass time is 82 minutes, with a minimum bypass time of 32 minutes and a maximum bypass time of 300 minutes. On the other hand, the median cross-clamp time is 64 minutes, with a minimum cross-clamp time of 22 minutes and a maximum cross-clamp time of 225 minutes. There was no statistically significant difference between the two groups in terms of bypass and cross-clamps times.\n\nUnivariate analysis was performed on every recorded variable to identify potential predictors to be included in the multivariate model. Age, Malay population, Indian population, history of CKD, left ventricular ejection fraction, beta-blocker, biguanide, left atrial enlargement and CABG plus valve surgery were identified as significant predictors in the univariate analysis (p<0.05).\n\nAge, Indian population, history of CKD, left ventricular ejection fraction and beta blocker were identified as independent predictors in our study after multi-variable adjustments. With every additional year a patient advances, there is a 4.5% additional odds of developing POAF (AOR=1.045, 95% CI, 1.019 – 1.072, p=0.001). The Indian population is less likely to develop POAF compared to the other populations, with a 56% reduction in odds (AOR=0.440, 95% CI, 0.253 – 0.765, p=0.004). The Malay population had a significant association with POAF in the univariate analysis, but the significance was lost after adjustment.\n\nCABG, coronary bypass grafting.\n\nHistory of CKD is another significant predictor of POAF as patients with CKD have more than double the odds of developing POAF (AOR = 2.124, 95% CI, 1.057 – 4.268, p=0.034). Left ventricular ejection fraction also has a significant association with POAF. With every 10% increment in the ejection fraction value, there is a 23% reduction of odds of developing POAF (AOR=0.977, 95% CI, 0.958 – 0.996, p=0.016). Lastly, pre-operative beta-blocker treatment has a positive association with POAF, as well with a 61% increase in odds over those who were not under beta-blocker treatment (AOR = 1.611, 95% CI, 1.049 – 2.467, p=0.029).\n\nThe mortality rate of our study stands at 2.7%. The mortality rate of the non-POAF group is 1.8% compared to 4.9% in the POAF group. There is a statistically significant difference between the mortality rates of both groups.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nFour patients in total developed stroke in the post-operative period. Three patients are from the POAF group and one from the non-POAF group, yielding incidence of stroke of 1.6% and 0.2%, respectively. However, the difference of stroke incidence between the two groups did not achieve statistical significance.\n\nOn top of that, 30 patients (16.4%) in the POAF group were re-operated while 29 patients (6.4%) in the non-POAF group were re-operated, a statistically significant difference was observed. Other post-operative complications that achieved statistical significant difference between two groups are renal failure requiring dialysis (2.0% in non-POAF group vs. 6.6% in the POAF group, p=0.003), pulmonary complications (6.4% in non-POAF group vs. 12.0 in the POAF group, p=0.018) and post-operative fever (1.8% in non-POAF group vs. 4.9% in POAF group, p=0.025).\n\nThe differences between the duration of ICU stay, HDU stay and hospital stay between the two groups were statistically significant. The median duration of ICU stay including readmission was 1648 minutes in the non-POAF group, compared to 3542 minutes in the POAF group, p<0.001. A total of 100 patients had prolonged ICU stay in the POAF group, accounting for more than half of the patients with POAF (54.9%). On the flipside, 136 patients in the non-POAF group had a prolonged ICU stay, accounting to 30.2% of those who did not develop POAF, p<0.001.\n\nICU, intensive care unit; HDU, high dependency unit.\n\n*Total n varies slightly for each item due to a small amount of missing data in each\n\nThe median duration of HDU stay including readmission in the non-POAF group was 1590 minutes, compared to 2430 minutes in the POAF group, p=0.003. In total, 38 patients (35.5%) experienced a prolonged HDU stay in the POAF group, compared to 36 (21.4%) in the non-POAF group, p=0.010.\n\nThe median duration of hospital stay was 7.2 days in the non-POAF group, while patients in the POAF group had a median duration of stay in the hospital of 9.0 days (p<0.001). A total of 37 patients (5.7%) in the POAF group had prolonged hospital stay, compared to 26 (20.2%) from the non-POAF group (p<0.001).\n\nThe duration of ventilation including reintubation was significantly different statistically in both groups as well. The median duration of ventilation in the non-POAF patients was 1155 minutes, compared to the 1277.5 experienced by the patients with POAF. In total, 62 (34.1%) patients had prolonged ventilation in the POAF group compared to 77 (17.1%) in the non-POAF group.\n\nThere is a significant difference in the incidence of readmission into ICU between the two groups. In total, 9 patients were readmitted into the ICU (4.9%) in the POAF group, while 8 patients were readmitted in the non-POAF group (1.8%), p=0.029. But the readmission rates into the HDU and hospital were not statistically significant between the two groups. However, reintubation rates between the two groups were significantly different statistically with a 4.9% reintubation rate in the POAF group and 0.4% reintubation rate in the non-POAF group, p<0.001.\n\n\nDiscussion\n\nAF is one of the most common adverse events after cardiac surgery occurring in slightly more than a quarter of patients in our study. Reported incidences range from 10–65%, depending on patient profile, type of surgery, method of arrhythmia surveillance, and definition of arrhythmia30–32.\n\nWhile the precise pathophysiology of POAF is unknown, most of the evidence suggests that it is multifactorial. A common underlying factor associated with POAF induced by mechanical, metabolic or pharmacologic stimuli is the redox changes in atrial tissue associated with tachyarrhythmia33.\n\nPOAF adversely affects mortality and morbidity, and consequently leads to a longer hospital stay and higher costs related to the cost of care26,34. There are certain pre-operative factors that predict the development of POAF. In our study, we found that advancing age, history of CKD, non-Indian populations, low left ventricular ejection fraction and pre-operative beta-blocker treatment are independently associated with the development of POAF.\n\nThe strength of our study lies on the large sample size with a good racial distribution among the three major populations in Malaysia, allowing good representation of the multiracial background. This being the first study of its kind in Malaysia provides a good background of the incidence of POAF and its outcomes for future studies in this country.\n\nThe incidence of POAF at the National Heart Institute (IJN), Malaysia, stands at 28.7% in the present study, which is within the range of most large series that report an incidence of around 30%26. Since this is, to our knowledge, the first study in Malaysia to investigate POAF, we do not have any previous studies to compare our figures to.\n\nSingapore is a country that lies in close proximity to Malaysia, shares the same multiracial population background as Malaysia and serves as a good point for comparison. In a recent prospective study conducted in heart institutes in Singapore, the recorded incidence of POAF was 17.3%, much lower than the incidence of 28.7% observed in our study. One possible explanation to this disparity between the incidences could be due to the relatively stringent criteria employed by the investigators in that study. Only AF episodes lasting more than 1 hour were considered as POAF in that study, whereas we included every episode of documented AF regardless of duration35.\n\nCompared to other studies conducted mainly in the Western population, our incidence of POAF is similar to those conducted in the United Kingdom, Europe and the United States34. The findings from our study supports the notion that POAF is a commonly occurring arrhythmia in the post-operative period.\n\nThe characteristics of POAF that were described in previous studies were transient episodes of AF, usually occurring between two to four days after operation with a high rate of recurrence36,37. The characteristics of POAF episodes observed in our study is similar to the description from previous studies.\n\nIn our study, the median time of development of POAF was 45 hours after operation with the majority of the first POAF episodes happen within three days after operation. The recurrence rate of POAF was 36.6%, comparable to the 40% recurrence rate seen in previous studies36,37. Knowing the characteristics of POAF will allow treating healthcare professionals to anticipate POAF better and be more vigilant in terms of monitoring and management.\n\nWe decided to select 48-hours as the POAF duration cut-off to categorize patients in favor of the more commonly used 24-hour cut-off period in an attempt to identify patients who required prophylactic anti-coagulation. In total, 8 of our patients had AF duration of more than 48-hours and were recommended to be started on prophylactic anti-coagulation for thromboembolic prevention in accordance to the management guideline38,39.\n\nIn addition, 2 patients passed away in the post-operative period, 3 were not started on any anti-coagulation therapy and the other 3 were started on warfarin, but this was most likely due to their respective valve replacements. The conservative approach of the treating team in IJN in terms of prophylactic anti-coagulation is most likely due to the relatively low incidence of stroke weighed against the high incidence of post-operative bleed in this hospital.\n\nUnivariate analysis was performed to every recorded pre-operative and operative variable in an attempt to identify potential predictors to be included in the final multivariate model. Variables that had significant association with POAF before adjustments were age, Malay population, Indian population, history of CKD, left ventricular ejection fraction, beta-blocker, biguanide, left atrial enlargement and CABG plus valve surgery (p<0.05) Other variables that were identified as potential predictors (p<0.25) include history of COPD, BMI, calcium channel blocker, insulin, diuretics, inhaled beta agonist, off-pump CABG surgery, bypass time, cross-clamp time and CABG plus valve plus other surgery.\n\nAll of the potential predictors that were mentioned above were included in the final multivariate logistic regression analysis to identify independent predictors. Hypertension, diabetes mellitus, hypercholesterolemia, smoking status and HMG CoA were included in the multivariate logistic model, as they were shown in previous studies34–36 to be significant independent predictors of POAF.\n\nAfter adjustment, age, Indian population, history of CKD, left ventricular ejection fraction and beta-blocker remained as variables with independent association with POAF.\n\nAge. Advancing age has been one of the most described pre-operative risk factor for POAF with strong associations established in previous studies. Mathew et al estimates that in patients over the age of 70, every 10-year increment in age will yield a 75% increase in the odds of developing POAF34. The associations were not as strong in our model, with a 45% increase in odds of developing POAF for every decade of increment in patient’s age. However, we would like to note that our model is not age restricted.\n\nIt is not surprising that the association between advancing age and AF is not restricted to the post-operative period. A strong association between advancing age and AF is seen in the general population as well with a higher incidence of 2–4% seen in the elderly population compared to the 0.4% seen in the younger population39. The strong association is due to the age-related changes in cardiac anatomy seen in the elderly, making them more prone to develop AF40. Along with a higher incidence of cardiac related co-morbid conditions acting as stressors to the heart, cardiac surgeries act as triggers to the eventual development of AF in the elder population.\n\nIndian population. With the majority of the POAF-related studies conducted in Western countries, the association between the multiracial background in Asian countries like Malaysia is still relatively unknown. Our study found a significantly lower-odds of developing POAF in the Indian population compared to their counterparts of other racial origins. The study conducted in Singapore found a similar trend, reporting a higher-odds of developing POAF in patients from Chinese and Malay populations compared to those of Indian ethnicity. To our knowledge this is the only other study that investigated this association35.\n\nThe relative resistance of developing POAF observed in the Indian population in our study could be traced back to their genetic lineage from India. According to a multinational study, the reported incidence of POAF in India was 15.7%, almost half of the incidence of POAF in Malaysia34. On top of that, the prevalence of AF in the general population in India is relatively low as well, with an overall incidence of 0.39% as reported by the PINNACLE registry41. More studies, both clinical and scientific, needs to be done to identify the reason behind this resistance of AF seen in the Indian population. Exploring this subject might bring about new theories on the still relatively unknown pathogenesis of POAF.\n\nChronic kidney disease (CKD). Among other pre-operative co-morbid conditions investigated in our study, CKD is the only one with significant association after adjustments. Patients with a history of CKD are more than two times more likely to develop POAF compared to those with normal kidney function (AOR = 2.124, 95% CI, 1.057 – 4.268, p=0.034). This association is consistent with those found in previous studies26,42.\n\nThe strong association found between CKD and AF in this study is most probably due to the pro-inflammatory state experienced by CKD patients. In fact, this association is not limited to chronic form of kidney disease. A recent study investigating the relationship between POAF and post-operative acute kidney failure found a significant association as well, suggesting a possible linkage between the pathogenesis of AF between the acute and chronic form of kidney disease42.\n\nLeft ventricular ejection fraction. Left ventricular ejection fraction is an estimate of the functioning status of the heart. Since congestive heart failure was described in previous study to be associated with an increased-odds of developing POAF, it does not come as a surprise that a high ejection fraction, indicating a high level of functioning of the heart, reduces the odds of developing POAF.\n\nBeta-blocker. Lastly, the surprising positive association between pre-operative beta-blocker treatment and POAF was found in our study. Patients who were prescribed beta-blockers were 61% more likely to develop POAF compared to those who were not prescribed beta-blockers. This result is in contradiction to other studies and meta-analyses43 that concluded that beta-blockers are effective in reducing POAF if given in the prophylactically in the pre-operative period. In fact, guidelines have recommended the prescription of beta-blockers to low-risk individuals to prevent POAF.\n\nOne of the possible explanations of this phenomenon could be due to the rebound effect when patients were taken off beta-blockers during the post-operative period. Usually the recommencement of beta-blockers in the post-operative period is dependent on the preference of the treating team and the stability of the patient. In fact, the cessation of beta-blocker was one of the risk factors of developing POAF34. Unfortunately, one of the limitations of our study was the lack of detection of the recommencement of beta-blockers in the post-operative period, rendering our study underpowered to confirm this association.\n\nOther variables. Certain variables, such as left atrial enlargement, longer bypass time and longer cross-clamp times, were associated with an increased-odds of developing POAF in previous studies. However, the association between these variables and POAF was not found to be significant in our study after adjustments.\n\nPre-operative co-morbid conditions, such as hypertension, diabetes, hypercholesterolemia and COPD, were also shown in previous studies to be independent predictors of POAF44–46; however this association could not be established in our study.\n\nAlthough the incidence of POAF in patients who underwent compound surgeries were higher compared to those who underwent CABG alone, the association was not significant after adjustment. However, we noted that the proportion of patients who underwent compound surgery was much less compared to those who underwent CABG alone. On top of that, it is likely that patients who are undergoing compound surgery, especially mitral valve related, to have underlying AF and were subsequently excluded from our study.\n\nMortality. The mortality rate observed in our study is 2.7%, slightly lower than the mortality rates seen in other studies. The mortality rate in the POAF group is 4.9%, compared to the 1.8% seen in the non-POAF group, a statistically significant difference. This was about three-fold increase in mortality rate in AF patients compared with patients without AF.\n\nA similar difference in mortality rates were seen in other studies7,47–48. POAF does not only affect the short-term mortality but long-term mortality as well. This finding demonstrates the not-so-benign nature of this arrhythmia.\n\nIt is thought that the mechanisms by which postoperative AF is associated with mortality are speculative. This might be due to hemodynamic compromise and heart failure directly as a result of the loss of atrial transport function may certainly contribute. In the long term, mechanistic and causal links are more difficult to establish. Possibilities include the development of heart failure with its attendant mortality risk, the occurrence of disabling stroke or other embolic catastrophes, and adverse drug effects, such as pro-arrhythmia with antiarrhythmic drugs or haemorrhage with anticoagulants.\n\nMorbidity. We observed that all of the investigated post-operative complications have higher rates in the POAF group compared to the non-POAF group. One of the most devastating complications of morbidity due to AF is the increased incidence of stroke. This is thought to be largely due to circulatory stasis in the left atrium resulting in the formation of an embolus.\n\nThere was a six-fold increase in the incidence of stroke, one of the feared sequelae of AF, in the POAF group compared to the non-POAF group (1.6% vs 0.2%). However, the difference did not achieve statistical significance due to the low incidence seen in this study.\n\nAttempts to investigate other AF-related complications, such as post-operative myocardial infarction and new heart failure, were quickly abandoned as we realized that a sufficiently accurate representation cannot be obtained from the medical records.\n\nThe investigated outcomes that achieved statistical significance are reoperation, renal failure requiring dialysis, pulmonary complications and post-operative fever. Due to the nature of our study, the interpretation of the outcomes with POAF needs to be perform with caution as a temporal relationship cannot be established. Since the development of AF does not affect the pathogenesis of these outcomes directly, the notion that these complications may trigger the development of AF cannot be ruled out. These complications are generally pro-inflammatory, on top of that patients who underwent reoperation are exposed to more inotropic agents and surgical insults while patients with renal failure are more likely to have electrolyte imbalance and fluid overload, conditions of which may play a role in the pathogenesis of AF.\n\nPost-operative stay. From our study, we noticed that patients with POAF experience significantly longer stay in the ICU, HDU and the hospital (p<0.001). Besides that, the duration of ventilation experienced by patients in the POAF group was significantly longer as well. Extended period of ICU and HDU stay for post-op AF patients may be due to further observation, management and nursing needed for stabilization of haemodynamic status, correction of hypoxia and for conversion of AF status to sinus rhythm.\n\nIn addition, there were higher rates of readmission into ICU, reintubation, readmission into HDU and hospital readmission in the POAF group, with the former two achieving statistical significance. This is common due to the nature of patients who developed AF after surgery are prone to develop other morbid events that need extended care facility, such as ventilatory support. Consequently, this led to a higher usage of hospital resources by the patients with POAF due to the prolonged stay and ventilation, which translates into increased cost to both the patients and the hospital.\n\nThe main limitation to this study lies in its observational nature. As such, temporal relationship was unable to be established on certain variables, especially the post-operative outcomes with regards to post-operative AF. Besides that, the amount of data collected in this study is strictly limited to the amount recorded in the patient medical records. Certain variables that will take our study a step above such as AF related complications, AF treatment related complications, and exact left atrial sizes were simply unavailable or inconsistently documented.\n\nBesides that, another limitation to our study would be the direct role the investigator played during the data collection process, giving rise to the possibility of observer bias. However, in order to reduce the influence of bias, part of the list of variables were collected by blinded data collectors, away from the influence of the investigator.\n\nMoving forward, to obtain higher quality data regarding the incidence of POAF with their respective outcome, a large prospective cohort study with long term follow-up would be indicated. Ideally, patients would be put on 24-hour cardiac monitoring for the first week to detect episodes of AF. Vital signs and post-operative complications have to be charted and cross checked with the time of onset of POAF to detect any trends or patterns. Date of cessation and recommencement of medication will be recorded. Patients will also have to be followed-up over years to detect the incidence of late AF. Resource utilization and cost will probably have to be recorded to allow accurate cost effectiveness analysis.\n\nBesides that, randomized controlled trials can also be conducted to investigate potential preventative interventions. Certain preventative interventions have yielded promising results in earlier studies49–54, more randomized controlled trials may give us enough evidence to adopt new methods in the prevention of POAF. Perhaps prevention of POAF using extracts from natural products like tocotrienol, which is a vitamin-E isomer derived from Palm Oil, that possesses potent anti-oxidative properties sounds promising53,54.\n\nFurthermore, healthcare professionals related to this field should be made aware of the potential implications of POAF on their patients’ health. Local management guidelines and protocols should be modified to reflect the findings of this study.\n\n\nConclusions\n\nAF is the most common complication after cardiac surgery and is associated not only with increased morbidity and mortality, but also with increased costs and longer hospital stay. A number of preoperative patients’ characteristics and intraoperative practice variables appear to affect the incidence of this arrhythmia.\n\nPre-operative and operative predictors identified in this study are consistent with the findings of studies conducted in Western countries. The multiracial population in Malaysia produced a unique study question. Previous conclusion of Asians having a lower incidence of POAF compared to Westerners will need to be revisited as our study suggests that the resistance to POAF is only limited to the Indian population.\n\nThe measurable outcomes of care, such as the need for re-intubation, ventilatory support, longer ICU, HDU and hospital stay, are all associated with postoperative AF, leading to an increase in resource utilization. Strategies to identify the patients at risk and to modify these risk factors by aggressive prophylactic measures, as well as changes in surgical techniques, should lead to lower incidence of AF and a reduced morbidity and mortality rate for patients undergoing cardiac surgery.\n\nWe hope that this study can spark a wave of interest on the subject of POAF as this condition has and will continue to affect the lives of patients. This study has highlighted the need of high quality studies and the need for prevention of this condition that is associated with an increase in morbidity and mortality, from the healthcare providers, especially in Malaysia.\n\nTherefore, we recommend a prospective randomized control trial on the preoperative identification and prophylactic intervention in order to develop an efficient prevention and management strategies in reducing the incidence of POAF.\n\n\nData availability\n\nDataset 1: Raw data for the study ‘A retrospective study on atrial fibrillation after coronary artery bypass grafting surgery at The National Heart Institute, Kuala Lumpur’ both in SAV and Excel formats. DOI, 10.5256/f1000research.13244.d19298255.",
"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\nSupplementary material\n\nSupplementary File 1: Pre-formulated data collection sheet.\n\nClick here to access the data.\n\n\nReferences\n\nJin R, Hiratzka LF, Grunkemeier GL, et al.: Aborted off-pump coronary artery bypass patients have much worse outcomes than on-pump or successful off-pump patients. Circulation. 2005; 112(9 Suppl): I332–7. PubMed Abstract\n\nCreswell LL, Schuessler RB, Rosenbloom M, et al.: Hazards of postoperative atrial arrhythmias. Ann Thorac Surg. 1993; 56(3): 539–549. PubMed Abstract | Publisher Full Text\n\nAsher CR, Miller DP, Grimm RA, et al.: Analysis of risk factors for development of atrial fibrillation early after cardiac valvular surgery. Am J Cardiol. 1998; 82(7): 892–895. PubMed Abstract | Publisher Full Text\n\nKannel WB, Abbot RD, Savage DD, et al.: Epidemiologic features of chronic atrial fibrillation: the Framingham study. N Engl J Med. 1982; 306(17): 1018–1022. PubMed Abstract | Publisher Full Text\n\nAndrews TC, Reimold SC, Berlin JA, et al.: Prevention of supraventricular arrhythmias after coronary artery bypass surgery. A meta-analysis of randomized control trials. Circulation. 1991; 84(5 Suppl): III236–244. PubMed Abstract\n\nHannes W, Fasol R, Zajonc H: Diltiazem provides anti-ischemic and anti-arrhythmic protection in patients undergoing coronary bypass grafting. Eur J Cardiothorac Surg. 1993; 7(5): 239–245. PubMed Abstract | Publisher Full Text\n\nVillareal RP, Hariharan R, Liu BC, et al.: Postoperative atrial fibrillation and mortality after coronary artery bypass surgery. J Am Coll Cardiol. 2004; 43(5): 742–8. PubMed Abstract | Publisher Full Text\n\nKaw R, Hernandez AV, Masood I, et al.: Short- and long-term mortality associated with new-onset atrial fibrillation after coronary artery bypass grafting: a systematic review and meta-analysis. J Thorac Cardiovasc Surg. 2011; 141(5): 1305–12. PubMed Abstract | Publisher Full Text\n\nMariscalco G, Engström KG: Postoperative atrial fibrillation is associated with late mortality after coronary surgery, but not after valvular surgery. Ann Thorac Surg. 2009; 88(6): 1871–6. PubMed Abstract | Publisher Full Text\n\nAuer J, Weber T, Berent R, et al.: Postoperative atrial fibrillation independently predicts prolongation of hospital stay after cardiac surgery. J Cardiovasc Surg (Torino). 2005; 46(6): 583–8. PubMed Abstract | Publisher Full Text\n\nShirzad M, Karimi A, Tazik M, et al.: Determinants of postoperative atrial fibrillation and associated resource utilization in cardiac surgery. Rev Esp Cardiol. 2010; 63(9): 1054–60. PubMed Abstract | Publisher Full Text\n\nOmae T, Kanmura Y: Management of postoperative atrial fibrillation. J Anesth. 2012; 26(3): 429–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPandit SV, Jalife J: Aging and atrial fibrillation research: where we are and where we should go. Heart Rhythm. 2007; 4(2): 186–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMathew JP, Fontes ML, Tudor IC, et al.: A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA. 2004; 291(14): 1720–9. PubMed Abstract | Publisher Full Text\n\nAllessie MA, Boyden PA, Camm AJ, et al.: Pathophysiology and prevention of atrial fibrillation. Circulation. 2001; 103(5): 769–77. PubMed Abstract | Publisher Full Text\n\nAbhayaratna WP, Seward JB, Appleton CP, et al.: Left atrial size: physiologic determinants and clinical applications. J Am Coll Cardiol. 2006; 47(12): 2357–63. PubMed Abstract | Publisher Full Text\n\nMaisel WH, Rawn JD: Atrial fibrillation after cardiac surgery. Ann Thorac Surg. 2001; 1064: 1061–1073.\n\nEchahidi N, Mohty D, Pibarot P, et al.: Obesity and metabolic syndrome are independent risk factors for atrial fibrillation after coronary artery bypass graft surgery. Circulation. 2007; 116(11 Suppl): I213–9. PubMed Abstract | Publisher Full Text\n\nAydin M, Susam I, Kilicaslan B, et al.: Serum cholesterol levels and postoperative atrial fibrillation. J Cardiothorac Surg. 2014; 9: 69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJacob KA, Nathoe HM, Dieleman JM, et al.: Inflammation in new-onset atrial fibrillation after cardiac surgery: a systematic review. Eur J Clin Invest. 2014; 44(4): 402–28. PubMed Abstract | Publisher Full Text\n\nWorku B, Tortolani AJ, Gulkarov I, et al.: Preoperative hypothyroidism is a risk factor for postoperative atrial fibrillation in cardiac surgical patients. J Card Surg. 2015; 30(4): 307–12. PubMed Abstract | Publisher Full Text\n\nChelazzi C, Villa G, De Gaudio AR: Postoperative atrial fibrillation. ISRN Cardiol. 2011; 2011: 203179. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGottlieb SS, Baruch L, Kukin ML, et al.: Prognostic importance of the serum magnesium concentration in patients with congestive heart failure. J Am Coll Cardiol. 1990; 16(4): 827–31. PubMed Abstract | Publisher Full Text\n\nToraman F, Karabulut EH, Alhan HC, et al.: Magnesium infusion dramatically decreases the incidence of atrial fibrillation after coronary artery bypass grafting. Ann Thorac Surg. 2001; 72(4): 1256–61; discussion 1261–2. PubMed Abstract | Publisher Full Text\n\nSvagzdiene M, Sirvinskas E, Benetis R, et al.: Atrial fibrillation and changes in serum and urinary electrolyte levels after coronary artery bypass grafting surgery. Medicina (Kaunas). 2009; 45(12): 960–70. PubMed Abstract\n\nHashemzadeh K, Dehdilani M, Dehdilani M: Postoperative Atrial Fibrillation following Open Cardiac Surgery: Predisposing Factors and Complications. J Cardiovasc Thorac Res. 2013; 5(3): 101–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTatsuishi W, Adachi H, Murata M, et al.: Postoperative hyperglycemia and atrial fibrillation after coronary artery bypass graft surgery. Circ J. 2015; 79(1): 112–8. PubMed Abstract | Publisher Full Text\n\nSample Size Calculator by Raosoft, Inc. [Internet]. Raosoft.com. 2016. [cited 11 June 2016]. Reference Source\n\nThe Asia-Pacific perspective: Redefining obesity and its treatment. IASO International Association for the Study of Obesity; World Health Organization, Western Pacific Region. 2000. Reference Source\n\nBradley D, Creswell L, Hogue CW Jr, et al.: Pharmacologic prophylaxis: American College of Chest Physicians guidelines for the prevention and management of postoperative atrial fibrillation after cardiac surgery. Chest. 2005; 128(2 Suppl): 39S–47S. PubMed Abstract | Publisher Full Text\n\nMartinez EA, Bass EB, Zimetbaum P, et al.: Pharmacologic control of rhythm: American College of Chest Physicians guidelines for the prevention and management of postoperative atrial fibrillation after cardiac surgery. Chest. 2005; 128(2 Suppl): 48S–55S. PubMed Abstract | Publisher Full Text\n\nAlqahtani AA: Atrial fibrillation post cardiac surgery trends toward management. Heart Views. 2010; 11(2): 57–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHogue CW Jr, Creswell LL, Gutterman DD, et al.: Epidemiology, mechanisms, and risks: American College of Chest Physicians guidelines for the prevention and management of postoperative atrial fibrillation after cardiac surgery. Chest. 2005; 128(2 Suppl): 9S–16S. PubMed Abstract | Publisher Full Text\n\nMathew JP, Parks R, Savino JS, et al.: Atrial Fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. MultiCenter Study of Perioperative Ischemia Research Group. JAMA. 1996; 276(4): 300–6. PubMed Abstract | Publisher Full Text\n\nZhang W, Liu W, Chew ST, et al.: A Clinical Prediction Model for Postcardiac Surgery Atrial Fibrillation in an Asian Population. Anesth Analg. 2016; 123(2): 283–9. PubMed Abstract | Publisher Full Text\n\nAranki SF, Shaw DP, Adams DH, et al.: Predictors of atrial fibrillation after coronary artery surgery. Current trends and impact on hospital resources. Circulation. 1996; 94(3): 390–7. PubMed Abstract | Publisher Full Text\n\nAlmassi GH, Schowalter T, Nicolosi AC, et al.: Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg. 1997; 226(4): 501–11; discussion 511–3. PubMed Abstract | Free Full Text\n\nFuster V, Rydén LE, Asinger RW, et al.: ACC/AHA/ESC Guidelines for the Management of Patients With Atrial Fibrillation: Executive Summary A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines and Policy Conferences (Committee to Develop Guidelines for the Management of Patients With Atrial Fibrillation) Developed in Collaboration With the North American Society of Pacing and Electrophysiology. Circulation. 2001; 104(17): 2118–50. PubMed Abstract\n\nEuropean Heart Rhythm Association; European Association for Cardio-Thoracic Surgery, Camm AJ, et al.: Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010; 31(19): 2369–429. PubMed Abstract | Publisher Full Text\n\nBiernacka A, Frangogiannis NG: Aging and Cardiac Fibrosis. Aging Dis. 2011; 2(2): 158–173. PubMed Abstract | Free Full Text\n\nGlusenkamp NY, Risch SA, Kerkar P, et al.: Atrial Fibrillation in India: Insights from the PINNACLE Indian Outpatient Registry. Circulation. 2014; 7: A161.\n\nNg RR, Tan GH, Liu W, et al.: The Association of Acute Kidney Injury and Atrial Fibrillation after Cardiac Surgery in an Asian Prospective Cohort Study. Medicine (Baltimore). 2016; 95(12): e3005. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBouri S, Shun-Shin MJ, Cole GD, et al.: Meta-analysis of secure randomised controlled trials of β-blockade to prevent perioperative death in non-cardiac surgery. Heart. 2014; 100(6): 456–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCağli K, Göl MK, Keles T, et al.: Risk factors associated with development of atrial fibrillation early after coronary artery bypass grafting. Am J Cardiol. 2000; 85(10): 1259–61. PubMed Abstract | Publisher Full Text\n\nMueller XM, Tevaearai HT, Ruchat P, et al.: Atrial fibrillation and minimally invasive coronary artery bypass grafting: risk factor analysis. World J Surg. 2002; 26(6): 639–42. PubMed Abstract | Publisher Full Text\n\nTadic M, Ivanovic B, Zivkovic N: Predictors of atrial fibrillation following coronary artery bypass surgery. Med Sci Monit. 2011; 17(1): CR48–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaPar DJ, Speir AM, Crosby IK, et al.: Postoperative atrial fibrillation significantly increases mortality, hospital readmission, and hospital costs. Ann Thorac Surg. 2014; 98(2): 527–33. discussion 533. PubMed Abstract | Publisher Full Text\n\nRostagno C, La Meir M, Gelsomino S, et al.: Atrial fibrillation after cardiac surgery: incidence, risk factors, and economic burden. J Cardiothorac Vasc Anesth. 2010; 24(6): 952–58. PubMed Abstract | Publisher Full Text\n\nCalò L, Bianconi L, Colivicchi F, et al.: N-3 Fatty acids for the prevention of atrial fibrillation after coronary artery bypass surgery: a randomized, controlled trial. J Am Coll Cardiol. 2005; 45(10): 1723–1728. PubMed Abstract | Publisher Full Text\n\nLeaf A: The electrophysiological basis for the antiarrhythmic actions of polyunsaturated fatty acids. Eur Heart J. 2001; 3(Suppl D): D98–105. Publisher Full Text\n\nKang JX, Leaf A: Antiarrhythmic effects of polyunsaturated fatty acids. Recent studies. Circulation. 1996; 94(7): 1774–1780. PubMed Abstract | Publisher Full Text\n\nLiu T, Korantzopoilos P, Li G: Antioxidant therapies for the management of atrial fibrillation. Cardiovasc Diagn Ther. 2012; 2(4): 298–307. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarling L, Rasoli S, Vecht JA, et al.: Do antioxidant vitamins have an anti-arrhythmic effect following cardiac surgery? A meta-analysis of randomised controlled trials. Heart. 2011; 97(20): 1636–1642. PubMed Abstract | Publisher Full Text\n\nRasoli S, Kakouros N, Harling L, et al.: Antioxidant vitamins in the prevention of atrial fibrillation: what is the evidence? Cardiol Res Pract. 2011; 2011: 164078. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarouk Musa A, Quan CZ, Xin LZ, et al.: Dataset 1 in: A retrospective study on atrial fibrillation after coronary artery bypass grafting surgery at The National Heart Institute, Kuala Lumpur. F1000Research. 2018. Data Source"
}
|
[
{
"id": "34013",
"date": "29 May 2018",
"name": "Johannes Steindl",
"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\nPostoperative atrial fibrillation after (POAF) is a common complication after coronary artery bypass grafting (CABG) and cardiac surgery in general.\nIn this article Musa and colleagues perform a retrospective, single-center analysis of 637 patients undergoing CABG in order to identify the incidence of POAF, its associated morbidity and mortality as well as its pre- and intraoperative risk factors. The authors compare their results to previous studies and also point out their regional relevance by comparison between Malaysia’s major ethnic groups.\nFrom a technical point of view, the paper is well written. The Introduction section includes a very detailed literature review summarizing available data about incidence, mortality and risk factors of POAF after CABG. This part of the article indicates the complexity of POAF with a high amount of variables possibly influencing its aetiology and pathogenesis. The Methods section describes the study design and sample size calculation, the inclusion and exclusion criteria are plausible and the statistical methods are well chosen. In the Results and Discussion section, a detailed analysis of the collected data is performed, comparing their own results to previous studies and literature. Finally, the Conclusion section includes a short recap and a future perspective for further studies.\n\nIn order to give a helpful feedback to the authors I want to divide my review in 2 sections: “technical aspects” and “intended relevance of the paper”.\n\nTechnical aspects:\n\nIntroduction:\nI want to point out that the Introduction offers an excellent insight into POAF literature and the current state of research on this topic.\n\nMethods:\nPage 4: The sample size was calculated in order to “investigate the outcome”. Please give a more detailed definition for the “outcome” you wanted to analyze. Which parameters did you look for in the first place (i.e. incidence, mortality, etc.)? Was the sample size large enough for all your endpoints? The inclusion/exclusion criteria as well as the statistical methods were chosen appropriately.\n\nResults:\nTo provide a clear design of the tables I would recommend to include the total number of patients (637), the number of Non-POAF patients (454) and POAF patients (183) into the tables, since some of the percentage numbers are a little bit confusing at first sight. So, for example, you may mention those numbers in the table’s head line or within each field (i.e. Table 1: Male gender 521/637 (81.8).\n\nPage 5, Table 2: “Time from surgery to POAF” should be corrected to “hours” instead of “minutes”\n\nPage 7, Operative details: The section about on-pump vs. off-pump CABG could be written a bit more understandable (i.e. the percentage of POAF in the on-pump group vs. the percentage of POAF in the off-pump group).\n\nPage 8, Postoperative stay: If available, the cause of prolonged stay on ICU and HDU (i.e. hemodynamic instability, rhythm control, non cardiac reasons) would be very interesting to know, in order to determine if there might be a causal relationship between POAF and the prolonged ICU stay or not.\n\nData about postoperative cardiac pacing and its influence on POAF would be very interesting. What is your protocol of postoperative atrial/ventricular pacing?\n\nDid you collect any data about the completeness of revascularization and its influence on POAF?\n\nDiscussion:\nPage 11: Please explain why you selected the 48-hour timeframe to identify patients for prophylactic anti-coagulation. What was the number of patients with a POAF episode of > 24 hours? What is your protocol for recurring POAF? Do you start anticoagulation in those patients? (and maybe for the discussion section: what do you think is a good follow up protocol for those patients?)\n\nConclusion:\n\nWhat are your thoughts about the importance of the difference of POAF incidence between the different ethical groups? Do you think there is any clinical relevance? Is it an interesting fact for research about genetic influences on POAF?\n\nPlease explain more detailed what surgical methods you think about, that could be improved? Bicaval cannulation (as you mentioned) is hardly done in isolated CABG. The reduction of cross clamp time is probably the goal of every surgical procedure. So in my opinion the current possibilities for surgical improvement are rather limited.\n\nWould it be possible to estimate the additional costs caused by POAF per year in your national health care system?\n\nRelevance of the paper: I think it is important to determine the target audience for this paper as well the article’s intended impact. If the author’s main intention is to present an analysis of POAF by using a local patient cohort, the paper is solidly written. However, despite the analysis of differences between Malayan ethnical groups, the novelty and scientific relevance of the information presented is limited.\nAlthough my limited knowledge about the Malayan health care system and its medical research institutions, in my opinion this article has huge potential to emphasize a discussion about improvement of the national research and data collection system. The authors mention on page 4 that there is currently no data available about the total number of CABG performed in Malaysia per year. Such information could be collected through manageable financial efforts by installing a national database for cardiac surgery, comparable to already existing databases in Europe or the United States. Such a database would profoundly facilitate further scientific research as well as quality control. Furthermore the author’s mention that prospective studies have to be performed in future. One of the main problems might be that the costs for that kind of studies often are enormous and that there might be a lack of industrial sponsors. If the authors were able to determine an estimation of the real-world additional costs of POAF for the the national health care budget, this might facilitate national sponsorship for research targeting to reduce these costs.\n\nHowever, in summary I support the indexing of the paper in regard of the revisions mentioned above.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3787",
"date": "01 Aug 2018",
"name": "Ahmad Farouk Musa",
"role": "Author Response",
"response": "Postoperative atrial fibrillation after (POAF) is a common complication after coronary artery bypass grafting (CABG) and cardiac surgery in general.In this article Musa and colleagues perform a retrospective, single-center analysis of 637 patients undergoing CABG in order to identify the incidence of POAF, its associated morbidity and mortality as well as its pre- and intraoperative risk factors. The authors compare their results to previous studies and also point out their regional relevance by comparison between Malaysia’s major ethnic groups.From a technical point of view, the paper is well written. The Introduction section includes a very detailed literature review summarizing available data about incidence, mortality and risk factors of POAF after CABG. This part of the article indicates the complexity of POAF with a high amount of variables possibly influencing its aetiology and pathogenesis. The Methods section describes the study design and sample size calculation, the inclusion and exclusion criteria are plausible and the statistical methods are well chosen. In the Results and Discussion section, a detailed analysis of the collected data is performed, comparing their own results to previous studies and literature. Finally, the Conclusion section includes a short recap and a future perspective for further studies. In order to give a helpful feedback to the authors I want to divide my review in 2 sections: “technical aspects” and “intended relevance of the paper”. Technical aspects: Introduction: I want to point out that the Introduction offers an excellent insight into POAF literature and the current state of research on this topic. Response: Thank you Professor. Methods: Page 4: The sample size was calculated in order to “investigate the outcome”. Please give a more detailed definition for the “outcome” you wanted to analyze. Which parameters did you look for in the first place (i.e. incidence, mortality, etc.)? Was the sample size large enough for all your endpoints? Response: We have revised the sample size calculation as follows: Primary endpoint is the incidence of POAF: Using the Raosoft® sample size calculator software28, the calculated sample size was 377 (margin of error: 5%; confidence interval: 95%; estimated population size: 20,000; and response distribution: 50%). This is the minimum recommended sample size for the study. However, using the formula for sample size calculation based on the incidence of POAF of 30% 9,10,12,13, the estimated sample size (n) = Z2P(1-P)/d2, where Z= Z statistic for a level of confidence, P=expected prevalence or proportion, and d=precision (Ref: Naing et al 2006). Therefore, n = [(1.96)2*0.3*0.7]/(0.05)2 = 323. Taking into account a possible 20% attrition rate, the minimum sample size required to determine the incidence of POAF = 323 + (0.2*323) = 388.Secondary endpoints are the mortality and morbidity rates, duration of ICU, HDU, and hospital stay. Sample size calculations for these endpoints are as follows (Omer et al 2016): Mortality rate (1.9%): n= Z2P(1-P)/d2 = [(1.96)2*0.019*0.981]/(0.05)2 = 28.6 approx. = 29 Length of Hospital Stay (LOHS): Using the P&S Software (Dupont & Plummer 1990), we calculate the sample size for a reported LOHS of mean 12.7+6.6 days for patients with POAF and mean of 10.3+8.9 days for those without (Omer et al 2016), as follows: α=0.05, 1-β=0.80, δ=2.4, σ=0.641, m=1. Therefore, n=4 Based on the above sample size calculations, we will adopt the highest sample size calculated, i.e., n=388 as the minimum sample size required for this study as this sample size is adequate for both primary and secondary endpoints. References:Dupont WD, Plummer WD: \"Power and Sample Size Calculations: A Review and Computer Program\", Controlled Clinical Trials 1990; 11:116-28.Naing L, Winn T, Rusli BN: Practical issues in calculating the sample size for prevalence studies. Archives of Orofacial Sciences 2006; (1):9-14.Omer S, Cornwell LD, Bakshi A et al: Incidence, predictors, and impact of postoperative atrial fibrillation after coronary artery bypass grafting in military veterans 2016; 43(5):397-403 The inclusion/exclusion criteria as well as the statistical methods were chosen appropriately. Response: Thank you. Results: To provide a clear design of the tables I would recommend to include the total number of patients (637), the number of Non-POAF patients (454) and POAF patients (183) into the tables, since some of the percentage numbers are a little bit confusing at first sight. So, for example, you may mention those numbers in the table’s head line or within each field (i.e. Table 1: Male gender 521/637 (81.8). Response: Thank you for pointing this out. I have made the necessary corrections. Page 5, Table 2: “Time from surgery to POAF” should be corrected to “hours” instead of “minutes” Response: Thank you for pointing out the mistake. I have changed it. Page 7, Operative details: The section about on-pump vs. off-pump CABG could be written a bit more understandable (i.e. the percentage of POAF in the on-pump group vs. the percentage of POAF in the off-pump group). Response: Thank you. I have done the necessary correction to make it more clearer. We noticed that 29.3% of on-pump patients developed AF post-operatively as compared to 18.2% in the off-pimp cases. However the difference is not statistically significant. Page 8, Postoperative stay: If available, the cause of prolonged stay on ICU and HDU (i.e. hemodynamic instability, rhythm control, non-cardiac reasons) would be very interesting to know, in order to determine if there might be a causal relationship between POAF and the prolonged ICU stay or not. Response: Since this is a retrospective review, we have not really looked into the related factors apart from POAF and the treatment to bring the patient back to sinus rhythm. However we would be interested to look into the causal relationship between POAF and the prolonged ICU stay in these subsets of patients. Data about postoperative cardiac pacing and its influence on POAF would be very interesting. What is your protocol of postoperative atrial/ventricular pacing? D in a future study to look in Response: We routinely placed prophylactic right ventricle temporary epicardial pacing wire. We don’t normally place atrial pacing wire. And our patients are not routinely paced post-operatively. Did you collect any data about the completeness of revascularization and its influence on POAF? Response: Unfortunately we do not have any data nor conduct any study intra- or post-operatively to determine the completeness of revascularization. We do not use Flowmeter during surgery to measure the blood flow in the graft though this idea is exciting. Discussion: Page 11: Please explain why you selected the 48-hour timeframe to identify patients for prophylactic anti-coagulation. What was the number of patients with a POAF episode of > 24 hours? What is your protocol for recurring POAF? Do you start anticoagulation in those patients? (and maybe for the discussion section: what do you think is a good follow up protocol for those patients?) Response: We noticed that there is a bleeding tendency in our population as compared to Western patients. And we hardly counter any thromboembolic patients in the first 48 hours. So the benefit of anticoagulation doesn’t outweigh the risk of bleeding in the first 48-hours.And if the POAF recurs after the first 48hours, patients will be started on LMWH overlapping oral Warfarin. Patients will be given Warfarin for the first 6-weeks with a target INR of 2.0-3.0. The patients will be followed up at six-weeks and if found to be in sinus rhythm, the Warfarin will be taken off.Conclusion: What are your thoughts about the importance of the difference of POAF incidence between the different ethical groups? Do you think there is any clinical relevance? Is it an interesting fact for research about genetic influences on POAF? Response: Our study showed a significantly lower incidence of POAF in the Indian community., a finding that was found in a similar study in Singapore. In fact, Indians in general are found to have a lower incidence of AF. Epidemiological studies conducted across multiple nations showed this trend previously. This shows that Indians are more resistant to develop AF in general, not just post-operative AF. Hence we think that genetic factor that confers immunity relatively to the Indian community could not be ruled out.It will be very interesting to investigate if the pre-op preventive strategies could significantly reduce the development of POAF in the Indian community. And if they do, will the benefit outweigh the risk. The number needed to treat will definitely be higher in the Indian community. So the cost-benefit analysis may not be generalised to this group of people. Please explain more detailed what surgical methods you think about, that could be improved? Bicaval cannulation (as you mentioned) is hardly done in isolated CABG. The reduction of cross clamp time is probably the goal of every surgical procedure. So in my opinion the current possibilities for surgical improvement are rather limited. Response: Yes Professor, the possibilities of improvement is rather limited. But we feel that despite the advancement in cardiothoracic surgery, perhaps using retrograde cardioplegia together with the routine antegrade has the potential to minimize the occurrence of POAF. Studies have shown that giving both antegrade and retrograde cardioplegia provides better myocardial protection that either technique alone – in our centre, the antegrade – and ensures good cardioplegic distribution to the left and right ventricle, and allows regional delivery of cardioplegic flow to region supplied by the occluded arteries. Though we have to admit that this was just our postulation based on our anecdotal experience in Melbourne, Australia. We have yet to come across any studies that suggest such a correlation. Would it be possible to estimate the additional costs caused by POAF per year in your national health care system? Response: Yes Professor. I believe this is possible but would definitely require a thorough study. We have done a study many years ago in looking at the cost-effectiveness of performing endoscopic vein harvesting to that of leg wound management and dressing in open saphenous vein harvesting. So looking at the additional cost in managing POAF patients is not impossible but would require time, effort, and full cooperation by the Institution. Relevance of the paper:I think it is important to determine the target audience for this paper as well the article’s intended impact. If the author’s main intention is to present an analysis of POAF by using a local patient cohort, the paper is solidly written. However, despite the analysis of differences between Malayan ethnical groups, the novelty and scientific relevance of the information presented is limited.Although my limited knowledge about the Malayan health care system and its medical research institutions, in my opinion this article has huge potential to emphasize a discussion about improvement of the national research and data collection system. The authors mention on page 4 that there is currently no data available about the total number of CABG performed in Malaysia per year. Such information could be collected through manageable financial efforts by installing a national database for cardiac surgery, comparable to already existing databases in Europe or the United States. Such a database would profoundly facilitate further scientific research as well as quality control. Furthermore the author’s mention that prospective studies have to be performed in future. One of the main problems might be that the costs for that kind of studies often are enormous and that there might be a lack of industrial sponsors. If the authors were able to determine an estimation of the real-world additional costs of POAF for the the national health care budget, this might facilitate national sponsorship for research targeting to reduce these costs. However, in summary I support the indexing of the paper in regard of the revisions mentioned above. Response: Thanks so much Professor for your advice and support. We definitely will take this matter up to the higher authorities."
}
]
},
{
"id": "33419",
"date": "30 May 2018",
"name": "Vlasta Bari",
"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\nAuthors propose a retrospective study about the incidence of post operative atrial fibrillation (POAF) in patients undergoing coronary artery bypass surgery in Malaysia.\nTo do so, 637 medical records have been considered, including several confounding factors that could be determinant of the outcome.\nResults show that the incidence of POAF in Malaysis is similar to the European one.\nThe paper is well written and factors are correctly described.\nMethods and statistics are appropriate.\nI only have some concerns:\nThe paper is quite long and it could be considered to try to revise some parts. I would like to suggest the authors try to revise some paragraphs, discarding the less necessary and focusing on the more important information, in order to improve readability\n\nTo strengthen the association with predictive factors of POAF it would be interesting to see also ROC curve showing the value of the area under the curve. That could result from the multivariate logistic regression analysis determined by including all factors resulting predictive for POAF.\n\nSome possible predictive factors, as the state of autonomic nervous system, as considered in the suggested works1,2, have not been taken into account and that should be added at least as limitation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3786",
"date": "01 Aug 2018",
"name": "Ahmad Farouk Musa",
"role": "Author Response",
"response": "Authors propose a retrospective study about the incidence of post operative atrial fibrillation (POAF) in patients undergoing coronary artery bypass surgery in Malaysia.To do so, 637 medical records have been considered, including several confounding factors that could be determinant of the outcome.Results show that the incidence of POAF in Malaysia is similar to the European one.The paper is well written and factors are correctly described.Methods and statistics are appropriate.I only have some concerns: The paper is quite long and it could be considered to try to revise some parts. I would like to suggest the authors try to revise some paragraphs, discarding the less necessary and focusing on the more important information, in order to improve readability. Response: Thank you for your constructive comments. I have revised the manuscript and made it more readable. To strengthen the association with predictive factors of POAF it would be interesting to see also ROC curve showing the value of the area under the curve. That could result from the multivariate logistic regression analysis determined by including all factors resulting predictive for POAF. Response: We employed three methods to assess the goodness-of-fit of the regression model to the data: Hosmer-Lemeshow test, Classification Table, and the Area Under the RIC Curve (AUC). Model Fit AssessmentThe Hosmer-Lemeshow test (below) indicated that the model is a good fit to the data (p=0.99) since the p value exceeded 0.05.Hosmer and Lemeshow TestStep Chi-square df Sig.1 1.664 8 .990Further, the contingency table for the Hosmer and Lemeshow test (below) also indicated that the expected/predicted counts of the model clearly shows small differences between the observed and expected counts, thereby confirming the goodness of model fit.Contingency Table for Hosmer and Lemeshow Test POAF = 0 POAF = 1 Observed Expected Observed Expected TotalStep 1 1 56 56.149 7 6.851 63 2 54 53.080 9 9.920 62 3 53 50.677 10 12.323 63 4 49 48.237 14 14.763 63 5 44 46.367 19 16.633 63 6 44 44.852 19 18.148 63 7 41 42.710 22 20.290 63 8 42 40.215 21 22.785 63 9 37 37.236 26 25.764 63 10 30 30.476 35 34.524 65Using the Classification Table (below), the table shows that the model correctly predicts POAF by more than 70% (71.4%), thereby indicating a good fit of the regression model to the data. Classification Tablea Predicted POAF Percentage Observed 0 1 CorrectStep 1 POAF 0 433 17 96.2 1 164 18 9.9 Overall percentage 71.4---------------------------------------------- a. The cut value is .500Finally, the Area Under the ROC Curve (AUC) (below) indicated that the AUC is 0.663, which is slightly below the cut-off level for goodness-of-fit (70%) with a 95% CI of between 0.617 to 0.709. Examining the lower level of the 95%CI reveals that it is greater than the cut-off level of 0.5, thereby confirming the good fit of the regression model to the data. Area Under the CurveTest Result Variable(s): Predicted probability Asymptotic 95% Confidence Interval Area Std. Errora Asymptotic Sig.b Lower Bound Upper Bound.663 .023 .000 .617 .709The test result variable(s): Predicted probability has at least one tie between the positive actual state group and the negative actual state group. Statistics may be biased.a. Under the nonparametric assumptionb. Null hypothesis: true area = 0.5The AUC is graphically illustrated in the figure here. Some possible predictive factors, as the state of autonomic nervous system, as considered in the suggested works1,2, have not been taken into account and that should be added at least as limitation. Response: Thanks for pointing this our Professor. Yes, we have omitted the autonomic nervous system as independent predictors of AF post-CABG since this is a retrospective review and we did not monitor such changes in our patients. This will be noted in the Limitation of the study."
}
]
}
] | 1
|
https://f1000research.com/articles/7-164
|
https://f1000research.com/articles/7-215/v1
|
22 Feb 18
|
{
"type": "Study Protocol",
"title": "A study protocol for a randomized controlled trial on the prevention of atrial fibrillation after coronary artery bypass grafting surgery using Tocotrienol, an isomer of Vitamin E derived from palm oil",
"authors": [
"Ahmad Farouk Musa",
"Jeswant Dillon",
"Mohamed Ezani Md Taib",
"Alwi Mohamed Yunus",
"Rusli Bin Nordin",
"Yuen Kah Hay",
"Jeswant Dillon",
"Mohamed Ezani Md Taib",
"Alwi Mohamed Yunus",
"Rusli Bin Nordin",
"Yuen Kah Hay"
],
"abstract": "Background: One of the most common complications following coronary artery bypass grafting (CABG) surgery is atrial fibrillation (AF), which contributes towards increasing morbidity and mortality, length of hospital stay (LoHS) and reduced quality of life (QoL) of patients. Objectives: To determine whether the intake of Tocotrienol, a Vitamin E isomer derived from palm oil, before and immediately following CABG prevents AF, reduces LoHS, and improves the QoL of patients. Protocol: The study is registered with the National Medical Research Register with a trial number NMRR-17-1994-34963 and designed as a prospective, randomized controlled trial (RCT) with parallel groups. The experimental group will receive two 200mg Tocotrienol capsules each day, while the control group will receive two identical placebo (palm Super Olein) capsules per day. ECG readings will be used to detect AF post operatively, LoHS will be measured by checking the records from the National Heart Institute Hospital register, and the health-related Quality of Life (HRQoL) analysis (the Malay version of the Short Form 36 Questionnaire) will be used to analyse QoL. The sample size was calculated to be 140 in each arm of the RCT for a power of 0.8 and a significance level of 0.05. Funding: HOVID Berhad funds this research project. Expected outcomes: The primary endpoint is the development of postoperative AF, whilst the secondary endpoints are the LoHS and HRQoL of patients post CABG. Future implications: Prevention of AF and its complications such as cardiovascular or cerebrovascular events, especially stroke, is an important output. Malaysia is one of the biggest producers and exporters of palm oil and palm oil products. Thus, the possibility of marketing Tocotrienol, in reducing AF post CABG surgery, is a very important proposition indeed. Trial number: NMRR-17-1994-34963",
"keywords": [
"Coronary Artery Bypass",
"Atrial Fibrillation"
],
"content": "Introduction\n\nAtrial fibrillation (AF) after coronary artery bypass grafting (CABG) surgery is common in clinical practice, with an incidence of up to 70%. It has the potential to double the risk of mortality and is associated with a 6-fold increase in the risk of stroke1.\n\nIn a recent local study by Musa et al. on post-CABG patients, it was demonstrated that patients who developed postoperative AF had a prolonged Intensive Care Unit (ICU) stay and High Dependency Unit (HDU) stay, and a prolonged total hospital stay. This leads to an increase in resource utilization. With an increased rate of morbidity and mortality, preoperative prophylactic strategies are necessary to reduce the incidence of AF post-operatively and improve the Quality of Life (QoL) of patients2.\n\nThe financial burden in managing AF is huge, and many new innovative ideas and materials have been tested in order to reduce the occurrence of post-operative AF. One of the very promising but not widely studied treatment alternatives is the use of Tocotrienol, an isomer of vitamin E.\n\nAF is thought to be initiated by rapid electrical activity arising from the muscular sleeves of pulmonary veins1,3. Multiple re-entrant wavelets, shorter refractory periods, conduction delays and fragmented electrograms maintain the arrhythmia1. If this arrhythmia continues, electrical remodeling occurs, which will further facilitate AF maintenance. The pathogenesis of post-operative AF is believed to be multifactorial. Previously it was thought to be caused by inflammatory pathways, but now it is known that oxidative stress plays a significant role4. Reactive oxygen species (ROS) are the cause of oxidative stress, and generate numerous pathological processes such as DNA damage, apoptosis, and cellular hypertrophy, as well as signal pathway modulation5.\n\nCardiac surgery is characterized by ischaemia and reperfusion injury, which leads to the release of ROS. This causes oxidative stress and initiates a systemic inflammatory response6. The extracorporeal circulation during on-pump surgery also activates inflammatory cells, increasing the oxidative stress7,8. Therefore it was always thought that an off-pump surgery would reduce the incidence of AF.\n\nOxidative stress causes the depletion of endogenous antioxidants, and this results in oxidative damage. This oxidative stress is mainly the result of on-pump surgery that requires aortic cross-clamping and the bypass machine9.\n\nAlmost all cardiac centers have guidelines on the pharmacological management of postoperative AF10. But the challenge has always been how to prevent AF from happening. Efforts have been made to develop alternative preventive strategies. With the recent understanding of the pathogenesis of AF via the oxidative pathway, perhaps antioxidant vitamins would be a promising treatment.\n\nPrevious data have shown that both Vitamin E and C are known to have antioxidant properties, and have the ability to counter the action of free radicals11. Vitamin E is able to maintain membrane stability. It could also prevent the process of lipid peroxidation. Vitamin C removes water-soluble free radicals and acts synergistically with Vitamin E.\n\nThere have been numerous mechanistic studies on the role of antioxidant vitamins in the prevention of postoperative AF. However, the therapeutic potential is uncertain and there is no universally accepted protocol. It is interesting to note that Carnes et al.12 reported a 16.3% incidence of postoperative AF in Vitamin C-treated patients compared to 34.9% in controls in their prospective study. In their meta-analysis, Harling et al.13 demonstrated a reduction in the incidence of postoperative AF and other forms of arrhythmia following antioxidant vitamin therapy, which remained significant when only randomized controlled trials (RCTs) were analyzed. They also showed that this reduction in incidence was not related to the bypass or cross-clamp time, where there was no significant difference between the two-groups. Lastly, the results of the meta-analysis comparing the groups of anti-oxidants and control also demonstrated a reduced stay at the ICU and at the hospital. While this finding cannot be attributed solely to a reduction in postoperative AF, it is possible that this reduction in both ICU and hospital stay was a result of the reduced incidence of AF in the treatment group.\n\nRasoli et al.14 came to a similar conclusion that the use of antioxidant vitamins plays a role in the reduction of post-operative AF. However, they argued that the effect was somewhat variable, because both vitamins, C and E, were weak anti-oxidants. They were limited in their capacity to cross the cell membrane in order to counter the superoxides.\n\nWe are proposing a more potent isomer of vitamin E, namely Tocotrienol, which is derived from palm oil, instead of the usual isomer of vitamin E, namely tocopherol, which was being used in all previous research. Tocotrienol has been proven to possess more potent antioxidant activity than tocopherols15. Laboratory studies have proven that Tocotrienol has the ability to scavenge peroxyl radicals in liposomes16. Tocotrienol is seen to be more evenly distributed in the phospholipid bilayer of the plasma membrane. It is also shown to have more efficient collision with radicals. These are some of the reasons why Tocotrienol is thought to have a more potent antioxidant properties15.\n\nAll these recent discoveries suggest the promising antioxidant properties for Tocotrienol. We hypothesize that it would be able to reduce the incidence of postoperative AF and subsequently reduce the complications associated with postoperative AF, along with the economic burden of prolonged intensive care stay and prolonged hospital stay. Overall, we also hypothesized that Tocotrienol will increase the health related quality of life (HRQoL) of patients. The potential value of marketing a national product that can be produced locally from palm oil and its potential uses, not only for patients, who are undergoing CABG, but also for patients with standalone AF, makes this an even more attractive product.\n\n\nProtocol\n\nThe National Heart Institute Ethics Committee has approved the study protocol (IJNREC/20112017). The ethics committee will also serve as the data safety committee. Similar approval has been obtained from MUHREC - Monash University Human Research Ethics Committee – (9277) and from MREC – the Malaysian Research Ethics Committee. The study is registered with the National Medical Research Registry (NMRR), Ministry of Health, Malaysia, (NMRR-17-1994-34963) and will be carried out in accordance with the ethical principles outlined in the Malaysian Good Clinical Practice Guidelines.\n\nWritten informed consent for participation in the trial will be obtained from each participant on recruitment (Supplementary File 1).\n\nThe clinical data of patients enrolled in this study will be de-identified. No personal data will be disclosed to anyone outside of the National Heart Institute Ethics Committee.\n\nThis is Protocol Version 2.0 dated on 15 June 2017.\n\nTrial registration: NMRR-17-1994-34963\n\nTrial registry: National Medical Research Register, Malaysia\n\nTrial registration date: 13th March 2017\n\nStudy status: The study has not yet started at the time of submission of this manuscript since we are still waiting for the Clinical Trial Exemption (CTX) from the National Pharmaceutical Regulatory Agency (NPRA).\n\nTocotrienol is a commercial product produced by Hovid Berhad, and is marketed as Tocovid Suprabio, containing 61.52mg alpha-Tocotrienol, 112.80mg gamma-Tocotrienol, 25.68mg delta-Tocotrienol and 91.60IU alpha-tocopherol. Its availability in Malaysia has allowed this innovative and promising study to be conducted. We hypothesize that:\n\nIntake of Tocotrienol capsules before and immediately following CABG prevents AF post CABG;\n\nIntake of Tocotrienol capsules before and immediately following CABG improves the QoL of patients post CABG;\n\nIntake of Tocotrienol capsules before and immediately following CABG shortens the length of hospital stay (LoHS) of patients post CABG.\n\nTo determine whether the intake of Tocotrienol capsules before and immediately following CABG prevents AF post CABG.\n\nTo determine whether the intake of Tocotrienol capsules before and immediately following CABG improves the QoL of patients post CABG.\n\nTo determine whether the intake of Tocotrienol capsules before and immediately following CABG shortens the LoHS of patients post CABG.\n\nStudy site: National Heart Institute, Kuala Lumpur, Malaysia\n\nThe study is designed as a prospective, randomized controlled trial, with two parallel groups of the same demographics and co-morbidities. The main aim is to look at the effect of Tocotrienol in reducing AF post CABG.\n\nPatients who are scheduled for CABG under the co-researchers at the National Heart Institute will be approached for their consent to be enrolled in this study once they are admitted to the wards. We will assign the patients enrolled in the study to one of two groups:\n\n1. The control group: Routine CABG surgery procedure plus placebo containing palm Super Olein capsules (supplier: HOVID Berhad, Malaysia). The placebo is an identical-looking capsule that contains tocotrienols-stripped palm oil (termed palm Super Olein), which is also the non-active excipient in the treatment capsule. The choice of control is to mimic the treatment capsule as much as possible, without the active component of tocotrienols.\n\n2. The study group: Routine CABG surgery procedure plus Tocotrienol capsules (supplier: HOVID Berhad, Malaysia)\n\nFor patients in the study group, Tocotrienol will be administered as two capsules of 200mg per day in two divided doses, and administration will start immediately after randomization and will continue until the first follow-up, which is normally six weeks after discharge.\n\nSince there is no available data for preferred dose of treatment, we decided on two divided doses of Tocotrienol of one capsule twice daily of 200mg Tocovid. This dosage was estimated based on the regime used by Olaf Stanger et al.17 at the Department of Cardiac Surgery, Paracelcus Medical University Salzburg, Austria, where three ampoules of 45IE Vitamin E were used, with 30mg per ampoule or 90mg in total. Because in our study we would be using an oral preparation instead of an IV preparation as in the Austrian study, and absorption of Tocotrienol has been shown to be low and incomplete via the oral route, we estimated a higher dosage18. Considering the bioavailability of Tocotrienol can be as low 10–30% if administered orally18; and taking into consideration that this is a pivotal study, it is reasonable to use the highest dose possible that is safe without any adverse effects. Many clinical studies with Tocotrienol have used 400mg daily in two divided doses and have been proven to be safe19,20.\n\nTocovid will be administered right after randomization, and will continue after the patient is discharged until the first follow-up, which is normally about six-weeks later. If the patient is on prolonged ventilation in the ICU, Tocotrienol will be administered through a nasogastric tube. The intensive care nurses and cardiothoracic ward nurses will monitor compliance.\n\nAll patients undergoing CABG surgery, either or with valve surgery, will be included. Similarly, this study will include participants undergoing both on-pump surgery using cold potassium cardioplegia, and off-pump beating heart surgery.\n\nAll patients will be admitted to the intensive care unit (ICU) after CABG surgery, with close monitoring on one-patient-one-nurse basis. They will then be transferred to a monitored high dependency unit (HDU) if their condition is stable, while some might be transferred straight back to the ward. Continuous rhythm monitoring will be performed using the 12-lead ECG for all patients in ICU and HDU. The monitoring will be continued for the first four to five postoperative days on the normal cardiothoracic wards using Holter monitors until the patients are discharged.\n\nWe will review the electrocardiographic (ECG) data on a daily basis. We will also review all the printouts of all abnormal rhythms, which will then be included in the clinical records. Additionally, an ECG will also be recorded when patients are symptomatic or when there is a suspicion of arrhythmia clinically. We will treat all AF episodes, in both the study arm and control arm, according to the protocols of the Cardiothoracic Department. The patients will be managed by the cardiothoracic surgeon and their team. The first-line drug used for treatment of new onset AF post-operatively by the Cardiothoracic team in the National Heart Institute (IJN) Hospital is Amiodarone, unless contraindicated, as a 300mg infusion over one hour to be followed by 900mg over the next 23 hours.\n\nPatients of both the study and control arms will come back for follow-up at the Cardiothoracic Clinic at the IJN usually six weeks after discharge. They will also be advised to report to the Outpatient Department at the IJN if they develop any symptoms. During follow-up, all patients will be assessed routinely with blood work and a 12-lead ECG.\n\nFor the study flowchart, see Figure 1.\n\nHovid Berhad follow strict GLP Guidelines in the manufacturing process of Tocotrienols and matching placebo capsules. Each 200mg capsule of Tocotrienol will contain 61.52mg alpha-Tocotrienol, 112.80mg gamma-Tocotrienol, 25.68mg delta-Tocotrienol and 91.60IU alpha-tocopherol. The placebo capsules will contain palm Super Olein oil.\n\nThe study is designed as a randomized controlled trial (RCT), where patients with coronary artery disease requiring CABG surgery at the IJN will be prospectively and randomly divided into two parallel groups by means of computer generated numbers using Excel 2016 (Microsof, Redmont, WA, USA) in a block of 10 experimental to 10 matching controls at enrollment. The experimental group will receive Tocovid capsules whilst the control group will receive identical placebo capsules containing palm Super Olein oil.\n\nThe unblinded pharmacist will generate the allocation sequence to assign the participants to the interventions. Surgeon investigators will enroll the participants. The participants will then be assign randomly to either the study arm or the control arm based on the sequence assigned to them.\n\nInclusion/exclusion criteria\n\nInclusion criteria:\n\n1. Males or females\n\n2. More than 18 years of age\n\n3. Elective, on-pump or off-pump CABG surgery of coronary artery revascularization, single or double procedure\n\nExclusion criteria:\n\n1. Less than 18 years of age\n\n2. Refusal to have surgery\n\n3. Urgent or emergency surgery\n\n4. Inability to give informed consent\n\n5. Documented allergy to palm oil\n\nPatients will be recruited from those who are scheduled electively for surgery under the Consultant Surgeons, JD, MEMT, and AMY, who are co-researchers in this project. All public and private patients are eligible to be enrolled in this study. Consent will be taken upon admission to the wards.\n\nPatients can choose to withdraw from the study at any time. Subjects may be withdrawn if the investigator deems that it is detrimental or risky for the subject to continue. The patients who withdraw will not be replaced.\n\nAs the primary study endpoint, we will look at the development of AF post-surgery. This will be documented with ECG. We take a 30 second duration as a cut-off point and define AF as when there is a loss of p-waves and irregular ventricular rate or a confirmed atrial flutter21.\n\nThe secondary endpoint would include the LoHS after surgery, which will be obtained from the IJN registry, and the HRQoL. Three measurements will be used to determine the LoHS:\n\n(1) Total number of days that the patients stay in the Intensive care unit (ICU)\n\n(2) Total number of days in the high-dependency unit (HDU); and\n\n(3) Total number of days in the hospital, overall21.\n\nWe will measure the HRQoL of patients using the validated Malay and English versions of Short-Form 36 Questionnaires (SF-36)22–24.\n\nSince the researchers are blinded to the trial, we will be evaluating all endpoints, primary and secondary, independently. We will also review the patients’ clinical records and all ECG tracings. In fact, all the trial participants, care-givers, outcome assessors and data analysts will be blinded to the study treatment allocation. An assigned pharmacist maintaining the randomisation list will be unblinded to the study treatment allocation. In the event of serious adverse events or emergency clinical treatment requiring knowledge of the study treatment, a request for unblinding shall be submitted to the pharmacist with reasonable justification for unblinding. The unblinding details and justification will be documented in the case record form. If unblinded, the participant will be withdrawn from the study in the event of unblinding, and post-treatment follow-up conducted as per protocol or as applicable. Where possible, the outcome assessor will remain blinded to all treatment allocation until end of study, even after participants are withdrawn from the study.\n\nPreoperative\n\nThe two randomized groups, based in the IJN, will be matched according to sex, age, NYHA criteria, ejection fraction, and diabetic status. Operative and peri-operative conditions will also be similar for both groups. All subjects will receive an identical prophylactic antibiotic regime consisting of Cefazolin 2gm at induction and 1gm 12 hourly for 48 hours. Gentamicin 2mg/kg will also be given at induction.\n\nIntra-operative\n\nOn-pump CABG\n\nA standard procedure of performing CABG will be done under general anaesthesia. Veins will be harvested from the legs and the left internal mammary will be taken down once the chest is opened. Titanium clips will be used to secure all side branches of saphenous veins and also the branches of the internal mammary artery. The patient will then be cannulated and placed on a heart-lung machine. Once the ascending aorta is cross-clamped, cold cardioplegic solution will be perfused in an ante-grade manner through the aortic root into the heart and the pericardial sac will be buried with ice sludge to create myocardial hypothermia.\n\nThe diseased coronary arteries will then be identified and arteriotomies will be performed beyond the level of blockages. The open ends of the saphenous veins and the internal mammary artery are sewn to the openings artertomies using Prolene 7/0 sutures. Once the distal anastomoses are constructed, proximal anastomoses will commence.\n\nOnce all the anastomoses are completed, the cross-clamp will be taken off and the heart-lung machine will then be gradually weaned off. Subsequently, the patient will be decannulated. Drainage catheters will be placed around the heart and temporary pacing will be sewn to the surface of the heart. Sternum will then be closed with steel wire and subcutaneous tissue and skin in the usual manner.\n\nOff-pump CABG\n\nIn this case, the bypass surgery will be done without the use of heart-lung machine. Surgery will be done on a beating heart. The procedure is similar to the on-pump surgery. Chest is opened and the left internal mammary artery is taken down. A stabilizer is placed on the heart to limit the motion of the heart. The anastomosis is done by sewing the open end of the internal mammary to the coronary artery, namely the left anterior descending artery. Chest is closed in the usual manner after placement of the drains and the temporary pacing wire.\n\nPost-operative\n\nBoth groups will continue taking either two capsules of Tocotrienol or two capsules of placebo (palm Super Olein) daily for the entire hospitalization period until follow-up at six weeks after discharge. The capsules will be taken on a bd – twice-daily dosing. If the patient is still ventilated in the cardiac ICU, the capsules will be broken and the content administered via a Ryle’s tube.\n\nAfter surgery, patients will be admitted to the ICU with close monitoring on one-patient-one-nurse basis and will be subsequently transferred to a monitored high dependency unit if their condition is stable. Continuous heart rhythm monitoring will be carried out using the 12-lead ECG. The monitoring will be continued for the first four to five postoperative days on the normal cardiothoracic wards using Holter monitors until the patients are discharged.\n\nThe electrocardiographic data will be stored for 24 h and reviewed on a daily basis by the cardiothoracic team involved in the research. The printouts of all abnormal rhythms will also be reviewed for any episodes of arrhythmia. All printouts will be included in the clinical records. Additionally, an ECG will also be recorded in case of symptoms or when arrhythmia is suspected on clinical grounds; AF episodes will always be treated under the direction of the attending cardiothoracic surgeon.\n\nAfter discharge, all patients will be asked to report to the outpatient department of our institution in case of any relevant symptom. Additionally, all patients will be followed up six weeks after discharge; this will include physical examinaton and a 12-lead ECG measurement.\n\nECG readings to detect AF post operatively\n\nDiagnostic criteria:\n\n1. Absent P waves\n\n2. Evidence of fibrillation (f) waves instead of P waves. These are irregular undulations of the base line in ECG.\n\n3. Irregular R-R intervals\n\nLength of hospital stay (LoHS)\n\nThis information will be obtained from the IJN registry of patients. Three measurements will be used to determine the LoHS:\n\n(1) Total Cardiac ICU length of stay;\n\n(2) Total HDU length of stay; and\n\n(3) Total hospital length of stay.\n\nHealth related quality of life (HRQoL)\n\nThe analysis of HRQoL will be performed using the validated Short-Form 36 Questionnaire (SF-36). SF-36 has demonstrated its efficiency in clinical studies and has been proven to be generally acceptable to patients, consistent and a valid measure of health outcome. We will be using the validated Malay version of SF-36, which is a thirty-six items questionnaire, which measures QoL across eight domains that encompass both physical and mental.\n\nThe questionnaire will be distributed to both groups of patients before CABG and at discharge and six weeks during clinic visit postoperatively. The questionnaires are going to be administered using the questionnaire-interview approach.\n\nWe will use the PS Power and Sample Size Calculation Software version 3.1.2 for sample size calculation. We plan to study a continuous response variable from independent control and experimental subjects with 1 control per experimental subject.\n\nCalculation of sample size in the present study (randomized controlled trial: RCT to rule out selection bias) requires precise specification of:\n\nThe primary hypothesis of the study (Tocotrienol consumption reduces the incidence of post-operative atrial fibrillation in subjects that had undergone CABG)\n\nThe method of analysis (using Relative Risk: RR).\n\nWe will also take into account the possibility of ‘loss to follow-up’ or attrition bias of subjects by analysing all subjects from the start to the completion of the study according to the groups that they were originally randomized. This is called an Intention-To-Treat (ITT) analysis25. To calculate the desired sample size based on the above consideration, we use the PS Power and Sample Size Calculation Software for sample size calculation26,27.\n\nIn the present RCT, estimated sample size for the primary endpoint (incidence of POAF) was computed on the basis of findings from Calo et al.4 In that study, the incidence of post-operative AF was 15.2% in the treatment group. Whereas in the placebo group, the incidence was 33.3% (RR=15.2/33.3 = 0.46).\n\nTo account for possible subject withdrawal/non-compliance (attrition bias), we will adopt the ITT analysis so that subjects who are ‘lost to follow-up’/non-compliant will be analyzed according to the groups that they were randomized at the beginning of the study. The formula suggested by Wittes25 is “1/1-c” where c is the proportion that is lost to follow-up/did not comply. When c= 12%, for a total enrollment of 260 patients. However, when c=20%, the estimated sample size is increased to 278 (rounded to 280).\n\nUsing the PS Power and Sample Size Calculator26,27 with α equivalent to 0.05 and power (1-β) is 0.8, the estimated sample size based on the two proportions above is 130 intervention subjects and 130 control subjects (Fisher’s exact test) if we assume the “lost to follow up” is 12%. However, if we assume that the ‘lost to follow up’ is as high as 20%, then the two proportions will be 140 (Intervention Group) and 140 (Control Group) with a total number of 280 patients.\n\nIBM SPSS Statistics version 25.0 will be used to analyze the results. The ITT analysis will be used to analyze all outcome. The RR of the two-binomial proportion analysis will be used to test the occurrence of postoperative AF in the two treatment groups. The Kaplan-Meier method will be used to look at the cumulative risk. Log-rank test will compare the survival curves of the two treatment groups. Continuous variables will use mean (±SD) while categorical variables will use frequencies (percentages). We will use unpaired Student-t test for continuous variables to look for differences between groups. And group differences will be examined by the chi-square or Fisher exact tests as appropriate for categorical variables. In case of an expected frequency of less than 5 in any cell in a 2×2 table, a Fischer Exact test will be applied. A p value of less than 0.05 will be considered as statistically significant.\n\nIn order to find the predictors of AF after surgery, a stepwise multiple logistic regression analysis will be done. To examine the mean (±SD) LoHS (number of days) differences between the two groups, we will use an unpaired Student-t test.\n\nTo examine the mean (+SD) QoL score differences between the two groups (pre-operative, six weeks, and three months), the one-way mixed-mode repeated measure ANOVA with post-hoc multiple comparison test (between and within subject) of the two groups will be performed.\n\nIn order to look for the factors influencing the change in the quality of life after CABG, we will perform the univariate simple logistic regression (SLogReg) initially, and examining the statistical significance of each independent variable such as the number of revascularization and duration of surgery on the outcome. Then, we will perform a multiple logistic regression (MLogReg) including variables with a level of significance less or equal to 0.25 in the univariate logistic regression and controlling for the effects of possible confounding variables (sex, age, NYHA Criteria, ejection fraction, and diabetic status). A p value of less than 0.05 will be taken as significant.\n\nThe Principal Investigator, AFM, will be responsible for ensuring participants’ safety on a daily basis and for reporting Serious Adverse Events and Unanticipated Problems to his IJN Review Board (IJNRB) as required. The study statistician prepares reports that list adverse events, serious adverse events, deaths, and disease-or treatment-specific events required for monitoring body review in order to ensure good clinical care and identify any emerging trends. The IJNRB will act in an advisory capacity to the IJN Ethics Committee to monitor participants’ safety, evaluate the progress of the study, to review procedures for maintaining the confidentiality of data, the quality of data collection, management, and analyses.\n\nData dissemination will be done via three ways. Firstly, via presentation at local and international conferences either in the form of posters and/or oral presentation. Secondly via writing for publication in scientific journals. Thirdly, via participation at international exhibitions and conventions on research and innovation.\n\nThe sponsor has the full discretion to decide on the termination of the study at any time. Patients will be informed if the study is terminated and follow-up visits will be arranged if needed.\n\n\nData Availability\n\nNo data are associated with this article.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nHOVID Berhad (contact information: 121, Jalan Tunku Abdul Rahman, 30010 Ipon, Perak, Malaysia) sponsored this study (grant number, MMRD-MS-1801, awarded to the principle investigator, Ahmad Farouk Musa).\n\nThe study design is solely developed by the main investigator. The sponsor, HOVID Berhad, supplies the materials for the research and the financial support.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: Informed consent form and participant information sheet.\n\nClick here to access the data.\n\n\nReferences\n\nNattel S, Allessie M, Haissaguerre M: Spotlight on atrial fibrillation-the ‘complete arrhythmia’. Cardiovasc Res. 2002; 54(2): 197–203. PubMed Abstract | Publisher Full Text\n\nMusa AF, Dillon J, Omar R: Atrial Fibrillation after Coronary Artery Bypass Grafting: A Restrospective Review of a Single Centre Experience. Heart Surg Forum. 2008; 2: 23.\n\nMarkides V, Schilling RJ: Atrial fibrillation: classification, pathophysiology, mechanisms and drug treatment. Heart. 2003; 89(8): 939–943. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBanach M, Kourliouros A, Reinhart KM, et al.: Postoperative atrial fibrillation - what do we really know? Curr Vasc Pharmacol. 2010; 8(4): 553–572. PubMed Abstract | Publisher Full Text\n\nHuang CX, Liu Y, Xia WF, et al.: Oxidative stress: a possible pathogenesis of atrial fibrillation. Med Hypotheses. 2009; 72(4): 466–467. PubMed Abstract | Publisher Full Text\n\nKim H, Kim KH: Role of nitric oxide in oxidative damage in isolated rabbit gastric cells exposed to hypoxia-reoxygenation. Dig Dis Sci. 1998; 43(5): 1042–1049. PubMed Abstract | Publisher Full Text\n\nFontaine D, Pradier O, Hacquebard M, et al.: Oxidative stress produced by circulating microparticles in on-pump but not in off-pump coronary surgery. Acta Cardiol. 2009; 64(6): 715–722. PubMed Abstract | Publisher Full Text\n\nMatata BM, Sosnowski AW, Galiñanes M: Off-pump bypass graft operation significantly reduces oxidative stress and inflammation. Ann Thorac Surg. 2000; 69(3): 785–791. PubMed Abstract | Publisher Full Text\n\nDe Vicchi E, Pala MG, Di Credico G, et al.: Relation between left ventricular function and oxidative stress in patients undergoing bypass surgery. Heart. 1998; 79(3): 242–247. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDunning J, Treasure T, Versteegh M, et al.: Guidelines on the prevention and management of de novo atrial fibrillation after cardiac and thoracic surgery. Eur J Cardiothorac Surg. 2006; 30(6): 852–872. PubMed Abstract | Publisher Full Text\n\nNiki E: Interaction of ascorbate and alpha-tocopherol. Ann N Y Acad Sci. 1987; 498: 186–199. PubMed Abstract | Publisher Full Text\n\nCarnes CA, Chung MK, Nakayama T, et al.: Ascorbate attenuates atrial pacing-induced peroxynitrite formation and electrical remodeling and decreases the incidence of postoperative atrial fibrillation. Circ Res. 2001; 89(6): E32–8. PubMed Abstract | Publisher Full Text\n\nHarling L, Rasoli S, Vecht JA, et al.: Do antioxidant vitamins have an anti-arrhythmic effect following cardiac surgery? A meta-analysis of randomised controlled trials. Heart. 2011; 97(20): 1636–1642. PubMed Abstract | Publisher Full Text\n\nRasoli S, Kakouros N, Harling L, et al.: Antioxidant vitamins in the prevention of atrial fibrillation: What is the evidence? Cardiol Res Prac. 2011; 2011: 164078. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPacker L, Weber S, Rimbach G: Molecular aspects of alpha-tocotrienol antioxidant action and cell signalling. J Nutr. 2001; 131(2): S369–S373. PubMed Abstract | Publisher Full Text\n\nSerbinova E, Kagan V, Had D, et al.: Free radical recycling and intramembrane mobility in the antioxidant properties of alpha-tocopherol and alpha-tocotrienol. Free Radic Biol Med. 1991; 10(5): 263–275. PubMed Abstract | Publisher Full Text\n\nStanger O, Aigner I, Schimetta W, et al.: Antioxidant Supplementation Attenuates Oxidative Stress In Patients Undergoing Coronary Artery Bypass Graft Surgery. Tohoku J Exp Med. 2014; 232(2): 145–154. PubMed Abstract | Publisher Full Text\n\nYap SP, Yuen KH, Lim AB: Influence of route of administration on the absorption and disposition of alpha-, gamma- and delta-tocotrienols in rats. J Pharm Pharmacol. 2003; 55(1): 53–58. PubMed Abstract | Publisher Full Text\n\nGopalan Y, Shuaib IL, Magosso E, et al.: Clinical investigation of the protective effects of palm vitamin E tocotrienols on brain white matter. Stroke. 2014; 45(5): 1422–1428. PubMed Abstract | Publisher Full Text\n\nMagosso E, Ansari MA, Gopalan Y, et al.: Tocotrienols for normalisation of hepatic echogenic response in nonalcoholic fatty liver: a randomised placebo-controlled clinical trial. Nutr J. 2013; 12(1): 166–174. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMozaffarian D, Marchioli R, Gardner T, et al.: The ω-3 Fatty Acids for Prevention of Post-Operative Atrial Fibrillation trial--rationale and design. Am Heart J. 2011; 162(1): 56–63.e3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJenkinson C: The SF-36 physical and mental health summary measures: an example of how to interpret scores. J Health Serv Res Policy. 1998; 3(2): 92–96. PubMed Abstract | Publisher Full Text\n\nSavelieva I, Camm AJ: Clinical relevance of silent atrial fibrillation: prevalence, prognosis, quality of life, and management. J Interv Card Electrophysiol. 2000; 4(2): 369–382. PubMed Abstract | Publisher Full Text\n\nSararaks S, Azman AB, Low LL, et al.: Validity and reliability of the SF-36: the Malaysian context. Med J Malaysia. 2005; 60(2): 163–179. PubMed Abstract\n\nWittes J: Sample size calculations for randomized controlled trials. Epidemiol Rev. 2002; 24(1): 39–56. PubMed Abstract | Publisher Full Text\n\nDupont WD, Plummer WD Jr: Power and sample size calculations: a review and computer program. Control Clinical Trials. 1990; 11(2): 116–28. PubMed Abstract | Publisher Full Text\n\nDupont WD, Plummer WD Jr: Power and sample size calculations for studies involving linear regression. Control Clinical Trials. 1998; 19(6): 589–601. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "31123",
"date": "02 Mar 2018",
"name": "Whady Armino Hueb",
"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\nThese reviewers suggest changing the title to: “Hypotheses, rationale, design, and methods for evaluation of a randomized controlled trial using Tocotrienol, an isomer of Vitamin E derived from palm oil, on the prevention of atrial fibrillation after coronary artery bypass grafting surgery”.\nWe appreciate the opportunity to review the manuscript “A study protocol for a randomized controlled trial on the prevention of atrial fibrillation after coronary artery bypass grafting surgery using Tocotrienol, an isomer of Vitamin E derived from palm oil”.\nIn this study, the authors aimed to test Tocotrienol, an isomer of Vitamin E derived from palm oil in the prevention of atrial fibrillation after myocardial revascularization surgery.\nThe study is well designed, well written and well detailed in definitions and methods.\nHowever, some serious methodological failures should be avoided in order not to irreversibly compromise the study.\nIt is known that myocardial revascularization surgery with the use of extracorporeal circulation adds non-physiological effects with direct damage to the myocardium.\nSuch effects include: cardiac arrhythmias, low cardiac output, SIRS, thrombocytopenia, among others. The systemic inflammatory process is usually treated with corticosteroids. Thus, authors cannot ignore the differences between on-pump or off-pump.\nAs suggestions, these reviewers recommend the following care in patient selection and data analysis.\nThe surgery should be only on-pump or just off-pump.\n\nPatients with valvular heart disease cannot be included in patient selection. Aortic stenosis is often accompanied by myocardial hypertrophy, and Mitral insufficiency accompanies an increase in the left atrium. Both conditions facilitate the development of atrial fibrillation.\n\nThe authors should include only patients with preserved left ventricular function. Ventricular dysfunction is one of the leading causes of AF.\n\nThe use of corticosteroids should be prespecified and, preferably, avoided because of its known anti-inflammatory actions.\n\nTiming from randomization and consequently beginning of Tocotrienol or placebo to surgery could be a relevant issue. A short time since medication start and surgery could compromise a possible significant effect of the drug on AF development.\n\nA Holter monitoring must be installed only in the first 72 hours. It is known that the most sensitive moment for the onset of AF is the immediate postoperative period. However, the use of the Holter monitoring for a longer period, even if uncomfortable, may be useful.\n\nIn statistical analysis, the stepwise multiple logistic regression analysis should include atrium data such as size and/or volume, and procedure-related variables, such as use of inotropes and other vasoactive drugs, cross-clamp and cardiopulmonary bypass time. All of these variables might be associated with the occurrence of AF after revascularization procedures and should be included in the final model. On the other hand, overfitting should be avoided.\n\nIn the multiple logistic regression model for the analysis of the factors influencing the change in quality of life after CABG, the authors should also include in the model angina data such as angina presence or frequency. In order to assess whether Tocotrienol is associated to better quality of life measures, it would be helpful if the authors control for angina relief after surgery.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": [
{
"c_id": "3564",
"date": "08 May 2018",
"name": "Ahmad Farouk Musa",
"role": "Author Response",
"response": "Thank you Prof Hueb and Prof Rezende. Your comments were highly beneficial for the improvement of this project. Since we have not started this project yet and waiting for CTX - Clinical Trial Exemption from NPRA - National Pharmaceutical Regulatory Agency, we would include all the suggestions in the Study Protocol which will be resubmitted for Ethical Clearance. And to make reading easier, I'll copy the comments and reply immediately below. The surgery should be only on-pump or just off-pump.Since the National Heart Institute is an on-pump center, only a small minority of less than 5% were off-pump cases. This was based on the previous retrospective research that we have done on atrial fibrillation after CABG. Since we have not started this study yet and waiting for the CTX –Clinical Trial Exemption approval, we will include this in the Exclusion Criteria. Refer: A retrospective study on atrial fibrillation after coronary artery bypass grafting surgery at the National Heart Institute, Kuala Lumpur. F1000Research 2018; 7:164. Patients with valvular heart disease cannot be included in patient selection. Aortic stenosis is often accompanied by myocardial hypertrophy, and Mitral insufficiency accompanies an increase in the left atrium. Both conditions facilitate the development of atrial fibrillation.In our center, less than 10% of cases are combined surgery based on the same retrospective research cited above. And based on that the same paper we cited above, we noticed that there was a higher incidence of AF in patients with combined surgery although the association was not significant after adjustment. However since some paper noticed this correlation, we have no issue in classifying this under the Exclusion Criteria. The authors should include only patients with preserved left ventricular function. Ventricular dysfunction is one of the leading causes of AF.Similarly we also noticed a statistically significant difference in the previous study we cited above in between patients with poor left ventricular function and preserved left ventricular function. And since these patients only constitute less than 10% (for EF<30%) and less than 1% (for EF<20%) of our total patients load of about 1800 CABG patients per year, we will also exclude them from the study. The use of corticosteroids should be prespecified and, preferably, avoided because of its known anti-inflammatory actions.It is not customary for our surgeons especially the co-researchers to add corticosteroids in the pump. We will specify that all patients must not be given corticosteroids. Patients with a history of long-term corticosteroids treatment at screening will be excluded. Timing from randomization and consequently beginning of Tocotrienol or placebo to surgery could be a relevant issue. A short time since medication start and surgery could compromise a possible significant effect of the drug on AF development.This is an issue which we have thought and discussed for quite some time. We wanted to load the patient a minimum of two days prior to surgery or preferably three days. The Institute protocol of Clinical Pathway for CABG does not allow patients to stay longer than 72hours prior to surgery. So we will ensure that all patients will receive their tocotrienol right after randomization and not less than two days prior to surgery. A Holter monitoring must be installed only in the first 72 hours. It is known that the most sensitive moment for the onset of AF is the immediate postoperative period. However, the use of the Holter monitoring for a longer period, even if uncomfortable, may be useful.We do not normally use Holter for our patients post-operatively but we will monitor them via continuous ECG monitoring. Our previous study cited above showed that 95% of patients developed AF within the first 48hours. We will monitor them using continuous ECG monitoring for 72hours post-operatively. In statistical analysis, the stepwise multiple logistic regression analysis should include atrium data such as size and/or volume, and procedure-related variables, such as use of inotropes and other vasoactive drugs, cross-clamp and cardiopulmonary bypass time. All of these variables might be associated with the occurrence of AF after revascularization procedures and should be included in the final model. On the other hand, overfitting should be avoided.We will include data on atrium size and/or volume, use of inotropes and other vasoactive drugs, cross-clamp and cardiopulmonary bypass time in the study proforma and the final model. We will modify and strengthen the proforma we used when collecting the data for our retrospective research before. Using multiple logistic regressions, we will choose the most parsimonious model and avoid overfitting of the model. Calibration of the model will be evaluated with the Hosmer-Lemeshow goodness-of-fit test, and discrimination with the Areas Under the ROC Curve (AUC) will be used as a guide towards model fitting strategy. In the multiple logistic regression model for the analysis of the factors influencing the change in quality of life after CABG, the authors should also include in the model angina data such as angina presence or frequency. In order to assess whether Tocotrienol is associated to better quality of life measures, it would be helpful if the authors control for angina relief after surgery.Yes, the frequency of angina will be included and controlled in the multiple logistic regression models for factors influencing the change in the quality of life after CABG. Thank you, Professors. Your comments are highly appreciated."
}
]
},
{
"id": "32234",
"date": "19 Apr 2018",
"name": "Olaf Stanger",
"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 the study protocol of a randomized controlled study to investigate the effect of a commercially available product (Tocovid Suprabio) that includes Tocotrienol, an Vitamin E isomer, against a placebo group, on the prevention of postoperative atrial fibrillation (AFib) in patients undergoing cardiac surgery.\n\nThe topic is of general interest. Postoperative AFib is common, is associated with complications and morbidity, thus consumes many resources and is, in part, potentially preventable. Because oxidative stress has been established as key mechanism, relatively cheap antioxidant treatment could be effective. The sample size under the current plan was calculated as 140 patients in each arm for a power of 0.8.\nThe ambitious study plan has some important shortcomings that must be addressed.\nAs the authors correctly note, postoperative AFib is assumed to be a multifactorial event and may be triggered by mechanisms other than oxidative stress. That has severe impact on exclusion criteria and statistical power calculations, that is difficult to assess for me without further information.\n\nThe authors correctly note that the use of cardiopulmonary bypass is involved in ischemia-reperfusion injury and radical oxygen species (ROS) generation. On page 3 it its stated that it is thought in consequence that off-pump surgery would reduce the incidence of AFib. I miss references, a clear statement whether or not this is thought to be a fact. In any case I would strongly suggest not to include off-pump patients and mix up the cohorts with on-pump cases. It is difficult to precisely quantify the difference from a metabolic point, and it appears impossible to draw a conclusion on the effect of a given dose with unknown blood and tissue concentrations on different levels and sources of ROS.\n\nWhat is known about the pharmacokinetics of the product? How quick and (in)complete must absorption be expected? What levels in blood and tissue can be achieved? Are interactions and competing drugs known?\n\nThe authors \"propose using a more potent isomer of vitamin E, namely Tocotrienol, instead of the usual Tocopherol\" (page 3). However in describing the commercially available drug on page 4 (Tocovid Suprabio by Hocid Berhad) it is clearly expressed, that the capsule contains Tocopherol (!) besides Tocopherol. So how will the investigators ever know which component, if any, was effective?\n\nThe authors aim at assessing the health-related quality of life in patients undergoing cardiac surgery; by definition a very sick group of elderly patients with usually many co-morbidities and with much drug intake. This aim is an extremely difficult matter involving many critical sources of definition, assessment tools (including SF-36), error and bias and I doubt that a meaningful result can be obtained.\n\nAccording to the study protocol patients are required to take the capsules \"before\" and immediately following CABG. Before is defined as \"immediately after randomization\" on page 4. It is quite normal that even elective cardiac surgery can not always be planned precisely on the exact day and hour, must be delayed for medical or organizational reasons. Therefore there will be patients with various time intervals of capsule intake and supposedly different blood and tissue concentrations of the study medication!\nCapsules intake should last for approx 6 weeks until the first follow-up. Nearly all cases of AFib occur within the first 4 days postoperative. It is extremely rare to observe first-time AFib later than that. Although the long intake is ambitious how do the investigators control full compliance?\n\nAgain, bioavailibility is an issue. The authors state on page 4 that bioavailibility of oral Tocotrienol is as low as 10-30%; how long before surgery must the drug be given to guarantee an effective availability in all patients of the non-placebo group? Is the trial controlled for interference with other drugs, malabsorption, possibly even gastric resections or concomitant intestinal disease etc that may\n\nprevent the study drug to be taken up effectively?\n\nDefinitely exclude patients undergoing concomitant valve surgery (page 4)! Operating times are longer, the heart will be OPEN with further exposure to ROS-producing surfaces and valve pathologies, stretch and surgical maneuvers are all known to cause and influence postoperative AFib. Up to 7% of patients following aortic valve replacement require permanent pacemakers anyway! Again, I would also suggest not to mix on- and off-pump patients.\n\nHolter ECG, very essential, will be monitored for the first 5 days, which makes sense. The study drug is given for approx 6 weeks. There will be no continuous ECG monitoring between discharge and follow-up. Unless the patient would be symptomatic and decides to see a physician who may be able to make a documentation, investigators are likely never to know whether there was an episode of AFib at any time between 1 and 6 weeks after surgery.\n\nI miss any information on potential PREoperative AFib assessment; or at least the exclusion of patients with AFib or other arrhythmias before surgery!\n\nAnd authors should exclude patients with vitamin or antioxidant intake before randomization! Check for all medications as many are known to exert antioxidant side effects.\n\nFinally I would consider it extremely valuable to obtain blood samples and analyse various ROS and vitamin concentrations (or at least the antioxidant capacity) to proof that any clinical outcome is directly linked to a mechanism as hypothesized. Otherwise, any outcome can be spurious and by chance, in principle and by definition.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Partly\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "3628",
"date": "08 May 2018",
"name": "Ahmad Farouk Musa",
"role": "Author Response",
"response": "The authors present the study protocol of a randomized controlled study to investigate the effect of a commercially available product (Tocovid Suprabio) that includes Tocotrienol, an Vitamin E isomer, against a placebo group, on the prevention of postoperative atrial fibrillation (AFib) in patients undergoing cardiac surgery. The topic is of general interest. Postoperative AFib is common, is associated with complications and morbidity, thus consumes many resources and is, in part, potentially preventable. Because oxidative stress has been established as key mechanism, relatively cheap antioxidant treatment could be effective. The sample size under the current plan was calculated as 140 patients in each arm for a power of 0.8.The ambitious study plan has some important shortcomings that must be addressed. As the authors correctly note, postoperative AFib is assumed to be a multifactorial event and may be triggered by mechanisms other than oxidative stress. That has severe impact on exclusion criteria and statistical power calculations, that is difficult to assess for me without further information. Thank you Prof for your comments and suggestions. In the revised version of this study, we will include a few other exclusion criteria namely poor LV, off-pump surgery, combined valve surgery and currently on or indicated for long-term corticosteroid treatment. The authors correctly note that the use of cardiopulmonary bypass is involved in ischemia-reperfusion injury and radical oxygen species (ROS) generation. On page 3 it its stated that it is thought in consequence that off-pump surgery would reduce the incidence of AFib. I miss references, a clear statement whether or not this is thought to be a fact. In any case I would strongly suggest not to include off-pump patients and mix up the cohorts with on-pump cases. It is difficult to precisely quantify the difference from a metabolic point, and it appears impossible to draw a conclusion on the effect of a given dose with unknown blood and tissue concentrations on different levels and sources of ROS. Our assertion that off-pump surgery would reduce the incidence of AF post-CABG was due to the fact that the extracorporeal circulation during on-pump surgery would activate the inflammatory cells thus increasing the oxidative stress that would later lead to AF. This was also acknowledged by Hashemzadeh K et al. in their paper published by the Journal of Cardiovascular and Thoracic Research 2013;5(2),45-49, with the title “Does off-pump coronary artery bypass reduce the prevalence of atrial fibrillation?, though we did not cite it. However, in the revised version of this Study Protocol, we have omitted Off-pump surgery. What is known about the pharmacokinetics of the product? How quick and (in)complete must absorption be expected? What levels in blood and tissue can be achieved? Are interactions and competing drugs known? Gan et al. published in Scientific Reports 2017; 7(1): p11542 on “Effect of palm-based tocotrienols and tocopherol mixture supplementation on platelet aggregation in subjects with metabolic syndrome: a randomised controlled trial” showed that a two-week supplementation of the same product and same dosing schedule resulted in plasma levels of approximately 0.58 ug/mL. And Patel et al. published in the Journal of Nutrition 2012; 142(3):513-519 on “Oral Tocotrienols are transported to human tissues and delay the progression of the model for end-stage liver disease score in patients” demonstrated that oral supplementation of tocotrienols in surgical patients up to four weeks before surgery significantly increased the tissue concentrations in blood, skin, adipose, brain, cardiac muscle, and liver. In our study, plasma tocotrienols levels will be measured to indicate the levels in the blood pre and post-surgery, and also on follow-up to determine if the blood levels are sufficient to exert clinical effects at the end of the study. And Tocotrienols have not been noted to have any interference with other drugs so far. The authors \"propose using a more potent isomer of vitamin E, namely Tocotrienol, instead of the usual Tocopherol\" (page 3). However in describing the commercially available drug on page 4 (Tocovid Suprabio by Hocid Berhad) it is clearly expressed, that the capsule contains Tocopherol (!) besides Tocopherol. So how will the investigators ever know which component, if any, was effective? The proposed investigation uses a commercially available product of natural mixed tocotrienols. Almost all natural sourced and marketed tocotrienols products contain a small fraction of tocopherols. In palm oil-derived tocotrienols, the composition been established as approximately 80% tocotrienols, 20% tocopherols. The investigation aims to evaluate the commercially available product (mixed tocotrienols, tocopherols and phytosterols) in the proposed effect on AF, and rather than isolated tocotrienols effect. We believe, the overall anti-oxidative effect may in part be contributed by the action of tocopherols with enhanced bioactivity from the much higher fraction of tocotrienols. The effects could potentially be synergistic to result in a clinically significant outcome. The authors aim at assessing the health-related quality of life in patients undergoing cardiac surgery; by definition a very sick group of elderly patients with usually many co-morbidities and with much drug intake. This aim is an extremely difficult matter involving many critical sources of definition, assessment tools (including SF-36), error and bias and I doubt that a meaningful result can be obtained. We acknowledged the concerns regarding the use of health-related quality of life (HRQOL) instruments, such as the SF-36, to document the QOL of principally sick, elderly patients undergoing CABG surgery. In order to improve the validity and reliability of the SF-36, a Malay version of the SF-36 has been developed and tested. Results suggest that the Malay-SF-36 is a valid and reliable tool to measure the HRQOL of patients undergoing CABG surgery at the National Heart Institute (IJN) Kuala Lumpur, Malaysia. Both English and Malay validated versions of SF-36 will be administered to the patients before surgery, at discharge, and at 6-weeks follow-up visit. A one-way repeated measure ANOVA analysis will be undertaken to assess changes in HRQOL across the three time points. According to the study protocol patients are required to take the capsules \"before\" and immediately following CABG. Before is defined as \"immediately after randomization\" on page 4. It is quite normal that even elective cardiac surgery can not always be planned precisely on the exact day and hour, must be delayed for medical or organizational reasons. Therefore there will be patients with various time intervals of capsule intake and supposedly different blood and tissue concentrations of the study medication!Capsules intake should last for approx 6 weeks until the first follow-up. Nearly all cases of AFib occur within the first 4 days postoperative. It is extremely rare to observe first-time AFib later than that. Although the long intake is ambitious how do the investigators control full compliance? The National Heart Institute has a protocol that must be adhered known as CABG Clinical Pathway where patients should not stay longer than seven days before surgery. Though we initially planned that patients should at least be on study medication three days before surgery, after randomisation, we find it rather difficult to achieve since there will be patients who will be admitted just two days prior to surgery at our centre. And to ensure that all study patients have a rather stable and sustainable level of the study medication, we have made it compulsory that the study patients to take the study drugs a minimum of seven days prior to surgery. This could be achieved by randomising the patients upon referral to the Cardiothoracic Wards for surgery once they have consented to be enrolled. We would take their blood samples at the beginning of randomization, pre-op, on Day-4 post-op, at discharge and on the first follow-up at six-weeks.The variability in dosing duration will be noted and supported with pre-surgery blood levels of tocotrienols. We will control for the dosing duration during analysis, if necessary. However, due to the short half-life of tocotrienols, the small variation of dosing duration is not expected to have a significant impact in the blood levels of tocotrienols. And we also realised the fact that almost all AF occurred during the first four days post-operatively and our previous retrospective study at our centre showed that the median time for the development of AF was 45 hours post-CABG where the majority developed within three days post-surgery with the second day being most common. But we would like patients to consume the study medication until the first follow-up which is six-weeks after discharge and will try to ensure compliance by asking the research assistants to remind the patients intermittently at home and that their blood will be taken upon their follow-up visit and pill-counting will be done as well. The patients will be informed before discharge about the blood taking and pill counting to ensure that they have good compliance.Apart from first-time AF, we will continue to monitor the safety endpoints, including symptoms requiring medical treatment during the post-discharge supplementation of 6 weeks. This will allow us to evaluate the effects of continued supplementation in post-CABG recovery. Again, bioavailibility is an issue. The authors state on page 4 that bioavailibility of oral Tocotrienol is as low as 10-30%; how long before surgery must the drug be given to guarantee an effective availability in all patients of the non-placebo group? Is the trial controlled for interference with other drugs, malabsorption, possibly even gastric resections or concomitant intestinal disease etc that may prevent the study drug to be taken up effectively? The investigational product is formulated with a patented delivery system that provides 100-200% enhanced bioavailability, of mixed tocotrienols. If taken on fasted state, tocotrienols, as with all oil-soluble vitamins, have been shown to have poor and erratic absorption due to dependence on food status. However, this is resolved using the enhanced absorption delivery system. The subjects are proposed to initiate supplementation at least one week from surgery in order to ensure levels of tocotrienols in the blood and tissues. All concomitant medication and medical history will be noted upon enrolment and throughout the study to evaluate any possible interaction effects on the study outcomes. The medical history and plasma levels will help us determine if gastrointestinal issues or any drugs may cause significant impact on the absorption and final clinical outcome. Definitely exclude patients undergoing concomitant valve surgery (page 4)! Operating times are longer, the heart will be OPEN with further exposure to ROS-producing surfaces and valve pathologies, stretch and surgical maneuvers are all known to cause and influence postoperative AFib. Up to 7% of patients following aortic valve replacement require permanent pacemakers anyway! Again, I would also suggest not to mix on- and off-pump patients. Yes Prof, we will exclude the concomitant valve surgery patients and Off-pump CABG. This is reflected in the revised version of the Study Protocol. Holter ECG, very essential, will be monitored for the first 5 days, which makes sense. The study drug is given for approx 6 weeks. There will be no continuous ECG monitoring between discharge and follow-up. Unless the patient would be symptomatic and decides to see a physician who may be able to make a documentation, investigators are likely never to know whether there was an episode of AFib at any time between 1 and 6 weeks after surgery. Based on our previous retrospective study, only 15% of our patients developed AF after more than three days post-surgery. Yes, we will never know if they would develop AF at home unless they present themselves to our centre again. But considering all our patients were discharged in sinus rhythm, we do not anticipate that they would develop AF after being discharged from the hospital. I miss any information on potential PREoperative AFib assessment; or at least the exclusion of patients with AFib or other arrhythmias before surgery! Yes, all patients documented with AF pre-operatively will be excluded from the study. And authors should exclude patients with vitamin or antioxidant intake before randomization! Check for all medications as many are known to exert antioxidant side effects. All pre-operative medications will be documented and patients on any form of anti-oxidant vitamins pre-operatively will be excluded. We will include this in our revised Study Protocol. Finally I would consider it extremely valuable to obtain blood samples and analyse various ROS and vitamin concentrations (or at least the antioxidant capacity) to proof that any clinical outcome is directly linked to a mechanism as hypothesized. Otherwise, any outcome can be spurious and by chance, in principle and by definition. Apart from anti-oxidative mechanisms, we postulate that tocotrienols may potentially inhibit the HMGCo-reductase to attenuate AF, similar to statins. We will consider including relevant blood biomarkers to support the findings of the study. However, due to transient effect of oxidative stress and whether blood levels are representative of cardiac condition remains a concern."
}
]
}
] | 1
|
https://f1000research.com/articles/7-215
|
https://f1000research.com/articles/7-1166/v1
|
31 Jul 18
|
{
"type": "Method Article",
"title": "To select relevant features for longitudinal gene expression data by extending a pathway analysis method",
"authors": [
"Suyan Tian",
"Chi Wang",
"Howard H. Chang",
"Chi Wang",
"Howard H. Chang"
],
"abstract": "The emerging field of pathway-based feature selection that incorporates biological information conveyed by gene sets/pathways to guide the selection of relevant genes has become increasingly popular and widespread. In this study, we adapt a gene set analysis method – the significance analysis of microarray gene set reduction (SAMGSR) algorithm to carry out feature selection for longitudinal microarray data, and propose a pathway-based feature selection algorithm – the two-level SAMGSR method. By using simulated data and a real-world application, we demonstrate that a gene’s expression profiles over time can be considered as a gene set. Thus a suitable gene set analysis method can be utilized or modified to execute the selection of relevant genes for longitudinal omics data. We believe this work paves the way for more research to bridge feature selection and gene set analysis with the development of novel pathway-based feature selection algorithms.",
"keywords": [
"Core subset",
"feature selection",
"gene set analysis",
"longitudinal microarray data",
"significance analysis of microarray (SAM)"
],
"content": "Introduction\n\nThe emerging field of pathway-based feature selection that incorporates biological information conveyed by gene sets/pathways to guide the selection of relevant genes1,2 has become increasingly popular and widespread. Here, a gene set or a pathway refers to a collection of genes that function together to influence and even regulate a specific biological process. In this study, the phrases “gene set” and “pathway” are used interchangeably.\n\nSince biological systems are dynamic, researchers are extremely interested in investigating gene expression patterns over a time course, in an effort to capture dynamical changes that are biologically meaningful and have casual implications. With the fast evolution of microarray technology and RNA-Seq technology, longitudinal experiments that collect gene expression profiles over a series of time points have become affordable and increasingly common in the fields of biomedicine and life science.\n\nThe analytical strategy typically employed for such longitudinal data is to stratify the whole dataset into separate subsets according to time points and then analyze the resulting subsets separately. This approach fails to consider the correlations among measures of a specific subject at different time points. Additionally, it overlooks those genes with trivial changes at individual points but non-marginal accumulated effects when taken together. Therefore, this approach is usually inefficient and lacks statistical power3–5.\n\nOn the other hand, some statistical methods that can analyze longitudinal gene expression data directly have been proposed. Among them, many have adopted a filter method to carry out the selection of relevant features for longitudinal gene expression profiles by screening genes one by one. For example, the GEE-based screening procedure by5 fits a GEE model6 to each gene and then excludes those non-significant genes (i.e., a gene with the corresponding p-value/q-value is larger than a pre-determined cutoff). By filtering genes one by one, this GEE-based screening is highly likely to include many redundant genes and thus to inflate the false positive rate. The redundant genes are irrelevant but suggested to be associated with the phenotype of interest by a feature selection method, mainly due to their correlations with the true relevant genes. Another example is the EDGE method proposed by Storey et al.3. The EDGE method is designed to identify differentially expressed genes over time between different phenotypes. This method utilized spline-based models to construct expression value-versus-time curves for individual genes and then screened genes one by one according to their significance levels. Again, this method has the same drawback as the GEE-based screening does, namely, the inclusion of many redundant genes.\n\nTo the best of our knowledge, there is no pathway-based feature selection algorithm for longitudinal gene expression data. Given the fact the pathway-based feature selection methods have been demonstrated to be superior to the conventional feature selection methods, there is an urgent need to develop pathway-based algorithms, in order to tackle longitudinal data.\n\nHere, we propose one extension to a pathway analysis method – significance analysis of microarray gene set reduction (SAMGSR)7 to conduct feature selection for longitudinal microarray data. In this modification, we extend SAMGSR by applying its reduction step twice. At the first reduction step, the core gene subsets corresponding to the selected gene sets are identified. Then, the essential time points of the selected genes are obtained subsequently. This extension is referred to as the two-level SAMGSR algorithm hereafter.\n\n\nMethods\n\nA previous version of this article is available as a pre-print on arXiv at Cornell University Library: https://arxiv.org/ftp/arxiv/papers/1511/1511.08272.pdf. In that version, the two-level SAMGSR algorithm and another extension we have made to the SAMGSR algorithm for longitudinal feature selection were included. After the preprint submission, we had made substantial modifications. Also we realized that the updated manuscript with two extensions together was easy to confuse the readers. Therefore, we decided to describe two extensions in separate manuscripts.\n\nData for the injury experiment were downloaded from the Gene Expression Omnibus repository. The accession number is GSE36809. This experiment was hybridized on Affymetrix HGU133 plus2 chips.\n\nIn this study, only patients with uncomplicated recoveries and patients with complicated recoveries were considered. According to Xia et al.8, uncomplicated recovery represents a recovery within 5 days while complicated recovery includes a recovery after 14 days, no recovery by 28 days, or death. If the recovery duration is longer than 14 days, the patient experienced complicated recovery for sure. Therefore, the possible time points for an uncomplicated recovery include days 1/2, 1, 4, 7 and 14, whereas those for a complicated recovery are days 1/2, 1, 4, 7,14, 21, and 28. Furthermore, we restricted our focus to the patients that had the full compliment of measurements (i.e., complicated patients with 7 measures and uncomplicated patients with 5 measures). The 25 uncomplicated patients and 18 complicated patients who met this request were put together and used as the training set. Since there were no measures for uncomplicated patients after 14 days, the data for patients with complications were truncated at 14 days.\n\nThe rest of patients including 50 uncomplicated patients and 23 complicated patients were used as a test set to validate the proposed method. In the test set, the time points considered were days 1/2, 1, 4, and 7. Of note, the characteristics of patients in the training set and the test set may be different since the test set includes patients who had been discharged early from the hospitals.\n\nSince different pre-processing procedures may impact the data analysis, we decided to download the summary expression values of the experimental data directly from the GEO database.\n\nThe gene sets were downloaded from the Molecular Signatures Database (MSigDB) (http://software.broadinstitute.org/gsea/msigdb). In this study, we considered the C2 category in this knowledgebase, which includes gene sets from curated pathways databases such as KEGG9 and those manually curated from the literature on gene expression.\n\nHere, we present a brief introduction to the SAMGSR algorithm7, and then discuss our extension for the purpose of feature selection for longitudinal gene expression data in detail.\n\nSAMGSR. The SAMGSR algorithm is an extension of the SAMGS method10 and provides additional reduction of significant gene sets into respective core subsets. According to Dinu et al.7, the SAMGSR method may result in an approximately 90% of reduction in the size of selected genes, in an effort to improve predictive performance and allow biological patterns to become more obvious.\n\nThe SAMGSR method consists of two major steps: the SAMGS process to identify important gene sets and the reduction step to refine those significant gene sets into their respective core subsets. In the SAMGS step10, an SAMGS statistic, the L2 norm of the SAM statistics11 for all genes within that gene set, is calculated. A p-value of the SAMGS statistic is computed using a permutation test by permuting phenotype labels of samples. Based on the p-value, the statistical significance of a gene set is determined. Second, focusing on those significant gene sets, the reduction step orders genes within a significant gene set j based on the magnitude of its SAM statistic and then gradually partitions the entire gene set into two subsets: the reduced subset Rk which includes the first k genes with largest SAM statistics, and the residual subset R¯k being the complement of Rk for k=1,…, |S-1|. Here S is the size of gene set j. Let ck be the SAMGS p-value of the residual subset R¯k. The optimal size of reduced set Rk is the smallest k such that ck is larger than a pre-specified cut-off for the first time.\n\nSAMGSR extension for longitudinal feature selection. In our modification to the SAMGSR method for the purpose of longitudinal feature selection, a gene set has two different meanings. First, it refers to a set of genes in a curated biological pathway. Second, it refers to a gene’s expression profiles across time. Correspondingly, the reduction step of SAMGSR is applied in an ordered manner at two levels: a lower level on gene sets and an upper level on the time points. First, the reduction procedure of SAMGSR is applied to identify the reduced core subsets within the selected/reduced gene sets obtained by the SAMGS step. Then upon the union of genes involved in those core subsets, the reduction procedure is applied again to determine exact time points where the expression values of those genes differ significantly between phenotypes. In the two-level SAMGSR algorithm, the SAMGS statistic is modified to as,\n\n\n\nwhere dij is the SAM statistic11 of gene i (i=1,…, |GSk|) at time point j (j=1,…, ti) and |GSk| is the size of gene set GSk (k=1,…,K, and K is the total number of gene sets). x¯d(ij) and x¯c(ij) are the sample averages of gene i at time point j for the diseased group and the control group, respectively. Parameter s(ij) is a pooled standard deviation that estimated by pooling all samples together, while s0j is a small positive constant used to offset the small variability in microarray expression measurements. Of note, both s(ij) and s0j are time-point specific because the variability of expression measurements differs at different time points. In the reduction step, SAMGS is calculated sequentially first to a series of subsets (for the size of 1,…, |GSk|-1) of a significant gene set, aiming to identify a core subset that makes an essential contribution to the statistical significance of this gene set. Then the algorithm moves to the level of time points, with the objective of determining which combination of time points contributes substantially to the importance of the specific gene. At this level, each gene’s expression profiles over time were viewed as a gene set. Our rational is that a gene’s expression values for the same individual over time are highly correlated, mimicking a gene set. Figure 1 elucidates the two-level SAMGSR algorithm.\n\nIn a separate study (unpublished study; Suyan Tian, Chi Wang, Howard H. Chang), we proposed another extension to the SAMGSR method for the purpose of longitudinal feature selection, which is referred to as the longitudinal SAMGSR method. The longitudinal SAMGSR method first applies the SAMGS step to select the relevant genes and then determines exact time point(s) that the expression values for a gene differ between two phenotypes. A potential disadvantage of this SAMGSR extension is it does not incorporate valuable biological information contained in pathways, which provides knowledge on how genes function in unison to impact on biology processes.\n\nIn both SAMGSR extensions, ck is regarded as a tuning parameter. Using the sequence of 0.05, 0.1, …, 0.5, the optimal value of ck corresponds to the one associated with the minimum 5-fold cross-validation (CV) error. In a 5-fold CV, a dataset was randomly divided into 5 roughly equal-sized folds, and 4 of these folds were used to train a classifier and the misclassification error rate was calculated upon the held-out fold. This step was repeated for each of the 5 folds as the held-out fold, and the error rates were averaged. Given the fact the SAMGSR extensions cannot estimate the coefficients of selected genes, support vector machine (SVM) models were fitted to estimate those coefficients. Then the posterior probability for a sample can be calculated for each time point.\n\nIn this study, we use four metrics - Belief Confusion Metric (BCM), Area Under the Precision-Recall Curve (AUPR), Generalized Brier Score (GBS), and the misclassified error – to evaluate the performance of a classifier. Our previous study1 provided detailed descriptions on those metrics. In summary, all these metrics have a range in between 0 and 1. For the first two, the closer to 1 the better a classifier is. In contrast, a value of 0 is optimal for the last two metrics. Given the SAMGSR extensions tend to identify those genes that are insignificant at isolated time points but significant jointly over time, an evaluation on individual time points using these statistical metrics might be unfair for the SAMGSR extensions, we also averaged the resulting posterior probabilities at each time point and then calculate the performance metrics using those averages.\n\nStatistical analysis was conducted in the R language version 3.1.2 (www.r-project.org). The Venn-diagram plot was made with the aid of an online bioinformatics tool. R codes of the two-level SAMGSR algorithm are given in the Supplementary File 1.\n\n\nResults and Discussion\n\nTraumatic injury with subsequent infection was a common cause of death in ancient times. Even today massive injury such as combat wounds remains life threatening12,13. A large clinical study that examined the genome-wide expression patterns of blood leukocytes in the immediate post-injury period was carried out several years ago8. A primary objective of that study was to explore if different patterns of gene expression existed for the two extremes of clinical recovery: the uncomplicated recovery and the complicated recovery. We used the longitudinal gene expression data collected specifically for this objective to evaluate our proposed method.\n\nFirst, the comparison between the two SAMGSR extensions and the SAMGSR separately at each time point was made. The two-level SAMGSR extension incorporates both the interaction information among genes inside a pathway and the correlations among the expression values of one specific gene over time. In contrast, the longitudinal SAMGSR extension only accounts for the correlations among the expression values of one specific gene over time, while the application of SAMGSR at individual time points only considers the interactions among genes inside a pathway. This comparison allows us to identify which factor - the interactions among genes inside pathways or the correlations among the same gene over time have significant impact on the performance of resulting signatures. The results were given in Table 1, from which we found the two-level SAMGSR method performs the best at the second and third time points while the longitudinal SAMGSR method does so at the first and fourth time points. Overall, the implementation of SAMGSR at separate time points has the worst performance.\n\nNote: 1 the posterior probabilities were calculated using an SVM classifier. Here, the cutoff for q-value in SAM-GS part is set at 0.05. # of genes represents the average number of genes over 5-fold cross-validated data selected by an algorithm at each time point for the five columns on the training set.\n\nAs we mentioned in the previous section, a separate evaluation on an algorithm at single time points might be unfair for both SAMGSR extensions, and an integrated evaluation that takes all time points into account is necessary. Therefore, we did such analysis. When taken all time points together, the superiority of the longitudinal SAMGSR method over the two-level SAMGSR method and the original SAMGSR method at separate time points has been established. The respective performance statistics are provided in Table 2.\n\nNote: 1 the posterior probabilities were calculated using an SVM classifier. Here, the cutoff for q-value in SAM-GS part is set at 0.05. # of genes represents the number of the union of individual genes selected at each time point. L-SAMGSR: the longitudinal SAMGSR method.\n\nIn summary, incorporation of the pathway information inside gene sets, the clusters of genes that might be potentially co-expressed/co-regulated together, did not result in the two-level SAMGSR having substantially superior performance. One possible explanation relates to the information quality of the pathway database itself. The canonically curated databases on pathways/gene sets are biased toward the well-studied diseases such as cancers, and with substantial less works investigating traumatic injury using gene expression profiles, the pathway knowledge contained inside those curated pathways may be uninformative for this specific disease.\n\nIn Figure 2, Venn-diagrams illustrate how the selected genes by the two-level SAMGSR method at different time points overlap. We observed that these two SAMGSR extensions perform comparably in term of overall model parsimony. Namely, the two-level SAMGSR algorithm identifies 94 unique genes while the longitudinal SAMGSR algorithm identifies 97 unique genes. Moreover, we observed that in the 94-gene signature identified by two-level SAMGSR, there is a substantial proportion of overlaps at all time points (24/94; 25.53%), while the number of genes being significant only at one specific time point is one half of this number. Again, this highlights the ability of our SAMGSR extensions (both the longitudinal SAMGSR method and the two-level SAMGSR method) to identify genes that present mild but concordant change across time points between two different phenotypes.\n\n(A) Venn-diagram illustrates the overlap of selected genes by the two-level SAMGSR method at different time points. (B) Venn-diagram illustrates the overlap of concordantly differentially expressed genes across all time points by the two-level SAMGSR algorithm and the longitudinal SAMGSR algorithm.\n\nUpon the overlapped 5 genes by both SAMGSR extensions, a plot (Figure 3) was made to compare the expression patterns over time between the complicated injury and the uncomplicated injury. It is observed that the pattern of change across time points for complicated patients versus uncomplicated ones is not quite unique or simple. Thus the results of our analysis provide no evidence on either the paradigm that complicated outcomes are associated with second hits or multiple inflammatory events which thus subsequently cause a secondary genomic response14,15, or its counterpart stating complication results from simultaneous and rapid induction of innate and suppression of adaptive immunity genes8. Further investigation is in demand.\n\nSubgroup sample means versus time plot for the 5 common genes that were identified as to be significant at all 5- time points between uncomplicated and complicated patients. Red line represents the complicated group while black line represents the uncomplicated group.\n\nIn order to investigate the properties of both SAMGSR extensions, we used observed expression values from the injury application to design two sets of simulations as in our previous study. Briefly, we chose two causal genes – F13A1 and GSTM1 – and then randomly selected 998 genes from the data serving as noise in the first simulation setting. Denote the expression value of gene i (F13A1 or GSTM1) at time j (j=1,…, 5) as Xi.j, the logit function of a complicated injury versus an uncomplicated injury is as following,\n\nlogitclu = 0.18XF13A1.1 + 0.57XF13A1.2 + 0.29XF13A1.3 + 0.41XF13A1.4 + 1.02XGSTM1.3\n\nHere, we considered one gene whose significance arises from its moderate joint contribution over time and the other whose association with the outcome is large at one specific time point. The aim of this simulation was to illustrate the inferred advantage possessed by the two SAMGSR extensions, namely, both of them incorporate the accumulated effect of genes over time, recognizing genes with mild or moderate change at each time point but with a coordinated change over time.\n\nIn the second simulation, we chose two genes – COX4I2 and RP9 as the relevant genes. The logit function was,\n\nlogitclu = 0.56XCOX4I2.1 − 0.91XRP9.5\n\nFor both simulation settings, 50 replicates were created. The results for these two simulations are given in Table 3 and Table 4, respectively. Unexpectedly, the longitudinal SAMGSR showed no inferiority to the two-level SAMGSR in both correctly selecting relevant genes and achieving a better model parsimony. Regarding model parsimony, the inferiority of two-level SAMGSR may stem from the pathway level; a relevant gene would be involved in many gene sets. Consequently, the number of highly correlated genes with the relevant ones might increase and since these genes are included in the final model, the parsimony of two-level SAMGSR naturally suffers. Regarding correct selection of causal genes, since two simulations are based on the injury data in which the biological knowledge might add no extra value to feature selection, as shown in the real-world application, it is unsurprising to have both algorithms correctly identify the causal genes.\n\nNote: # of genes represents the average number of genes selected by either the longitudinal SAMGSR algorithm or the two-level SAMGSR algorithm at each time point over 50 replicates. Ave # represents the average number of unique genes across 5 time points. The percentages of the causal genes being correctly selected at each time point over these 50 replicates are presented in the corresponding cells.\n\nNote: # of genes represents the average number of genes selected by either the longitudinal SAMGSR algorithm or the two-level SAMGSR algorithm at each time point over 50 replicates. Ave # represents the average number of unique genes across 5 time points. The percentages of the causal genes being correctly selected at each time point over these 50 replicates are presented in the corresponding cells.\n\nAlthough in the second simulation the number of relevant time points was less than that in the first one, the number of selected genes by both algorithms was dramatically larger in the second simulation. This might be because the relevant genes in the second simulation were highly correlated with other genes compared to the first simulation. The highly correlated structure between relevant features and irrelevant ones produced a large number of redundant features that both SAMGSR extensions, especially the two-level SAMGSR, cannot exclude. To our best knowledge, however, many feature selection algorithms, especially those based on filtering, may suffer from this drawback. As illustrated in our previous work16,17, an additional filtering using a relevant algorithm such as bagging18 may provide a solution to alleviate or eliminate this problem.\n\n\nConclusions\n\nBoth of the SAMGSR extensions incorporated the correlated structure of an expression’s profiles over time in the framework of gene sets, and were more likely to identify genes with coordinated and aggregated effects over time, while their effect size at individual time points may be insignificant. The naïve strategy of implementing feature selection separately at individual time points would overlook these genes. The employed process explains why the overlaps among the selected genes by both extensions over time were very large.\n\nThe curated pathways in major databases such as KEGG9 and GO19 tend to be enriched in the most prevalently studied diseases, e.g., cancers. Moreover, the pathways are far from completeness even for these diseases20. These facts potentially introduce biases and unfairness to an algorithm that utilizes pathway information to guide feature selection, e.g., the two-level SAMGSR method. One solution is to consider a statistical method to construct data-driven gene sets e.g., 21. Future work to construct such gene sets for longitudinal microarray data is needed, particularly to determine whether gene sets are stable or dynamic over time. Based on these facts, we suggest the longitudinal SAMGSR algorithm should be considered first, especially for an entry-level data analyst. If the diseases under investigation are cancers or the lab has its own customized pathways for the diseases, the two-level SAMGSR algorithm is recommended because the biological information contained in those pathways could provide more values on selecting relevant genes.\n\nIn this article, we adapted the SAMGSR method for feature selection of longitudinal gene expression profiles. To the best of our knowledge, this study is one of the few efforts to explore how to execute feature selection for longitudinal gene expression data, with additional consideration on pathway knowledge. Given that the two-level SAMGSR extension only performs comparable at individual time points and is even outperformed by the longitudinal SAMGSR method when considerer all time points together, our try here is not fruitful. Nevertheless, we believe this work paves the way for more research to incorporate pathway information to guide feature selection, with the development of novel algorithms to tackle longitudinal gene expression data.\n\n\nData availability\n\nThe microarray data were downloaded from the Gene Expression omnibus (GEO) repository, accession number GSE36809. The R codes of the two-level SAMGSR algorithm were given in Supplementary File 1.",
"appendix": "Competing interests\n\n\n\nNo competing interests have been declared.\n\n\nGrant information\n\nThis study was supported by a fund (No. 31401123) from the Natural Science Foundation of China.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Dr. Catherine Anthony for the English editing.\n\n\nSupplementary material\n\nSupplementary File 1: The R-codes for the two-level SAMGSR method.\n\nClick here to access the data.\n\n\nReferences\n\nTian S, Chang HH, Wang C: Weighted-SAMGSR: combining significance analysis of microarray-gene set reduction algorithm with pathway topology-based weights to select relevant genes. Biol Direct. 2016; 11(1): 50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang A, Tian S: Classification of early-stage non-small cell lung cancer by weighing gene expression profiles with connectivity information. Biom J. 2018; 60(3): 537–546. PubMed Abstract | Publisher Full Text\n\nStorey JD, Xiao W, Leek JT, et al.: Significance analysis of time course microarray experiments. Proc Natl Acad Sci U S A. 2005; 102(36): 12837–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang L, Zhou J, Qu A: Penalized generalized estimating equations for high-dimensional longitudinal data analysis. Biometrics. 2012; 68(2): 353–360. PubMed Abstract | Publisher Full Text\n\nXu P, Zhu L, Li Y: Ultrahigh dimensional time course feature selection. Biometrics. 2014; 70(2): 356–365. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZeger SL, Liang KY, Albert PS: Models for longitudinal data: a generalized estimating equation approach. Biometrics. 1988; 44(4): 1049–60. PubMed Abstract | Publisher Full Text\n\nDinu I, Potter JD, Mueller T, et al.: Gene-set analysis and reduction. Brief Bioinform. 2009; 10(1): 24–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiao W, Mindrinos MN, Seok J, et al.: A genomic storm in critically injured humans. J Exp Med. 2011; 208(13): 2581–2590. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOgata H, Goto S, Sato K, et al.: KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 1999; 27(1): 29–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDinu I, Potter JD, Mueller T, et al.: Improving gene set analysis of microarray data by SAM-GS. BMC Bioinformatics. 2007; 8: 242. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A. 2001; 98(9): 5116–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nProbst C, Pape HC, Hildebrand F, et al.: 30 years of polytrauma care: An analysis of the change in strategies and results of 4849 cases treated at a single institution. Injury. 2009; 40(1): 77–83. PubMed Abstract | Publisher Full Text\n\nBe NA, Allen JE, Brown TS, et al.: Microbial profiling of combat wound infection through detection microarray and next-generation sequencing. J Clin Microbiol. 2014; 52(7): 2583–2594. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeel M, Trentz O: Pathophysiology of polytrauma. Injury. 2005; 36(6): 691–709. PubMed Abstract | Publisher Full Text\n\nNast-Kolb D, Aufmkolk M, Rucholtz S, et al.: Multiple organ failure still a major cause of morbidity but not mortality in blunt multiple trauma. J Trauma. 2001; 51(5): 835–41; discussion 841-2. PubMed Abstract | Publisher Full Text\n\nTian S, Suárez-Fariñas M: Multi-TGDR: a regularization method for multi-class classification in microarray experiments. PLoS One. 2013; 8(11): e78302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTian S, Chang HH, Wang C, et al.: Multi-TGDR, a multi-class regularization method, identifies the metabolic profiles of hepatocellular carcinoma and cirrhosis infected with hepatitis B or hepatitis C virus. BMC Bioinformatics. 2014; 15: 97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBreiman L: Bagging predictors. Mach Learn. 1996; 24(2): 123–140. 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): 25–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCreixell P, Reimand J, Haider S, et al.: Pathway and network analysis of cancer genomes. Nat Methods. 2015; 12(7): 615–621. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangfelder P, Horvath S: WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008; 9: 559. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "36655",
"date": "17 Aug 2018",
"name": "Hulin Wu",
"expertise": [
"Reviewer Expertise Longitudinal data analysis",
"gene expression data analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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, authors intend to propose an extension of a pathway analysis method, SAMGSR, for longitudinal gene expression data. Both real data analysis and computer simulations are used to compare the modified method with the existing methods, which demonstrates some benefits of the proposed method. The proposed method is interesting, although the novelty is limited. Some other detailed comments and concerns are:\nMore explanation on “a gene’s expression profiles over time can be considered as a gene set”. Right top on Page 2: When the SAMGSR is used for the first time, the reference should be provided. Bottom right on Page 2: The sample sizes (the number of patients) are inconsistent on two different paragraphs. English may need to be further polished.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "37528",
"date": "03 Sep 2018",
"name": "Shengping Yang",
"expertise": [
"Reviewer Expertise High throughput data analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 proposes to use a two-level SAMGSR method for feature selection in longitudinal microarray data analysis. The idea of considering the over-time expressions of a gene as the expressions of genes in a gene set is quite interesting. The following are a few concerns that can be further clarified in the revised version of this paper.\n\nThe longitudinal SAMGSR model, which has not been published before, was only briefly introduced. Since it has not been published, I would suggest to formally introduce the longitudinal model and include it as an alternative option in this paper.\n\nData normalization is considered an important step in high throughput data analysis. It would great to provide a brief description. The t-like statistic is used in SAMGSR, and the focus of this paper is microarray data. I am wondering if it is possible to have a very brief discussion on the feasibility of applying the proposed method to RNA-seq data after certain data transformation, e.g., “voom” normalization. The inserted texts in Figure 1 do not seem to be consistent in size and font. Please provide the link (and last access date) to the online Venn-diagram plot tool. The first sentence in the abstract and introduction sections are identical. I would suggest making some minor modifications, e.g., using a more concise version in the abstract.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "37148",
"date": "12 Sep 2018",
"name": "Irina Dinu",
"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\nConsidering the intense method developments in time-course microarray data analysis, the authors should include a more rigorous literature review in background section supporting their claim about the existence of no pathway-based feature selection method.\n\nAs the proposed method is selecting both significant genes and significant time-points, the authors should explain “feature selection” in introduction in order to clarify the study objectives for the reader. In method description, the authors should describe how they calculated p-value and explain the corresponding permutation procedure in each of the steps. This method is able to detect the genes with different “levels” of expression over time and it fails to detect differential temporal patterns of gene expressions. The figure 3 also shows that the significant genes are different in expression levels and similar in temporal patterns. This limitation should be discussed by the authors. How the method performs when one gene expression is a noisy realisation of the other one? Does the method fail to detect the similarity of these two expression trajectories? If yes, please discuss. More specifically, if xbar(D)=xbar(C)+E and mean(E)=0; then xbar(D) is a noisy realisation of xbar(C) and the test should fail to reject the null hypothesis. However, by this method, we may get very large statistic and reject the null hypothesis.\n\nIn the second step, the proposed method treats time as quantitative variable and disregards the temporal order of measurements. According to the authors: “each gene’s expression profile over time was viewed as a gene-set”. What limitations will this feature impose to the analysis?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1166
|
https://f1000research.com/articles/7-1158/v1
|
30 Jul 18
|
{
"type": "Research Article",
"title": "Characteristics and management of ventricular shunt infections in children, 2000-2015: a single centre retrospective chart review",
"authors": [
"Iris C. Feijen",
"Charlene M.C. Rodrigues",
"Christopher J.A. Cowie",
"Claire Nicholson",
"Muhammad Raza",
"Marieke Emonts",
"Iris C. Feijen",
"Charlene M.C. Rodrigues",
"Christopher J.A. Cowie",
"Claire Nicholson",
"Muhammad Raza"
],
"abstract": "Background: Infections are a common and serious complication of ventricular shunts that can lead to significant mortality and morbidity. Treatment consists of surgical and antimicrobial therapy, but there is a lack of evidence regarding optimal management. We therefore aimed to analyse the current practice and patient outcomes within a large tertiary referral centre. Methods: We identified cases of infection in ventriculoperitoneal shunts from January 2000 until April 2015 in our institution. All patients were under 18 years at the time of infection. Clinical, microbiological and radiological data were collected with the use of a standardised proforma. Non-parametric tests were used for statistical analysis. Results: There were 92 episodes of infection in 65 patients. The most common microorganisms were coagulase-negative staphylococci (47%), followed by Staphylococcus aureus (16%). Surgical treatment included shunt externalisation (15%) and complete removal (67%). Antibiotics were given in 97% of the patients in addition to surgery. Vancomycin, linezolid, cefotaxime, meropenem and rifampicin were used most frequently. The median duration of antibiotic treatment was 18 days (IQR 14-25 days). Two patients died from consequences of a shunt infection and seven had recurrent infection. Conclusions: It would be beneficial to develop a guideline for recognition and treatment of shunt infections. Complete removal of the shunt and placement of an EVD seems the safest surgical treatment. Empirical antibiotic treatment should be started as soon as possible. A combination of linezolid and ceftriaxone would be appropriate first line antibiotics, with meropenem as second line. Antibiotics can be rationalised once the CSF culture results are known.",
"keywords": [
"Ventriculopertioneal shunt infection",
"Hydrocephalus",
"children",
"Staphylococcus aureus",
"Coagulase-negative staphylococci",
"Cerebrospinal fluid shunt"
],
"content": "Introduction\n\nHydrocephalus is a common and serious neurosurgical condition. The incidence is estimated between 0.4-4 cases per 1000 births worldwide1,2. In some cases there is a treatable underlying cause of hydrocephalus (e.g. a posterior fossa tumour) but in most cases cerebrospinal fluid (CSF) diversion is required. CSF diversion usually requires some kind of shunt, although endoscopic third ventriculostomy is an option in certain types of obstructive hydrocephalus. Different types of shunts are available, of which ventriculoperitoneal (VP) shunts are most frequently used. Other types of shunts include; ventriculoatrial (VA), ventriculopleural and lumboperitoneal (LP)2–5.\n\nShunt insertion procedures are not without risk and infection is a frequent complication6,7. Rates of infection per shunt vary widely, between 3 and 13%3,6,8–12. Overall shunt infection rate in the UK is estimated at 5.2%7. Shunt infections have a mortality of 10.1% and worse Glasgow Outcome Scale scores, as well as worse school performance in the long-term13. From a public health perspective, shunt infections are associated with major hospital costs, with one episode costing approximately £29,00014. Therefore, prevention or prompt recognition and treatment of infection are imperative.\n\nThere are recognised factors known to increase the risk of shunt infection. In general, younger and premature children have a higher risk due to immaturity of their immune system and differences in skin flora compared to older children15. Premature infants who have had shunt surgery before 40 weeks gestational age have the greatest risk of all ages (Hazard radio (HR): 4.72, 95% CI 1.71-13.06)9,10,14,16. Children who have multiple shunt revision procedures, either because of shunt blockage, malfunction, fracture or infection are also at greater risk as with each revision, the cumulative risk rises11. There are technical aspects of shunt surgery that increase risk, including handling of the shunt system with sterile surgical gloves contaminated early in the procedure (HR 1.07, 95% CI 1.02-1.12, p=0.01). The risk is higher still if there is a post-operative CSF leak through the wound, allowing skin flora to access the shunt system (HR 19.16, 95% CI 6.96-52.91, P<0.0001)4,16.\n\nRecognition of shunt infections in children can be difficult as they present with a wide variety of non-specific symptoms such as fever, vomiting, drowsiness, headaches, irritability, seizures and abdominal pain3,10,17. Up to 90% of shunt infections occur within the first three to nine months after shunt insertion. During this period there should be a low threshold for investigating patients who display any of these symptoms3,4,7,12.\n\nThe organisms most frequently identified are skin flora, with staphylococcal species accounting for up to 80% of shunt infections9,14,18. Staphylococcus epidermidis is the most frequent coagulase-negative staphylococci (CoNS)7,12. Other common organisms are Staphylococcus aureus (5-26%) and other CoNS (36.1-53%)9–11,17,19. Gram-negative organisms are also frequent findings, accounting for up to 54% of infections in some series20. Gram-negative rods were identified in approximately 7–9% of infections11,17.\n\nTreatment of shunt infection requires complete removal of the device in combination with temporary drainage, unless infection is confined to the abdomen in which case there is a period of shunt externalisation before replacement18,20. Early surgical intervention is especially important since staphylococci commonly form a biofilm after colonising an implant, making it more difficult for antibiotics to penetrate and eradicate infection21. Additionally, early and prolonged intravenous and/or intrathecal antibiotic therapy are needed to sterilise the CSF. It is recommended to wait until the cultures have been sterile for at least 72 hours before implanting a new shunt5. There is some evidence that in case of low-grade inflammation the patient can sometimes be treated with antibiotics alone, but success rates are reported to be low3,4,13,14.\n\nOptimal duration and choice of antibiotic treatment are unclear in the current literature. There is no correlation between recurrence of infection and the duration of antibiotic therapy given, suggesting that a shorter treatment course could be as effective as a longer one22. Due to the increasing concern of multidrug-resistant bacteria, the use of penicillin has largely been replaced by vancomycin18. Linezolid is a good alternative, effective against most Gram-positive microorganisms, including multidrug-resistant strains23. There are currently no randomized studies, large case series, recent prospective studies or guidelines regarding the treatment of ventricular shunt infections in children14.\n\n\nStudy objectives\n\nGiven the severity of this condition and the lack of evidence-based management strategies, there is a need for evaluation of current practice to establish a standard of care. Therefore, this study aimed to retrospectively analyse the management of ventricular shunt infections in children in the regional referral centre for the North East of England, the Great North Children’s Hospital in Newcastle upon Tyne, UK. We aimed to describe the clinical presentation, diagnostic efficacy, as well as surgical and medical treatment regimens including antibiotic use. In addition, we hope to identify areas for improvement in clinical practice to reduce mortality, re-infection rates or neurological sequelae in these children.\n\n\nMethods\n\nAll children (≤18 years) diagnosed with shunt infections presenting to the Great North Children’s Hospital, Newcastle Upon Tyne, UK over the last 15 years were included in this study. Only children with a permanent shunt device were included. This could either be a VP shunt, a VA shunt, a ventriculopleural shunt or a LP shunt. Patients who had the infection before the year 2000 were excluded from the study, as were patients who were over 18 years old at the time of infection. A shunt infection was defined as any patient coded as ‘shunt infection’ at discharge or any patient who received treatment for suspected shunt infection, with or without microbiological confirmation.\n\nPatient lists were obtained from clinical coding department of the hospital. Using broad search terms ‘shunt infection’, ‘CSF infection’ and patients aged ≤18 years old, 708 possible patients were identified. After screening for eligibility with a digital archive for discharge and referral letters, using the criteria listed above, 65 patients were included in the study (Figure 1).\n\nA proforma (Supplementary File 1) was compiled including all relevant clinical, microbiological and radiological information. It was completed by medical staff using clinical records, laboratory (biochemistry and microbiology) results, radiology scans and reports, discharge and referral letters. Data was analysed using Microsoft Excel 2010. Statistical methods used were non-parametric.\n\nPrior to this study, there was no standard protocol at this institution for the management of CSF shunt infections. It was common practice to give a single dose of cefuroxime at induction in theatre as prophylaxis prior to surgery for shunt insertion. Additionally, gentamicin was used occasionally to flush through the shunt system prior to implantation, depending upon surgeon preference. Antibiotics were not routinely given post-operatively for a first time shunt insertion without suspected infection. Shunt infections were managed on a case-by-case basis. Usually the shunt was removed completely and a new shunt inserted as soon as possible after CSF sterilisation. For the purposes of this study, CSF sterilisation was defined as the first negative CSF culture that remained negative in repeated cultures.\n\nEthical approval was not required as this was a quality improvement study. The Caldicott principles, a framework of good practice in the use of patient information, were adhered to throughout the study in agreement with local hospital policies. The Clinical Governance Department of the Trust provided Caldicott approval under ID no. 4251. All data were collected and stored securely on password protected hospital computers and all data were anonymised.\n\n\nResults\n\nA total of 65 patients were included in this study, 55.4% were male (n=36, p=0.19), 46.1% were preterm, 33.8% were born at term and 20.0% unknown due to lack of clinical information. Three different types of shunt were used, VP shunts (n=61, 95.3%) were the most frequently used, two patients had LP shunts and one a VA shunt. Only one shunt was known to be antibiotic impregnated. The majority (n=49, 83.0%) had their first shunt before the age of one year. The median age at which patients received their first shunt was two months (IQR 1-6 months) (Table 1).\n\nOf the 92 episodes, diagnosis relied on a positive CSF culture in 61 (66%) cases, a positive shunt culture in 8 (9%) cases and a positive wound culture in 3 (3%) cases. A further 8 (9%) episodes had negative cultures, but were diagnosed by white blood cell count in the CSF >5/mm3. In the rest of the episodes (n=12, 13%) diagnosis was made solely on clinical data.\n\nThe median number of leukocytes in CSF was 46/mm3. A blood culture was obtained in 75% (n=68) of episodes, of which only 3% (n=2) had a positive result. In 43% (n=39) of episodes a CT scan was performed to exclude shunt blockage. In only 5% (n=2) of these episodes the scan was suggestive of infection. Ventriculitis may be seen as irregular enhancement of the ependymal lining of the ventricles. For three of the scans no final report was available.\n\nOf the 65 patients, 42 (64.6%) received prophylactic antibiotics prior to first shunt insertion. The other 23 patients either did not receive antibiotics or had incomplete or missing documentation. Of the patients who received a single dose of antibiotics (n=35), the majority received prophylactic cefuroxime (n=23, 65.7%). Other antibiotics included; vancomycin (n=4. 11.4%), flucloxacillin (n=4, 11.4%), cefotaxime (n=2, 5.7and benzylpenicillin (n=2, 5.7%).\n\nThe median age at the first infection was 36 months (IQR 4-100 months) and 40.6% (n=26) had their first infection in the first year of life. Patients presented with a variety of non-specific symptoms. Of all the episodes of infection for which information was available (n=91), 61 (67.0%) presented with fever. Other common symptoms were irritability (n=27, 29.7%), vomiting (n=26, 28.6%), cutaneous manifestations, for example rashes or erythema over the shunt tract, (n=21, 23.1%) and abdominal pain (n=17, 18.7%). Infections with S. aureus presented more often with cutaneous manifestations (n=5, 14%), headache (n=3, 9%) and diarrhoea, (n=3, 9%), while CoNS infections presented more often with irritability (n=13, 18%) (Figure 2).\n\nInformation regarding treatment regimens was available for 91 episodes. A total of 97 microorganisms could be identified in 73 (80.2%) of the episodes, with 31 different microorganisms causing shunt infection (Figure 3). The two largest groups were human skin flora, CoNS (n=46, 47.4%) and S. aureus (n=16, 16.5%). In 19.8% (n=18) of episodes no microorganisms were identified. Eighty-eight patients (96.7%) received antibiotic treatment. Complete information about the treatment course could be collected from 62 episodes from 45 patients. The median number of days these patients received antibiotics per episode was 18 days (IQR 14-25 days). The shortest period was one day and the longest 52 days. Most patients received more than one type of antibiotic during their treatment course. Only 4.8% (n=3) of episodes were treated with one antibiotic. The median number of different antibiotics received per episode was three, with the maximum of nine in one patient (Table 2).\n\nIn total 25 different types of antibiotics were used to treat 88 episodes. Vancomycin was used most frequently, in 58.0% (n=51) of episodes. Furthermore, linezolid was used in 48.9% (n=43), cefotaxime in 43.2% (n=38), meropenem in 38.6% (n=34) and rifampicin in 33.0% (n=29). Other antibiotics were used in less than 20% of the episodes (Figure 4). A schematic representation of the antibiotic treatment per patient was generated, which showed no overall trends in antibiotic use. Table 3 and Table 4 show the different combinations of antibiotics used for five days or more in S. aureus and CoNS infections respectively to give a clearer view of overall use. Empirical use of antibiotics at the start of treatment is highlighted for comparison. We have limited this information to CoNS and S. aureus infections due to a wide variation in other pathogens (n= 1-2 maximum).\n\nX are antibiotics used for ≥5 days overall, highlighted are antibiotics were started empirically at diagnosis.\n\nVanc = vancomycin, Line = linezolid, Rifa = rifampicin, Fluc = flucloxacillin, Mero = meropenem, Cotr = co-trimoxazole, Cefo = cefotaxime, Cefu = cefuroxime, Metr =metronidazole, Teic = teicoplanin, Ceft =ceftriaxone\n\nX are antibiotics used for ≥5 days overall, highlighted are antibiotics were started empirically at diagnosis.\n\nVanc = vancomycin, Line = linezolid, Rifa = rifampicin, Fluc = flucloxacillin, Mero = meropenem, Cotr = co-trimoxazole, Cefo = cefotaxime, Cefu = cefuroxime, Cipr = ciprofloxacin, Metr = metronidazole, Teic = teicoplanin, Ceft = ceftriaxone, Cefa = cefalexin, Cefta = ceftazidime, Gent = gentamicin\n\nRegarding surgical management, 67.0% (n=61) had shunt removal, 15.4% (n=14) had shunt externalisation and 26.4% (n=24) were treated without any surgical intervention. Date of sterilisation of the CSF was available for 57 episodes. The time between infection of the shunt and sterilisation of the CSF varied and was between 1 and 28 days. The median was 4 days (IQR 2-10 days). The median number of days between sterilisation of the CSF and insertion of a new shunt was 11 days (IQR 5-17 days).\n\nRecurrent infections occurred in 18/64 patients (28.1%). In seven of these episodes the reinfection was with the same microorganism within 90 days of the previous infection. All seven patients were treated with appropriate antibiotics for the first infection and six of them had sterilised CSF after treatment. Three of the patients had no surgery, one had a shunt externalisation and three had shunt removal for this episode of infection.\n\nAmong these 92 infections in 65 patients, a total of 146 new shunts were implanted and a total of 178 shunts were revised, of which 120 revisions were not due to infection, but predominantly due to shunt blockage. Most of the infections took place shortly after insertion or revision of the shunt. 58.8% (n=47) took place within 100 days after insertion of a shunt and 70.4% (n=57) within 100 days of last contact with the shunt (either placement or revision) (Figure 5). 67.5% (n=54) of infections took place within nine months after placement of a new shunt and 58.8% (n=47) within three months. 74.1% (n=60) of the infections took place within nine months of last contact with the shunt and 70.4% (n=57) within three months. There was a significant positive correlation between the number of infections and shunts, Spearman rho 0.247 (p=0.05) and a non-significant correlation between the number of infections and revisions, Spearman rho 0.153 (p=0.23).\n\nFrom January 2000 until May 2015 all-cause mortality was 10.8% (n=7) and shunt infection related mortality was 3.1% (n=2).\n\n\nDiscussion\n\nInfants and children born prematurely are known to have a higher risk of shunt infection9,10,14,16. In our population 46% of the patients were born prematurely and 41% had the first infection before the age of one year. This reaffirms that in these groups the suspicion of a shunt infection should be promptly investigated and treated. Gender does not seem to be a risk factor for developing a shunt infection11,13,16 and there was no difference between the number of males and females infected (55% male, p=0.19) in our study. In our population 68% of infections occurred within nine months of shunt insertion and 59% within three months. This suggests that the shorter the period after shunt insertion, the higher the suspicion of an infection. Our percentages are lower than those reported in the literature; 90% within three to nine months4,7,12, which may suggest fewer episodes might be attributed to microbial colonisation during surgery in our cohort. It has been shown that the cumulative risk of infection increases with every revision11. Our data also show a significant correlation between number of shunts and number of infections. However, we found no correlation between the number of revisions and the number of infections. This could be because the definition of ‘revision’ is complex and used for a large range of surgical procedures in our records.\n\nPreventative strategies during surgery may influence the incidence of shunt infections. One of the most important actions shown to reduce shunt infections is perioperative prophylactic antibiotics, with the literature supporting their use in all patients undergoing a shunt insertion2,24. A systematic review and meta-analysis by Ratilal et al. demonstrated that antibiotic prophylaxis can significantly decrease the rate of shunt infection (OR 0.51, 95% CI 0.36-0.73)25. Only 65% of the patients in our populations had documented evidence of receiving antibiotics before their first shunt insertion. However, due to incomplete availability of records and a lack of documentation this number is very likely underestimated. Clearer documentation will be necessary in the future to be able to accurately determine this modifiable factor. In the Great North Children’s Hospital it is standard practice to use cefuroxime as prophylaxis before surgery. Cefuroxime covers for infections with S. aureus, but does not cover for infections with CoNS, which was the most frequently found organism causing shunt infection. It may therefore be beneficial to consider using teicoplanin as prophylaxis which is effective against both based on local resistance patterns.\n\nMany different, non-specific symptoms can occur as a consequence of shunt infection3,10,17. In this study, fever was the most common symptom (67% of the episodes), with cutaneous manifestations more frequent with S. aureus infections and irritability associated with CoNS infections. In children with a shunt who complain of fever, irritability, vomiting, cutaneous manifestations or abdominal pain the suspicion of an infection should be high.\n\nS. epidermidis, S. aureus and other CoNS were the most frequently observed and together accounted for 54% of the organisms found. This correlated with findings from other groups from the United Kingdom, the United States of America, Taiwan, Korea, Spain and Germany7,9–12,14,17–20. Another relatively frequent organism in the literature was Propionibacterium acnes19,26,27 which was not identified in our cohort possibly due to the fact this organism is mainly found in patients over the age of 14–15 years old, whereas 88% (n=80) of the episodes in this study occurred before 14 years.\n\nIn this cohort, 67% underwent shunt removal and 15% underwent shunt externalisation (n=14, of which 9 went on to have shunt removal). In CSF infections, it is best practice to remove the infected device completely and insert an EVD until the CSF is sterilised18 ,20. However, in infections very soon after insertion of a shunt, treatment with antibiotic alone could be justified, as the microorganisms have not had the time yet to form a biofilm, but there is a paucity of evidence for this at present and currently the safest option is to remove the infected device completely. There are no specific guidelines, but it is considered safe to insert a new shunt when the CSF has been sterile for 48–72 hours5. Since antibiotic treatment should be started as soon as possible, it may be beneficial to develop protocols for the use of empirical antibiotic therapy after CSF sampling at presentation. The agent chosen should cover for at least CoNS and S. aureus, since they were the most common findings. If there is abdominal pain, it may be advisable to choose an agent that also covers for Gram-negative bacteria. This agent should also be able to penetrate the CSF. A combination of linezolid and ceftriaxone would fulfil these requirements, with meropenem as a second line agent. However, specific local resistance patterns should be taken into account. Antibiotics can be rationalised once the CSF culture results are known. The optimal duration of treatment is still unclear,but should be at least seven to twelve days after the last positive culture, and depending on the microorganisms identified, according to expert opinion. It is best to start treatment IV, as this will allow CSF delivery of antibiotic. Once a patient with confirmed ventriculitis has an EVD, antibiotics should be switched to intrathecal administration when possible, since this seems to be the optimal method of antibiotic delivery to CSF28,29.\n\nLimitations of this study included, the incomplete availability of clinical records and incomplete documentation, resulting in information that was not possible to acquire retrospectively, in particular antibiotic prophylaxis, treatment courses and surgical notes. We were unable to obtain the total number of CSF shunts inserted during this period and were therefore unable to calculate the incidence of shunt infections in our region.\n\n\nData availability\n\nDataset 1: Data outlining the demographics and treatment course for each anonymised patient. 10.5256/f1000research.15514.d21161230\n\nThe results presented here have previously been presented as a part of an abstract for the 2016 34th Annual European Society for Paediatric Infectious Diseases meeting and can be found here.",
"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\nAcknowledgments\n\nSpecial thanks to the clinical records department of the Royal Victoria Infirmary for their help in obtaining clinical records.\n\n\nSupplementary material\n\nSupplementary File 1: Proforma\n\nClick here to access the data.\n\n\nReferences\n\nScadding JW, Losseff N: Clinical neurology. 4th ed. London: Hodder Arnold, 2012. Reference Source\n\nKaye AH: Essential neurosurgery. 3rd ed. Malden, Mass.: Blackwell, 2005. Reference Source\n\nPople IK: Hydrocephalus and shunts: what the neurologist should know. J Neurol Neurosurg Psychiatry. 2002; 73 Suppl 1: i17–22. PubMed Abstract | Free Full Text\n\nForsyth R, Newton RW: Paediatric neurology. Second edition. ed. Oxford, United Kingdom: Oxford University Press, 2012. Reference Source\n\nEllenbogen RG, Abdulrauf SI, Sekhar LG: Principles of neurological surgery. Philadelphia, PA: Saunders/Elsevier, 2012. Reference Source\n\nWong JM, Ziewacz JE, Ho AL, et al.: Patterns in neurosurgical adverse events: cerebrospinal fluid shunt surgery. Neurosurg Focus. 2012; 33(5): E13. PubMed Abstract | Publisher Full Text\n\nRichards HK, Seeley HM, Pickard JD: Efficacy of antibiotic-impregnated shunt catheters in reducing shunt infection: data from the United Kingdom Shunt Registry. J Neurosurg Pediatr. 2009; 4(4): 389–93. PubMed Abstract | Publisher Full Text\n\nSimon TD, Hall M, Riva-Cambrin J, et al.: Infection rates following initial cerebrospinal fluid shunt placement across pediatric hospitals in the United States. Clinical article. J Neurosurg Pediat. 2009; 4(2): 156–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang KW, Chang WN, Shih TY, et al.: Infection of cerebrospinal fluid shunts: causative pathogens, clinical features, and outcomes. Jpn J Infect Dis. 2004; 57(2): 44–8. PubMed Abstract\n\nLee JK, Seok JY, Lee JH, et al.: Incidence and risk factors of ventriculoperitoneal shunt infections in children: a study of 333 consecutive shunts in 6 years. J Korean Med Sci. 2012; 27(12): 1563–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcGirt MJ, Zaas A, Fuchs HE, et al.: Risk factors for pediatric ventriculoperitoneal shunt infection and predictors of infectious pathogens. Clin Infect Dis. 2003; 36(7): 858–62. PubMed Abstract | Publisher Full Text\n\nSamandouras G: Shunts and endoscopic third ventriculostomy. The Neurosurgeon's Handbook: Oxford University Press (OUP), 2010; 696–702.\n\nVinchon M, Dhellemmes P: Cerebrospinal fluid shunt infection: risk factors and long-term follow-up. Childs Nerv Syst. 2006; 22(7): 692–7. PubMed Abstract | Publisher Full Text\n\nGutiérrez-González R, Boto GR, Pérez-Zamarrón Á: Cerebrospinal fluid diversion devices and infection. A comprehensive review. Eur J Clin Microbiol Infect Dis. 2012; 31(6): 889–97. PubMed Abstract | Publisher Full Text\n\nPople IK, Bayston R, Hayward RD: Infection of cerebrospinal fluid shunts in infants: a study of etiological factors. J Neurosurg. 1992; 77(1): 29–36. PubMed Abstract | Publisher Full Text\n\nKulkarni AV, Drake JM, Lamberti-Pasculli M: Cerebrospinal fluid shunt infection: a prospective study of risk factors. J Neurosurg. 2001; 94(2): 195–201. PubMed Abstract | Publisher Full Text\n\nvon der Brelie C, Simon A, Gröner A, et al.: Evaluation of an institutional guideline for the treatment of cerebrospinal fluid shunt-associated infections. Acta Neurochir (Wien). 2012; 154(9): 1691–7. PubMed Abstract | Publisher Full Text\n\nAdams DJ, Rajnik M: Microbiology and Treatment of Cerebrospinal Fluid Shunt Infections in Children. Curr Infect Dis Rep. 2014; 16(10): 427. PubMed Abstract | Publisher Full Text\n\nFulkerson DH, Sivaganesan A, Hill JD, et al.: Progression of cerebrospinal fluid cell count and differential over a treatment course of shunt infection. J Neurosurg Pediatr. 2011; 8(6): 613–9. PubMed Abstract | Publisher Full Text\n\nDrew RJ, Cole TS, Lee MK, et al.: Antimicrobial treatment options for neurosurgical ventricular shunt infections in children from 1993 to 2012: a systematic review. Childs Nerv Syst. 2014; 30(5): 841–50. PubMed Abstract | Publisher Full Text\n\nFey PD: Modality of bacterial growth presents unique targets: how do we treat biofilm-mediated infections? Curr Opin Microbiol. 2010; 13(5): 610–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKestle JR, Garton HJ, Whitehead WE, et al.: Management of shunt infections: a multicenter pilot study. J Neurosurg. 2006; 105(3 Suppl): 177–81. PubMed Abstract | Publisher Full Text\n\nYılmaz A, Dalgic N, Müslüman M, et al.: Linezolid treatment of shunt-related cerebrospinal fluid infections in children. J Neurosurg Pediatr. 2010; 5(5): 443–8. PubMed Abstract | Publisher Full Text\n\nKlimo P Jr, Van Poppel M, Thompson CJ, et al.: Pediatric hydrocephalus: systematic literature review and evidence-based guidelines. Part 6: Preoperative antibiotics for shunt surgery in children with hydrocephalus: a systematic review and meta-analysis. J Neurosurg Pediatr. 2014; 14 Suppl 1: 44–52. PubMed Abstract | Publisher Full Text\n\nRatilal B, Costa J, Sampaio C: Antibiotic prophylaxis for surgical introduction of intracranial ventricular shunts: a systematic review. J Neurosurg Pediatr. 2008; 1(1): 48–56. PubMed Abstract | Publisher Full Text\n\nArnell K, Cesarini K, Lagerqvist-Widh A, et al.: Cerebrospinal fluid shunt infections in children over a 13-year period: anaerobic cultures and comparison of clinical signs of infection with Propionibacterium acnes and with other bacteria. J Neurosurg Pediatr. 2008; 1(5): 366–72. PubMed Abstract | Publisher Full Text\n\nConen A, Walti Laura N, Merlo A, et al.: Characteristics and treatment outcome of cerebrospinal fluid shunt-associated infections in adults: a retrospective analysis over an 11-year period. Clin Infect Dis. 2008; 47(1): 73–82. PubMed Abstract | Publisher Full Text\n\nRagel B, Browd S, Schmidt R: Surgical shunt infection: significant reduction when using intraventricular and systemic antibiotic agents. J Neurosurg. 2006; 105(2): 242–7. PubMed Abstract | Publisher Full Text\n\nMoussa WM, Mohamed MA: Efficacy of postoperative antibiotic injection in and around ventriculoperitoneal shunt in reduction of shunt infection: A randomized controlled trial. Clin Neurol Neurosurg. 2016; 143: 144–9. PubMed Abstract | Publisher Full Text\n\nFeijen IC, Rodrigues CMC, Cowie CJA, et al.: Dataset 1 in: Characteristics and management of ventricular shunt infections in children, 2000–2015: a single centre retrospective chart review. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15514.d211612"
}
|
[
{
"id": "39134",
"date": "15 Oct 2018",
"name": "Andrej Trampuz",
"expertise": [
"Reviewer Expertise Biofilm infections associated with implants and devices"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors analyzed 92 episodes of ventriculoperitoneal shunts in patients aged under 18 years at the time of infection. The article is well written, however, I have some concerns regarding the content and conclusions.\nPatient data were collected with the use of a standardised proforma, however, no definition criteria were presented. Simply citing that “A shunt infection was defined as any patient coded as ‘shunt infection’ at discharge or any patient who received treatment for suspected shunt infection, with or without microbiological confirmation” is not sufficient. There are different established definition criteria and at least one of them should be used and the frequency of each defining criteria should be presented.\nThe second main concern is that conclusions do not follow the results. As this was a retrospective observational study, no recommendations can be given regarding the antibiotic or surgical treatment. It is just a description of the treatment children received in this institution. In particular, I strongly disagree with recommending a combination of linezolid and ceftriaxone as first-line treatment. For the evaluation of the outcome, only one sentence was provided, which insufficient – how was the outcome evaluation performed, what was the definition of failure, what was the follow-up time etc.\nAnother point is that the author recommend complete removal of the shunt and placement of an EVD seems the safest surgical treatment. The question is, however, when the shunt can be retained an eradication of biofilm-associated infection with biofilm-active antibiotics can be achieved. Rifampin was used in about half of patients, but detailed data are lacking.\nThat it would be beneficial to develop a guideline for recognition and treatment of shunt infections and that empirical antibiotic treatment should be started as soon as possible is not a conclusion of this study and was already known from previous studies, so it should be deleted from the conclusion section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1158
|
https://f1000research.com/articles/7-1157/v1
|
30 Jul 18
|
{
"type": "Research Article",
"title": "Prevalence of aerobic pathogenic bacteria isolated from patients with burn infection and their antimicrobial susceptibility patterns in Al-Najaf City, Iraq- a three-year cross-sectional study.",
"authors": [
"Ahmed Abduljabbar Jaloob Aljanaby",
"Israa Abduljabbar Jaloob Aljanaby",
"Israa Abduljabbar Jaloob Aljanaby"
],
"abstract": "Background: Burn infections are one of the most common serious illnesses caused by pathogens, mainly by both gram-negative and gram-positive bacteria. The aim of this study was to detect of the prevalence of multi-drug resistant and extended-spectrum β-lactamase-producing (ESBL) bacteria isolated from inpatients with burn infection and the antimicrobials sensitivity patterns of all bacterial isolates during three years. Methods: This cross-sectional study was performed in Al-Najaf Central Hospital in Al-Najaf City, Iraq from January 2015 to December 2017. A total of 295 burns swabs were collected from hospitalized patients with burn infection. All grown bacterial isolates were identified by standardized microbiological tests. Antimicrobials susceptibility testing was done using the disc diffusion method. Multi-drug, extensive-drug and pan-drug resistant bacteria and extended-spectrum β-lactamase-producing bacteria were determined according to standardized methods and guidelines. Results: Of the 295 burn swabs, 513 different bacteria strains were isolated. Pseudomonas aeruginosa was the most common bacteria with 142 isolates (27.6%) followed by methicillin resistance Staphylococcus aureus 106 isolates (20.6%), while Staphylococcus typhi was the least common bacteria with only 17 isolates (3.3%). 323 (63%) different bacterial strains were isolated from patients who stayed in hospital for 15 days. Most bacterial isolates were resistant to most antimicrobials with high percentages. Out of the 513 bacterial isolates; only 33 isolates (6.4%) were resistant to imipenem 10µg and 464 isolates (90.4%) were multi-drug resistant, 20 isolates (14%) were extensive-drug resistant and 17 isolates (3.3%) were pan-drug resistant. Pseudomonas aeruginosa was the most common ESBL-producing bacteria (51 isolates-35.9%). Conclusions: There was a high prevalence of multi-drug resistant bacteria in burn infection in Al-Najaf hospital. Pseudomonas aeruginosa was the most common multi-drug resistant bacteria, and the most common of ESBL bacteria causing burn infection over the three years.",
"keywords": [
"Burn infection",
"Pathogenic bacteria",
"Antimicrobials susceptibility patterns",
"ESBL."
],
"content": "Introduction\n\nBurn infection caused by pathogenic bacteria is one of the most common hospital problems worldwide, particularly in developing countries1. Fire leads to skin destruction and simultaneous suppression of both humoral and cellular immune system subsequently resulting burn infection2. Complications of burn infection are responsible for more than 70% of death cases among inpatients with burns3. These infections mainly caused by multi-drug resistant gram-negative and gram-positive bacteria such as Pseudomonas aeruginosa (P. aeruginosa), Klebsiella pneumoniae (K.pneumoniae) and Staphylococcus aureus (S.aureus)4,5. Non-sterile burns halls and duration of patients stay in hospital in addition to the surface area of burned skin, are the most important factors related to the increase of persistent and multiplication of pathogenic bacteria in the burned areas6,7. Multi-drug resistant (MDR) bacteria is one of the most common pathogens causing burn infection in hospitalized patients worldwide8,9. These pathogens are resisting to at least three different classes of antimicrobials such as, penicillin’s, beta-lactams, cephems, 3rd and 4th generation cephalosporins, aminoglycosides, tetracyclines and quinolones, and is becoming one of the most dangerous health issues in hospitals10. In addition, extended-spectrum β-lactamase (ESBL)-producing bacteria are considered as a potent pathogens due to it their resistance to a wide range of antimicrobials like, cefotaxime, ceftriaxone and ceftazidime, that lead to difficulty in the treatment of most infections such as burn infection and urinary tract infection11,12. Burn infection is characterized by difficult healing due to administration of unsuitable treatment, long stays in hospital and the contaminates of hospital environments lead to the emergence of new multi-drug resistant bacterial isolates causing dangerous complications such as, bacteremia, septicemia and death13,14. Therefore, we must pay attention to all safety standards in hospitals, especially in burns wards through sterilization, performing antimicrobial susceptibility test on all pathogenic bacteria isolated from burn infections, and keeping the burned skin in sterile conditions to prevent the emergence of these pathogens. According to the above, the aim of this work was to investigate of the prevalence of multi-drug resistant bacteria and extended-spectrum β-lactamase-producing bacteria isolated from inpatients with burn infection in Al-Najaf central hospital in Al-Najaf City, Iraq over three years, from January 2015 to December 2017 to increase our understanding of the most prominent bacteria and their resistance to different antimicrobials to prevent the emergence of these isolates in the future.\n\n\nMethods\n\nWe confirm that we received approval for this study including: patient’s swabs and consent from the participants. All swabs were taken by physician and consent by the hospital treatment and care team responsible and then handed the samples over to us. All swabs were provided from the participants physician in Al-Najaf central hospital in Al-Najaf City, (Burns Department). All swabs were immediately transported to the Laboratory of Microbiology in Faculty of Science, University of Kufa to process. Note: Each swab was labeled with the following items: age, sex, duration of stay in the hospital after burning. Al Najaf Central Hospital is part of the University of Kufa and therefore written approval was not sought as there is a pre-existing agreement between the university and hospital regarding clinical sample collection sample collection. Oral consent to take swabs was taken from each patient.\n\nPatients will be considered eligible for registration into this study if they fulfill all the inclusion criteria and none of the exclusion criteria as defined below.\n\n1- Patients (Male or female) at least more than 18 years old.\n\n2- Patients should have sufficient capacity for informed consent.\n\n3- Patients should don’t have any other infections.\n\nThis is a cross-sectional descriptive study performed in Al-Najaf central hospital in Al-Najaf City, Iraq, from January 2015 to December 2017. A total of 295 swabs (emulsion with normal saline) were collected from the burned area of hospitalized patient with burn infections (2nd degree, shown the signs of infection during the change of dressings), ages ranges 18-45 years old (males and females), 3 swabs were taken from each patient at 5, 10 and 15 days of stay. Immediately, all collected swabs were incubated with brain heart infusion broth (Oxoid™, USA, CM1135R) for 24h at 37°C to encourage bacterial growth and then streaked onto blood agar (Oxoid™, USA, CM0055B) using a swab (Himedia, India, PW1210G) and chocolate agar (Oxoid™, USA, R01293) surface and incubated aerobically at 37°C for 24-48 h. All emerged bacterial isolates were identified according to colony morphology and standard microbiological tests such as; colony morphology, blood hemolysis onto blood agar surface (Oxoid™, USA, CM0055B), gram stain, oxidase test, catalase test, imvic test, motility test, coagulase test, growth on MacConkey agar (Oxoid™, USA, R061322) and Mannitol salt agar (Oxoid™, USA, CM0085B)15.\n\nAntimicrobials susceptibility testing was performed by disc diffusion method according to Kirby-Bauer method onto Mueller Hinton agar (Oxoid™, USA, PO5007A) surface16. Fourteen different antimicrobial discs were used in this study provide from OxoidTM, USA as follow: penicillin 10IU (P) (CT0043B), amoxicillin 25µg (AX) (CT0161B), Amoxiclav 30µg (AMC) (CT0538B), ceftriaxone 30µg (CRO) (CT0417B), cefotaxime 30µg (CTX) (CT0166B), ceftazidime 30µg (CAZ) (CT0412B), gentamicin 10µg (GM) (CT0024B), tobramycin 10µg (TM) (CT0056B), amikacin 30µg (NA) (CT0107B), ciprofloxacin 5µg (CIP) (CT0425B), imipenem 10µg (IMP) (CT0455B), Streptomycin 10µg (S) (CT0047B), erythromycin 30µg (E) (CT0021B), tetracycline 30 µg (TE) (CT0054B). The diameters of inhibition zones (mm) were measured using a caliper measure each zone with the unaided eye, and compared with clinical and laboratory standards institute (CLSI) guideline 201717. Any bacterial isolate was resistance to at least three different antimicrobials classes considered as MDR, if any bacterial isolate was resistance to all antimicrobial classes except two or three antimicrobial classes considered as extensive-drug resistant (XDR) and when any bacterial isolate was resistance to all antimicrobials class considered as pan-drug resistant (PDR)17.\n\nAll S.aureus isolates growth was adjusted according to turbidity of standard McFarland tube 0.5 (measured by Vis-Nir spectrophotometer, Biobase, UK, bk-S410). All isolates were streaked onto Mueller Hinton agar (Oxoid™, USA, PO5007A) surface supplemented with 4% NaCl. Five µg of methicillin disc (Oxoid™, USA, CT0159B) was placed at the surface of Mueller Hinton agar and incubated aerobically at 37°C for 24h. All S. aureus isolates that were resistant to methicillin with diameter of inhibition zone < 17 mm were considered as methicillin resistant S.aureus (MRSA), while those isolates with diameters of inhibition zone ≥ 17 mm considered methicillin sensitive S.aureus (MSSA)18.\n\nThis test was performed according to modified double disc synergy test (MDDST)19 as follows: all bacterial isolates (turbidity was adjusted according to McFarland tube 0.5) were streaked by sterile swab (Himedia, India) onto Mueller Hinton agar (OxoidTM, USA) surface, AMC disc 30µg was placed in the center of agar plate, CRO 30µg, CTX 30µg and CAZ 30µg were placed around AMC disc 30µg (15 mm from center to center). All plates were incubated aerobically at 37 °C for 24h. Any increase in the inhibition zone towards AMC disc 30µg was considered as positive for the extended spectrum beta-lactamase.\n\nPercentages were used in this study to compare between the prevalence of pathogenic bacteria and their resistant to antimicrobials using Graphpad-prism V.10 computer software.\n\n\nResults\n\nOf the 295 burn swabs, 513 different bacterial strains were isolated, 335 isolates (65.3%) were gram negative bacteria and 178 isolates (34.7%) were gram positive bacteria (Figure 1). Pseudomonas aeruginosa was one of the most common bacteria causing burn infection, 142 isolates (27.6%), followed by methicillin resistant S.aureus 106 isolates (20.6%), K. pneumoniae 93 isolates (18.2%), methicillin sensitive S.aureus 72 isolates (14.1%), E.coli 51 isolates (10%), A.baumannii 32 isolates (6.2%) and S.typhi 17 isolates (3.3%). 323 different bacterial strains (63%) were isolated from patients with burn infection who stayed in hospital for 15 days (Table 1). Out of total 513 bacterial isolates, 122 (23.8%) were isolated as single growth, while 391(76.2%) were isolated as mixed growth (Table 2). According to the results of antimicrobial susceptibility tests, most bacterial isolates were resistant to most antimicrobials with high percentages. Out of the 513 bacterial isolates, only 33 isolates (6.40%) were resistant to imipenem 10µg. The results of antimicrobials susceptibility test and overall resistant of 513 bacterial strains to 14 antimicrobials are shown in Table 3 and Figure 2. Of the total 513 bacterial isolates, 464 isolates (90.4%) were MDR, 20 isolates (14%) were XDR and 17 isolates (3.3%) were PDR (Table 4). Pseudomonas aeruginosa was the most common MDR-bacteria, 130 strains (91.5%), while 4 strains (2.8%) were XDR and 8 strains (5.6%) were PDR. All resistant types of all bacterial isolates are shown in Table 4. According to MDDST, Pseudomonas aeruginosa was the most common ESBL-producing bacteria 51 isolates (35.9%) (Figure 3) followed by K.pneumoniae 21 isolates (2.6%), while, no strain of S.typhi was ESBL. All ESBL- producing gram negative bacteria are shown in Figure 4.\n\nN=513.\n\nN=513.\n\nData were presented as numbers (NO.) and percentages (%) of bacterial isolates. MRSA: Methicillin resistance S.aureus, MSSA: Methicillin sensitive S.aureus.\n\nN=513.\n\nData were presented as numbers (No.) and percentages (%) of bacterial isolates.\n\nData were presented as numbers (No.) and percentages (100%) of pathogenic bacteria that were resistant to antimicrobials. AB: Antimicrobials, P: Penicillin 10IU, AX: amoxicillin 25µg, AMC: Amoxiclav 30µg, CRO: Ceftriaxone 30µg, CTX: Cefotaxime 30µg, CAZ: Ceftazidime 30µg, GM: Gentamicin 10µg, TM: Tobramycin 10µg, NA: Amikacin 30µg, CIP: Ciprofloxacin 5µg, IMP: Imipenem 10µg, S: Streptomycin 10µg, E: Erythromycin 30µg, TE: Tetracycline 30µg, MRSA: Methicillin resistance S.aureus, MSSA: Methicillin sensitive S.aureus.\n\nN=513.\n\nN=513.\n\nData were presented as numbers (No.) and percentages (%) of pathogenic bacteria. MDR: multi-drug resistant, XDR: extensive-drug resistant, PDR: pan-drug resistant, MRSA: Methicillin resistance S.aureus, MSSA: Methicillin sensitive S.aureus.\n\nESBL: extended spectrum beta-lactamase-producing bacteria.\n\n\nDiscussion\n\nBurn infection is one of the most serious problems in hospitals caused by different pathogens that infect most patients who stay in hospitals for prolonged periods. In this study, 513 different bacterial strains were isolated from 295 swabs of hospitalized patients with burn infections over the three years. Gram negative bacteria were responsible for more than half of infections while gram positive bacteria accounted for 34.7% of overall bacterial isolates. Pseudomonas aeruginosa was the most common bacteria, accounting for 27.6% of total isolates. These results are in agreement with previous studies20–23. Pseudomonas aeruginosa is one of the most important pathogens causing different infections such as bacteremia and burn infections24. This pathogen is well adapted to the hospital environments due to biofilm formation that provides long survival advantages for the pathogen, and effectively prevent eradication by the host immune system or antimicrobial drug treatment25. Pseudomonas aeruginosa has become responsible for more than 70% of mortality in burn patients26,27 . The results of this study showed that MRSA was the second most common bacteria isolated from patients with burn infection, 106 isolates (20.6%) of total bacterial isolates, while, MSSA was the most 4th common pathogen with 72 isolates (14.1%). These results are similar with many previous studies28,29. Staphylococcus aureus is the most common bacteria causing both hospital and community associated infections, including bacteremia, pneumonia and burn infection30. Hospital-associated MRSA accounts for a high proportion of hospitalized infected with S.aureus31. As compare with different bacteria that cause burn infection in hospitalized patients, MRSA-infections is associated with higher morbidity and mortality32. Klebsiella pneumoniae was the most 3rd common bacteria isolated from burn infection in this study with, 93 isolates (18.2%). This result is similar to past studies33,34. Klebsiella pneumoniae is an opportunistic pathogen which causes serious infections like, urinary tract infection, pneumonia, burn infection, and soft tissue infections in compromised and hospitalized patients. It has number of virulence factors such as a capsule that enable this pathogen to colonize and provides phagocytosis resistance35,36. The results of this study showed that Escherichia coli, A.baumannii and S.typhi were prevalent in different percentages, 10%, 6.2% 3.3%, respectively. Our results are similar with some previous studies37–39. On the other hand, the results of the current study proved that there was a positive relationship between a longer stay in hospital and the high prevalence of pathogenic bacteria causing burn infections. Contaminated burning wards and duration of patients stay in hospital, in addition to the size of surface area of burned skin are the most important reasons to increase of persistent and multiplication of pathogenic bacteria in the burned areas40. There are a number of factors that influence the emergence of infection in burns patient including; prolonged hospital stays, contamination of burns wards, nature of burn injury itself, as well as intensive diagnostic and therapeutic procedures41,42. Some studies suggested that burn infection is the most common type of infection, while, others studies reports show it to be bacteremia and pneumonia43,44. In this study, most pathogenic bacteria isolated from burn infection were highly resistant to most antimicrobials, especially against beta-lactams and 3rd generation cephalosporins., All pathogenic bacteria were MDR with high percentages and most of them were XDR,. P. aeruginosa was the most common PDR- bacteria followed by MRSA and A.baumannii. These results are similar with many previous studies5,45–47. Biofilm formation by microorganisms is one of the most important mechanisms in antimicrobials resistant, consisting of the irreversible assemblage of bacterial cells associated with a surface and enclosed in matrix of polysaccharides material48. Biofilms are regarding as a major factor contribution to many chronic inflammatory diseases such as burn infection due to enabling bacteria to colonize the burned skin, altering growth rate and allowing genes to be transcribed that provide these pathogens to high resistance to antimicrobials and host immune system. The overuse and unsuitability of different antimicrobials to treat burn infections has led to the emergence of new MDR, XDR and PDR-bacterial strains that are able to resistant a wide range of many antimicrobials such as aminoglycosides, beta-lactams, cephalosporins, streptomycin and tetracycline49,50. Burn infection in hospitalized patients caused by MDR, XDR and PDR-gram negative and gram positive bacteria such as; P. aeruginosa, K. pneumoniae, MRSA, MSSA and A.baumannii may lead to delays in burn healing, graft lose, as well as development of sepsis and death; therefore, determination of the risk factors for these pathogens infections is essential for infection control47. The results of this study showed that P. aeruginosa and K. pneumoniae were the most common ESBL-producing gram negative bacteria followed by A.baumannii and E.coli while there was no any strain of ESBL- S.typhi. These results are similar to previous studies1,51,52. Infections caused by ESBL-producing gram negative bacteria are associated with an increase of health care costs, morbidity and mortality53,54. Extended spectrum beta-lactamases (ESBLs) have been reported as one of the most important hospital-acquired infections such as burn infection and bacteremia9,55. Most bacteria harboring ESBLs are usually resistant to beta-lactam antibiotics and other classes of antimicrobials56. These enzymes are carried in and transferred from bacteria to bacteria by plasmids57,58. The most important steps to ensure the safety of patients with burn infections are: to control the spread of ESBL-producing bacteria, isolation of colonized patients in sterile wards, and continuously performing antimicrobial sensitivity tests59.\n\n\nConclusions\n\nThere was a high incidence of MDR-bacteria causing burn infections in Al-Najaf hospital in Al-Najaf City, Iraq. Pseudomonas aeruginosa was the most common MDR, XDR and PDR-bacteria, and the most common of ESBL-producing bacteria causing burn infection over three years followed by MRSA. Imipenem 10µg had good antibacterial activity against more than 93% of bacterial isolates. There was positive correlation between a long stay in hospital and high prevalence of pathogenic bacteria causing burn infection.\n\n\nLimitation of the study\n\nIn this study, some gram negative and gram positive bacterial isolates are excluded because of the small number of isolates (less than seven isolates over the three years) such as proteus spp (5 isolates), citrobacter spp (4 isolates), enterobacter spp (6 isolates) and enterococcus spp (5 isolates). We think this small number of bacterial isolates don’t has any significant effect on the results of this study.\n\n\nData availability\n\nDataset 1: Test results for patient swabs 10.5256/f1000research.15088.d21154160",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgement\n\nThe authors are very thankful to all staff of laboratory of Al-Najaf central hospital in Al-Najaf City for provided all swabs during three years.\n\n\nReferences\n\nAljanaby AAJ, Alhasnawi HMRJ: Phenotypic and Molecular Characterization of Multidrug Resistant Klebsiella pneumoniae Isolated from Different Clinical Sources in Al-Najaf Province-Iraq. Pak J Biol Sci. 2017; 20(5): 217–232. PubMed Abstract | Publisher Full Text\n\nSabzghabaee AM, Abedi D, Fazeli H, et al.: Antimicrobial resistance pattern of bacterial isolates from burn wounds in an Iranian University Hospital. J Res Pharm Pract. 2012; 1(1): 30–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEkrami A, Kalantar E: Bacterial infections in burn patients at a burn hospital in Iran. Indian J Med Res. 2007; 126(6): 541–4. PubMed Abstract\n\nChurch D, Elsayed S, Reid O, et al.: Burn wound infections. Clin Microbiol Rev. 2006; 19(2): 403–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nForson OA, Ayanka E, Olu-Taiwo M, et al.: Bacterial infections in burn wound patients at a tertiary teaching hospital in Accra, Ghana. Ann Burns Fire Disasters. 2017; 30(2): 116–120. PubMed Abstract | Free Full Text\n\nAgnihotri N, Gupta V, Joshi RM: Aerobic bacterial isolates from burn wound infections and their antibiograms--a five-year study. Burns. 2004; 30(3): 241–3. PubMed Abstract | Publisher Full Text\n\nWardhana A, Djan R, Halim Z: Bacterial and antimicrobial susceptibility profile and the prevalence of sepsis among burn patients at the burn unit of Cipto Mangunkusumo Hospital. Ann Burns Fire Disasters. 2017; 30(2): 107–115. PubMed Abstract | Free Full Text\n\nvan Langeveld I, Gagnon RC, Conrad PF, et al.: Multiple-Drug Resistance in Burn Patients: A Retrospective Study on the Impact of Antibiotic Resistance on Survival and Length of Stay. J Burn Care Res. 2017; 38(2): 99–105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAljanaby AAJ: In vitro antibacterial activity of an aqueous extracts of Matricaria chmomilla flowers against pathogenic bacteria isolated from pregnant women with urinary tract infection. Biomed Res. 2018; 29(11): 2395–2400. Reference Source\n\nAljanaby AAJ: Antibacterial activity of an aqueous extracts of Alkanna tinctoria roots against drug resistant aerobic pathogenic bacteria isolated from patients with burns infections. ROMJ. 2018; 7(1): 1–6. Publisher Full Text\n\nBennett JW, Robertson JL, Hospenthal DR, et al.: Impact of extended spectrum beta-lactamase producing Klebsiella pneumoniae infections in severely burned patients. J Am Coll Surg. 2010; 211(3): 391–9. PubMed Abstract | Publisher Full Text\n\nWalker KJ, Lee YR, Klar AR: Clinical Outcomes of Extended-Spectrum Beta-Lactamase-Producing Enterobacteriaceae Infections with Susceptibilities among Levofloxacin, Cefepime, and Carbapenems. Can J Infect Dis Med Microbiol. 2018; 2018: 3747521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLachiewicz AM, Hauck CG, Weber DJ, et al.: Bacterial Infections After Burn Injuries: Impact of Multidrug Resistance. Clin Infect Dis. 2017; 65(12): 2130–2136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlik J, Kawecki M, Gazdzik T, et al.: The impact of the types of microorganisms isolated from blood and wounds on the results of treatment in burn patients with sepsis. Pol Przegl Chir. 2012; 84(1): 6–16. PubMed Abstract | Publisher Full Text\n\nMac Faddin JF: Biochemical Tests for Identification of Medical Bacteria. 3rd edn., Williams and Wilkins, Philadelphia 2000; 912. Reference Source\n\nBauer AW, Kirby WM, Sherris JC, et al.: Antibiotic susceptibility testing by a standardized single disk method. Am J Clin Pathol. 1966; 45(4): 493–6. PubMed Abstract\n\nClinical and Laboratory Standards Institute (CLSI): Performance Standards for Antimicrobial Susceptibility Testing; 25ed. Informational Supplement. PA, USA. 2017. Reference Source\n\nFrench GL: Methods for screening for methicillin-resistant Staphylococcus aureus carriage. Clin Microbiol Infect. 2009; 15 Suppl 7: 10–6. PubMed Abstract | Publisher Full Text\n\nPaterson DL, Bonomo RA: Extended-spectrum beta-lactamases: a clinical update. Clin Microbiol Rev. 2005; 18(4): 657–86. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHodle AE, Richter KP, Thompson RM: Infection control practices in U.S. burn units. J Burn Care Res. 2006; 27(2): 142–151. PubMed Abstract | Publisher Full Text\n\nArmour AD, Shankowsky HA, Swanson T, et al.: The impact of nosocomially-acquired resistant Pseudomonas aeruginosa infection in a burn unit. J Trauma. 2007; 63(1): 164–171. PubMed Abstract | Publisher Full Text\n\nNichols DP, Caceres S, Caverly L, et al.: Effects of azithromycin in Pseudomonas aeruginosa burn wound infection. J Surg Res. 2013; 183(2): 767–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAljanaby AAJ, aljanaby IAJ: Profile of Antimicrobial Resistance of Aerobic Pathogenic Bacteria isolated from Different Clinical Infections in Al-Kufa Central Hospital–Iraq During period from 2015 to 2017. Research J Pharm and Tech. 2017; 10(10): 3264–3270. Publisher Full Text\n\nMaraolo AE, Cascella M, Corcione S, et al.: Response to: 'Letter to the Editor: \"Management of multidrug-resistant Pseudomonas aeruginosa in the Intensive Care Unit: state of the art\"'. Expert Rev Anti Infect Ther. 2018; 16(5): 369–371. PubMed Abstract | Publisher Full Text\n\nGroenewold MK, Massmig M, Hebecker S, et al.: A phosphatidic acid-binding protein is important for lipid homeostasis and adaptation to anaerobic biofilm conditions in Pseudomonas aeruginosa. Biochem J. 2018; 475(11): pii: BCJ20180257, 1885–1907. PubMed Abstract | Publisher Full Text\n\nMcManus AT, Mason AD Jr, McManus WF, et al.: Twenty-five year review of Pseudomonas aeruginosa bacteremia in a burn center. Eur J Clin Microbiol. 1985; 4(2): 219–223. PubMed Abstract | Publisher Full Text\n\nRoham M, Momeni M, Saberi M, et al.: Epidemiologic analysis of central vein catheter infection in burn patients. Iran J Microbiol. 2017; 9(5): 271–276. PubMed Abstract | Free Full Text\n\nAmissah NA, van Dam L, Ablordey A, et al.: Epidemiology of Staphylococcus aureus in a burn unit of a tertiary care center in Ghana. PLoS One. 2017; 12(7): e0181072. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu Y, Du FL, Liu PP, et al.: Molecular Epidemiology and Virulence Features of Staphylococcus aureus Bloodstream Isolates in a Regional Burn Center in China, 2012–2016. Microb Drug Resist. 2018. PubMed Abstract | Publisher Full Text\n\nLuzzaro F, Ortisi G, Larosa M, et al.: Prevalence and epidemiology of microbial pathogens causing bloodstream infections: results of the OASIS multicenter study. Diagn Microbiol Infect Dis. 2011; 69(4): 363–9. PubMed Abstract | Publisher Full Text\n\nStefani S, Chung DR, Lindsay JA, et al.: Meticillin-resistant Staphylococcus aureus (MRSA): global epidemiology and harmonisation of typing methods. Int J Antimicrob Agents. 2012; 39(4): 273–82. PubMed Abstract | Publisher Full Text\n\nCosgrove SE, Sakoulas G, Perencevich EN, et al.: Comparison of mortality associated with methicillin-resistant and methicillin-susceptible Staphylococcus aureus bacteremia: a meta-analysis. Clin Infect Dis. 2003; 36(1): 53–9. PubMed Abstract | Publisher Full Text\n\nEftekhar F, Naseh Z: Extended-spectrum β-lactamase and carbapenemase production among burn and non-burn clinical isolates of Klebsiella pneumoniae. Iran J Microbiol. 2015; 7(3): 144–9. PubMed Abstract | Free Full Text\n\nPerween N, Prakash SK, Siddiqui O: Multi Drug Resistant Klebsiella Isolates in Burn Patients: A Comparative Study. J Clin Diagn Res. 2015; 9(9): DC14–6. PubMed Abstract | Free Full Text\n\nHussein K, Sprecher H, Mashiach T, et al.: Carbapenem resistance among Klebsiella pneumoniae isolates: risk factors, molecular characteristics, and susceptibility patterns. Infect Control Hosp Epidemiol. 2009; 30(7): 666–71. PubMed Abstract | Publisher Full Text\n\nRiquelme SA, Ahn D, Prince A: Pseudomonas aeruginosa and Klebsiella pneumoniae Adaptation to Innate Immune Clearance Mechanisms in the Lung. J Innate Immun. 2018. PubMed Abstract | Publisher Full Text\n\nAsati S, Chaudhary U: Prevalence of biofilm producing aerobic bacterial isolates in burn wound infections at a tertiary care hospital in northern India. Ann Burns Fire Disasters. 2017; 30(1): 39–42. PubMed Abstract | Free Full Text\n\nHuang G, Peng Y, Yang Y, et al.: Multilocus sequence typing and molecular characterization of β-lactamase genes among Acinetobacter baumannii isolates in a burn center. Burns. 2017; 43(7): 1473–1478. PubMed Abstract | Publisher Full Text\n\nOli AN, Eze DE, Gugu TH, et al.: Multi-antibiotic resistant extended-spectrum beta-lactamase producing bacteria pose a challenge to the effective treatment of wound and skin infections. Pan Afr Med J. 2017; 27: 66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAL-Aali KY: Microbial Profile of Burn Wound Infections in Burn Patients, Taif, Saudi Arabia. Arch of clin microbial. 2016; 7(2:15): 1–9. Reference Source\n\nPeck MD, Weber J, McManus A, et al.: Surveillance of burn wound infections: a proposal for definitions. J Burn Care Rehabil. 1998; 19(5): 386–389. PubMed Abstract | Publisher Full Text\n\nCoban YK: Infection control in severely burned patients. World J Crit Care Med. 2012; 1(4): 94–101. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarner JS, Jarvis WR, Emori TG, et al.: CDC definitions for nosocomial infections, 1988. Am J Infect Control. 1988; 16(3): 128–140. PubMed Abstract | Publisher Full Text\n\nLari AR, Alaghehbandan R: Nosocomial infections in an Iranian burn care center. Burns. 2000; 26(8): 737–740. PubMed Abstract | Publisher Full Text\n\nAli U, Rehan M, Khan MS, et al.: Prevalence of different bacteria and their antibiotic susceptibilities in BCC Pims. JSM Burns Trauma. 2017; 2(3): 1021. Reference Source\n\nShuping LL, Kuonza L, Musekiwa A, et al.: Hospital-associated methicillin-resistant Staphylococcus aureus: A cross-sectional analysis of risk factors in South African tertiary public hospitals. PLoS One. 2017; 12(11): e0188216. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTekin R, Dal T, Bozkurt F, et al.: Risk factors for nosocomial burn wound infection caused by multidrug resistant Acinetobacter baumannii. J Burn Care Res. 2014; 35(1): e73–80. PubMed Abstract | Publisher Full Text\n\nWolcott RD, Rhoads DD, Dowd SE: Biofilms and chronic wound inflammation. J Wound Care. 2008; 17(8): 333–41. PubMed Abstract | Publisher Full Text\n\nAljanaby AAJJ: Antibacterial activity of an aqueous extract of Petroselinum crispum leaves against pathogenic bacteria isolated from patients with burns infections in Al-najaf Governorate, Iraq. Res Chem Intermed. 2013; 39(8): 3709–3714. Publisher Full Text\n\nAmissah NA, Buultjens AH, Ablordey A, et al.: Methicillin Resistant Staphylococcus aureus transmission in a ghanaian burn unit: The importance of active surveillance in resource-limited settings. Front Microbiol. 2017; 8: 1906. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeyestanaki DK, Mirsalehian A, Rezagholizadeh F, et al.: Determination of extended spectrum beta-lactamases, metallo-beta-lactamases and AmpC-beta-lactamases among carbapenem resistant Pseudomonas aeruginosa isolated from burn patients. Burns. 2014; 40(8): 1556–61. PubMed Abstract | Publisher Full Text\n\nOwlia P, Azimi L, Gholami A, et al.: ESBL- and MBL-mediated resistance in Acinetobacter baumannii: a global threat to burn patients. Infez Med. 2012; 20(3): 182–7. PubMed Abstract\n\nOverdevest IT, Willemsen I, Elberts S, et al.: Laboratory detection of extended-spectrum-beta-lactamase-producing Enterobacteriaceae: evaluation of two screening agar plates and two confirmation techniques. J Clin Microbiol. 2011; 49(2): 519–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAljanaby AAJ, Medhat AR: Prevalence of Some Antimicrobials Resistance Associated-genes in Salmonella typhi Isolated from Patients Infected with Typhoid Fever. J Biol Sci. 2017; 17(4): 171–184. Publisher Full Text\n\nAibinu IE, Ohaegbulam VC, Adenipekun EA, et al.: Extended-spectrum beta-lactamase enzymes in clinical isolates of Enterobacter species from Lagos, Nigeria. J Clin Microbiol. 2003; 41(5): 2197–200. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDaniel M, Paul M, Andrew N, et al.: Antimicrobial resistance patterns in extended-spectrum β- lactamase producing Escherichia coli and Klebsiella pneumoniae isolates in a private tertiary hospital, Kenya. Microbiology Discovery. 2013; 1(5): 1–4. Publisher Full Text\n\nAljanaby AAJ, Tuwaij NSS, Al-khilkhali HJB: Antimicrobial susceptibility patterns of Klebsiella pneumoniae isolated from older smokers and non-smokers of inpatients in intensive care unit infected with chronic pneumonia in AL-Najaf hospital, Iraq. J Pharm Sci & Res. 2018; 10(5): 1093–1097. Reference Source\n\nAljanaby AAJ: Antibiotics susceptibility pattern and virulence-associated genes in clinical and environment strains of Pseudomonas aeruginosa in Iraq. Asian J Sci Res. 2018; 11(3): 401–408. Publisher Full Text\n\nDelerue T, de Pontual L, Carbonnelle E, et al.: The potential role of microbiota for controlling the spread of extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBL-PE) in neonatal population [version 1; referees: 2 approved]. F1000Res. 2017; 6: 1217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAljanaby AAJ, Aljanaby IAJ: Dataset 1 in: Prevalence of aerobic pathogenic bacteria isolated from patients with burn infection and their antimicrobial susceptibility patterns in Al-Najaf City, Iraq- a three-year cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15088.d211541"
}
|
[
{
"id": "45345",
"date": "02 Jul 2019",
"name": "Vikrant Chitnis",
"expertise": [
"Reviewer Expertise Biomedical waste disposal",
"hospital infection control",
"bacteriology",
"molecular biology",
"disinfectants."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract needs rewriting with focus on role of hospital infection control committee (HICC), if existing in hospital. How HICC has helped or not or will help in near future in preventing such infections in burn unit. Name of organism in abstract: staphylococcus typhi? No such organism exists either it is staphylococcus aureus or salmonella typhi. There are many grammatical errors throughout the manuscript and very long sentences needs revision. Study is not useful if you do not involve hospital infection practices and corrective action. Is there an antibiotic policy in the hospital? If yes, are they drawing data to take care of antibiotic resistance? Are microbiologist active? Could be mentioned as corrective action in discussion or as it is suitable. Role of hospital infection control committee? What disinfectants were used? Why they did not go for anaerobic culture? Inclusion criteria of study needs redefining. Why patients above 60 were not chosen - justification needed, is it due to weak immunity or no such patient was admitted during study time?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "4716",
"date": "03 Jul 2019",
"name": "ahmed abduljabbar jaloob aljanaby",
"role": "Author Response",
"response": "Greetings We think the abstract does not need rewriting with focus on role of hospital infection control committee (HICC), because the our work focused on types of bacteria and antimicrobials. The name of organism in abstract: Staphylococcus typhi. Yes, this is a print mistake, we are so sorry the right name of organism in abstract must be Salmonella typhi. We can correct this mistake. There was an antibiotic policy in the hospital according to antimicrobial sensitivity test protocol. The role of hospital infection control committee is not in our study, because the only goal of our study is the type of aerobic bacteria and antimicrobials. About your question: what disinfectants were used in hospitals? We don't know, this is hospital work only and we can not know about these security things. About your question: why they did not go for anaerobic culture? This is another different study, our study focused on aerobic bacteria only. About your question: why patients above 60 were not chosen. Because we did not find patients above 60 in our study."
},
{
"c_id": "4743",
"date": "10 Jul 2019",
"name": "ahmed abduljabbar jaloob aljanaby",
"role": "Author Response",
"response": "Greetings, We think the abstract does not need rewriting with focus on role of hospital infection control committee (HICC), because our work focused on types of bacteria and antimicrobials. The name of organism in abstract: staphylococcus typhi. Yes, this is print mistake, we are so sorry. The write name of organism in abstract must be salmonella typhi. We can correct this mistake. There was an antibiotic policy in the hospital according to antimicrobial sensitivity test protocol. The role of hospital infection control committee is not in our study, because the only goal of our study is the type of aerobic bacteria and antimicrobials. About your question, what disinfectants were used in hospitals? We don't know, this is hospital work only and we can not know about these security things. About your question. Why they did not go for anaerobic culture? This is another different study, our study focused on aerobic bacteria only. About your question, Why patients above 60 were not chosen? Because we did not find patients above 60 in our study."
}
]
},
{
"id": "59146",
"date": "17 Feb 2020",
"name": "Dana M. Walsh",
"expertise": [
"Reviewer Expertise Toxicology",
"microbiology",
"metagenomics",
"bioinformatics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper provides an observational report of the number of drug resistant bacteria isolated from infected burn wounds in patients in a hospital in Iraq. Increasing numbers of drug resistant organisms are identified with increasing length of hospital stay. The most commonly identified organism was Pseudomonas aeruginosa, a bacteria that commonly identified in burn wound infections.\nOverall, the science in this article is sound. Appropriate MDR testing is used and compared to accepted standards.\nMethods section: It is unclear if an institutional review board approved this study; this needs to be directly mentioned in the methods section. No statistics were performed on this study; however, a simple Student's T test could be performed to determine statistical significance between two groups or analysis of variance (ANOVA) could be performed to show statistically significant differences between the number of isolates depending on length of hospital stay. This could also be done for number and type of drug resistant organisms to show which antibiotic has the least resistance. Including statistics would strengthen this work.\nDiscussion section: Why are biofilms mentioned at all? This seems out of place, especially since no testing was done to determine if these microbes are capable of making biofilms. This either needs to be removed or more appropriately addressed in the context of the study.\nConclusions section: This could be strengthened. For example, it seems that a majority of the organisms were not resistant to imipenem. This is an important finding for this hospital and should be mentioned as an antibiotic that could help this facility control MDR infections. It would be helpful if the authors addressed the following questions in this section: What new information does this study provide? Is this important for the hospital’s burn ward? Will this increase awareness of the prevalence of MDR organisms and encourage physicians to use imipenem for treatment of infections?\nOther comments: Metagenomic sequencing and biofilm testing would both be interesting additions to this study. Metagenomics could be used to determine what other bacteria were present in the infections that were not culturable.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1157
|
https://f1000research.com/articles/7-1155/v1
|
30 Jul 18
|
{
"type": "Research Article",
"title": "Characterization of Panton–Valentine leukocidin-positive Staphylococcus aureus from skin and soft tissue infections and wounds in Nigeria: a cross-sectional study",
"authors": [
"Olayemi O. Ayepola",
"Nurudeen A. Olasupo",
"Louis O. Egwari",
"Frieder Schaumburg",
"Nurudeen A. Olasupo",
"Louis O. Egwari",
"Frieder Schaumburg"
],
"abstract": "Background: Staphylococcus aureus is a significant pathogen implicated in numerous nosocomial and community-acquired infections. The Panton–Valentine leukocidin (PVL) can be associated with severe necrotizing diseases such as pneumonia, skin and soft tissue infection (SSTI). Methods: In total, 96 S. aureus isolates were obtained from patients presenting with wounds (n=48) and soft tissue infections (SSTIs, n=48). These were characterized based on their antimicrobial susceptibility profile, the possession of virulence genes (e.g. capsular type, PVL), accessory gene regulator (agr) type, and the staphylococcal protein A (spa) type. The production of the PVL protein was assessed by western blotting. Results: All isolates were susceptible to methicillin. The resistance was highest to penicillin (97.9%), followed by trimethoprim/sulfamethoxazole (85.4%) and tetracycline (10.4%). The PVL gene was found in 83.3% of isolates from SSTIs and in 79.2% of isolates from wound. Of these, 53 (68%) produced PVL as assessed by western blotting. The most prevalent spa type was the t084 (78.1%, n=75) and, majority of the isolates carried agr2 (82.3%, n=79). Conclusions: Prevalence of antibiotic resistant PVL-positive methicillin susceptible S. aureus strains has severe implications on PVL mediated infections.",
"keywords": [
"Staphylococcus aureus",
"PVL"
],
"content": "Introduction\n\nStaphylococcus aureus is an important human pathogen that causes significant hospital and community acquired infections1. S. aureus producing Panton-Valentine leukocidin (PVL) is linked to a broad array of necrotizing diseases such as pneumonia and skin and soft tissue infections (SSTIs)2. PVL is more frequently associated with community isolates3. PVL is a pore-forming toxin that can kill myeloid cells by forming channels in the plasma membrane, leading to loss of osmotic balance that ultimately lyses the cel4. Earlier reports have shown PVL to be one of the most important virulence determinants in S. aureus from sub Saharan Africa5. This study was conducted to investigate the presence of virulence genes including lukS-PV/lukF-PV, the production of the PVL protein and the antibiotic resistance in methicillin-susceptible S. aureus strains isolated from wounds and SSTIs between 2010 and 2011.\n\n\nMethods\n\nEthical approval for this study was obtained from the Ethics Committee of the Department of Biological Sciences, Covenant University, Ota, Ogun State, Nigeria (CUNG-2010-035). All participants signed a written informed consent before the commencement of the study.\n\nIn this study we made use of an already existing database which has been published6. The study was conducted in four health facilities in Ogun and Lagos States of Nigeria between June 2010 and May 2011. Samples were collected from patients presenting with SSTIs and wound infections. The isolation and identification of the isolates were done by culture and genotyping. A total of 96 S. aureus isolates were obtained from wounds (n=48) and SSTIs (n=48). The Vitek automated systems (bioMérieux, Marcy L’Étoile, France) was employed to determine the antibiotic susceptibility profile. The PVL gene (lukS-PV/lukF-PV), capsular polysaccharides (cap 5, cap 8), exfoliative toxins (eta, etb), the toxic shock syndrome toxin (tst) and the agr type were detected by PCR. All amplifications was done in a thermocycler (Bio-Rad, Munich, Germany). The cycling conditions and primers used are as earlier published. Detection of the lukS-PV/lukF-PV gene was carried out using primer sequences: luk-PV-1(5'-ATCATTAGGTAAAATGTCTGGACATGATCCA-3') and luk-PV-2 (5' GCATCAASTGTATTGGATAGCAAAAGC- 3')7. The negative control was S. aureus ATCC 49230 (MSSA) and the positive control was sta 635/636 (a PVL-positive CA-MRSA strain). Primers specific for the variable segment of the cap locus. Cap5-f: (5'-GAAAGTGAACGATTAGTAGAA-3') Cap5-r: (5'-GTACGAAGCGTTTTGATAGTT-3') Cap 8-f: (5'-GTGGGATTTTTGTAGCTTTT-3') Cap 8-r: (5'-CGCCTCGCTATATGAACTAT-3') was used for the capsular typing8. Sequences specific for exfoliative toxins; eta, etb and the toxic shock syndrome toxin; tst were detected by multiplex PCR9. The agr types of the S. aureus strains were determined by the multiplex PCR strategy10. Extracellular production of PVL by lukS-PV/lukF-PV –positive strains was evaluated by a Western blot using in-house antibodies raised in rabbits (anti-lukF-PV: 334 µg/ml, anti-lukS-PV: 900 µg/ml11. The nitrocellulose membrane (Schleicher & Schüll, Dassel, Germany) was first incubated with rabbit anti-lukS-PV/lukF-PV antibodies (in-house antibodies, 1:1000 in TBST) and later incubated with polyvalent goat alkaline-phosphatase-conjugated anti-rabbit antibodies (1:1000 in TBST, DAKO, Germany, D0487). The membranes were washed and the bands visualized using alkaline phosphatase color development substrate (BCIP/NBT, Thermo Fischer Scientific, 34042)11. The production of PVL was determined semi-quantatively in four categories: no PVL production; low PVL production, high and very high PVL production. The genetic diversity of all isolates was determined by the staphylococcal protein A (spa) typing12. The highly polymorphic region X of the protein A gene, which is composed of a variable number of 24-bp repeats, was amplified by PCR. spa types were determined with the Ridom StaphType software version 1.5 beta (Ridom GmbH, Würzburg, Germany). All statistical computations were performed in SPSS Version 25. Data is explored using relevant descriptive analysis alongside chi2 to measure any association between antibiotic resistance, virulence genes and lukS-PV/lukF-PV. P<0.05 is deemed to be statistically significant.\n\n\nResults and discussion\n\nWe analyzed the characteristics of the PVL-positive S. aureus isolates as well as the relationship between antibiotic resistance, virulence genes and PVL gene (Table 1). Antibiotic resistance was highest to penicillin (100% in SSTI isolates and 94% in wound isolates), followed by trimethoprim/sulfamethoxazole (84% in SSTI isolates and 83% in wound isolates) and tetracycline (8% in SSTI isolates and 10% in wound isolates (Table 1). This is consistent with an earlier study which showed similar resistance rates for penicillin (98%), trimethoprim/sulfamethoxazole (80%) and tetracycline (18%) in Nigeria6. All isolates were methicillin-susceptible. The lukS-PV/lukF-PV gene was detected in 83.3% (n=40) of SSTI isolates and 79.2% (n=38) of wound isolates. Reports from other African countries have shown high rates of PVL positive MSSA ranging from 17% to 74%5. For example, a study in an Algiers hospital reported a prevalence of 72% among clinical isolates13. A multi-center study reported that deep-seated SSTIs associated with the PVL gene resulted in more hospitalizations of patients and this led more often to incision and drainage14. A meta-analysis showed PVL to be consistently associated with SSTIs than invasive diseases15. In a study carried out in Gabon, PVL-positive isolates were found to occur more in SSTIs, and PVL was also associated with resistance to trimethoprim/sulfamethoxazole16\n\nNote: R=resistant, S=susceptible\n\nThe presence of the PVL gene does not necessarily guarantee that the protein will be expressed and, if it is, toxin levels could vary widely from strain to strain. The production of PVL (in contrast to the sole presence of lukS-PV/lukF-PV) was observed in 75% of lukS-PV/lukF-PV SSTI isolates and 60.5% of lukS-PV/lukF-PV wound isolates. In vitro variation in the production of PVL by different strains of S. aureus has been reported and this suggests important differences in transcriptional and/or translational control of gene expression17. In this study, the level of PVL produced by lukS-PV/lukF-PV positive S. aureus isolates varied from strain to strain (Figure 1). It was observed in that none of the PVL-positive strains harboured other toxin genes such as eta, etb and tst. Seven different spa types were identified (Table 1). The most prevalent spa type was t084 (78.1%, n=75). An earlier study revealed a significant association of the spa-CC 084 PVL-positive isolates with PVL-positive isolates6. Typing of the agr locus, which controls the expression of many S. aureus virulence factors, showed that most isolates (82.3%, n=79) possessed the agr2, while none carried agr3. Other studies have linked isolates carrying an agr4 allele to exfoliatin-related diseases and usually carry eta and/or etb18,19. These were absent in this study.\n\nIn conclusion, this study showed that many S. aureus isolates in Nigeria carry the PVL genes but few produced PVL in vitro. Antibiotic resistance combined with the presence of the PVL genes, has serious implications in the treatment of S. aureus infections. This study is limited by the few study locations. A larger study population is needed to provide a better understanding of the clones of S. aureus in Nigeria. The results is however significant for regional surveillance.\n\n\nData availability\n\nDataset 1: Results of Vitek assay, PCR results for virulence genes, agr typing and spa typing. 10.5256/f1000research.15484.d21182720\n\nDataset 2: Results of PCR experiments. Gel photo for amplification of lukS-pv and lukF-pv gene. 10.5256/f1000research.15484.d21182821\n\nDataset 3: Results of PCR experiments. Gel photo for amplification of agr group. 10.5256/f1000research.15484.d21182922\n\nThe results were previously presented at the 4th International Conference on Prevention & Infection Control (ICPIC 2017) Geneva, Switzerland. 20–23 June 2017. Antimicrobial Resistance and Infection Control 2017, 6(Suppl 3):52. DOI 10.1186/s13756-017-0201-4. Poster 261.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the European Molecular Biology Organization (EMBO) [ASTF 18–2011].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank Mr Bode Onile-Ere for assistance with the statistical analysis.\n\n\nReferences\n\nHolmes A, Ganner M, McGuane S, et al.: Staphylococcus aureus isolates carrying Panton-Valentine leucocidin genes in England and Wales: frequency, characterization, and association with clinical disease. J Clin Microbiol. 2005; 43(5): 2384–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTristan A, Bes M, Meugnier H, et al.: Global distribution of Panton-Valentine leukocidin--positive methicillin-resistant Staphylococcus aureus, 2006. Emerg Infect Dis. 2007; 13(4): 594–600. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBocchini CE, Hulten KG, Mason EO Jr, et al.: Panton-Valentine leukocidin genes are associated with enhanced inflammatory response and local disease in acute hematogenous Staphylococcus aureus osteomyelitis in children. Pediatrics. 2006; 117(2): 433–40. PubMed Abstract | Publisher Full Text\n\nYoong P, Torres VJ: The effects of Staphylococcus aureus leukotoxins on the host: cell lysis and beyond. Curr Opin Microbiol. 2013; 16(1): 63–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchaumburg F, Alabi AS, Peters G, et al.: New epidemiology of Staphylococcus aureus infection in Africa. Clin Microbiol Infect. 2014; 20(7): 589–596. PubMed Abstract | Publisher Full Text\n\nAyepola OO, Olasupo NA, Egwari LO, et al.: Molecular Characterization and Antimicrobial Susceptibility of Staphylococcus aureus Isolates from Clinical Infection and Asymptomatic Carriers in Southwest Nigeria. PLoS One. 2015; 10(9): e0137531. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLina G, Piémont Y, Godail-Gamot F, et al.: Involvement of Panton-Valentine leukocidin-producing Staphylococcus aureus in primary skin infections and pneumonia. Clin Infect Dis. 1999; 29(5): 1128–1132. PubMed Abstract | Publisher Full Text\n\nGoerke C, Esser S, Kümmel M, et al.: Staphylococcus aureus strain designation by agr and cap polymorphism typing and delineation of agr diversification by sequence analysis. Int J Med Microbiol. 2005; 295(2): 67–75. PubMed Abstract | Publisher Full Text\n\nBecker K, Friedrich AW, Lubritz G, et al.: Prevalence of genes encoding pyrogenic toxin superantigens and exfoliative toxins among strains of Staphylococcus aureus isolated from blood and nasal specimens. J Clin Microbiol. 2003; 41(4): 1434–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLina G, Boutite F, Tristan A, et al.: Bacterial competition for human nasal cavity colonization: role of Staphylococcal agr alleles. Appl Environ Microbiol. 2003; 69(1): 18–23, (accessed July 14, 2018). PubMed Abstract | Publisher Full Text | Free Full Text\n\nLöffler B, Hussain M, Grundmeier M, et al.: Staphylococcus aureus panton-valentine leukocidin is a very potent cytotoxic factor for human neutrophils. PLoS Pathog. 2010; 6(1): e1000715. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMellmann A, Friedrich AW, Rosenkötter N, et al.: Automated DNA sequence-based early warning system for the detection of methicillin-resistant Staphylococcus aureus outbreaks. PLoS Med. 2006; 3(3): e33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamdani-Bouguessa N, Bes M, Meugnier H, et al.: Detection of methicillin-resistant Staphylococcus aureus strains resistant to multiple antibiotics and carrying the Panton-Valentine leukocidin genes in an Algiers hospital. Antimicrob Agents Chemother. 2006; 50(3): 1083–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlabi A, Kazimoto T, Lebughe M, et al.: Management of superficial and deep-seated Staphylococcus aureus skin and soft tissue infections in sub-Saharan Africa: a post hoc analysis of the StaphNet cohort. Infection. 2018; 46(3): 395–404. PubMed Abstract | Publisher Full Text\n\nShallcross LJ, Fragaszy E, Johnson AM, et al.: The role of the Panton-Valentine leucocidin toxin in staphylococcal disease: a systematic review and meta-analysis. Lancet Infect Dis. 2013; 13(1): 43–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKraef C, Alabi AS, Peters G, et al.: Co-detection of Panton-Valentine leukocidin encoding genes and cotrimoxazole resistance in Staphylococcus aureus in Gabon: implications for HIV-patients’ care. Front Microbiol. 2015; 6: 60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamilton SM, Bryant AE, Carroll KC, et al.: In vitro production of panton-valentine leukocidin among strains of methicillin-resistant Staphylococcus aureus causing diverse infections. Clin Infect Dis. 2007; 45(12): 1550–8. PubMed Abstract | Publisher Full Text\n\nJarraud S, Lyon GJ, Figueiredo AM, et al.: Exfoliatin-producing strains define a fourth agr specificity group in Staphylococcus aureus. J Bacteriol. 2000; 182(22): 6517–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJarraud S, Mougel C, Thioulouse J, et al.: Relationships between Staphylococcus aureus genetic background, virulence factors, agr groups (alleles), and human disease. Infect Immun. 2002; 70(2): 631–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAyepola OO, Olasupo NA, Egwari LO, et al.: Dataset 1 in: Characterization of Panton–Valentine leukocidin-positive Staphylococcus aureus from skin and soft tissue infections and wounds in Nigeria: a cross-sectional study. F1000Research. 2018. Data Source\n\nAyepola OO, Olasupo NA, Egwari LO, et al.: Dataset 2 in: Characterization of Panton–Valentine leukocidin-positive Staphylococcus aureus from skin and soft tissue infections and wounds in Nigeria: a cross-sectional study. F1000Research. 2018. Data Source\n\nAyepola OO, Olasupo NA, Egwari LO, et al.: Dataset 3 in: Characterization of Panton–Valentine leukocidin-positive Staphylococcus aureus from skin and soft tissue infections and wounds in Nigeria: a cross-sectional study. F1000Research. 2018. Data Source"
}
|
[
{
"id": "37150",
"date": "13 Sep 2018",
"name": "Solayide Abosede Adesida",
"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\nThe PVL production did not appear to correlate with the presence of the gene. Therefore, the title should be modified to reflect the variability concerning the production of pvl, its detectable level, presence of the pvl genes and resistance rates.\n\nAbstract The authors investigated 96 S. aureus isolates were obtained from patients presenting with wounds and soft tissue infections in four health facilities in two States in Nigeria. Resistance to penicillin, trimethoprim/sulfamethoxazole was more than 80% and all isolates were susceptible to methicillin. The PVL gene was found in 83.3% and 79.2% of isolates from SSTIs and wound respectively. 53 (68%) produced PVL by western blotting. The most prevalent spa type was the t084 (78.1%, n=75) and, majority of the isolates carried agr2 (82.3%, n=79). The objective (s) of the study was not stated.\n\nIntroduction Change “cel” to “cell”\nMethods\nThere should be comma after the statement “In this study” Since the isolates have been described as part of a larger study, in my opinion, extensive presentation of the previous methods used for characterising the isolates is not required. The current emphasis should be on the core findings which have not been presented in the existing or published database (ref.6). I suggest, you present a table summarizing the characteristics of the 96 isolates as obtained in the database.\n\nResults and discussion Information on pvl production and the presence of the genes is limited for the region under surveillance. In my opinion, this aspect of the work is relatively novel. Therefore, kindly, reconcile your findings on level of pvl production and the presence of the genes. You may also correlate these with the antibiotic resistance and perhaps the other results in the database. However, there is a lot more reports on S. aureus from skin and soft skin infection (wounds) with reference to pvl genes/production than you have cited. Please, see Sharma-Kuinkel et al1, Zhang et al, 2018, www.nature.com/scientific reports, Hamilton et al. Clinical Infectious Diseases, 2007, 45 (2007). Other studies in Nigeria should also be considered to enrich the discussion. Overall, the discussion should be revised appropriately. However, I approve subject to the corrections suggested.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "38327",
"date": "27 Sep 2018",
"name": "Funmilola A. Ayeni",
"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 The authors characterised PVL phenotypically and also investigate the presence of gene coding for PVL production, antibiotic susceptibility and agr production. The study is interesting as it goes further from detection of PVL gene to semi quantify the product.\n\nIntroduction Line 3, remove first `and` or recast in ``such as pneumonia and skin and soft tissue infections`` Spa typing is not part of the stated objectives.\nMethod Characterization of isolates ``In this study we made use of an already existing database which has been published6.`` It should be clearly stated the data used from the referenced study above. The Method section should be rewritten so that all previously described method in Ayepola et al (2015) should not be rewritten in the present article but only referred to.\nResults and discussion It will be interesting to also expatiate on the proportion of isolates that are PVL positive molecularly and phenotypically i.e. the isolates with the genes and also producing PVL and those with the genes without PVL. Spa typing reported in this study should be discussed too in relation to its epidemiological significance.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1155
|
https://f1000research.com/articles/7-253/v1
|
01 Mar 18
|
{
"type": "Research Note",
"title": "Characterization and stability evaluation of nanoencapsulated epoxylignans",
"authors": [
"Yusnita Rifai",
"Radhia Riski",
"Gemini Alam",
"Magdalena Litaay",
"Latifah Rahman",
"Radhia Riski",
"Gemini Alam",
"Magdalena Litaay",
"Latifah Rahman"
],
"abstract": "3',6-dimethoxy-3'',4''-(methylenedioxy)-2,5-epoxylignan-4'-ol (DMEO), an epoxylignan isolated from Piper nigrum, has currently captured attention for its potential antitumor effect. However, low stability is limiting its therapeutic application. The application of nanocapsulation would be the main strategy for overcoming this problem. DMEO-loaded nanocapsules were prepared by an emulsion-diffusion method using Eudragit RL 100 (at concentrations of 1, 1.5 and 2%) and polyvinyl alcohol. As the polymer content increased, the encapsulation efficiency and mean particle size also increased. After 6 months of storage at 25°C (0% RH), no crystalline peaks were observed in the diffraction patterns of all nanocapsules, thereby suggested that the physical stability of nanoencapsulated DMEO was not affected by the concentration ratio of the polymer−stabilizer combinations.",
"keywords": [
"nanocapsules",
"Eudragit RL 100",
"Glioma",
"Piper nigrum"
],
"content": "Introduction\n\nThe hedgehog (Hh) pathway is required for the growth and proliferation of various cancers1. The signaling begins with the binding of Hh protein ligand to its membrane receptor Ptch, which represses the activity of Smoothened (Smo). Smo in turn promotes the expression of the GLI (Glioma-associated oncogene) family of transcription factors, leading to tumor development2. The direct association of GLI with a specific binding site (5’-GACCACCCA-3’) in the promoter region of the target gene has been reported to contribute to pancreatic and prostate cancer cells3. We previously reported Hedgehog/GLI inhibitors from various plants4–6. We also isolated five lignans ((8R*,8’R*)-9-hydroxy-3,4-dimethoxy-3’,4’-methylenedioxy-9,9’-epoxylignan, kusunokinin, haplomyrfolol, dihydroclusin) including a new epoxylignan dmeo from Piper nigrum using the immobilization of GLI-GST on carboxylic acid magnetic dynabeads. DMEO was confirmed to have Hh signaling inhibitory activity and to be selectively cytotoxic against PANC1. Meanwhile the synthetic epoxylignan of DMEO related compound was reported to inhibit the mRNA expression of protein patched homolog (Ptch) in human pancreatic cancer cells (PANC1) and thus is considered to be a prospective drug candidate to treat cancer related to the GLI signaling pathway. However, poor solubility remains the main limitation of DMEO7, hence making drug administration in vivo difficult. The nanoencapsulation of DMEO is one of the ways of overcoming the problem.\n\nNanoencapsulation techniques are particularly important to protect drugs from degradation in biological fluids and improve their penetration into cells. The techniques are also beneficial for hydrophobic molecules because the ultra-dispersed pharmaceutical dosage forms that nanoencapsulation provides allow rapid drug dissolution8. The nanoencapsulation of DMEO within a suitable polymer is considered to be a good way to ameliorate its poor solubility, because the polymer acts as a rate-controlling membrane to obtain the desired controlled release. The physical stability of pure compounds remains the greatest challenge for pharmaceutical scientists seeking to exploit higher solubility properties. A gold standard revealed by the International Council for Harmonization (ICH) stated that physical stability tests of compounds should be performed within accelerated (6 months) and/or long-term (12 months) storage conditions9. The physical stability was examined using powder X-ray diffraction (PXRD) analysis.\n\nThe objective of this study was to characterize the physical stability of DMEO-loaded nanocapsules, which were optimized by varying the polymer concentration to obtain stable spherical particles. The most stable particles would result from the best concentration ratio between polymers and stabilizers, ultimately improving the dissolution rate of poorly water soluble drugs.\n\n\nMethods\n\nEudragrit RL 100 and polyvinyl alcohol were purchased from Sigma-Aldrich Ltd. (St Louis, MO, USA). DMEO was obtained from the Pharmaceutical Chemistry Laboratory of Hasanuddin University (Indonesia). Methanol, ethyl acetate, acetonitrile, chlorophorm and demineralized water were purchased from Merck, Indonesia. All chemicals and solvents were of analytical or pharmaceutical grade.\n\nNanocapsules were prepared by an emulsion-diffusion method using Eudragit RL (ERL, Merck Ltd) at various concentrations (1%, 1.5% and 2%). Briefly, the ERL polymer (100, 150 and 200 mg) were dissolved respectively in 10 mL of ethyl acetate saturated with water. Each of this organic phase was then emulsified with 40 mL of aqueous phase, saturated with ethyl acetate, containing 300 mg of Polyvinyl alcohol (PVA) using a high speed homogenizer (ultra-turax T 25, Germany) at 1500 rpm for 60 minutes. Deionized water (150 mL) was then added to the emulsion to induce the diffusion of ethyl acetate into the continuous phase leading to the formation of nanocapsules. The organic solvent and the water phase were evaporated under reduced pressure to obtain a concentrated suspension of 40 mL.\n\nDMEO was previously synthesized and characterized as a white powder10. 10 mg of DMEO was dissolved in 10 mL of methanol, which was then added to polymerized nanocapsules. The DMEO-loaded nanocapsules were dried in a desiccator until constant weight. They were then kept in a closed glass vial and stored at 25°C.\n\nThe size of DMEO-loaded nanocapsules was analyzed by laser diffraction using a Partica LA-950 laser diffraction particle size analyzer (Horiba Ltd, Japan). Dried particles (5 mg) were dispersed in Miglylol 812 using a UP50H ultrasound processor (Hielscher, Germany) and analyzed in triplicate. The surface characteristics of the particles were observed by scanning electron microscopy (SEM) (Jeol, JSM-5600 LV, Japan). The yields of the particles were calculated by the sum of the weights of all components, discounting the content of water in the suspensions. After stirring the powders in acetonitrile for 90 minutes at room temperature followed by centrifugation and filtration (GVWP membrane, 0.45 µm, Millipore), the nanoencapsulation efficiency of DMEO was determined by UV-Vis spectroscopy (Shimadzu; 229 nm absorbance detector). The nanoencapsulation efficiency (%) of the powder was calculated from the correlation of the theoretical and experimental DMEO concentration. The onset of its crystallization was examined before and after 6 months of storage at 25°C, using an X-Ray Diffractometer (XRD-7000, SHIMADZU) with Cu K α radiation (λ = 0.154 nm). The diffraction patterns were analyzed using X’Pert Highscore Plus (version 2.2d).\n\n\nResults\n\nSEM analysis reveals that all DMEO-loaded nanocapsules are mostly spherical (Figure 1). As the polymer content increased, the encapsulation efficiency and mean particle size also increased, as seen in Table 1, suggesting that nanoencapsulation minimized deactivation of the drugs during the delivery process due to the protection from the polymer shell. This might ensure sufficient amounts of drug reaching the targeted areas. ERL is a copolymer of partial esters of acrylic acid containing low amounts of a quaternary ammonium group. ERL is a water-soluble polymer that plays a crucial role in controlling drug release because its water uptake characteristic contributes to the swelling of the polymers.\n\nSEM images of sprayed-dried powder in formulations (A) F1 (B) F2 and (C) F3.\n\nIn crystalline materials, atoms are periodically arranged, but in non-crystalline materials, atoms are randomly arranged. Figure 2 shows that the peaks of DMEO do not possess that periodicity, while the characteristic peak falls at 19.3° 2θ. After the physical mixtures were subjected to various concentrations of polymers and stabilizers, all the peaks were re-observed at 0 and 6 months of storage. No crystalline peaks were observed in the diffraction patterns, as shown in Figure 2a, 2b and 2c (upper, down), suggesting that the particles remain spherical throughout the duration of the stability study. The unchanged position of the largest peak of DMEO, which remains at 19.3° 2θ (Figure 2a, 2b and 2c (upper, down)), indicates that there is no change in the diffraction pattern of DMEO after 6 months of storage. It is important for the DMEO particles to remain stable during 6 months of storage to ensure that the resulting DMEO-loaded nanocapsules meet the minimal standard requirements of drug formulation stability.\n\n(A) X-ray diffraction patterns of DMEO-loaded nanocapsules of F1 after 0 and 6 months of storage (B) X-ray diffraction patterns of DMEO-loaded nanocapsules of F2 after 0 and 6 months of storage (C) X-ray diffraction patterns of DMEO-loaded nanocapsules of F3 after 0 and 6 months of storage.\n\n\nDiscussion\n\nThe polymer-based encapsulation may be beneficial regarding improved short-term physical stability. The nanoencapsulation of DMEO yielded entrapment efficiencies of 89.53±0.307, 89.46±0.211, 90.31±0.352% with particle sizes of 227.55±0.231, 253.96±0.012, and 255.09±0.455 nm, respectively. It is generally considered that higher molecular weight polymers have less entropy loss related to their fast motion, which results in a higher affinity to drug surfaces10, and they therefore perhaps have better steric formation. The charge along the polymer chains can build a strong double layer surrounding the particles, granting electrostatic stabilization and therefore smaller particle size11.\n\nThe results showed that the polymer concentration influenced the particle size and entrapment efficiency. The 1% polymer concentration formula (F1) fulfilled the requirement of stable nanocapsules, including spherical and uniform surface morphology. This formula exhibited the greatest percentage of nanocapsules’ weight (10.24%) but had the smallest particle size (227.55 nm). Meanwhile, F3 showed the greatest entrapment efficiency (90.31%). We chose F1, F2 and F3 for the further physical stability study using XRD to see whether the different concentrations of polymer interfered with the stability of DMEO in nanocapsules. Polymers were expected to adsorb to the surface of the nanocapsules, providing both electrostatic and steric repulsion among particles, therefore producing nanoparticles with decreased particle sizes and improved physical stability. The polymers may adsorb to the drug surfaces through several points along the polymer chains, enabling the loops and tails of the polymers to extend into the liquid medium, providing a steric effect.\n\nThe key finding in this work was that higher concentrations of polymer in the stirring solution during the production process yielded higher encapsulation efficiencies and smaller particle sizes. Tuning the polymer ratio was essential to obtaining spherical particles, without necessarily affecting the stability of the obtained nanocapsules.\n\n\nData availability\n\nDataset 1: Raw data for particle size, particle yields, nanoencapsulation efficiencies and X-ray diffraction pattern values. 10.5256/f1000research.13047.d19534812",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Ministry of Research, Technology and Higher Education of Indonesia through a grant from WCU-UNHAS 2016-2017.\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 are grateful to Dahlan Tahir from the Department of Physics, Faculty of Life Sciences for providing necessary research facilities.\n\n\nSupplementary material\n\nSupplementary File 1: Structure of 3’,6-dimethoxy-3’’,4’’-(methylenedioxy)-2,5-epoxylignan-4’-ol (DMEO). 1H and 13C NMR Data of DMEO\n\nClick here to access the data.\n\n\nReferences\n\nLauth M, Bergström A, Shimokawa T, et al.: Inhibition of GLI-mediated transcription and tumor cell growth by small-molecule antagonists. Proc Natl Acad Sci USA. 2007; 104(20): 8455–8460. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee Y, Kawagoe R, Sasai K, et al.: Loss of suppressor-of-fused function promotes tumorigenesis. Oncogene. 2007; 26(44): 6442–6447. PubMed Abstract | Publisher Full Text\n\nAltaba AR, Sánchez P, Dahmane N: Gli and Hedgehog in cancer: tumours, embryos and stem cells. Nat Rev Cancer. 2002; 2(5): 361–367. PubMed Abstract | Publisher Full Text\n\nRifai Y, Arai MA, Koyano T, et al.: Terpenoids and a Flavonoid Glycoside from Acacia pennata Leaves as Hedgehog/GLI-Mediated Transcriptional Inhibitors. J Nat Prod. 2010; 73(5): 995–997. PubMed Abstract | Publisher Full Text\n\nRifai Y, Arai MA, Sadhu SK, et al.: New Hedgehog/GLI signaling inhibitors from Excoecaria agallocha. Bioorg Med Chem Lett. 2011; 21(2): 718–722. PubMed Abstract | Publisher Full Text\n\nRifai Y, Arai MA, Koyano T, et al.: Acoschimperoside P, 2′-acetate: a Hedgehog signaling inhibitory constituent from Vallaris glabra. J Nat Med. 2011; 65(3–4): 629–632. PubMed Abstract | Publisher Full Text\n\nRifai Y, Tani HB, Nur M, et al.: Synthesis, Molecular Mechanism and Pharmacokinetic Studies of New Epoxy Lignan-Based Derivatives. Arch Pharm (Weinheim). 2016; 349(11): 848–852. PubMed Abstract | Publisher Full Text\n\nBrigger I, Dubernet C, Couvreur P: Nanoparticles in cancer therapy and diagnosis. Adv Drug Deliv Rev. 2002; 54(5): 631–651. PubMed Abstract | Publisher Full Text\n\nHuynh-Ba K, Zahn M: Understanding ICH Guidelines Applicable to Stability Testing. Handbook of Stability Testing in Pharmaceutical Development. 2009; 1st ed. New York: Springer. Publisher Full Text\n\nChoi JY, Park CH, Lee J: Effect of polymer molecular weight on nanocomminution of poorly soluble drug. Drug Deliv. 2008; 15(5): 347–353. PubMed Abstract | Publisher Full Text\n\nDuro R, Alvarez C, Martínez-Pacheco R, et al.: The adsorption of cellulose ethers in aqueous suspensions of pyrantel pamoate: effects on zeta potential and stability. Eur J Pharm Biopharm. 1998; 45(2): 181–188. PubMed Abstract | Publisher Full Text\n\nRifai Y, Riski R, Alam G, et al.: Dataset 1 in: Characterization and stability evaluation of nanoencapsulated epoxylignans. F1000Research. 2018. Data Source"
}
|
[
{
"id": "31917",
"date": "10 Apr 2018",
"name": "Rani Sauriasari",
"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 overall construct of the manuscript is generally concise. The authors highlight the higher concentrations of polymer in the stirring solution during the production process yielded higher encapsulation efficiencies and smaller particle sizes. However, it is better to add statistical analysis to show the correlation between concentration of polymer with encapsulation efficiency and particle size.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3852",
"date": "23 Jul 2018",
"name": "Yusnita Rifai",
"role": "Author Response",
"response": "Thanks for the valuable feedback. Statistical analysis is important to correctly interpret the data. However, because of the small number of sample formula and their repetitious treatment (n =3), we did not do any statistical analysis."
}
]
},
{
"id": "33856",
"date": "23 May 2018",
"name": "Heni Rachmawati",
"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 describes the strategy to improve the lack of physical characteristic of such natural active compound by developing the nanocarrier system based on polymer. The authors discussed the correlation between polymer concentration and the drug loading efficiency. The main objective as well as the concept is acceptable in term of scientific perspective. However, some lack important information in particular in the method session: how the drug-loaded nanocapsules was formed, the sample preparation on such evaluation e.g PXRD, and the detail on SEM analysis (magnification etc). The important data to conclude the successful approach are missing: solubility and/or dissolution rate. Also, the only PXRD data is not sufficient to conclude the stability issue of the drug-loaded nanocapsules. The description presented in the result section is not sufficient to explain the phenomenon found in this study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3853",
"date": "23 Jul 2018",
"name": "Yusnita Rifai",
"role": "Author Response",
"response": "Thanks for valuable scientific feedback. We add some more explanation in the method session of the revised manuscript associated with how the drug-loaded nanocapsules are formed, the sample preparation and the detail on SEM analysis (magnification at 600x). As for the preparation of nanocapsules, the DMEO can be encapsulated into nanocapsules at a maximum concentration of 1 mg/ml with the ratio maintained at the 1:40; 1:45; and 1:50 drug/polymer ratio.Meanwhile, the diffraction patterns of the DMEO-loaded nanocapsules exhibit a diffraction peak at 19.3° 2θ. These results indicate that DMEO-loaded nanocapsules are not present as a crystalline state, but are probably dissolved within the nanocapsule’s core. It is noted that the only PXRD data is not sufficient to conclude the stability issue of the drug-loaded nanocapsules, the PXRD data aimed to characterize the crystallinity of DMEO in the formulated nanocapsules."
}
]
}
] | 1
|
https://f1000research.com/articles/7-253
|
https://f1000research.com/articles/7-130/v1
|
31 Jan 18
|
{
"type": "Antibody Validation Article",
"title": "Validation of a yeast malate dehydrogenase 2 (Mdh2) antibody tested for use in western blots",
"authors": [
"Shiran Gabay-Maskit",
"Maya Schuldiner",
"Einat Zalckvar",
"Shiran Gabay-Maskit"
],
"abstract": "Malate dehydrogenases (Mdhs) reversibly convert malate to oxaloacetate and serve as important enzymes in several metabolic pathways. In the yeast Saccharomyces cerevisiae there are three Mdh isozymes, localized to different compartments in the cell. In order to identify specifically the Mdh2 isozyme, GenScript USA produced three different antibodies that we further tested by western blot. All three antibodies recognized the S. cerevisiae Mdh2 with different background and specificity properties. One of the antibodies had a relatively low background and high specificity and thus can be used for specific identification of Mdh2 in various experimental settings.",
"keywords": [
"Malate dehydrogenase",
"Saccharomyces cerevisiae",
"antibody",
"western blot"
],
"content": "Introduction\n\nMalate dehydrogenases (Mdhs) catalyse the interconversion of malate and oxaloacetate using NAD+ or NADH1. In the yeast Saccharomyces cerevisiae there are three known isozymes: Mdh1 is located in mitochondria, Mdh2 is mostly cytosolic, and Mdh3 is localized to peroxisomes. The residues of the three isozymes are between 43–50% identical2. It is thus very important for purposes of isolation and identification to utilize specific antibodies that will recognize only one specific isozyme. Using three Mdh2 peptides, which were specifically designed to unique regions in the Mdh2 protein, GenScript USA produced three different antibodies that should have high specificity for S. cerevisiae Mdh2 relative to Mdh1 and Mdh3. We then tested all three antibodies by western blotting and found one with specific binding. Due to its specificity, this antibody has the potential to also work in other experimental assays such as immunoprecipitation, immunohistochemistry and ELISA (Enzyme-Linked Immunosorbent Assay).\n\n\nMethods\n\nThree antibodies for S. cerevisae Mdh2 were custom produced for us by GenScript USA : 1. Purified antibody, Anti-peptide #1, item number: U1684BK300_2, LOT number: A416120074, catalogue number: SC1195. 2. Purified antibody, Anti-peptide #2, item number: U1684BK300_5, LOT number: A416120094, catalogue number: SC1195. 3. Purified antibody, Anti-peptide #3, item number: U1684BK300_8, LOT number: A416120072, catalogue number: SC1195.\n\nAll three antibodies were raised in New Zealand Rabbit and are polyclonal. The immunogen for all three antibodies is conjugated KLH Peptide. To find specific Mdh2 peptides, the three S. cerevisiae malate dehydrogenases (Mdh1, Mdh2 and Mdh3) were aligned and analysed by GenScript, using their Antigen Design Tool. From this analysis, three peptides (corresponding to the three antibodies) were selected: #1: CHPQSRNSGIERRIM; #2: CINIESGLTPRVNSM; #3: MPHSVTPSIEQDSLC. The cysteines in the N’ terminus (peptides number 1 and 2) or C’ terminus (peptide number 3) were added for KLH conjugation.\n\nYeast strains are all based on the BY4741 laboratory strain3. Manipulations were performed using a standard PEG/LiAC transformation protocol4. The GFP-Mdh1, GFP-Mdh2, GFP-Mdh3 and OE-mCherry-Mdh3 strains were picked from the SWAT N’ GFP or N’ mCherry yeast libraries that were recently prepared in our lab5. All Strains with a fluorescent tag, including the strains that were picked from the SWAT libraries were verified using PCR (one primer from within the endogenous Open Reading Frame (ORF) and one from within the tag sequence) as well as by fluorescent microscopy. Primers for creating strains with deletions or C’ tagging of genes (Δmdh1, Δmdh2, Δmdh3 and Mdh2-mCherry) were designed using the Primers-4-Yeast web tool6. All deletions were verified using primers from within the endogenous ORF. All primers are summarized in Table 1.\n\nYeast proteins were extracted by the NaOH protocol as previously described7 and resolved on polyacrylamide gels with the following modifications:\n\n1. Liquid cultures were grown over night at 30°C in glucose containing media, and then were diluted and incubated for an additional 4–6 hours to get to mid-log phase. Cells that were grown in a different carbon source than glucose were grown in synthetic media with 0.1% glucose (6.7 g/l yeast nitrogen base and 0.1% glucose) for the overnight incubation, then diluted in synthetic media containing galactose (6.7 g/l yeast nitrogen base and 2% galactose) or oleic acid (6.7 g/l yeast nitrogen base, 0.2% oleic acid and 0.1% Tween-80) and incubated for an additional 4–6 hours. Cells that were grown in glucose were grown in yeast extract peptone dextrose (YPD) rich medium (1% Yeast Extract, 2% peptone, 2% glucose) for the whole process.\n\n2. The extractions were incubated for 10 minutes at 70°C and 15 µl supernatant was loaded per lane of the gel.\n\nThe gels were transferred to nitrocellulose membrane blots, blocked for 1 hour in SEA block diluted in TBS-T (1:5) at room temperature (RT), and probed with primary rabbit/mouse antibodies against Mdh2, Histone H3, Actin and mCherry (Table 2) for 1 hour at RT. Final concentrations of the primary antibodies were: 1 µg/ml for anti-Mdh2 peptides #1 and #3, 0.5 µg/ml for anti-Mdh2 peptide #2, anti-mCherry and anti-Actin and 0.2 µg/ml for anti-Histone H3. The membranes were washed 3 times without incubation in TBS-T, followed by 3x3 minute washes in TBS-T. The membranes were then probed with a secondary goat anti-rabbit/mouse antibody conjugated to IRDye800 or to IRDye680 (Table 2) for 30 minutes at RT and washed as before. Final concentrations of the secondary antibodies were 0.1 µg/ml for IRDye 800CW Goat anti-Rabbit IgG and IRDye 680RD Goat anti-mouse IgG, and 0.2 µg/ml for Goat anti-mouse HRP conjugated. Membranes were scanned for infrared signal using the Odyssey Imaging System (Li-Cor). For detecting the anti-Actin primary antibody (Table 2), a goat anti-mouse HRP conjugated secondary antibody (Table 2) was used. The membrane was then washed with ECL reagents and scanned using imageQuant LAS 4000 system (GE Healthcare). Images were transferred to ImageJ 1.51s for slight adjustments of contrast and brightness. For further information about the western blot reagents see Table 3.\n\nSeveral controls were used in this study. Loading controls for the total amount of protein were performed using anti-Histone H3 or anti-Actin antibodies. A Δmdh2 strain, alongside Δmdh1 and Δmdh3 strains, were used to verify the specificity of the Mdh2 antibodies. An anti-mCherry antibody was used as a control for the bands detected in Mdh2-mCherry or Tef2-mCherry-Mdh3 (over expression) compared to the bands detected by the anti-Mdh2 antibodies. An anti-GFP antibody was used as a control for the bands detected in GFP-Mdh1, GFP-Mdh2 and GFP-Mdh3 with anti Mdh2 antibody peptide #3.\n\n\nResults\n\nFor first examination of the three antibodies, we grew control cells in three different carbon sources: glucose, galactose and oleic acid. Since transcription of S. cerevisiae Mdh2 is repressed by glucose8, we hypothesized that the antibodies will not detect Mdh2 in extracts of cells grown in glucose, but will detect it in the galactose and oleic acid samples. We then performed protein extraction using the NaOH approach, and examined the three GenScript anti-Mdh2 antibodies.\n\nIndeed, all three antibodies could not detect a specific band at the correct size of Mdh2 (41 KDa) in protein extractions from cells grown in glucose (Figure 1, Glu). Although the general protein levels of the control strain grown in oleic acid were much lower than from the levels of controls grown in glucose, a prominent band at the correct size for Mdh2 was detected by all three antibodies. In galactose there was no prominent band at the correct size, although this could be due to the general low levels of protein in this condition, as can be seen by the Histone H3 loading control. Endogenous Mdh2 was also recognized in protein extractions from cells grown on other carbon sources such as EtOH (not shown). This suggests that the upregulation of Mdh2 in cells grown in oleic acid is due to the removal of the glucose repression and not because of an oleic acid-specific up-regulation.\n\nWestern blot analysis was performed on protein extracts of control cells (BY4741) grown in glucose (Glu), oleic acid (Ole) or galactose (Gal) using antibodies 1–3 (Ab1, Ab2, Ab3). Native Mdh2 (41 KDa) is detected by all three anti-Mdh2 antibodies in protein extractions from cells grown in oleic acid as indicated by yellow arrows. Histone H3 was used as a loading control. In total we detected seven times a band at the size of Mdh2 when cells were grown on Oleate: One time using antibody1, three times using antibody 2 and three times using antibody 3.\n\nAfter the first comparison, we focused on antibodies 2 and 3 as they displayed the lowest background signal. We verified the affinity and specificity of the antibodies using several strains, in addition to the control strain: Δmdh1, Δmdh2, Δmdh3, as well as strains that had a mCherry tag fused to either Mdh2 or Mdh3 or GFP tag fused to Mdh1, Mdh2 or Mdh3. Both antibody 2 and antibody 3 detected endogenous Mdh2 in wild type cells and in Δmdh3 cells grown in oleic acid (Figure 2 and Figure 3), but were specific to Mdh2 as they did not cross react in the Δmdh2 strain. Another way to demonstrate specificity is to tag Mdh2 with mCherry, thus shifting only this specific isoform in size. Indeed, under these conditions both antibodies 2 and 3 detected a higher protein form only (Mdh2 tagged with mCherry = ∼70 kDa), though with lower affinity. Reciprocally, neither antibody detected over expressed mCherry-Mdh3 (expressed under a Tef2 promotor), although it was highly expressed as could be verified by an anti-mCherry antibody. In addition, antibody 3 detected endogenous Mdh2 in Δmdh1 cells grown in oleic acid, as well as a bigger band size only in GFP-Mdh2 tagged cells (Mdh2 tagged with GFP = ∼68 kDa) (Figure 4). Although both antibodies fit the requirement of specific identification of the yeast Mdh2, we have decided to use antibody 3, as it has a better specificity as seen from the lower background signal (Figure 3).\n\nWestern blot analysis was performed on protein extractions from different strains grown on oleic acid (Ole) or in glucose (only for OE-mCherry-Mdh3) using antibody 2 and anti-mCherry antibody. Antibody 2 recognizes the native Mdh2 in control cells and in Δmdh3 cells grown in oleic acid, but does not recognize it in Δmdh2 strain (second lane). The antibody also recognizes Mdh2 tagged with mCherry (∼70 kDa), but not the over-expressed and mCherry tagged Mdh3. The last one can be recognized in the control with the anti-mCherry. Histone H3 was used as a loading control. Yellow arrows indicate bands corresponding to Mdh2. (OE = over expression). We saw that antibody 2 specifically recognized Mdh2 and not Mdh3 in three experiments.\n\nWestern blot analysis was done on protein extractions from different strains grown on oleic acid (Ole) or glucose (Glu) (OE-mCherry-Mdh3 was grown in glucose) using antibody 3 and anti-mCherry. Similarly to antibody 2, antibody 3 does not recognize Mdh2 in Δmdh2 strain or in cells grown on glucose. This antibody has a good specificity to Mdh2, as seen by the low background signal. Actin was used as a loading control. Yellow arrows indicate bands corresponding to Mdh2. (OE = over expression). We saw that antibody 3 specifically recognizes Mdh2 in three experiments.\n\nWestern blot analysis was done on protein extractions from different strains grown on oleic acid (Ole) using antibody 3 and an anti-GFP antibody. As shown in Figure 3, this antibody does not recognize Mdh2 in Δmdh2 strain or in cells grown on glucose but does recognize Mdh2 in Δmdh3 cells grown on oleic acid. Here we show that antibody 3 also recognizes Mdh2 in Δmdh1 cells grown on oleic acid. Actin was used as a loading control. Yellow arrows indicate bands corresponding to Mdh2. We saw that antibody 3 specifically recognizes Mdh2 and not Mdh3 in two experiments and that antibody 3 does not recognize Mdh1 in one experiment.\n\n\nConclusion\n\nThree antibodies against endogenous S. cerevisiae Mdh2 were prepared by GenScript, using specific peptides of Mdh2 as the immunogens. The antibodies were checked by western blot and were shown to have the ability to detect endogenous Mdh2, when cells are grown under conditions in which Mdh2 is expressed. Although two of the three antibodies demonstrated the specificity qualities required from such an antibody, antibody 3 had the best specificity qualities, and the lowest background signal. We thus recommend antibody 3 as the best option to detect endogenous S. cerevisiae Mdh2. This antibody can be used in western blot, as shown in this manuscript, and should be tested for other uses (e.g. ELISA, FACS, IP).\n\n\nData availability\n\nDataset 1: raw images of all western blots included in figures presented 10.5256/f1000research.13396.d1908959",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a European Union ERC CoG #646604 and an SFB 1190 from the DFG. MS is an incumbent of the Dr. Gilbert Omenn and Martha Darling Professorial Chair in Molecular Genetics.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Pierce Feng from GenScript for help in antigen design and for rapid turnover time from order to delivery.\n\n\nReferences\n\nMcAlister-Henn L, Steffan JS, Minard KI, et al.: Expression and function of a mislocalized form of peroxisomal malate dehydrogenase (MDH3) in yeast. J Biol Chem. 1995; 270(36): 21220–21225. PubMed Abstract | Publisher Full Text\n\nSmall WC, McAlister-Henn L: Metabolic effects of altering redundant targeting signals for yeast mitochondrial malate dehydrogenase. Arch Biochem Biophys. 1997; 344(1): 53–60. PubMed Abstract | Publisher Full Text\n\nBrachmann CB, Davies A, Cost GJ, et al.: Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast. 1998; 14(2): 115–132. PubMed Abstract | Publisher Full Text\n\nGietz RD, Woods RA: Transformation of yeast by lithium acetate/single-stranded carrier DNA/polyethylene glycol method. Methods Enzymol. 2002; 350: 87–96. PubMed Abstract | Publisher Full Text\n\nYofe I, Weill U, Meurer M, et al.: One library to make them all: streamlining the creation of yeast libraries via a SWAp-Tag strategy. Nat Methods. 2016; 13(4): 371–378. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYofe I, Schuldiner M: Primers-4-Yeast: a comprehensive web tool for planning primers for Saccharomyces cerevisiae. Yeast. 2014; 31(2): 77–80 . PubMed Abstract | Publisher Full Text\n\nKushnirov VV: Rapid and reliable protein extraction from yeast. Yeast. 2000; 16(9): 857–860. PubMed Abstract | Publisher Full Text\n\nMinard KI, McAlister-Henn L: Glucose-induced degradation of the MDH2 isozyme of malate dehydrogenase in yeast. J Biol Chem. 1992; 267(24): 17458–17464. PubMed Abstract\n\nGabay-Maskit S, Schuldiner M, Zalckvar E: Dataset 1 in: Validation of a yeast malate dehydrogenase 2 (Mdh2) antibody tested for use in western blots. F1000Research. 2018. Data Source"
}
|
[
{
"id": "30437",
"date": "15 Feb 2018",
"name": "Andreas Hartig",
"expertise": [
"Reviewer Expertise Organelle biogenesis",
"glyoxylate cycle",
"yeast"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAs an antibody validation study the manuscript presented is complete, sufficient controls are included demonstrating the specificity of the antibodies for the Mdh2-protein with Ab3 sticking out as the best one because of low background. In figures 3 and 4 an unspecific band at the same size as the fusion proteins can be seen in many lanes, which is unfortunate raising doubts whether the Mdh2-mCherry or the GFP-Mdh2 are really specifically recognized. On the other hand, the native Mdh2-band at 41kD is missing in these lanes as expected increasing confidence into the specificity of the antibodies.\nSince Mdh2p is the isoform involved in the glyoxylate cycle it can be expected that this protein will be well expressed under conditions requiring this metabolic pathway, e.g. on ethanol, acetate or oleate. From the data presented in this manuscript it is not possible to conclude that “the upregulation of Mdh2 in cells grown on oleic acid is due to removal of the glucose repression and not because of an oleic acid-specific up-regulation”. For such a statement one would need to compare and present quantitative expression data from cells grown on non-repressing carbon sources such as ethanol and on oleate. It may well be that the expression of Mdh2 is derepressed on ethanol and further induced when grown on oleate. Since this is neither essential nor the primary subject of the antibody validation study I would suggest removing this sentence from the final version of the manuscript.\nI did have troubles in assessing the original data of the Westernblots, but this might be due to the old version of photoshop I’m using on my Computer.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nAre sufficient details of materials, methods and analysis provided to allow replication by others? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "31221",
"date": "06 Mar 2018",
"name": "Ewald H Hettema",
"expertise": [
"Reviewer Expertise Yeast molecular cell biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe title of this manuscript describes exactly its contents. The validation of a S. cerevisiae malate dehydrogenase 2, Mdh2, antibody for use in western blots.\n\nThe study is informative and complete and clearly shows that, of the three antisera tested, the third antiserum is of highest specificity. In total lysates, no cross reactivity is found to related proteins Mdh1 and Mdh3. Conditions of sample preparation and western blot analysis are well described and should ensure for reproducibility. The only thing that could further enhance reproducibility is information on how much protein or culture OD600 equivalents were loaded per lane.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nAre sufficient details of materials, methods and analysis provided to allow replication by others? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-130
|
https://f1000research.com/articles/7-1151/v1
|
30 Jul 18
|
{
"type": "Research Article",
"title": "Extraction, purification, and activity of protease from the leaves of Moringa oleifera",
"authors": [
"Swarnali Banik",
"Shrutidhara Biswas",
"Srabani Karmakar",
"Swarnali Banik",
"Shrutidhara Biswas"
],
"abstract": "Background: Proteases cleave proteins, thereby providing essential amino acids for protein synthesis, and degrade misfolded and damaged proteins to maintain homeostasis. Proteases also serve as signaling molecules, therapeutic agents and find wide applications in biotechnology and pharmaceutical industry. Plant-derived proteases are suitable for many biomedical applications due to their easy availability and activity over a wide range of pH, temperature, and substrates. Moringa oleifera Lam (Moringaceae) is a very common food plant with medicinal property and geographically distributed in tropical countries. Here, we isolate proteases from the leaves of Moringa oleifera and characterize its enzymatic activity. Methods: Proteases were isolated from the aqueous leaf extract of Moringa oleifera by ammonium sulfate precipitation and purified by ion exchange chromatography. Subsequently, the enzyme kinetics was determined using casein as a substrate and calibrated over different pH and temperature range for maximal activity. Results: We obtained purified fraction of the protease having a molecular weight of 51 kDa. We observed that for the maximal caseinolytic activity of the protease, a pH of 8 and temperature of 37ºC was found to be most effective. Conclusion: The plant-derived proteolytic enzymes are finding increasing clinical and industrial applications. We could extract, purify and characterize the enzymatic activity of proteases from the leaves of Moringa oleifera. Further molecular characterization, substrate specificity and activity of the extracted protease are required for determining its suitability as a proteolytic enzyme for various applications.",
"keywords": [
"Casein",
"Enzyme activity",
"Leaf extract",
"Moringa Oleifera",
"Plant-derived proteases",
"Protein purification."
],
"content": "Introduction\n\nAll organisms contain proteases that hydrolyze peptide bonds in order to maintain systemic homeostasis and for its normal growth and development1,2. Proteases derived from plants, animals and microbes find wide industrial applications including in the leather, food, brewery and pharmaceutical industry2–4 corresponding to approximately 60% of the total worldwide enzyme sales5.\n\nMoringa oleifera is one of the best known medicinal plants widely distributed in the tropical regions6. It contains a mixture of several hydrolytic enzymes, in which proteases are the key enzymes reported to show pharmacological activity7. We attempted to investigate the protease activity of aqueous extracts of Moringa oleifera leaf. Here, we have isolated and purified the protease from Moringa leaves and carried out enzyme kinetics study and find that the protease exhibited optimal caseinolytic activity in alkaline pH.\n\n\nMethods\n\nMature Moringa Oleifera leaves were collected from a plant located near TIU campus, Salt Lake Kolkata and crushed along with 20mM phosphate buffer (pH 7.5) and 0.1% tween 20 detergent and protease cocktail inhibitor followed by centrifugation with plastocraft table top refrigerated centrifuge machine (Rota 4RV/FM) at 10000 rpm for 10 mins at 4°C. The crude soup was mixed with 40% ammonium sulphate to obtain the protein precipitate, which was then dissolved in 20 mM tris buffer for further evaluation.\n\nThe total protein content of the solutions at different stages of protein purification was determined by Bradford methods8 using Sigma’s Bradford reagent (B6916). In this assay, a series of BSA standard solutions (0.1 – 1.2mg/ml) were used to prepare the standard curve. Bradford assay was performed by adding 1 mL of Bradford reagent to 20 μl of each standard solutions or unknown solution, and homogenized by using vortex mixer. The samples were incubated in dark conditions for 10 minutes and the absorbance was read at 595 nm.\n\nWe performed sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) using 12% resolving and 5% stacking gels for separating proteins. We followed the Laemmli’s method9 for gel electrophoresis. The samples were mixed with equal volume of gel loading buffer and heated at 95°C in dry heating bath for 2 mins. The electrophoresis process was run with 90 V for first 10 mins and then run at 150 V with Biorad mini protean gel electrophoresis system. After complete run the gel was stained with Coomassie Brilliant Blue. We have used protein marker (10kD to 250 kD) from GCC biotech (Pre-stained protein marker GCR-P4B) for determination of molecular weight. We imaged the gels in Biorad gel documentation system. Acrylamide, bis acrylamide, Tris and TEMED (T9281) are from Sigma Aldrich. Coomassie Brilliant Blue R250 (93473) and Ammonium per sulphate (28575) was from SRL (Sisco Research Laboratories).\n\nDialysis: The pellet dissolved in Tris buffer as obtained above was then dialyzed in 3.5cm/ml dialysis tubing (SIGMA Aldrich D6066 overnight in a magnetic stirrer by immersing the tubing in a buffer containing Tris (pH 8) and phenylmethylsulfonyl fluoride (PMSF) SRL, which was repeated thrice for complete exchange of buffer.\n\nDiethylaminoethyl (DEAE) cellulose ion exchange chromatography: The protein sample was loaded in the DEAE cellulose (SIGMA Aldrich 30477) column. Ion exchange column chromatography was carried out by using an assembly of Biorad’s Econo pump model EP-1, UV monitor and chart recorder from Atto, Japan and Biorad’s fraction collector model 2110. A gradient of 0.05 M to 0.5 M NaCl was used to elute the protein from the column. The gradient was run for 150 min with a flow rate of 1ml/min. Optical density (OD) of all the fractions were taken at 280 nm with Schimadzu 2401 UV Vis Spectrophotometer.\n\nSamples at different stages of purification were tested for albuminolytic property of protease by using BSA SIGMA as substrate. BSA digestion was performed at 37°C and pH 7.5 for 1 hour. Further, each of the samples were mixed with protein gel loading dye in 1:1 ratio and loaded in SDS PAGE and the gel was imaged with Biorad gel documentation system.\n\nIn this assay, β-casein was used as substrate. If protease digests casein, the amino acid tyrosine is liberated along with other peptide fragments. Folin’s reagent reacts with free tyrosine to generate a blue colored product, which is quantifiable and measured as an absorbance value on the Schimadzu UV 2401 spectrophotometer at 660 nm. A tyrosine standard calibration curve is constructed to determine the amount of tyrosine released after the proteolytic activity. A series of tyrosine standard solutions at different concentrations (5 – 50 μg/mL) were prepared from the 0.18mg/mL L-tyrosine stock solution with deionized water. L-tyrosine was purchased from Himedia, Fohlin’s reagent was obtained from SRL and β-casein from SIGMA.\n\nWe have assayed the protease activity in terms of caseinolytic activity with plant leaf extracts at different stages of purification (crude soup is the initial supernatant after homogenization and centrifugation, 40% ammonium soup is the phosphate dissolved pellet after 40% ammonium sulphate fractionation and pooled soup is the final collection of pure fractions came from DEAE cellulose column). All the three samples were dialysed to remove protease inhibitor and EDTA before the protease assay. The protease activity of pure protein was examined at different pH range 4–9 and temperature range 4–70°C.\n\nThe enzyme activity assay for protease was conducted with different concentrations of β-casein as substrate, at pH-8 in 37°C respective optimum conditions as determined with the previous experiments described above (optimum temperature and pH conditions). Here the substrate concentration (β-casein) varied in the range (0.81, 1.6, 2.4, 4.03, 5.2) mg/ml keeping the enzyme concentration fixed.\n\n\nResults\n\nMoringa oleifera leaves are reported to contain protease but there are no detailed studies on the purification and kinetic parameters of the enzyme. Here, we obtain partially purified protease from the aqueous extract of the leaves by ion exchange chromatography such that in anion exchange the proteins show a peak at 280 nm implying a positively charged protein.\n\nThe protein concentration from mature Moringa oleifera leaves at various stages of purification is shown in Table 1, which was purified by DEAE cellulose ion exchange column chromatography. The chromatogram for purification is shown in Figure 1A. The purified fractions were observed in 12% SDS PAGE (Figure 1B). The protein was of 51 kDa according to molecular weight markers.\n\nA. Chromatogram for the purification of protein from Moringa oleifera shows the elution time versus absorbance at 280 nm and the corresponding NaCl gradient profile (ranging from 0.04M to 0.25M) for maximal elution B. SDS PAGE of the crude extract and fractions after purification by DEAE cellulose ion exchange chromatography. Lane 1 shows extract after 40% ammonium sulphate precipitation, lane 2 shows the prestained molecular weight marker from GCC biotech marking 140, 100, 91, 71, 51, 25 and 10 kDa bands, lanes 3 to 8 represent fractions after column purification, lanes 9 and10 show the bands from crude leaf extract.\n\nResults from Figure 2 shows that both crude extract and 40% ammonium sulfate fractionated sample possesses protease activity and is able to produce fragments of BSA (lane 5, 6 and 9).\n\nLane 1 is BSA and trypsin, lane 2 is BSA + trypsin + PMSF, lane 3 is BSA, lane 4 is BSA +40% Ammonium Sulphate cut + PMSF, lane 5 is BSA + 40% Ammonium Sulphate cut, lane 6 is BSA + Pooled pure protein, lane 7 is BSA, lane 8 is prestained molecular weight marker from GCC biotech showing 140, 100, 91, 71, 51, 25 and 10 kDa bands and lane 9 is BSA + crude leaf extract.\n\nUV-vis absorption spectra of the pure protease were shown in Figure 3. A single peak at 280 nm can be observed for the pure protein.\n\nIn both crude extract and purified protein, protease activity was measured as described in methods. Reactions in different pH 4, 5, 6, 7, 8 and 9 were done (Figure 4). The results showed maximum activity at the pH 8.0. Therefore, the enzyme is an alkaline protease.\n\nProtease activity of the pooled pure fractions on β-casein degradation is plotted against different pH (4, 5, 6, 7 and 8) at 37ºC. Free tyrosine liberated due to β-casein degradation was monitored with Folin-Ciocalteau reagent at 660 nm and the corresponding amount was measured from the tyrosine standard curve.\n\nThe protease assay with β-casein as substrate was performed at a range of temperatures; 4°C, 25°C 37°C, 55°C and 70°C (Figure 5) according to the methods described above. The enzyme activity was found to be maximum at 37°C.\n\nProtease activity of the pooled pure fractions on β-casein degradation is plotted against different temperature (4, 25, 37, 55 and 70°C) at pH 8. Free tyrosine liberated due to β-casein degradation was measured as described earlier.\n\nSpecific activity of the protease was calculated by enzyme activity from the protease assay using β-casein as substrate and the total protein content of the protease solution. We can see a large increase in specific activity after the final purification (Figure 6).\n\nFree tyrosine liberated due to β-casein degradation was measured as described earlier. Enzyme activity present per amount of enzyme is calculated as specific activity of the protease.\n\nWe have seen increasing protease activity in the initial substrate concentration range and then saturation of protease activity above concentration of 4.03 mg/ml β-casein (Figure 7). The graph as a result follows conventional Michaelis Menten kinetics. We calculated KM and Vmax from the corresponding double reciprocal plot i.e. Lineweaver Burk plot as shown in the inset graph (Table 2). KM is 5.47 mg/ml and Vmax is 588.23 μM/min.\n\nKM and Vmax obtained from the double reciprocal plot are 5.47 mg/ml and 588.23 µM/min respectively.\n\n\nDiscussion\n\nOur study concludes that mature leaves from Moringa oleifera contains a protease with an approximate molecular weight of 51kD, with an optimum temperature of 37°C and optimum pH of 8.0 for its caseinolytic property. This is the first report of purification of a protease from Moringa oleifera to our knowledge. Further determination of molecular characterization, substrate specificity and activity of the protease are required to determine its suitability for industrial applications.\n\n\nData availability\n\nDataset 1: Enzyme kinetics data. Zip file containing underlying data of all enzyme activity assays with raw gel images 10.5256/f1000research.15642.d212249",
"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\nWe would like to acknowledge Mr Dipak Chandra Konar of Department of Chemistry, Bose Institute, for his help in using the protein purification set up, and spectrophotometric assays. Special thanks go to Prof. K. P Das, Bose Institute, for his support in this project.\n\n\nReferences\n\nLópez-Otín C, Overall CM: Protease degradomics: a new challenge for proteomics. Nat Rev Mol Cell Biol. 2002; 3(7): 509–519. PubMed Abstract | Publisher Full Text\n\nTurk B: Targeting proteases: successes, failures and future prospects. Nat Rev Drug Discov. 2006; 5(9): 785–799. PubMed Abstract | Publisher Full Text\n\nLi Q, Yi L, Marek P, et al.: Commercial proteases: present and future. FEBS Lett. 2013; 587(8): 1155–1163. PubMed Abstract | Publisher Full Text\n\nLópez-Otín C, Bond JS: Proteases: multifunctional enzymes in life and disease. J Biol Chem. 2008; 283(45): 30433–30437. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi S, Yang X, Yang S, et al.: Technology prospecting on enzymes: application, marketing and engineering. Comput Struct Biotechnol J. 2012; 2(3): e201209017. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStohs SJ, Hartman MJ: Review of the Safety and Efficacy of Moringa oleifera. Phytother Res. 2015; 29(6): 796–804. PubMed Abstract | Publisher Full Text\n\nSatish A, Sairam S, Ahmed F, et al.: Moringa oleifera Lam.: Protease activity against blood coagulation cascade. Pharmacognosy Res. 2012; 4(1): 44–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976; 72(1–2): 248–254. PubMed Abstract | Publisher Full Text\n\nLaemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970; 227(5259): 680–5. PubMed Abstract | Publisher Full Text\n\nBanik S, Biswas S, Karmakar S: Dataset 1 in: Extraction, purification, and activity of protease from the leaves of Moringa oleifera. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15642.d212249"
}
|
[
{
"id": "38180",
"date": "18 Sep 2018",
"name": "Rajat Banerjee",
"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 research article entitled “Extraction, purification, and activity of protease from the leaves of Moringa oleifera by Banik S, Biswas S and Karmakar S” explains the extraction, purification and activity of a protease that has been isolated from the leaves of the plant Moringa oleifera. The work is unique and can be of great applications since Moringa oleifera is reported to have various medicinal properties. This work may be recommended for indexing only after incorporation of a few changes listed as follows:\nMajor Revision:\nFigure 2 – Lane numbers are missing; Marker positioning wrong; data of BSA+POOLED+PMSF and BSA+CRUDE+PMSF missing. Try to include these data for a better clarification to the readers. Figure 4 – What happens after pH 8, is not reflected in the graph? Then how can this be concluded as optimum pH? A range of pH 4 – 12 at least, would be better for any such conclusion. Figure 5 – The data points are very scattered. More data points must be included in the graph for proper conclusion, especially between 25°C and 37°C and 37°C and 55°C. Figure 7 – This is very confusing. The graph does not look like a Michaelis – Menten graph. Number of data points must be increased. Km is outside the substrate concentration range taken in the graph. Km is usually expressed in units of μM or mM and not mg/ml. The Lineweaver-Burk plot is also erroneous. If the linear line is extrapolated backwards, it will give a negative y-intercept, i.e., a negative 1/Vmax value. The Vmax values obtained by the authors can neither be correlated to the Michaelis – Menten graph nor to the Lineweaver-Burk plot. Then how is this value obtained?\nMinor Revision:\nUnder the Methods section, in Preparation of crude extract and Purification of protein, the authors have used “Tris” buffer. It is usually written as Tris-HCl buffer. Moreover, under the two above mentioned sub-headings, somewhere the pH of the buffer is missing and elsewhere its concentration. Clear information should be provided. Under the Enzyme kinetics assay, it is written “keeping the enzyme concentration fixed”. The concentration used must be specified. Figure 1B – is a bit confusing. Marker should be preferably loaded into any of the side lanes. AS 40% pellet also has multiple bands, quite similar to the ladder. If possible, change the gel picture with proper loading arrangement and labelling. Figure 3 – The peak at 280 nm is quite blunt and the 260/280 ratio is close to 1, suggesting the presence of impurities. Purer fractions must be used. Figure 6 – There is no need to write “Moringa samples” in the X-axis title. Instead, it should be mentioned in the figure legend as “Samples from Moringa oleifera”. Why is this paper cited as a reference as well? Can this be done? Details about the datasheet is provided in the manuscript itself, after conclusion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "38435",
"date": "18 Sep 2018",
"name": "Sujit Roy",
"expertise": [
"Reviewer Expertise DNA damage repair mechanism in plants under abiotic stress"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 entitled ‘Extraction, purification, and activity of protease from the leaves of Moringa oleifera’ by Banik et al. appears to be an interesting topic related to purification of protease from an economically and medicinally important plant species Moringa oleifera.\nOverall, the manuscript is well composed, the background and objectives of the study has been mentioned clearly and also justified. The introduction part also appropriately reviews some of the relevant information. The results are sound and explained properly and supported with further explanation in the Discussion part. Therefore, in summary, I’m recommending acceptance and publication of this article.\n\nHowever, for future research interest, the introduction section could have been more extensive, indicating relevant information from recent past and current studies for getting meaningful insight into the background of this research as well as the lacunae which motivated to set the objectives for carrying out this study.\n\nThe methods part is quite sufficient. However, for the purification part, describing the isolation of the indicated protease activity from Moringa leaf extracts may be more extensive and apart from Coomassie blue staining, quality of purification may be assessed by silver staining procedure to compare the results and enrichment of purification after the column chromatographic techniques.\n\nFinally, a discussion section may be included to compare and discuss the findings in light of the related research. In concluding note, application and future research possibilities using the gained knowledge and information may be mentioned.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36602",
"date": "25 Sep 2018",
"name": "Somsubhra Nath",
"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 Banik et al. describes the isolation and further characterization of protease activity of partially purified protein fraction from the leaves of Moringa oleifera. The manuscript is well-written with detailed description of materials, methods and results. It, indeed, identified the protease activity with sufficient detailing. It has the potential for future investigation of biochemical characterization and functional relevance of this plant derived protease.\nHowever, there are a few concerns, as pointed below:\nIn Figure 2, it is apparent that the first lane is not a part of original figure. Adding a separate lane in an otherwise complete picture is not accepted. A repeat experiment with all the lanes is suggested.\n\nThe referencing requires to be more elaborate.\n\nThe discussion should describe potential application of the findings, based on literature review.\nAddressing these issues will strengthen the acceptability of the finding described in this manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1151
|
https://f1000research.com/articles/7-576/v1
|
14 May 18
|
{
"type": "Research Article",
"title": "Status of drowning in Nepal: A study of central police data",
"authors": [
"Bhagabati Sedain",
"Puspa Raj Pant",
"Puspa Raj Pant"
],
"abstract": "Background: Drowning is a serious and mostly preventable injury-related cause of death. Low-and-middle income countries represent 90% of total drowning deaths worldwide. There is lack of epidemiological studies of drowning in Nepal. The aim of this paper is to describe the status of drowning in Nepal. Methods: Cases of drowning, occurring between January 2013 and December 2015 were extracted from the Daily Incident Recording System of Nepal Police. Variables on age, sex of the deceased, types of water bodies, places, season when drowning occurred and activities of deceased were extracted and descriptive analysis was done. Results: A total of 1,507 drowning cases were recorded over a 3 year period. The rate of drowning was 1.9 per 100,000 (2.95 for males and 0.92 for females). Majority of drowning occurred among males (76%) and more than half were (53%) under 20 years of age. Mostly drowning occurred in rivers (natural water bodies). The findings provide strong indication that drowning occurs throughout the year in Nepal. Children were highly vulnerable to drowning. The magnitude of drowning was found to be lower than estimated by global burden of disease (GBD) study. Conclusion: The burden of drowning in Nepal is considerable, but mostly unknown to the public. Despite only having access to a limited data source, this study provides useful evidence that comprehensive research in Nepal is needed urgently.",
"keywords": [
"Drowning",
"Nepal",
"Public Health",
"Natural water bodies",
"Injury Prevention"
],
"content": "Introduction\n\nDrowning is gradually being recognized as a leading cause of death in the low-and-middle-income countries (LMIC) yet it remains a neglected problem in many countries in the absence of adequate data (Peden et al., 2008; Rahman et al., 2009). The World Health Organization (WHO) estimated 372,000 deaths every year occur due to drowning, making it the world’s third leading cause of fatal unintentional injuries (WHO Global Report on Drowning). About 91% of the drowning deaths occur in low- and middle- income countries (Peden et al., 2008). Similarly, 40% of the world's total drowning deaths occur in children below 15 years and most of these occurred in low- and middle-income countries; 29% in South-East Asia region alone (IHME, 2016). A study from Bangladesh found the fatal drowning rate to be as high as 15.8 per 100,000 (Rahman et al., 2017).\n\nA systematic review on epidemiology of drowning in LMICs has found that most studies on drowning were only from some countries in Asia: Bangladesh, China, India, Iran, Pakistan, Sri Lanka, Thailand and Vietnam (Tyler et al., 2017). Nepal being a mountainous country; drowning does not come instantly to mind when considering the main causes of deaths. In Nepal, the summer monsoon season accounts for 80% of the annual rainfall, and the winter monsoon accounts for the remaining 20% (DoHM, 2017). Flash floods, ditches, irrigation canals, and open wells are the chief hazards for drowning. Nepal has the poor coverage of vital event registration (Gautam, 2016). According to the GBD estimates, an average of 1,300 people died in 2016 from drowning in Nepal with a mortality rate of 4.0 (95%UI 3.2 – 5.6) per 100,000 population (IHME, 2016).\n\nDetailed community-based studies on drowning have not been conducted in Nepal; however, there are some small-scale hospital-based post-mortem reports which present the cases of drowning. Drowning was the cause of death for 5% of total reported autopsy cases in a study of external causes of death reported in the Autopsy Centre in Kathmandu (Lunetta et al., 2004; Sharma et al., 2006). Another hospital based study found drowning as a cause of death for 144 cases out of total 205 post-mortem cases in Kathmandu during 2005 to 2010. This study also found that 39% of the drowning victims were below 18 years of age and 24% below 11 years during 2005–2010 (Wasti et al., 2010). However, these studies only included autopsy cases referred to hospital by the police as 'unnatural deaths,' which included deaths caused by drowning (Sharma et al., 2006; Wasti et al., 2010).\n\nSimilarly, an analysis of media reports on drowning deaths conducted for April 2010 to April 2011 showed that over 200 cases of drowning deaths in Nepal were reported to media, with more than half of the total drowning deaths were children (Pant et al., 2011). This study describes the situation of drowning in Nepal utilizing media report as secondary source of data.\n\n\nMethods\n\nThis study utilised drowning cases reported to Nepal Police during January 2013 to December 2015. Family members or community people must report the incidents of drowning to the Daily Incidence Reporting Center of Nepal Police in each village development committee (VDC) level. Polices’ database records the facts and evidence pertinent to an incident.\n\nPolice records of drowning deaths is the only national source of drowning deaths recorded in Nepal. Drowning deaths are recorded in the form of event records (narrative) including the details of the place of the residence, age, date, sex, place of drowning, type of water body, intent (intentional or accidental drowning) and activity during drowning. The described variables included; location/ geographical region, water bodies, months of drowning were also extracted, entered and analysed using SPSS version 16.0.\n\nDescriptive study was performed in order to show the distribution of fatal drowning by age, gender, type of water body, months of the year, geographical location, and activity of the victim before the drowning incident. Possible core victim identifiers i.e. age, gender, date, activity and location of the incident and describing variables i.e. water bodies, months of drowning, and activity before drowning were also analysed and described. Information about the 'time', 'distance' from the victim’s home and the companion were not available in the database.\n\n\nEthics\n\nEthical approval was not required for this study, because it did not involve human participants. This study used secondary data registered by police. To access the data the request latter on behalf of Tribhuvan University, Padma Kanya Campus was submitted to the Crime Investigation Department. As per the request, the police department granted accesses to the police record. In this study, personal identifiers or confidential information of deceased were not disclosed.\n\n\nResults\n\nThis study identified 1,507 cases of drowning deaths in three years (2013–2015). Among these drowning cases highest deaths were in Terai (low land) 700 (46.4%). Drowning deaths were found to vary between geographical regions (Table 1). Most of the drowning deaths in Nepal occurred in the open water. The highest number of drowning deaths were observed in the Plains followed by the Hills.\n\nThe proportion of males that died from drowning was higher than females. Table 2 shows the sex differentials in drowning deaths. Younger people (<20 years) comprised over half of the drowning deaths, making up 52.7% of all drowning deaths. Similarly, age specific drowning deaths were higher for 10–19 years closely followed by 0–9 years and then declined as the age increases. The most common drowning age was 10–19 years for males and 0–9 years for females. Age reporting was missing for 27 cases.\n\n§Age was not recorded for 27 cases\n\nThis study also identified a seasonal pattern of drowning, which showed high incidents of drowning during monsoon season, with peaks in July, with the winters (December–January) being relatively low for drowning. The monsoon season (June –August) claimed about 43% of the total drownings and over 15% drownings occurred in the summer season i.e. during April-May (Figure 1).\n\nMore than 88% of the drowning deaths occurred in natural bodies of water. Rivers, ponds and lakes were most common places for drowning in Nepal (Table 3). A smaller proportion of drowning also occurred in man-made water containing bodies like canals, water-filled pits, safety tanks, water tanks, and wells.\n\nSwimming, bathing, and crossing the river were the mostly reported activities before drowning, and occurring mainly in the summer season. In Nepal, whether people go near the river, bathe by the bank or go into the water for swimming depends upon age, gender and location of the river. Activity before drowning for 181 (12%) cases were not mentioned at all (Table 4).\n\n¶Others were the cases where activity before drowning were not clearly mentioned\n\n\nDiscussions\n\nA systemic review on fatal river drownings showed that river drowning deaths are an issue in many regions and countries around the world (Peden et al., 2016; Rahman et al., 2009). However, there has been limited research that has explored drowning in locations other than beaches and swimming pools. Natural water bodies such as rivers, streams and lakes regularly account for large proportions of drowning deaths (Peden et al., 2016; Rahman et al., 2017). Drowning, along with other injuries is neglected in Nepal. Most of the drowning occur in the natural water bodies such as rivers, canals, streams, lakes. In the plains, the number of drowning deaths were high although it was proportional to the population size. Nature of water bodies and landscape are different in three (Mountain, Hill and Terai) geographical regions, same as that the risk factors vary by geographical regions. Similarly, more than half of the drowning deaths occur in Hilly and Mountain regions.\n\nA study by Pant and colleagues (Pant et al., 2011) has a number of limitations; even though that demonstrated a similar type of results reported in this paper. Nepal is mountainous country, rivers are the major natural water body and rivers accounted for the largest number of fatal drownings (80%) for the period between 2013 and 2015. This was unlike other studies from LMICs (Hyder et al., 2009; Rahman et al., 2009; Soko, 2012) where most of the drownings occurred in wells, water storage facilities or reservoirs, ditches and drains.\n\nThe monsoon season in Nepal takes place during the months of June–September when 80% of the yearly rainfall occurs throughout the country. During the months of April – June (summer), it is very warm during the day and more people use rivers, streams or other water bodies for swimming and bathing. Drownings occur while performing these activities. However, it was found drowning deaths occur throughout the year.\n\nIn the absence of household pipe water supplies and bridges, people are forced to be in contact with these unprotected water bodies every day. Data revealed that drowning often occurs as a result of the activities associated with daily life i.e. bathing, crossing river and fishing in the river. Surprisingly, no cases were reported to have drowned while fetching water which implies the need for more in-depth research.\n\nDrowning varies greatly with age (Pant et al., 2011; Quan & Cummings, 2003). This study also revealed that drowning is more common in children rather than older age groups. A similar finding has also been highlighted in several global (World Health Organization, 2014), regional (Quan & Cummings, 2003; World Health Organization, 2014) and country level studies (Pant et al., 2011; Rahman et al., 2017). Drowning due to accidental falls mostly occurred in 0–4 year old children. These findings were similar to the study by Royal Life Saving Society in Australia (RLSSA, 2016). This study also found that drowning claimed more lives of children below 18 years than adults.\n\nSimilarly, the proportion of drowning deaths were higher among males in comparison to females in all age groups. Nearly three quarters of the deaths from drowning occurred in males, which is similar to the findings of WHO’s Global Report on Drowning (Rahman et al., 2017; Soko, 2012; World Health Organization, 2014; World Health Organization, 2014). This might be due to increased exposure to water and riskier behaviour. Findings show that female drowning was common in early ages (less than 9 years) and for the male it was high during the age 10–19 years. Overall, males of all ages outnumbered females for drowning.\n\nThese patterns of drowning in terms of age, gender, places of drowning, activity before drowning and seasonal patterns are consistent with findings of studies elsewhere (IHME, 2016; Pant et al., 2011; Peden et al., 2016). Further study is needed to identify the tendency to exposure and age-related drowning by activity before drowning.\n\n\nStrengths\n\nThe strength of the study is that nationally recorded information from the Nepal Police was obtained. To our knowledge, this is the first study that attempted to study about drowning throughout the country which is able to present the patterns of drowning by district, age, gender and place of occurrence.\n\n\nLimitations\n\nInformation was limited to those drowning cases reported within 24 hours of occurrence through the daily incident reporting system of Nepal Police. This information doesn't take into account those who have gone missing into water. There is a lack of information on distance from home, time of the event, person accompanying the victims, and intent and influence of substances or illness responsible for the drowning incidents, which can be very helpful for designing drowning prevention interventions.\n\nThe data collected by the police was not be for the purpose of drowning prevention rather was for a forensic investigation or criminality. Some of the reported variables were incomplete (i.e. age of the person, location of drowning etc.). The data used in this manuscript has very limited information for every component of: a) victim information, b) scene information, c) any emergency medical services provided, and d) any Hospital care received by the victim.\n\n\nConclusion\n\nThe findings of this study suggest that drowning occurs in many parts of Nepal and not necessarily only in the plains; and children are highly vulnerable to drowning. Rivers were the most common place of drowning in Nepal and the rate of drowning increased in rainy season. There is a need for better understanding of people's contact with rivers by gender and age to inform prevention.\n\n\nData availability\n\nDataset 1: Data file containing drowning data gathered from police records 10.5256/f1000research.14563.d202804 (Sedain & Pant, 2018)",
"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\nThe authors would like to thank the Office In-charge of the Crime Investigation Department of Nepal Police who kindly provided access to the data.\n\n\nReferences\n\nDepartment of Hydrology and Meteorology (DoHM): Observed climate trend analysis of Nepal (1971–2014). Department of Hydrology and Meteorology. 2017. Reference Source\n\nGautam BR: Civil registration and vital statistics: policy and practices in Nepal. Department of Civil Registration, Babarmahal, Kathmandu, Nepal, 2016. Reference Source\n\nHyder AA, Sugerman DE, Puvanachandra P, et al.: Global childhood unintentional injury surveillance in four cities in developing countries: a pilot study. Bull World Health Organ. 2009; 87(5): 345–352. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIHME: Global Burden of Disease Study 2015. (GBD 2015) Results. 2016.\n\nLunetta P, Smith GS, Penttilä A, et al.: Unintentional drowning in Finland 1970-2000: a population-based study. Int J Epidemiol. 2004; 33(5): 1053–1063. PubMed Abstract | Publisher Full Text\n\nPant PR, Baset K, Towner E: Fatal drownings in Nepal: As reported in the media, 2011. Abstract World Conference on Drowning Prevention, 2010. International Lifesaving Federation. 2011. Reference Source\n\nPeden AE, Franklin RC, Leggat PA: Fatal river drowning: the identification of research gaps through a systematic literature review. Inj Prev. 2016; 22(3): 202–209. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeden M, Oyegbite K, Ozanne-Smith J, et al.: World Report on Child Injury Prevention. Geneva. 2008. PubMed Abstract\n\nQuan L, Cummings P: Characteristics of drowning by different age groups. Inj Prev. 2003; 9(2): 163–168. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRahman A, Alonge O, Bhuiyan AA, et al.: Epidemiology of Drowning in Bangladesh: An Update. Int J Environ Res Public Health. 2017; 14(5): pii: E488. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRahman A, Mashreky SR, Chowdhury SM, et al.: Analysis of the childhood fatal drowning situation in Bangladesh: exploring prevention measures for low-income countries. Inj Prev. 2009; 15(2): 75–79. PubMed Abstract | Publisher Full Text\n\nRoyal Life Saving Society Australia (RLSSA): Royal Life Saving National Drowning Report. 2016. Reference Source\n\nSedain B, Pant PR: Dataset 1 in: Status of drowning in Nepal: A study of central police data. F1000Research. 2018. Data Source\n\nSharma G, Shrestha PK, Wasti H, et al.: A review of violent and traumatic deaths in Kathmandu, Nepal. Int J Inj Contr Saf Promot. 2006; 13(3): 197–199. PubMed Abstract | Publisher Full Text\n\nSoko B: Fiji Drowning Report 2012. Suva. 2012. Reference Source\n\nTyler MD, Richards DB, Reske-Nielsen C, et al.: The epidemiology of drowning in low- and middle-income countries: a systematic review. BMC Public Health. 2017; 17(1): 413. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWasti H, Lunetta P, Chaudhary G: Epidemiological profile of drowning in Nepal. Preliminary data from the Kathmandu region, 19127447. 2010. Reference Source\n\nWorld Health Organization: World Health Organization. Drowning Prevention in the South-East Asia Region-2014. New Delhi. 2014. Reference Source"
}
|
[
{
"id": "34009",
"date": "15 May 2018",
"name": "Amy E. Peden",
"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 the opportunity to review the article entitled “Status of drowning in Nepal: A study of central police data”. This is a well-written study detailing drowning in a country with little previous published research, using a novel data source being police data.\nI have some general comments for the authors that I hope they find useful.\nFor someone unfamiliar with the geography of Nepal, I suggest adding a map as an additional figure that depicts the regions in Table 1 (Terai, Plains and Hills) geographically.\nYou need to clarify somewhere in the methods (and abstract) that this is all drowning, both intentional and unintentional.\nI also have some specific comments for the authors as follows:\nAbstract - Methods – can you specify the dates of the study – e.g. 1 January 2013 and 31 December 2015\n\nAbstract – Methods – can you confirm what places means? Its not aquatic locations so is it geographical location within Nepal?\n\nAbstract – Methods – suggest replacing ‘done’ with ‘undertaken’ or ‘conducted’\n\nMain manuscript – Introduction – this study needs a clearer aim – the sentence that currently ends the introduction may be about the study referenced before. A clearer aim e.g. this study aims to …. Using police report as the main data source, with media reports as a secondary source of data..\n\nMethods – how many don’t report to police? Add to limitations.\n\nMethods – again confirm what place of drowning means if not type of body of water\n\nEthics – usually even though the participants are deceased, this research would be considered as including human participants and ethics would be required. Is it possible to get ethics approval retrospectively? It is definitely helpful that you specify that the data received was de-identified.\n\nResults – can a definition for the term ‘open water’ as per the first paragraph of the results be added to the methods?\n\nDiscussion – ‘systemic’ should be ‘systematic’.\n\nDiscussion – please also cite dedicated epidemiological study of rivers that have been conducted in Australia1. These further support the statement that natural water bodies such as rivers regularly account for large proportions of drowning deaths.\n\nDiscussion – when discussing riskier behaviour in males – you might want to note the involvement of alcohol (see Peden et al. (2017)2). I’m not sure whether alcohol is a big issue in Nepal, and I note there was no toxicological data from police supporting or disproving alcohol involvement but it might be worth mentioning as an additional area of research for drowning in Nepal, especially among males.\n\nStrengths - By using police records, you also avoid issues with traditional ICD drowning codes (W65-74) not capturing all drownings (see Peden et al. (2017)3). This is a strength of the study.\n\nLimitations – are there ever cases of people not reporting a drowning at all or not reporting it within 24 hours? If so this study likely underreports drowning in Nepal and this would be worth mentioning.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3819",
"date": "30 Jul 2018",
"name": "Bhagabati Sedain",
"role": "Author Response",
"response": "For someone unfamiliar with the geography of Nepal, I suggest adding a map as an additional figure that depicts the regions in Table 1 (Terai, Plains and Hills) geographically. Author's response: Thank you! We found your comment extremely helpful and have revised accordingly. You need to clarify somewhere in the methods (and abstract) that this is all drowning, both intentional and unintentional. Author's Response: Both the intentional and unintentional drowning deaths were analysed in this study. So, in this revised paper, it is clearly mentioned that the study included both the intentional and unintentional drowning deaths. I also have some specific comments for the authors as follows: Abstract - Methods – can you specify the dates of the study – e.g. 1 January 2013 and 31 December 2015 Author's Response: We also realised it is better to mention the study date in such way and have revised accordingly. Abstract – Methods – can you confirm what places means? Its not aquatic locations so is it geographical location within Nepal? Author's Response: In the study, the places means the geographical locations i.e. Mountain, Hill and Terai. Instead of places, the term geographical location is used in the revised paper. Abstract – Methods – suggest replacing ‘done’ with ‘undertaken’ or ‘conducted’ Author's Response: We really appreciated your comments and since terminology would be more accurate and precise, we have taken your advice. Main manuscript – Introduction – this study needs a clearer aim – the sentence that currently ends the introduction may be about the study referenced before. A clearer aim e.g. this study aims to …. Using police report as the main data source, with media reports as a secondary source of data.. Author's Response: The study used police report as the main source of data. Objectives of the study was mentioned in the abstract. We also noticed that the objective of the study was missing in the introduction section of the main text. At the end of the introduction of the main text, the objective of the study is stated in the revised article. Methods – how many don’t report to police? Add to limitations. Author's Response: Review of literature showed that larger number of unnatural deaths were not reported in the police. As Nepal is a less developed country, there is the high probability of people not reporting the unnatural deaths to the police. So, it is challenging to discuss on the missing cases. In case of drowning deaths, Nepal Police’s record is the only one source information. Methods – again confirm what place of drowning means if not type of body of water Author's Response: The word drowning place was mistakenly written in the main text so it is deleted. Ethics – usually even though the participants are deceased, this research would be considered as including human participants and ethics would be required. Is it possible to get ethics approval retrospectively? It is definitely helpful that you specify that the data received was de-identified. Author's Response: Ethical approval was obtained from University Review Committee. We used secondary data which was completely de-identified right from Data Extraction which was done on the Police Office premises. No identifiable data for the deceased person was transferred by any means. Results – can a definition for the term ‘open water’ as per the first paragraph of the results be added to the methods? Author's Response: Thank you for this excellent observation. Both of the reviewers commented on this word ‘open water’. So the word replace with natural water bodies. Discussion – ‘systemic’ should be ‘systematic’. Author's Response: It was a typo in the manuscript and changed accordingly. Discussion – please also cite dedicated epidemiological study of rivers that have been conducted in Australia1. These further support the statement that natural water bodies such as rivers regularly account for large proportions of drowning deaths. Author's Response: We agree that it will be better if it was included in the discussion. We have added the suggested reference in the discussion section. Discussion – when discussing riskier behaviour in males – you might want to note the involvement of alcohol (see Peden et al. (2017)2). I’m not sure whether alcohol is a big issue in Nepal, and I note there was no toxicological data from police supporting or disproving alcohol involvement but it might be worth mentioning as an additional area of research for drowning in Nepal, especially among males. Author's Response: In the police record, the use alcohol use was not clearly mentioned. However, there is the high prevalence of drinking behaviour among the male. So we are grateful to know about this article and it was realised that the adding of the suggested reference would validate our result. So the suggested reference has been included in the revised manuscript. Limitations – are there ever cases of people not reporting a drowning at all or not reporting it within 24 hours? If so this study likely underreports drowning in Nepal and this would be worth mentioning. Author's Response: . There is high possibility of not reporting all deaths within 24 hrs. The information recorded in the Nepal Police through daily incidence reporting system was analysed in this study. It is likely to observe underreporting of the drowning deaths. In the revised manuscript this issue is clearly mentioned in the limitation section."
}
]
},
{
"id": "34006",
"date": "16 May 2018",
"name": "David Meddings",
"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 the opportunity to review this paper.\nGeneral comments\nThe paper is an important and useful contribution, as there is very little in the peer-reviewed literature about drowning in Nepal. The authors should go into more detail about the police data source they used - particularly whether there are estimates of the number of deaths that occur but which are not reported to this source, whether there are biases etc. The presentation of results seems somewhat haphazard - the authors could benefit from taking one step back from their data and deciding what are the 3-5 main findings we have come across here that are important and deserve to be amplified and framed clearly for the reader. There is little in the way of serious follow up discussion on some aspects that would seem to invite more critical thinking - the disparity between this study's findings and GBD estimates for example. There are numerous spelling and grammatical errors in the text, so this should be edited prior to publication.\nSpecific comments\nResults section - drop \"the open water\" - which could be confused as meaning waters far from any coastline - and replace with either natural water bodies or rivers. Proportion of males is cited as being higher than females - it is simple enough to give this as \"roughly 3 times as many deaths were reported in males versus females\", or words to that effect. It communicates with greater precision one of the findings of interest. I take the seasonality pattern but believe there is somewhat muddled writing about the effect of seasons here. Clearly there are some drownings that occur year round, but the authors should be clear on drawing attention to the surge during the monsoon season as this has implications for prevention programmes. The references appear incomplete - for example, the Global Report on Drowning from the WHO is referenced within the body of the text but does not appear within the references, although a WHO regional report does.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-576
|
https://f1000research.com/articles/7-1144/v1
|
27 Jul 18
|
{
"type": "Method Article",
"title": "Revealing HIV viral load patterns using unsupervised machine learning and cluster summarization",
"authors": [
"Samir A. Farooq",
"Samuel J. Weisenthal",
"Melissa Trayhan",
"Robert J. White",
"Kristen Bush",
"Peter R. Mariuz",
"Martin S. Zand",
"Samir A. Farooq",
"Samuel J. Weisenthal",
"Melissa Trayhan",
"Robert J. White",
"Kristen Bush",
"Peter R. Mariuz"
],
"abstract": "HIV RNA viral load (VL) is an important outcome variable in studies of HIV infected persons. There exists only a handful of methods which classify patients by VL patterns. Most methods place limits on the use of viral load measurements, are often specific to a particular study design, and do not account for complex, temporal variation. To address this issue, we propose a set of four unambiguous computable characteristics (features) of time-varying HIV viral load patterns, along with a novel centroid-based classification algorithm, which we use to classify a population of 1,576 HIV positive clinic patients into one of five different viral load patterns (clusters) often found in the literature: durably suppressed viral load (DSVL), sustained low viral load (SLVL), sustained high viral load (SHVL), high viral load suppression (HVLS), and rebounding viral load (RVL). The centroid algorithm summarizes these clusters in terms of their centroids and radii. We show that this allows new VL patterns to be assigned pattern membership based on the distance from the centroid relative to its radius, which we term radial normalization classification. This method has the benefit of providing an objective and quantitative method to assign VL pattern membership with a concise and interpretable model that aids clinical decision making. This method also facilitates meta-analyses by providing computably distinct HIV categories. Finally we propose that this novel centroid algorithm could also be useful in the areas of cluster comparison for outcomes research and data reduction in machine learning.",
"keywords": [
"Machine learning",
"HIV",
"viral load",
"feature extraction",
"HIV categories",
"centroid",
"cluster summarization",
"clinical interpretability"
],
"content": "Introduction\n\nThe primary clinical goal of HIV treatment and patient engagement is suppression of the HIV viral load (VL), as measured by low or undetectable circulating HIV RNA levels. However, VL most often fluctuates over repeated measurements, with a range that spans 8 orders of magnitude from 0 (undetectable) - 107 copies/mL. VL is regularly monitored for signs of progression of HIV infection. Standard HIV treatment protocols are based on VL measurements1, especially when monitoring responses to antiretroviral therapy (ART). Monitoring of VL helps to determine whether ART therapy was able to successfully suppress patient VL2. Individuals with sustained high viral loads (SHVL) are at greater risk of secondary transmission, clinical progression to AIDS, and death3–6. In contrast, significant reduction in VL or high viral load suppression (HVLS) both lead to immune recovery, as measured by CD4 T cell levels7, and can reduce or eliminate the risks of SHVL. Furthermore, patients sustaining low-level viral load (SLVL), or with a rising VL after previous suppression, have a high incidence of treatment failure8. Thus, developing an objective measure of VL status, and categorization of patients by time varying patterns of VL, is critical for standardizing both therapy and comparing research protocol efficacy.\n\nReports in the current literature differ in the definition “high viral load\"9–12, and their findings of how long it takes a patient on highly active anti-retroviral therapy (HAART) to suppress their VL2,9,10,13. We summarize some of the published approaches here (for greater detail see Supplementary File 1). With respect to VL levels, Terzian et al. defined SHVL as two consecutive viral load measurements (VLM) ≥100,000 copies/mL9. Durably suppressed viral load (DSVL) was defined as all VLM <400 copies/mL. In contrast, Greub et al. focused on detecting low level viral rebound (LLVR) by first considering patients with an initial consecutive VLM pair <50 copies/mL, and classified LLVR as having subsequent maximum VLM between 51–5008. Alternatively, Rose et al. investigated the use of five different frameworks to categorize suppressed versus not-suppressed VL10. Their approach excluded patients with VLM<200 at baseline, and stratified the remainder with regard to VL suppression using an 8 month window centered around 24 months after the start of VLM (18–30 months). Another approach was used by Phillips et al., and characterized VLM responses to ART13, utilizing a 24–40 week window and a rule-based method to identify two populations of HIV patients (Viral Failure and Viral Rebound). Despite these studies, no formal standard has been adopted by the field to classify a patient as having DSVL, SHVL, HVLS, SLVL, or rebounding viral load (RVL) patterns.\n\nClassifying patient VL states outside of research studies is further complicated in that real-world VL measurements are taken intermittently over time, and missing data is common due to a variety of factors (e.g. travel, social circumstances, non-adherence). This leads to incomplete and irregularly spaced data points. In addition, differences in the sensitivity of the multiple VL clinical assays available results in multiple cut off points for “undetectable\" viral loads analyzed at different facilities, further complicating analyses. Thus, there is a need for analytic techniques that can adjust for these details and classify VL states, both across research studies using different methodologies and to consistently classify patients in clinical practice.\n\nMachine learning methods can provide objective, unsupervised classification of patient clinical status14. These methods begin by collecting a set of features from patient data (e.g. demographics, laboratory measurements, therapies) and then performing computational clustering to identify similar patient classes. Some groups have applied machine learning methods to HIV research studies15 to predict HIV VL responses16 or CD4 T cell counts17, to distinguish between suppressed and viremic patients18, and to select therapeutic regimens19. None, however, have used machine learning to create a standard classification for VL status with irregularly sampled VL measurements across a cohort of patients.\n\nTo address these issues, we propose a set of unambiguous features which, when combined as a feature vector, capture the distinct dynamic patterns present in VL measurements over time. In addition, we have developed a novel centroid algorithm to cluster HIV positive subjects based on these patterns. Here we present the derivation of this method, and demonstrate its application to clustering 1,576 HIV patients with repeated VL measurements over a 5 year period. We found that patient VL measurements can be clustered into five time-varying patterns that correspond well to clinically relevant states. We note that the method and resulting categories can be used to standardize definitions of VL patterns across research studies, and potentially for clinical classification.\n\n\nMethods\n\nThis proposal was reviewed and approved by the University of Rochester Human Subjects Review Board (protocol number RSRB00068884). Consent was waived by the review board due to de-identification of the data set. The analysis in this paper is presented in compliance with Center for Medicare Services (CMS) current cell size suppression policy20. Data were coded such that patients could not be identified directly in compliance with the Department of Health and Human Services Regulations for the Protection of Human Subjects (45 CFR 46.101(b)(4)).\n\nWe obtained medical encounter data from all patients with an HIV diagnosis in the University of Rochester Medical Center’s electronic medical record system (EMR) between 2011–2016, including age, gender, race, ethnicity, and VL measurements. There were a total of 1,892 patients with at least one VL measured, with 1,576 of these patients having at least three VL measurements.\n\nMeasurements ≤48 copies/mL, present as categorical values “NEG\", “POS < 20\", or “POS < 48\" were transformed into numerical values of 0, 20, and 48 respectively. The deidentified study data containing only viral load measurements and relative time are available at https://doi.org/10.5281/zenodo.131324521.\n\nAnalyses were performed on a Windows 8 server with Intel(R) Xeon(R) CPUs E5-2620 v2 @ 2.10GHz and 256GB of RAM. Python 2.7 was used for most data mining and machine learning under Spyder v.3 installed from Anaconda2 (64-bit). The default packages available in Anaconda were used for analysis, including, but not limited to: NumPy, scikit-learn, SciPy, datetime, csv, math, Matplotlib, pip, operator, copy, random, and time. Using pip we installed the webcolors and pydotplus packages for rendering a decision tree. SQLite was used to store, query, and clean ~the data. Analytic code is available for download at https://github.com/SamirRCHI/Viral_Load_Data_Categorization.\n\nSince VL data is asynchronous and noisy, with variable numbers of data points for each subject, we excluded patients with ≤ 2 VL measurements as too few to accurately assess VL patterns. Based on temporal patterns of VL described in the literature, the VL pair distribution of our data (Figure S1), and a further extensive investigation into the data, we hypothesized six potential temporal VL patterns, defined in Table 1 and illustrated in Figure 1.\n\n*Colors are used throughout the manuscript to identify clusters\n\nExamples of each type of viral load pattern. Note that actual viral load patterns are noisier and may often be more difficult to distinguish. The magnitudes of viral load values reflect those found in the dataset.\n\nIt is important to note that these definitions are pattern based, and do not explicitly select absolute VL cutoff levels or a specific temporal window, as other reports have done8–10,13. This has the advantage of allowing the absolute VL levels and critical time windows to emerge from the analysis. It also does not preclude incorporation of absolute levels (e.g. VLM>400) at a later stage into the pattern specification.\n\nMathematical notations for this work are described in Table 2. We next designed a feature vector to capture characteristics that would allow us to distinguish between VL patterns. VL values at the lower limit of detection are a function of the specific assay used, and appear in our data set as 0, 20, and 48 copies/mL (Figure S1). Thus, plots of the log10 transformed data have discretely spaced values at the lower level of detection, capturing the undetectable range of viral load. Additionally, we adjusted the data by log10[V L + 10] to avoid log10[0]. The addition of 10 to VL (instead of 1) is used to minimize the distance between the undetectable values: 0, 20, and 48 (copies/mL). Thus, in our notation, all the values related to viral load are assumed to have been adjusted to this measure. For example, minV L = log10[0 + 10] = 1 and maxV L = log10[107 + 10] ≈ 7.\n\n*This is after selecting for patients with ≥ 3 measurements.\n\n**This value changes after transformation of the data.\n\nUsing the transformed VL data, we next extract several relevant features of the VL measurements over time. These features are used for machine learning classification of individual patient VL time series, and designed to distinguish patterns in VL change while minimizing the effects of noise. We do not limit feature extraction based on the total elapsed time of viral load measurements because the optimal time-point for determining viral load class is not well established. The attributes for feature extraction are: relative area of viral exposure, weighted recency reliability, adjusted maximal difference, and interquartile range. The definitions include:\n\n1. Relative area of viral exposure (Area) - the area under the viral load curve relative to the total viral load area possible, which has a range between [0,1]. We choose a normalized, relative score, as the total time span between the first and last viral load measurement, which differs between patients. This feature is similar to finding the mean and median, except it is sensitive to the dimension of time, hence yielding more information. The feature is calculated by summing the area of each trapezoid created by each pair of viral load values, followed by dividing by the total possible area (Equation 1).\n\n\n\n2. Weighted recency reliability (wRR) - Due to viral load noise, the last measurement may not be an accurate reflection of a patient’s viral load trend. For example, a patient may have a VL whose average slope is negative, indicating high viral load suppression over time (HVLS). If, however, the last measurement is slightly higher than the trend, heavily weighting this last measurement could lead to mis-classifying the patient as rebounding viral load (RVL). To account for this, we calculate a weighted mean where the weight of the VL measurement increases with time. More specifically, the weight function follows an inverse square root function (f(x)=1x) rather than an inverse function (g(x)=1x). This has the advantage of avoiding rapid convergence of g(x) to zero when time is measured in units of days (Equation 2). Weighted recency is then calculated as the dot product of the viral loads and weights divided by the sum of the weights (Equation 3).\n\n\n\n\n\nWe were also interested in how reliable wR is as a representation of the patient’s viral load trend. To this end, we calculated the absolute deviations from the viral load measurements to wR (Equation 4). Rather than averaging the deviations, we take the median to reduce the effects of outliers and call this our weighted recency reliability measure (Equation 5). We take the inverse to force the range of the result to be between [0,1]; a property made to use in our next proposed feature, adjusted maximal difference.\n\n\n\n\n\n3. Adjusted maximal difference (Adj MD) - this is time-independent the difference between the “peak” and last VL measurements. To distinguish between viral load suppression or emergence, we calculate the “peak” as the maximum of the absolute deviations (Equation 4) and retain the sign of the result. We expected the positive scores to effectively isolate the EVL group, however, we instead found that retaining the positive (emergent) scores lead to mis-categorization of SHVL and RVL groups without clearly identifying EVL patterns. This, along with other investigation into the data, led us to conclude that the EVL pattern may not exist in our data, but we refrain to make generalizations to all healthcare facilities. With this consideration, we force (ground) the positive scores down to zero for proper labeling of SHVL and RVL (Equation 6).\n\nDue to the varying nature in viral load measurements, we are hesitant to use the final viral load measurement as a means of judging suppression. Thus we propose to use wR instead. To reduce the effects of rebounding patients being falsely labeled as suppressed patients, we multiply our result by wRR - as rebounding patients are expected to have a low score in the range [0,1]. The maximal difference is necessary in order to ensure that the suppression type of viral load patterns are classified appropriately (Equation 7).\n\n\n\n\n\n4. Interquartile range (IQR) - This feature is added to further segregate the rebounding patients and follows the standard interquartile range calculation (Equation 8).\n\n\n\nMachine learning methods for cluster classification were compared by calculating F1 scores, the harmonic mean of precision and recall22, defined by Equation 9–Equation 11.\n\n\n\n\n\n\n\nHere we formally define keywords appearing in the analysis: Let Feature extraction be the process of determining the values Ȧ, wRR, D˅, and IQR from a set of patients (using their viral load patterns) with the formulations given above. Then a feature vector (F→p) contains the values Ȧp, wRRp, D˅p, and IQRp extracted from patient p’s viral load pattern. The words sample or point are also used here RVL (black; n= 237) and HVLS (purple; n=316) clusters. interchangeably. The term feature (F) can be thought of as a column vector for all patients in the dataset consisting of the four attributes: FȦ, FwRR, FD˅, and FIQR. Finally, the terms label assignment, VL pattern membership assignment, patient categorization, and prediction, all refer to the same principle: To assign the most appropriate label which characterizes the viral load pattern of a patient. However, while the principle is the same, the method of assigning such an appropriate label differs depending on the categorization or the learning method used.\n\n\nResults\n\nWe began by transforming viral load data by min-max normalization22 to equally weight the temporal features of the VL series (Equation 12). That is, we normalize the features, F, to a range between [0, 1] using Equation 12 where F* = f(F).\n\n\n\nNext, we examined each of the four features for all patients with ≥ 3 viral load measurements (N = 1,576 patients), and did not find distinct bi-variate clustering (Figure S2). A feature correlation coefficient analysis (Supplementary Table S1) revealed that the Adj MD feature is linearly independent of Area and wRR. In contrast, there is modest linear dependence between IQR and Adj MD, and between Area and both wRR and IQR. As expected, the largest linear dependency is between wRR and IQR. These results suggest the separation between viral load patterns will be most noticeable between the Area and the Adj MD features - as we designed them to be. Also, although Adj MD is dependent upon wRR, we find that their correlation coefficient is very low (0.033).\n\nWe then performed hierarchical clustering of the individual subject VL patterns using a Euclidean distance metric and Ward’s linkage criterion23 to minimize the total within-cluster variance. Patients showed a clear separation into 5 distinct groups, which had clinical significance (Figure 2 and Figure 3). The cluster with the lowest viral loads and the highest weighted recency reliability (n=442) corresponds to the DSVL patient group. The patients corresponding to the SHVL group (orange; n=46) exhibited the highest relative Area and very low IQR. Compared to the DSVL cluster, the blue cluster (n=535) has slightly greater area and IQR with a significant difference in the weighted recency reliability. Using this information, along with the general patterns shown by Figure 3, we identify this as the SLVL group. The algorithm also identifies the RVL (black; n=237) and HVLS (purple; n=316) clusters. The RVL cluster has a low weighted recency reliability and high IQR. In contrast, the HVLS cluster has a lower area, higher weighted recency reliability, indicating little variation in the terminal portion of the VL time series, and most importantly very low adjusted maximal differences (Figure 4).\n\nClustered using the Euclidean distance along with Ward’s method. Numbers on the bottom axis show number of patients in each cluster. The corresponding viral load pattern plots can be found in Figure 3. DSVL = Durably Suppressed Viral Load, SHVL = Sustained High Viral Load, SLVL = Sustained Low Viral Load, RVL = Rebounding Viral Load, HVLS = High Viral Load Suppression\n\nFor each cluster categorization of the patient from Figure 2, the days since first viral load measurement are plotted against the viral load counts. The points on the plots indicate the last viral load measurement.\n\nEach patient is colored corresponding to the results from the hierarchical clustering in Figure 2. The artificial line of points is a result of the grounding function used in Adj MD. Area = relative area of viral load exposure, wRR= weighted recency reliability, IQR = interquartile range, AdjMD = adjusted maximal viral load difference.\n\nVL patterns are similar within clusters, and dissimilar between clusters Figure 3. Interestingly, there are patients within each cluster whose last VL measurement occurs near 1,827 days. This is equivalent to the full span of five years of VL monitoring data set. This suggests that these clusters don’t disappear after some elapsed time, but rather each type of pattern can be found at virtually any time point.\n\nWe found large VL spikes within the time series of the HVLS group. We hypothesize that this may be due to the asynchronous timing of measurements between subjects, the natural variation in biological responses, or patient variability in adherence to therapy. This observation also reflects one limitation of asynchronous outcomes data sampling, which lacks a “completion\" endpoint characteristic of most prospective, randomized clinical trials. If measurements ended at a spike, the adjusted maximal difference feature may be weighted in the favor of the patient being classified as RVL. This may indicate that some patients classified as suppressing their viral loads should have been classified as having rebounding viral loads. Alternatively, may indicate that these features do not restrict a patient to forever to one category, but allow for dynamic classification as a function of biological or therapeutic responses.\n\nUsing the same data set, we next compared our VL pattern categorization method to those previously published in the literature. Visually, we find that the SLVL group detected by our method is very similar to the LLVR group defined by Greub et al. (Figure 5). Furthermore, it appears that the methods trying to capture SHVL, viral rebound, and viral failure patients did not succeed as well as the identification of SHVL and RVL patients in our method. RMVL repeat continuous visually appears to have performed very well in identifying patients whom have suppressed their viral loads. However, the results suggest that our analysis performs slightly better in identifying the suppression group (HVLS), as we find that the last VL measurements (black dots in Figure 5) are consistently low using our method.\n\nA 2D binning of VLM counts for every patient category. Each row uses a different categorization method, and the method name is located to the right of the row, and the title of each subplot is the category assigned by the indicated method. The columns of each 2D bin are normalized based on the maximum number of logged viral load measurement (VLM) counts in the column: log10[1 + V L M ]. Bin color for a count of 0 is copper, and other bin colors range from white to teal (the maximum of the log10[V L M counts] in the column of the bin). The black dots represent the last viral load measurement for the patient (opacity ≥ 0.3; 2D bins have variable opacity for the dots). The bottom row is our analysis is the same as Figure 3, but represented as a 2D bin. DSVL = durably suppressed viral load; LLVR = low level viral rebound; SLVL = sustained low viral load; SHVL = sustained high viral load; HVLS = high viral load suppression, RVL = rebounding viral load.\n\nThe other methods may not have performed as well as they rely on a window or a consecutive pair measure, which may be too subjective for assigning VL pattern membership. Furthermore, notice that patients with baseline VL<200 (Figure 5) contain VL patterns which can reach as high as 106 copies/ml, which is in contrast to Rose et al.’s assumption that these patients have consistently low viral variation. Lastly we wish to emphasize that while some of these categorization methods are successful in identifying a specific group of patients, our method is unique as it attempts to associate each VL pattern to a specific category, without using categories such as “Not Suppressed”, “Unspecified”, or “Omitted”.\n\nWe next used the classes identified by hierarchical clustering to compare several machine learning models, with the goal of identifying methods that could be trained to prospectively assign HIV patients to VL categories (i.e. SHVL, SVL, SLVL, DSVL, and RVL). Unsupervised learning (e.g. hierarchical clustering) is useful for establishing the data structure of VL categories and their locations in the feature vector space. Once the model is established (e.g. cluster boundaries), supervised learning methods are better suited for prospective cluster assignment, given a robust \"ground truth\" for model training, as they do not depend on re-analysis of the entire population.\n\nTo this end, we compared the predictive power of several supervised learning methods for HIV cluster assignment, including: k-nearest neighbors (kNN), decision tree, support vector machine (SVM), Adaboost, and random forests. Models were trained on the original data set, and we then ranked their prediction power by their average F1 score derived by leave-one-out cross-validation (LOOCV) on the clustered results (Table 3). We compared the ability of these methods to reconstruct the originally identified clusters, even when allowing for variability in cluster numbers (e.g. kNN with k={7, 9}, or DT without a maximum depth specification). All methods performed comparably well, with the notable exception of Adaboost. This generally high performance was expected because the VL pattern categories are well-separated as a result of the clustering. k-Nearest Neighbors and k=5, was computationally efficient and yielded the best results in Table 3.\n\nLOOCV = leave-one-out cross validation, kNN = k nearest neighbor, DT = decision tree\n\nSVM = support vector machine, DSVL = Durably Suppressed Viral Load\n\nSHVL = Sustained High Viral Load, SLVL = Sustained Low Viral Load\n\nRVL = Rebounding Viral Load, HVLS = High Viral Load Suppression\n\nWe next considered the trade-off of predictive precision versus model interpretability. Critical clinical evaluation of machine learning results is important to protect against mis-categorization and clinical error. For this reason, many have advocated using models that are more clinically interpretable. kNN is dependent upon the entire training set for prediction, as it does not inherently “learn” patterns24, hence it does not meet our interpretability criteria. In comparison, SVM offers a simpler model, but it’s results could be non-intuitive for clinicians. And although Decision Trees offer the best interpretability, overly complex trees may be generated, as occurred in our study (Figure S3).\n\nWe found that pruned decision tree rules, with a maximum depth of 5 levels, met this interpretable criteria, however at a slight cost to the predictive power (Table 3). The extracted decision rules are shown in Figure 6. Each category has a rule with a high proportion of true positive samples following the rule relative to all samples for the category (support). Similarly, a high proportion of the predicted class was found in the rule (precision), indicating that the rules can be summarized into a majority rule. Note that the sum of the support does not necessarily add up to one for each class because some samples belonging to that class may have been otherwise placed into a different rule, making the precision of that rule weaker.\n\nSupport is the fraction of true positives satisfying the rule relative to all samples of the class. Precision is the proportion of true positives versus all positives in the rule. Rules are sorted in order of application, first by the level of the decision tree depth (Depth), and then by descending precision. The colored regions represent the the values for which the rule holds (rule feature space). For the centroid method (CM; shaded gray) bounds were calculated by the polyhedron method, where the rectangular bar is the center and the radius is the area inside the parentheses. Area = relative area of viral load exposure, wRR= weighted recency reliability, IQR = interquartile range, AdjMD = adjusted maximal viral load difference.\n\nAs an alternative interpretable model we explored the use of centroid cluster summarization, which is often used in clustering algorithms, and is flexible enough to accommodate different centroid determination methods22,25. To determine the effects of different centroid determination algorithms, we compared seven different methods: multidimensional mean, multidimensional median, best representative center, bounding box method, smallest disk method, polyhedral center, and a novel “push and pull\" (PnP) method inspired by force-directed graph drawing such as the Fruchterman-Reingold’s algorithm26,27 (see Supplementary File 1). Force directed clustering methods maximize inter-cluster center distances, while minimizing intra-cluster distance, and are the basis for modularity clustering in graph theory28.\n\nWe then combined the centroid cluster summarization approach with a radius-based classification prediction algorithm. Let ci be the ith cluster center with corresponding radius ri, where ri is calculated as the distance to the farthest intra-cluster sample from ci, then for a new sample s choose its predicted cluster membership j such that ‖s−cj‖2rj is a minimum. We refer to this method as radial normalization classification.\n\nComparing the representative F1 power of the centroid radial normalization methods (italicized in Table 3) to common machine learning algorithms, we find that the centroid interpretation loses some predictive power. However, the centroid summary is highly interpretable because the entire model can be expressed concisely (Supplementary Table S2), and understood clearly. For example, a clinician classifying a patient by VL time series values would compare observed feature values with the ranges given in Figure 6, and find which classification the patient’s data fits best within. In the case of the centroid method, if an observed value appears to fall in multiple categories, then they should be assigned to the one closest to the center (this allows a clinician to cross-check model predictions).\n\nHIV patient viral load states are often fluid, with class changes (e.g. SHVL → HVLS) occurring due to therapy, viral genetics, social and other factors. To examine this aspect of classes, we use the k-Nearest Neighbors (k=5) model, fit to the original clusters, to predict the class state of each patient with ≥ 3 VL measurements using only partially retained VL data. For example if a patient has 6 viral load measurements, then we predict the class state at 3, 4, 5, and 6 VLM, which may yield SHVL → SHVL → RVL → HVLS as its prediction. We then constructed a state-transfer network using the trace-route method29, revealing several interesting relationships:\n\n1. Patients on therapy appear to suppress their viral loads at a positive linear rate throughout the entire 900 day span. This is quite different from the literature which suggests that if a patient is going to suppress their VL, it will be within 32 weeks, or 224 days13 (Figure 7A).\n\nA) Classification using kNN, with k=5, trained on the original five clusters to predict on partially retained viral load for patients ≥ 900 days of data. The number of patients in one class between 0–900 days are shown relative to the first state classification (i.e. third viral load measurement). B) A trace-route map of class state transfers (class1 → class2) as a function of partially retained viral load derived from model. Nodes represent viral load classification and arrows reflect the volume of state transitions between successive VL measurements (e.g. SHVL→DSVL). Self-loops (e.g. RVL→RVL) indicate no change in state reflecting stable classification.\n\n2. DSVL classification appears unstable for the first 400 days, suggesting that patients in this class should be monitored carefully during this initial period (Figure 7A).\n\n3. The number of patients classified as SHVL drops considerably until ∼500 days after first classification. After this point, those patients who have not yet left the SHVL category, may not do so (Figure 7A).\n\n4. The two sets of classes {DSVL, SLVL} and {SHVL, RVL, HVLS} are well separated (i.e. without much transfer between sets; Figure 7B). This appears to suggest that patients whose viral load is consistently low or durably suppressed tend not to transfer into a high viral load state (i.e. RVL or SHVL), at least in this data cohort.\n\n5. SHVL patients in this cohort tended to transfer out of the class at a much higher rate than the transfer in, suggesting positive patient care (Figure 7B). This observation is consistent with reports in the literature that entry into treatment, with adherence to a HAART regimen, generally results in viral load suppression.\n\n6. The state transfer diagram illustrates that the most frequent state transition over time is remaining within the same cluster (Figure 7B) assignment.\n\n\nDiscussion\n\nResearchers have previously performed HIV population case studies using differing schema to classify VL patterns2,9,10,13. We have developed a unique method for standardizing the algorithmic classification of VL patterns using a set of optimally segregating features. These features have been specifically engineered to optimize unsupervised clustering of temporal sequences of VL data that are asynchronous and noisy. Our findings demonstrate their success in identifying five viral load patterns often reported in the literature7–12. It is possible that additional viral load patterns may emerge in the future, for example due to new HIV variants that are resistant to current therapies. The method reported here is flexible enough to recognize such new temporal patterns of VL responses. It is also general enough that models could be trained on other viral infections that have patterns of natural or treatment related patient responses (e.g. hepatitis B and C, parvovirus B19), although this may require defining new features that capture disease specific pattern variants.\n\nA common practice in data analytics is to calculate the centroid as the average of the points22,30. However, Table 3 suggests that the mean is not necessarily the best centroid for HIV viral load data. We note two advantages of the centroid algorithm: First, we can choose the centroid best corresponding to the shape of the data, and second, we can use it to mathematically determine the amount of over-lap between n-dimensional cluster spheres (i.e. viral load categories). This method may facilitate cross-comparison of HIV research studies by providing a standard for VL pattern classification. Such standardization would be immensely useful in meta-analyses31–35, potentially revealing the influence of different patient care strategies or new relationships between different patient populations.\n\nOur work also explored the trade-off with respect to predictive accuracy between model interpretability and more complex, \"black box\" approaches to classification. The interpretability versus predictability problem is well known in the deep learning literature36–38. Interpretability is a desirable attribute in clinical classification systems, allowing clinicians to integrate causal physiology and diagnostic information with data features in a way promotes clearer bedside clinical reasoning. Using an interpretable model for assigning viral load pattern membership may be advantageous when a clinician wishes to use the assigned pattern membership to aid in making a critical clinical decision (e.g. choosing between treatment options), or when examining features that may be linked to a mechanism (e.g. slope of VL decline and viral genotype). A \"black box\" or more complex model may make such decisions or interpretations more difficult39, and can favor the use of simpler models at the expense of some predictive power.\n\nAlong these lines, we have also proposed a novel centroid-based algorithm for summarization of clustering results. This algorithm is not meant to supplant other well defined supervised learning algorithms, but rather to aid in interpretable assignment of VL patterns from other data sets into one of the five categories. The algorithm results are concise, allowing investigators to build the model in their preferred programming language. Hence this method may improve and standardize HIV population research by giving precise definitions to the varying temporal VL patterns, and potentially improving patient care.\n\nSeveral caveats apply to our work. As noted, this is a single center study, and thus our method should be tested with a much larger data set to cross-validate the categories represented by the clusters. In addition, our feature vector was designed specifically based on observed VL patterns previously reported in the literature rather than objectively clustering the data using a standard time-series based clustering method40–42. This may limit generalizability to other VL analyses. In addition, some of our features are slightly collinear - with the greatest correlation coefficient being between IQR and wRR (-0.717). However, while HVLS and RVL both have a varied range of IQR, it is clear that the HVLS class has greater wRR than the RVL class due to HVLS patients having a long consistent viral load tail. Furthermore IQR helps distinguish the HVLS and the SLVL or SHVL class, hence both IQR and wRR are necessary despite the slight correlation. Finally, because our method normalizes time into number of days since first VL measurement, we lose the ability to look for seasonal or yearly patterns in the data.\n\nOur data set did not have patients in whom VL was initially suppressed, and then rebounded (EVL). We originally hypothesized the existence of six distinct VL patterns, we found that the emergent VL group was not a pattern identified in our data. Perhaps this is a consequence of a high rate of local patient engagement in therapy in this cohort study, access to care, or the effectiveness of highly active anti-retroviral therapy regimens. We hypothesize that these conditions may not always exist (e.g. in areas where HAART is expensive, when people may lose the ability to pay for therapy), and that in such cases the EVL pattern may indeed be present and significant. Based on the formulation of the adj MD and wRR features, we hypothesize that a consequence of the grounding function is that any EVL pattern, if exists, will be grouped under RVL. This grouping may be appropriate as one can argue that going from a suppressed state to a high VL state is a form of rebounding. Clinical treatment of these patterns is likely to be similar. Further work with data sets that contain RVL patterns will need to be done to test these hypotheses. Unfortunately, we are not aware of any such data currently in the public domain.\n\nOur method used hierarchical clustering to define groups, with a cutoff for group specification at a high level in the branching tree (i.e. level 5). Such thresholds or tuning parameters are characteristic of most unsupervised clustering algorithms22,43. However, identification of important sub-clusters by using a lower threshold is also possible. Clustering results may change depending on the parameter chosen, revealing finer between-cluster differences as the number of clusters increase. The hierarchical clustering algorithm has the advantage that a proper cut-off can be easily visualized. For example, choosing a lower cut-off may reveal that the suppression group splits itself into categories with different rates of HIV viral load suppression during treatment. Researchers wishing to engineer a new feature vector for VL pattern segregation may find useful the Supplementary material on features we considered but subsequently removed due to poor performance.\n\n\nConclusions\n\nWe have proposed a set of four unambiguous features which have been successfully used in segregating five different types of temporal viral load patterns: durably suppressed viral load (DSVL), sustained low viral load (SLVL), sustained high viral load (SHVL), high viral load suppression (HVLS), and rebounding viral load (RVL). We have also proposed a novel centroid-based cluster summary algorithm. The use of this algorithm may improve meta-analyses or population studies of viral load patterns by standardizing the classification of HIV patient categories. Furthermore, the segregation process used in this paper (i.e. identifying domain specific features, performing unsupervised clustering, interpreting the results with a cluster summary) can be used to model other viral infections and the response of VL levels over time to treatment or natural disease progression. We also found that using a temporal state variation method is important when considering patient viral load classifications, as changes in patient response can continue to occur beyond previously estimated time frames.\n\n\nAbbreviations\n\nAdjMD = adjusted maximal viral load difference, Area = relative area of viral load exposure, ART = anti-retroviral therapy, DT = decision tree, EVL = emerging viral load, DSVL = Durably Suppressed Viral Load, HAART = highly active retroviral therapy, HIV = human immunodeficiency virus, IQR = interquartile range, kNN = k nearest neighbor, LLVL = low level viral load, LOOCV = leave-one-out cross validation, SHVL = Sustained High Viral Load, SLVL = Sustained Low Viral Load, RVL = Rebounding Viral Load, HVLS = High Viral Load Suppression, SVM = support vector machine, VL = viral load, VLM = viral load measurement, wRR= weighted recency reliability.\n\n\nData availability\n\nFull access to the data is available on GitHub (Data S1): https://doi.org/10.5281/zenodo.131324521\n\nData S1: Viral load data. The data set used for this study is provided in a completely deidentified format, CSV format where the first column represents a unique subject, with a random identifier. The subsequent values are as ti,j, V Li,j, where ti,j is the time from a universal T0 for the VL measurement j for patient i, and V Li,j is the corresponding VL measurement. Each record (row) is of a unique length, depending on the number of VL measurements present for that subject. The study data, and code used for analysis, can be found at https://doi.org/10.5281/zenodo.131324521.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was partially funded by the University of Rochester Clinical and Translational Science Institute grants UL1 TR002001, and TL1 TR002000 from the National Center for Advancing Translational Sciences (NCATS), a component of the National Institutes of Health (NIH). This publication was also made possible through core services and support from the University of Rochester Center for AIDS Research (CFAR), an NIH-funded program (P30 AI078498).\n\nThe content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health (NIH).\n\n\nAcknowledgements\n\nWe would like to thank Yusuf Bilgic (State University of New York at Geneseo) and James Java (University of Rochester), for discussions regarding the statistical analyses.\n\n\nSupplementary material\n\nSupplementary Figures:\n\nClick here to access the data.\n\nFigure S1: Viral load distribution. For each pair of viral load measurements, we calculate the change in days and the change in viral load counts for all patients and plot it as a scatter. The horizontal line of dots which appears between 0 and 2 are an artifact of using 20 and 48 in data to replace the “Pos <20\" and “Pos <48\" values which appeared in our data. The sequential range of viral load measurements shows that VL measurements taken within 10 days of each other may vary by ±105 copies/mL.\n\nFigure S2: Patient feature extraction. Feature extraction on 1576 patients displayed as 2D splicing of the 4 dimensional feature space. Each splice plots a dimension versus another in the form of a scatter plot.\n\nFigure S3: Decision Tree. While some useful rules may be pruned, the tree is otherwise complicated and difficult to draw useful conclusions from.\n\nFigure S4: Seven centroid calculations on clustered viral load data. For each cluster, the seven methods of calculating a globular cluster center are shown in comparison to each other (calculated on the normalized and clustered viral load data). Since the PnP method can have a center outside the range of [0,1], an indicator is shown for when the center goes beyond the range.\n\nFigure S5: Centroid methods. Gives a visual of how the seven methods work on an example point set. The green target signifies the exact center which is found according to the different methods in our algorithm.\n\nSupplementary File 1: Review of existing viral load categorization methods and features and centroid detection methodologies that were considered but not used. A review of currently published viral load categorization methods.\n\nClick here to access the data.\n\nSupplementary Tables:\n\nClick here to access the data.\n\nSupplementary Table S1: Correlation coefficinet matrix features.\n\nSupplementary Table S2: Centroids and radii from polyhedral CM.\n\n\nReferences\n\nCenters for Disease Control and Prevention (CDC): Vital signs: HIV prevention through care and treatment--United States. MMWR Morb Mortal Wkly Rep. 2011; 60(47): 1618–23. PubMed Abstract\n\nYehia BR, Fleishman JA, Metlay JP, et al.: Sustained viral suppression in HIV-infected patients receiving antiretroviral therapy. JAMA. 2012; 308(4): 339–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMellors JW, Muñoz A, Giorgi JV, et al.: Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med. 1997; 126(12): 946–54. PubMed Abstract | Publisher Full Text\n\nSterling TR, Vlahov D, Astemborski J, et al.: Initial plasma HIV-1 RNA levels and progression to AIDS in women and men. N Engl J Med. 2001; 344(10): 720–725. PubMed Abstract | Publisher Full Text\n\nDybul M, Fauci AS, Bartlett JG, et al.: Guidelines for using antiretroviral agents among HIV-infected adults and adolescents: recommendations of the Panel on Clinical Practices for Treatment of HIV. MMWR Recommendations and reports: Morbidity and mortality weekly report Recommendations and reports/Centers for Disease Control. 2002; 51(RR-7): 1–55. Reference Source\n\nAttia S, Egger M, Müller M, et al.: Sexual transmission of HIV according to viral load and antiretroviral therapy: systematic review and meta-analysis. AIDS. 2009; 23(11): 1397–1404. PubMed Abstract | Publisher Full Text\n\nViard JP, Burgard M, Hubert JB, et al.: Impact of 5 years of maximally successful highly active antiretroviral therapy on CD4 cell count and HIV-1 DNA level. AIDS. 2004; 18(1): 45–49. PubMed Abstract\n\nGreub G, Cozzi-Lepri A, Ledergerber B, et al.: Intermittent and sustained low-level HIV viral rebound in patients receiving potent antiretroviral therapy. AIDS. 2002; 16(14): 1967–1969. PubMed Abstract | Publisher Full Text\n\nTerzian AS, Bodach SD, Wiewel EW, et al.: Novel use of surveillance data to detect HIV-infected persons with sustained high viral load and durable virologic suppression in New York City. PLoS One. 2012; 7(1): e29679. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRose CE, Gardner L, Craw J, et al.: A Comparison of Methods for Analyzing Viral Load Data in Studies of HIV Patients. PLoS One. 2015; 10(6): e0130090. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Jong MD, Simmons CP, Thanh TT, et al.: Fatal outcome of human influenza A (H5N1) is associated with high viral load and hypercytokinemia. Nat Med. 2006; 12(10): 1203–1207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYlitalo N, Sørensen P, Josefsson AM, et al.: Consistent high viral load of human papillomavirus 16 and risk of cervical carcinoma in situ: a nested case-control study. Lancet. 2000; 355(9222): 2194–2198. PubMed Abstract | Publisher Full Text\n\nPhillips AN, Staszewski S, Weber R, et al.: HIV viral load response to antiretroviral therapy according to the baseline CD4 cell count and viral load. JAMA. 2001; 286(20): 2560–7. PubMed Abstract | Publisher Full Text\n\nKononenko I: Machine learning for medical diagnosis: history, state of the art and perspective. Artif Intell Med. 2001; 23(1): 89–109. PubMed Abstract | Publisher Full Text\n\nDubey A: Applications of Machine Learning: Cutting Edge Technology in HIV Diagnosis, Treatment and Further Research. Computational Molecular Biology. 2016; 6(3): 1–6. Publisher Full Text\n\nRosa RS, Santos RH, Brito AY, et al.: Insights on prediction of patients’ response to anti-HIV therapies through machine learning. In: Neural Networks (IJCNN), 2014 International Joint Conference on. IEEE; 2014; 3697–3704. Publisher Full Text\n\nRodríguez JO, Prieto SE, Correa C, et al.: Predictions of CD4 lymphocytes' count in HIV patients from complete blood count. BMC Med Phys. 2013; 13(1): 3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamirez CM, Sinclair E, Epling L, et al.: Immunologic profiles distinguish aviremic HIV-infected adults. AIDS. 2016; 30(10): 1553–1562. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParbhoo S, Bogojeska J, Zazzi M, et al.: Combining Kernel and Model Based Learning for HIV Therapy Selection. AMIA Jt Summits Transl Sci Proc. 2017; 2017: 239–248. PubMed Abstract | Free Full Text\n\nCenter for Medicare Services: CMS Cell Size Suppression Policy. 2015. [Online; accessed 29-November-2017]. Reference Source\n\nSamirRCHI: Samir-RCHI/Viral_Load_Data_Categorization: HIV Viral Load Categorization Release (Version v0.1-alpha).Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1313245\n\nHan J, Pei J, Kamber M: Data mining: concepts and techniques. Elsevier; 2011. Reference Source\n\nPunj G, Stewart DW: Cluster Analysis in Marketing Research: Review and Suggestions for Application. J Mark Res. 1983; 20(2): 134–148. Publisher Full Text\n\nKeller JM, Gray MR, Givens JA: A fuzzy K-nearest neighbor algorithm. IEEE Transactions on Systems, Man, and Cybernetics. 1985; SMC-15(4): 580–585. Publisher Full Text\n\nMaire F: An algorithm for the exact computation of the centroid of higher dimensional polyhedra and its application to kernel machines. In: Third IEEE International Conference on Data Mining. IEEE Comput Soc. 2003. Publisher Full Text\n\nKobourov SG: Spring Embedders and Force Directed Graph Drawing Algorithms. arXiv preprint arXiv 12013011. 2012. Reference Source\n\nFruchterman TMJ, Reingold EM: Graph drawing by force‐directed placement. Software: Practice and Experience. 1991; 21(11): 1129–1164. Publisher Full Text\n\nNoack A: Modularity clustering is force-directed layout. Phys Rev E Stat Nonlin Soft Matter Phys. 2009; 79(2 Pt 2): 026102. PubMed Abstract | Publisher Full Text\n\nZand MS, Trayhan M, Farooq SA, et al.: Properties of healthcare teaming networks as a function of network construction algorithms. PLoS One. 2017; 12(4): e0175876. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbdi H: Centroids. Wiley Interdiscip Rev Comput Stat. 2009; 1(2): 259–260. Publisher Full Text\n\nEtter P, Landovitz R, Sibeko S, et al.: Recommendations for the follow-up of study participants with breakthrough HIV infections during HIV/AIDS biomedical prevention studies. AIDS. 2013; 27(7): 1119–1128. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOlsen CM, Knight LL, Green AC: Risk of melanoma in people with HIV/AIDS in the pre- and post-HAART eras: a systematic review and meta-analysis of cohort studies. PLoS One. 2014; 9(4): e95096. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlaser N, Wettstein C, Estill J, et al.: Impact of viral load and the duration of primary infection on HIV transmission: systematic review and meta-analysis. AIDS. 2014; 28(7): 1021–1029. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoender TS, Sigaloff KC, McMahon JH, et al.: Long-term Virological Outcomes of First-Line Antiretroviral Therapy for HIV-1 in Low- and Middle-Income Countries: A Systematic Review and Meta-analysis. Clin Infect Dis. 2015; 61(9): 1453–1461. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoerma RS, Boender TS, Bussink AP, et al.: Suboptimal Viral Suppression Rates Among HIV-Infected Children in Low- and Middle-Income Countries: A Meta-analysis. Clin Infect Dis. 2016; 63(12): 1645–1654. PubMed Abstract | Publisher Full Text\n\nBologna G: Symbolic Rule Extraction from the DIMLP Neural Network. In: Lecture Notes in Computer Science. Springer Berlin Heidelberg; 2000; 240–254. Publisher Full Text\n\nBologna G: A model for single and multiple knowledge based networks. Artif Intell Med. 2003; 28(2): 141–163. PubMed Abstract | Publisher Full Text\n\nIntrator O, Intrator N: Interpreting neural-network results: a simulation study. Comput Stat Data Anal. 2001; 37(3): 373–393. Publisher Full Text\n\nShickel B, Tighe PJ, Bihorac A, et al.: Deep EHR: A Survey of Recent Advances in Deep Learning Techniques for Electronic Health Record (EHR) Analysis. IEEE J Biomed Health Inform. 2017; 1–1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlapper-Rybicka M, Schraudolph NN, Schmidhuber J: Unsupervised Learning in LSTM Recurrent Neural Networks. In: Artificial Neural Networks — ICANN 2001. Springer Berlin Heidelberg; 2001; 684–691. Publisher Full Text\n\nBahadori MT, Kale D, Fan Y, et al.: Functional subspace clustering with application to time series. In: International Conference on Machine Learning. 2015; 228–237. Reference Source\n\nKontaki M, Papadopoulos AN, Manolopoulos Y: Continuous subspace clustering in streaming time series. Inf Syst. 2008; 33(2): 240–260. Publisher Full Text\n\nKarypis G, Han EH, Kumar V: Chameleon: hierarchical clustering using dynamic modeling. Computer. 1999; 32(8): 68–75. Publisher Full Text"
}
|
[
{
"id": "37937",
"date": "28 Sep 2018",
"name": "Sally Blower",
"expertise": [
"Reviewer Expertise mathematical modeling of HIV"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 extremely interesting study that proposes a novel quantitative methodology for classifying HIV patients by viral load patterns. The authors propose four computable characteristics of time-varying viral load patterns and a novel classification algorithm. They demonstrate their approach by classifying a group of 1,576 HIV positive patients into five categories based on viral load patterns.\n\nThis is an extremely well written interesting paper with excellent figures. The proposed methodology has great importance and utility for both research studies and clinical programs.\nMy only very minor comments are:\nFor the descriptions given in Table 2 for mathematical notation. I suggest changing the description of \"refers to a single patient\" to \"refers to a specific patient\". Equation 10, please clarify what the denominator means; i.e., how does \"positive\" differ from \"true positive\" or \"false positive\". In the paragraph on page 6, under the heading Analytic terminology, editing is needed on line 8, where it says: clusters. interchangeably.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "39645",
"date": "29 Oct 2018",
"name": "Amalio Telenti",
"expertise": [
"Reviewer Expertise Host and pathogen genomics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report brings machine learning approaches to the classification of patterns of viral control in HIV infected individuals. This is welcome because, although this is a mature field in HIV, it signals the opportunity for new models in this and in other infections.\nStrengths: very well developed models and excellent reporting of the results through figures and documentation. The code and datasets are available in Github.\nWeakness: the model is trained and implemented on a suboptimal dataset. Treatment response in HIV infection (and thus the modeling of viral response) is well understood and best modeled with the knowledge of the time of treatment initiation, and a full understanding of variable influencing treatment response. Having a cohort that is described solely by “time from first measured viral load” is to all purposes, suboptimal. An additional issue is the reliance of a limited number of viral load determinations for an unclear number of individuals. Depending on the circumstances of sampling, having three viral load over an undisclosed time period is note devoid of many uncontrolled biases. Lastly, the text is equivocal in the utilization of the last time point – the reviewer understands that the information contained in the last point may be weighted because of the possibility that it is noisy. Unfortunately, in real life, that is the moment where strong predictive models are needed. It is possible that this was actually the goal of the authors.\nSummary: this work is a valuable contribution to the field, and the basic concepts and models will hopefully be deployed in the study of datasets that are more appropriate for this exercise. It is desirable that future modeling includes a more ambitious plan to move from the current train-test approach to one that establishes the generalization of the model. It will also be critical to observe the predictive value of the model on longer term outcomes.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1144
|
https://f1000research.com/articles/7-1143/v1
|
27 Jul 18
|
{
"type": "Research Article",
"title": "The effect of aerobic exercise on relative leukocyte telomere length in male Sprague-Dawley rats given a high fat-diet",
"authors": [
"Dewi Irawati Soeria Santoso",
"Nurul Paramita",
"Ani Retno Prijanti",
"Thressia Hendrawan",
"Swandito Wicaksono",
"Dewi Irawati Soeria Santoso",
"Ani Retno Prijanti",
"Thressia Hendrawan",
"Swandito Wicaksono"
],
"abstract": "Background: There is an increasing number of studies showing that physical activity and aerobic exercise have a positive effect on telomere length. Some studies also show that dynamics of telomere length is influenced by various environmental factors such as lifestyle and diet. However, the association between exercise and diet with telomere length is still questionable. The aim of this study was to examine the effects of aerobic physical exercise on relative telomere length changes in high fat-diet condition in rat animal models. Methods: This study was an in vivo experimental study using twelve Sprague-Dawley male white rats (12-month-old). Subjects were evenly and randomly divided into two groups (n=6): (1) high fat-diet fed control group; (2) high fat-diet fed and aerobic exercise treatment group. Aerobic exercise was conducted using animal treadmill with intensity of 20 m/min, 5 days/week. At weeks 4 and 8, relative telomere length was compared with week 0 control group, using q-RT-PCR. Results: Lengthening of relative telomere length was observed in both control and treatment groups at weeks 4 and 8, when compared to week 0 control group. The lengthening in the control group was much greater than the treatment group. Conclusions: Excessive increase of relative telomere length was seen in high fat-diet conditions. Aerobic exercise for 8 weeks suppresses excessive increase of relative telomere length in high fat-diet conditions.",
"keywords": [
"aerobic exercise",
"telomere length",
"high fat-diet"
],
"content": "Introduction\n\nObesity is a global problem that is associated with high mortality and morbidity. Several studies has shown that obesity increases the risk of cardiovascular disease, renal impairment, diabetes, and even certain cancers1–3. Meanwhile, the World Health Organization (WHO) showed that in the last 4 decades obesity rates have increased 10 times worldwide3. This phenomenon is worrisome because there is an increasing number of people who have a high risk of developing various diseases associated with obesity.\n\nSome studies show that there is a strong association between the incidence of obesity with the level of physical activity and high fat diet intake. Individuals with low physical activity levels are known to have a higher risk of developing obesity4,5. Physical activity is also known to have a role on lowering blood glucose levels, improving homeostasis in people with diabetes mellitus, decreasing production of oxidative stress, lowering triglyceride levels in the body, increasing endogenous antioxidants, and can also maintain telomere length, thus reducing cardiovascular metabolic disease risk6–8.\n\nThe role of physical activity in decreasing the risk of cardiovascular metabolic mortality and morbidity is thought to be mediated by maintaining the length of telomeres9. A study performed by Goglin et al. demonstrates the association between telomere shortening and increased 5-year mortality in patients with acute coronary syndromes. This study also showed that a decrease in telomere shortening rate would be followed by a decrease in mortality rate10.\n\nThe rate of shortening of the telomere can be suppressed through a healthy lifestyle such as healthy diet and physical activity9,11. Certain types of food have been shown to have a correlation with telomere length. A Mediterranean-rich diet of olive oil (38% of total energy as fat) has been observed to maintain telomere length12. In contrast, a high western type diet of sugar and red meat is associated with shortening of the telomere12.\n\nThe aim of this study was to explore the effect of aerobic physical exercise on telomere length under high fat-diet conditions to provide information for further research.\n\n\nMethods\n\nTwelve male Sprague-Dawley rats (Rattus norvegicus), weighing 250–450g, aged 12–13 months, were obtained from Central Animal Facility (Bogor Agricultural University) and divided into two groups: control and trained high fat diet. They were acclimatized for 1 week in a controlled room temperature of 24+1°C, with a 12-hour light/dark cycle, and access to food pellets and filtered water ad libitum to adapt to the new environment. They were housed in plastic cages (50×34×25 cm), two animals in each cage. All protocols used in this experiment received approval from the Ethical Animal Care and Use Committee of Faculty of Medicine Universitas Indonesia with approval number 164/UN2.F1/ETIK/2017.\n\nBefore the beginning of the study all rats were acclimatized with high fat-diet for 10 weeks, consisting of 19% fat, 24% protein, and 47.77% carbohydrate. After 10 weeks of acclimatization, the rats were evenly and randomly assigned using a random number into two groups (n=6 per group): (1) the control group (without aerobic exercise) and (2) the treatment group (with aerobic physical exercise). The treatment group received aerobic exercise for 8 weeks. During 8 weeks of intervention, both treatment and control groups were still given a high-fat diet.\n\nAerobic exercise was conducted using an animal treadmill, with a speed of 20 m/min for 20 minutes, 5 days/week, every morning around 6 am until 8 am. The intervention was carried out at the Biochemistry and Molecular Laboratory, Faculty of Medicine, Universitas Indonesia. All aerobic exercise protocols were supervised by experienced researchers.\n\nBlood was collected from both groups after an overnight fasting. All animals were anesthetized intraperitoneally with a ketamine-xylazine (KX) solution before blood was taken. Approximately 1 ml of whole blood was taken from the sinus orbitalis on week 0, week 4 and week 8. Genomic DNA in leukocytes was extracted from peripheral blood. DNA isolation were then performed using DNA isolation kit (GeneAll® ExgeneTM Clinic SV mini). Relative telomere length from isolated DNA were measured on a real-time PCR detection system using a Quantitative PCR method. Kit used for qPCR were pipettes (and tips), optical PCR plates and caps, and master mix. The type of taq used was AmpliTaq Gold DNA polymerase. The model number/name of the PCR machine was Applied Biosystems 7300. The cycling conditions used were 10 min at 95°C, followed by 40 cycles of 95°C for 15 sec, 60°C for 1 min, followed by a dissociation (or melt) curve. The primer sequences were as follows:\n\nTelo F: CGGTTTGTTTGGGTTTGGGTTTGGGTTTGGG TTTGGGTT\n\nTelo R: GGCTTGCCTTACCCTTACCCTTACCC TTACCCTTACCCT\n\n36B4 F: ACTGGTCTAGGACCCGAGAAG\n\n36B4 R: TCAATGGTGCCTCTGGAGATT\n\nThe primers were obtained from rodent (GenScript®). Relative telomere length was calculated using the formula of 2-ΔΔCt13.\n\nAll statistical analysis was performed using SPSS 20 for Windows. Because the distribution of data is not normal, the data was assessed using a nonparametric Kruskal-Wallis test.\n\n\nResults\n\nThe characteristics of the animals are shown in Table 1. All the animals were in good condition throughout the length of study. There was no significant difference in age, body weight, Lee index and telomere length between control and treatment groups.\n\nThere was an increase in relative telomere length at weeks 4 and 8 compared to week 0 in both groups. At week 4, the relative telomere length of the control group (2.231) did not differ much with the treatment group (1.802) when compared to week 0 of control group. At week 8, there was a progressive increase of relative telomere length in both groups compared to week 0 and week 4. Relative telomere length increase in week 8 of the control group was much higher (178.62) compared to week 8 treatment group (74.86) (Figure 1; Table 2).\n\nMedian (minimum-maximum).\n\n\nDiscussion\n\nOne important structures located at the ends of the linear chromosomes is the telomere. In human cells, they are composed of TTAGGG repeats and a number of proteins. Their function is to protect the integrity and stability of the DNA14.\n\nMany studies showed that telomere length is influenced by a number of factors9,12,15. Sedentary lifestyle, high blood glucose levels, and increased percentage of body fat have a negative influence on telomere length. The underlying mechanism have been suggested as being mediated through oxidative stress and inflammation16.\n\nExercise as a lifestyle intervention has been associated with longer leukocytes telomere length15. A study by Cherkas on 2401 subject showed that telomere length has a direct relation with increased level of physical activity17. An observation of physical activity and telomere length by Du et al. on 7813 adult women concluded that moderate and vigorous intensity activity increased telomere length compared with least active women18. Ludlow et al. studied the effect of physical activity on telomere length in three different groups: sedentary, moderate and overtraining. The result showed a positive effect on telomere length in the moderate group19.\n\nTo date, to our knowledge, very few studies have investigated the relative effect of a specific diet on telomere length. Cassidy et al. found that total fat intake was only inversely associated with leukocytes telomere length and higher polyunsaturated fatty acids (PUFA) intake, specifically linoleic acid intake, was inversely associated with leukocytes telomere length11. Li et al. found that there was no difference in telomere length between consumption of fish oil-rich diet and soy oil-rich diet20. Kiecolt-Glaser et al. compared telomere length between subjects with n-6 PUFA and n-3 PUFA supplementation. They found that telomere length is longer in subject with omega 3 or n-3 PUFA supplementation (high in fish oil) compared to n-6 PUFA supplementation21.\n\nResults from our current study showed a lengthening of relative telomere length in both groups in week 4 and week 8. This was in contrast with studies showing telomeres usually shortened with age9., but there are also studies which indicate that in vivo, telomere may shorten or elongate, and leukocyte telomere length may fluctuates within months22,23.\n\nIn general, preserved telomere length and lengthening of telomere are considered as something good because it is thought to play an important role in extending the biological age of cells12,21,24.\n\nNevertheless, studies have also shown that telomere lengthening can be an initial response that arises after exposure to low doses of various carcinogenic chemicals in vitro and in experimental animals23. Zhang et al. concluded from their study that subject with longer telomere length had a higher risk of getting lung cancer, and this was especially true for men25.\n\nThe positive associations between high fat-diet conditions and telomere length is difficult to explain because a high fat-diet is associated with increased risk of various diseases. Telomeres generally shorten with age, thus, the discovery of telomer elongation in the provision of high-fat diet can be regarded as something that deviates from normal condition. Therefore, the positive associations between high fat-diet conditions and telomere length observed in this study are notable.\n\nTelomere will shorten at each cell division. Telomeres that elongate excessively in both groups in this study may indicate a prolonged period before apoptosis, and this could indicates a change from normal cell function. Currently, the implications of excessive telomere lengthening are still unknown. Our result shows that aerobic exercise can act as a barrier to progressive changes that occur in the relative telomere length caused by a high-fat diet condition. Modulation of oxidative stress in the body is one possible mechanism that may explain how aerobic exercise resist relative telomere changes. Aerobic exercise upregulate genes that encode various antioxidant enzymes. Several studies shows that regular physical exercise increase the body's endogenous antioxidant activity and thus increase body’s resistance to oxidation events15,26.\n\n\nConclusions\n\nOur study showed that exposure to a high fat-diet plays an important role to the emergence of altered telomere length, and aerobic exercise could reduce the progression of the alteration in length. Our results support the hypothesis that leukocyte telomere length is associated with daily dietary intake and physical activities. Further investigation is still needed to explore the mechanism and implications of telomere length changes found in this study.\n\n\nData availability\n\nDataset 1: Raw data including the relative telomere length for control and treatment groups at week 0, 4 and 8, and 2-ΔΔCt calculations. DOI, 10.5256/f1000research.15127.d21168127.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was funded by Hibah Publikasi Internasional Terindeks untuk Tugas Akhir (PITTA) 2017.\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\nArroyo-Johnson C, Mincey KD: Obesity Epidemiology Worldwide. Gastroenterol Clin North Am. 2016; 45(4): 571–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nProspective Studies Collaboration, Whitlock G, Lewington S, et al.: Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet. 2009; 373(9669): 1083–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerrington de Gonzalez A, Hartge P, Cerhan JR, et al.: Body-mass index and mortality among 1.46 million white adults. N Engl J Med. 2010; 363(23): 2211–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRachmi CN, Li M, Baur LA: Overweight and obesity in Indonesia: prevalence and risk factors-a literature review. Public Health. 2017; 147: 20–9. PubMed Abstract | Publisher Full Text\n\nRauner A, Mess F, Woll A: The relationship between physical activity, physical fitness and overweight in adolescents: a systematic review of studies published in or after 2000. BMC Pediatr. 2013; 13: 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nColberg SR, Sigal RJ, Fernhall B, et al.: Exercise and type 2 diabetes: the American College of Sports Medicine and the American Diabetes Association: joint position statement. Diabetes Care. 2010; 33(12): e147–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaßenroth D, Meyer A, Salewsky B, et al.: Sports and Exercise at Different Ages and Leukocyte Telomere Length in Later Life--Data from the Berlin Aging Study II (BASE-II). Saretzki G, editor. PLoS One. 2015; 10(12): e0142131. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBotha M, Grace L, Bugarith K, et al.: The impact of voluntary exercise on relative telomere length in a rat model of developmental stress. BMC Res Notes. 2012; 5(1): 697. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDenham J, O’Brien BJ, Charchar FJ: Telomere Length Maintenance and Cardio-Metabolic Disease Prevention Through Exercise Training. Sports Med. 2016; 46(9): 1213–37. PubMed Abstract | Publisher Full Text\n\nGoglin SE, Farzaneh-Far R, Epel ES, et al.: Change in Leukocyte Telomere Length Predicts Mortality in Patients with Stable Coronary Heart Disease from the Heart and Soul Study. Wellinger RJ, editor. PLoS One. 2016; 11(10): e0160748. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCassidy A, De Vivo I, Liu Y, et al.: Associations between diet, lifestyle factors, and telomere length in women. Am J Clin Nutr. 2010; 91(5): 1273–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFreitas-Simoes TM, Ros E, Sala-Vila A: Nutrients, foods, dietary patterns and telomere length: Update of epidemiological studies and randomized trials. Metabolism. 2016; 65(4): 406–15. PubMed Abstract | Publisher Full Text\n\nCawthon RM: Telomere measurement by quantitative PCR. Nucleic Acids Res. 2002; 30(10): e47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Sullivan RJ, Karlseder J: Telomeres: protecting chromosomes against genome instability. Nat Rev Mol Cell Biol. 2010; 11(3): 171–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArsenis NC, You T, Ogawa EF, et al.: Physical activity and telomere length: Impact of aging and potential mechanisms of action. Oncotarget. 2017; 8(27): 45008–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim S, Parks CG, DeRoo LA, et al.: Obesity and weight gain in adulthood and telomere length. Cancer Epidemiol Biomarkers Prev. 2009; 18(3): 816–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCherkas LF, Hunkin JL, Kato BS, et al.: The association between physical activity in leisure time and leukocyte telomere length. Arch Intern Med. 2008; 168(2): 154–8. PubMed Abstract | Publisher Full Text\n\nDu M, Prescott J, Kraft P, et al.: Physical activity, sedentary behavior, and leukocyte telomere length in women. Am J Epidemiol. 2012; 175(5): 414–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLudlow AT, Zimmerman JB, Witkowski S, et al.: Relationship between physical activity level, telomere length, and telomerase activity. Med Sci Sports Exerc. 2008; 40(10): 1764–71, [cited 2018 May 10]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Y, Zhao F, Wu Q, et al.: Fish oil diet may reduce inflammatory levels in the liver of middle-aged rats. Sci Rep. 2017; 7(1): 6241, [cited 2018 May 14]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiecolt-Glaser JK, Epel ES, Belury MA, et al.: Omega-3 fatty acids, oxidative stress, and leukocyte telomere length: A randomized controlled trial. Brain Behav Immun. 2013; 28: 16–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShin JY, Choi YY, Jeon HS, et al.: Low-dose persistent organic pollutants increased telomere length in peripheral leukocytes of healthy Koreans. Mutagenesis. 2010; 25(5): 511–6. PubMed Abstract | Publisher Full Text\n\nDuan X, Yang Y, Wang S, et al.: Changes in the expression of genes involved in cell cycle regulation and the relative telomere length in the process of canceration induced by omethoate. Tumour Biol. 2017; 39(7): 1010428317719782. PubMed Abstract | Publisher Full Text\n\nRoake CM, Artandi SE: DNA repair: Telomere-lengthening mechanism revealed. Nature. 2016; 539(7627): 35–6. PubMed Abstract | Publisher Full Text\n\nZhang C, Doherty JA, Burgess S, et al.: Genetic determinants of telomere length and risk of common cancers: a Mendelian randomization study. Hum Mol Genet. 2015; 24(18): 5356–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPowers SK, Radak Z, Ji LL: Exercise-induced oxidative stress: past, present and future. J Physiol. 2016; 594(18): 5081–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSantoso DIS, Paramita N, Prijanti AR, et al.: Dataset 1 in: The effect of aerobic exercise on relative leukocyte telomere length in male Sprague-Dawley rats given a high fat-diet. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15127.d211681"
}
|
[
{
"id": "36736",
"date": "13 Aug 2018",
"name": "Ronny Lesmana",
"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\nStudy is interesting, however there are some points need to be added:\nTraining protocol should followed and compared with the protocol of the aerobic exercise in previous report, please cross check and add reference (lesmana et al, 20171). Please present the change in level of profil lipid Discuss more about factor change and induce telomere characteristic Reasons the using high fat diet connected to the change telomere.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "40574",
"date": "27 Nov 2018",
"name": "Andrew T. Ludlow",
"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\nSantoso et al. explore the relationship between a high fat diet, aerobic exercise and telomere length. In this small-scale animal research study, the authors make the unexpected observation that after 8 weeks of high fat diet, telomeres are elongated in circulating blood cells (leukocytes). They observed in aerobically trained animals that the apparent telomere elongation was significantly attenuated compared to a sedentary control. These interesting findings observations call into question the strict dogmatic view that telomere shortening is the rule in pathology. Human epidemiological studies have reported similar discrepancies, where telomere shortening is not always observed in pathological scenarios but rather telomere elongation can also be apparent in disease situations. While these observations are provocative, I have several concerns with the sample size, the model organism, the method used to measure telomere length, and the interpretation of the data.\n\nThis study is very positive in terms that it points out the need to perform longitudinal analysis of telomere biology in both human and model organism studies. Repeat sampling will be critical to moving the field of telomere biology and environmental/lifestyle impacts forward. For instance, what happens to telomere biology or cell biology after 12 weeks on a high fat diet following the apparent rapid and massive telomere elongation event at 8 weeks?\n\nThe negatives of this study are the model organism of choice, the measurement biochemistry, the data analysis, and the lack of mechanistic insights provided by the authors.\n\n1. Sample size, model organism, and statistical treatment of the data.\nWhile the effect observed in this study is large, the sample size is small. Further the authors utilize a repeated-measures design and do not analyze the data as such. Each rat’s telomere length was measured three times. Thus, change in telomere length per individual rat should be reported between each time point and analyzed via a repeated measures analysis. Further, with the assay of choice, 6 animals would be borderline powered enough to detect a meaningful change. What were the authors expectations in terms of a priori predictions on the magnitude of change in telomere length?\n\nAre the rats inbred or outbred? This needs to be reported in the document.\n\nTranslatability of findings to humans. Rats use a different life history/ life course strategy to promote survival of the species. These animals grow quickly, have a small body size, do not suppress telomerase, reproduce young, and are cancer susceptible while humans grow slowly, have a large body size, suppress telomerase, reproduce late in life, and are cancer resistant. Thus, could this extreme lengthening be due to evolutionary differences between humans and the chosen model organism? That is, humans might not have a similar response compared to rodents.\n\n2. Assay choice.\nThe Cawthon method or T/S ratio to determine relative telomere length is a valid assay when performed by certain labs, but it can have issues. A critical validation step in the assay that I did not see in the online data is the reference DNA sample. Did the authors have a reference sample of known telomere length that was run on each plate to ensure their assay was performing as expected? This should have been done on each plate and the inter and intra plate coefficient of variation should be reported for the assay in general. This is critical to be sure the findings are not erroneous. The authors could purchase a cell line with a known telomere length, isolate DNA and repeat the T/S ratio on all samples (both known and unknowns/experimental) to ensure the measurement is correct.\n\nI would have expected higher T values across the board. Rats have 25-40 kb telomeres and a single copy of 36B4. Thus, the ratio of telomere copy to 36B4 copy should have been more like 2 or 3, not 1.\n\n3. Interpretation concerns.\nWhat is known about rapid telomere elongation to my knowledge is as follows: E. Blackburn’s group showed that yeast mutants for telomerase RNA result in ultra-elongated telomeres rapidly and impair cellular survival. Further, some mutants acted as a ‘time-bomb’ and killed cells later in life (slow on set of telomere elongation 1). In human cancer cells telomere elongation induces cell differentiation and less aggressive tumors 2. Another line of evidence from the L. Harrington lab is that telomere elongation reduced cell survival after radiation treatment 3. From epidemiological evidence, telomere trajectory in CVD can be maintained, shorten or elongated in follow up 4. Whether or not these changes correlate to health outcomes or are correlated with secondary CVD events is unknown. Thus, it appears that ultra-long telomere lengths (greater than 15 kb in human) could be equally as ‘bad’ as short telomeres for cell fate decisions. The major cellular and molecular issues of long telomeres are stalled replication forks at telomeres and double strand breaks in the telomeres that a refractory to repair. Do the authors suspect that the immune cells with ultra-long telomeres will die soon after the 8-week period? Is there a major shift in immune cell distribution after 8 weeks on a high fat diet that would indicate a lasting impact of this observation? Are the cells with long telomeres dying?\n\nThe concept that telomere length must be maintained at the correct telomere length and that telomere structure and sensing by the cell may be more important that absolute length is an emerging concept in the field. The results from the current study lend to the hypothesis that absolute length is not critical but rather tight regulation and sensing of telomere length is critical to cellular health and in turn gene expression regulation and cellular physiology. This idea needs and deserves more investigation as much of our evidence to date comes from cells grown in culture which may be very different than cells in vivo when it comes to telomere biology.\n\n4. Short duration and extreme impact.\nMy hypothesis would have been no or very subtle changes in telomere length for this duration of intervention. What are the speculative mechanisms the authors see occurring? Are the leukocytes entering a stress mode whereby they activate telomerase and elongate telomeres dramatically to try to survive? Could this be the beginning of alternative lengthening of telomeres in these cells? Can the author speculate a bit about what they think might be happening in terms of mechanisms?\n\nIssue of timing – in terms of physiology – after 8 weeks on a high fat diet what else is changed in the immune system of these animals? Can the authors provide about how long 8 weeks in a rats’ life would be equivalent to how long on a high fat ‘western’ diet for humans would be?\n\nWhat other physiological changes occur? Fatty liver? Higher cholesterol? Higher fasting blood glucose? Hormone levels (epi, nor-epi, insulin, IGF-1, glucagon, cortisol, growth hormone, TNF alpha, IL-6, IL-10, etc.). All of these measures could identify correlates to the change in telomere length.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1143
|
https://f1000research.com/articles/7-1138/v1
|
25 Jul 18
|
{
"type": "Research Article",
"title": "Effect of antenatal care on low birth weight prevention in Lao PDR: A case control study",
"authors": [
"Latsamy Oulay",
"Wongsa Laohasiriwong",
"Teerasak Phajan",
"Supat Assana",
"Kritkantorn Suwannaphant",
"Latsamy Oulay",
"Teerasak Phajan",
"Supat Assana",
"Kritkantorn Suwannaphant"
],
"abstract": "Background: Low Birth Weight (LBW) is a worldwide public health problem, which subsequently may affect the health status of the child. Lao PDR has high incidence of LBW. Antenatal care (ANC) is provided to improve maternal and child health outcomes. The aim of this study was to identify the effect ANC on LBW prevention in Lao PDR. Methods: This case control study was conducted in tertiary hospitals of Lao PDR. The ratio of case: control was 1:3, of which there were 52 cases and 156 controls that passed the inclusion criteria included in the study. In our analysis information on pregnancy and ANC including height of mother, maternal weight gain during pregnancy, maternal gestational age at delivery, type of delivery, supplementary vitamins, and other covariates including age, marital status, educational attainment, occupation, family income, health insurance, family size and living condition were described and determine their association with LBW using multiple logistic regression analysis. Results: There were only 32.69 % of complete ANC among cases and 57.69% in control. Incomplete ANC (<4 times) were significant increased the odds of having LBW (adj. OR=2.97; 95%CI: 1.48 to 5.93; p-value =0.002). Other covariates which also influenced LBW were having maternal weight gain during pregnancy less than 10 kg. (adj.OR=2.28; 95%CI: 1.16 to 4.49; p-value = 0.017), maternal gestation age at delivery less than 40 weeks (adj. OR=3.33; 95%CI: 1.52 to 7.32; p-value =0.003). Conclusion: Complete ANC could help both mother and child in term of weight gain and full term delivery which may effect on LBW reduction.",
"keywords": [
"Infant",
"Low Birth Weight (LBW)",
"Maternal Risk Factors",
"Case – Control Study."
],
"content": "Introduction\n\nLow birth weight (LBW) refers to a baby who has a birth weight of le<2,500 grams1. LBW may cause birth asphyxia, amniotic fluid aspiration, hypoglycemia and hyponatremia. An infant weight of 1,500–2,500 grams has been shown to have 5–10 times higher mortality rate than normal infants2. During 2005–2010, LBW incidences in countries of the Association of Southeast Asian nations ASEAN were 21.0% in Philippines, 11.0% in Malaysia, Cambodia and Lao PDR, 9.0% in Indonesia, 7.0% in Thailand and 5.0% in Vietnam3. The incidence of LBW reflects a country’s socio-economic development4. Mortality of LBW babies is as high as 1% when compared with 0.2% among normal children5. At the 2005 World Summit a plan was announced to improve quality of life called Millennium Development Goals (MDGs)\", covered goals to be achieved worldwide by the year 20155.\n\nMaternal and child health problems have been considered as indicators of the health service performances. In Southeast Asia, 28% of all deaths of 1 month-old infants were from infection, and 20% from preterm birth and LBW6. LBW is commonly used as an indicator of health status and is important for national health policy development7. Important factors associated with LBW are maternal factors, such as socioeconomic status, food consumption behaviors, calorie intake, urinary tract infection and prenatal care, smoking, genital infections, maternal health and stress8,9.\n\nAntenatal care (ANC) is care provided to pregnant women by health personal. Care includes risk identification, prevention and management of pregnancy-related or concurrent disease, and health education and health promotion10. The World Health Organization’s ANC model, also known as focused or basic ANC is a goal-orientated approach to delivering evidence-based interventions at four critical times for ANC during pregnancy11; therefore completed ANC in this study is having ANC for at least 4 times.\n\nHigh incidence of LBW and high mortality of both mothers and children in Lao PDR has been hypothesized as being caused by various influencing factors, such as low socio economic status and limited access to qualified health services related to pregnancy12. However, there are limited studies identifying the role of ANC in reducing LBW incidence in Lao PDR. Therefore this study aimed to determine whether having ANC at least four times during pregnancy could help reduce LBW when controlling for other covariates related to socioeconomic and pregnancy factors. The results could be used to develop appropriate measures for prevention of LBW and obtaining better maternal and child health statuses.\n\n\nMethods\n\nPostpartum mothers who had delivered babies and came for regular checkup after delivery in four tertiary hospitals in Vientiane, Lao PDR between July and December 2016 were included in this study. The four hospitals were Mahosot Hospital, Sethathirat Hospital, Mothers and Child Center Hospital, and Mittraphab Hospital.\n\nThe sample size was calculated using the formula for the analysis of a relationship in a case-control study13. The formula indicated that the sample size of the study group (cases) should be 52 participants. The control group was 3 times the size of the study group (case:control ratio of 1:3). Therefore, the control group included 156 participants, which made a final total of 208 participants. Cases and controls were not matched for demographics.\n\nInclusion criteria were mothers who had delivered babies in the four hospitals during the study period, who were 18 – 49 years of age. This age range was chosen as the reproductive age of women is 15–49 years old. However, those aged below 18 are considered as a vulnerable group, therefore we selected 18–49 year-olds. The exclusion criteria were mothers who delivered twins, did not live in the study area and who were not willing to participate. The samples in this study were divided into two groups:\n\n1) Case group: mothers of babies whose birth weight was <2,500 grams (LBW);\n\n2) Control group: mothers of babies whose births weight was ≥2,500 grams who were born during the same period as the cases.\n\nThe tool for data collection was a structured questionnaire interview that consisted of seven parts (Supplementary File 1): Part I, general information; Part II, sociodemographic characteristics; Part III, knowledge of health care during pregnancy; Part IV, maternal factors and pregnancy status; Part V, environment factors and support for ANC; Part VI, prenatal distress (as assessed using the Edinburgh Prenatal Depression Scale); Part VII, obstetric information at delivery (gathered from the mother’s medical records).\n\nThe questionnaire was content validity tested by five experts in terms of theory and understanding. Unclear questions were edited and some information that was missing was added. Reliability was tested among 30 mothers (from Xaythany and Sisattanak Hospitals), indicating the high reliability with the Cronbach’s alpha coefficient of 0.84.\n\nData was collected by four physicians who were trained in using the questionnaire. These interviewers were blinded for the infant status of normal or LBW. The interview of the participants took place in a postpartum patient room or nursing room within 5 days of agreement to participate.\n\nSTATA version 10.0 was used to analyze data14. Descriptive statistics frequency, percentage, means, standard deviation, minimum, maximum were used to present data on the following characteristics: characteristics of mothers, knowledge on health care during pregnancy of the mother, pregnancy status, environment factors and support for ANC, prenatal distress and obstetric information at delivery.\n\nSimple logistic regression was used to determine factors associated with LBW. The factors which had association with LBW (p-value <0.25) were analyzed using multiple logistic regression. Multiple logistic regression was applied to identify the association of ANC and LBW when controlling for other covariates, presenting adjusted OR, 95% confidence interval (95% CI) with the levels of significance at 0.0515.\n\nWe submitted the proposal and questionnaire to the Ethical Committee of Khon Kaen University, Thailand (reference No. HE 592087) and the Ethical Committee of University of Health Sciences, Lao PDR (reference No.13/16) for approval. Both committees granted approval of this study, including the reliability test of the questionnaire. Written informed consent was obtained from all participants.\n\n\nResults\n\nThe study involved a total of 208 mothers, of which 52 were cases and 156 were controls. There were a higher proportion of younger mothers (<20 years) among cases than controls (17.31% and 11.54%, respectively). Controls had higher educational attainment of upper secondary (43.59%) compared to cases (28.85%). More controls lived in urban areas than cases (60.26% and 51.92%, respectively), and 21.15% of cases had average monthly family income ≤1,000,000 (Kips) which was 9.62% among controls (see Table 1).\n\nOnly 32.69 % of cases completed ANC (≥ 4 times) check-ups while 57.69% controls completed these. The bivariable analysis of ANC, socioeconomic and maternal factors with LBW showed that going to ANC check-ups <4 times had higher odds of having LBW than those who went to at least 4 times (OR=2.80; 95%CI: 1.44 to 5.43; p-value = 0.002).\n\nIn the bivariate analysis, which considers the association of one independent variable with the outcome (LBW), we selected factors with a p-value <0.25 to proceed to the multivariable analysis. These factors were height of mother (cm), taking any supplementary vitamins, prenatal depression, maternal weight gain during pregnancy (kg), maternal gestational age during delivery (weeks), and type of delivery (see Table 2).\n\nThe multivariate analysis of maternal risk factors of LBW show that after adjusting for the effect of covariates there was significant association between ANC check-ups <4 times and LBW (adjusted OR = 2.97; 95% CI: 1.48 to 5.93; p-value =0.002). Other covariates that were also significantly associated with LBW were maternal weight gain during pregnancy <10 kg (adjusted OR =2.28; 95% CI: 1.16 to 4.49; p-value = 0.017), maternal gestation at delivery <40 weeks (adjusted OR= 3.33; 95% CI: 1.52 to 7.32; p-value =0.003) (see Table 3).\n\n\nDiscussion\n\nOur study demonstrated that inadequate ANC, poor maternal weight gain during pregnancy and maternal gestational age at delivery were significant independent determinants of LBW in Lao PDR. Only about one third of the mothers with LBW babies (cases) had completed ANC (≥4 times), whereas about half of the mothers with normal weight babies (control) had completed ANC. This finding supports the results of other studies16,17, which indicated that ANC (times) were found to be significant maternal risk factors for LBW babies. In this study, the mothers who attended ANC fewer than 4 times had almost 3 times higher odd of having LBW. Previous studies in general hospitals have indicated that ANC visits <4 times were LBW risk factors18–21. Antenatal visits of the pregnant mothers are very important as they provide chances for monitoring the fetal wellbeing and allow timely intervention for feto-maternal protection. Little ANC could increase prenatal fetor-maternal complications22. In the present study, among mothers who did not receive ANC, there was 1.122 times higher changes than those who received ANC23 and ANC < 4 visits OR was 1.41 (95% CI: 1.02 to 1.94). These are similar findings to results found in Thailand24–26.\n\nMaternal age in some studies has no significant association with LBW19; however other studies had different results27,28. For example, Fariha et al. found that maternal age was found significantly associated with LBW infants29, and some studies reported that the older the mother, the higher the risk of having LBW infants30. LBW babies among older mothers, whose age is 35 years and above, was 23.89%. It was significantly higher than the percentage of LBW babies for mothers in other age categories (p = 0.004)16,31. For maternal weight gain during pregnancy it was found that weight gain <10 kg were risk factors for LBW in the present study, which is similar to the result of a study in Indonesia among others26,32–34. Low maternal weight gain reflects poor child growth, which puts both mother and child at risk for morbidity and mortality35,36.\n\nMaternal gestational age at delivery of <40 weeks was another associated factor for LBW in the present study, which is similar to a result found in Malaysia37. Some studies showed that the risk factor of LBW infants was a gestational ages of <37 weeks24,38. When the mother delivers a baby before the baby is at full term, the baby is not fully grown. Therefore the babies are more likely to be small (LBW) and have higher risk for mortality since some organs such as lung is not fully functioning.\n\nOne limitation in this case-control study could be data collection bias due to interviewer prejudices. However, we minimized this by blinding the interviewers; therefore the manner in which they asked the questions were the same in both case and control groups.\n\n\nConclusion\n\nThis hospital-based case-control study was conducted in Vientiane, Lao PDR and indicated that ANC checkups at least 4 times could help reduce LBW of babies. Consequently, policy should improve coverage and quality of ANC of at least 4 times for all pregnant women in this population.\n\n\nData availability\n\nDataset 1: Raw data supporting the results presented. Since this study did not analyse knowledge of health care of the mothers or environmental factors and support of ANC, answers to Parts III and V of the questionnaire have not been included in the dataset. DOI, 10.5256/f1000research.15295.d21014839",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by the China Medical Board Project and the University of Health Sciences, Vientiane, Lao PDR; and the Faculty of Public Health, Khon Kaen University.\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 are thankful to all participants, directors and heads of the Department of Obstetrics and Gynecology in participating hospitals.\n\n\nSupplementary material\n\nSupplementary File 1: Questionnaire used in the study in English.\n\nClick here to access the data.\n\n\nReferences\n\nWorld Health Organization: International statistical classification of disease and related health problems, tenth revision. World health Organization, Geneva. 1992. Reference Source\n\nManji KP, Massawe AW, Mgone JM: Birthweight and neonatal outcome at the Muhimbili Medical Centre, Dar es Salaam, Tanzania. East Afr Med J. 1998; 75(7): 382–7. PubMed Abstract\n\nWardlaw TM: Low birth weight: country, regional and global estimates. United Nations Children’s Fund and World Health Organization, UNICEF, New York. 2004. Reference Source\n\nWorld Health Organization: Neonatal and perinatal mortality: Country, regional and global estimates 2004. Geneva, WHO. 2004. Reference Source\n\nWorld Health Organization: Maternal mortality in 2005. Estimates developed by WHO UNICEF, UNFPA and the World Bank. Geneva. 2007. Reference Source\n\nOECD/WHO: Health at a Glance: Asia/Pacific 2016: Measuring Progress towards Universal Health Coverage. Paris. 2016. Publisher Full Text\n\nUnited Nations: The millennium development goals report 2015. United Nations. 2015. Publisher Full Text\n\nDeshpande JD, Phalke DB, Bangal VB, et al.: Maternal risk factors for Low birth weight neonates: A hospital based case-control study in rural area of Western Maharashtra, India. National J Comm Med. 2011; 2(3): 394–8. Reference Source\n\nMetgud CS, Naik VA, Mallapur MD: Factors affecting birth weight of a newborn--a community based study in rural Karnataka, India. PLoS One. 2012; 7(7): e40040. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: WHO recommendations on antenatal care for a positive pregnancy experience. World health Organization, Geneva. 2016. Reference Source\n\nWorld Health Organization: WHO antenatal care randomized trial: manual for the implementation of the new model. World health Organization, Geneva. 2002. Reference Source\n\nLao PDR: Lao Social Indicator Survey (LSIS) 2011-2012 (Multiple indicator cluster survey /demographics and health). Ministry of Health and Lao Statistics Bureau. 2010. Reference Source\n\nSchlesselman JJ: Case control studies: design, control, analysis. Oxford University Press-Monographs in epidemiology and biostatistics. New York. 1982. Reference Source\n\nStata Corp: Stata statistical software: Release 10. College Station,Texas. 2007. Reference Source\n\nDirek L: Data Analysis by STATA program. Chulalongkorn University Press (CUP). Bangkok, Thailand. 2011.\n\nMumbare SS, Maindarkar G, Darade R, et al.: Maternal risk factors associated with term low birth weight neonates: a matched-pair case control study. Indian Pediatr. 2012; 49(1): 25–28. PubMed Abstract | Publisher Full Text\n\nAnant P, Durgesh K: Maternal factors associated with low birth weight: a case control study in rural Kerala. Int J Community Med Public Health. 2017; 4(10): 3793–3795. Publisher Full Text\n\nNegandhi PH, Negandhi NH, Zodoey SP, et al.: Risk factors for low birth weight in an Indian urban setting: a nested case control study. Asia Pac J Public Health. 2014; 26(5): 461–9. PubMed Abstract | Publisher Full Text\n\nKositkulaporn S: Antenatal care and self-care behaviors of pregnant women who received services at health promoting hospital. Regional Health Promotion Center 11, Nakhon Si Thammarat Province. 2014.\n\nKader M, Perera NK: Socio-economic and nutritional determinants of low birth weight in India. N Am J Med Sci. 2014; 6(7): 302–308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeron M, Sutton PD, Xu J, et al.: Annual summary of vital statistics: 2007. Pediatrics. 2010; 125(1): 4–15. PubMed Abstract | Publisher Full Text\n\nUnited Nations Children’s Fund and World Health Organization: Low birth weight country, regional and global estimates. UNICEF, New York. 2004. Reference Source\n\nMadiha KM, Shumaila Y, Ayesha I, et al.: Maternal anemia is a risk factor for low birth weight babies at term. PJMHS. 2016; 10(3): 741–743. Reference Source\n\nPhonglopisit S: Maternal risk factors of low birth weight infant at Thoen Hospital, Lampang. Med J. 2009; 30(3): 146–153. Reference Source\n\nTewelde G, Kahsay G, Endeshaw A, et al.: Determinants of low birth weight among mothers who gave birth in Debremarkos Referral Hospital, Debremarkos Town, East Gojam, Amhara Region, Ethiopia. Journal of Neonat Pediatr Med. 2018; 4(1): 145. Reference Source\n\nPrakong T: Related Factors of Low Birth Weight Infant. Journal of Yala Rajabhat University. 2011; 6(2): 113–122. Reference Source\n\nOladeinde HB, Oladeinde OB, Omoregie R, et al.: Prevalence and determinants of low birth weight: the situation in a traditional birth home in Benin City, Nigeria. Afri Health Sci. 2015; 15(4): 1123–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhaskar RK, Deo KK, Neupane U, et al.: A Case Control Study on Risk Factors Associated with Low Birth Weight Babies in Eastern Nepal. Int J Pediatr. 2015; 2015: 807373. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFariha A, Tahir J, Muhammad FA, et al.: Maternal risk factors associated with low birth weight: A case control study. Annals. 2011; 17(3): 223–228. Reference Source\n\nGanesh Kumar S, Harsha Kumar HN, Jayaram S, et al.: Determinants of low birth weight: a case control study in a district hospital in Karnataka. Indian J Pediatr. 2010; 77(1): 87–9. PubMed Abstract | Publisher Full Text\n\nMichael OF, Iddrisu AR, Riskatu Y: Maternal risk factors for low birth weight in a district hospital in Ashanti Region of Ghana. Research in Obstetrics and Gynecology. 2013; 2(4): 48–54. Reference Source\n\nEdi P, Sirikul I, Jiraporn C: Maternal risk factors for low birth weight infant at Famawati General Hospital, Jakarta, INDONESIA. Journal of Public Health and Development. 2008; 6(1): 123–133. Reference Source\n\nCita YP, Resmiati N: The relationship between demographic factors and low birth weight infants. Int J Res Nurs. 2010; 1(1): 25–28. Publisher Full Text\n\nAgarwal K, Agarwal A, Agarwal VK, et al.: Prevalence and determinants of \"low birth weight\" among institutional deliveries. Annals of Nigerien Medicine. 2011; 5(2): 48–52. Publisher Full Text\n\nUsawadee J, Piyathida K, Orawun N: Pregnancy factor affects low birth weight at Nong Bua Lum Phu Hospital, Nong Bua Lum Phu Province. Graduate Research Conference, Khon Kaen University. 2014; 1791–1800. Reference Source\n\nNahid K, Saeid S, Samira A, et al.: Predictors of low birth weight infants in the North West Province of Iran: a Case-control study. Int J Pediatr. 2016; 4(6): 1983–1991. Publisher Full Text\n\nRosnah S, Mazlina M, Aimi NM, et al.: Determinant of low birth weight infants: a matched case control study. OJPM. 2014; 4(3): 91–99. Publisher Full Text\n\nNittaya P: Maternal risk factors of low birth weight infants in Khumuang District, Buriram Province. J Health Sci. 2017; 26(2): 346–352. Reference Source\n\nOulay L, Laohasiriwong W, Phajan T, et al.: Dataset 1 in: Effect of antenatal care on low birth weight prevention in Lao PDR: A case control study. F1000Research. 2018. Data Source"
}
|
[
{
"id": "36472",
"date": "03 Aug 2018",
"name": "Metha Songthamwat",
"expertise": [
"Reviewer Expertise Ob-Gyn"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article is a case control study, aims to study the association between the complete antenatal care (ANC > 4 times) and the low birth weight (LBW) in Lao PDR. It demonstrated the significance of ANC on the prevention of neonatal complication however there are a number of issues with the methods and analysis section that need to be clarified with some of the conclusion are not correspond with the results. The detail of comments by section is below:\nAbstract: The conclusion that complete ANC help mother and child in term of weight gain and full term delivery is not from the analysis of the results. The multiple logistic regression from this study only showed that antenatal care and maternal weight gain and GA at delivery had effect to the LBW.\nIntroduction:\nPage 3, Paragragh 1– second sentence birthweight of le< 2500 grams, the reference about LBW incidence in ASEAN was in 2004 that should be update. The last sentence about MDGs in 2015 is out of date and irrelevant which should be removed. Paragraph 2- some sentences are repeat of first paragraph Paragraph 3- four critical times of ANC was recommended from WHO in 2002, the update of recommendation from WHO is in 2016 which can be found in http://www.who.int/reproductivehealth/publications/maternal_perinatal_health/anc-positive-pregnancy-experience/en/ ,the new guideline recommend minimal 8 times of antenatal care to reduce perinatal mortality (page xvi) Paragraph 4- the authors mentioned there are limited studies identifying the role of ANC in reducing LBW incidence in Lao PDR, please give the detail, reference and gap of knowledge.\n\nMethods: The method was well described, however some points should be clarify\nThe study period was 6 months in four tertiary hospitals in capital city which mean a lot of patients, with the only 52 cases of LBW and 156 control cases, the authors should mentioned about the selection method especially the control cases to avoid the selection bias. Some pregnancy factors that affect the LBW such as associated maternal diseases, complication during pregnancy, gravida and preterm birth should be added.\nResults:\nThe subjects were caesarean delivery in 90.38% of cases and 79.49% of control, 19.23% of cases and 13.46% of controls were government officer with only 1.92% of cases and 3.21% of controls were farmer. This might not be the good representative of Lao PDR pregnant women and the inferential of this study to general population might be limited. Paragraph 1-The control showed higher education, urban and higher income. This might cause from the convenience selection of control group and cause the better health status, more times ANC attend and higher birthweight of control group. Please clarify the method of control selection. Us dollar might be better than Kips for the understanding of the reader. Total birth, total LBW, total preterm birth and total cesarean deliveries in the study period should be added. Table 1- Maternal age should be 18-20 instead of <20 (<18 was excluded from this study), Famer should be changed to farmer. Only mean or median should be selected depend on the distribution of data. If median was selected, the min-max should be changed to interquartile range. Maternal medical and obstetrics complications should be added because it was an important factor for LBW. Table 2-The gestational age normally divided to <37, 37-<42, > 42 instead of 40. The BMI should be used instead of maternal height and why 10 kg was used instead of the recommended weight gain during pregnancy which depends on maternal BMI. Very high cesarean section in both groups was shown, please add the reasons or indications.\n\nThe continuous data such as age, income should be analyzed in continuous form, especially income which was statistical difference between two groups if analyzed in continuous form and should be analyzed in multiple logistic regression analysis. The downgrade of continuous data to ordinal data should be avoided. Normally, the premature birth is the birth before 37 weeks GA and is the most important cause of LBW. This should be used instead of 40 weeks.\n\nConclusion: As mentioned above, the new WHO recommendation in 2016 recommend at least 8 times of antenatal care to reduce perinatal mortality.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "42672",
"date": "09 Jan 2019",
"name": "Zohra S. Lassi",
"expertise": [
"Reviewer Expertise Public Health",
"Perinatal Epidemiology",
"Adolescent Health",
"Global Health"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMajor:\nPoorly written paper and needs English editing throughout. Many of the sentences are not clear and I could not follow many things because of grammatical errors. Abstract results need to be rewritten correctly. For the calculation of sample size, what indicators were used? Also provide the formula. Was the questionnaire by those physicians completed in the hospital premises? Although you did say those physicians were blinded but if they filled the questionnaire in health facilities then those physicians could have accessed the medical record data? This may add bias to the study. Facts such as women in cases who were young, lived in non urban areas, and had lower monthly family income were not discussed in the discussion section. Reference 2 is from 1998 and from Tanzania. Should use an updated citation and global figure. Reference 3 is very old (from 2004). Reference 5 indicates it is on maternal mortality. However, authors have used it for LBW burden. Please check if this is correctly cited?\nMinor: Several minor points has been raised in the attached.\nI would suggest changing the title to “Does antenatal care has any impact on reducing low birth weight. – a case-control study.” Several grammatical errors, would suggest getting a good proof for language and grammar. Abstract: suggest changing the word “problem” to concern. “Background: Low Birth Weight (LBW) is a worldwide public health problem..” Abstract: suggest changing this sentence from: Antenatal care (ANC) is provided to improve maternal and child health outcomes.” to “Women are offered antenatal care (ANC) to improve maternal, pregnancy and newborn outcomes”. Introduction: second line “le” should be “less than” Rewrite this sentence: “During 2005–2010, LBW incidences in countries of the Association of Southeast Asian nations ASEAN were 21.0% in Philippines, 11.0% in Malaysia, Cambodia and Lao PDR, 9.0% in Indonesia, 7.0% in Thailand and 5.0% in Vietnam 3.” Whenever you are saying compared with normal children, it should be normal weight babies.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1138
|
https://f1000research.com/articles/7-1135/v1
|
25 Jul 18
|
{
"type": "Research Article",
"title": "Molecular characterization of Pseudomonas aeruginosa isolates from Sudanese patients: A cross-sectional study",
"authors": [
"Reem H. Amoon",
"Amna H. Abdallha",
"Ahmed Osman Sharif",
"Ehssan H. Moglad",
"Hisham N. Altyb",
"Salaheldein G. Elzaki",
"Mohamed A. Salih",
"Reem H. Amoon",
"Amna H. Abdallha",
"Ahmed Osman Sharif",
"Hisham N. Altyb",
"Salaheldein G. Elzaki",
"Mohamed A. Salih"
],
"abstract": "Background: 16S rRNA gene sequence analysis is a robust tool for characterization of new pathogens in clinical specimens with suspected bacterial disease. The aim of this study was to characterize Pseudomonas aeruginosa isolated from clinical specimens by sequencing the 16S rRNA gene. Methods: Forty bacterial isolates were obtained from different clinical specimens (wound, urine and sputum) using enrichment selective media and biochemical tests to characterize and identify the bacteria as P. aeruginosa. DNA was extracted from P. aeruginosa using the Chelex method. A universal primer was used to amplify 16S rRNA genes by a conventional PCR technique. The amplified PCR products were sequenced, and the sequences were viewed by Finch TV program version 1.4.0. The identity and similarity of the nucleotide sequence of the isolated strains was detected by comparing them with published sequences using BLASTn. Phylogenetic trees were constructed using Phylogeny.fr software. Results: Sequence analysis by BLASTn displayed high similarity and identity with P. aeruginosa from China KX461910, Australia JN609194 and with other P. aeruginosa isolates from the GenBank database. Conclusions: Our observation of isolates from different origin sites, further show the utility of 16s rRNA PCR amplification. This reveals the high specify of the primers and accuracy of the PCR. Thus, 16S rRNA sequencing can be used to identify genetically atypical P. aeruginosa isolates from different origins.",
"keywords": [
"Pseudomonas aeruginosa",
"16S rRNA",
"Alignment",
"PCR",
"phylogenetic tree",
"Blast",
"Sudan."
],
"content": "Introduction\n\nPseudomonas aeruginosa is a gram-negative bacterium that is found widely in the environment and engages in various forms of interactions with eukaryotic host organisms. It is an opportunistic pathogen that is widely spread in humans, giving rise to a broad spectrum of infections in community and healthcare facilities1,2. Due to the extended spread of P. aeruginosa habitat, the control of the organism in a hospital setting is very difficult, and makes it practically impossible to prevent contamination3. The major threat is the infection of patients who are immunocompromised or those in burns, neonatal and cancer wards4,5. Infection of P. aeruginosa is still one of the main causes of death among the critically ill and patients with impaired immune systems in spite of the development of newer and stronger antibiotics.\n\nSequencing of 16S rRNA worldwide has provided interesting and useful information6–9. For instance, with the use of 16S rDNA sequencing, 215 novel bacterial species, 29 of which belong to novel genera, have been discovered from human specimens in the past 7 years of the 21st century (2001–2007). One hundred of the 215 novel species, 15 belonging to novel genera, have been found in four or more subjects10. In Sudan, there is deficient data on sequencing about bacteria; therefore it is important to investigate what kinds of bacteria affect Sudanese people. Consequently, the objective of this study was to isolate and characterize P. aeruginosa from samples obtained from Sudanese patients by sequencing the 16SrRNA gene.\n\n\nMethods\n\nThis was cross sectional laboratory based study, conducted in Khartoum state in the period from January to April 2016. The study was approved by the Ethical and Scientific Committee of the Medicinal and Aromatic Plants and Traditional Medicine Research Institute, National Center of Research, Khartoum, Sudan (approval number 03-16) which ensures that all ethical considerations for conducting the research in a way that protects patient’s confidentiality and privacy are followed. Informed consent was obtained from the hospital laboratories (laboratory manager) after providing them with the ethical approval to collect samples during routine procedures from the microbiology laboratories. Participants’ privacy and confidentially was protected for all samples; personal information was not of great value in the current study and was thus not taken. Consequently, the Ethical and Scientific Committee waived the need for patient consent.\n\nA total of 40 isolates of P. aeruginosa (all samples available) were obtained from three hospitals in Khartoum State (Al Ribat Hospital, Bahri Hospital and Souba Hospital).\n\nThe bacterial isolates were collected from sputum culture, urine culture and wound infection because these types of cultures were dominant. Standard biochemical tests7 were performed on all samples for the isolation and identification of the bacterial isolates and were performed at the Medicinal and Aromatic Plants and Traditional Medicine Research Institute (MAPTMRI), Department of Microbiology. P. aeruginosa presence was confirmed in all 40 samples. All media required for biochemical tests were obtained from LAB M, UK and Laboratories Flow Media, Sudan.\n\nBacterial DNA was isolated by the Chelex-based protocol7 for all samples without deviation from the methodology.\n\nAll bacterial genomic DNA were used as templates for PCR amplification of the 16S rRNA gene. The two primers used were 27F (5ʹ AGAGTTTGATCCTGGCTCAG-3ʹ) and 1495R (5ʹ CTACGGCTACCTTGTTACGA- 3ʹ) for forward primer and reverse primer, respectively (Macrogen, South Korea). The 25μL PCR reaction mixture (Intro Biotechnology, South Korea) contained 1μL DNA, 1x reaction buffer (10x) with 3mM MgCl2, 2.5U i-TaqTM DNA polymerase (5 U/μL), 2.5mM dNTPs, 1μL of 10 pmol of each primer, and 1x of gel loading buffer, followed by completing the volume to 25μL with nuclease free water. Thermal cycling conditions were as follows: 94°C for 5 min; 30 cycles of denaturation (94°C for 1 min), annealing (58°C for 1 min), extension (72°C for 2 min); final extension at 72°C for 10 min. PCR was performed on a Bio-Rad (DNA engine/Dyad Peltier) automatic thermal cycler. Duplicate PCR of every sample were carried out for confirmation11,12. PCR products (5μl) were analyzed by gel electrophoresis in 1.0% agarose stained with ethidium bromide. The results were photographed under ultraviolet light machine (Transillumnator; Uvite, UK) to detect the specific amplified product by comparing it with 100 base pairs standard DNA ladder (Figure 1) and the remains from PCR products were store at -20°C until sequencing.\n\nLanes: 1- 8 positives 16SrRNA gene molecular weight marker (O’Range Ruler 100 DNA Ladder, SM1143-Fermentas).\n\nIsolates were packaged according to the International Air Transport Association guidelines and shipped with authorized permission to Macrogen Company (Seoul, South Korea). Purification and standard forward sequencing of 16S rRNA were done by ABI Genetic Analyzer (Applied Biosystems). We randomly selected 20 isolates only for DNA sequencing due to the limitation of resources.\n\nThe chromatogram sequences were visualized using Finch TV program version 1.4.013. The nucleotide sequences of the 16S rRNA gene were searched for sequences similarity using online BLASTn14. Highly similar sequences (accession numbers KX108935.1, KT943978.1, JN609194.1, KX214108.1, KU672378.1, KU764451.1 and FJ648815.1) were retrieved from NCBI GenBank and subjected to multiple sequence alignment using BioEdit software version 7.2.5. Newick format was withdrawn from ClustalW12, to create phylogenetic trees in Phylogeny.fr software15. MEGA6 software version 0.06 was used for confirmation of phylogentic trees16.\n\n\nResults\n\nA total of 40 clinical isolates were identified as P. aeruginosa by conventional methods, including growth characteristics, colony morphology, and biochemical tests. The results revealed that P. aeruginosa were dominant in wound infection cultures (42.5%), followed by (32.5%) from urine cultures and (25%) from sputum cultures.\n\nThe 16SrRNA in the isolates was amplified by PCR. In total, 20 were sent for characterization by sequencing of PCR products, 5 out of these 20 had a clear chromatogram, leading to further sequence analysis. Sequence analysis by BLASTn revealed 100% identity with P. aeruginosa from Iran (KX108935.1), China (KT943978.1), Australia (JN609194.1), Austria (KX214108.1), India (KU672378.1) and South Africa (KU764451.1), and slightly different from a sample from Australia (FJ648815): there was one substitution of A to G in position 148 (Figure 2). The phylogenetic tree of the 16S rRNA gene and other 16S rRNA genes obtained from the database revealed that the tree is classified into two branches. All isolates have a common ancestor. The isolate 120 is at the upper branch, isolate 113 and isolate 114 are sister groups, as shown in Figure 3.\n\nThe isolates had 100% identify with samples from Iran, China, Austria, India, and South Africa. There was one substitution of A to G in position 148 from a sample from Australia (FJ648815).\n\nThe tree is classified into two branches. All isolates have a common ancestor. The isolate 120 is at the upper branch, isolate 113 and isolate 114 are sister groups.\n\n\nDiscussion\n\nP. aeruginosa can cause various clinical infections. In the present study, 42.5% of isolates were identified from wound infection. Beside the highly preserved conserved primer binding sites, 16S rRNA gene sequences include hypervariable regions with high conservation that have the ability to characterize species-specific signature sequences beneficial to the characterization of bacteria17,18. The present results agree with Didelot et al., who reported that 16S rRNA gene sequencing is now common in medical microbiology as a quick and inexpensive alternative to phenotypic approaches of bacterial identification19. We investigated the phylogenetic affiliation of the pseudomonads and reexamined the 16S rRNA sequence data available in public databases. According to an alignment of 16S rRNA sequences, we identified P. aeruginosa-specific signature sequences.\n\n16S rRNA sequencing can be used to identify genetically atypical P. aeruginosa isolates from different origins. Our observation of 40 isolates from 40 samples, further showed the utility of 16s rRNA PCR amplification. This reveals the high specify of the primers and accuracy of the PCR. Sequencing of the 16S (rRNA) gene was used as an effective tool for determining phylogenetic relationships between bacteria. The features of this molecular target make it valuable and useful for bacterial detection and identification in the clinical laboratory.\n\n\nConclusions\n\nIn conclusion, 16S rRNA-based PCR assay and sequencing is highly specific, sensitive and reliable for identification of P. aeruginosa and its differentiation from other genotypic closely related Pseudomonas species. DNA sequencing of the 16S rRNA gene has been used as an effective tool to study bacterial phylogeny and taxonomy relationships between bacteria and for bacterial detection.\n\n\nData availability\n\nThe results of the nucleotide sequences of the 16S rRNA gene were submitted in the GenBank database. Accession numbers: KX650648, KX650649, KX650650, KX650651 and KX650652.",
"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\nAcknowledgments\n\nThe authors are grateful to Africa City of Technology, Medicinal and Aromatic Plants and Traditional Medicine Research Institute (MAPTMRI), Microbiology and Parasitology department, National Center for Research, Khartoum, Sudan and department Epidemiology, Molecular Epidemiology Laboratory Biology, Tropical Medicine Research Institute, National Centre for Research, Khartoum, Sudan for their kind support.\n\n\nReferences\n\nDriscoll JA, Brody SL, Kollef MH: The epidemiology, pathogenesis and treatment of Pseudomonas aeruginosa infections. Drugs. 2007; 67(3): 351–368. PubMed Abstract | Publisher Full Text\n\nTalbot GH, Bradley J, Edwards JE Jr, et al.: Bad bugs need drugs: an update on the development pipeline from the Antimicrobial Availability Task Force of the Infectious Diseases Society of America. Clin Infect Dis. 2006; 42(5): 657–668. PubMed Abstract | Publisher Full Text\n\nDavies J: Inactivation of antibiotics and the dissemination of resistance genes. Science. 1994; 264(5157): 375–382. PubMed Abstract | Publisher Full Text\n\nKhalil MA, Ibrahim Sonbol F, Mohamed AF, et al.: Comparative study of virulence factors among ESβL-producing and nonproducing Pseudomonas aeruginosa clinical isolates. Turk J Med Sci. 2015; 45(1): 60–69. PubMed Abstract | Publisher Full Text\n\nEmpel J, Filczak K, Mrówka A, et al.: Outbreak of Pseudomonas aeruginosa infections with PER-1 extended-spectrum beta-lactamase in Warsaw, Poland: further evidence for an international clonal complex. J Clin Microbiol. 2007; 45(9): 2829–2834. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoussier S, Vandewalle P, Luisetti J: Genetic diversity of african and worldwide strains of ralstonia solanacearum as determined by PCR-restriction fragment length polymorphism analysis of the hrp gene region. Appl Environ Microbiol. 1999; 65(5): 2184–2194. PubMed Abstract | Free Full Text\n\nLozupone CA, Knight R: Global patterns in bacterial diversity. Proc Natl Acad Sci U S A. 2007; 104(27): 11436–11440. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlindworth A, Pruesse E, Schweer T, et al.: Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013; 41(1): e1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHugenholtz P, Pace NR: Identifying microbial diversity in the natural environment: a molecular phylogenetic approach. Trends Biotechnol. 1996; 14(6): 190–197. PubMed Abstract | Publisher Full Text\n\nWoo PC, Lau SK, Teng JL, et al.: Then and now: use of 16S rDNA gene sequencing for bacterial identification and discovery of novel bacteria in clinical microbiology laboratories. Clin Microbiol Infect. 2008; 14(10): 908–934. PubMed Abstract | Publisher Full Text\n\nVilgalys R, Hester M: Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species. J Bacteriol. 1990; 172(8): 4238–4246. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCocolin L, Rantsiou K, Iacumin L, et al.: Study of the ecology of fresh sausages and characterization of populations of lactic acid bacteria by molecular methods. Appl Environ Microbiol. 2004; 70(4): 1883–1894. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOzdilek A, Cengel B, Kandemir G, et al.: Molecular phylogeny of relict-endemic Liquidambar orientalis Mill based on sequence diversity of the chloroplast-encoded matK gene. Plant Syst Evol. 2012; 298(2): 337–349. Publisher Full Text\n\nSeth-Smith HM, Rosser SJ, Basran A, et al.: Cloning, sequencing, and characterization of the hexahydro-1,3,5-Trinitro-1,3,5-triazine degradation gene cluster from Rhodococcus rhodochrous. Appl Environ Microbiol. 2002; 68(10): 4764–4771. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBorriss R, Chen XH, Rueckert C, et al.: Relationship of Bacillus amyloliquefaciens clades associated with strains DSM 7T and FZB42T: a proposal for Bacillus amyloliquefaciens subsp. amyloliquefaciens subsp. nov. and Bacillus amyloliquefaciens subsp. plantarum subsp. nov. based on complete genome sequence comparisons. Int J Syst Evol Microbiol. 2011; 61(Pt 8): 1786–1801. PubMed Abstract | Publisher Full Text\n\nOsman NAM, Alrayah IE, Mohamed YM, et al.: Molecular study of Panton-Valentine Leukocidin genes among Staphylococcus aureus clinical isolates in Khartoum State, Sudan. Am J Microbiol Res. 2015; 3(3): 107–111. Reference Source\n\nAdékambi T, Colson P, Drancourt M: rpoB-based identification of nonpigmented and late-pigmenting rapidly growing mycobacteria. J Clin Microbiol. 2003; 41(12): 5699–5708. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuasp C, Moore ER, Lalucat J, et al.: Utility of internally transcribed 16S-23S rDNA spacer regions for the definition of Pseudomonas stutzeri genomovars and other Pseudomonas species. Int J Syst Evol Microbiol. 2000; 50(4): 1629–1639. PubMed Abstract | Publisher Full Text\n\nDidelot X, Bowden R, Wilson DJ, et al.: Transforming clinical microbiology with bacterial genome sequencing. Nat Rev Genet. 2012; 13(9): 601–612. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "39641",
"date": "29 Oct 2018",
"name": "Giuseppe Comi",
"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 at the moment is only a description of a group of Pseudomonas aeruginosa. The strains come from different areas, the genomic study was made by a Korean lab. What did the authors want to say or to communicate with the paper?\nThe paper is interesting, and gives additional data about the P. aeruginosa species. However :\nReading the paper I cannot find the aim and the meaning of the work. It would be better to explain the aim of the paper.\n\nIt would be better to explain the aim of the paper.\n\nAdditional literature to explain the aim and the meaning of the work must be included.\n\nWhat is the meaning of the results?\n\nHow can the results be used?\n\nMaterial and methods must be better written.\n\nWhat did the authors do? The collected data and applied a software and no other skills.\n\nThe conclusion must be improved.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "42041",
"date": "02 Jan 2019",
"name": "Juan Alfonso Ayala Serrano",
"expertise": [
"Reviewer Expertise Molecular Microbiology. Peptidoglycan characterization",
"Antibiotic resistance",
"betalactamases"
],
"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\nManuscript deals with the characterizarion of Pseudomonas clinical isolates by PCR and 16S rRNA gene sequencing. Only twenty isolates were used for sequence comparison, and actualy only five sequences are compared by BLAST and deposited on Genbank DATABASE. Authors claim that sequencing of 16S rRNA gene is useful for Pseudomonas identification, but this fact has been shown in hundred of papers. It is not a novelty. Moreover, no other molecular analysis is done with the isolates, as for example, identification of resistance mechanisms. In addition, isolates were already identified as Psedomonas aeruginosa by biochemical methods. A more useful analysis would be to take the full number of isolates and without previous identification, characterize them as P. aeruginosa.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1135
|
https://f1000research.com/articles/7-1134/v1
|
25 Jul 18
|
{
"type": "Research Article",
"title": "Osteogenic potential of gingival stromal progenitor cells cultured in platelet rich fibrin is predicted by core-binding factor subunit-α1/Sox9 expression ratio (in vitro)",
"authors": [
"Alexander Patera Nugraha",
"Ida Bagus Narmada",
"Diah Savitri Ernawati",
"Aristika Dinaryanti",
"Eryk Hendrianto",
"Igo Syaiful Ihsan",
"Wibi Riawan",
"Fedik Abdul Rantam",
"Alexander Patera Nugraha",
"Ida Bagus Narmada",
"Diah Savitri Ernawati",
"Aristika Dinaryanti",
"Eryk Hendrianto",
"Igo Syaiful Ihsan",
"Wibi Riawan"
],
"abstract": "Background: Alveolar bone defect regeneration has long been problematic in the field of dentistry. Gingival stromal progenitor cells (GSPCs) offer a promising solution for alveolar bone regeneration. In order to optimally differentiate and proliferate progenitor cells, growth factors (GFs) are required. Platelet rich fibrin (PRF) has many GFs and can be easily manufactured. Core-binding factor subunit-α1 (CBF-α1) constitutes a well-known osteogenic differentiation transcription factor in SPCs. Sox9, as a chondrogenic transcription factor, interacts and inhibits CBF-α1, but its precise role in direct in vitro osteogenesis remains unknown. GSPCs cultured in vitro in PRF to optimally stimulate osteogenic differentiation has been largely overlooked. The aim of this study was to analyze GSPCs cultured in PRF osteogenic differentiation predicted by CBF-α1/Sox9. Methods: This study used a true experimental with post-test only control group design and random sampling. GPSCs isolated from the lower gingiva of four healthy, 250-gram, 1-month old, male Wistar rats (Rattus Novergicus) were cultured for two weeks, passaged every 4-5 days. GSPCs in passage 3-5 were cultured in five M24 plates (N=108; n=6/group) for Day 7, Day 14, and Day 21 in three different mediums (control negative group: αModified Eagle Medium; control positive group: High Glucose-Dulbecco’s Modified Eagle Medium (DMEM-HG) + osteogenic medium; Treatment group: DMEM-HG + osteogenic medium + PRF). CBF-α1 and Sox9 were examined with ICC monoclonal antibody. A one-way ANOVA continued with Tukey HSD test (p<0.05) based on Kolmogorov–Smirnov and Levene's tests (p>0.05) was performed. Results: The treatment group showed the highest CBF-α1/Sox9 ratio (16.00±3.000/14.33±2.517) on Day 7, while the lowest CBF-α1/Sox9 ratio (3.33±1.528/3.67±1.155) occurred in the control negative group on Day 21, with significant difference between the groups (p<0.05). Conclusion: GSPCs cultured in PRF had potential osteogenic differentiation ability predicted by the CBF-α1/sox9 ratio.",
"keywords": [
"Core-Binding Factor Subunit-α",
"Gingival Stromal Progenitor Cells",
"Osteogenic Differentiation",
"Platelet Rich Fibrin",
"Sox9."
],
"content": "Introduction\n\nDental caries represents a major global dental public health problem because of their high prevalence. The World Health Organization reported that almost 90% of people worldwide suffered from caries1. Basic Health Research of National Health (RISKESDAS) in 2013 reported that 93,998,727 Indonesians, 53.2% of the population, suffered from active caries2. Dental caries must be treated appropriately because, if neglected, they become so severe that the affected teeth must be extracted. Indeed, the most common cause of tooth loss is dental caries3. Populations experiencing low socioeconomic conditions demonstrate higher prevalence and extent of tooth loss because of extremely limited access to dental treatment4.\n\nTooth extraction has been the most common form of dental treatment performed in Indonesia that can lead to bone defects. RISKESDAS statistics dating from 2014 indicated that treatment involving tooth extraction reached as high as 79.6% of cases5. A previous study of tooth extraction-related complications revealed the prevalence of fractures (31.82%), bleeding (4.54%) and swelling (2.27%)6. Tooth extraction can lead to alveolar bone resorption and the destruction of alveolar bone components. Moreover, it may lead to resorption of the jawbone7. Tooth extraction followed by buccolingual and apicocoronal dimension reduction of the alveolar ridge at the edentulous site might be performed due to bone defects8.\n\nAlveolar bone defect regeneration has long represented a challenge in the field of dentistry. Various efforts have been made to accelerate bone regeneration, such as using bone grafts. The most current treatment performed in relation to the alveolar bone involves the use of platelet rich fibrin (PRF). PRF materials encouraging bone regeneration therapy have significantly improved the clinical outcomes stemming from the treatment of infrabony defects. PRF has achieved this through the maintaining of space for tissue regeneration by inducing an osteoinductive and osteoconductive effect in the alveolar bone defect area9.\n\nNowadays, alveolar bone defect treatment involving PRF using stromal progenitor cells (SPCs) is becoming increasingly widespread. SPCs have the advantage of being able to repair and regenerate various organs and tissue, and have been considerably used in bone tissue engineering, which offers encouraging solutions for bone regeneration10,11. SPCs are non-hematopoietic stromal cells. They have multipotent capabilities, including immunomodulators and immunoregulators, paracrine, autocrine action, and migrate directly to the tissue initiating healing and regeneration making SPCs particularly suitable for regenerative medicine development12–14.\n\nThe orofacial region is a unique and rich source of SPCs. Those contained in the oral cavity and tooth tissue represent an emerging interesting and topical object for investigation because isolating progenitor cells from the oral tissues can be achieved with minimal invasive procedures compared to bone marrow mesenchymal stem cell (BMSC) obtainment. The utilization of progenitor cells from the oral cavity is still rarely studied and applied. However, it is potentially advantageous for tissue regeneration and, therefore, merits further investigation15.\n\nThe SPCs that are potentially useful as part of regenerative alveolar bone therapy are gingival stromal progenitor cells (GSPCs) derived from hyperplastic gingival tissue (gum overgrowth) by means of a gingivectomy. GSPCs have phenotypic characteristics and abilities similar to those of BMSCs. GSPCs possess self-renewal capabilities and also demonstrate the specific ability to regenerate into alveolar bone when transplanted into immunocompromised mice. GSPCs also specifically induce bone matrix formation in lamellar structures by recruiting host cells11,15–17. The osteogenic ability of GSPCs needs to be explored for further application and therapy.\n\nDuring skeletal formation, master transcription genes such core-binding factor subunit-α (CBF-α1) and Osterix have been identified18. However, their specific and distinct roles in various tissue types are still unclear. Sox9 is well known as a master gene regulator during chondrogenic differentiation, while CBF-α1 plays an important role during osteogenic differentiation. GSPCs, as osteoprogenitors and chondroprogenitors, express Sox9 and Runx2 during skeletal formation condensation. There is also a direct interaction between Sox9 and CBF-α1, which inhibits Sox9 activity18. Sox9 inhibitory effect on osteoblast maturation through CBF-α1 is an essential mechanism for osteo-chondroprogenitor fate determination19.\n\nIn order for GSPCs to differentiate and proliferate optimally they require growth factors (GFs), various varieties of which are shown to promote osteogenic differentiation of SPCs in vitro. PRF is predicted to be combined to promote SPCs osteogenic differentiation and ensure mineralization in vitro19. PRF can be easily produced by centrifuging without anticoagulants. PRF is rich in GFs consisting of platelet derived growth factor-β (PDGF-β), transforming growth factor-β1 (TGFβ-1), vascular endothelial growth factor (VEGF) and insulin growth factor (IGF-I). PRF provides an effective scaffold to facilitate osteogenic differentiation of GSPCs20–23.\n\nThe osteogenic differentiation of GSPCs can be detected by various osteogenic marker expressions, such as CBF-α1 expression. The observed osteogenic markers of GSPCs are CBF subunit-α1 (CBF-α1) and Sox924,25. Nonetheless, there is insufficient information regarding Sox9’s role in osteogenesis of GPSCs in vitro. A study conducted by Stockl et al. mentioned that Sox9 plays a positive proliferative role in inhibiting and delaying osteogenic differentiation in rat SPCs26.\n\nThe hypothesis of the current study is that GSPCs cultured in PRF can increase the CBF-α1/Sox9 expression ratio during osteogenic differentiation. Furthermore, a second objective was to analyze GSPCs cultured in PRF osteogenic differentiation predicted by CBF-α1/Sox9 expression ratio.\n\n\nMethods\n\nThis study received ethical clearance relating to animal subjects from the Ethics Research Committee, Faculty of Dental Medicine, Universitas Airlangga, Surabaya, East Java, Indonesia (number 289/HRECC.FODM/XII/2017). The research was conducted at an experimental laboratory within the Stem Cell and Tissue Engineering Development Centre, Universitas Airlangga.\n\nThe research was fully experimental with a post-test only control group design. Sample groups were selected by means of simple random number sampling. Each animal was assigned a unique number, which were picked out of a hat by a blindfolded researcher.\n\nThe subjects consisted of male Wistar rats (Rattus norvegicus; n=4), who were adapted to the environment for 7 days. Wistar rats were obtained and cared for at the Stem Cell Animal Laboratory, Universitas Airlangga. All animals were housed in polycarbonate cages, subjected to a 12-hour light-dark cycle at the constant temperature of 23°C, and fed a standard pellet diet (expanded pellets; Stepfield, UK) with tap water ad libitum at a temperature of 22°C±2°C.\n\nGPSCs were isolated from the lower gingival tissue of four 1-month old, healthy, mean weight = 250g, male rats through a gingivectomy, before the rats were euthanized with doses 60mg/body weight of ketamine and xylazine. Animal suffering was reduced when removing the GPSCs using rodent’s anesthesia (intramuscular injection at 0.05–0.1ml/10g body weight rodent anesthesia: ketamine, xylazine, acepromazine, and sterile isotonic saline; Sigma Aldrich, USA) following Duan et al’s method23.\n\nGPSCs was passaged every 4–5 days following Rantam et al’s SPCs culture method27. GSPCs in passage 3–5 were cultured in five M24 plates (Sigma-Aldrich) (N=108; n=6/group) until Day 7, Day 14 and Day 21 in three different culture mediums (control negative group, control positive group and treatment group; see below for details).\n\nSample size (n=4 for GPSCs isolation; n=36 for PRF isolation) was based on Lemeshow's formula to determine minimum sample size\n\nA different population of rats were used for PRF isolation (n=36; 36 month old; mean weight = 250g). These male Wistar rats were maintained as above. Blood was aspirated through the left ventricle of each animals’ heart, after anesthesia had been administered by injection using a 60mg/body weight dose of ketamine and a 3mg/body weight dose of xylazine (Sigma Aldrich). 1.5ml of blood was aspirated using a 3ml disposable syringe and then inserted in a vacutainer tube without an anticoagulant before being centrifuged at 3000 rpm/min for 10min (Kubota, Tokyo, Japan). The centrifuging was performed by inserting two balance tubes containing water with the same weight as the tube of blood. When the tube is removed from the centrifuge, three layers will appear that are divided into three sections; the lower section consists of red blood cells, the middle section contains PRF and the upper section is formed of acellular plasma. The PRF was then isolated after which the PRF was cut into small pieces using sterile scissors and inserted into each culture plate of the treatment group22,28,29.\n\nThe analysis was conducted on three groups, consisting of two control groups and one experimental group.\n\nGSPC treatment group: GSPCs were cultured with PRF and containing ITS plus, 2mM L-glutamine, 100μg/ml sodium pyruvate, 0.2mM ascorbic acid-2 phosphate, dexamethasone 10-7 M (GeneTex, Taiwan), 10ng/ml TGF-β3 and high-dose glucose-Dulbecco's Modified Eagle Medium (DMEM-HG) (Sigma Aldrich).\n\nPositive control group: GSPCs were placed on an osteogenic medium culture plate of ITS plus, 2 mM L-glutamine, 100μg/ml sodium pyruvate 0.2mM ascorbic acid-2 phosphate, dexamethasone 10-7 M (GeneTex).\n\nNegative control group: GSPCs were cultured with αModified Eagle Medium (αMEM) (Sigma Aldrich).\n\nEvery three days, every group cell medium was replaced. Osteogenic differentiation was evaluated on Day 7, 14, 21 culture cells groups16.\n\nGSPCs cultured cells were coated with coverslips and, after incubation at 37°C for 1 - 2 hours, were fixed using 10% formaldehyde for 15 min. The coverslips were then rinsed four times with PBS and dried for several minutes. The cells were blocked with PBS and FBS 1% for 15–30 minutes and washed with PBS four times. The samples were then examined following immunocytochemical staining by indirect technique using a 3.3'-diaminobenzidine stain kit (Pierce DAB Substrate Paint Kit 34002, ThermofisherTM, Waltham, MA, USA) and monoclonal antibodies (Santa Cruz Biotechnology, Dallas, TX, USA): anti-CBF-α1 (mouse monoclonal; sc-101145) and anti-Sox9 (mouse monoclonal; sc-166505). CBF-α1 and Sox9 expression was read using a light microscope (CX22 Binocular, Olympus) at 200x magnification. Every cell expressing CBF1-α or Sox9 in one field was examined three times by three experts (WR, EH and FAR) and the mean was then calculated27,30,31.\n\nThe data obtained was analyzed using ANOVA continued with Tukey HSD test (p<0.05) based on a Saphiro-Wilk normality test and a Levene's variance of homogeneity test (p>0.05). Data were analyzed using SPSS version 20.0 (IBM SPSS, Chicago, USA).\n\nThe experiments were replicated 3 times (n=54). The data was then duplicated (n=108) using an estimation formula and SPSS (see Supplementary File 1 and Supplementary File 2)32.\n\n\nResults\n\nThe highest average CBF-α1 expression was in the treatment group on Day 7, whereas the lowest was in the control (-) group on Day 21 (Figure 1 and Figure 2). Sox9 expression had the highest mean value in the treatment group on Day 7, while its lowest value was in the negative control group on Day 21 (Figure 3 and Figure 4).\n\n(A-C) CBF-α1 expression in the negative control group; (D-F) CBF-α1 expression in the positive control group; (G-I) CBF-α1 expression in the treatment group. CBF-α1 expression in GSPCs was observed on Days 7, 14 and 21. Positive CBF-α1 expression is highlighted in brown (red arrow) following an examination at 200x magnification (n=1).\n\n(A-C) Sox9 expression in the negative control group; (D-F) Sox9 expression in the positive control group; (G-I) Sox9 expression in the treatment group. Sox9 expression in GSPCs was observed on Days 7, 14 and 21. Positive Sox9 expression is highlighted in brown (red arrow) following examination at 200x magnification (n=1).\n\nThe treatment group recorded the highest CBF-α1/Sox9 ratio (16.00±3.000/14.33±2.517/) on Day 7 while the lowest CBF-α1/Sox9 ratio (3.33±1.528/3.67±1.155) was registered by the control negative group on Day 21 (Table 1). The data obtained was normal with homogeneous distribution (p>0.05). There was significant difference between CBF-α1 and Sox9 expression in each group (p<0.05) (Supplementary Table 1 and Supplementary Table 2).\n\nResults are presented as the mean ± standard deviation. *One-way ANOVA, significant at p<0.05.\n\nCBF-α1: core-binding factor subunit-α1.\n\n\nDiscussion\n\nGSPCs cultured in PRF expressed CBF-α1 strongly. In this study, the highest CBF-α1 expression was recorded by the treatment group on Day 7, with significant difference between groups. The CBF-α1 expression declined between Day 14 and Day 21. The results of this study were in line with the research by Zou et al., which suggested that CBF-α1 expression is used to detect the osteogenic ability of SPCs using yellow fluorescent protein33.\n\nCBF-α1 is a master key gene transcription factor associated with osteoblast differentiation, which initiates temporally and spatially controlled osteogenesis. Disturbances to CBF-α1 result in obstacles to bone formation because osteoblast differentiation cannot occur. Loss of CBF-α1 expression gene function in the early stages will interfere with osteogenic differentiation and homeostasis in bone development. CBF-α1 is often expressed strongly between Day 7 and Day 1423,25. Osterix and CBF-α1 periodically regulate osteoblast differentiation processes34,35. A study conducted by Loebel et al. showed that CBF-α1 expression increased on Day 7, while Duan et al.’s study demonstrated that CBF-α1 expression increased on Day 12 as detected by RT-PCR19,23. Such findings differed from the results of this study due to the contrasting methods and samples employed, but there were similarities in that CBF-α1 was an early marker of osteogenic differentiation.\n\nCBF-α1 plays an important role in the early stages of BMSCs differentiation into preosteoblasts. CBF-α1 is generally a preliminary regulator and Osterix is a regulator activator during osteoblast differentiation. Both of these osteoblastogenic coding genes are stimulated and regulated by various signaling pathways, such as the canonical Wnt signaling pathway and bone morphogenetic protein (BMP). Wnt/Cytosolic β-catenin stimulates osteoblastogenesis through the activation of osteogenic transcription factors CBF-α1 and Osterix36. CBF-α1 is known as an important regulatory gene during osteogenic development by enhancing specific osteoblastic differentiation by inducing osteogenic extracellular matrix gene expression during osteoblast maturation, such as collagen-Iα, alkaline phosphatase, and osteocalcin33.\n\nIn the present study, the GSPCs cultured in PRF stimulates CBF-α1 expression because PRF is rich in various GFs, such as TGFβ-1, PDGF, IGF, VEGF, FGF, EGF, and HGF. PRF promotes migration, proliferation and differentiation of mesenchmymal stem cells as well as neovascularization and collagen synthesis. PRF also promotes, accelerates and improves the quality of soft and bone tissue regeneration23. According to Li et al., PRF significantly promotes the induction of mineralization of progenitor cells in alveolar bone, and endogenous stem cells present in the dental tissue that increases exclusively in CBF-α1 expression37,38.\n\nInterestingly, GSPCs cultured PRF in this study increased Sox9 even in an osteogenic culture medium with significant difference with the control groups. In this study, the highest Sox9 expression occurred in the treatment group on Day 7. The results of this study were supported by those of a study by Sumarta et al., which stated that SPCs cultured in PRF stimulate Sox9 expression22. Sox9 expression showed a positive expression, thereby establishing the role of Sox9 during bone formation. In a knockout Sox9 animal model, osteogenic differentiation was also delayed39. Significantly, recent genetics studies stated that Sox9 in SPCs could eventually differentiate into osteoblasts40. Therefore, the inhibitory effect of Sox9 on osteoblastic and chondrocyte maturation via repression of CBF-alpha1 function is an essential mechanism for osteo-chondroprogenitor cell fate determination41.\n\nIn this study, GSPCs-cultured PRF regulated and stimulated both CBF-α1/Sox9 expression ratio on Day 7 with significant difference between groups. The interaction and cooperation between CBF-α1/Sox9 is a mandatory master transcription gene for cartilage and bone development18. As Sox9 inhibited and downregulated CBF-α1 on Day 7, it may be even more sensitive to predict osteogenic differentiation ability of SPCs. Furthermore, while Sox9 expression was downregulated, osteogenic differentiation ability was stimulated during early osteogenic differentiation in vitro. Nevertheless, CBF-α1/Sox9 expression ratio on Day 7 could be used to predict the osteogenic differentiation ability of GMSCs, suggesting a balance between CBF-α1/Sox9 in the earlier regulatory bone formation and regeneration19,41.\n\n\nConclusion\n\nGSPCs cultured in PRF increased CBF-α1/Sox9 expression on Day 7. GSPCs cultured in PRF possessed potential osteogenic differentiation ability as predicted by the CBF-α1/sox9 expression ratio. CBF-α1/Sox9 expression constitutes a promising future in vitro screening method employed to detect the earliest osteogenic differentiation of SPCs. Further study is required to analyze any association with CBF-α1/Sox9 expression ratio in vivo.\n\n\nData availability\n\nDataset 1: Raw results for CBF-α1 and Sox9 expression for all time points for all treatment groups (N=108; n=6/group). DOI, 10.5256/f1000research.15423.d21063842\n\nDataset 2: Raw image data. DOI, 10.5256/f1000research.15423.d21063943",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe research was funded by the Progam Menuju Doktor Sarjana Unggul (PMDSU) Batch III of the Ministry of Research, Technology and Higher Education of the Republic of Indonesia (Kemenristekdikti RI) (letter of appointment agreement number, 1035/D3/PG/2017; grant number 2146/D3/PG/2017).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors would like to thank the Postgraduate School, Department of Dental Medicine, Faculty of Medicine, Stem Cell Research and Development Centre, Universitas Airlangga for its support of the research reported here.\n\n\nSupplementary material\n\nSupplementary Table 1. Tukey HSD multiple comparison between groups of CBF-α1.\n\nClick here to access the data.\n\nSupplementary Table 2. Tukey HSD multiple comparison between groups of Sox9 expression.\n\nClick here to access the data.\n\nSupplementary File 1: Estimation resume of CBF-α1 expression.\n\nClick here to access the data.\n\nSupplementary File 2: Estimation resume of Sox9 expression.\n\nClick here to access the data.\n\n\nReferences\n\nPetersen PE, Bourgeois D, Ogawa H, et al.: The global burden of oral diseases and risks to oral health. Bull World Health Organ. 2005; 83(9): 661–69. PubMed Abstract | Free Full Text\n\nMinistry of Health and National Institute of Health Research and Development: National report on basic health research, RISKESDAS, 2013. Jakarta, Indonesia, 2014 (and additional analysis). Ministry of Health, Republic of Indonesia, Jakarta.\n\nLee JH, Oh JY, Choi JK, et al.: Trends in the incidence of tooth extraction due to periodontal disease: results of a 12-year longitudinal cohort study in South Korea. J Periodontal Implant Sci. 2017; 47(5): 264–272. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSusin C, Oppermann RV, Haugejorden O, et al.: Tooth loss and associated risk indicators in an adult urban population from south Brazil. Acta Odontol Scand. 2005; 63(2): 85–93. PubMed Abstract | Publisher Full Text\n\nMinistry of Health Republic of Indonesia: Health Profile in Indonesia 2014. Ministry of Health Republic of Indonesia. Sekretaris Jenderal. Jakarta, 2015. Reference Source\n\nLande R, Kepel BJ, Siagian KV: Profile of Risk Factor and Complication of Tooth Extraction at RSGM PSPDG-FK Unsrat. Jurnal E-Gigi (Eg). 2015; 3(2): 1–6.\n\nHamzah Z, Kartikasari N: Irrational Tooth Extraction increasing structural and functional bone resorption. Stomatognatic (J. K. G Unej). 2015; 12(2): 61–6.\n\nSchropp L, Wenzel A, Kostopoulos L, et al.: Bone healing and soft tissue contour changes following single-tooth extraction: a clinical and radiographic 12-month prospective study. Int J Periodontics Restorative Dent. 2003; 23(4): 313–323. PubMed Abstract\n\nChang IC, Tsai CH, Chang YC: Platelet-rich fibrin modulates the expression of extracellular signal- regulated protein kinase and osteoprotegerin in human osteoblasts. J Biomed Mater Res A. 2010; 95: 327–32. PubMed Abstract | Publisher Full Text\n\nBarzilay R, Melamed E, Offen D: Introducing transcription factors to multipotent mesenchymal stem cells: making transdifferentiation possible. Stem Cells. 2009; 27(10): 2509–2515. PubMed Abstract | Publisher Full Text\n\nEgusa H, Sonoyama W, Nishimura M, et al.: Stem cells in dentistry--part I: stem cell sources. J Prosthodont Res. 2012; 56(3): 151–165. PubMed Abstract | Publisher Full Text\n\nAugello A, Kurth TB, De Bari C: Mesenchymal stem cells: a perspective from in vitro cultures to in vivo migration and niches. Eur Cell Mater. 2010; 20: 121–133. PubMed Abstract | Publisher Full Text\n\nLaw S, Chaudhuri S: Mesenchymal stem cell and regenerative medicine: regeneration versus immunomodulatory challenges. Am J Stem Cells. 2013; 2(1): 22–38. PubMed Abstract | Free Full Text\n\nGao F, Chiu SM, Motan DA, et al.: Mesenchymal stem cells and immunomodulation: current status and future prospects. Cell Death Dis. 2016; 7: e2062. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEgusa H, Sonoyama W, Nishimura M, et al.: Stem cells in dentistry--Part II: Clinical applications. J Prosthodont Res. 2012; 56(4): 229–248. PubMed Abstract | Publisher Full Text\n\nNiibe K, Suehiro F, Oshima M, et al.: Challenges for stem cell-based \"regenerative prosthodontics\". J Prosthodont Res. 2017; 61(1): 3–5. PubMed Abstract | Publisher Full Text\n\nMiran S, Mitsiadis TA, Pagella P: Innovative Dental Stem Cell-Based Research Approaches: The Future of Dentistry. Stem Cells Int. 2016; 2016: 7231038. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou G, Zheng Q, Engin F, et al.: Dominance of SOX9 function over RUNX2 during skeletogenesis. Proc Natl Acad Sci U S A. 2006; 103(50): 19004–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoebel C, Czekanska EM, Bruderer M, et al.: In vitro osteogenic potential of human mesenchymal stem cells is predicted by Runx2/Sox9 ratio. Tissue Eng Part A. 2015; 21(1–2): 115–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaek HS, Lee HS, Kim BJ, et al.: Effect of platelet-rich fibrin on repair of defect in the articular disc in rabbit temporomandibular joint by platelet-rich fibrin. Tissue Engineering and Regenerative Medicine. 2011; 8(6): 530–535. Reference Source\n\nKazemi D, Fakhrjou A, Dizaji VM, et al.: Effect of autologous platelet rich fibrin on the healing of experimental articular cartilage defects of the knee in an animal model. Biomed Res Int. 2014; 2014: 486436. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSumarta NPM, Pramono DC, Hendrianto E, et al.: Chondrogenic Differentiation Capacity of Human Umbilical Cord Mesenchymal Stem Cells with Platelet Rich Fibrin Scaffold in Cartilage Regeneration (In Vitro Study). Bali Med J. 2016; 5(3): 420–426. Publisher Full Text\n\nDuan X, Lin Z, Lin X, et al.: Study of platelet-rich fibrin combined with rat periodontal ligament stem cells in periodontal tissue regeneration. J Cell Mol Med. 2018; 22(2): 1047–1055. PubMed Abstract | Free Full Text\n\nMarie PJ, Fromigué O: Osteogenic differentiation of human marrow-derived mesenchymal stem cells. Regen Med. 2006; 1(4): 539–48. PubMed Abstract | Publisher Full Text\n\nShui C, Spelsberg TC, Riggs BL, et al.: Changes in Runx2/Cbfa1 expression and activity during osteoblastic differentiation of human bone marrow stromal cells. J Bone Miner Res. 2003; 18(2): 213–21. PubMed Abstract | Publisher Full Text\n\nStöckl S, Göttl C, Grifka J, et al.: Sox9 modulates proliferation and expression of osteogenic markers of adipose-derived stem cells (ASC). Cell Physiol Biochem. 2013; 31(4–5): 703–17. PubMed Abstract | Publisher Full Text\n\nRantam FA, Ferdiansyah P: Stem cell mesenchymal, Hematopoetic Stem Cells and Application Model. 2nd Ed. Surabaya: Airlangga University Press, 2014; 2: 38.\n\nBorie E, Oliví DG, Orsi IA, et al.: Platelet-rich fibrin application in dentistry: a literature review. Int J Clin Exp Med. 2015; 8(5): 7922–7929. PubMed Abstract | Free Full Text\n\nRahmawati D, Roestamadji RI, Yuliati A, et al.: Osteogenic ability of combined hematopoetic stem cell, hydroxyapatite graft and platelet rich fibrin on rats (Rattus novergicus). Journal of Krishna Institute of Medical Sciences University. 2017; 6(4): 88–95. Reference Source\n\nKamadjaja MJK, Salim S, Rantam FA: Osteogenic Potential Differentiation of Human Amnion Mesenchymal Stem Cell with Chitosan-Carbonate Apatite scaffold (in vitro study). Bali Med J. 2016; 5(3): 427–433. Publisher Full Text\n\nEkizer A, Yalvac ME, Uysal T, et al.: Bone marrow mesenchymal stem cells enhance bone formation in orthodontically expanded maxillae in rats. Angle Orthod. 2015; 85(3): 394–399. PubMed Abstract | Publisher Full Text\n\nKarangwa I, Kotze D: Using the Markov Chain Monte Carlo Method to Make Inferences on Items of Data Contaminated by Missing Values. American Journal of Theoretical and Applied Statistics. 2013; 2(3): 48–53. Publisher Full Text\n\nZou L, Kidwai FK, Kopher RA, et al.: Use of RUNX2 expression to identify osteogenic progenitor cells derived from human embryonic stem cells. Stem Cell Reports. 2015; 4(2): 190–198. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrunetti G, Mori G, D’Amelio P, et al.: The crosstalk between the bone and the immune system: Osteoimmunology. Clinical and Developmental Immunology. 2013; 2013: 617319. Publisher Full Text\n\nCrotti TN, Dharmapatni AA, Alias E, et al.: Osteoimmunology: Major and Costimulatory Pathway Expression Associated with Chronic Inflammatory Induced Bone Loss. J Immunol Res. 2015; 2015: 281287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGraves DT, Kayal RA, Oates T: Chapter 19 - Osteoimmunology in the Oral Cavity (Periodontal Disease, Lesions of Endodontic Origin, and Orthodontic Tooth Movement). Osteoimmunology. 2016; 325–344. Publisher Full Text\n\nLi Q, Pan S, Dangaria SJ, et al.: Platelet-rich fibrin promotes periodontal regeneration and enhances alveolar bone augmentation. Biomed Res Int. 2013; 2013: 638043. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q, Reed DA, Min L, et al.: Lyophilized platelet-rich fibrin (PRF) promotes craniofacial bone regeneration through Runx2. Int J Mol Sci. 2014; 15(5): 8509–8525. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkiyama H, Lyons JP, Mori-Akiyama Y, et al.: Interactions between Sox9 and beta-catenin control chondrocyte differentiation. Genes Dev. 2004; 18(9): 1072–1087. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkiyama H, Kim JE, Nakashima K, et al.: Osteo-chondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci U S A. 2005; 102(41): 14665–14670. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou G, Zheng Q, Engin F, et al.: Dominance of SOX9 function over RUNX2 during skeletogenesis. Proc Natl Acad Sci U S A. 2006; 103(50): 19004–19009. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNugraha AP, Narmada IB, Ernawati DS, et al.: Dataset 1 in: Osteogenic potential of gingival stromal progenitor stem cells cultured in plasma rich fibrin is predicted by core-binding factor subunit-a1/Sox9 expression ratio (in vitro). F1000Research. 2018. Data Source\n\nNugraha AP, Narmada IB, Ernawati DS, et al.: Dataset 2 in: Osteogenic potential of gingival stromal progenitor stem cells cultured in plasma rich fibrin is predicted by core-binding factor subunit-a1/Sox9 expression ratio (in vitro). F1000Research. 2018. Data Source"
}
|
[
{
"id": "36488",
"date": "13 Aug 2018",
"name": "Guang 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\nThis paper needs major revision. I would make the following comments.\n\nIntroduction: 1. In final part, you said “Furthermore, a second objective was …”, however I cannot find your first objective. What is the first objective of your study? I think you should change this part to “The objective of this study was to analyze GSPCs cultured in PRF osteogenic differentiation predicted by CBF-α1/Sox9 expression ratio. The hypothesis of the current study is that GSPCs cultured in PRF can increase the CBF-α1/Sox9 expression ratio during osteogenic differentiation.”\n\n2. The reason that why you do this research in necessary in Introduction section. And also please add the that is the problem of current and past research regarding relationship of osteogenic and CBF-α1/Sox9 ratio?\n\nMethods:\n3. The materials section is too long, you can try to simplify this section.\n\nDiscussion:\n4. The answer of hypothesis of this study should be included in the discussion section.\n\n5. The discussion about the experiment method and selection of material is necessary. There was no discussion why expression differences existed. The authors need to discuss based upon the mechanism otherwise the paper is going to be a laboratory report.\n\n6. Please add the clinical implications of your study and clinical significance of the findings in discussion section.\n\n7. What is the limitation of this study? You should indicate at this section.\n\nFigure:\n8. Fig. 1 and Fig. 3: Please indicate which pic is A, B, C……. and H, I. 9. Fig. 2 and Fig. 4: Please add the SD (standard deviation) to your bar graph. And if possible also add the results of statistical analysis to your graph.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "36776",
"date": "15 Aug 2018",
"name": "Ananto Ali Alhasyimi",
"expertise": [
"Reviewer Expertise Biological tooth movement"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 study have investigated the role of GSPCs cultured in PRF in increasing the CBF-α1/Sox9 expression ratio during osteogenic differentiation. Based on the study results, the authors have concluded that the GSPCs cultured in PRF increased CBF-α1/Sox9 expression on Day 7. The results are certainly interesting and the authors are commended for executing this study. However, the authors need to address the following minor concerns: General:\nPaper is well organized and easy to follow. Novelty is sufficient and high impact. To improve the readability, It is recommended that the text is checked by a native English speaking person as many of the sentences might be misunderstood. I suggest a revision of the English grammar structures by an expert editor in revising manuscripts.\nIntroduction:\nThe authors have mentioned “Furthermore, a second objective was …”, however, I didn’t find the first objective. Please change this section The Introduction section is too long, you can try to simplify this section (please move some of the key mechanism of the materials in enhancing CBF-α1/Sox9 expression in the discussion part\n\nResults:\nPlease provide and add some information about standard deviation of the data in the diagrams. Please indicate whether the difference is significant or not in your graph/ diagrams as well, using some symbols (asteric).\n\nDiscussion:\nDiscussion of the results is quite comprehensive. In analyzing the results, the authors also show citations from the previous study to support the explanation of these results. The answer to the hypothesis of this study should be included at the beginning of the discussion section. Please mention the limitation of this study in the discussion section for\n\nReferences: The supporting references are too long for the medium article\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36777",
"date": "16 Aug 2018",
"name": "Benny M. Soegiharto",
"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\nFirst of all, please allow me to congratulate the authors for attempting to undertake this project which I found very interesting and of valuable additional knowledge. The manuscript itself is well-written and well-structured and I have to also commend the authors for this matter. However, I may require some clarifications on the following issues:\n\nWhat would be the arguments of dividing into the 3 groups as well as using such media treatment for each group (GSPC, + and - groups). This would be beneficial for readers to replicate in future use. What would be the reason for determining the evaluation of osteogenic differentiation on Day 7, 14 and 21 cell groups? At the end of that sentence, there was a quote of another work. However, I think it would make this manuscript even clearer if this could be briefly mentioned.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1134
|
https://f1000research.com/articles/7-1122/v1
|
23 Jul 18
|
{
"type": "Review",
"title": "Tourette syndrome research highlights from 2017",
"authors": [
"Andreas Hartmann",
"Yulia Worbe",
"Kevin J. Black",
"Yulia Worbe",
"Kevin J. Black"
],
"abstract": "This is the fourth yearly article in the Tourette Syndrome Research Highlights series, summarizing research from 2017 relevant to Tourette syndrome and other tic disorders. The authors briefly summarize reports they consider most important or interesting. The highlights from 2018 article is being drafted on the Authorea online authoring platform, and readers are encouraged to add references or give feedback on our selections using the comments feature on that page. After the calendar year ends, the article is submitted as the annual update for the Tics collection on F1000Research.",
"keywords": [
"Tourette syndrome",
"tic disorders",
"review",
"natural history",
"etiology",
"pathophysiology"
],
"content": "Introduction\n\nThis article is meant to disseminate recent scientific progress on Gilles de la Tourette Syndrome (TS).\n\n\nMethods\n\nWe searched PubMed from time to time using the search strategy “(\"Tic Disorders\"[MeSH] OR Tourette NOT Tourette[AU]) AND 2017[PDAT] NOT 1950:2016[PDAT]”. On 06 July 2018 this search returned 212 citations. Colleagues also recommended articles, and we attended medical conferences. We selected material to be discussed in this review subjectively, guided by our judgment of possible future impact on the field.\n\n\nResults\n\nA number of tic experts contributed to a review article on TS (Robertson et al., 2017).\n\nSchaefer et al. (2017) describe 16 people with TS who had experienced a clinical remission or marked improvement of more than 1 year’s duration, followed by symptomatic worsening as adults, leading them again to seek treatment. On average the “latent period” (the absence or substantial reduction in tics) had lasted 16 years. Seven of them had worse tics when returning for care than they recalled as children. New substance use was reported as a trigger for exacerbation in 5 patients. This report strengthens evidence that even long-lasting symptomatic improvement in tic disorders (TDs) may not always be permanent, and that in fact the typical course of TS “is one of occasional recurrences of mild tics throughout adult life” (Bruun & Budman, 1997); see also (Black et al., 2016; Shapiro et al., 1988, p. 188; Stárková, 1990; Singer, 2006).\n\nRegarding natural course and history, the fate of non-tic symptoms in TS has remained less well explored. A large Danish study reported follow-up data 6 years after enrolling 314 children and teenagers with TS, assessing tics and comorbidities (mainly obsessive–compulsive disorder (OCD) and Attention-Deficit/Hyperactivity Disorder (ADHD); N=226 at follow-up) (Groth et al., 2017). Most patients’ tics improved over time, but almost a quarter of those over age 16 still had severe tics, and only a sixth had no tics. The severity of OCD and ADHD declined significantly during adolescence, suggesting a shift towards so-called “pure” TS with age. Furthermore, the authors expected tic-related impairment to improve with an age-related decline in tic frequency and severity, but surprisingly the impairment score did not reflect the improvement in tics.\n\nA report on 606 patients with a movement disorder starting in childhood produced an estimate for tic onset of 7.4 ± 3.8 years with a mean delay to diagnosis of 9.9 ± 11 years (Bäumer et al., 2016).\n\nEpidemiology. New and important findings this year involve the previously unappreciated risk of death in TDs. Meier et al. (2017) demonstrated that mortality rates are elevated in TS and other TDs, with or without comorbidities. In a very large epidemiological study, TDs in adults were associated with a four-fold higher risk of suicide, with the risk not explained by other psychiatric illness such as major depression (Fernández de la Cruz et al., 2017b). These researchers analyzed 7736 TS/chronic tic disorder (CTD) cases in the Swedish National Patient Register over a 44-year period (1969–2013) and compared them with control subjects from the general population. An increased risk of both suicide and attempted suicide was observed in TS/CTD patients, which was not solely dependent on psychiatric comorbidities. Tics that persisted beyond young adulthood significantly predicted completed suicide. Thus, TS/CTD is a serious medical condition that requires careful monitoring of suicidality. The same group also found a 7- to 10-fold higher risk of completed suicide in people with OCD, even after controlling for comorbid diagnoses (Fernández de la Cruz et al., 2017a).\n\nMartino et al. (2017c) review screening instruments and rating scales for TDs; see also (Augustine et al., 2017; Martino & Pringsheim, 2017).\n\nTransient environmental effects on tic severity. A study of 45 children with TS supported the typical antecedent–behavior–consequence behavioral psychology model (Eaton et al., 2017). Specifically, consequences of tics, “such as receiving accommodations or attention from others,” explained significantly more variance in tic severity than did the child’s level of separation anxiety, though the latter was also a significant factor. This study provides supportive evidence for the approach taken by “CBIT-Jr,” a behavior therapy designed for younger children with TS (Piacentini et al., 2015).\n\nOther. Non-tic symptoms in TS are reviewed by Martino et al. (2017b). Lee et al. (2017) used Taiwan’s National Health Insurance Research Database to compare 1124 newly diagnosed TS patients to controls in a 1:3 match. Sleep disorders were twice as common in TS, and remained significantly higher in TS after accounting for anxiety disorders, which were the comorbid conditions associated with the highest risk. In a sample of 811 TS subjects recruited for a genetics study, hair pulling (3.8%) and skin picking (13.0%) disorders by DSM-5 were surprisingly common (Greenberg et al., 2018).\n\nAutism spectrum disorders (ASD) comprise an underexplored comorbidity of TD. There are few epidemiological studies on the subject but the prevalence of ASD in children with TDs is estimated at 20% (Khalifa & von Knorring, 2006). In a large study including patients with TD (n=535) and their family members (n=234), Darrow et al. (2017a) used the Social Responsiveness Scale Second Edition (SRS) to characterize ASD symptoms, and compared them to historical ASD samples. SRS scores in participants with TD were similar to those observed in other clinical samples but lower than in ASD samples. This is mostly but not entirely explained by elevations in the RRB (restricted interests and repetitive behaviors) subscale, which may be indicating tics rather than other stereotypic movements. The presence of OCD was associated with higher scores on the social cognition and RRB subscales. Complex tics and OCD symptoms (repetitive behaviors) can also be hard to discriminate from core ASD symptoms, especially those related to social communication.\n\nA national database study found that parents of children with chronic tics had significantly higher rates of psychiatric illness: more than twice as high in mothers and 39% higher in fathers (Leivonen et al., 2017). The maternal risk, which included a range of psychopathology, was significantly higher than the paternal risk, which comprised primarily OCD and anxiety disorders. Further work will be needed to clarify whether results reflect maternal-specific environmental risks, genetic risks, factors related to parental care-seeking, or (to a modest extent, given the typical ages of onset for various parental disorders identified) parental stress.\n\nGenetics. A large mixed genetic sample yielded two heritable collections of symptoms that cross diagnostic boundaries, here named symmetry (including some other obsessions and compulsions) and disinhibition (including complex verbal tics) (Darrow et al., 2017b). Whole exome sequencing from over 500 trios identified a clear excess of likely gene-disrupting de novo mutations, and 4 risk genes that were altered by different mutations in multiple probands (Willsey et al., 2017). Another report described whole exome sequencing in a 3-generation family with TS, using induced pluripotent stem cells converted into neuronal cell types and assessing protein expression levels (Sun et al., 2018). The PNKD gene product was expressed at a lower rate in family members with TS or OCD.\n\nRoger Albin wrote a very thoughtful review of TS as “a disorder of the social decision-making network” (Albin, 2018). Beste & Münchau (2018) describe tics in the context of the Theory of Event Coding, which describes the brain’s bidirectional pairing of stimulus and movement. They note the salience of urges to most tic patients, and the fact that most tics are individually relatively normal movements, but occur repeatedly and out of context. Similarly, Shafer et al. (2017) focus on the strong connections between movement and sensory function to propose a theory that encompasses both the development of stereotypies as part of typical development and their persistence in maladaptive forms.\n\nAnimal models. A special issue on animal models of TS appeared in the Journal of Neuroscience Methods (Bortolato & di Giovanni, 2017). Articles reviewed gait and sensorimotor function in the D1CT-7 mouse model of TS (Fowler et al., 2017), stress mediating the timing of abnormal movements in animal models (Godar & Bortolato, 2017), and chemogenetic and optogenetic models (Burton, 2017). Deer mice have behavior that has been discussed as a natural model of OCD (Wolmarans et al., 2018).\n\nRecently postnatal ablation of the TrkB receptor in cells expressing parvalbumin was shown to produce dramatic changes in cortex and cerebellum and “profound hyperactivity, stereotypies, motor deficits and learning/memory defects” (Xenos et al., 2017). This result may help explain why autopsies in TS show a lower number of parvalbumin-containing interneurons in the striatum, but this model also produces much more substantial neuroanatomical changes than are seen in TS.\n\nA gene identified in human OCD, slc1a1, which codes an excitatory amino acid transporter, was altered in mice to prevent its expression and function (Zike et al., 2017). The loss of this protein resulted in mice with reduced extracellular dopamine concentrations and reduced movement and stereotypic behavior after challenge with amphetamine or a dopamine D1 receptor agonist. Restoring the gene’s expression in the midbrain, but not in the striatum, partially rescued the exogenous dopamine-induced stereotypies. This research is important for its direct links to human illness and its anatomical specificity, and lends additional support to testing dopamine D1 antagonists in TS (see Medication section, below).\n\nNew insights from computational modelling. Using the neurophysiological data obtained from a TS animal model of pharmacological striatal disinhibition, Caligiore et al. (2017) proposed a computational model of the basal ganglia-cerebellar-thalamo-cortical system to address the mechanisms of motor tic generation in TS. Overall, the model suggested that interplay between dopaminergic signal and cortical activity triggered the occurrence of a tic and was able to predict the number of tics generated when striatal dopamine increases and when the cortex is externally stimulated. Maia & Conceição (2017) further discussed the role of tonic and phasic dopamine in tics learning and expression. Based on the existing literature of habit formation and reinforcement learning in TS, the authors proposed a model of tics as exaggerated and persistent motor habits reinforced by aberrant, increased phasic dopamine responses. According to this model, tonic dopamine release would serve to amplify the tendency to execute learned tics. The authors also proposed the mechanism of antipsychotics’ action on tics: increased activity of indirect pathway due to antipsychotic administration could result in tic reduction, but at the same time potentially also could increase the propensity for reinforcing tics due to plasticity in the indirect pathway. In contrast, the authors also suggested that low-dose dopamine agonists could decrease both phasic and tonic dopamine and thus reduce both tic learning and tic expression. Both of these reports assume increasing tics with increasing dopamine concentrations, an assumption that seems to contradict observations that tics do not worsen with exogenous levodopa (Black & Mink, 2000; Gordon et al., 2013) nor improve with development of Parkinson disease (Kumar & Lang, 1997; Martinez-Torres et al., 2009; Shale et al., 1986). Furthermore, low-dose pergolide in children and adolescents with TS that reduced tic severity suppressed prolactin rather than increasing it, consistent with overall enhancement of dopamine transmission, at least in the hypothalamic-pituitary dopaminergic pathway (Gilbert et al., 2000a; Gilbert et al., 2000b).\n\nBased largely on available functional anatomical studies, Conceição et al. (2017) also proposed a computational model of premonitory urges in TS. According to this model, premonitory urges and in particular their termination, like termination of other aversive stimuli, might elicit positive prediction errors, supported by phasic dopamine release that would then reinforce tics. The insula may play a central role in aversive feeling associated with premonitory urges and their learned negative value. The insula might send this information via direct or indirect projections to dopamine neurons, which might use it for calculation of the positive prediction errors that occur with termination of the premonitory urge. In short, the authors provide a more detailed neurobiological explanation for the classic model that premonitory urges may strengthen tics through negative reinforcement.\n\nCognitive function and decision-making in TS. Morand-Beaulieu et al. (2017a) provide an updated and exhaustive review of neuropsychological aspects of TS (compare Eddy et al. (2009)). The review highlighted the slight alteration of social cognition as well as more frequent learning difficulties and disabilities in children with TS. Recent data also seem to confirm the deficit in executive function in TS as indexed by poor performance in continuous performance test and Stroop tests. Interestingly, using longitudinal evaluation of executive function in children with TS, Yaniv et al. (2018) showed that adults with TS showed response inhibition deficits, that tic reduction over time was significantly associated with development of response inhibition, and that former TS patients whose tics had remitted performed as well as, or on some tests better than, healthy control subjects. In contrast, the attentional and memory capacity seems to be impacted by comorbid symptoms more than TS per se. Sample size (n=122) or the healthy control group may explain different results obtained by Abramovitch et al. (2017), who studied treatment response in TS. They concluded that “the finding that significant change in symptom severity of TS/CTD patients is not associated with impairment or change in inhibitory control regardless of treatment type suggests that inhibitory control [as measured by the tests selected] may not be a clinically relevant facet of these disorders in adults.” Given these contradictory results, Morand-Beaulieu et al. (2017b) review the puzzling question of inhibitory control in TS in a recent meta-analysis, and find larger inhibitory deficits in TS + ADHD patients, but this deficit was also present in “pure” TS. This deficit in TS was most prominent in verbal responses, was associated with tics severity as assessed with Yale Global Tic Severity Scale-Total Tics Score (YGTSS-TTS) and was larger in studies that included medicated TS patients.\n\nSalvador et al. (2017) study decisional capacities in TS and specifically the ability to learn from the outcomes of alternative courses of action (known as counterfactual learning). Unmedicated patients with TS showed normal performance on this task, whereas alteration was found in TS patients treated with the dopamine D2-like receptor (D2R) partial agonist aripiprazole, suggesting that modulating D2Rs may impair certain aspects of human reinforcement learning.\n\nThe Committee on Research of the American Neuropsychiatric Association published a systematic review on the neurobiology of the premonitory urge in TS (Cavanna et al., 2017).\n\nElectrophysiology. Brandt et al. (2017) reported on enhanced multi-component behavior in TS, which was also reflected in a smaller P3 event-related potential measured by EEG and potentially related to chronic tic control in these patients. An EEG study during simulated driving found that “mind wandering” may be quantifiable using EEG measures of alpha power and the P3a component of an auditory event-related potential (Baldwin et al., 2017). Since mind wandering is a defining feature of ADD, in addition to being ubiquitous during repetitive tasks, these measures may prove useful in studying ADHD phenomenology in TDs.\n\nNeuroimaging studies. Polyanska & colleagues. (2017) provide a meta-analysis of task-based neuroimaging reports in TS. Adults with TS show some impairment in lateralized sequential finger tapping movements, and this impairment was compared to fractional anisotropy of white matter tracts connecting primary motor cortex (M1) to M1 and supplementary motor area (SMA) to SMA (Martino et al., 2017a).\n\nPrevious in vivo MRI studies of basal ganglia volume and shape in TS have produced differing results (reviewed in Greene et al. (2017)). A study of 47 children age 8-12 with TS, and controls with or without ADHD, found no significant group differences (Forde et al., 2017).\n\nA PET study of the microglial activating marker TSPO in OCD found elevated TSPO concentrations in dorsal caudate, orbitofrontal cortex, thalamus, ventral striatum, dorsal putamen, and anterior cingulate cortex (Attwells et al., 2017). Concentrations in patients were about 1/3 higher than in controls. This report implicates low-level brain inflammation in the pathophysiology of OCD.\n\nClinical and neuropsychological studies. A study using the alternating serial reaction time task found no impairment of procedural learning in TS or ADHD (Takács et al., 2017). This result is surprising, given that habit learning on a “weather prediction” task is slower in TS (Kéri et al., 2002; Marsh et al., 2004), but may indicate that the two tasks engage different learning systems, a conclusion supported by the generally normal cognitive function in TS.\n\nMünchau and colleagues extended their previous work on echopraxia (imitation) to children with TS (Brandt et al., 2017). Participants were asked to lift either their index or pinky finger when prompted by an auditory tone; simultaneously they were shown a compatible or incompatible visual stimulus. Children with TS were slower overall but thereby gained less interference from the incompatible stimulus. The authors conclude that these results suggest that children with TS may employ “different or additional inhibition strategies” than children without tics. Again, on this task children with TS show superior, not deficient, action inhibition.\n\nGanos et al. (2017) review treatments for TDs in children.\n\nPsychological interventions. Houghton et al. (2017) analyzed their existing data to test the longstanding theory that habituation to premonitory urges during adequate periods of tic suppression is the mechanism by which behavior therapies for tic disorders exerted their beneficial effects. In two previous randomized trials, 126 children and adolescents and 122 adults with TS or a persistent TD had been assigned to Comprehensive Behavioral Intervention for Tics (CBIT) or to education and supportive therapy. The prediction was that urge severity would decrease in CBIT responders. Surprisingly, however, children showed no significant reduction in premonitory urges with treatment, even though CBIT was quite effective in the child study, and declines in urges in adults had no association with clinical improvement or group assignment. The authors conclude that habituation cannot be the underlying process by which CBIT exerts its beneficial effects. This important result warrants further study.\n\nAnother report from the same data set examined what clinical features at baseline predicted improvement in tics during treatment (Sukhodolsky et al., 2017). Importantly, those treated with CBIT improved regardless of medication status, while in the control (supportive therapy) group, tics improved only in those taking medication for tics. Also importantly for the understanding of behavior therapy in TS and patient selection, other psychiatric symptoms, age, sex, family functioning, and expectation of improvement had no significant effects on benefit from therapy. And contrary to some opinions, patients with worse tic severity had significantly better improvement in tics with CBIT. Anxiety disorders and, surprisingly, more severe premonitory urges predicted less improvement.\n\nO’Connor and colleagues published a book describing their combined psychotherapeutic approach to tics (O’Connor et al., 2017).\n\nSeveral groups reported efforts to increase dissemination and use of behavior therapy for tics. An internet-based, therapist-guided behavior therapy for tics is being studied in the BiP-TIC project (Karlsson, 2016). The TicHelper.com internet-based CBIT program went live in late 2017 and received a positive review (Conelea & Wellen, 2017).\n\nMedication. A group of international experts provided a status report and recommendations for using brain imaging for the rational development of novel psychopharmacological interventions (Suhara et al., 2017). As an example, a fascinating study in pediatric ADHD showed that fMRI response to a cognitive task (Go/No-Go) strongly predicted better clinical response to methylphenidate than to atomoxetine (Schulz et al., 2017). As the authors conclude, “These data do not yet translate directly to the clinical setting, but the approach is potentially important for informing future research and illustrates that it may be possible to predict differential treatment response using a biomarker-driven approach.”\n\nInitial results from the first randomized, controlled trial (RCT) with a dopamine D1 receptor antagonist in pediatric TS were released by the sponsor in January, 2017 (Chipkin, 2017). These new results supported the positive results from a pilot study in adults with TS (Gilbert et al., 2014), and suggest a novel treatment mechanism for tics.\n\nIn April, the US FDA approved the presynaptic dopamine depleting agent valbenazine (Ingrezza®) for treatment of tardive dyskinesia (Neurocrine Biosciences, Inc. 2017b). TS is a likely off-label use for the drug, as the company has been conducting studies in children and adults with TS (ClinicalTrials.gov). The FDA designated valbenazine an orphan drug for pediatric patients with TS (Neurocrine Biosciences, Inc. 2017a). Another VMAT2 inhibitor, tetrabenazine, has been used for some time in the treatment of TS (Marsden, 1973; Sweet et al., 1974; Jankovic, 2016). A related compound, deutetrabenazine (Paton, 2017), showed initial positive results in TS (Jankovic et al., 2016), and the FDA approved it for treatment of tardive dyskinesia in August (Business Wire, 2017).\n\nAripiprazole has become a drug of choice in treating tics and comorbidities in TS over the past decade. However, large scale trials have been missing. Moreover, aripiprazole is not marketed for children and adolescents in many countries, regardless of the indication. Sallee et al. (2017) report on a phase 3, randomized, double-blind, placebo-controlled trial in 133 pediatric patients randomized in a 1:1:1 ratio to low-dose aripiprazole (5 mg/day if <50 kg; 10 mg/day if ≥50 kg), high-dose aripiprazole (10 mg/day if <50 kg; 20 mg/day if ≥50 kg), or placebo for 8 weeks. The primary efficacy endpoint was mean change from baseline to week 8 in the YGTSS-TTS. The Clinical Global Impression-Tourette’s Syndrome improvement score was also evaluated. High-dose aripiprazole was more effective than low-dose aripiprazole, and both were superior to placebo. Importantly, tolerance was overall good and no serious adverse events or deaths occurred, indicating that oral aripiprazole is a safe and effective treatment for tics in children and adolescents. These results provide important reassurance to clinicians who have been using aripiprazole for TS for years now. The placebo response rate (for the Clinical Global Impression scale), though half the response rate in the active treatment groups, was nevertheless surprisingly high (38%).\n\nA small (N=34) RCT of guanfacine showed no meaningful difference in effects on tic ratings or clinical impressions of improvement between the drug and placebo groups (Murphy et al., 2017). This result is important and surprising, given that adrenergic α2 agonists have been seen as first-line treatment for TS, especially in TS patients with ADHD (Hollis et al., 2016). Both these studies show how important RCTs are to clinical care in TS. Speaking of RCTs and guanfacine, a press release reported that extended-release guanfacine showed superiority to placebo in adults with ADHD (discussed here). The importance of this report comes primarily from the fact that data on ADHD treatments are scarcer in adults than in children.\n\nCannabinoids for TS are increasingly being studied; a brief summary of some of this work appears in Black (2017).\n\nNeurosurgery. A fascinating study demonstrated in mice that interfering electrical “beats” (similar to the beats one hears when tuning one instrument to another) can be used to steer neuron activation to focal sites in the brain without surgical electrode implantation (Grossman et al., 2017). Much work remains to be done to demonstrate feasibility, safety and efficacy in humans, but this approach potentially could lead to noninvasive, focal brain stimulation.\n\nWelter et al. (2017) performed a randomized, double-blind, controlled trial of deep brain stimulation (DBS) of the posterior and anterior internal globus pallidus for severe TS. The design was very similar to that reported previously by Kefalopoulou et al. (2015), and so were the results. The primary endpoint was the difference in YGTSS score between the beginning and end of the 3 month double-blind period. No significant differences between groups were noted in YGTSS score change between the beginning and the end of the 3 month double-blind period, despite a slight improvement in the stimulated condition. However, after the end of the open stimulation period and further parameter adjustment, results become significant compared to baseline. Also, when turning the stimulators off in a blinded fashion, YGTSS scores increased again to reach near baseline levels. Overall, this study is most important in terms of optimal study design. The stimulator programming period preceding the blinded phases was likely too short and the parameters suboptimal (a choice intended to reduce unblinding). Future studies will need to consider these results carefully.\n\nA London center reported an analysis of GPi DBS data, looking for the “sweet spot” for DBS for tic improvement (Akbarian-Tefaghi et al., 2017). They report that “a region within the ventral limbic GPi, specifically on the medial medullary lamina in the pallidum at the level of the AC-PC, was significantly associated with improved tics but not mood or OCB outcome.” Two patients with DBS to the fields of Forel experienced a good outcome (Neudorfer et al., 2017).\n\nTS patients undergoing DBS, independent of the surgical target, usually require high stimulation parameters, leading to short (2–3 years) battery life, meaning frequent and costly battery replacements. Michael Okun and colleagues showed proof of the principle that triggered rather than continuous DBS may be helpful (Molina et al., 2017). In a single patient with medically refractory TS, a spectral feature in the 5- to 15-Hz band was used as the control signal from bilateral leads in the centromedian-parafascicular (Cm-Pf) region of the thalamus. Significant tic improvement compared to baseline was observed 12 months after the procedure, similar to that obtained with continuous DBS, and resulted in a 63% improvement in the neurostimulator’s projected mean battery life. These so called closed-loop systems are gaining traction in various CNS diseases where DBS is applied, and this case report paves the way in TS.\n\nA report from the DBS group in the Netherlands called attention to side effects over the course of treatment in TS patients with thalamic DBS (Smeets et al., 2018). A subthalamic nucleus (STN) DBS study in OCD reminds us that DBS can cause side effects; STN stimulation at higher voltages caused chorea-ballismus (Mulders et al., 2017).\n\nQuality of life is lower in parents of children with TS and is more related to factors other than tic severity (Jalenques et al., 2017). Stewart et al. (2015) previously reported similar results. These studies emphasize the need to assess and treat symptoms other than tics in TS patients, and to care for the whole patient.\n\n\nConclusions\n\n2017 has again seen a rise in publications on TS, reflecting the increased interest this field receives, both from clinicians and researchers devoted to this disorder but also from adjoining fields, given the substantial psychopathology associated with TDs.\n\nOne important question raised is the natural course of TS and the rate of (non-)remission. Recent data suggest that the optimistic outlook prevalent with regard to tic severity when gliding from adolescence into adulthood might not be completely justified. Here, large longitudinal studies are warranted, and also to better understand the prognostic factors associated with tic remission or persistence. The fate of comorbidities also needs to be better understood, even though data have begun to emerge. We also still face important delays in the diagnosis of TS, and we need to deal with the recently demonstrated increased suicide risk in this condition, which underlines that TDs are far from benign.\n\nOf note, there is a rise in computational models of TS, which provide important leads to understanding its pathophysiology and will likely fuel more targeted treatment approaches for tics. In the therapy field, 2017 has mainly seen studies centered on well-known pharmacological targets such as the dopaminergic system; however, new antidopaminergic medications were studied or came to the market in 2017, and increasing data supports existing off-label prescription practices. Behavioral therapy continues to emerge as a main pillar of tic treatment, with an improved understanding of its mechanisms and response factors. In the surgical field, deep brain stimulation for severe TS cases continues to draw interest; in particular, target choice, length of stimulator programming for optimal outcome, and closed-loop systems are under active investigation.\n\nOverall, the field is active and burgeoning. Breakthroughs are to be expected in the upcoming years, especially with regard to large-scale efforts in the field.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Competing interests\n\n\n\nKJB participated in clinical trials supported by Neurocrine Biosciences, Inc., and Psyadon Pharmaceuticals, Inc.\n\n\nGrant information\n\nThis work was supported in part by the U.S. National Institutes of Health, grant R01 MH104030.\n\nThe authors confirm that the funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAbramovitch A, Hallion LS, Reese HE, et al.: Neurocognitive Predictors of Treatment Response to Randomized Treatment in Adults with Tic Disorders. Prog Neuropsychopharmacol Biol Psychiatry. 2017; 74: 9–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkbarian-Tefaghi L, Akram H, Johansson J, et al.: Refining the Deep Brain Stimulation Target within the Limbic Globus Pallidus Internus for Tourette Syndrome. Stereotact Funct Neurosurg. 2017; 95(4): 251–58. PubMed Abstract | Publisher Full Text\n\nAlbin RL: Tourette Syndrome: a Disorder of the Social Decision-Making Network. Brain. 2018; 141(2): 332–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAttwells S, Setiawan E, Wilson AA, et al.: Inflammation in the Neurocircuitry of Obsessive-Compulsive Disorder. JAMA Psychiatry. 2017; 74(8): 833–840. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAugustine EF, Adams HR, Mink JW: Screening Tools for Tic Disorders-Focus on Development or Implementation? Mov Disord. 2017; 32(6): 946. PubMed Abstract | Publisher Full Text\n\nBaldwin CL, Roberts DM, Barragan D, et al.: Detecting and Quantifying Mind Wandering during Simulated Driving. Front Hum Neurosci. Frontiers Media SA: 2017; 11: 406. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBäumer T, Sajin V, Münchau A: Childhood-Onset Movement Disorders: A Clinical Series of 606 Cases. Mov Disord Clin Pract. Wiley-Blackwell: 2016; 4(3): 437–40. Publisher Full Text\n\nBeste C, Münchau A: Tics and Tourette Syndrome - Surplus of Actions Rather than Disorder? Mov Disord. 2018; 33(2): 238–42. PubMed Abstract | Publisher Full Text\n\nBlack KJ: Smoke gets in your tics. Authorea. (updated 16 Jul 2018). Publisher Full Text\n\nBlack KJ, Black ER, Greene DJ, et al.: Provisional Tic Disorder: What to tell parents when their child first starts ticcing [version 1; referees: 3 approved]. F1000Res. 2016; 5: 696. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlack KJ, Mink JW: Response to levodopa challenge in Tourette syndrome. Mov Disord. 2000; 15(6): 1194–1198. PubMed Abstract | Publisher Full Text\n\nBortolato M, Di Giovanni G: Editorial on Special Issue: Animal Models of Tourette Syndrome. J Neurosci Methods. 2017; 292: 1. PubMed Abstract | Publisher Full Text\n\nBrandt VC, Moczydlowski A, Jonas M, et al.: Imitation inhibition in children with Tourette syndrome. J Neuropsychol. 2017. PubMed Abstract | Publisher Full Text\n\nBrandt VC, Stock AK, Münchau A, et al.: Evidence for enhanced multi-component behaviour in Tourette syndrome - an EEG study. Sci Rep. 2017; 7(1): 7722. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBruun RD, Budman CL: The course and prognosis of Tourette syndrome. Neurol Clin. 1997; 15(2): 291–98. PubMed Abstract | Publisher Full Text\n\nBurton FH: Back to the Future: Circuit-testing TS & OCD. J Neurosci Methods. 2017; 292: 2–11. PubMed Abstract | Publisher Full Text\n\nBusiness Wire: “Teva Announces FDA Approval of AUSTEDO® (Deutetrabenazine) Tablets for the Treatment of Tardive Dyskinesia in Adults”. 2017. Reference Source\n\nCaligiore D, Mannella F, Arbib MA, et al.: Dysfunctions of the basal ganglia-cerebellar-thalamo-cortical system produce motor tics in Tourette syndrome. PLoS Comput Biol. 2017; 13(3): e1005395. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCavanna AE, Black KJ, Hallett M, et al.: Neurobiology of the Premonitory Urge in Tourette’s Syndrome: Pathophysiology and Treatment Implications. J Neuropsychiatry Clin Neurosci. 2017; 29(2): 95–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChipkin RE: Psyadon Announces Positive Results from Phase 2b Clinical Study of Ecopipam for the Treatment of Tourette’s Syndrome in Children. 2017. Reference Source\n\nConceição VA, Dias Â, Farinha AC, et al.: Premonitory urges and tics in Tourette syndrome: computational mechanisms and neural correlates. Curr Opin Neurobiol. 2017; 46: 187–99. PubMed Abstract | Publisher Full Text\n\nConelea CA, Wellen BCM: Tic Treatment Goes Tech: A Review of TicHelper.com. Cogn Behav Pract. 2017; 24(3): 374–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDarrow SM, Grados M, Sandor P, et al.: Autism Spectrum Symptoms in a Tourette’s Disorder Sample. J Am Acad Child Adolesc Psychiatry. 2017a; 56(7): 610–617.e1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDarrow SM, Hirschtritt ME, Davis LK, et al.: Identification of Two Heritable Cross-Disorder Endophenotypes for Tourette Syndrome. Am J Psychiatry. 2017b; 174(4): 387–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEaton CK, Jones AM, Gutierrez-Colina AM, et al.: The Influence of Environmental Consequences and Internalizing Symptoms on Children’s Tic Severity. Child Psychiatry Hum Dev. 2017; 48(2): 327–34. PubMed Abstract | Publisher Full Text\n\nEddy CM, Rizzo R, Cavanna AE: Neuropsychological aspects of Tourette syndrome: a review. J Psychosom Res. 2009; 67(6): 503–13. PubMed Abstract | Publisher Full Text\n\nFernández de la Cruz L, Rydell M, Runeson B, et al.: Suicide in obsessive-compulsive disorder: a population-based study of 36 788 Swedish patients. Mol Psychiatry. 2017a; 22(11): 1626–1632. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFernández de la Cruz L, Rydell M, Runeson B, et al.: Suicide in Tourette’s and Chronic Tic Disorders. Biol Psychiatry. 2017b; 82(2): 111–18. PubMed Abstract | Publisher Full Text\n\nForde NJ, Zwiers MP, Naaijen J, et al.: Basal ganglia structure in Tourette’s disorder and/or attention-deficit/hyperactivity disorder. Mov Disord. 2017; 32(4): 601–4. PubMed Abstract | Publisher Full Text\n\nFowler SC, Mosher LJ, Godar SC, et al.: Assessment of gait and sensorimotor deficits in the D1CT-7 mouse model of Tourette syndrome. J Neurosci Methods. 2017; 292: 37–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGanos C, Martino D, Pringsheim T: Tics in the Pediatric Population: Pragmatic Management. Mov Disord Clin Pract. 2017; 4(2): 160–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGilbert DL, Budman CL, Singer HS, et al.: A D1 receptor antagonist, ecopipam, for treatment of tics in Tourette syndrome. Clin Neuropharmacol. 2014; 37(1): 26–30. PubMed Abstract | Publisher Full Text\n\nGilbert DL, Sethuraman G, Sine L, et al.: Tourette’s syndrome improvement with pergolide in a randomized, double-blind, Crossover trial. Neurology. 2000a; 54(6): 1310–15. PubMed Abstract | Publisher Full Text\n\nGilbert DL, Salleee FR, Sine L, et al.: Behavioral and Hormonal Effects of Low-Dose Pergolide in Children and Adolescents with Gilles De La Tourette’s Syndrome. Curr Ther Res Clin Exp. 2000b; 61(6): 378–87. Publisher Full Text\n\nGodar SC, Bortolato M: What makes you tic? Translational approaches to study the role of stress and contextual triggers in Tourette syndrome. Neurosci Biobehav Rev. 2017; 76(Pt A): 123–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGordon M, Cologne SE, Hartlein JM, et al.: A Pilot Study of Levodopa for Treatment of Tics in Children and Adults. In F1000Posters. 2013; 4: 668. Reference Source\n\nGreenberg E, Tung ES, Gauvin C, et al.: Prevalence and predictors of hair pulling disorder and excoriation disorder in Tourette syndrome. Eur Child Adolesc Psychiatry. 2018; 27(5): 569–579. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreene DJ, Williams Iii AC, Koller JM, et al.: Brain structure in pediatric Tourette syndrome. Mol Psychiatry. 2017; 22(7): 972–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrossman N, Bono D, Dedic N, et al.: Noninvasive Deep Brain Stimulation via Temporally Interfering Electric Fields. Cell. 2017; 169(6): 1029–41.e16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGroth C, Mol Debes N, Rask CU, et al.: Course of Tourette Syndrome and Comorbidities in a Large Prospective Clinical Study. J Am Acad Child Adolesc Psychiatry. 2017; 56(4): 304–12. PubMed Abstract | Publisher Full Text\n\nHollis C, Pennant M, Cuenca J, et al.: Clinical effectiveness and patient perspectives of different treatment strategies for tics in children and adolescents with Tourette syndrome: a systematic review and qualitative analysis. Health Technol Assess. 2016; 20(4): 1–450, vii–viii. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoughton DC, Capriotti MR, Scahill LD, et al.: Investigating Habituation to Premonitory Urges in Behavior Therapy for Tic Disorders. Behav Ther. 2017; 48(6): 834–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJalenques I, Auclair C, Morand D, et al.: Health-related quality of life, anxiety and depression in parents of adolescents with Gilles de la Tourette syndrome: a controlled study. Eur Child Adolesc Psychiatry. 2017; 26(5): 603–17. PubMed Abstract | Publisher Full Text\n\nJankovic J, Jimenez-Shahed J, Budman C, et al.: Deutetrabenazine in Tics Associated with Tourette Syndrome. Tremor Other Hyperkinet Mov (N Y). 2016; 6: 422. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJankovic J: Dopamine depleters in the treatment of hyperkinetic movement disorders. Expert Opin Pharmacother. 2016; 17(18): 2461–70. PubMed Abstract | Publisher Full Text\n\nKarlsson V: Using the internet to treat children with tics. Karolinska Institutet; previously published in Medicinsk Vetenskap number 3. (https://issuu.com/karolinska_institutet/docs/mv_nr_3_2016). Accessed: 2018-07-11. (Archived by WebCite® at http://www.webcitation.org/70pzwqN2o). 2016. Reference Source\n\nKefalopoulou Z, Zrinzo L, Jahanshahi M, et al.: Bilateral globus pallidus stimulation for severe Tourette’s syndrome: a double-blind, randomised crossover trial. Lancet Neurol. 2015; 14(6): 595–605. PubMed Abstract | Publisher Full Text\n\nKéri S, Szlobodnyik C, Benedek G, et al.: Probabilistic classification learning in Tourette syndrome. Neuropsychologia. 2002; 40(8): 1356–62. PubMed Abstract | Publisher Full Text\n\nKhalifa N, von Knorring AL: Psychopathology in a Swedish population of school children with tic disorders. J Am Acad Child Adolesc Psychiatry. 2006; 45(11): 1346–53. PubMed Abstract | Publisher Full Text\n\nKumar R, Lang AE: Coexistence of tics and parkinsonism: evidence for non-dopaminergic mechanisms in tic pathogenesis. Neurology. 1997; 49(6): 1699–1701. PubMed Abstract | Publisher Full Text\n\nLee WT, Huang HL, Wong LC, et al.: Tourette Syndrome as an Independent Risk Factor for Subsequent Sleep Disorders in Children: A Nationwide Population-Based Case-Control Study. Sleep. 2017; 40(3): zsw072. PubMed Abstract | Publisher Full Text\n\nLeivonen S, Scharf JM, Mathews CA, et al.: Parental Psychopathology and Tourette Syndrome/Chronic Tic Disorder in Offspring: A Nationwide Case-Control Study. J Am Acad Child Adolesc Psychiatry. 2017; 56(4): 297–303.e4. PubMed Abstract | Publisher Full Text\n\nMaia TV, Conceição VA: The Roles of Phasic and Tonic Dopamine in Tic Learning and Expression. Biol Psychiatry. 2017; 82(6): 401–12. PubMed Abstract | Publisher Full Text\n\nMarsden CD: Drug treatment of diseases characterized by abnormal movements. Proc R Soc Med. 1973; 66(9): 871–73. PubMed Abstract | Free Full Text\n\nMarsh R, Alexander GM, Packard MG, et al.: Habit learning in Tourette syndrome: a translational neuroscience approach to a developmental psychopathology. Arch Gen Psychiatry. 2004; 61(12): 1259–68. PubMed Abstract | Publisher Full Text\n\nMartinez-Torres I, Hariz MI, Zrinzo L, et al.: Improvement of tics after subthalamic nucleus deep brain stimulation. Neurology. 2009; 72(20): 1787–1789. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartino D, Pringsheim TM: Reply to \"Screening tools for tic disorders - Focus on development or implementation?\" Mov Disord. 2017; 32(6): 947. PubMed Abstract | Publisher Full Text\n\nMartino D, Delorme C, Pelosin E, et al.: Abnormal lateralization of fine motor actions in Tourette syndrome persists into adulthood. PLoS One. 2017a; 12(7): e0180812. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartino D, Ganos C, Pringsheim TM: Tourette Syndrome and Chronic Tic Disorders: The Clinical Spectrum Beyond Tics. Int Rev Neurobiol. 2017b; 134: 1461–90. PubMed Abstract | Publisher Full Text\n\nMartino D, Pringsheim TM, Cavanna AE, et al.: Systematic review of severity scales and screening instruments for tics: Critique and recommendations. Mov Disord. 2017c; 32(3): 467–473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeier SM, Dalsgaard S, Mortensen PB, et al.: Mortality risk in a nationwide cohort of individuals with tic disorders and with tourette syndrome. Mov Disord. 2017; 32(4): 605–9. PubMed Abstract | Publisher Full Text\n\nMolina R, Okun MS, Shute JB, et al.: Report of a patient undergoing chronic responsive deep brain stimulation for Tourette syndrome: proof of concept. J Neurosurg. 2017; 1–7. PubMed Abstract | Publisher Full Text\n\nMorand-Beaulieu S, Leclerc JB, Valois P, et al.: A Review of the Neuropsychological Dimensions of Tourette Syndrome. Brain Sci. 2017a; 7(8): pii: E106. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorand-Beaulieu S, Grot S, Lavoie J, et al.: The puzzling question of inhibitory control in Tourette syndrome: A meta-analysis. Neurosci Biobehav Rev. 2017b; 80: 240–62. PubMed Abstract | Publisher Full Text\n\nMulders AEP, Leentjens AFG, Schruers K, et al.: Choreatic Side Effects of Deep Brain Stimulation of the Anteromedial Subthalamic Nucleus for Treatment-Resistant Obsessive-Compulsive Disorder. World Neurosurg. 2017; 104: 1048.e9–1048.e13. PubMed Abstract | Publisher Full Text\n\nMurphy TK, Fernandez TV, Coffey BJ, et al.: Extended-Release Guanfacine Does Not Show a Large Effect on Tic Severity in Children with Chronic Tic Disorders. J Child Adolesc Psychopharmacol. 2017; 27(9): 762–70. PubMed Abstract | Publisher Full Text\n\nNeudorfer C, El Majdoub F, Hunsche S, et al.: Deep Brain Stimulation of the H Fields of Forel Alleviates Tics in Tourette Syndrome. Front Hum Neurosci. 2017; 11: 308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeurocrine Biosciences Inc: Neurocrine Granted FDA Orphan Drug Designation for Valbenazine for the Treatment of Pediatric Patients with Tourette Syndrome. 2017a. Reference Source\n\nNeurocrine Biosciences Inc: Press Release: Neurocrine Announces FDA Approval of INGREZZA (Valbenazine) Capsules as the First and Only Approved Treatment for Adults with Tardive Dyskinesia (TD). 2017b. Reference Source\n\nO’Connor KP, Lavoie ME, Schoendorff B: Managing Tic and Habit Disorders: A Cognitive Psychophysiological Approach with Acceptance Strategies. Hoboken, NJ: Wiley-Blackwell. 2017. Reference Source\n\nPaton DM: Deutetrabenazine: Treatment of hyperkinetic aspects of Huntington's disease, tardive dyskinesia and Tourette syndrome. Drugs Today (Barc). 2017; 53(2): 89–102. PubMed Abstract | Publisher Full Text\n\nPiacentini J, Bennett S, Capriotti M, et al.: CBIT-Jr: BEhavior Therapy for Chronic Tics in 5-8 Year Olds. Conference Proceedings. In 1st World Congress on Tourette Syndrome and Tic Disorders. 2015; 182–83 (Poster P101). Publisher Full Text\n\nPolyanska L, Critchley HD, Rae CL: Centrality of prefrontal and motor preparation cortices to Tourette Syndrome revealed by meta-analysis of task-based neuroimaging studies. Neuroimage Clin. 2017; 16: 257–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobertson MM, Eapen V, Singer HS, et al.: Gilles de la Tourette syndrome. Nat Rev Dis Primers. 2017; 3: 16097. PubMed Abstract | Publisher Full Text\n\nSallee F, Kohegyi E, Zhao J, et al.: Randomized, Double-Blind, Placebo-Controlled Trial Demonstrates the Efficacy and Safety of Oral Aripiprazole for the Treatment of Tourette’s Disorder in Children and Adolescents. J Child Adolesc Psychopharmacol. 2017; 27(9): 771–781. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalvador A, Worbe Y, Delorme C, et al.: Specific effect of a dopamine partial agonist on counterfactual learning: evidence from Gilles de la Tourette syndrome. Sci Rep. 2017; 7(1): 6292. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchaefer SM, Chow CA, Louis ED, et al.: Tic Exacerbation in Adults with Tourette Syndrome: A Case Series. Tremor Other Hyperkinet Mov (N Y). 2017; 7: 450. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchulz KP, Bédard AV, Fan J, et al.: Striatal Activation Predicts Differential Therapeutic Responses to Methylphenidate and Atomoxetine. J Am Acad Child Adolesc Psychiatry. 2017; 56(7): 602–9.e2. PubMed Abstract | Publisher Full Text\n\nShafer RL, Newell KM, Lewis MH, et al.: A Cohesive Framework for Motor Stereotypy in Typical and Atypical Development: The Role of Sensorimotor Integration. Front Integr Neurosci. Frontiers Media SA. 2017; 11: 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShale H, Fahn S, Mayeux R: Tics in a patient with Parkinson’s disease. Mov Disord. 1986; 1(1): 79–83. PubMed Abstract | Publisher Full Text\n\nShapiro AK, Shapiro E, Young JG, et al.: Gilles De La Tourette Syndrome. 2nd ed. Raven Press. 1988. Reference Source\n\nSinger HS: Discussing outcome in Tourette syndrome. Arch Pediatr Adolesc Med. 2006; 160(1): 103–5. PubMed Abstract | Publisher Full Text\n\nSmeets AYJM, Duits AA, Leentjens AFG, et al.: Thalamic Deep Brain Stimulation for Refractory Tourette Syndrome: Clinical Evidence for Increasing Disbalance of Therapeutic Effects and Side Effects at Long-Term Follow-Up. Neuromodulation. 2018; 21(2): 197–202. PubMed Abstract | Publisher Full Text\n\nStárková L: [Tics in Childhood]. Cesk Psychiatr. 1990; 86(5): 304–10. PubMed Abstract\n\nStewart SB, Greene DJ, Lessov-Schlaggar CN, et al.: Clinical correlates of parenting stress in children with Tourette syndrome and in typically developing children. J Pediatr. Elsevier BV: 2015; 166(5): 1297–1302.e3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuhara T, Chaki S, Kimura H, et al.: Strategies for Utilizing Neuroimaging Biomarkers in CNS Drug Discovery and Development: CINP/JSNP WOrking GRoup Report. Int J Neuropsychopharmacol. 2017; 20(4): 285–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSukhodolsky DG, Woods DW, Piacentini J, et al.: Moderators and predictors of response to behavior therapy for tics in Tourette syndrome. Neurology. 2017; 88(11): 1029–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSun N, Nasello C, Deng L, et al.: The PNKD gene is associated with Tourette Disorder or Tic disorder in a multiplex family. Mol Psychiatry. 2018; 23(6): 1487–1495. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSweet RD, Bruun R, Shapiro E, et al.: Presynaptic catecholamine antagonists as treatment for Tourette syndrome. Effects of alpha methyl para tyrosine and tetrabenazine. Arch Gen Psychiatry. 1974; 31(6): 857–61. PubMed Abstract | Publisher Full Text\n\nTakács Á, Shilon Y, Janacsek K, et al.: Procedural learning in Tourette syndrome, ADHD, and comorbid Tourette-ADHD: Evidence from a probabilistic sequence learning task. Brain Cogn. 2017; 117: 33–40. PubMed Abstract | Publisher Full Text\n\nWelter ML, Houeto JL, Thobois S, et al.: Anterior pallidal deep brain stimulation for Tourette's syndrome: a randomised, double-blind, controlled trial. Lancet Neurol. 2017; 16(8): 610–19. PubMed Abstract | Publisher Full Text\n\nWillsey AJ, Fernandez TV, Yu D, et al.: De Novo Coding Variants Are Strongly Associated with Tourette Disorder. Neuron. 2017; 94(3): 486–99.e9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWolmarans W, Scheepers IM, Stein DJ, et al.: Peromyscus maniculatus bairdii as a naturalistic mammalian model of obsessive-compulsive disorder: current status and future challenges. Metab Brain Dis. 2018; 33(2): 443–55. PubMed Abstract | Publisher Full Text\n\nXenos D, Kamceva M, Tomasi S, et al.: Loss of TrkB Signaling in Parvalbumin-Expressing Basket Cells Results in Network Activity Disruption and Abnormal Behavior. Cereb Cortex. 2017; 1–15. PubMed Abstract | Publisher Full Text\n\nYaniv A, Benaroya-Milshtein N, Steinberg T, et al.: Executive control development in Tourette syndrome and its role in tic reduction. Psychiatry Res. 2018; 262: 527–35. PubMed Abstract | Publisher Full Text\n\nZike ID, Chohan MO, Kopelman JM, et al.: OCD candidate gene SLC1A1/EAAT3 impacts basal ganglia-mediated activity and stereotypic behavior. Proc Natl Acad Sci U S A. 2017; 114(22): 5719–24. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "36381",
"date": "06 Aug 2018",
"name": "Keith A. Coffman",
"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 thoroughly enjoyed this manuscript, except for two paragraphs.\n\nFirst, on Page 2, right column, the following sentence should be incorporated into another paragraph or eliminated as it is awkward standing alone.\n\"Martino et al. (2017c) review screening instruments and rating scales for TDs; see also (Augustine et al.2017; Martino & Pringsheim, 2017).\"\nSecond, on Page 5, the paragraph that starts, \"A small (N=34) RCT of guanfacine showed no meaningful difference in effects on tic ratings or clinical impressions of improvement between the drug and placebo groups (Murphy et al., 2017)\" is incorrect. The study that Murphy and colleagues did was using EXTENDED RELEASE guanfacine, which has significantly less bioavailability than immediate release quanfacine. The point of their study was that the extended release preparation was ineffective, not that guanfacine was ineffective.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "3881",
"date": "07 Aug 2018",
"name": "Kevin J Black",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Thanks for the thoughtful review. We can address these points in the revision after we receive comments from other reviewers."
}
]
},
{
"id": "37059",
"date": "23 Aug 2018",
"name": "Lorena Fernández de la Cruz",
"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 written, concise, and useful manuscript summarising the literature published in TS during 2017. I commend the authors’ efforts and I am grateful for the opportunity to review this piece. I only have a few minor comments, listed below.\n1. The choice of articles is obviously subjective, but I think that the review could also cite the following register-based study from Finland, including 1195 cases, about parental age and the (lack of) association with tic disorders. I believe that the study is relevant from an etiological point of view since it contrasts with what it is known in other disorders (e.g., ASD, ADHD):\nChudal R, Leivonen S, Rintala H, Hinkka-Yli-Salomäki S, Sourander A (2017)1\n2. Regarding the Fernández de la Cruz et al. (2017b) paper on suicide, cited in the Epidemiology section, the authors write: “TDs in adults were associated with a four-fold higher risk of suicide…” I suggest that the authors delete “in adults”. This large epidemiological cohort included individuals of all ages, followed up for different periods of time, up to 44 years, and the outcomes may have occurred at any time during the follow-up, including childhood/adolescence.\n3. When mentioning the paper by Darrow et al. (2017a), the authors explain: “The presence of OCD was associated with higher scores on the social cognition and RRB subscales”. I assume that a higher score in the RRB subscale means more repetitive behaviours, but does an elevation in the social cognition subscale reflect more (better) or less (worse) social cognition? I take from the following sentence that it is probably worse, but please clarify.\n4. When referring to the BIP TIC project in the Psychological interventions section, you may want to cite the communication cited below, which was presented at the ESSTS conference in 2017, and/or the project’s registration (https://clinicaltrials.gov/ct2/show/NCT02864589) instead of the Karlsson, 2016 reference.\nAndrén, P. (2017, June). Development and evaluation of a therapist-guided, Internet-based behaviour therapy programme for young people with Tourette syndrome and chronic tic disorder: The BIP TIC programme. 10th European Conference on Tourette Syndrome and Tic Disorders / 2017 Annual Meeting of the European Society for the Study of Tourette Syndrome, Seville, Spain. [oral and poster presentation]\nOther minor points:\nThe acronym ADD is only used once in the text (in the Electrophysiology section) and is never spelled out. Please write “attention deficit disorder” instead. Please spell out TSPO the first time that is mentioned (Neuroimaging studies section).\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
},
{
"id": "37057",
"date": "10 Sep 2018",
"name": "Thomas Fernandez",
"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 brief review of the literature pertaining to Tourette Syndrome in 2017. This review appears to include the main studies from 2017, summarizing them accurately. They provide a conclusion section that highlights what the authors believe are the most important themes in research for 2017 and important areas to investigate for 2018. Overall, I feel that this is a nice way for researchers in the field to quickly come up to speed with what is being done in TS research, and a good collection of references for closer review.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1122
|
https://f1000research.com/articles/7-195/v1
|
15 Feb 18
|
{
"type": "Software Tool Article",
"title": "BED: a Biological Entity Dictionary based on a graph data model",
"authors": [
"Patrice Godard",
"Jonathan van Eyll",
"Jonathan van Eyll"
],
"abstract": "The understanding of molecular processes involved in a specific biological system can be significantly improved by combining and comparing different data set and knowledge resources. However these information sources often use different identification systems and an identifier conversion step is required before any integration effort. Mapping between identifiers is often provided by the reference information resources and several tools have been implemented to simplify their use. However these tools cannot be easily customized and optimized for any specific use. Also the information provided by different resources is not combined to increase the efficiency of the mapping process and deprecated identifiers from former version of databases are not taken into account. Finally finding automatically the most relevant path to map identifiers from one scope to the other is often not trivial. The Biological Entity Dictionary (BED) addresses these challenges by relying on a graph data model describing possible relationships between entities and their identifiers. This model has been implemented using Neo4j and an R package provides functions to query the graph but also to create and feed a custom instance of the database.",
"keywords": [
"genomics",
"transcriptomics",
"proteomics",
"RNA-seq",
"microarray",
"database",
"identifiers"
],
"content": "Introduction\n\nSince the advent of genome sequencing projects, many technologies have been developed to get access to different molecular information on a large scale and with high throughput. DNA micro-arrays are probably the archetype of such technology because of their historical impact on gathering data related to nucleic acids: genomic DNA and RNA. They triggered the emergence of “omics” fields of research such as genomics, epigenomics or transcriptomics. Lately massive parallel sequencing further increased the throughput of data generation related to nucleic acids by several orders of magnitude. In a different way, mass spectrometry-related technologies allow the identification and the quantification of many kinds of molecular entities such as metabolites and proteins. Many information systems have been developed to manage the exploding amount of data and knowledge related to biological molecular entities. These resources manage different aspects of the data. For example some are genome or proteome centered, whereas others are focused on molecular interactions and pathways. Thus all these resources rely on different identifier systems to organize the concepts of interest. The value of all the experimental data and all the knowledge collected in public or private resources is very high as such but is also often synergistically leveraged by their cross comparison in a dedicated manner. Indeed many datasets can be relevant when addressing the understanding of a specific biological system, a phenotypic trait or a disease for example. These datasets can focus on different biological entities such as transcripts or proteins in different tissues, conditions or organisms. Comparing all these data and integrating them with available knowledge requires the ability to map the identifiers on which each resource relies.\n\nTo achieve this task public and proprietary information systems provide mapping tables between their own identifiers and those from other resources. Furthermore many tools have been developed to facilitate the access to this information. Ensembl BioMarts (Kinsella et al., 2011), mygene (Wu et al., 2013), and g:Profiler (Reimand et al., 2016a) are popular examples among many others. However, as pointed out by van Iersel et al. (2010), these tools are generally dedicated to a particular domain and not necessarily relevant or complete for all research projects, and keeping them up-to-date can also be an issue. Recognizing these challenges van Iersel et al. (2010) proposed the BridgeDb framework providing to bioinformatics developers a standard interface between tools and mapping services and also allowing the easy integration of custom data by a transitivity mechanism.\n\nHere we present BED: a biological entity dictionary. BED has been developed to address three main challenges. The first one is related to the completeness of identifier mappings. Indeed direct mapping information provided by the different systems are not always complete and can be enriched by mappings provided by other resources. More interestingly direct mappings not identified by any of these resources can be indirectly inferred by using mappings to a third reference. For example, many human Ensembl gene identifiers are not directly mapped to any Entrez gene identifiers but such mapping can be inferred using respective mappings to HGNC identifiers. The second challenge is related to the mapping of deprecated identifiers. Indeed entity identifiers can change from one resource release to another. The identifier history is provided by some resources, such as Ensembl or the NCBI, but it is generally not used by mapping tools. The third challenge is related to the automation of the mapping process according to the relationships between the biological entities of interest. Indeed mapping between gene and protein identifier scopes should not be done the same way than two scopes of gene identifiers. Also converting identifiers from different organism should be possible using gene ortholog information.\n\nTo meet these challenges we designed a graph data model describing possible relationships between different biological entities and their identifiers. This data model has been implemented with the Neo4j® graph database (Neo4j inc, 2017) and conversion rules have been defined and coded in an R (R Core Team, 2017) package. We provide an instance of the BED database focused on human, mouse and rat organism but many functions are available to construct other instances tailored to other needs.\n\n\nMethods\n\nThe BED (Biological Entity Dictionary) system relies on a data model inspired by the central dogma of molecular biology (Crick, 1970) and describing relationships between molecular concepts usually manipulated in the frame of genomics studies (Figure 1). A biological entity identifier (BEID) can identify either a Gene (GeneID), a Transcript (TranscriptID), a Peptide (PeptideID) or an Object (ObjectID). Object entities can correspond to complex concepts coded by any number of genes (i.e. a protein complex or a molecular function). BEID are extracted from public or private databases (BEDB). BEDB can provide an Attribute related to each BEID. For example it can be the sequencing region provided by the Ensembl database (Zerbino et al., 2018) or the identifier status provided by Uniprot (The UniProt Consortium, 2017). BEID can have one or several associated names (BENames) and symbols (BESymbol). GeneID can have one or several homologs in other organisms belonging to the same GeneIDFamily. Many genomics platforms, such as micro-array, allow the identification of biological entity by using probes identified by ProbeID. In general, BEID can be targeted by several probes belonging to a Platform which is focused on one, and only one, type of entity (BEType) among those described above: Gene, Transcript, Peptide or Object. A BEType can have several BEType products but can be the product of at most one BEType. This constraint allows the unambiguous identification of the most relevant path to convert identifiers from one scope to another and is fulfilled by the current data model: peptides are only produced from transcripts, which are only produced from genes, which can also code for objects.\n\nThe model is shown as an Entity/Relationship (ER) diagram: entities correspond to graph nodes and relationships to graph edges. “ID” and “idx” indicate if the corresponding entity property is unique or indexed respectively. Some redundancies occur in this data model. Indeed some “value” properties are duplicated in upper case (“value_up”) in order to improve the performance of case-insensitive searches. Also the database of a BEID node is provided as a property to ensure uniqueness of the couples of “database” and “value” properties. The same approach has been applied for the “platform” property of ProbeID nodes.\n\nBEID identifying the same biological entity are related through three different kinds of relationship according to the information available in the source databases, and to the decision made by the database administrator about how to use them. Two BEID which corresponds_to each other both identify the same biological entity. A BEID which is_associated_to or which is_replaced_by another BEID does not directly identify any biological entity: the link is always indirect through one or several other BEID. Therefore, by design a BEID which is_associated_to or which is_replaced_by another BEID can be related to several different biological entities. It is not the case for other BEID which identify one and only one biological entity. This set of possible relationship allows the indirect mapping of different identifiers not necessarily provided by any integrated resource.\n\nIn order to efficiently leverage an indirect path through these different relationships the data model has been implemented in a Neo4j® graph database (Neo4j inc, 2017).\n\nTwo R (R Core Team, 2017) packages have been developed to feed and query the database. The first one, neo2R, provides low level functions to interact with Neo4j®. The second R package, BED, provides functions to feed and query the BED Neo4j® graph database according to the data model described above.\n\nMany functions are provided within the package to build a tailored BED database instance. These functions are not exported in order not to mislead the user when querying the database (which is the expected most frequent usage of the system). An R markdown document showing how to build a BED database instance for human, mouse and rat organisms is provided within the package. It can be adapted to other organisms or needs.\n\nBriefly these functions can be divided according to three main levels:\n\nThe lowest level function is the bedImport function which loads a table in the Neo4j® database according to a Cypher® query.\n\nFunctions of the second level allow loading identifiers and relationships tables ensuring the integrity of the data model.\n\nHighest level functions are helpers for loading information provided by some public resources in different specific formats.\n\nThe BED R package provides several functions to retrieve identifiers from different resources, and also to convert identifiers from one reference to another. These functions generate and call Cypher® queries on the Neo4j® database. Converting thousands of identifiers can take some time (generally a few seconds). Also such conversions are often recurrent and redundant. In order to improve the performance for such recurrent and redundant queries, a cache system has been implemented. The first time, the query is run on Neo4j® for all the relevant ID related to user input and the result is saved in a local file. Next time similar queries are requested, the system does not call Neo4j® but loads the cached results and filters it according to user input. By default the cache is flushed when the system detects inconsistencies with the BED database. It can also be manually flushed if needed.\n\nMinimal system requirements for running BED and neo2R R packages:\n\nR ≥ 3.4\n\nOperating system: Linux, macOS, Windows\n\nMemory ≥ 4GB RAM\n\nThe graph database has been implemented with Neo4j® version 3 (Neo4j inc, 2017). The BED R package depends on the following packages available in the Comprehensive R Archive Network (CRAN):\n\nvisNetwork (Almende et al., 2017)\n\ndplyr (Wickham et al., 2017)\n\nhtmltools (RStudio inc, 2017)\n\nDT (Xie, 2016)\n\nshiny (Chang et al., 2017)\n\nminiUI (Cheng, 2016)\n\nrstudioapi (Allaire et al., 2017)\n\n\nUse cases\n\nAn instance of the BED database (UCB-Human) has been built using the script provided in the BED R package and made available in a Docker® image (Docker inc, 2017) available here: https://hub.docker.com/r/patzaw/bed-ucb-human/\n\nThis instance used to exemplify the following use cases is focused on Homo sapiens, Mus musculus and Rattus norvegicus organisms and it has been built from the following resources:\n\nEnsembl (Zerbino et al., 2018)\n\nNCBI (NCBI Resource Coordinators, 2017)\n\nUniprot (The UniProt Consortium, 2017)\n\nbiomaRt (Durinck et al., 2009)\n\nGEOquery (Davis & Meltzer, 2007)\n\nClarivate Analytics MetaBase® (Clarivate Analytics, 2017)\n\nThe numbers of biological entity (BE) identifiers (BEID) available in this BED database instance and which can be mapped to each other are shown in Table 1. In total, 3,519,181 BEID are available in this BED instance. This number includes deprecated identifiers without successor and which therefore cannot be mapped to any other identifier. All the genomics platforms included in this BED database instance are shown in Table 2. They provide mapping to BEID from 354,205 ProbeID in total.\n\nNumbers have been split according to the BE type and the organism. Only BEID which can be mapped to each other are taken into account (e.g. excluding deprecated identifiers without successor).\n\nThe getBeIds function returns all BE identifiers from a specific scope. A scope is defined by the type of BE or probe, the source of the identifiers (database or platform) and the organism. For example, the following code returns all the Ensembl identifiers of human genes.\n\n\n\n\n\nThe id column corresponds to the BEID from the source of interest. The column named according to the BE type (in this case Gene) corresponds to the internal identifiers of the related BE. This internal identifier is not a stable reference that can be used as such. Nevertheless, it is useful to identify BEID identifying the same BE. In the example above even if most of Gene BE are identified by only one Ensembl gene BEID, many of them are identified by two or more (5,809 / 59,515 = 10%); 277 BE are even identified by more than 10 Ensembl BEID (Figure 2.a). In this case, most of these redundancies come from deprecated ID from former versions of the Ensembl database (version in use here: 91) and can be excluded by setting the restricted parameter to TRUE when calling the getBeIds function (Figure 2.b). However many BE are still identified by two or more current Ensembl BEID (2,715 / 59,515 = 5%). This result comes from the way the BED database is constructed: When two identifiers from the same resource correspond to the same identifier in another resource (correspond_to relationship in the data model), all these BEID are considered to identify the same BE.\n\na) All Ensembl gene ID. b) Current Ensembl gene ID (version 91).\n\nA complex example of such mapping is shown in Figure 3 mapping all the BEID of the human TAS2R8 gene which codes for a protein of the family of candidate taste receptors. There are three identifiers corresponding to this gene symbol in Ensembl. All these three identifiers correspond to the same Entrez gene and the same HGNC identifiers. All these BEID are thus considered to identify the same gene. It turns out that the three Ensembl BEID correspond to the same gene mapped on different sequence version of the chromosome 12: the canonical (ENSG00000121314), CHR_HSCHR12_2_CTG2 (ENSG00000272712) and CHR_HSCHR12_3_CTG2 (ENSG00000277316). This information provided by Ensembl is encoded in the seq_region attribute for each Ensembl BEID (see data model) and is used to define preferred BEID which are mapped on canonical version of chromosome sequences. The ENSG00000272712 identifier shows also a complex history in former Ensembl versions.\n\nBEID are shown as circle and gene symbol in the rounded box. The color legend is shown to the left of the figure. BEID surrounded in bold correspond to preferred identifiers. Solid arrows represent correspond_to and is_known_as relationships. Dotted arrows represent is_replaced_by and is_associated_to relationships. This graph has been drawn with the exploreBe function.\n\nThe main goal of BED is to convert identifiers from one scope to another easily, rapidly and with high completeness. It has been thought in order to allow recurring comparisons to each other of many lists of biological entities from various origins.\n\nThe function guessIdOrigin can be used to guess the scope of any list of identifiers. A simple example regarding the conversion of human Ensembl gene to human Entrez gene identifiers is shown below and discussed hereafter. By setting the restricted parameter to TRUE the converted BEID are restricted to current - non-deprecated - version of Entrez gene identifiers. Nevertheless all the input BEID are taken into account, current and deprecated ones.\n\n\n\nAmong all the 68,460 human Ensembl gene identifiers available in the database, 21,718 (32%) were not converted to any human Entrez gene identifier: 21,073 (33%) of the 64,661 non-deprecated and 645 (17%) of the 3,799 deprecated identifiers.\n\nThree other tools were used on January 04, 2018 to perform the same conversion task: biomaRt (Durinck et al., 2009; Kinsella et al., 2011), mygene (Mark et al., 2014; Wu et al., 2013), and gProfileR (Reimand et al., 2016a; Reimand, 2016b). At that time, biomaRt and mygene were based on the Ensembl 91 release whereas gProfileR was based on release 90.\n\nThe numbers of human Ensembl gene identifiers successfully converted by each method are compared in Figure 4. Five identifiers were only converted by gProfileR. They were provided by former versions of Ensembl or NCBI but are now deprecated in the current releases of these two resources. All the other gene identifiers converted by the different methods were also converted by BED. However, BED was able to map at least 17,912 more identifiers than all the other tools (Figure 4.a). A few of these mappings (3,154) are explained by the fact that BED is the only tool mapping deprecated identifiers to current versions. Nevertheless, even when focusing on the mapping of current versions of Ensembl identifiers BED was able to map 14,758 more identifiers than all the other tools (Figure 4.b). A few of these mappings (627) are directly provided by the NCBI. But most of them (14,131) are inferred from a mapping of the Ensembl and Entrez gene identifiers to the same HGNC (Gray et al., 2015) identifier.\n\nVenn diagrams showing the number of human Ensembl gene identifiers mapped to at least one human Entrez gene identifier by the different tested tools when focusing (a) on all 68,460 or (b) on current 64,661 BEID (Ensembl 91 release).\n\nA rough approximation of running times of the different methods is provided in Table 3. The aim of this table is to show that BED, as a dedicated and locally available tool, is a very efficient option to convert large lists of identifiers on the fly and recurrently. The aim of BED is to improve the efficiency of identifier conversion in a well defined context (organism, information resources of interest. . .) and not to replace biomaRt, mygene, gProfileR or other tools which provide many more features for many organisms and which should not be narrowed to this task for a complete comparison.\n\nThe BED convBeIds function can be used to convert identifiers from any available scope to any other one. It automatically find the most relevant path according to the considered biological entities. It allows elaborate mapping such as the conversion between probe identifiers from a platform focused on mouse transcripts into human protein identifiers. Because such mappings can be intricate, BED also provides a function to show the shortest relevant path between two different identifiers (Figure 5).\n\nThe legend is shown to the left of the figure. The red arrow represents the is_homolog_of relationship. This graph has been drawn with the exploreConvPath function.\n\nSome additional use cases and examples are provided in the BED R package vignette. Several functions are available for annotating BEID with symbols and names, again taking advantage of information related to connected identifiers. Other functions are also provided to seek relevant identifiers of a specific biological entity. These functions are used by a shiny (Chang et al., 2017) gadget (Figure 6) providing an interactive dictionary of BEID which is also made available as an Rstudio add-in (Allaire et al., 2017; Cheng, 2016).\n\nIn this example the user is looking after human Ensembl transcript identifiers corresponding to “il6”.\n\n\nConclusions\n\nBED is a system dedicated to the mapping between identifiers of molecular biological entities. It relies on a graph data model implemented with Neo4j® and on rules coded in an R package. BED leverages mapping information provided by different resources in order to increase the mapping efficiency between each of them. It also allows the mapping of deprecated identifiers. Rules are used to automatically convert identifiers from one scope to another using the most appropriate path.\n\nThe intent of BED is to be tailored to specific needs, and beside functions for querying the system, the BED R package provides functions to build custom instances of the database. Database instances can be locally installed or shared across a community. This design combined with a cache system makes BED efficient for converting large lists of identifiers from and to a large variety of scopes.\n\nBecause of our research field we provide an instance focused on human, mouse and rat organisms. This database instance can be directly used in relevant projects but it can also be enriched depending on user or community needs.\n\n\nSoftware availability\n\nLatest source code is available at:\n\nhttps://github.com/patzaw/BED\n\nhttps://github.com/patzaw/neo2R\n\nArchived source code as at time of publication:\n\nhttps://zenodo.org/badge/latestdoi/119707445 (Godard, 2018a)\n\nhttps://zenodo.org/badge/latestdoi/119698430 (Godard, 2018b)\n\nSoftware is available to use under a GPL-3 license",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was entirely supported by UCB Pharma. The authors declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nWe are grateful to Frédéric Vanclef, Malte Lucken, Liesbeth François, Matthew Page, Massimo de Francesco, and Marina Bessarabova for fruitful discussions and constructive criticisms.\n\n\nReferences\n\nAllaire JJ, Wickham H, Ushey K, et al.: rstudioapi: Safely Access the RStudio API. 2017. Reference Source\n\nAlmende BV, Thieurmel B, Robert T: visNetwork: Network Visualization using ’vis.js’ Library. 2017. Reference Source\n\nChang W, Cheng J, Allaire JJ, et al.: shiny: Web Application Framework for R.2017. Reference Source\n\nCheng J: miniUI: Shiny UI Widgets for Small Screens. 2016. Reference Source\n\nClarivate Analytics: MetaCore delivers high-quality biological systems content in context. 2017. Reference Source\n\nCRAN: The Comprehensive R Archive Network. Reference Source\n\nCrick F: Central dogma of molecular biology. Nature. 1970; 227(5258): 561–563. PubMed Abstract | Publisher Full Text\n\nDavis S, Meltzer PS: GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics. 2007; 23(14): 1846–1847. PubMed Abstract | Publisher Full Text\n\nDocker inc: Docker Community Edition. 2017. Reference Source\n\nDurinck S, Spellman PT, Birney E, et al.: Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009; 4(8): 1184–1191. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGodard P: patzaw/BED: Publication release (Version v1.0.0). Zenodo. 2018a. Data Source\n\nGodard P: patzaw/neo2R: Publication release (Version v1.0.0). Zenodo. 2018b. Data Source\n\nGray KA, Yates B, Seal RL, et al.: Genenames.org: the HGNC resources in 2015. Nucleic Acids Res. 2015; 43(Database issue): D1079–1085. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKinsella RJ, Kähäri A, Haider S, et al.: Ensembl BioMarts: a hub for data retrieval across taxonomic space. Database (Oxford). 2011; 2011: bar030. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMark A, Thompson R, Afrasiabi C, et al.: mygene: Access MyGene.Info_ services. 2014. Publisher Full Text\n\nNCBI Resource Coordinators: Database Resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2017; 45(D1): D12–D17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeo4j inc: Neo4j Community Edition. 2017. Reference Source\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria; 2017. Reference Source\n\nReimand J, Arak T, Adler P, et al.: g:Profiler-a web server for functional interpretation of gene lists (2016 update). Nucleic Acids Res. 2016a; 44(W1): W83–89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReimand J, Kolde R, Arak T: gProfileR: Interface to the ’g:Profiler’ Toolkit. 2016b. Reference Source\n\nRStudio inc: htmltools: Tools for HTML. 2017. Reference Source\n\nThe UniProt Consortium: UniProt: the universal protein knowledgebase. Nucleic Acids Res. 2017; 45(D1): D158–D169. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Iersel MP, Pico AR, Kelder T, et al.: The BridgeDb framework: standardized access to gene, protein and metabolite identifier mapping services. BMC Bioinformatics. 2010; 11: 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWickham H, Francois R, Henry L, et al.: dplyr: A Grammar of Data Manipulation. 2017. Reference Source\n\nWu C, Macleod I, Su AI: BioGPS and MyGene.info: organizing online, gene-centric information. Nucleic Acids Res. 2013; 41(Database issue): D561–565. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXie Y: DT: A Wrapper of the JavaScript Library ’DataTables’. 2016. Reference Source\n\nZerbino DR, Achuthan P, Akanni W, et al.: Ensembl 2018. Nucleic Acids Res. 2018; 46(D1): D754–D761. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "31026",
"date": "05 Mar 2018",
"name": "Denise Slenter",
"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 introduces BED a new identifier mapping tool. Using a graph database like Neo4j provides a fast way to query relationships between the biological entities and retrieve mappings of interest. The available source code is nicely documented and for bioinformaticians, setting up the database and running queries should be straight-forward.\n\nNevertheless, there are several major issues that we would like to comment on:\nAlready in the abstract, it is indicated that current tools cannot be easily customized and optimized for any specific use. It is unclear what the authors actually mean with this statement and how this is solved through BED. Further on it is also stated that current tools are generally dedicated to a particular domain, which is also true for BED. BED only focuses on gene related identifies (genes, transcripts, proteins) similar to mygene, Ensembl BioMart and g:Profiler.\n\nIn the introduction, three main challenges are mentioned which are addressed by BED.\n(1) Integration of mappings from different resources - very relevant but the difficult question is if transitive mappings are always biological meaningful. They can also lead to conflicting statements when resources show inconsistent relationships (we have experienced this when comparing Ensembl → UniProt and UniProt → Ensembl mappings) - how are you dealing with that? We want to state that mygene is also integrating mappings from multiple resources. (2) Mapping of deprecated identifiers - this is indeed an interesting problem when analysing older datasets and the visualization in Figure 3 can be very useful when running into such issues. While you mention that BED contains all deprecated identifiers, it is not discussed why g:Profiler has five deprecated identifiers that are not in BED (Figure 4). (3) Mapping scope - It is not clear why the automation of mapping between different scopes needs to be done differently and how BED is solving this. Importantly, BioMarts and mygene also provide easy ways to map between the different scopes (gene - gene / gene - protein / gene - homolog).\n\nFigure 3 - we believe that it would make sense to use two different edge styles for is_replaced_by and is_associated_to since they have very different meaning. Also check the layout (in this example, it looks like the blue node is placed over the edge from the purple to the light-purple node).\n\nFigure 5 - what do the bold borders of nodes mean in the network? Preferred identifiers? How are those selected? Additionally, when talking about the shortest relevant path, the arrows on the edges might be misleading and confusing (since there is no path from ILMN_1220595 to Q16552 taking the directionality into account).\n\nThe authors shortly mention the neo2R package to build the database. The functionality is not discussed in detail and it is unclear why the existing R package provided by Neo4j (https://neo4j.com/developer/r/ was not used. Neo4j can also be easily queried from other programming languages. Are you planning to provide APIs in other languages that would allow the integration in tools other than R?\n\nWhile the conversion rate from Ensembl to Entrez Gene is very interesting, we are missing a comparison between the tools for real research examples, e.g. selection of several datasets and mapping from probe to Ensembl identifier / Entrez Gene identifier (one of the most common use cases in R workflows). This is also mentioned under the criteria for a software tool article in F1000: “The article should provide examples of suitable input data sets and include an example of the output that can be expected from the tool and how this output should be interpreted.”\n\nIs it possible to only include edges from certain resources when performing the identifier conversion? Or do the users need to build their own database with only those selected resources?\n\nAs a final comment, we think that structure of the article is sometimes hard to follow and paragraphs are often not linked to each other. In the section “Converting identifiers” you state the following: “The aim of BED is to improve the efficiency of identifier conversion in a well defined context (organism, information resources of interest. . .) and not to replace biomaRt, mygene, gProfileR or other tools which provide many more features for many organisms and which should not be narrowed to this task for a complete comparison.” We believe that this efficiency, especially in the context of run time, is the key advantage of this tool and this should be made more clear in the article (abstract/intro/conclusion).\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3620",
"date": "27 Apr 2018",
"name": "Patrice Godard",
"role": "Author Response",
"response": "Thanks for having taken the time to review this article and for your constructive comments that will help us to improve its quality. We are working on a second version. In the mean time we would like to provide you some feedback about the different issues you arose and how we are going to take your comments into account in the second version of our manuscript. Also we would like to inform you that we are going to use an updated version of the BED instance based on version 92 of Ensembl (released in April). Thus numbers provided in the article will slightly change in the next version. Already in the abstract, it is indicated that current tools cannot be easily customized and optimized for any specific use. It is unclear what the authors actually mean with this statement and how this is solved through BED. Further on it is also stated that current tools are generally dedicated to a particular domain, which is also true for BED. BED only focuses on gene related identifies (genes, transcripts, proteins) similar to mygene, Ensembl BioMart and g:Profiler. This statement has also been questioned by the other referee although slightly differently. We wanted to highlight the point that the way the mapping is done by most of these resources (excepted BridgeDB) cannot be customized, optimized or extended by the user according to his knowledge or to internal, non-public or non-standard information. These tools are dedicated to a particular domain: they are focused on species, type of identifiers, and update frequencies (as stated in BridgeDB publication by van Iersel et al. (2010)). It’s convenient because ready to use but not flexible as BridgeDB or BED which allow an empowered user to focus on required information. We are going to modify this statement to make it less ambiguous in the next version. In the introduction, three main challenges are mentioned which are addressed by BED. (1) Integration of mappings from different resources - very relevant but the difficult question is if transitive mappings are always biological meaningful. They can also lead to conflicting statements when resources show inconsistent relationships (we have experienced this when comparing Ensembl → UniProt and UniProt → Ensembl mappings) - how are you dealing with that? We want to state that mygene is also integrating mappings from multiple resources. The transitivity mechanism is managed by the 2 following relationships: “corresponds_to” and “is_associated_to”. On one hand the “corresponds_to” relationships make the mapping transitive since 2 BEIDs connected through this kind of relationship are considered to identify the same BE. On the other hand a BEID which “is_associated_to” to another one does not automatically “identify” the same BE making this kind of relationship not available for indirect mappings. When the BED database is fed, the user chooses which relationship should be of type “corresponds_to” or of type “is_associated_to” for each resource taken into account. For example, in the instance we provide, cross references provided by Ensembl from Ensembl gene ID to Entrez, HGNC and Vega gene ID are considered as “corresponds_to” relationships whereas cross references to miRbase, Unigene and OMIM are considered as “is_associated_to” relationship. In Ensembl the Hs.745351 Unigene ID is mapped to ENSG00000184033 and to ENSG00000268651 Ensembl gene IDs which correspond to 2 different genes in Ensembl but also in Entrez and in HGNC and these genes are located on the same chromosome but at different positions. This Unigene identifier will be mapped to both Ensembl gene IDs but another external identifier mapped to only one of these 2 Ensembl gene ID won’t be mapped to the other (the association to Hs.745351 won’t be used indirectly). The cross references provided by Ensembl and Uniprot between Ensembl peptide IDs and Uniprot IDs are considered as “corresponds_to” relationship in the BED instance we provide. If mygene integrates mapping from multiple resources it does not apply transitive mapping between Ensembl, Entrez and HGNC gene IDs (as shown in figure 4) and it does not allow the user to do it. (2) Mapping of deprecated identifiers - this is indeed an interesting problem when analysing older datasets and the visualization in Figure 3 can be very useful when running into such issues. While you mention that BED contains all deprecated identifiers, it is not discussed why g:Profiler has five deprecated identifiers that are not in BED (Figure 4). These deprecated identifiers are not associated to any up-to-date identifier in Ensembl and as such they are not considered anymore for mapping in BED. We will develop this point in the next version of the article in order to make it clearer. (3) Mapping scope - It is not clear why the automation of mapping between different scopes needs to be done differently and how BED is solving this. Importantly, BioMarts and mygene also provide easy ways to map between the different scopes (gene - gene / gene - protein / gene - homolog). BED use the biological relationship between genes, transcript and peptides to convert identifiers. For example, when converting peptides identifiers from the same species it will use only mapping done at the peptide level and won’t use mapping to transcript and gene mapping. This strategy seems to be applied by biomaRt but not by mygene nor by gProfileR which map for example one Uniprot ID to all the Ensembl peptide ID coded by the same gene. For example the A6NI28 Uniprot identifier is unambiguously mapped to the ENSP00000298815 Ensembl peptide identifier by BED and biomaRt but is mapped to three additional Ensembl peptide identifiers (ENSP00000431776, ENSP00000434304 and ENSP00000435961 which are encoded by the same gene: ENSG00000165895) by mygene and gProfileR. Mapping biological entities identifier which are not genes from two different organisms using ortholog information requires at least two steps in biomaRt, mygene and gProfileR: one for find the ortholog gene and the other to find the relevant biological entity identifier. These two steps are integrated and transparent in BED. We will add clarifying sentences in the next version of the article to address this. Figure 3 - we believe that it would make sense to use two different edge styles for is_replaced_by and is_associated_to since they have very different meaning. Also check the layout (in this example, it looks like the blue node is placed over the edge from the purple to the light-purple node). The visNetwork library only provides 2 types of edges: solid or dash. And we would prefer not using too many colors for different types of relationships. The “is_replaced_by” and “is_associated_to” relationship can easily be differentiated using the colors of the nodes: if the nodes have the same color it is an “is_replaced_by” relationship”; if the nodes have different colors it is an “is_associated_to” relationship. In this kind of graph “is_known_as”, “identifies” (optional) or “targets” (optional) relationships can also be differentiated according to the shapes of the nodes. We will clarify this point in the figure legend. We will also fix the layout issue of figure 3 in the next version of the article. Figure 5 - what do the bold borders of nodes mean in the network? Preferred identifiers? How are those selected? Additionally, when talking about the shortest relevant path, the arrows on the edges might be misleading and confusing (since there is no path from ILMN_1220595 to Q16552 taking the directionality into account). Bold borders in this figure indeed meant preferred identifiers. In the first version of the BED instance we provided (bed-ucb-human:2018.01.03), the preferred status of RefSeq transcripts and peptides is determined according to the status field provided in the gene2refseq file provided by the NCBI. The ID is preferred if the status is “MODEL”. The way to define the preferred status of Entrez gene, RefSeq transcripts and peptides will change in the next version of this instance where we will consider the assembly information also provided in the gene2refseq file: identifiers associated to non-alternative assembly will be “preferred”. We will remove the arrows from the edges in figure 5 to avoid the confusion about the use of directionality to find a path between two identifiers. The authors shortly mention the neo2R package to build the database. The functionality is not discussed in detail and it is unclear why the existing R package provided by Neo4j (https://neo4j.com/developer/r/) was not used. Neo4j can also be easily queried from other programming languages. Are you planning to provide APIs in other languages that would allow the integration in tools other than R? Two reasons motivated our choice to develop neo2R : (i) The development of Rneo4j package was on hold for a long time period (according to github commits) and (ii) it used legacy cypher HTTP endpoint. We wanted to use the transactional HTTP endpoint as recommended in the neo4j documentation (https://neo4j.com/docs/rest-docs/current/#rest-api-cypher). As the scope of the article is on the biological entity mapping and not on Neo4j as such, we don’t want to put emphasis on this point because we don’t think we provide strong additional value at this level. We do not plan to provide API in other languages but we would be happy if it is done by other developers and we would be ready to help in this frame. Indeed one of the reason to make the BED package publicly available under a GPL-3 license is to allow the community to build on it and to improve it. While the conversion rate from Ensembl to Entrez Gene is very interesting, we are missing a comparison between the tools for real research examples, e.g. selection of several datasets and mapping from probe to Ensembl identifier / Entrez Gene identifier (one of the most common use cases in R workflows). This is also mentioned under the criteria for a software tool article in F1000: “The article should provide examples of suitable input data sets and include an example of the output that can be expected from the tool and how this output should be interpreted.” We will provide such an example in the next version. The example will be focused on the comparison of results from different experiments with different designs: different microarray platforms and organisms. Is it possible to only include edges from certain resources when performing the identifier conversion? Or do the users need to build their own database with only those selected resources? As mentioned here-above, the conversion strategy is defined when feeding the BED database and the use of the relationships: “corresponds_to” and “is_associated_to”. At the end-user level, refinements of mapping can be achieved by using the “restricted” (which focus the mapping to non-deprecated identifiers) and the “preFilter” (which focus the mapping to preferred identifiers) parameters. Also the “getDirectProduct” and “getDirectOrigin” functions allow the user to find direct products or direct origins of molecular biology processes. For example the direct products of an Ensembl gene ID will be Ensembl transcript IDs. This is particularly useful when the user wants to focus on canonical transcription or translation events when this information is available (this is the case for Ensembl transcripts and peptides). As a final comment, we think that structure of the article is sometimes hard to follow and paragraphs are often not linked to each other. In the section “Converting identifiers” you state the following: “The aim of BED is to improve the efficiency of identifier conversion in a well defined context (organism, information resources of interest. . .) and not to replace biomaRt, mygene, gProfileR or other tools which provide many more features for many organisms and which should not be narrowed to this task for a complete comparison.” We believe that this efficiency, especially in the context of run time, is the key advantage of this tool and this should be made more clear in the article (abstract/intro/conclusion). We adopted the structure recommended by F1000Research for a “software tool article”. Nevertheless we take note of this comment and we will try to improve the flow of the text in the next version of the article. We will put higher emphasis on the efficiency statement in the next version of the article."
}
]
},
{
"id": "31928",
"date": "26 Mar 2018",
"name": "T. Ian Simpson",
"expertise": [
"Reviewer Expertise Biological informatics",
"computational biology",
"neuroscience",
"statistics",
"machine learning"
],
"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 article the authors present BED, a biological entity database implemented as a Neo4J labelled property graph. In addition, they provide two R-packages (BED & neo4J) for the construction and query of such graphs that adhere to their data model. These packages include utility functions to facilitate graph construction from a range of commonly used and publicly available data sources. The software and database are well documented and available through GitHub and Docker (as a Docker image) respectively and proved straight forward to install and run.\nThere are several elements of the current manuscript that warrant commentary:\nMotivation/Rationale. The authors have correctly identified an important problem with the integration of biological data that has been addressed before, not least by the resources/tools mentioned in the manuscript (Biomart, my gene, g:Profiler amongst others). They have chosen a particularly good approach (labelled property graphs) to build the data architecture to address such a problem and one that has recently been used to great effect by the EMBL-EBI Reactome team to model data related to biological pathways. Currently this manuscript somewhat undersells the potential for the tools that have been developed. Whilst allusion is made at various points to the fact that the software developed could be used by others to develop custom resources very little is presented as to the suitability of their approach for such. The \"Abstract\" states that existing resources \"cannot be customised and optimised for any specific use\" which is not correct and should be removed or re-worded to clarify the author's meaning. Whilst the implementation presented here is focussed primarily on gene level mappings it should be made clear throughout the manuscript that the general approach used could be (and indeed has been, see citation) used in other really quite different biological data modelling scenarios.\n\nIntroduction. The issue of \"transitivity\" is raised here, this is a complex issue for many biological data types that are far removed from the rigid structures of ontologies that commonly enforce it by definition. The meaning of \"transitivity\" in the context used here is not clear and warrants further explanation. This is particularly important later in the article where decisions are being made about inferring mappings where they don't exist in the data. Some such inferences are entirely logical (e.g. using HGNC ids two link gene_ids between two resources that don't map directly to each other) but others are far more complex (e.g. mapping between species). The inclusion of deprecated identifiers is excellent and will help to close a notable gap in many existing resources for which mapping older data into more recent datasets can be extremely time consuming and frustrating. The authors comment on \"mapping between different scopes\" is unclear and should be clarified.\n\nMethods. The sections \"Feeding the Database\" and \"Querying the Database\" are very brief and would benefit from much more detail about the functionality of the database creation and query system. Whilst these are covered in detail in the various pieces of documentation (including some very nice working examples) there is not enough in the manuscript itself to allow the reader to assess the available functionality.\n\nUse Cases.\n\nThere appears to be a discrepancy in gene counts from the Ensembl examples used in this section; the first example calls human Ensembl genes and returns 59,515 genes the second states the total number of human Ensembl genes to be 68,460. Figure 3. illustrates a relationship graph including deprecated BEIDs. Whilst is_replaced_by is clear, it is not clear (or defined anywhere) what the meaning of is_associated_to is and how that differs from corresponds_to. This should be clarified in the text. The sentence \"The function guessIdOrigin...\" appears out of place, unconnected to the surrounding text. The statement \"Five identifiers were only...\" and the following sentence should be combined and re-worded so that the explanation as to why 5 BEIDs were uniquely found by gProfiler is clearer. No validation or commentary has been presented to test the efficacy of inferences made by the query system. I would like to have seen an attempt made to check the veracity of mappings made in this way especially when the majority (c.80%) of extra Ensmembl->EntrezID mappings recovered via BED were inferred. A \"rough approximation\" of timings for queries within BED and across other systems is not particularly informative. It would have been straightforward to automate a sampling approach to generate a mean response time (and a variance) to a defined set of query sizes/complexities to give the user a better understanding of how variable these response times are between the systems in practice. In addition, it would have been nice to see some analysis/discussion about the \"scalability\" of the system as this is likely to be of particular interest to end-users considering a similar modelling approach in other domains. Figure5. The meaning of directionality here is not clear. Whilst I can see the benefit for provenance reasons i.e. a mapping from EntrezGene to RefSeq it's meaning here is somewhat moot.\n\nThe work presented in this manuscript promises to be very useful for researchers wanting to use LPGs for data integration. The implementation and deployment have been very well executed so that they can be readily adopted and modified by end-users.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3621",
"date": "27 Apr 2018",
"name": "Patrice Godard",
"role": "Author Response",
"response": "Thanks for having taken the time to review this article and for your constructive comments that will help us to improve its quality. We are working on a second version. In the mean time we would like to provide you some feedback about the different issues you arose and how we are going to take your comments into account in the second version of our manuscript. Also we would like to inform you that we are going to use an updated version of the BED instance based on version 92 of Ensembl (released in April). Thus numbers provided in the article will slightly change in the next version. Motivation/Rationale. The authors have correctly identified an important problem with the integration of biological data that has been addressed before, not least by the resources/tools mentioned in the manuscript (Biomart, my gene, g:Profiler amongst others). They have chosen a particularly good approach (labelled property graphs) to build the data architecture to address such a problem and one that has recently been used to great effect by the EMBL-EBI Reactome team to model data related to biological pathways. Currently this manuscript somewhat undersells the potential for the tools that have been developed. Whilst allusion is made at various points to the fact that the software developed could be used by others to develop custom resources very little is presented as to the suitability of their approach for such. We will address this point in the conclusion of the next version of the article by mentioning in which context BED can be used. The \"Abstract\" states that existing resources \"cannot be customised and optimised for any specific use\" which is not correct and should be removed or re-worded to clarify the author's meaning. This statement has also been questioned by the other referee although slightly differently. We wanted to highlight the point that the way the mapping is done by most of these resources (excepted BridgeDB) cannot be customized, optimized or extended by the user according to his knowledge or to internal, non-public or non-standard information. These tools are dedicated to a particular domain: they are focused on species, type of identifiers, and update frequencies (as stated in BridgeDB publication by van Iersel et al. (2010)). It’s convenient because ready to use but not flexible as BridgeDB or BED which allow an empowered user to focus on required information. We are going to modify this statement to make it less ambiguous in the next version. Whilst the implementation presented here is focused primarily on gene level mappings it should be made clear throughout the manuscript that the general approach used could be (and indeed has been, see citation) used in other really quite different biological data modelling scenarios. We will mention in the introduction that graph databases, specially Neo4j, have been used to model different kind of biological data. Introduction. The issue of \"transitivity\" is raised here, this is a complex issue for many biological data types that are far removed from the rigid structures of ontologies that commonly enforce it by definition. The meaning of \"transitivity\" in the context used here is not clear and warrants further explanation. This is particularly important later in the article where decisions are being made about inferring mappings where they don't exist in the data. Some such inferences are entirely logical (e.g. using HGNC ids two link gene_ids between two resources that don't map directly to each other) but others are far more complex (e.g. mapping between species). The inclusion of deprecated identifiers is excellent and will help to close a notable gap in many existing resources for which mapping older data into more recent datasets can be extremely time consuming and frustrating. The authors comment on \"mapping between different scopes\" is unclear and should be clarified. The transitivity mechanism is managed by the 2 following relationships: “corresponds_to” and “is_associated_to”. On one hand the “corresponds_to” relationships make the mapping transitive since 2 BEIDs which are connected through this kind of relationship are considered to identify the same BE through an “identifies” relationship. On the other hand a BEID which “is_associated_to” to another one does not automatically “identify” the same BE making this kind of relationship not available for indirect mappings. When the BED database is fed, the user chooses which relationship should be of type “corresponds_to” or of type “is_associated_to”. For example, in the instance we provide, cross references provided by Ensembl from Ensembl gene ID to Entrez, HGNC and Vega gene ID are considered as “corresponds_to” relationships whereas cross references to miRbase, Unigene and OMIM are considered as “is_associated_to” relationship. In Ensembl the Hs.745351 Unigene ID is mapped to ENSG00000184033 and to ENSG00000268651 Ensembl gene IDs which correspond to 2 different genes in Ensembl but also in Entrez and in HGNC and these genes are located on the same chromosome but at different positions. This Unigene identifier will be mapped to both Ensembl gene IDs but another external identifier mapped to only one of these 2 Ensembl gene ID won’t be mapped to the other (the association to Hs.745351 won’t be used indirectly). An identifier scope is defined by the type of BE or probe, the source of the identifiers (database or platform) and the organism. Two scopes are different when at least one of these three elements is different. Mapping is the process to identify equivalent identifiers in two different scopes. This definition comes too late and is spread in the current version of the article. We will improve it in the next version. Methods. The sections \"Feeding the Database\" and \"Querying the Database\" are very brief and would benefit from much more detail about the functionality of the database creation and query system. Whilst these are covered in detail in the various pieces of documentation (including some very nice working examples) there is not enough in the manuscript itself to allow the reader to assess the available functionality. We will list the available functions (at least the most relevant ones) in the next version of the article. Use Cases. There appears to be a discrepancy in gene counts from the Ensembl examples used in this section; the first example calls human Ensembl genes and returns 59,515 genes the second states the total number of human Ensembl genes to be 68,460. 59,515 corresponds to the number of BE (Gene in this case). 68,460 corresponds to the number of BEIDs (Ensemble gene IDs in this case). As explained multiple BEID can identify the same BE. In other words 59,515 BE are identified by 68,460 BEID. Figure 3. illustrates a relationship graph including deprecated BEIDs. Whilst is_replaced_by is clear, it is not clear (or defined anywhere) what the meaning of is_associated_to is and how that differs from corresponds_to. This should be clarified in the text. See our answer about transitivity. We will make it clearer in the next version of the article. The sentence \"The function guessIdOrigin...\" appears out of place, unconnected to the surrounding text. We agree and it will be moved to another place in the next version of the article (probably in the “Additional features” section). The statement \"Five identifiers were only...\" and the following sentence should be combined and re-worded so that the explanation as to why 5 BEIDs were uniquely found by gProfiler is clearer. We will refine this part to address your comment and the similar one raised by the other reviewer. No validation or commentary has been presented to test the efficacy of inferences made by the query system. I would like to have seen an attempt made to check the veracity of mappings made in this way especially when the majority (c.80%) of extra Ensembl->EntrezID mappings recovered via BED were inferred. This kind of validation is quite difficult. We propose to use gene coordinates provided by the NCBI and Ensembl in order to compare the position on chromosomes of genes which are mapped by the different tool. Two mapped gene identifiers should have identical or similar locations. We will add these results in the next version of the article. A \"rough approximation\" of timings for queries within BED and across other systems is not particularly informative. It would have been straightforward to automate a sampling approach to generate a mean response time (and a variance) to a defined set of query sizes/complexities to give the user a better understanding of how variable these response times are between the systems in practice. In addition, it would have been nice to see some analysis/discussion about the \"scalability\" of the system as this is likely to be of particular interest to end-users considering a similar modelling approach in other domains. We will make the analysis of mean response time with different kinds of queries and we will incorporate the results in the next version of the article. Scalability is not discussed because it highly depends on the graph database system. Here we use Neo4j for which scalability depends on the edition, community or enterprise (https://neo4j.com/subscriptions/). Figure5. The meaning of directionality here is not clear. Whilst I can see the benefit for provenance reasons i.e. a mapping from EntrezGene to RefSeq its meaning here is somewhat moot. This comment has also been made by the other reviewer. We will remove the arrows from the edges in figure 5 to avoid the confusion about the use of directionality to find a path between two identifiers."
}
]
}
] | 1
|
https://f1000research.com/articles/7-195
|
https://f1000research.com/articles/7-159/v1
|
07 Feb 18
|
{
"type": "Research Note",
"title": "Seasonality of birth defects in West Africa: could congenital Zika syndrome be to blame?",
"authors": [
"Maimuna S. Majumder",
"Rosanna Hess",
"Ratchneewan Ross",
"Helen Piontkivska",
"Maimuna S. Majumder",
"Rosanna Hess",
"Ratchneewan Ross"
],
"abstract": "The link between Zika virus infection during pregnancy and microcephaly and other neurodevelopmental defects in infants, referred to as congenital Zika syndrome (CZS), was recently discovered. One key question that remains is whether such neurodevelopmental abnormalities are limited to the recently evolved Asiatic ZIKV strains or if they can also be induced by endemic African strains. Thus, we examined birth registries from one particular hospital from a country in West Africa, where ZIKV is endemic. Results showed a seasonal pattern of birth defects that is consistent with potential CZS, which correspond to a range of presumed maternal infection that encompasses both the peak of the warm, rainy season as well as the months immediately following it, when mosquito activity is likely high. While we refrain from definitively linking ZIKV infection and birth defects in West Africa at this time, in part due to scant data available from the region, we hope that this report will initiate broader surveillance efforts that may help shed light onto mechanisms underlying CZS.",
"keywords": [
"Zika virus",
"ZIKV",
"birth defects",
"microcephaly",
"congenital Zika syndrome",
"West Africa",
"seasonality"
],
"content": "Introduction\n\nSince 2015, when an initial link between Zika Virus (ZIKV) and microcephaly was discovered in Brazil, the term congenital Zika syndrome (CZS) has been coined to reflect a broad range of Zika-linked neurodevelopmental damages1 beyond microcephaly1–3, including ocular4,5 and auditory defects6,7. The overall risk of Zika-linked birth defects has been estimated at ~10% and 15% for infections during the 1st trimester8, potentially impacting thousands of infants in the US and US territories alone. Concerns exist that ZIKV-related outcomes are underreported, particularly when ZIKV infections result in neurodevelopmental abnormalities without visible microcephaly (e.g., developmental delays and learning disabilities that would not be immediately noticeable9–11). Such outcomes are likely not recorded, particularly if the causative infection was asymptomatic12. Further, a recent CDC report that examined birth defect records from 15 US jurisdictions showed a statistically significant increase in prevalence of birth defects potentially consistent with CZS in areas with documented local ZIKV transmission in the second half of 201613. Notably, the majority of infants and fetuses with birth defects potentially related to ZIKV infection in this report lacked ZIKV infection testing, which may be in part attributed to lack of known maternal exposure or other such indicators14. Nonetheless, these findings are alarming and underscore not only the need for continued monitoring and surveillance, but also the need to better understand the full extent – as well as mechanisms – of neurodevelopmental defects associated with CZS.\n\n\nCan we expect to find cases of congenital Zika syndrome in the African Continent?\n\nCurrently our understanding of the mechanisms that underlie CZS remains limited, including the possibility that ZIKV infection is a necessary but insufficient condition for CZS15. One key question is whether ZIKV-mediated birth defects are associated with a specific strain of ZIKV, which, for example, could have evolved after ZIKV migrated from South East Asia to French Polynesia and Brazil16. However, other studies suggest that African strains are likely to be as pathogenic as the Asiatic strains (e.g., 17,18). Thus, the lack of connection between ZIKV infection in pregnancy and birth defects prior to the 2015–16 Brazilian outbreak may instead be attributable to the benign nature of ZIKV infection in adults19 and lack of surveillance, among other factors. This can be illustrated by a report of a handful of birth defects in Hawaii in 2009–2012, which now has been shown to be associated with ZIVK infection, but was undetected as such until the Brazilian microcephaly epidemic brought ZIKV into the spotlight20.\n\nEarlier studies from the African Continent – where ZIKV is endemic – have documented a relatively high prevalence of ZIKV antibodies in human populations (e.g., Nigeria21, Sierra-Leone22). Despite this, there exists negligible data regarding CZS across the African Continent; nevertheless, lack of evidence should not be taken as definitive proof of absence17. Thus, here we examine birth registries from one particular hospital in West Africa from a country considered by the WHO to be at medium risk of a ZIKV outbreak23. As there are ongoing security concerns in this location, to ensure the safety of the hospital and staff, the hospital name and location have been kept anonymous. The study was approved by the Committee on Administration of the hospital in lieu of a functioning Ethics Committee.\n\n\nCase study: seasonality of CZS-type birth defects in a hospital in West Africa\n\nRisk of major neurodevelopmental defects, such as microcephaly, appears to be particularly high if vertical transmission occurs during the first trimester, especially within a “vulnerability window” around 12 weeks (10 to 14 weeks) post-conception8,24. Thus, we hypothesized that – similar to seasonal malaria infections, which peak a few weeks following abundant rainfalls during the “rainy” season, typically from August through October in the study region25,26 – seasonal variations in the number of CZS-type birth defects would be detectable from the aforementioned hospital data. Such expectations are consistent with prior findings from Senegal (West Africa)27 and Kenya (East Africa)28, where Rift Valley Fever epizootics were associated with heavy rainfalls. Our hypothesis was further informed by the temporal relationships between the number of ZIKV infections and microcephaly cases reported in Brazil29. We expected that the peak of CZS-type birth defects (such as documented cases of microcephaly and/or stillbirth) would coincide with vertical ZIKV transmission at around 12-weeks post-conception during the peak of the rainy season, assuming an approximate 3-week lag between maternal infection and vertical transmission30, as suggested by data from Brazil in 201531.\n\n\nMethods\n\nA total of 13445 birth registries (2009–2015) from a non-governmental hospital in West Africa were examined to determine whether we could identify a seasonal pattern of birth defects potentially attributable to ZIKV infection (i.e., CZS-type birth defects). The number of births and respective outcomes (i.e., live birth versus stillbirth) and reported complications (i.e., fetal malformation, breech, etc.) were collated by month/year. Reporting standards for birth complications varied between years; thus, we focused exclusively on visible neurodevelopmental complications (such as microcephaly) and pregnancy losses (such as stillbirth) that could be attributed to potential CZS1,2 (Supplementary Table 1 and Supplementary Table 2). To infer the “vulnerability window” of 12 weeks (spanning 10 to 14 weeks) post-conception, we assumed that births that were not reported as premature in the records were full-term, thus enabling us to infer the likely month of conception29.\n\nWe also considered national average monthly temperature and rainfall data for the study years, collected from the World Bank Climate Change Knowledge Portal database. These values were treated as proxy indicators for mosquito activity in the hospital catchment area at time of maternal infection. To visualize any relevant trends, we plotted these data, as well as the average percentage of birth defects consistent with potential CZS by month.\n\n\nResults and discussion\n\nAs shown in Figure 1, the average percentage of births consistent with potential CZS demonstrates a marked peak between March and July, which places maternal month of infection between August and December of the previous year. These months encompass the peak and latter half of the warm, rainy season (August–October) as well as the first half of the cool, dry season that immediately follows (October–December) in the study region, which likely represent months with considerable mosquito activity. Notably, the hospital from which these birth defects data were acquired generally experiences a peak in childhood malaria cases every October, which falls squarely in the middle of the August to December range of presumed maternal month of ZIKV infection determined here.\n\nWith this in mind, the early months of the cool, dry season (October–December) are likely hospitable enough for mosquito vectors to thrive and spread pathogens, including ZIKV; this may explain why the range of our presumed maternal month of infection (August–December) extends past the rainy season, given that mosquitoes require a balanced environment for survival, including both moderate rainfall and optimal temperatures32.\n\nOur results suggest that a seasonal pattern exists with respect to CZS-type birth defects reported in the study region (Figure 1), where the largest fraction of said defects appear to occur in the months of March through July. Furthermore, this pattern can be linked to ecological evidence, such as rainfall and temperature trends that likely facilitate maternal ZIKV infection. While consistent with the expectation that some of these defects might be attributable to (unreported) ZIKV infections that occurred early in pregnancy (and indeed resemble temporal patterns from studies in Brazil29), our findings stop short of definitively linking ZIKV infection and birth defects in the study region, in part due to scant data. Instead, by reviewing the potential limitations of the data analyzed here, we hope that this report will initiate broader surveillance efforts that may help shed light onto mechanisms that underlie CZS, including utilization of data that might already be available across various African countries where ZIKV is endemic and/or competent vectors exist (e.g., Gabon33, Central African Republic34). For example, a recent WHO Bulletin on Outbreaks and Other Emergencies35 reported a number of microcephaly cases from Angola (a country listed in the High risk category23) that appear to be linked to ZIKV infection, despite the lack of direct PCR confirmation from the specimens. Due to only recent implementation of active surveillance in the country, the true magnitude of the event is not yet clearly understood. Nonetheless, it is important for insights into a broader pattern of potential CZS defects, despite the lack of experimental ZIKV infection confirmation or lack of evidence of ongoing active ZIKV transmission in Angola (i.e. only two ZIKV cases were reported from Angola in early 201736).\n\nSeveral conservative assumptions made in this analysis, such as assuming a gestation period of ~9 months, or not classifying “low birth weights” (which may represent full-term births of ZIKV-infected fetuses) as a CZS-type birth defect, would likely lead to underestimation of potential trends, if any. We also assumed that the available birth records were representative of the pregnancy/birth patterns that occur across the entire region. Other limitations of the available data are related to the standard of care that is feasible in much of West Africa, including (i) lack of family history and/or genetic testing for mutations in loci responsible for primary microcephaly; (ii) lack of laboratory evidence or testing for ZIKV and/or other infections, including TORCH agents37,38, often due to inability to pay for testing (e.g., 39); and (iii) lack of detailed clinical prenatal history, including whether rash and/or other symptoms of Zika infection were present at any point during pregnancy. This final limitation may be considered minor, given that the majority of ZIKV infections are asymptomatic19,24. Additionally, no data were available regarding other clinically relevant factors that are also associated with microcephaly40, such as history of excessive alcohol consumption or recreational drug use, and/or prolonged exposure to pesticides, such as pyriproxyfen. However, the former life-style factors are unlikely to have a seasonal effect spanning several years, and the role of the latter factor as a causative agent of microcephaly remains unclear41. There is also a lack of precise ecological data, including estimates of rainfalls in the hospital catchment area, the distribution and feeding habits of mosquitoes, and whether or not said mosquitoes carry ZIKV, as well as data regarding ZIKV prevalence in the human population.\n\nDespite these limitations, our findings suggest that using the data we already have – even in the absence of formal surveillance systems for CZS – can provide compelling, introductory insights. In the future, work that employs existing data from hospitals across the African continent – which encompasses countries with a variety of climates, dry and rainy seasons, and suitability for widespread mosquito habitats – should be pursued.\n\n\nData availability\n\nFigshare : Data for Figure 1. Average birth defects consistent with congenital Zika syndrome per month, with corresponding average daily temperatures and rainfall during presumed maternal month of infection. doi: 10.6084/m9.figshare.5387029. Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Competing interests\n\n\n\nAll authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work was partially supported by the Research Seed Award from Kent State University (to HP) and by Research For Health, Inc. (to RH).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplemental Table 1. A list of birth complications (as recorded in the health records) consistent with potential congenital Zika syndrome.\n\nClick here to access the data.\n\nSupplemental Table 2. A list of birth complications (as recorded in the health records) that are not consistent with potential congenital Zika syndrome and/or missing.\n\nClick here to access the data.\n\n\nReferences\n\nMelo AS, Aguiar RS, Amorim MM, et al.: Congenital Zika Virus Infection: Beyond Neonatal Microcephaly. JAMA Neurol. 2016; 73(12): 1407–1416. PubMed Abstract | Publisher Full Text\n\nLucey D, Cummins H, Sholts S: Congenital Zika Syndrome in 2017. JAMA. 2017; 317(13): 1368–1369. PubMed Abstract | Publisher Full Text\n\nMoore CA, Staples JE, Dobyns WB, et al.: Characterizing the Pattern of Anomalies in Congenital Zika Syndrome for Pediatric Clinicians. JAMA Pediatr. 2017; 171(3): 288–295. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSingh PK, Guest JM, Kanwar M, et al.: Zika virus infects cells lining the blood-retinal barrier and causes chorioretinal atrophy in mouse eyes. JCI Insight. 2017; 2(4): e92340. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCuljat M, Darling SE, Nerurkar VR, et al.: Clinical and Imaging Findings in an Infant With Zika Embryopathy. Clin Infect Dis. 2016; 63(6): 805–811. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlantz G: Zika and Hearing Loss: The Race for a Cure Continues. Hear J. 2016; 69(12): 22–24. Publisher Full Text\n\nLeal MC, Muniz LF, Ferreira TS, et al.: Hearing Loss in Infants with Microcephaly and Evidence of Congenital Zika Virus Infection - Brazil, November 2015–May 2016. MMWR Morb Mortal Wkly Rep. 2016; 65(34): 917–919. PubMed Abstract | Publisher Full Text\n\nReynolds MR, Jones AM, Petersen EE, et al.: Vital Signs: Update on Zika Virus-Associated Birth Defects and Evaluation of All U.S. Infants with Congenital Zika Virus Exposure - U.S. Zika Pregnancy Registry, 2016. MMWR Morb Mortal Wkly Rep. 2017; 66(13): 366–373. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVogel G: INFECTIOUS DISEASE. Experts fear Zika’s effects may be even worse than thought. Science. 2016; 352(6292): 1375–1376. PubMed Abstract | Publisher Full Text\n\nHotez PJ: What Does Zika Virus Mean for the Children of the Americas? JAMA Pediatr. 2016; 170(8): 787–789. PubMed Abstract | Publisher Full Text\n\nSoares de Souza A, Moraes Dias C, Braga FD, et al.: Fetal Infection by Zika Virus in the Third Trimester: Report of 2 Cases. Clin Infect Dis. 2016; 63(12): 1622–1625. PubMed Abstract | Publisher Full Text\n\nHotez P, Aksoy S: Will Zika become the 2016 NTD of the Year? Blog, posted January 7, 2016 by Peter Hotez and Serap Aksoy. 2016. Reference Source\n\nDelaney A, Mai C, Smoots A, et al.: Population-Based Surveillance of Birth Defects Potentially Related to Zika Virus Infection - 15 States and U.S. Territories, 2016. MMWR Morb Mortal Wkly Rep. 2018; 67(3): 91–96. PubMed Abstract | Publisher Full Text\n\nFitzgerald B, Boyle C, Honein MA: Birth defects potentially related to Zika virus infection during pregnancy in the United States. JAMA. 2018. PubMed Abstract | Publisher Full Text\n\nde Oliveira WK, Carmo EH, Henriques CM, et al.: Zika Virus Infection and Associated Neurologic Disorders in Brazil. N Engl J Med. 2017; 376(16): 1591–1593. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFaria NR, Quick J, Claro IM, et al.: Establishment and cryptic transmission of Zika virus in Brazil and the Americas. Nature. 2017; 546(7658): 406–410. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNutt C, Adams P: Zika in Africa-the invisible epidemic? Lancet. 2017; 389(10079): 1595–1596. PubMed Abstract | Publisher Full Text\n\nZhu Z, Chan JF, Tee KM, et al.: Comparative genomic analysis of pre-epidemic and epidemic Zika virus strains for virological factors potentially associated with the rapidly expanding epidemic. Emerg Microbes Infect. 2016; 5: e22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuffy MR, Chen TH, Hancock WT, et al.: Zika virus outbreak on Yap Island, Federated States of Micronesia. N Engl J Med. 2009; 360(24): 2536–2543. PubMed Abstract | Publisher Full Text\n\nKumar M, Ching L, Astern J, et al.: Prevalence of Antibodies to Zika Virus in Mothers from Hawaii Who Delivered Babies with and without Microcephaly between 2009–2012. PLoS Negl Trop Dis. 2016; 10(12): e0005262. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFagbami AH: Zika virus infections in Nigeria: virological and seroepidemiological investigations in Oyo State. J Hyg (Lond). 1979; 83(2): 213–219. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobin Y, Mouchet J: [Serological and entomological study on yellow fever in Sierra Leone]. Bull Soc Pathol Exot Filiales. 1975; 68(3): 249–258. PubMed Abstract\n\nWHO, World Health Organization: Zika Virus risk assessment in the WHO African Region: a technical report. February 2016. 2016. Reference Source\n\nHonein MA, Dawson AL, Petersen EE, et al.: Birth Defects Among Fetuses and Infants of US Women With Evidence of Possible Zika Virus Infection During Pregnancy. JAMA. 2017; 317(1): 59–68. PubMed Abstract | Publisher Full Text\n\nBrewster DR, Greenwood BM: Seasonal variation of paediatric diseases in The Gambia, west Africa. Ann Trop Paediatr. 1993; 13(2): 133–146. PubMed Abstract | Publisher Full Text\n\nSivakumar MVK: Predicting rainy season potential from the onset of rains in Southern Sahelian and Sudanian climatic zones of West Africa. Agr Forest Meteorol. 1988; 42(4): 295–305. Publisher Full Text\n\nWilson ML, Chapman LE, Hall DB, et al.: Rift Valley fever in rural northern Senegal: human risk factors and potential vectors. Am J Trop Med Hyg. 1994; 50(6): 663–675. PubMed Abstract | Publisher Full Text\n\nDavies FG, Linthicum KJ, James AD: Rainfall and epizootic Rift Valley fever. Bull World Health Organ. 1985; 63(5): 941–943. PubMed Abstract | Free Full Text\n\nReefhuis J, Gilboa SM, Johansson MA, et al.: Projecting Month of Birth for At-Risk Infants after Zika Virus Disease Outbreaks. Emerg Infect Dis. 2016; 22(5): 828–832. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLessler J, Chaisson LH, Kucirka LM, et al.: Assessing the global threat from Zika virus. Science. 2016; 353(6300): aaf8160. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Q, Sun K, Chinazzi M, et al.: Spread of Zika virus in the Americas. Proc Natl Acad Sci U S A. 2017; 114(22): E4334–E4343. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlanford JI, Blanford S, Crane RG, et al.: Implications of temperature variation for malaria parasite development across Africa. Sci Rep. 2013; 3: 1300. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrard G, Caron M, Mombo IM, et al.: Zika virus in Gabon (Central Africa)--2007: a new threat from Aedes albopictus? PLoS Negl Trop Dis. 2014; 8(2): e2681. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerthet N, Nakouné E, Kamgang B, et al.: Molecular characterization of three Zika flaviviruses obtained from sylvatic mosquitoes in the Central African Republic. Vector Borne Zoonotic Dis. 2014; 14(12): 862–865. PubMed Abstract | Publisher Full Text\n\nWHO: Weekly Bulletin on Outbreaks and Other Emergencies, Week 48: 25 November – 1 December 2017. 2017. Reference Source\n\nWHO: Zika situation report. 20 January 2017. 2017. Reference Source\n\nCoyne CB, Lazear HM: Zika virus - reigniting the TORCH. Nat Rev Microbiol. 2016; 14(11): 707–715. PubMed Abstract | Publisher Full Text\n\nVouga M, Baud D: Imaging of congenital Zika virus infection: the route to identification of prognostic factors. Prenat Diagn. 2016; 36(9): 799–811. PubMed Abstract | Publisher Full Text\n\nEballé AO, Ellong A, Koki G, et al.: Eye malformations in Cameroonian children: a clinical survey. Clin Ophthalmol. 2012; 6: 1607–1611. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCalvet G, Aguiar RS, Melo ASO, et al.: Detection and sequencing of Zika virus from amniotic fluid of fetuses with microcephaly in Brazil: a case study. Lancet Infect Dis. 2016; 16(6): 653–660. PubMed Abstract | Publisher Full Text\n\nBar-Yam Y, Nijhout HF, Parens R, et al.: The Case for Pyriproxyfen as a Potential Cause for Microcephaly; From Biology to Epidemiology. arXiv preprint arXiv: 170303765. 2017. Publisher Full Text"
}
|
[
{
"id": "32204",
"date": "03 Apr 2018",
"name": "Matthew T. Aliota",
"expertise": [
"Reviewer Expertise Virology",
"medical entomology",
"evolution and transmission dynamics of arthropod-borne pathogens"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have produced a paper speculating that African lineage Zika virus has the same capacity to cause congenital Zika syndrome (CZS) as more recent Asian/American outbreak isolates. This was accomplished by analyzing birth outcome reports from one hospital in West Africa and temporally linking outcomes consistent with CZS to seasonal patterns when mosquito populations would be abundant. The paper is well written and certainly addresses an outstanding question in the field. In sum, this article's major purpose is to create a discussion and provide a rationale for broader Zika surveillance activities in Africa, which I I believe has been accomplished, however, there are a few minor details that could be changed to facilitate understanding by the reader.\nI understand the need for security and thus not naming the hospital but would naming the country truly jeopardize security?\n\nWhile I agree that it is possible that African ZIKV has always had the capability of causing CZS and that it is likely underreported or unrecognized, an alternative explanation is that in Africa where ZIKV is endemic girls and women are exposed early in life and subsequent immunity provides protection against CZS during child bearing years. Is there any age distribution that can be associated with the outcomes presented here?\n\nThroughout comparisons are made to Malaria which is transmitted by Anopheles mosquitoes, whereas, ZIKV is likely transmitted by an Aedes species mosquito. Therefore, the same environmental factors may not drive transmission of both equally. Do the authors have any data on, for example, dengue cases that were reported at the same hospital during the study period?\n\nThe data are from a country that is at \"medium risk for a Zika outbreak\". I believe it is more important to know what the estimated seroprevalence of Zika exposure is in this country. This suggests that perhaps Zika is not endemic in the country.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3831",
"date": "19 Jul 2018",
"name": "Helen Piontkivska",
"role": "Author Response",
"response": ">>> 1. I understand the need for security and thus not naming the hospital but would naming the country truly jeopardize security?We agree with you in principle; however, the Committee on Administration of the hospital required us not to name the hospital or the country. We added a statement clarifying this. >>> 2. While I agree that it is possible that African ZIKV has always had the capability of causing CZS and that it is likely underreported or unrecognized, an alternative explanation is that in Africa where ZIKV is endemic girls and women are exposed early in life and subsequent immunity provides protection against CZS during child bearing years. Is there any age distribution that can be associated with the outcomes presented here?We agree with the reviewer that the question of whether early exposure to ZIKV and/or other flaviviruses may offer subsequent long-term protection against CZS during childbearing years warrants further attention. CZS-type defects analyzed in our study were found in women of ages 15-37 years old; the average number of cases per maternal age class is 2.39 (± standard error of 0.42). Thus, these data are not well suited to test this hypothesis, which we hope can be tested in the future with appropriate power. >>> 3. Throughout comparisons are made to Malaria which is transmitted by Anopheles mosquitoes, whereas, ZIKV is likely transmitted by an Aedes species mosquito. Therefore, the same environmental factors may not drive transmission of both equally. Do the authors have any data on, for example, dengue cases that were reported at the same hospital during the study period?There are no data on dengue cases available from the hospital. However, the temperature ranges of the peak transmission of malaria and dengue are overlapping [mid 20's ºC for malaria, per (Beck-Johnson et al. 2013); 23-34ºC for DENV, per (Mordecai et al. 2017)]. Therefore, we expect that the transmission patterns would be similar.Beck-Johnson, L. M., et al. (2013). \"The effect of temperature on Anopheles mosquito population dynamics and the potential for malaria transmission.\" PLoS One 8(11): e79276.Mordecai, E. A., et al. (2017). \"Detecting the impact of temperature on transmission of Zika, dengue, and chikungunya using mechanistic models.\" PLoS Negl Trop Dis 11(4): e0005568.>>> 4. The data are from a country that is at \"medium risk for a Zika outbreak\". I believe it is more important to know what the estimated seroprevalence of Zika exposure is in this country. This suggests that perhaps Zika is not endemic in the country.Unfortunately, there are relatively scant data available from the region, most of which comes from the decades-old studies. Specifically, antibodies to ZIKV were found in human blood samples collected in 1960-ies in Senegal, Burkina Faso, Cote D'Ivoire, Guinea-Bissau, Cameroon, Mali, Niger, Benin, Gabon (Brès 1970), in Senegal (Renaudet et al. 1978) and Nigeria (Adekolu-John and Fagbami 1983) in 1970-ies and 1980-ies, supporting the premise that ZIKV is present across multiple countries in West Africa region (reviewed in (Kindhauser et al. 2016)). We hope that this study will motivate the collection of such data in the future. We added these references to the manuscript, to the \"Can we expect to find cases of congenital Zika syndrome in the African Continent?\" section.Adekolu-John, E. O. and A. H. Fagbami (1983). \"Arthropod-borne virus antibodies in sera of residents of Kainji Lake Basin, Nigeria 1980.\" Trans R Soc Trop Med Hyg 77(2): 149-151.Brès, P. (1970). \"Données récentes apportées par les enquêtes sérologiques sur la prévalence des arbovirus en Afrique, avec référence spéciale à la fièvre jaune [Recent data from serological surveys on the prevalence of arbovirus infections in Africa, with special reference to yellow fever].\" Bulletin of the World Health Organization 43(2): 223-267.Kindhauser, M. K., et al. (2016). \"Zika: the origin and spread of a mosquito-borne virus.\" Bull World Health Organ 171082.Renaudet, J., et al. (1978). \"[A serological survey of arboviruses in the human population of Senegal].\" Bull Soc Pathol Exot Filiales 71(2): 131-140."
}
]
},
{
"id": "35010",
"date": "25 Jun 2018",
"name": "Diana Valencia",
"expertise": [
"Reviewer Expertise Birth defects surveillance in low and middle income countries"
],
"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 the opportunity to review this article. I have the following questions/suggestions for the consideration of the authors:\nThe following important reference is missing from the article, please include: Rasmussen et al. (2016)1\n\nAuthors should mention that Zika migrated not only to Brazil but to the Americas.\n\nWhat are the characteristics of the chosen hospital? Why only one hospital was chosen? Does the hospital has a good reliable birth defect registry?\n\nThere is no description of the birth defects taken into account for the paper; which CZS-type birth defects were chosen form the hospital registry? Were the birth defects confirmed? Was any description available in medical records? Were X-rays available for review?\n\nI agree with the authors on the importance to have birth defects surveillance systems in place. I hope that this article will contribute to the awareness of public health authorities regarding this important issue.\n\nBirth defects surveillance is needed to identify Zika related defects. As the authors mention, the majority of infected pregnant women are asymptomatic.\n\nI think it will be important to differentiate in the article between Zika virus surveillance and birth defects surveillance, and the difficulties of implementing them in Africa.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3830",
"date": "19 Jul 2018",
"name": "Helen Piontkivska",
"role": "Author Response",
"response": ">>> 1. The following important reference is missing from the article, please include: Rasmussen et al. (2016)Thank you for pointing out our oversight. We added this important reference to the paper.>>>> 2. Authors should mention that Zika migrated not only to Brazil but to the Americas.Done. >>>> 3. What are the characteristics of the chosen hospital? Why only one hospital was chosen? Does the hospital has a good reliable birth defect registry?Additional identifiable characteristics were not included in order to protect the identity of the hospital; however, we now added several broad characteristics of the hospital to the manuscript.These were the only data that were available to us, and we are confident in the quality of the birth defect registry. We are excited that we were able to demonstrate this trend even with the relatively small sample available to us. We hope this study will spur further comprehensive analyses of larger datasets.>>> 4. There is no description of the birth defects taken into account for the paper; which CZS-type birth defects were chosen form the hospital registry? Were the birth defects confirmed? Was any description available in medical records? Were X-rays available for review?The description of birth complications that we categorize as a CZS-type (i.e., birth defects consistent with potential congenital Zika syndrome) are listed in the Supplementary table 1, while Supplementary table 2 lists the non-CZS-type complications. We did not have access to individuals' medical records beyond the descriptions listed in Supplementary tables. The hospital does not routinely perform X-rays on newborns with birth defects; thus, no further description of cases or X-rays were available to us. >>> 5. I agree with the authors on the importance to have birth defects surveillance systems in place. I hope that this article will contribute to the awareness of public health authorities regarding this important issue.6. Birth defects surveillance is needed to identify Zika related defects. As the authors mention, the majority of infected pregnant women are asymptomatic.Thank you, we completely agree. >>> 7. I think it will be important to differentiate in the article between Zika virus surveillance and birth defects surveillance, and the difficulties of implementing them in Africa.We agree, and have added statements to that effect. Enhanced birth defect surveillance within hospitals are potentially feasible with the current resources, although they may require additional personnel and/or training. However, the surveillance of birth defects outside of hospital settings would be problematic, though potentially possible. On the other hand, Zika virus surveillance per se, as well as surveillance of other viral diseases, would require substantial financial investments globally, as well as overcoming other challenges related to the implementation of such surveillance."
}
]
},
{
"id": "35015",
"date": "25 Jun 2018",
"name": "Catherine M. 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\nThe authors present an analysis of birth defects by season from a single hospital in an unnamed West African country relative to a purported mosquito season. Their conclusion is that the seasonality of certain types of birth defects, consistent with CZS, could be explain by Zika virus transmission in the area. This is an interesting attempt to use available data to speculate about the impact of Zika virus on congenital defects in an endemic area. Although this is certainly thought-provoking and supports the authors recommendation to improve surveillance for Zika virus in endemic areas, there is insufficient discussion of all of the limitations to this particular analysis which, if included, would cast doubt on their proposed conclusion.\nThe limitation that the data are from a single hospital in an un-named location make it impossible to make any generalizable conclusions and to understand any limitations completely.\nData do not exist on the underlying sero-prevalence of Zika virus infection in this population (I assume). However, given that Zika virus infection likely results in lifelong immunity, one might not expect to see any change in CZS-like defects simply due to a low number of susceptible pregnant women. Were the women with infants with birth defects generally younger and therefore less likely to have been previously infected? And why wouldn't we expect children to have a high rate of infection resulting in immunity as they reach child-bearing age?\nAgain, although difficult to determine given the lack of geographic specificity provided, there is minimal discussion in the paper of other potential causes of microcephaly; most notably rubella and malnutrition. Although rubella-related birth defects might not be expected to show any seasonality, malnutrition might. As might toxoplasmosis.\nThe time period covered by this report encompasses both the Ebola outbreak and significant Yellow fever activity; both of these could cause confounding in the data.\nMosquito species are quite variable in their feeding habits, optimal climactic conditions and in their vectorial capacity. Relying on patterns of malarial disease to represent what is likely to happen with a flavivirus like Zika which is transmitted by different species of mosquitoes should be considered to be a limitation. What about using Yellow fever or dengue disease patterns?\nLastly, while I understand the argument the authors are making, I found the assertion that the maternal infection period of August through October encompassed the peak and latter half of the warm, rainy season confusing. Based on the presented data, August through October had minimal rain and lower than average temperatures. October through December was described as the cool, dry season although temperatures peak for the year in November through January (according to the data presented by the authors).\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3829",
"date": "19 Jul 2018",
"name": "Helen Piontkivska",
"role": "Author Response",
"response": ">>>The limitation that the data are from a single hospital in an un-named location make it impossible to make any generalizable conclusions and to understand any limitations completely. We appreciate these limitations, and they were requirements for us to access the data. Specifically, the Committee on Administration of the hospital required us not to name the hospital or the country. We added a statement clarifying this. Our goal was to spur further discussion and further research from larger sample sizes in multiple locations in order to access the generality of this pattern.>>> Data do not exist on the underlying sero-prevalence of Zika virus infection in this population (I assume). However, given that Zika virus infection likely results in lifelong immunity, one might not expect to see any change in CZS-like defects simply due to a low number of susceptible pregnant women. Were the women with infants with birth defects generally younger and therefore less likely to have been previously infected? And why wouldn't we expect children to have a high rate of infection resulting in immunity as they reach child-bearing age?It is still unclear whether in humans (versus non-human primates) Zika infection confers life-long immunity (Siedner, Ryan and Bogoch 2018). However, age distribution of mothers of infants with CZS-like birth defects indicates that such birth defects appear across multiple age cohorts, from 15 to 37, with the mean number of CZS-like cases per each age group of 2.39 (± standard error of 0.42). Further research of larger cohorts across different locations is needed to definitively test whether there are any maternal-age related effects.>>> Again, although difficult to determine given the lack of geographic specificity provided, there is minimal discussion in the paper of other potential causes of microcephaly; most notably rubella and malnutrition. Although rubella-related birth defects might not be expected to show any seasonality, malnutrition might. As might toxoplasmosis. The time period covered by this report encompasses both the Ebola outbreak and significant Yellow fever activity; both of these could cause confounding in the data.We agree that there are many human health issues that could display seasonality. This underscores the need for widespread comprehensive test for infectious agents and prenatal care, although implementation of such tests in Africa will require significant investments to be feasible.>>> Mosquito species are quite variable in their feeding habits, optimal climactic conditions and in their vectorial capacity. Relying on patterns of malarial disease to represent what is likely to happen with a flavivirus like Zika which is transmitted by different species of mosquitoes should be considered to be a limitation. What about using Yellow fever or dengue disease patterns?In our inferences we did not rely on the malaria pattern to infer seasonality; instead, it was referenced to indicate consistency in the inferred activities from the temperature and rainfall data. Regretfully, no data regarding yellow fever or dengue infections were available. However, the temperature ranges of the peak transmission of malaria and dengue are overlapping [mid 20's ºC for malaria, per (Beck-Johnson et al. 2013); 23-34ºC for DENV, per (Mordecai et al. 2017)]. Therefore, we expect that the transmission patterns would be similar.Beck-Johnson, L. M., et al. (2013). \"The effect of temperature on Anopheles mosquito population dynamics and the potential for malaria transmission.\" PLoS One 8(11): e79276.Mordecai, E. A., et al. (2017). \"Detecting the impact of temperature on transmission of Zika, dengue, and chikungunya using mechanistic models.\" PLoS Negl Trop Dis 11(4): e0005568.>>> Lastly, while I understand the argument the authors are making, I found the assertion that the maternal infection period of August through October encompassed the peak and latter half of the warm, rainy season confusing. Based on the presented data, August through October had minimal rain and lower than average temperatures. October through December was described as the cool, dry season although temperatures peak for the year in November through January (according to the data presented by the authors). We revised the figure 1 legend to make it clear that the bottom axis corresponds with the birth month (depicted with yellow bars) and therefore with the left Y axis that shows % CZS-type birth defects, while the top axis corresponds with the month of maternal infection (i.e., with the rainfall and temperature shown on the left, depicted in green and red, respectively). We have also added color guides to the figure to make it visually clear."
}
]
}
] | 1
|
https://f1000research.com/articles/7-159
|
https://f1000research.com/articles/7-1107/v1
|
18 Jul 18
|
{
"type": "Method Article",
"title": "A convenient protocol for establishing a human cell culture model of the outer retina.",
"authors": [
"Savannah A. Lynn",
"Eloise Keeling",
"Jennifer M. Dewing",
"David A. Johnston",
"Anton Page",
"Angela J. Cree",
"David A. Tumbarello",
"Tracey A. Newman",
"Andrew J. Lotery",
"J. Arjuna Ratnayaka",
"Savannah A. Lynn",
"Eloise Keeling",
"Jennifer M. Dewing",
"David A. Johnston",
"Anton Page",
"Angela J. Cree",
"David A. Tumbarello",
"Tracey A. Newman",
"Andrew J. Lotery"
],
"abstract": "The retinal pigment epithelium (RPE) plays a key role in the pathogenesis of several blinding retinopathies. Alterations to RPE structure and function are reported in Age-related Macular Degeneration, Stargardt and Best disease as well as pattern dystrophies. However, the precise role of RPE cells in disease aetiology remains incompletely understood. Many studies into RPE pathobiology have utilised animal models, which only recapitulate limited disease features. Some studies are also difficult to carry out in animals as the ocular space remains largely inaccessible to powerful microscopes. In contrast, in-vitro models provide an attractive alternative to investigating pathogenic RPE changes associated with age and disease. In this article we describe the step-by-step approach required to establish an experimentally versatile in-vitro culture model of the outer retina incorporating the RPE monolayer and supportive Bruch’s membrane (BrM). We show that confluent monolayers of the spontaneously arisen human ARPE-19 cell-line cultured under optimal conditions reproduce key features of native RPE. These models can be used to study dynamic, intracellular and extracellular pathogenic changes using the latest developments in microscopy and imaging technology. We also discuss how RPE cells from human foetal and stem-cell derived sources can be incorporated alongside sophisticated BrM substitutes to replicate the aged/diseased outer retina in a dish. The work presented here will enable users to rapidly establish a realistic in-vitro model of the outer retina that is amenable to a high degree of experimental manipulation which will also serve as an attractive alternative to using animals. This in-vitro model therefore has the benefit of achieving the 3Rs objective of reducing and replacing the use of animals in research. As well as recapitulating salient structural and physiological features of native RPE, other advantages of this model include its simplicity, rapid set-up time and unlimited scope for detailed single-cell resolution and matrix studies.",
"keywords": [
"Retinal Pigment Epithelium (RPE)",
"In-vitro",
"Cell culture",
"Disease modelling",
"Retinopathy"
],
"content": "\n\nWe provide a step-by-step protocol to rapidly establish an in-vitro model of the outer retina incorporating the Retinal Pigment Epithelium (RPE) and the supportive Bruch’s membrane.\n\nWe discuss the advantages and limitations of RPE cells (the ARPE-19 cell-line) used in this work.\n\nThis in-vitro model allows the use of powerful confocal microscopes (fast, high-resolution imaging) and new platforms such as 3View and Lightsheet.\n\nAllows a high degree of experimental manipulation.\n\nThis in-vitro culture model can be used as an alternative to in-vivo experiments in spontaneously arising, acutely-induced or transgenic mouse models of retinal degeneration, or be used in parallel with animal studies.\n\nThis model enables users to obtain functional RPE monolayers with desirable physiological and structural features of the native RPE tissue after only 2–4 months in culture.\n\nSuch in-vitro RPE monolayers can therefore be used to model disease features which do not manifest in some mouse models for as long as 18 months.\n\nThis in-vitro culture model has a relatively fast set-up period enabling studies after 2–4 months.\n\nThe well-characterised ARPE-19 cell-line used in this work facilitates reproducibility and comparisons with a large body of published literature.\n\nCost effective compared to carrying out similar studies in-vivo.\n\nAllows first-line investigations of specific disease pathways in-vitro, hence suited to high-throughput drug discovery screens.\n\nThis set-up can be used to culture/model primary RPE cells (from porcine, rodent and humans sources), cell-lines such as ARPE-19 as well as human foetal and stem-cell derived RPE from patients.\n\nSuitable for single-cell level studies and those investigating dynamic changes to the RPE monolayer or the underlying extracellular matrix.\n\nUsed by researchers to study RPE changes in different types of blinding diseases in which the retina irreversibly degenerates including Age-related Macular Degeneration (AMD), Sorsby Fundus Dystrophy and Retinitis Pigmentosa.\n\nFurther refinements to this model can be made by incorporating stem cell-derived RPE directly from patients, synthetic membranes to better mimic the underlying Bruch’s membrane and microfluidic devices to simulate the choroid.\n\nSets the standard to recapitulate structural and functional features for future in-vitro 3D retinal models.\n\n\nIntroduction\n\nThe retinal pigment epithelium (RPE) consists of a monolayer of largely cuboidal-shaped pigmented cells found beneath the neuroretina and overlying the vascular blood supply of the choriocapillaris. Occupying this strategic position in the outer retina the RPE performs multiple functions which are essential for retinal homeostasis and maintenance of life-long vision. This includes the daily phagocytosis of shed Photoreceptor Outer Segments (POS), re-isomerization of all-trans-retinal to 11-cis-retinal in the visual cycle, protection against effects of photo-oxidation, trans-epithelial transport as well as the polarised secretion of molecules towards the overlying neuroretina and the underlying choroid. The RPE also forms part of the outer blood-retinal barrier (BRB) which functions to confer an immune privileged state within the ocular environment1. Dysfunction or abnormalities of the RPE monolayer is correlated with early stages of pathology linked to a range of ocular conditions such as Age-related Macular Degeneration (AMD), Sorsbys fundus dystrophy, Stargardt disease and Best disease, diabetic retinopathy as well as pattern dystrophies1–3. However, the origins of RPE dysfunction and how they contribute to such diverse ocular conditions remains incompletely understood.\n\nNumerous in-vivo models including non-human primates, pigs, sheep, rabbits and rodents have been used to study retinal pathobiology4. Of these, the most widely used are mice, which show regional differences in Bruch’s membrane (BrM) thickness and photoreceptor density, a similar rod to cone ratio at locations comparable to the peripheral human macula as well as a similar RPE monolayer to humans5. Mice also offer advantages in terms of costs compared to the use of larger animals and the possibility of studying salient disease features in a matter of months. This has led to the use of spontaneously arising6, acutely-induced7 and transgenic mouse models8, or indeed combined models where genetics and diet has been manipulated and mice aged for long periods to bring about disease features9,10. However, given the lack of anatomical specialisation equivalent to humans, rodent models are of limited value for studies into macular conditions. Moreover, no single mouse model is capable of replicating the full disease spectrum observed in human retinopathies. This has often led to the unnecessary and over-use of poorly characterised rodent models, many of which show only limited disease features and/or have to be aged for long periods before any obvious retinal pathology is detected4,11. The arrangement of ocular tissues such as the RPE also makes them difficult to image, particularly for studies requiring dynamic, real-time imaging or data at single-cell resolution. Welfare concerns and severity limits of mouse models, other than basic information on animal husbandry, are also poorly reported in the literature. For instance, there is limited data on how a particular genetic alteration or mice maintained over long periods (>18 month) might affect their behaviour and quality of life. In contrast, in-vitro models, although simplistic by comparison, are not limited by these issues and boast distinct advantages over mouse models for delineating cellular pathways of damage, or for drug screens to identify effects on a given cell type. Cells cultured under in-vitro conditions that recapitulate their in-situ environment have been shown to reproduce a phenotype that closely resemble native tissues. These cells not only adopt native-like structural and physiological characteristics, but also a genetic profile closely matching their in-situ counterpart. RPE cells were initially grown on plastic substrates and did not exhibit a fully differentiated phenotype. Investigators therefore started culturing RPE monolayers on commercially-sourced transwell inserts with varying pore sizes which mimics important features of the underlying BrM2,12. The culture of RPE cells on 0.4μm pore-size inserts is now widely regarded to produce the most desirable RPE phenotype13. The presence of a porous underlying substrate allows the RPE to undertake activities such as matrix deposition14 and directional secretion of molecules15 which are key features of these cells. Transwell inserts are now widely used to culture primary porcine16, murine17,18, human foetal RPE (hfRPE)19–23 and adult human RPE24 as well as numerous cell-lines including ARPE-19 cells18,19,25,26. Studies have shown that RPE cultured under such conditions display structural and functional characteristics of native RPE cells, albeit to differing extents16–19,21,24,25. New developments are also incorporated into transwell systems. For instance, recent advances in stem-cell technology, which allows the generation of pluripotent stem-cell derived RPE (PSC-RPE) directly from patients27,28, are routinely modelled in transwells. This approach has resulted in what many consider to be the current gold-standard in RPE modelling. However, several unresolved issues remain as respective labs use different protocols and appear to differentiate cells to different extents before studies are undertaken. Human PSCs are also limited by effects of senescence and variability between clones. These may cause future difficulties with reproducibility29. RPE cell-lines by contrast, which some may consider to be the poor cousin of PSC-RPE, continue to offer some advantages. The rat immortalised RPE-J cell-line30 and the spontaneously arising human ARPE-19 cell-line26 are two noteworthy examples that have been extensively used in transwell systems. ARPE-19 cells have certain advantages as the RPE-J cells will only proliferate if maintained at 32°C and require retinoic acid for contact inhibition. ARPE-19 cells are also highly characterised, well understood by researchers and widely used for over 2 decades to gain key insights into RPE pathobiology since first described by Dunn and colleagues in 199626. The consistency of ARPE-19 cells sourced from commercial suppliers has also contributed to generating comparable and reproducible data across different labs that is as yet unmatched in the field. The culture of ARPE-19 was further optimised in recent years25 such that they exhibit a normal karyotype31, apical-basolateral specialisation18 as well as pigmentation and express characteristic proteins18,25 including components of the BRB18,25,32–34, polarised secretion18,19,25 and phagocytosis25,26,35. In fact, new evidence show that when cultured for 4 months in optimised medium, ARPE-19 cells exhibit a comparable transcriptome to the native RPE36. Further refinement in modelling RPE cells on transwell inserts can be anticipated due to advances in new artificial BrM substitutes37,38.\n\nA literature search using PubMed Central in May 2018 using the terms ‘Retinal Pigment Epithelium’ and ‘Transwell’, reported a 193% increase of in-vitro studies within the past 5 years compared to the previous 5-year period. In contrast, search terms ‘Retinal Pigment Epithelium’ and ‘in-vivo’ revealed a 16% reduction in the number of reports using in-vivo models encompassing rodents, rabbits, porcine, bovine and non-human primates over a similar period. Based on these findings and an average annual increase of 4.3% in citations with RPE studies, we estimate that at least 45 publications reporting in-vivo work will be replaced by in-vitro RPE modelling studies. Given the rapid rate at which RPE modelling work is progressing, this will yield at the very least 188 annual citations by 2023. In this article we provide a detailed step-by-step approach for optimising an in-vitro RPE cell model using the spontaneously arisen ARPE-19 cell-line. We also provide steps used to characterise this model, discuss its advantages and limitations as well as how it fulfils the 3Rs objectives.\n\n\nMethods\n\nCulture of ARPE-19 cells. ARPE-19 cells26 were obtained from the American Tissue Culture Collection (CRL-2302, ATCC, USA) and maintained in a 37°C humidified incubator with 5% CO2 atmosphere and 95% air. Cells were cultured in an optimised medium comprising Dulbecco’s modified Eagle’s Medium (DMEM) with 4.5 g/l L-D glucose (high glucose), L-glutamine and pyruvate (41966–029, Life Technologies, UK) supplemented with 1% heat inactivated foetal calf serum (N4762, Sigma Aldrich, UK) and 1% of the penicillin streptomycin stock solution (10,000 units/ml penicillin, 10mg/ml streptomycin in 0.85% saline; P4333, Sigma Aldrich, UK)25. Cells cultured in T25cm2 flasks were maintained in a 5ml volume with a complete media change performed every 2–3 days. Cells cultured on Corning® 12mm, 0.4μm pore, PET Transwell® Permeable Supports (CLS3460, Sigma Aldrich, UK) were maintained in 0.5ml and 2ml of volume of media in apical and basal chambers. Cells grown on Corning® 24mm, 0.4μm pore, PET Transwell® Permeable Supports (CLS3450, Sigma Aldrich, UK) were maintained in 1.5ml and 3ml volume of media in apical and basal chambers. A complete media change in the apical chamber and a 20% (v) change in the basal compartment was performed every 2–3 days. Cells were used between passages 23–27.\n\nFibronectin coating of transwell inserts. Lyophilised fibronectin (F2006, Sigma Aldrich, UK) was prepared to a final concentration of 50μg/ml in double distilled water (ddH20), and applied to the apical surface of transwell inserts. A volume of 0.25ml or 0.6ml was used for 12mm and 24mm inserts respectively. Transwells were partially covered in a laminar flow hood and allowed to dry overnight, after which any residual fibronectin was aspirated and inserts washed in 1x phosphate buffered saline (PBS, Sigma Aldrich, UK).\n\nPassage of ARPE-19 cells. ARPE-19 cells were grown in T25cm2 flasks for up to 3 weeks prior to passaging at a 1:3 ratio. This was achieved by washing cells in Ca2+ and Mg2+ free Hank’s Balanced Salt Solution (HBSS; 14065049, Life Technologies, UK) and incubation with 1.5ml 0.25% Trypsin/EDTA (25200056, Life Technologies, UK) for 6 minutes, followed by neutralisation/trituration with 7ml volume of complete medium. The cell suspension was centrifuged at 125g for 5 minutes after which the pellet was suspended in fresh medium. Cells were seeded on fibronectin coated/uncoated 0.4μm PET transwell inserts (Corning, UK) at a density of 1.25×104 and 5×104 for 12mm or 24mm inserts, and left undisturbed for 4 days to facilitate cell adhesion prior to media change. The culture media used in apical and basal transwell compartments were identical. Long-term cultures were maintained for 2–4 months before assessing characteristic structural RPE features or expression of cell-specific and barrier proteins. Functional assays were carried out in 4 month old cultures.\n\nConfocal Immunofluorescence microscopy. ARPE-19 cultures were rinsed in 1xHBSS, fixed in ice-cold PBS containing 4% formaldehyde for 30 minutes at 4°C and washed three times in 1x PBS. Blocking and permeabilisation were achieved by incubation with 5% normal goat serum (NGS; G9023, Sigma Aldrich, UK) in 1x Phosphate Buffered Saline with Triton X-100 (PBST, Sigma Aldrich, UK) for one hour prior to the addition of primary antibody (prepared in blocking buffer) which was incubated overnight at 4°C. The following primary antibodies were used: ZO-1 (1:100, RRID: AB_2533456, Thermo Fisher Scientific, UK), Occludin (1:100, RRID: AB_2533977, Thermo Fisher Scientific, UK), RPE65 (1:100, RRID: AB_2181006, Abcam, UK) and alpha 1 Na+/K+ ATPase (1:100, RRID:AB_306023, Abcam, UK). The following day, cells were washed in 0.05% PBST and incubated with the appropriate Alexa Fluor® labelled secondary antibody (RRID: AB_2534115, RRID: AB_2534085, RRID: AB_2534116, RRID: AB_2534087, RRID: AB_2534114, RRID: AB_2534064, Life technologies, UK) at a dilution of 1:200 prepared in 0.05% PBST for 1 hour. This was followed by three washes in 1× PBS and one wash with ddH20 after which cells were incubated with 1μg/ml 4’, 6’-diamino-2-phenylindole (DAPI; D9542, Sigma Aldrich, UK) for 10 minutes. Inserts were washed an additional three times in ddH2O and mounted between two glass coverslips with Mowiol® mounting medium (Harco Chemical Company Ltd., UK) prior to imaging with a Leica SP5 or SP8 laser-scanning confocal microscope (Leica Microsystems, UK).\n\nTransmission electron microscopy. Transwell inserts were processed for transmission electron microscopy (TEM) by first washing in 1X HBSS, followed by immersion in primary fixative (3% glutaraldehyde and 4% formaldehyde in 0.1M PIPES [Agar Scientific, UK]; pH 7.2) for 1 hour, after which they were rinsed twice for 10 minutes in 0.1M PIPES before post fixation for 1 hour with 1% osmium tetroxide in 0.1M PIPES. Inserts were rinsed twice in 0.1M PIPES and once in ddH2O, after which they were stained in 2% (aqueous) uranyl acetate for 20 minutes. Samples were subsequently dehydrated by passing the inserts through a series of ethanol gradients (30%, 50%, 70% and 95%) for 10 minutes each, followed by two 20 minute incubations in absolute ethanol (100%). A link reagent acetonitrile (Agar Scientific, UK) was then applied to inserts for 10 minutes, after which samples were incubated overnight with at a 1:1 ratio of acetonitrile to Spurr resin (Agar Scientific, UK). The following day, cells were incubated with Spurr resin for 6 hours and embedded and polymerised in fresh resin at 60°C for 24 hours. Ultrathin sections were cut using a Reichert Ultracut E (Leica Microsystems, UK), collected on 200 mesh carbon and formvar coated copper grids and stained with Reynolds Lead Stain (Agar Scientific, UK). Cross sections of cultures on transwell inserts were visualised using a Hitachi 7000 transmission electron microscope (Hitachi, Germany) fitted with a SIS Megaview III camera (EMSIS, Germany).\n\nTrans-epithelial Electrical Resistance measurements. Trans-epithelial Electrical Resistance (TEER) were carried out over a 3 month period using an EVOM2 epithelial voltohmmeter and 4mm STX2 chopstick electrode (EVOM2; 300523, World Precision Instruments Inc., USA). Briefly, the electrode was sterilised in 70% ethanol, rinsed in ddH2O and equilibrated in pre-warmed culture medium, before being simultaneously introduced into the apical and basal transwell compartments. Measurements were recorded from at least three separate wells per experiment. In each case, five measurements were recorded per well to obtain an average value. The reference value from a fibronectin-coated insert and devoid of cells was subtracted from the average value to yield a net TEER measurement. This was subsequently corrected for the growth area using the following formula (Equation 1). Measurements were performed at room temperature within 6 minutes of removing cultures from the incubator. A full media change was also performed after weekly measurements to minimise the risk of contamination.\n\nFinal TEER (Ω/cm2) = Net TEER (Ω) × Area of Transwell insert (cm2) Equation 1\n\nELISA studies. Secreted levels of human vascular endothelial growth factor (VEGF: isoforms 165, 121) and pigment epithelium derived factor (PEDF) in apical and basal transwell compartments were quantified using the Novex® human VEGF (KHG0111, Life Technologies, UK) and human PEDF (RD191114200R, BioVendor, Germany) solid-phase sandwich ELISA. Conditioned media was collected from 2 month old ARPE-19 cultures (n=3) and diluted 1:1 and 1:9 prior to VEGF and PEDF ELISAs, respectively. Experiments were carried out in triplicate and followed the manufacturers’ instructions. Optical densities were determined by measuring the absorbance at 450nm using a micro-titre plate reader (FLUOstar Optima; BMG LABTECH, UK) and accounting for the 570nm wavelength correction.\n\nPreparation of POS-FITC. Porcine eyes (maximum of 2 days post mortem) were sourced from a butcher. An incision was made proximate to the ora serata after which the anterior ocular portion was removed and retinae detached gently from the underlying RPE. These were subsequently pooled in KCl buffer (0.3M KCl, 10mM HEPES, 0.5mM CaCl2, 1mM MgCl2; pH 7.0) with 48% sucrose (w/v), agitated for 2 minutes and centrifuged at 5000g for 5 minutes to facilitate POS detachment. The resultant supernatant containing isolated POS was filtered through a sterile gauze into 1.5ml Eppendorf tube containing an equal volume of KCl buffer without sucrose. POS were pelleted by centrifugation at 4,000g for 7 minutes, washed three times in 1× PBS and re-suspended in DMEM with 2.5% sucrose (w/v). POS were covalently tagged to Fluorescein isothiocyanate (FITC) by incubation with 5ml labelling buffer (20Mm phosphate buffer pH 7.2, 5mM taurine with 10% w/v sucrose) and 1.5ml FITC stock solution (2mg/mL FITC isomer I in 0.1M Na2CO3 buffer; pH 9.5) at room temperature for 1 hour in the dark on a Stuart SB2 Rotator (Camlam Ltd, UK). POS-FITC conjugates were pelleted by centrifugation at 3,000g for 5 minutes and re-suspended in DMEM with 2.5% sucrose (w/v). Isolated POS can be stored for up to 6 months at -80°C. The total protein content in preparations was quantified using a BCA assay (23225, Thermo Fisher, UK) following the manufacturer’s instructions.\n\nPOS feeding assay and assessment of trafficking dynamics. ARPE-19 monolayers on transwell inserts were incubated at 17°C for 30 minutes prior to exposure with 4mg/cm2 POS-FITC for a further 30 minutes at 17°C. This facilitates maximal POS binding with minimal internalisation39 to initiate a pulse-chase assay. The POS-FITC solution was aspirated to remove unbound POS. Cultures were supplemented with fresh media and returned to a humidified 37°C incubator with 5% CO2 and 95% air. Cells were subsequently fixed at 2, 4, 6, 12, 24 and 48 hours with 1×PBS containing 4% formaldehyde for 30 minutes at 4°C, after which they were incubated with 1% BSA in PBS-Tween to block/permeabilise cells for 30 minutes. Cultures were probed overnight at 4°C with the following primary antibodies prepared in blocking buffer. Rab 5 (1:200, RRID: AB_470264, Abcam, UK), Rab 7 (1:200, RRID: AB_2629474, Abcam, UK), LAMP1 (1:1000, RRID: AB_775978, Abcam, UK), LAMP2A (1:1000, RRID: AB_775981, Abcam, UK) and LC3B (1:200, RRID: AB_881433, Abcam, UK). Excess antibodies were washed three times in 1xPBS, following which cultures were incubated with the appropriate Alexa Fluor® secondary antibody (RRID: AB_2534115, RRID: AB_2534085, RRID: AB_2534116, RRID: AB_2534087, RRID: AB_2534114, RRID: AB_2534064, Life technologies, UK) for one hour at room temperature. Inserts were finally washed three times in 1×PBS, once in ddH2O and counterstained with 1μg/ml DAPI for 10 minutes. Samples were mounted between two glass coverslips with Mowiol® mounting medium and imaged using an SP8 laser-scanning confocal microscope (Leica Microsystems, UK). Quantification of POS-FITC co-localisation with various endocytic/phagocytic, lysosomal and autophagy compartments (n=10 cells/compartment/time point) was performed using Volocity software, version 6.1.1 (Perkin Elmer, UK), which employs the Costes et al. automated statistical algorithm40. Co-localisation values were plotted for each compartment as a function of time.\n\nStatistical Analysis. Statistical analyses were conducted using the GraphPad Prism 7 Software (GraphPad, US). Values were first assessed to ensure data met assumptions of the selected statistical test. Tests for each experiment appear in figure legends. Briefly, ELISA quantification was assessed using the unpaired student’s t-test, whilst TEER were evaluated using a one way ANOVA and Tukey’s multiple comparisons tests. In both cases, a single well corresponded to an experimental unit. Data is presented as means ± standard error of the mean (SEM) where n represents independent experiments. Statistical significance is denoted as * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 and **** p ≤ 0.0001.\n\n\nProtocol\n\nHere we describe the step-by-step procedure required for establishing and validating long-term cultures of ARPE-19 monolayers on transwell inserts. A schematic highlighting the sequence of steps and timelines are summarised in Figure 1.\n\nKey points in the establishing and validating RPE cultures on transwells. The steps and time periods required for this procedure are indicated alongside each outcome measure.\n\nStep 1: Preparation of cell culture media\n\nARPE-19 cells (CRL-2302™ ATCC®, USA) require growth in optimised culture media, which can be prepared according to Table 1. Freshly prepared cell culture media is passed through a vacuum filter to maximise sterility and used within 2 weeks. A volume of 250ml is sufficient for the culture of 6 transwell plates of 12mm diameter inserts or 4 plates of 24mm diameter inserts (Table 4) for a period of approximately 2 weeks. Volumes should be scaled according to the size and number of desired transwells as storage of media for longer periods is not recommended.\n\nAbbreviations: Dulbecco’s Modified Eagle Medium (DMEM)\n\nStep 2: Culture of ARPE-19 cells\n\nARPE-19 cells should be maintained in a humidified incubator set at 37°C with an atmosphere of 5% CO2. Cells should be cultured in a T25cm2 flask containing 5ml of freshly prepared media and passaged at a 1:3 ratio when confluent or approximately every 3 weeks. A complete media change should be performed every 2–3 days to retain physiological glucose levels41. Although it is possible to maintain ARPE-19 cells in T25cm2 flasks for longer periods, cells becoming increasingly difficult to passage following the formation of an underlying extracellular matrix. Passaging involves removal of conditioned medium and washing cells with 1xHBSS followed by exposure to 1.5ml 0.25% trypsin/EDTA for 6 minutes in an incubator. The trypsin/EDTA solution is neutralised using 7ml of complete media and triturated to obtain single cells after which the resulting suspension is centrifuged at 300g for 5 minutes. The pellet is re-suspended in a volume of freshly prepared culture media and split at a 1:3 ratio between T25cm2 flasks. The importance of correct cell passaging to maintaining an epithelial phenotype is discussed elsewhere42. For the long term storage in liquid nitrogen (-195.8°C) cells should be suspended in the desired concentration in freezing medium comprising 75% complete culture medium and 25% dimethylsulfoxide (DMSO; S-002-M, Sigma Aldrich, UK). For ARPE-19 cells we recommend 1ml aliquots of 1x106 cells. These must first be frozen in a suitable container (Mr Frosty™, Thermo Fisher Scientific UK) or equivalent with 100% isopropanol at -80°C overnight prior to storage in liquid nitrogen.\n\nStep 3: Coating transwell inserts with fibronectin\n\nPrepare lyophilised fibronectin (F2006, Sigma Aldrich, UK) to a final concentration of 50μg/ml by adding 20ml of sterile ddH2O. We recommend preparing an initial 5ml solution using sterile ddH2O followed by transfer to a 50ml falcon containing 15ml sterile ddH2O. The fibronectin solution should be used immediately and is sufficient to coat six 12mm diameter or five 24mm diameter transwell plates, or should be divided into aliquots for storage at -20°C. Users should apply the stock solution to the apical transwell compartment as indicated in Table 2. Ensure that the entire surface of the membrane is covered after which transwells are left partially covered in a laminar flow hood overnight. The following day any residual fibronectin should be aspirated and wells washed with 1x sterile PBS for cell seeding (Table 3). We observed that ARPE-19 cells readily proliferate and mature to form in-situ RPE-like monolayers on an underlying fibronectin matrix although others had used a laminin substrate26. Consequently, work presented here are carried out on transwells coated with fibronectin. It is also possible to culture cells in the absence of an underlying coating on transwell membranes (Figure 2). ARPE-19 cultured without an extracellular matrix (ECM) substrate develop pigmentation, establish a trans-epithelial barrier and secrete proteins directionally25. The culture of cells without an underlying coating is particularly useful for studies in which de-novo synthesis/deposition of extra cellular components and their turnover can be monitored without any influence of artificial substrates.\n\nAbbreviations: Polyethylene terephthalate (PET)\n\nSeeding densities used to start ARPE-19 cultures on transwell inserts of different diameters.\n\nAbbreviations: Polyethylene terephthalate (PET)\n\nMedia changed every 2–3 days.\n\nAbbreviations: Polyethylene terephthalate (PET)\n\nWe tested effects of an extracellular matrix (ECM) such as fibronectin on the ability of cells to form a monolayer on transwell membranes. Inserts coated with fibronectin [A,C,E] or those without any coating [B,D,F] were imaged over several days after seeding. No differences were observed in cell attachment or growth. Contact inhibition proceeded cell differentiation after approximately 10 days in culture. Scale bars correspond to 200μm.\n\nTechnical tip: Rapid thawing of soluble fibronectin can result in the irreversible precipitation of proteins. To avoid this we suggest that the stock solution should be gradually thawed at 4°C.\n\nStep 4: Seeding and culture of ARPE-19 cells on transwell inserts\n\nCells are passaged as described in step 2. A confluent flask of T25cm2 ARPE-19 cells yield between 3–5 million cells. Consequently, one T25cm2 flask is sufficient to seed 10 plates of 24mm diameter transwell inserts or 20 plates of 12mm diameter inserts. Cells are seeded on fibronectin coated transwells (Table 3), although the benefit of using uncoated transwell membranes have also been discussed. Culture media are applied to the apical and basal transwell compartment at least one hour prior to cell seeding (Table 4). Following seeding, we recommend leaving cultures undisturbed for approximately 4 days prior to the first media change. Cultures are maintained in a 37°C incubator with an atmosphere of 5% CO2 and media changed every 2–3 days (Table 4). After approximately 1 week cultures appear confluent and by 2 weeks exhibit characteristic cobblestone morphology. Obvious signs of pigmentation will develop after 3–4 months, although some evidence of pigmentation is apparent under light microscopy after 2 months. The size of pores in transwell membranes are known to influence RPE morphology13 hence we suggest users adhere to these recommendations. Cultures are maintained for a minimum of 2 months prior to validation studies. For functional experiments, such as POS feeding assays, we recommend maintaining cultures for approximately 4 months.\n\nTechnical tip: Pipette solutions along plastic walls of the transwell chamber at a steady state during media changes to avoid inducing cell stress or cell detachment. When removing media we recommend first tilting the transwell plate to a 45° angle to avoid disturbing the RPE monolayer.\n\nIn this section we describe the steps used to validate and characterise ARPE-19 monolayers on transwell inserts. These approaches however may be adopted for the culture of RPE from different sources, although the time taken to obtain monolayers displaying physiological and structural features akin to the native RPE may vary depending on the specific type/source of RPE cells.\n\nStep 5: Confocal immunofluorescence studies of ARPE-19 monolayers\n\nCells that have been in culture for at least 2 month are used to ensure RPE monolayers had adopted structural and physiological features of native RPE. Transwell inserts are washed with 1x sterile PBS prior to fixation in 4% PFA for 30 minutes. Each transwell membrane is removed from its insert by running a blade along the circumferential ring (Figure 3). The amount of material (RPE monolayers) required to carry out experiments may be maximised by sectioning transwell membranes into multiple sections, although caution must be exercised to prevent disturbing the delicate cell layer. We recommend using a sharp razor blade to guillotine sections of the membrane outright as cutting or slicing generates sheer forces which rucks membranes leading to cell detachment (Figure 3). Transwell membranes are washed three times in 1x PBS and blocked/permeabilised in 5% NGS in 0.1% PBST for 1 hour. A battery of antibodies are used to probe for components of the BRB (ZO-1 and Occludin), to assess RPE polarisation (Na+/K+ ATPase) and to detect expression of the cell-specific marker (RPE65), although other proteins such as CRALBP may also be included. Readers are also referred to studies described in Ahmado et al., 201125. Primary Antibodies are diluted 1:100 in blocking buffer for incubation at 4°C overnight (Table 5). The following day membranes are washed three times in 0.05% PBST followed by incubation with the appropriate Alexa Fluor® labelled secondary antibody (RRID: AB_2534115, RRID: AB_2534085, RRID: AB_2534116, RRID: AB_2534087, RRID: AB_2534114, RRID: AB_2534064, Life Technologies, UK) prepared in 0.05% PBST for 1 hour. Membranes are washed three times in ddH2O to remove any unbound secondary antibody and mounted between two glass coverslips with Mowiol® mounting medium (Harco Chemical Company Ltd., UK). We recommend sandwiching membranes between 2 glass coverslips as opposed to pairing a single coverslip with a thicker glass slide as this allows either side of the sample to be imaged without potential optical interference from the porous membrane. Z-stack image are captured with a laser-scanning confocal microscope (Supplementary Figure S1 and Figure 4). We recommend a minimum optical slice thickness of 1μm through z-stacks to help assess the polarised expression of RPE markers.\n\nTranswell membranes can be detached by making incisions with a scalpel following the circumference of the plastic chamber. The insert lip may be used as an indicator of the boundary. Once removed, the detached membrane can be guillotined using a sharp razorblade. We advise against slicing or cutting using sheer forces as this is likely to damage or dislodge the Retinal Pigment Epithelial monolayer.\n\nCultures were probed for [A–D] the early tight junctional marker Zonula Occludens-1, which showed cobblestone morphology characteristic of Retinal Pigment Epithelial (RPE) cells. [E–H] We also observed expression of the mid-late barrier protein occludin, although their staining was somewhat weaker. [I–L] The epithelial transporter Na+/K+ ATPase was observed in APPE-19 monolayers (arrows) but was not evident in all cells. Staining was however observed predominantly on the apical RPE surface, a feature reported in highly differentiated RPE cells. [M–P] The cell-specific marker RPE-specific 65 kDa protein (RPE65) was also observed after 2 months in culture. Nuclei were counterstained with DAPI (blue). [A–C, E–G, I–K, M–O] show representative en-face confocal images whilst [D,H,L,P] show corresponding z-plane reconstructions. Scale bars correspond to 20μm.\n\nTechnical tip: Pores within the membrane may be used as a reference point to help orient the position of apical and basolateral RPE surfaces. However, users may have to account for small undulations in the membrane which will alter the focal plane across the sample.\n\nStep 6: Transmission electron microscopy studies of ARPE-19 cultures\n\nTransmission Electron Microscopy (TEM) studies are performed on monolayers cultured for at least 2 months to ensure structural specialisation of the apical and basolateral RPE surfaces. Inserts are washed in 1x HBSS immediately following removal of conditioned media to prevent dehydration and fixed in primary fixative (3% glutaraldehyde and 4% formaldehyde in 0.1M PIPES; pH 7.2) for one hour. Samples are subsequently washed twice in 0.1M PIPES and post fixed in 1% buffered osmium tetroxide (Oxkem, UK) prepared in 0.1M PIPES for 1 hour. Following fixation, samples are rinsed twice in 0.1M PIPES, once in ddH20 and stained with 2% (aqueous) uranyl acetate (Agar Scientific, UK) for 20 minutes. At this stage membranes may be submerged in a 30% ethanol solution for removal from transwells. It is vital that this process is carried under an appropriate liquid (buffer, ethanol) to prevent the cells drying out which would render them useless for microscopy. Samples are passed successively through a graded series of ethanol concentrations (30%, 50%, 70% and 95% ethanol) for 10 minutes each, followed by two successive incubation periods in absolute ethanol for 20 minutes to achieve optimal dehydration. The link reagent acetonitrile (Fisher Scientific, UK) is applied for 10 minutes and membranes incubated in a mixture containing an equal ratio of acetonitrile to Spurr resin overnight. In our experience, Spurr resin appears to be the best medium to effectively bond filters, although sections often split along the resin/filter interface during sectioning and when viewing under the microscope. The following day, samples are incubated for an additional 6 hours in Spurr resin and embed in fresh Spurr resin for polymerisation at 60°C for 24 hours. Samples should be embedded in Spurr resin as triangular slices with the apex of the triangle positioned towards the bottom of the embedding capsule (Figure 5), which facilitates ease of cutting. This procedure is carried out without a rotator as cells could otherwise detach from the underlying membrane. Ultrathin/silver TEM sections are prepared using a Reichert Ultracut E ultramicrotome and collected on 200 mesh carbon and formvar coated copper palladium grids. We advise against the use of chloroform to stretch sections which exacerbates potential separation of the resin/membrane interface during sectioning. Sections are subsequently stained with Reynold’s lead stain and visualised using a Hitachi 7000 transmission electron microscope fitted with a SIS Megaview III plate EMSIS camera (Figure 6).\n\nSchematic outlining steps carried out to embed segmented transwell membranes into capsules containing fresh Spurr resin. The apex is positioned downwards, which greatly assists with cutting sections for TEM.\n\nCultures were assessed by transmission electron microscopy to determine the extent of Retinal Pigment Epithelium (RPE) structural specialisation on transwell inserts. [A] Cross-section of the RPE monolayer showing microvilli (Mv) on apical surface (arrows). Pores within the polyethylene terephthalate (PET) membrane are also visible. Scale bar corresponds to 5μm. [B] Basolateral infolds (Bi) in RPE cells adjacent to the transwell membrane (arrows) are evident, under which we have previously observed the accumulation of sub-RPE deposits in long-term culture18. Scale bar corresponds to 500nm. [C–E] Intracellular organelles including mitochondria (Mt), vesicular compartments such as phagosomes/endosomes and lysosomes, rough endoplasmic reticulum (RER), Golgi and pigment molecules (P) were evident, indicating apical-basolateral specialisation recapitulating arrangement of in-situ RPE. Scale bars correspond to 200nm in [C,D] and 1μm in [D]. [F–G] We also observed the presence of tight Junctions (Tj) and adherens Junctions (Aj) in apical borders of RPE cells (arrows). Scale bars correspond to 200nm. TEM micrograph in panel B was published previously18 under the Creative Commons licence.\n\nTechnical tip: We recommend viewing samples starting at a lower magnification with the electron beam spread widely in order to minimise shrinkage or movement across sections, and to protect against the possibility of splitting at the resin/membrane interface.\n\nStep 7: Trans-epithelial electrical resistance measurement of ARPE-19 cultures\n\nTEER studies are carried out after a minimum of 6 weeks in culture as ARPE-19 cells do not form an effective barrier before this time (Figure 7A–B). ARPE-19 cells also generate relatively poor barriers compared to hfRPE or PSC-RPE. However, the method describe herein can be adopted to test barriers created by RPE cells from different sources. Electrical recordings are obtained using an EVOM2 epithelial voltohmmeter and a 4mm STX2 chopstick electrode (EVOM2; 300523, World Precision Instruments Inc., USA). As importance is given to maintaining sterility in RPE cultured for long periods, electrodes are first sterilised in 70% ethanol, rinsed in ddH2O and equilibrated in pre-warmed culture medium prior to use. For this reason we also recommend performing a complete media change in both transwell compartments after measurements. Electrodes are inserted perpendicularly into the apical and basal compartments so that the tip of each arm is immersed in media. Five recordings are taken from each transwell at set time intervals (10 seconds) to calculate the average TEER value. Measurements are recorded from at least three separate transwell inserts. The reference value from a fibronectin coated transwell without cells is subtracted from initial measurements (Equation 2) and the net recording corrected for area of cell growth to yield a final TEER value (Equation 3, Table 2). All measurements are performed at room temperature within 6 minutes of removing cells from the incubator.\n\nNet TEER (Ω) = Measured TEER (Ω) - Reference TEER (Ω) [Equation 2]\n\nFinal TEER (Ω/cm2) = Net TEER (Ω) × Area of transwell membrane (cm2) [Equation 3]\n\nTrans-epithelial Electrical Resistance (TEER) measurements were obtained from long-term cultures to evaluate effectiveness of the Retinal Pigment Epithelial barrier. [A] Measurements were conducted from transwells (n=3) at weekly time intervals after seeding. Values were plotted as a percentage change from the previous measurement which show a gradual increase as junctions form and mature. [B] Fluctuations between average weekly TEER were observed prior to week 6 (p=0.009 at 3 weeks, p=0.013 at 4 weeks and p=0.001 at 6 weeks, one-way ANOVA with Tukey’s multiple comparisons) after which a stable value of 40.72 Ω.cm2 was achieved. Data is presented as mean ± SEM. Next, we quantified polarised secretion of Vascular Endothelial Growth Factor (VEGF) and Pigment Epithelium Derived Factor (PEDF) by ARPE-19 cells. Conditioned media was collected (n=3) after 72 hours and proteins quantified by ELISA. [C] The apical compartment was found to contain 0.942 ± 0.035ng/ml of VEGF compared to 2.852 ± 0.145ng/ml in the basal chamber, which was statistically significant (p=0.0002). [D] PEDF concentrations in the apical compartment was 16.95 ± 0.72ng/ml compared to 25.05 ± 3.93ng/ml in the basal chamber. There were no significant differences (p= 0.112) although more PEDF was secreted via the basolateral RPE surface. Data is presented as mean ± SEM with statistical comparisons made using the unpaired student’s t-test and sourced in part from material published previously18 under the Creative Commons licence.\n\nTechnical tip: Care should be taken to prevent electrodes from touching chamber walls as this results in inconsistent TEER values.\n\nStep 8: ELISA studies of ARPE-19 cultures\n\nThe capacity to secrete proteins directionally can be assessed by performing an ELISA on conditioned media harvested from apical and basal transwell compartments (Figure 1). A Novex® human VEGF solid-phase sandwich ELISA kit (Life Technologies, UK) is used to measure secreted levels of VEGF, whilst a Biovendor Human PEDF solid-phase sandwich ELISA kit (Biovendor, UK) is used to measure secreted levels of PEDF. A complete media change is performed prior to quantifying soluble protein levels during a 2–3 day period. Collected samples are kept at 4°C or on ice before quantification to prevent protein degradation or stored at -80°C for future use. ELISA quantification is carried out in triplicate on a minimum of three separate wells (Figure 7C–D). The volume of media lost due to sampling is restored afterwards by the addition of freshly prepared media into apical and basal transwell compartments. Assays are carried out following the manufacturers’ guidelines (Table 6).\n\nUsers should check assay detection thresholds and suitability for use with conditioned culture media.\n\nStep 9: Photoreceptor outer segment phagocytosis assay\n\nPost-confluent ARPE-19 cells are reported to exhibit phagocytic activity after 2 weeks in culture43, although we recommend using cultures of approximately 4 months so cells exhibit a gene profile comparable to native RPE36.\n\nPorcine eyes are obtained from a butcher or abattoir within 2 days of post mortem and POS isolated on the same day. An incision is made at the ora serata to remove the anterior eye portion after which the retina can be gently detached. We find this is best achieved by teasing the retina away from the RPE in a circular fashion. The optic nerve is severed at the nerve head to detach the retina. The retinae are pooled in KCl buffer (0.3M KCl, 10mM HEPES, 0.5mM CaCl2, 1mM MgCl2; pH 7.0) with 48% sucrose (w/v) and agitated vigorously for 2 minutes on a rotation mixer after which the solution is centrifuged for 5 minutes at 5,000g. At this point isolated POS appear in the supernatant and the pellet can be discarded. Filter the POS containing supernatant through a sterile surgical gauze positioned on a 1.5ml Eppendorf tube into an equal volume of KCl buffer without sucrose and incubate at room temperature for 5 minutes. Centrifuge the suspension at 4,000g for 7 minutes to pellet isolated POS and discard the supernatant. Wash POS pellets three times in 1xPBS and re-suspend in DMEM with 2.5% (w/v) sucrose44. POS is covalently conjugated to fluorescein isothiocyanate (FITC). This is achieved by incubating pooled POS in 5ml labelling buffer (20Mm phosphate buffer pH 7.2, 5mM taurine with 10% w/v sucrose) and 1.5ml FITC stock solution (2mg/mL FITC isomer I in 0.1M Na2CO3 buffer; pH 9.5) on a rotator mixer at room temperature for 1 hour in the dark. Pellet the POS-FITC by centrifugation at 3,000g for 5 minutes, suspend in DMEM with 2.5% sucrose (w/v) and store for a maximum of 6 months at -80°C. Once thawed isolated POS should not be refrozen. The total protein content of POS preparations can be quantified using a BCA assay prior to use.\n\nCultures are incubated at 17°C for 30 minutes after which 4mg/cm2 POS-FITC is applied to RPE cultures for 30 minutes to maximise binding with minimal internalisation39. This concentration is sufficient to challenge each RPE cell with approximately 10 isolated POS molecules44. Alternatively, if cultures cannot be chilled to 17°C they may be incubated with isolated POS for 2 hours at 37°C to achieve a similar effect45. Following the feeding assay, wash inserts once in fresh medium and return to an incubator set at 37°C and 5% CO2. Transwells are removed at desired time points after which they are washed once in 1xHBSS followed by fixation in 1xPBS containing 4% formaldehyde for 30 minutes at 4°C. Wash cells three times in 1x PBS and store at 4°C until use. Immunostaining is performed by blocking/permeabilising cells in PBS-Tween containing 1% BSA for 30 minutes followed by incubation at 4°C overnight with the desired antibody (Table 5) prepared in the same solution. The following day, wash cells three times with 1xPBS to remove any unbound primary antibodies and incubate with the appropriate secondary antibody (step 5) for 1 hour at room temperature. Wash samples as before and incubate with 1μg/ml DAPI (prepared in ddH2O) for 10 minutes before performing three final washes in 1xPBS. Mount the sample between two glass coverslips using Mowiol® mounting medium for confocal microscopy studies (step 5).\n\nFor co-localisation studies we use an unbiased statistical algorithm described by Costes et al.40 and performed using Volocity Software (Perkin Elmer, UK). Considerations prior to undertaking co-localisation studies include careful selection of suitable fluorophores to avoid bleed through and chromatic aberration as well as pixel saturation (Figure 8). We also suggest selecting non-overlapping and non-adjacent fluorophores and refer to several excellent articles on co-localisation studies40,46–48.\n\nCultures were pulsed with photoreceptor outer segments (POS) bound to FITC. Cargo internalisation and trafficking via phagosomes/endosomes, lysosomes and autophagy bodies were quantified at 2, 4, 6, 12, 24 and 48 hours. Following internalisation, POS were detected [A–B] in Rab 5 early vesicles by 2–4 hours. These diminished over time as cargo appeared [C–D] in Rab 7-positive compartments. [E–F] At 6 hours, a large proportion of POS were present in early lysosomes and [G–H] in LAMP2 vesicles by 12–24 hours. [I–J] Cargos appeared in LC3B-positive autophagy bodies afterwards which were present up to 48 hours. Images show representative confocal images with quantification (n=10 cells/compartment/time point) in 2 independent experiments. Inserts show extent of co-localisation in red (R) and green (G) channels using an unbiased quantification method by Volocity. Data is presented as mean ± SEM and sourced in part from material published previously18,50 under the Creative Commons licence.\n\n\nResults\n\nARPE-19 monolayers on transwell inserts can be easily maintained in long term culture (Figure 1). This allows them to mature and express structural and physiological features of native RPE. Our experiments were carried out on monolayers that had been in culture for 2–4 months. We also tested the ability of ARPE-19 cells to attach and spread on transwell membranes with or without the presence of fibronectin; the preferred substrate for these cells in our experience18,49,50. Our findings show that cells were capable of attachment and growth to confluence on PET membranes irrespective of the presence/absence of an underlying fibronectin matrix (Figure 2). Prior to carrying out imaging studies transwell membranes were carefully removed from their plastic wells. We show a convenient method by which even a small transwell insert can be sectioned into several segments so that the investigator is able to probe for multiple markers and thus maximise the possibility of obtaining data from each transwell (Figure 3). In-vitro RPE monolayers were studied for physiological and structural features characteristic of RPE cells. We first probed for junctional complexes zonula occludens (ZO-1) (Figure 4A–D) and occludin (Figure 4E–H). We also looked for evidence of apically expressed Na+/K+ ATPase (Figure 4I–L) and the cell-specific marker RPE65 (Figure 4M–P). After 2 months in culture ARPE-19 monolayers expressed the early tight-junction protein ZO-1 with a border demarcating cell-to-cell contact, and 3D imaging revealing polarisation towards the apical cellular region. ZO-1 staining was also observed in the cytoplasm and prominently in the nucleus, which is consistent with reported literature51. Expression of occludin during mid-late stages of barrier formation was also observed. Next, we probed for expression of the Na+/K+ ATPase transporter to assess the presence of a polarised plasma membrane. Na+/K+ ATPase is predominantly expressed on the apical RPE surface where it facilitates the process of photo-transduction. Apically expressed Na+/K+ ATPase is also linked with a highly differentiated, polarised RPE phenotype52. We detected Na+/K+ ATPase in some but not all cells, although expression appeared to be limited to the apical RPE surface (Figure 4L). We also probed for the cell-specific retinoid isomerohydrolase RPE65 marker to confirm identity of RPE cells in the monolayer (Figure 4M–P). RPE65 was observed as punctate, cytoplasmic staining as reported by others53 and confirmed the identity of RPE cells in long-term culture. Next, we assessed the extent to which ARPE-19 monolayers on transwells adopt ultrastructural features of native RPE. We describe a convenient technique by which transwell membranes can be sectioned into smaller segments to be embedded in resin blocks for TEM studies (Figure 5). We observed evidence of numerous microvilli on the apical RPE surface (Figure 6A) and infolded/convolutions of the basolateral cell membrane (Figure 6B), characteristic of native RPE. Micrographs also showed details of intracellular organelles including mitochondria, compartments in the endocytic-lysosomal pathway and pigment molecules (Figure 6C–E). Mitochondria, for instance, appear in cross-section as a double membrane-bound structures with luminal cristae, whilst vesicles contained cargos of varying electron densities. The arrangements of these organelles conformed to the apical-basolateral axis of native RPE. Junctional complexes between RPE, detected previously by immunofluorescence studies (Figure 4A–H), were also observed at ultrastructural resolution as tight junctions and adherens junctions along membranes at the apical region of RPE cells (Figure 6F–G). These were observed as electron-dense regions indicating points of cell-to-cell contact and associated with desmosomes in some instances.\n\nEstablishment of an effective trans-epithelial barrier is a key feature of native RPE, and one that can be readily measured in transwell cultures. We carried out TEERs of ARPE-19 cultures over an approximately 3 month period. A stable electrical gradient was established following 6 weeks in culture (Figure 7A), after which there were no appreciable changes to the barrier (Figure 7B). An average TEER value of 40.72 Ω/cm2 was noted once cultures had established a stable barrier in-line with previous reports22,25. Polarised secretion of molecules towards the overlying neuroretina and the underlying choroid is an important feature of RPE cells15. Proteins such as VEGF and PEDF that are synthesised/secreted by RPE are known to possess pro-angiogenic and neuroprotective effects, respectively1. Directional secretion of such molecules can easily be quantified in transwell compartments once cells establish an effective trans-epithelial barrier. To assess if this was achieved in culture we measured VEGF and PEDF levels in conditioned media using two different ELISAs. ARPE-19 cells secreted VEGF through both apical and basolateral surfaces at concentrations of 0.942 ± 0.035ng/ml and 2.852 ± 0.145ng/ml, respectively. VEGF secretion towards the choroid was therefore significantly higher compared to amounts released towards the neuroretina (Figure 7C). PEDF levels were also secreted via both surfaces at concentrations of 16.95 ± 0.72ng/ml (apical) and 25.05 ± 3.93ng/ml (basal). Statistically, there were no differences in amounts of PEDF secreted towards the choroid or neuroretina (Figure 7D). Next, we assessed the ability of cultured ARPE-19 cells to bind and internalise POS cargo. In-situ RPE daily internalises and proteolytically degrade POS from overlying photoreceptors, the impairment of which plays a key role in retinopathy1. POS-FITC cargos were fed to 4 month old monolayers using a pulse-chase method described previously39. Each compartment in the endosome/phagosome-lysosomal and autophagy pathway was assessed at 2, 4, 6, 12, 24 and 48 hours for the extent of co-localisation with fluorescently-labelled POS. Cargos initially appeared in early Rab 5 compartments (Figure 8A–B), which by ~6 hours had trafficked to Rab 7 late vesicles (Figure 8C–D). Between 6–12 hours, a large proportion of cargo had co-localised to early LAMP1 (Figure 8E–F) and mature LAMP2A lysosomes (Figure 8G–H). 48 hours after the pulse-chase assay was initiated, a large proportion of cargo appeared in LC3B-positive autophagy bodies (Figure 8I–J).\n\n\nDiscussion\n\nIn this article, we describe a convenient protocol by which users can rapidly establish and study RPE cells in long-term culture. We used the human ARPE-19 cell-line, although the approaches described herein may be adopted for studying RPE from a variety of sources. We provide examples from our own laboratory as well as other groups to highlight the type of questions which investigators could realistically address using this model. These are by no means exhaustive as there is a considerable amount of literature that is beyond the scope of this article. Readers are directed to accompanying citations as well as to a special issue of Experimental Eye Research for detailed reviews covering specific aspects of RPE biology54; Retinal Pigment Epithelium cell culture: Current standards and technical criteria for model systems (2014, Volume 126; 1–84, Edited by Bruce A. Pfeffer and Nancy J. Philp). We also discuss the versatility of transwell culture models as well as advantages and limitations of APRE-19 cells in particular, and where in-vitro models could replace similar work carried out in animals.\n\nThe use of transwell inserts to culture RPE is widely accepted as the best method to study this cell-type and to model dysfunction of this important tissue in-vitro. This has led to a plethora of studies in which RPE cells from different sources have been cultured on transwell systems16,17,19–22,24. The use of transwell supports appear to mimic important structural features of the BrM, as cells displayed desirable structural and functional features of native RPE. Limitations to this approach are driven largely by the source/type of RPE cells as well as differences in culture conditions. For instance, the culture of hfRPE on transwell inserts produced what some consider to be the holy grail of RPE culture by mimicking drusen formation associated with complement activation55. Further advances in RPE modelling were made by culturing PSC-RPE from Sorsby fundus dystrophy, Doyne honeycomb retinal dystrophy/malattia Leventinese and autosomal dominant radial drusen patients, which recapitulated important RPE-associated disease features in-vitro56. Use of the ARPE-19 cell-line by contrast, which had been in use for significantly longer, appeared to be less attractive. When cultured under certain growth conditions these cells failed to replicate directional secretion of proteins, showed limited evidence of pigmentation and impaired retinoid metabolism as well as poor expression of markers PMEL17, BEST1, CRALBP and MerTK22,25,57. More recently however, it has been demonstrated that when cultured under optimised growth medium on transwell inserts for extended periods, ARPE-19 cells regain a phenotype and gene expression profile comparable to that of native RPE cells25,36. Moreover, recent improvements to ARPE-19 culture meant that features that were difficult to reproduce previously including directional secretion of proteins, apical-basolateral morphology, pigmentation, internalisation of POS and expression of mRNA/proteins in retinoid metabolism have all been successfully recapitulated18,25. In fact, when cultured for ~4 months, the genetic profile of ARPE-19 cells were found to be comparable to hfRPE and PSC-RPE36. Through the adoption of high glucose and sodium pyruvate media described by Ahmado and colleagues25 as well as other advances described herein, we and others have collectively improved the capacity to exploit ARPE-19 cells to model specific aspects of retinopathy under culture conditions. For example, we have shown that after two months in culture, VEGF and PEDF are secreted by ARPE-19 at concentrations similar to levels reported in hfRPE18,22. In-vitro models such as those described herein provide a convenient method to assay directional secretion of molecules which would otherwise be challenging to study using in-situ RPE in mouse models. We also show the presence of desirable ultrastructural features after 2 months in culture including apical-basolateral specialisation and presence of apically distributed pigment which typically become visible without a microscope after 3–4 months in culture18,25. Similarly, for the first time, we report the presence of RPE65 and the apical expression of Na+/K+ ATPase after just 2 months in culture18. Na+/K+ ATPase expression however was limited to a sub-population of RPE cells. A similar observation was made by others after 15 weeks in culture25, suggesting a potential limitation of this cell-line. Given that the expression of this transporter is correlated with the increased pigment25, we recommend maintaining ARPE-19 cultures for 2–4 months before carrying out any studies on trans-epithelial transport.\n\nARPE-19 cells express the appropriate surface receptors and ligands for binding to POS. Investigators have therefore exploited these cells to study cargo trafficking and impairment of this process which is associated with disease50,58. However, some caution is advised as post confluent ARPE-19 cells were reported to internalise cargo after only 2 weeks without expressing MerTK (for POS internalisation) or ITGAV (for POS recognition/binding) receptors until at least 4 months in culture36. We therefore recommend that POS binding and trafficking studies should only be carried out after this period. Investigators should also note that cargo trafficking rates differ somewhat between primary RPE and ARPE-19 cells with slower speeds reported in the latter58. Using a pulse-chase assay we used cultures to obtain a detailed timeline of POS trafficking in ARPE-19 cells18,50. Our findings reveal the time wise entry of POS to Rab 5-positive early compartments followed by Rab 7 late vesicles and appearance in lysosomes between 6–12 hours. Cargos were detected in LC3B autophagy bodies as late as 48 hours after pulse chase, although a majority of cargo were degraded within 16–20 hours as reported before58. Importantly, such new information defining trafficking rates in heathy cells may be used as a reference point in modelling specific disease conditions, and how they affects POS processing and degradation linked to retinopathies. These studies are far easier to manipulate and carryout under in-vitro conditions hence investigators may prefer to eschew mouse models for this type of work.\n\nAnother important feature of the RPE modelled under in-vitro conditions is their capacity to develop and form junctional complexes integral to creating the BRB14. Although we observed the expression of early and mid-late barrier complexes after 2 months in culture by confocal microscopy as well as tight junctions and adherens junctions at ultrastructural resolution, ARPE-19 cells appear to be less well suited to barrier studies. ZO-1 is reported to shuttle to/from the nucleus depending on the extent and maturity of tight junctions51,59. We observed strong ZO-1 staining in cell margins as well as in nuclei after 8 weeks, suggesting a potentially incomplete maturation of the barrier. Staining for occludin was also limited to circumferential tight junctions. 15 week old ARPE-19 cells however displayed a stronger pattern of staining for ZO-1 and occludin25, suggesting a degree of on-going barrier maturation at earlier time points (8 weeks) at which our experiments were carried out. Immature and/or incomplete barrier complexes appear to be reflected in low TEER values for ARPE-19 cells, which is seldom reported to exceed 50 Ω/cm225. We recorded an average TEER of 40.7 Ω/cm2 over a 3 month period, which was substantially less compared to values of 200–1500 Ω/cm2 reported in RPE cells from other sources33. For these reasons we do not advocate the use of the ARPE-19 cell-line for barrier studies.\n\nIn summary, this article provides an in-depth set-up and validation protocol for establishing a culture model of the outer retina using the widely utilized ARPE-19 cell-line. We also discussed advantages and limitations of transwell models in general and ARPE-19 cells in particular, so that users may best exploit this versatile system for their studies. Advantages over mouse models such as (1) its use as a viable alternative, (2) ability to rapidly generate functional RPE monolayers akin to native tissues, and (3) ability to reproduce disease features that could only be previously studied in mice18,55,56, makes in-vitro models of the outer retina especially attractive. Their versatility is further demonstrated by studies in which the surface of transwell membranes are directly modified to mimic effects of aging23. Investigators are also well-placed to take advantage of new developments in stem-cell technology and refinements to in-vitro cultures described herein60–62 as well as a plethora of artificial BrM substrates on offer. The latter may be set-up and assembled similarly to transwells by using commercially available products such as CellCrownTM inserts (Sigma, UK). The growing interest in microfluidic devices allow laboratories to model relationships between the RPE vs. choroidal endothelial cells and blood flow by incorporating the latter into transwell devices. These advances combined with the development of fast/high-resolution imaging and new 3D imaging platforms such as serial block face scanning electron microscopy and Lightsheet are likely to usher in further opportunities to exploit in-vitro models. Consequently, investigators may wish to consider these culture models as attractive alternatives to using animals, or at least as powerful new tools to be exploited in parallel that will also have the benefit of reducing and replacing animals used in research.\n\n\nData availability\n\nDataset 1: Raw data underlying Figure 2 10.5256/f1000research.15409.d20925263\n\nDataset 2: Raw data underlying Figure 4 and S1 10.5256/f1000research.15409.d20925364\n\nDataset 3: Raw data underlying Figure 6 10.5256/f1000research.15409.d20925465\n\nDataset 4: Raw data underlying Figure 7 10.5256/f1000research.15409.d20925566\n\nDataset 5: Raw data underlying Figure 8 10.5256/f1000research.15409.d20925667\n\nFigure 4, Figure 6, Figure 7 and Figure 8 has been published previously either in part or in whole (Ratnayaka et al., IOVS 201549; Lynn et al., 2017: doi:10.1016/j.tice.2017.06.00318; Keeling et al., 2018: doi:10.3390/cells702001650), and presented here under the terms of the Creative Commons (CC) license.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was funded by awards to JAR from the National Centre for the Replacement, Refinement & Reduction of Animals in Research [NC/L001152/1], the Macular Society, RP Fighting Blindness [GR590] and the Gift of Sight Appeal.\n\n\nAcknowledgements\n\nThe authors thank Dr Elizabeth Angus and Dr Suzanne Brooks (Biomedical Imaging Unit, University of Southampton) for technical assistance with TEM and Mr Alexander Rodgers (Longitude Engineers, Southampton) for help with preparing images.\n\n\nSupplementary material\n\nFigure S1: Secondary antibody controls used to detect Retinal Pigment Epithelium proteins in ARPE-19 cultures on transwells. [A–C] Probing with goat anti-rabbit Alexa 594 (red) or [D–F] goat anti-mouse Alexa 594 (red) produced no signals. Nuclei were labelled with DAPI (blue). Scale bars correspond to 20 µm.\n\nClick here to access the data.\n\n\nReferences\n\nKhandhadia S, Cherry J, Lotery AJ: Age-related macular degeneration. Adv Exp Med Biol. 2012; 724: 15–36. PubMed Abstract | Publisher Full Text\n\nBhutto I, Lutty G: Understanding age-related macular degeneration (AMD): relationships between the photoreceptor/retinal pigment epithelium/Bruch's membrane/choriocapillaris complex. Mol Aspects Med. 2012; 33(4): 295–317. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBonilha VL: Age and disease-related structural changes in the retinal pigment epithelium. Clin Ophthalmol. 2008; 2(2): 413–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPennesi ME, Neuringer M, Courtney RJ: Animal models of age related macular degeneration. Mol Aspects Med. 2012; 33(4): 487–509. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVolland S, Esteve-Rudd J, Hoo J, et al.: A comparison of some organizational characteristics of the mouse central retina and the human macula. PLoS One. 2015; 10(4): e0125631. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNagai N, Lundh von Leithner P, Izumi-Nagai K, et al.: Spontaneous CNV in a novel mutant mouse is associated with early VEGF-A-driven angiogenesis and late-stage focal edema, neural cell loss, and dysfunction. Invest Ophthalmol Vis Sci. 2014; 55(6): 3709–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTobe T, Ortega S, Luna JD, et al.: Targeted disruption of the FGF2 gene does not prevent choroidal neovascularization in a murine model. Am J Pathol. 1998; 153(5): 1641–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMata NL, Tzekov RT, Liu X, et al.: Delayed dark-adaptation and lipofuscin accumulation in abcr+/- mice: implications for involvement of ABCR in age-related macular degeneration. Invest Ophthalmol Vis Sci. 2001; 42(8): 1685–90. PubMed Abstract\n\nMalek G, Johnson LV, Mace BE, et al.: Apolipoprotein E allele-dependent pathogenesis: a model for age-related retinal degeneration. Proc Natl Acad Sci U S A. 2005; 102(33): 11900–05. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToomey CB, Kelly U, Saban DR, et al.: Regulation of age-related macular degeneration-like pathology by complement factor H. Proc Natl Acad Sci U S A. 2015; 112(23): E3040–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamkumar HL, Zhang J, Chan CC: Retinal ultrastructure of murine models of dry age-related macular degeneration (AMD). Prog Retin Eye Res. 2010; 29(3): 169–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmin S, Chong NH, Bailey TA, et al.: Modulation of Sub-RPE deposits in vitro: a potential model for age-related macular degeneration. Invest Ophthalmol Vis Sci. 2004; 45(5): 1281–8. PubMed Abstract | Publisher Full Text\n\nKlingeborn M, Dismuke WM, Skiba NP, et al.: Directional Exosome Proteomes Reflect Polarity-Specific Functions in Retinal Pigmented Epithelium Monolayers. Sci Rep. 2017; 7(1): 4901. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenedicto I, Lehmann GL, Ginsberg M, et al.: Concerted regulation of retinal pigment epithelium basement membrane and barrier function by angiocrine factors. Nat Commun. 2017; 8: 15374. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKay P, Yang YC, Paraoan L: Directional protein secretion by the retinal pigment epithelium: roles in retinal health and the development of age-related macular degeneration. J Cell Mol Med. 2013; 17(7): 833–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToops KA, Tan LX, Lakkaraju A: A detailed three-step protocol for live imaging of intracellular traffic in polarized primary porcine RPE monolayers. Exp Eye Res. 2014; 124: 74–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFernandez-Godino R, Garland DL, Pierce EA: Isolation, culture and characterization of primary mouse RPE cells. Nat Protoc. 2016; 11(7): 1206–18. PubMed Abstract | Publisher Full Text\n\nLynn SA, Ward G, Keeling E, et al.: Ex-vivo models of the Retinal Pigment Epithelium (RPE) in long-term culture faithfully recapitulate key structural and physiological features of native RPE. Tissue Cell. 2017; 49(4): 447–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSonoda S, Spee C, Barron E, et al.: A protocol for the culture and differentiation of highly polarized human retinal pigment epithelial cells. Nat Protoc. 2009; 4(5): 662–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu J, Bok D: A cell culture medium that supports the differentiation of human retinal pigment epithelium into functionally polarized monolayers. Mol Vis. 2000; 7: 14–19. PubMed Abstract\n\nMaminishkis A, Chen S, Jalickee S, et al.: Confluent Monolayers of Cultured Human Fetal Retinal Pigment Epithelium Exhibit Morphology and Physiology of Native Tissue. Invest Ophthalmol Vis Sci. 2006; 47(8): 3612–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAblonczy Z, Dahrouj M, Tang PH, et al.: Human retinal pigment epithelium cells as functional models for the RPE in vivo. Invest Ophthalmol Vis Sci. 2011; 52(12): 8614–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKay P, Yang YC, Hiscott P, et al.: Age-related changes of cystatin C expression and polarized secretion by retinal pigment epithelium: potential age-related macular degeneration links. Invest Ophthalmol Vis Sci. 2014; 55(2): 926–34. PubMed Abstract | Publisher Full Text\n\nBlenkinsop TA, Salero E, Stern JH, et al.: The culture and maintenance of functional retinal pigment epithelial monolayers from adult human eye. Methods Mol Biol. 2013; 945: 45–65. PubMed Abstract | Publisher Full Text\n\nAhmado A, Carr AJ, Vugler AA, et al.: Induction of differentiation by pyruvate and DMEM in the human retinal pigment epithelium cell line ARPE-19. Invest Ophthalmol Vis Sci. 2011; 52(10): 7148–59. PubMed Abstract | Publisher Full Text\n\nDunn KC, Aotaki-Keen AE, Putkey FR, et al.: ARPE-19, a human retinal pigment epithelial cell line with differentiated properties. Exp Eye Res. 1996; 62(2): 155–69. PubMed Abstract | Publisher Full Text\n\nCarr AJ, Vugler A, Lawrence J, et al.: Molecular characterization and functional analysis of phagocytosis by human embryonic stem cell-derived RPE cells using a novel human retinal assay. Mol Vis. 2009; 15: 283–95. PubMed Abstract | Free Full Text\n\nChristensen DRG, Brown FE, Cree AJ, et al.: Sorsby fundus dystrophy - A review of pathology and disease mechanisms. Exp Eye Res. 2017; 165: 35–46. PubMed Abstract | Publisher Full Text\n\nOsterloh JM, Mullane K: Manipulating cell fate while confronting reproducibility concerns. Biochem Pharmacol. 2018; 151: 144–56. PubMed Abstract | Publisher Full Text\n\nNabi IR, Mathews AP, Cohen-Gould L, et al.: Immortalization of polarized rat retinal pigment epithelium. J Cell Sci. 1993; 104(Pt 1): 37–49. PubMed Abstract\n\nKuznetsova AV, Kurinov AM, Aleksandrova MA: Cell models to study regulation of cell transformation in pathologies of retinal pigment epithelium. J Ophthalmol. 2014; 2014: 801787. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRizzolo LJ, Peng S, Luo Y, et al.: Integration of tight junctions and claudins with the barrier functions of the retinal pigment epithelium. Prog Retin Eye Res. 2011; 30(5): 296–323. PubMed Abstract | Publisher Full Text\n\nRizzolo LJ: Barrier properties of cultured retinal pigment epithelium. Exp Eye Res. 2014; 126: 16–26. PubMed Abstract | Publisher Full Text\n\nMannermaa E, Reinisalo M, Ranta VP, et al.: Filter-cultured ARPE-19 cells as outer blood-retinal barrier model. Eur J Pharm Sci. 2010; 40(4): 289–96. PubMed Abstract | Publisher Full Text\n\nFinnemann SC, Bonilha VL, Marmorstein AD, et al.: Phagocytosis of rod outer segments by retinal pigment epithelial cells requires alpha(v)beta5 integrin for binding but not for internalization. Proc Natl Acad Sci U S A. 1997; 94(24): 12932–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSamuel W, Jaworski C, Postnikova OA, et al.: Appropriately differentiated ARPE-19 cells regain phenotype and gene expression profiles similar to those of native RPE cells. Mol Vis. 2017; 23: 60–89. PubMed Abstract | Free Full Text\n\nThomson HA, Treharne AJ, Backholer LS, et al.: Biodegradable poly(α-hydroxy ester) blended microspheres as suitable carriers for retinal pigment epithelium cell transplantation. J Biomed Mater Res A. 2010; 95(4): 1233–43. PubMed Abstract | Publisher Full Text\n\nThomson HA, Treharne AJ, Walker P, et al.: Optimisation of polymer scaffolds for retinal pigment epithelium (RPE) cell transplantation. Br J Ophthalmol. 2011; 95(4): 563–68. PubMed Abstract | Publisher Full Text\n\nHall MO, Abrams T: Kinetic studies of rod outer segment binding and ingestion by cultured rat RPE cells. Exp Eye Res. 1987; 45(6): 907–22. PubMed Abstract | Publisher Full Text\n\nCostes SV, Daelemans D, Cho EH, et al.: Automatic and quantitative measurement of protein-protein colocalization in live cells. Biophys J. 2004; 86(6): 3993–4003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhone A, Lo J, Han AH, et al.: Glucose and Pyruvate Toxicity for Human Retinal Pigment Epithelial Cells ARVO 2018. Honolulu, Hawaii, 2018.\n\nMao Y, Finnemann SC: Analysis of photoreceptor outer segment phagocytosis by RPE cells in culture. Methods Mol Biol. 2013; 935: 285–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDejos C, Kuny S, Han WH, et al.: Photoreceptor-induced RPE phagolysosomal maturation defects in Stargardt-like Maculopathy (STGD3). Sci Rep. 2018; 8(1): 5944. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrohne TU, Holz FG, Kopitz J: Apical-to-basolateral transcytosis of photoreceptor outer segments induced by lipid peroxidation products in human retinal pigment epithelial cells. Invest Ophthalmol Vis Sci. 2010; 51(1): 553–60. PubMed Abstract | Publisher Full Text\n\nFinnemann SC, Silverstein RL: Differential roles of CD36 and alphavbeta5 integrin in photoreceptor phagocytosis by the retinal pigment epithelium. J Exp Med. 2001; 194(9): 1289–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZinchuk V, Zinchuk O, Okada T: Quantitative colocalization analysis of multicolor confocal immunofluorescence microscopy images: pushing pixels to explore biological phenomena. Acta Histochem Cytochem. 2007; 40(4): 101–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLagache T, Sauvonnet N, Danglot L, et al.: Statistical analysis of molecule colocalization in bioimaging. Cytometry A. 2015; 87(6): 568–79. PubMed Abstract | Publisher Full Text\n\nDunn KW, Kamocka MM, McDonald JH: A practical guide to evaluating colocalization in biological microscopy. Am J Physiol Cell Physiol. 2011; 300(4): C723–C42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRatnayaka JA, Lynn SA, Griffiths H, et al.: An ex-vivo platform for manipulation and study of Retinal Pigment Epithelial (RPE) cells in long-term culture. Investigative Ophthalmology & Visual Science. 2015; 56(7): 2332–32.\n\nKeeling E, Lotery AJ, Tumbarello DA, et al.: Impaired Cargo Clearance in the Retinal Pigment Epithelium (RPE) Underlies Irreversible Blinding Diseases. Cells. 2018; 7(2): pii: E16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBauer H, Zweimueller-Mayer J, Steinbacher P, et al.: The Dual Role of Zonula Occludens (ZO) Proteins. J Biomed Biotechnol. 2010; 2010: 402593. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLehmann GL, Benedicto I, Philp NJ, et al.: Plasma membrane protein polarity and trafficking in RPE cells: past, present and future. Exp Eye Res. 2014; 126: 5–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang J, Possin DE, Saari JC: Localizations of visual cycle components in retinal pigment epithelium. Mol Vis. 2009; 15: 223–34. PubMed Abstract | Free Full Text\n\nPfeffer BA, Philp NJ: Cell culture of retinal pigment epithelium: Special Issue. Exp Eye Res. 2014; 126: 1–4. PubMed Abstract | Publisher Full Text\n\nJohnson LV, Forest DL, Banna CD, et al.: Cell culture model that mimics drusen formation and triggers complement activation associated with age-related macular degeneration. Proc Natl Acad Sci U S A. 2011; 108(45): 18277–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGalloway CA, Dalvi S, Hung SSC, et al.: Drusen in patient-derived hiPSC-RPE models of macular dystrophies. Proc Natl Acad Sci U S A. 2017; 114(39): E8214–E8223. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarr AJ, Vugler AA, Yu L, et al.: The expression of retinal cell markers in human retinal pigment epithelial cells and their augmentation by the synthetic retinoid fenretinide. Mol Vis. 2011; 17: 1701–15. PubMed Abstract | Free Full Text\n\nMazzoni F, Safa H, Finnemann SC: Understanding photoreceptor outer segment phagocytosis: use and utility of RPE cells in culture. Exp Eye Res. 2014; 126: 51–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGottardi CJ, Arpin M, Fanning AS, et al.: The junction-associated protein, zonula occludens-1, localizes to the nucleus before the maturation and during the remodeling of cell-cell contacts. Proc Natl Acad Sci U S A. 1996; 93(20): 10779–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHazim RA, Volland S, Williams DS: A Rapid protocol for the differentiation of human ARPE-19 cells. ARVO 2018. Honolulu, Hawaii IOVS, 2018.\n\nSmith J, Wardani S, Carr AJ: Differentiation of ARPE-19 cells in X-Vivo culture medium. ARVO 2018. Honolulu, Hawaii: IOVS, 2018.\n\nCarr AJ, Swann C, Radeke MJ, et al.: Using ARPE-19 cells to investigate pathways associated with retinal pigment epithelium differentiation. ARVO 2018. Honolulu, Hawaii: IOVS, 2018.\n\nLynn SA, Keeling E, Dewing JM, et al.: Dataset 1 in: A convenient protocol for establishing a human cell culture model of the outer retina. F1000Research. 2018. Data Source\n\nLynn SA, Keeling E, Dewing JM, et al.: Dataset 2 in: A convenient protocol for establishing a human cell culture model of the outer retina. F1000Research. 2018. Data Source\n\nLynn SA, Keeling E, Dewing JM, et al.: Dataset 3 in: A convenient protocol for establishing a human cell culture model of the outer retina. F1000Research. 2018. Data Source\n\nLynn SA, Keeling E, Dewing JM, et al.: Dataset 4 in: A convenient protocol for establishing a human cell culture model of the outer retina. F1000Research. 2018. Data Source\n\nLynn SA, Keeling E, Dewing JM, et al.: Dataset 5 in: A convenient protocol for establishing a human cell culture model of the outer retina. F1000Research. 2018. Data Source"
}
|
[
{
"id": "36346",
"date": "14 Aug 2018",
"name": "David Y.S. Chau",
"expertise": [
"Reviewer Expertise Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 article aims to address the utilisation of a viable, in vitro, retinal pigment epithelium (RPE) cell culture model, as an alternative to animal models, for disease elucidation and the ageing mechanism. Specifically, the article describes a novel, step-by-step, technique to establish an experimentally versatile model of the outer retina- incorporating the RPE monolayer and associated Bruch’s membrane (BrM).\n\nGeneral Assessment: Introduction: Concise overview and summary of RPE cells and aligns the importance of this layer and to disease aetiology. Examples of current in vivo models used for the assessment of retinal pathobiology is mentioned and the exploitation of mice is emphasised appropriately- although the inclusion of absolute specific numbers of rodents/mice should be stated if possible. For completeness of 3R, a sentence regarding in silico/computational alternate retinal models should be included in the Introduction. The authors correctly highlight the importance of a suitable microenvironment to the induction/differentiation of the (RPE) cells and emphasis is placed on Transwells. However, a brief mention of additional stimuli/factors should also be elaborated upon including use of hydrogels, co-culture, surface topography, surface geometry/curvature, growth factors- see literature. Good overview of ARPE-19 cells- with emphasis of it being human cell line. Should a critique of the use of human donor tissue/cells be mentioned- from a 3R perspective?\n\nM&M: It may be worth stating the original passage number of these cells, if known and/or since purchase, as the culture methodology is dependent on low/high status c.f. Dunn et al., 1996, Ahmado et al., 2011, Samuel et al., 2017. Would using 1% human serum and/or serum-alternative be possible- considering the implication of the 3R? Why fibronectin coating and not the other ECM proteins/laminin/Matrigel- laminin is a component of basal RPE lamina? The coating exploits surface adsorption- why do you let the Transwell “dry” overnight and how can you aspirate residual fibronectin if so- rather to “coat” overnight? Confirmation of the presence of fibronectin coating was achieved by? Agitation during the coating procedure? Is the ELISA a single timepoint at 2-months or cumulative collection of media- how is this achieved when there are frequent media changes required? FBS contains exogenous (ECM) proteins as well as VEGF and PEDF? Please state “room temperature” for TEER measurement (and POS)- I assume that this was performed in a hood rather than “on the bench” hence airflow/room temperature appropriate? Capital “T” for Transwell throughout.- registered product.\n\nProtocol:\nStep 1: Additionally filter step may be seen as “overkill”- an additional step could actually increase infection risk- reagents can simply be added using aseptic technique? Please mention storage condition and “warming-up” protocol for media. Discrepancy to name of “FBS” mentioned in the M&M and Table 1: FBS = foetal calf serum; N4762 is newborn calf serum (NBCS)- difference in composition especially antibodies and exogenous protein profile.\nStep 2: Amend to ensure that the FN-coating step is mentioned before the seeding of cells- the former is required before the latter and should also tally with Figure 1. What % confluency is reached before media change? The 6-minute trypsin protocol is very exact- did you confirm visually under a microscope? Over/under-incubating can be problematic as it may lead to selective population isolation- also, as the author stated, longer-term cells generate their own ECM which will enhance the cells’ adhesiveness to the substrate. Was a simple viability and cell count i.e. trypan blue performed- needs to be mentioned in Step 4- determine initial seeding densities?\nStep 3: any agitation used during the coating process? Step 3 (technical tip): fibronectin solution should never be vortexed/centrifuged either- also results in “crashing out” of solution. Note that “steps” in Figure 1 and main text are out of synch- please amend e.g. Step 3 in Figure 1 is “seeding” whereas in the text, Step 4 is seeding the Transwell.\nStep 4: State volume applied to apical and basal compartments prior to seeding? What volume was the cell suspension when adding to the Transwells- possible sink conditions? Clarify what “obvious signs of pigmentation is”. Step 4 (technical tip): was there a slight agitation when cells were first added to ensure uniform coating of substrate?\nStep 5: Steps out of sync with Figure legend. Can anything about morphology be discussed from the images? The study emphasises the long-term ability of the culture conditions i.e. 2-4 months- do you have any associated images for the extended timepoints to indicate viability and/or correct phenotype? Previous researchers have exploited wax embedding as a way to preserve/protect the samples before imaging- similar to your TEM preparation- any reason/justification/mention of the suitability? Capital A in antibodies. What confocal was used and associated operating conditions required. Could Figure 4, panel I, J and K be recomposed with the same magnification/scale as the other panels- allow easier comparison. State timepoint of image acquisition- assuming 2 months based on RPE65 sentence? Full name for NGS required- normal goat serum?\nStep 6: State the dimensions/manufacturer of the embedding capsule used. Details for Reynold’s Lead stain required i.e. citrate? concentration, time.\nIsolation of POS: Please state pore size/manufacturer for the surgical gauze used for filtration.\nPOS feeding assay: how were the samples chilled/incubated to 17°C- state instrument used and/or humidified/gaseous incubator? Optimal internalisation occurs at 37°C- so why 2 hours at 37°C same as 30 minutes at 17°C?\n\nResults: Authors comment on attachment and spreading characteristics but have not alluded to what is ideal/typical of RPE cells- expansion needed regarding morphology of cells. Growth, per se, has not really been documented- absolute cell count, mitochondrial activity assessment i.e. MTT/XTT/MTS assays required alongside a complementary LDH-release profile would fully confirm that the cells are active, alive and not subject to (induced) death i.e. apoptosis or necrosis. No cell numbers are mentioned throughout the long-term culture aside from the initial cell seeding density and justification of contact inhibition. Appropriate representative markers have been selected and results suggest that the cells are of the correct phenotype, barrier formation and correct functionality.\n\nDiscussion: A good descriptive and critical overview is provided by the authors. However, there is actually no direct incorporation of a Bruch’s membrane e.g. ex vivo tissue or biomimetic collagen-elastin substitute within this study which contrasts the statement(s) made in the Abstract and Research Highlights. The authors also point out the shortfall of their model e.g. ATPase expression, TEER/barrier function, POS binding. I would like a bit more discussion with regards to how this technique/protocol could be applied to mimic diseased states- is it simply culturing cells from (diseased) donor tissues/cells? Would a full (re)validation of culture conditions be required or is the assumption that Transwell plus culture media optimisation would suffice? Could this be incorporated within a 3D model? Multilayer/co-culture format? Slight expansion on future integration in ocular toxicity assessment and drug discovery would be beneficial.\n\nReferences: All appropriate- suitable and up-to-date\n\nSupplementary material: All appropriate and relevant\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "37636",
"date": "13 Sep 2018",
"name": "Ian J. Jackson",
"expertise": [
"Reviewer Expertise Mouse models of human disease"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper is describing the culture of ARPE-19 cells on fibronectin-coated membranes. Cultures of ARPE-19 cells have been used for some time as models of retinal pigment epithelium. Other cell types are used: human fetal RPE cells, pig or mouse primary RPE cells and latterly human induced pluripotent stem cells and embryonic stem cells. The method described overall is not novel, but it is very well documented in considerable detail in this paper. The use of ARPE-19 cells has some merits in terms of reproducibility.\n\nWhilst differentiated iPS cells from patients is a means of investigating diseases processes in vitro, this is time consuming and technically challenging. On the other hand ARPE-19 cells have been edited by CRISPR and could be used to the same ends possibly more easily. Further engineering of ARPE-19 cells could introduce, for example, reporters for live imaging.\n\nThe methodology is clear and the protocols are detailed.\n\nThe results as presented are clear. Once downloaded and unzipped, the source data are clear\n\nIn the Abstract the authors say that they are describing a model of the outer retina incorporating the RPE and the supportive Bruch's membrane. I think it is debatable whether this is a good model of Bruch's membrane; there isn't a full barrier established, for example. In their characterisation they compare their data to previously published data from other systems; this isn't ideal as there could be protocol and lab differences. It would also be worth noting in the Discussion that previous work (Ablonczy et al 2011, their ref 22) has compared ARPE-19 cells to human fetal RPE cells grown on permeable filters and concluded that hfRPE cells were a better model. They concluded that ARPE-19 cells resembled a pathological RPE.\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1107
|
https://f1000research.com/articles/7-1105/v1
|
18 Jul 18
|
{
"type": "Software Tool Article",
"title": "drawProteins: a Bioconductor/R package for reproducible and programmatic generation of protein schematics",
"authors": [
"Paul Brennan"
],
"abstract": "Protein schematics are valuable for research, teaching and knowledge communication. However, the tools used to automate the process are challenging. The purpose of the drawProteins package is to enable the generation of schematics of proteins in an automated fashion that can integrate with the Bioconductor/R suite of tools for bioinformatics and statistical analysis. Using UniProt accession numbers, the package uses the UniProt API to get the features of the protein from the UniProt database. The features are assembled into a data frame and visualized using adaptations of the ggplot2 package. Visualizations can be customised in many ways including adding additional protein features information from other data frames, altering colors and protein names and adding extra layers using other ggplot2 functions. This can be completed within a script that makes the workflow reproducible and sharable.",
"keywords": [
"protein",
"schematic",
"BIOCONDUCTOR",
"R package",
"visualization."
],
"content": "Introduction\n\nProtein schematics are abundant in research papers, reviews, text books and on the internet1. Thus, they represent one of the most common molecular visualizations shown and seen by researchers and students. Constructing protein schematics is often time consuming and is not performed in a reproducible and easily shared manner. The schematics frequently reflect domain expertise, but often also reflect the opinions of an individual researcher in a manner that is not obvious.\n\nThere are solutions in other languages: a Java and JavaScript tool1,2 that can be used for protein visualization. For visualization on the internet, there is also the BioJS solution, which can be used for proteins3. Both of these tools are useful but not easily integrated into the Bioconductor workflow. The GenVisR package contains the option to produce highly customisable publication-quality graphics for genomic data4. The focus on genomic data reduces the usefulness of drawing protein schematics, particularly those illustrating multiple proteins and protein families.\n\nFor these reasons, a protein visualization package was produced using R to allow compatibility with the Bio- conductor suite of bioinformatics packages. It uses the UniProt Proteins API5,6 as a resource of protein features and the ggplot2 package7 as a basis for drawing the schematics. Multiple proteins can be drawn from similar or different families. The only limitation is the availability of UniProt entries.\n\nSchematic customisation is possible. Protein chains, domains, regions, motifs or phosphorylation sites can be drawn separately or together. Colors can be altered and protein names (labels) can be changed. All of this can be done in a scripted manner that facilitates code sharing, visualization reproducibility and good practice in scientific computing8.\n\n\nMethods\n\nThe Bioconductor pacakge, drawProteins, is designed to work with Bioconductor 3.7 and R version 3.5.\n\nThis package has been created to allow the creation of protein schematics based on the data obtained from the Uniprot Protein Database.\n\nThe basic workflow of drawProteins is:\n\n1. To provide one or more Uniprot IDs\n\n2. Get a list of feature from the Uniprot API\n\n3. Draw the chains of the proteins\n\n4. Add features as desired\n\ndrawProteins uses the package httr to interact with the Uniprot API and extract a JSON object into R. The JSON object is used to create a data frame. Adaptations of the graphing package ggplot2 are then used to create the protein schematic.\n\nGetting the data from Uniprot. Currently, drawProteins obtains the protein feature information from the UniProt Protein API5,6. At least one working Uniprot accession number must be provided. More than one can be provided but they must be in the same vector, separated by a space. The space is replaced to create a url that can be used to query the Uniprot API9.\n\nThe get_features() function uses the Uniprot API to return the features of a protein - the chain, domain information and other annotated features such as “repeats” and “motifs”. Post-translational modifications, such as phosphorylations, are also provided by the UniProt API.\n\nThe httr::content() function is then used to extract the content. From the get_features() function, this will provide lists of lists. The length of the parent lists corresponds to the number of accession numbers provided. Interestingly, the order is different to that of the UniProt accession numbers provided. The lists inside the parent list are a list of six, one for each protein, that contains names of the proteins and the features.\n\nAs an example, the script below will retrieve, from UniProt, the details of a the human version of a protein called RelA or NF-kappaB, p65, a well-studied transcription factor10.\n\nWith internet access, this can be retrieved from Uniprot with this code:\n\n\n\n\n\nTurning Uniprot data into a dataframe. The next step in the workflow is to convert the data from the Uniprot API into a dataframe that can be used with ggplot2.\n\nThe feature_to_dataframe() function will convert the list of lists of six provided by the get_features() function to a dataframe, which can then be used to plot the schematics.\n\nThe feature_to_dataframe() function will also add an “order” value to allow plotting. The order goes from the bottom in the manner of a graph.\n\n\n\nThe rel_data object is a data frame with 9 variables and observations that include protein features. The variables are show below. A data frame of this type could be created independently of UniProt.\n\n\n\n\n\nDraw the canvas, protein chains and domains. The first step is to create the plot area with the draw_canvas() function. The x-axis of the canvas is based on the length of the protein (or the longest protein in the case of drawing multiple proteins). The y-axis is based on the number of proteins being drawn. The draw_canvas() function requires a data frame.\n\nUsually, the next step is to draw the protein chains using the draw_chains() function. This requires a ggplot2 object and a data frame in that order. The data frame does not have to be the same as that used for draw_canvas() but must contain the variables type, description, begin, end, entryName, order.\n\nProtein domains can be added with the draw_domains() function, which also requires a ggplot2 object and a data frame in that order. Again, the data frame does not have to be the same as that obtained from UniProt but must contain the variables type, description, begin, end, and order. Thus custom domains can be added with the draw_domains() function. Note that the chain and the domain are drawn to scale in terms of their number of amino acids (Figure 1).\n\n\n\nThe default output gives a grey background and labels the domain. RHD = Rel Homology Domain.\n\nTo show this visualization better, a white background helps, as does removing the y-axis and the grid (Figure 2). Changing the size of the text using the base_size argument also aids visualization. This can be done with this code:\n\n\n\nThe background can be customized using theme functions from ggplot2. RHD = Rel Homology Domain.\n\n\nUse cases\n\nThe UniProt API provides information on protein regions, protein motifs and protein phosphorylation sites. By using the functions shown in the script below, it is possible to show the features of the protein desired to create. Altering colors and adding customisation is possible.\n\nFor the human protein RelA, also known as the p65 subunit of NFkappaB, a transcription factor with diverse functions including a role in leukaemia, inflammation and cancer, here is a good workflow that generates a nice schematic of the protein showing domains and phosphorylation sites (Figure 3).\n\n\n\nBy drawing the domains, regions and motifs a more detailed protein schematic is generated. RHD = Rel Homology Domain; TAD = Transactivation Domain. Yellow circles denote phosphorylation sites.\n\nWith internet access, the script below shows the workflow and generates a visualization of the five human proteins of the NF-kappaB transcription factor family (Figure 4).\n\n\n\n\n\n\n\nThe five members of the NF-kappaB transcription factors family can be illustrated by drawing the domains, regions and motifs as detailed on the UniProt database. The lengths of the chains, domains and motifs are proportional to the number of amino acids. RHD = Rel Homology Domain; TAD = Transactivation Domain. Yellow circles denote phosphorylation sites.\n\nThe proteins, domains and phosphorylation sites (yellow circles) are drawn and positioned according to amino acid number.\n\nIt is possible to use bioMart11,12 to pull out the UniProt accession numbers for a Gene Ontology (GO) term. For example, the GO term for “MAP kinase activity”. This has a GO number of GO:0004707. This example script borrows heavily on the biomaRt users guide written by Steffen Durinck, Wolfgang Huber and Mike Smith13. The script below generates a visualization containing 14 protein schematics (Figure 5).\n\n\n\n\n\n\n\n\n\nUsing bioMart with the Gene Ontology term for \"MAP kinase activity\", it is possible to draw multiple human MAP kinases using data from UniProt. Yellow circles denote phosphorylation sites.\n\nVarious customizations are possible:\n\n1. Alter chain color and outline.\n\n2. Change the labels to a custom list (but remember that the plots are drawn from the bottom up)\n\n3. Change the size and color of the phosphorylation symbols.\n\nThese are illustrated with the script below which generates Figure 6.\n\n\n\nUsing arguments in the draw_chains() and draw_phospho() functions, it is possible customize colors and labels.\n\n\nDiscussion\n\nAfter 20 years of manual drawing of protein schematics and experience in proteomic studies14–17, the need for a more sustainable and programmatic method seemed important and worthwhile. It seemed wise to develop an approach that would integrate protein visualizations with other bioinformatic tools available in Bioconductor. This package represents an approach to enabling the reproducible and programmatic generation of protein schematics.\n\nThe plan is to develop this package further in terms of generating use cases and adding features. A list of issues for future development has already been added by the author on GitHub. Bug reports, feedback on desired features or code contributions can be made through GitHub.\n\nThe challenge with protein visualization is that specialist domain knowledge sometimes trumps databases. Thus, while integrating knowledge from UniProt is an excellent starting point, it is also essential to allow customisation of plots. This can be done by adding or removing information about the proteins, protein features and post-translational modification to the dataframe object made with R.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nThe drawProteins package is available at: http://bioconductor.org/packages/drawProteins/\n\nSource code is available at: https://github.com/brennanpincardiff/drawProteins\n\nArchived source code as at time of publication: https://github.com/brennanpincardiff/drawProteins/tree/v1.0.2 and https://doi.org/10.5281/zenodo.130661918\n\nLicense: MIT+",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nPB has been supported by funding from Bloodwise, UK.\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\nThanks to Steph Locke (https://github.com/stephlocke) and Dave Parr (https://github.com/DaveRGP) from the CaRdiff R User Group for help, advice and support that led to the development of this package. Without them the package just wouldn’t have happened.\n\nSome of the Use Cases above have been shown on the authors blog, R for Biochemists (http://rforbiochemists.blogspot.co.uk/).\n\n\nReferences\n\nRen J, Wen L, Gao X, et al.: DOG 1.0: illustrator of protein domain structures. Cell Res. 2009; 19(2): 271–3. PubMed Abstract | Publisher Full Text\n\nLiu W, Xie Y, Ma J, et al.: IBS: an illustrator for the presentation and visualization of biological sequences. Bioinformatics. 2015; 31(20): 3359–3361. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorpas M: The BioJS article collection of open source components for biological data visualisation [version 1; referees: not peer reviewed]. F1000Res. 2014; 3: 56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSkidmore ZL, Wagner AH, Lesurf R, et al.: GenVisR: Genomic Visualizations in R. Bioinformatics. 2016; 32(19): 3012–3014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThe UniProt Consortium: Uniprot: the universal protein knowledgebase. Nucleic Acids Res. 2017; 45(D1): D158–D169. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNightingale A, Antunes R, Alpi E, et al.: The Proteins API: accessing key integrated protein and genome information. Nucleic Acids Res. 2017; 45(W1): W539–W544. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2009. Reference Source\n\nWilson G, Bryan J, Cranston K, et al.: Good enough practices in scientific computing. PLoS Comput Biol. 2017; 13(6): e1005510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThe uniprot proteins api. 2018. Reference Source\n\nLawrence T: The nuclear factor NF-kappaB pathway in inflammation. Cold Spring Harb Perspect Biol. 2009; 1(6): a001651. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDurinck S, Moreau Y, Kasprzyk A, et al.: BioMart and Bioconductor: a powerful link between biological databases and microarray data analysis. Bioinformatics. 2005; 21(16): 3439–3440. PubMed Abstract | Publisher Full Text\n\nDurinck S, Spellman PT, Birney E, et al.: Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009; 4(8): 1184–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith M, Durinck S, Huber W: The biomart users guide. 2017. Reference Source\n\nBrennan P, Babbage JW, Thomas G, et al.: p70s6k integrates phosphatidylinositol 3-kinase and rapamycin-regulated signals for E2F regulation in T lymphocytes. Mol Cell Biol. 1999; 19(7): 4729–4738. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrennan P, Floettmann JE, Mehl A, et al.: Mechanism of action of a novel latent membrane protein-1 dominant negative. J Biol Chem. 2001; 276(2): 1195–1203. PubMed Abstract | Publisher Full Text\n\nFielding CA, Siebert S, Rowe M, et al.: Analysis of human tumour necrosis factor receptor 1 dominant-negative mutants reveals a major region controlling cell surface expression. FEBS Lett. 2004; 570(1–3): 138–142. PubMed Abstract | Publisher Full Text\n\nAlsagaby SA, Khanna S, Hart KW, et al.: Proteomics-based strategies to identify proteins relevant to chronic lymphocytic leukemia. J Proteome Res. 2014; 13(11): 5051–5062. PubMed Abstract | Publisher Full Text\n\nBrennan P: brennanpincardiff/drawProteins: F1000 publication release (Version v1.0.2). Zenodo. 2018. Data Source"
}
|
[
{
"id": "37074",
"date": "22 Aug 2018",
"name": "Matej Lexa",
"expertise": [
"Reviewer Expertise bioinformatics",
"plant biology and genomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper describes a protein annotation visualization tool for R/Bioconductor that has already been in use for more than a year. The software is mature, I was able to install it easily, all the examples of code in the manuscript are functional. I only noticed a small deviation in the biomaRt example, most likely caused by slightly different data returned by biomaRt.\nThe manuscripts describes the purpose and functioning of the package in sufficient detail. However, I found the introduction a bit too brief, especially in terms of mentioning other similar software and advantages of drawProteins. For example, I found a Bioconductor package called Pviz that may be able to generate protein annotation visualizations.\nAs a minor remark, the last sentence in this paragraph on page 3 seems unclear to me: \"The feature_to_dataframe() function will also add an “order” value to allow plotting. The order goes from the bottom in the manner of a graph.\"\nFinally, a suggestion. The software would be even more useful if it could handle reading and writing of standard protein annotation files in GFF or similar format. UniProt wikipedia page describes how it is applied to proteins in this example: https://www.uniprot.org/uniprot/P0A7B8.gff Although GFF is much more popular with genomic data, it seems to be ocassionally used for annotating protein sequences.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3935",
"date": "31 Aug 2018",
"name": "Paul Brennan",
"role": "Author Response",
"response": "Dear Matej, Thanks for your review. I will look at extending the introduction over the next little while and also clarifying my remark about the order of the proteins. I was inspired by your suggestion about GFF formats. I have created a gff_parser() function that is now available on Github and in the development version of drawProteins. It will be interesting to see if this is found useful. Interestingly for RelA/p65, the GFF format only includes annotation about the protein folding and not annotation about the domains. For that reason, I have also created a draw_folding() function which colours the regions of the protein according to STRAND, HELIX and TURN types which denote regions of the proteins that assemble as beta-strands, alpha helicies or make a turn in the 3D structure of the protein. Thanks again, Paul ---- Paul Brennan, Cardiff University"
}
]
},
{
"id": "37075",
"date": "22 Aug 2018",
"name": "Malachi Griffith",
"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 author describes the creation of a Bioconductor/R package for generic visualization of 2D protein diagrams. The tool does not address 3D visualizations or visualization of protein sequences or alignments. Rather it allows the user to view the position of protein annotations within an arbitrary number of proteins. By default, UniProt sequence identifiers and associated annotations are supported. The demonstrated examples depict protein domains/motifs and phosphorylation sites. Identifying a set of proteins to view from the Gene Ontology via bioMart is also demonstrated. There are likely to be many additional useful queries that could be achieved by similar bioMart queries.\n\nThe package leans on ggplot2 functionality for the actual drawing of each protein schematic. The drawings consist of simple elements: colored rectangles and circles that can be nested, overlaid, and labeled. A few simple use cases are illustrated. The writing is brief and straightforward. The code is available under an open access document.\n\nI was able to install the package and run through the examples in the paper in a few minutes with only minor difficulties.\n\nIn particular, I had trouble getting this line to work in the example: ``` output <- getBM(attributes = c('uniprotswissprot', 'hgnc_symbol'), filters = 'go', values = 'GO:0004707', mart = ensembl) ```\n\nPerhaps some problem with the way the single quotes are being encoded in that section?\n\nOverall the tool is quite basic but also appears to be customizable. I anticipate it could be useful to many who wish to automate visualization of protein drawings and integrate these with visualization of various types omics data.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1105
|
https://f1000research.com/articles/7-327/v1
|
15 Mar 18
|
{
"type": "Research Article",
"title": "Association of fat mass profile with natriuretic peptide receptor alpha in subcutaneous adipose tissue of medication-free healthy men: A cross-sectional study",
"authors": [
"Petros C. Dinas",
"Eleni Nintou",
"Dimitra Psychou",
"Marnie Granzotto",
"Marco Rossato",
"Roberto Vettor",
"Athanasios Z. Jamurtas",
"Yiannis Koutedakis",
"George S. Metsios",
"Andreas D. Flouris",
"Petros C. Dinas",
"Eleni Nintou",
"Dimitra Psychou",
"Marnie Granzotto",
"Marco Rossato",
"Roberto Vettor",
"Athanasios Z. Jamurtas",
"Yiannis Koutedakis",
"George S. Metsios"
],
"abstract": "Background: Atrial natriuretic peptide increases lipolysis in human adipocytes by binding to natriuretic peptide receptor-A (NPRA). The aim of the current study was to examine the associations of NPRA mRNA of subcutaneous adipose tissue with fat mass, fat-free mass, body mass index (BMI) and arterial blood pressure in medication-free healthy men. Method: Thirty-two volunteers [age (years): 36.06±7.36, BMI: 27.60±4.63 (kg/m2)] underwent assessments of body height/weight, % fat mass, fat-free mass (kg), blood pressure, and a subcutaneous adipose tissue biopsy via a surgical technique. Results: We found that NPRA mRNA was negatively associated with % fat mass (r=-0.40, R2=0.16, p=0.03) and BMI (r=-0.45, R2=0.20, p=0.01). Cohen’s f2 effect size analyses showed a small effect size between NPRA mRNA and BMI (f2=0.25). One-way analysis of variance with Bonferroni post-hoc tests showed a tendency for mean differences of NPRA mRNA across BMI categories (p=0.06). This was confirmed by Cohen’s d effect size analyses revealing a large effect size of NPRA mRNA between obese individuals (BMI≥30 kg/m2) and either normal weight (BMI=19-25 kg/m2; d=0.94) or overweight (BMI=25-30 kg/m2; d=1.12) individuals. Conclusions: NPRA mRNA is negatively associated with % fat mass and BMI in medication-free healthy men, suggesting a possible role of NPRA in the control of fat mass accumulation.",
"keywords": [
"NPRA",
"atrial natriuretic peptide",
"lipolysis",
"BMI",
"obesity"
],
"content": "Introduction\n\nAtrial natriuretic peptide (ANP) lowers arterial pressure to maintain fluid volume homeostasis, thus protecting against renal and cardiac pathogenesis1. ANP also increases lipolysis in human adipocytes2 by binding to natriuretic peptide receptor-A (NPRA)3. NPRA is less expressed in subcutaneous adipose tissue (SAT) in obese individuals and type 2 diabetes patients than in normal glucose tolerant individuals4. Also, NPRA signalling in skeletal muscle may enhance long-term insulin sensitivity5. Collectively, NPRA may potentially treat obesity-related disorders while ANP may play a role in the therapeutic mechanisms of beta-adrenoceptor antagonists in the mitigation of heart dysfunction and utilization of lipid mobilization6. However, the role of ANP in lipolysis has been primarily investigated mainly in vitro models7–9, in human blood cells from individuals under medication treatment10, and in animal models9. To our knowledge, no such information is currently available in relation to the role of its receptor (NPRA) on the adipocytes of healthy individuals. Therefore, the aim of the current study was to examine the associations of NPRA mRNA of SAT with fat mass, fat-free mass (FFM), body mass index (BMI) and arterial blood pressure (BP) in medication-free healthy men.\n\n\nMethods\n\nThe study was approved by the Ethics Committee of the University of Thessaly (protocol no. 698/2013). The inclusion criteria were: healthy adult men, non-smokers, no chronic disease and/or being under medication treatment. The participants were recruited by advertisements in a local newspaper in Trikala, Thessaly, Greece and the data collection started in July 2013 and ended in June 2014. Written consent was obtained from the 32 healthy men recruited [age (years): 36.06±7.36, BMI: 27.60±4.63 (kg/m2)].\n\nTo avoid misleading results, the participants refrained from exercise, alcohol, and passive smoking 72h prior the measurements, while they followed an overnight fast before they visit the physiology laboratory in the School of Exercise Science between 07:00–09:00 am. PCD and DP performed the following measurements: body height using a Seca 220 (Hamburg, Germany) stadiometer, body weight using a precision scale (KERN & Sohn GmbH, Version 5.3, Germany) and blood pressure using an Aneroid sphygmomanometer. Fat mass percentage (%FM) and FFM were measured via bioelectrical impedance using a body composition monitor (Fresenius Medical Care AG & Co. KGaA D-61346 Bad Hamburg, Germany).\n\nFollowing the aforementioned measurements, the participants underwent a SAT biopsy in the physiology laboratory by a trained physician, as previously described11. Briefly, the site of the incision was disinfected and a 10 ml of 2% xylocaine (no adrenaline) was injected for local anaesthesia. An incision of 2–2.5 cm was made 3–5 cm to the left of the navel. Nearly 500 mg of subcutaneous adipose tissue was captured and removed. The NPRA mRNA analysis of SAT samples is described elsewhere12. Briefly, total RNA was extracted using RNeasy Lipid Tissue mini kit (QIAGEN). First-strand cDNAs were synthesized from equal amounts of total RNA using random primers and M-MLV reverse transcriptase (Promega). Quantitative real-time polymerase chain reaction was performed using Sybr Green fluorophore. 18S rRNA gene was used as a reference for normalization.\n\nFollowing previous methodology, we removed two mean values (i.e. outliers) of NPRA mRNA that were at a distance of more than two standard deviations from the mean of the distribution13,14. Also, there were three missing values in the NPRA mRNA analysis of SAT samples due to failure to extract RNA from adipose tissue. Eventually, 27 NPRA mRNA values were included in the statistical analysis using SPSS (version 24; SPSS Inc., Chicago, IL, USA). Normal distribution was determined using Shapiro-Wilk test, whereas Pearson’s correlation coefficient, linear regression, and Cohen’s f2 effect size (R2/1-R2)15 were used to detect associations between NPRA mRNA, %FM, FFM (kg), BMI, and BP. We also used one-way analysis of variance (ANOVA) with Bonferroni post-hoc tests, and Cohen’s d effect size analyses to explore the mean differences of NPRA mRNA across different BMI categories [normal weight <25 kg/m2 (n=9); overweight 25–30 kg/m2 (n=9); obese >30 kg/m2 (n=9)]. The level of significance was set at p≤0.05.\n\n\nResults\n\nThe participants’ characteristics are provided in Table 1. The NPRA mRNA was negatively correlated with %FM (r=-0.40, p=0.03) (Figure 1) and BMI (r=-0.45, p=0.01) (Figure 2). No associations were found between NPRA mRNA and FFM, systolic or diastolic BP (p>0.05).\n\nLinear regression analyses confirmed the associations between NPRA mRNA and %FM (R2=0.16, p=0.03) as well as BMI (R2=0.20, p=0.01). Cohen’s f2 effect size analyses showed a small effect size between NPRA mRNA and BMI (f2=0.25), however, no effect size was detected between NPRA mRNA and %FM (f2<0.20). ANOVA demonstrated a strong tendency for mean differences in NPRA mRNA across BMI categories (p=0.06). This was confirmed by Cohen’s d effect size analyses in NPRA mRNA, revealing large effect sizes between obese individuals (BMI≥30 kg/m2) and either normal weight (BMI <25 kg/m2; d=0.94) or overweight (BMI=25–30 kg/m2; d=1.12) individuals.\n\nBMI: Body mass index\n\n\nDiscussion and conclusions\n\nWe have shown that the NPRA mRNA is negatively associated with %FM and BMI in medication-free healthy men and that it is expressed less in obese compared to lean individuals. Previous evidence showed that NPRA mRNA is expressed less in normal glucose tolerant individuals than in type 2 diabetes patients4, while it is positively associated with insulin sensitivity4. Given that insulin sensitivity is negatively associated with excessive FM in humans16,17 the inverse association of NPRA mRNA with %FM and BMI observed in the current study suggests a possible role of NPRA in lowering FM in humans. Indeed, it has been established that natriuretic peptides by binding to NPRA, increase cyclic guanosine monophosphate – a well-known intracellular second messenger – which phosphorylates protein kinase G leading to activation of hormone sensitive lipase18,19. This process mediates triglyceride degradation (i.e. lipolysis), which subsequently increases fatty acid availability18. Furthermore, findings in mice showed that the lack of NPRA gene may increase FM9.\n\nThe current study may be affected by methodological limitations such as the lack of insulin sensitivity measurements and a priori power calculation to determine the sample size. However, a post-measurements power calculation was conducted using an online software (DSS Research) to test statistical power. This revealed 89% of statistical power for the 27 available samples, based on the NPRA mRNA value (1.02±0.38) we detected in our study and expected NPRA mRNA value (0.81±0.08) from a previous similar study that examined NPRA in SAT in humans4.\n\nIn conclusion, NPRA mRNA is negatively associated with %FM and BMI in medication-free healthy men, suggesting a possible role of ANP/NPRA axis in the control of FM accumulation. To date, the investigation of NPRA has mainly focused either on circulating and muscle NPRA20–22 or on medication-dependent NPRA measurements4,10. Our study indicates that NPRA may also play role in FM profile of healthy individuals, which should be further explored in a cause-and-effect research setting.\n\n\nData availability\n\nDataset 1: Subcutaneous adipose tissue NPRA mRNA of medication-free healthy men 10.5256/f1000research.14198.d19769423\n\nBMI=Body mass index; SBP=Systolic blood pressure; DBP=Diastolic blood pressure; FFM=Fat-free mass; NPRA= Natriuretic peptide receptor-A.\n\nBMI categories= 1. <25 kg/m2, 2. 25–30 kg/m2, 3. >30 kg/m2.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by funding from the European Union 7th Framework Program (FP7-PEOPLE-2012-IRSES grant no. 319010; FP7-PEOPLE-2013-IRSES grant no. 612547).\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\nTokudome T, Horio T, Kishimoto I, et al.: Calcineurin-nuclear factor of activated T cells pathway-dependent cardiac remodeling in mice deficient in guanylyl cyclase A, a receptor for atrial and brain natriuretic peptides. Circulation. 2005; 111(23): 3095–3104. PubMed Abstract | Publisher Full Text\n\nSengenès C, Berlan M, De Glisezinski I, et al.: Natriuretic peptides: a new lipolytic pathway in human adipocytes. FASEB J. 2000; 14(10): 1345–1351. PubMed Abstract | Publisher Full Text\n\nKumar P, Bolden G, Arise KK, et al.: Regulation of natriuretic peptide receptor-A gene expression and stimulation of its guanylate cyclase activity by transcription factor Ets-1. Biosci Rep. 2009; 29(1): 57–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKovacova Z, Tharp WG, Liu D, et al.: Adipose tissue natriuretic peptide receptor expression is related to insulin sensitivity in obesity and diabetes. Obesity (Silver Spring). 2016; 24(4): 820–828. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoué M, Badin PM, Vila IK, et al.: Defective Natriuretic Peptide Receptor Signaling in Skeletal Muscle Links Obesity to Type 2 Diabetes. Diabetes. 2015; 64(12): 4033–4045. PubMed Abstract | Publisher Full Text\n\nLafontan M, Moro C, Berlan M, et al.: Control of lipolysis by natriuretic peptides and cyclic GMP. Trends Endocrinol Metab. 2008; 19(4): 130–137. PubMed Abstract | Publisher Full Text\n\nSengenes C, Bouloumie A, Hauner H, et al.: Involvement of a cGMP-dependent pathway in the natriuretic peptide-mediated hormone-sensitive lipase phosphorylation in human adipocytes. J Biol Chem. 2003; 278(49): 48617–48626. PubMed Abstract | Publisher Full Text\n\nSengenès C, Zakaroff-Girard A, Moulin A, et al.: Natriuretic peptide-dependent lipolysis in fat cells is a primate specificity. Am J Physiol Regul Integr Comp Physiol. 2002; 283(1): R257–265. PubMed Abstract | Publisher Full Text\n\nBordicchia M, Liu D, Amri EZ, et al.: Cardiac natriuretic peptides act via p38 MAPK to induce the brown fat thermogenic program in mouse and human adipocytes. J Clin Invest. 2012; 122(3): 1022–1036. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoro C, Polak J, Hejnova J, et al.: Atrial natriuretic peptide stimulates lipid mobilization during repeated bouts of endurance exercise. Am J Physiol Endocrinol Metab. 2006; 290(5): E864–869. PubMed Abstract | Publisher Full Text\n\nChachopoulos V, Dinas PC, Chasioti M, et al.: A Technique for Subcutaneous Abdominal Adipose Tissue Biopsy via a Non-diathermy Method. J Vis Exp. 2017; (127). PubMed Abstract | Publisher Full Text\n\nDinas PC, Valente A, Granzotto M, et al.: Browning formation markers of subcutaneous adipose tissue in relation to resting energy expenditure, physical activity and diet in humans. Horm Mol Biol Clin Investig. 2017; 31(1): pii: /j/hmbci.2017.31.issue-1/hmbci-2017-0008/hmbci-2017-0008.xml. PubMed Abstract\n\nRatcliff R: Methods for dealing with reaction time outliers. Psychol Bull. 1993; 114(3): 510–532. PubMed Abstract | Publisher Full Text\n\nDinas PC, Nikaki A, Jamurtas AZ, et al.: Association between habitual physical activity and brown adipose tissue activity in individuals undergoing PET-CT scan. Clin Endocrinol (Oxf). 2015; 82(1): 147–154. PubMed Abstract | Publisher Full Text\n\nCohen J: Statistical Power Analysis for the Behavioral Sciences. 2nd edn; 1988. Reference Source\n\nSjöström LV: Mortality of severely obese subjects. Am J Clin Nutr. 1992; 55(2 Suppl): 516S–523S. PubMed Abstract\n\nPouliot MC, Després JP, Nadeau A, et al.: Visceral obesity in men. Associations with glucose tolerance, plasma insulin, and lipoprotein levels. Diabetes. 1992; 41(7): 826–834. PubMed Abstract | Publisher Full Text\n\nLafontan M, Langin D: Lipolysis and lipid mobilization in human adipose tissue. Prog Lipid Res. 2009; 48(5): 275–297. PubMed Abstract | Publisher Full Text\n\nRukavina Mikusic NL, Rosón MI, Kouyoumdzian NM, et al.: Natriuretic Peptide Receptor Type A (NPRA). In: Encyclopedia of Signaling Molecules. Edited by Choi S. Cham: Springer International Publishing; 2018; 3344–3351. Publisher Full Text\n\nBuglioni A, Cannone V, Cataliotti A, et al.: Circulating aldosterone and natriuretic peptides in the general community: relationship to cardiorenal and metabolic disease. Hypertension. 2015; 65(1): 45–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiyashita K, Itoh H, Tsujimoto H, et al.: Natriuretic peptides/cGMP/cGMP-dependent protein kinase cascades promote muscle mitochondrial biogenesis and prevent obesity. Diabetes. 2009; 58(12): 2880–2892. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang TJ, Larson MG, Levy D, et al.: Impact of obesity on plasma natriuretic peptide levels. Circulation. 2004; 109(5): 594–600. PubMed Abstract | Publisher Full Text\n\nDinas PC, Nintou E, Psychou D, et al.: Dataset 1 in: Association of fat mass profile with natriuretic peptide receptor alpha in subcutaneous adipose tissue of medication-free healthy men. F1000Research. 2018. Data Source"
}
|
[
{
"id": "33152",
"date": "24 Apr 2018",
"name": "Marcelo Roberto Choi",
"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 present work by Dinas et al. is a descriptive study that investigated the association of NPRA mRNA of subcutaneous adipose tissue (SAT) with fat mass, fat-free mass, body mass index (BMI) and arterial blood pressure in thirty-two medication-free healthy volunteers. The authors observed a negative association between NPRA mRNA with %fat mass and BMI. Based on these results the authors suggest a possible role of ANP-NPRA axis in the control of FM accumulation. Although the topic is of interest, some issues need to be addressed in order to improve the quality of the manuscript.\nGiven that NPRA plays a role in lipolysis, it would have been interesting if the authors could determine the levels of triglycerides and cholesterol as well as ANP plasma levels in order to verify a possible correlation between these parameters and the percentage of fat and/or NPRA mRNA levels in SAT. This issue should be considered in the discussion section and mentioned as a limitation of this study.\n\nPlease provide more details about blood pressure registers (for example: number of measures, guideline used to define hypertension, brand of the device used).\n\nYu et al.1 suggested that adipocyte size is an important determinant of ANP-stimulated lipolysis. In this way, expressed higher mRNA levels of receptor (NPR)-A and hormone sensitive lipase as well as more NPR-A on the membrane than small It would have been interesting if the authors had determined the NPRA mRNA according to the size of the adipocyte. This issue should be considered in the discussion section and mentioned as a limitation of this study. (see also Laurencikiene et al.2)\n\nIt has been reported that obesity is associated with low circulating levels of BNP and recent evidences suggest an altered expression of BNP receptors-both the signaling receptors (NPR)-A and the clearance NPR-C receptor-in Pivovarova et al.3; Gentili et al.4). Moreover, increased tissue secretion of adipokines and cytokines has been implicated in the chronic low-grade state associated with obesity (Schindler et al.5; Moro et al.6). This issue should be considered in the discussion section and mentioned as a limitation of this study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3827",
"date": "18 Jul 2018",
"name": "Petros Dinas",
"role": "Author Response",
"response": "We thank you very much for your comments. Please see below our responses to your comments: Comment 1: Thank you for this suggestion. Indeed, this would have been a very interesting analysis to perform. Unfortunately, there are no samples available at this stage to perform the suggested work. Therefore, we have included this information in the limitations of the Discussion section (Page 6, paragraph 2). Comment 2: We have added further details about blood pressure registers and we provide a relevant reference (24) in methods section (Page 4, paragraph 1). Comment 3: As suggested, we have included this information in the limitations of the Discussion section and we provide the relevant reference (25). (Page 6, paragraph 2). Comment 4: As suggested, we have added this information in the limitations along with the relevant references (26,27, 28,29). (Page 6, paragraph 2)."
}
]
},
{
"id": "34721",
"date": "25 Jun 2018",
"name": "Kailash N. Pandey",
"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 manuscript by P.C. Dinas et al., describes the association of fat mass profile with the mRNA levels of natriuretic peptide receptor-A (NPRA) in the subcutaneous adipose tissues (SAT) of medication-free healthy individuals. The major goal of the present study was to determine the association of NPRA mRNA in SAT with fat mass (FM), fat-free mass (FFM), body mass index (BMI), and arterial blood pressure in 32 volunteers with average age of 36 years. The subcutaneous adipose tissues biopsy was done by surgical procedures. The authors report that NPRA mRNA was negatively associated with % FM as well as BMI. Based on the present findings, the authors suggest that NPRA signaling mechanism might play a critical role in the control of fat mass accumulation. The results of the current study are interesting and critical in understanding the versatility of the NPRA signaling mechanisms in the health and disease; however, there are some concerns, which need to be considered to improve the contents of this manuscript.\nThe present study establishes a link between NPRA mRNA, % FM, and BMI. The results indicate that atrial and brain natriuretic peptides (ANP, BNP) and their cognate receptor NPRA might play a vital role in lipolysis. Thus, it could have been a logical extension to determine the hormone-sensitive lipases. Although, ANP and BNP have been shown to act as the lipolytic hormones and also to affect energy use and metabolism in adipocytes, thus their actions seem to be associated with hormonal stimulation of lipases (Sengenes et al., 20031; Lafontan et al., 20082). It would be helpful that these issues and limitation might be discussed in the manuscript.\n\nThe ethnicity of the Greek subjects under the study should be indicated and if there is any limitation that could be mentioned. The study is based on the small number of subjects and the limitations might also be indicated.\n\nThe cGMP- dependent protein kinase 1 (CGK1) phosphorylates perilipin-1, a hormonal-regulated lipase that initiates lipolysis (Sengenes et al., 2003; Lafontan et al., 20053). Nonetheless, it has been suggested that the potent lipolytic function of ANP and BNP seems to be restricted to primates (Sengenes et al., 20024; Engeli et al., 20125). Intriguing was the finding that a lower concentrations of ANP and BNP seem to be associated with obesity, insulin resistance, and metabolic consequences, thereby, ANP-BNP/NPRA/cGMP axis might regulate fat oxidation to prevent obesity and glucose intolerance (Wang et al., 20046; Wang et al., 20077; Mitsuishi et al., 20088; Miyashita et al., 20099). These paradigms and issues of the ANP/NPRA/cGMP signaling could be discussed.\n\nInterestingly, ANP-BNP/NPRA signaling axis has been shown to induce a browning of white adipocytes in humans, which seems to be physiologically significant (Enerback, 201010; Collins et al., 201411; Liu et al., 201812). A brief discussion on this aspect of ANP/NPRA signaling mechanisms might be helpful.\n\nThe previous findings have suggested that NPRA acts as a determinant of insulin sensitivity; however, the upregulation of natriuretic peptide receptor-C (NPRC) decreases the glucose tolerance in obase subjects, which seems to repress ANP-BNP/NPRA signaling axis and thereby the lipolytic effects of ANP was completely rescued in Npr3 (coding for NPRC) gene-knockout mice (Bordicchia et al., 201213). The significance of these previous findings and the relationship to the current work could be discussed in the current manuscript.\n\nThe conclusions of the present study are based on the findings in male gender. The limitations on the inclusion of female gender should be mentioned.\n\nI have read and reviewed the manuscript and believe that I have the expertise to evaluate and state that the current work is of an acceptable scientific quality. Indeed, there are some concerns in the manuscript, which are summarized above.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3826",
"date": "18 Jul 2018",
"name": "Petros Dinas",
"role": "Author Response",
"response": "We thank you very much for your comments. Please see below our responses to your comments: Comment 1: Thank you for this suggestion. Indeed, this would have been an interesting analysis to perform. Unfortunately, there are no samples available at this stage to perform the suggested work. Therefore, we have included in the limitations (Page 6, paragraph 2) the lack of hormone sensitive lipase measurements. Comment 2: We have added this information in the limitations (Page 6, paragraph 2). Comment 3: The issue regarding the obesity prevention via ANP/BNP axis has been added in the Discussion section (Page 6, paragraph 1). Insulin sensitivity measurements in the current study has also been indicated as a limitation. (Page 6, paragraph 2). Comment 4: We have added this information at the end of the first paragraph of the Discussion section (Page 6, paragraph 1). Comment 5: We have included a brief discussion on these topics in the first paragraph of the Discussion section (Page 6, paragraph 1). Comment 6: We have added this information in the limitations (Page 6, paragraph 2)."
}
]
}
] | 1
|
https://f1000research.com/articles/7-327
|
https://f1000research.com/articles/7-1101/v1
|
18 Jul 18
|
{
"type": "Brief Report",
"title": "Real time portable genome sequencing for global food security",
"authors": [
"Laura Boykin",
"Ammar Ghalab",
"Bruno Rossitto De Marchi",
"Anders Savill",
"James M. Wainaina",
"Tonny Kinene",
"Stephen Lamb",
"Myriam Rodrigues",
"Monica Kehoe",
"Joseph Ndunguru",
"Fred Tairo",
"Peter Sseruwagi",
"Charles Kayuki",
"Deogratius Mark",
"Joel Erasto",
"Hilda Bachwenkizi",
"Titus Alicai",
"Geoffrey Okao-Okuja",
"Phillip Abridrabo",
"Emmanuel Ogwok",
"John Francis Osingada",
"Jimmy Akono",
"Elijah Ateka",
"Brenda Muga",
"Samuel Kiarie",
"Ammar Ghalab",
"Bruno Rossitto De Marchi",
"Anders Savill",
"James M. Wainaina",
"Tonny Kinene",
"Stephen Lamb",
"Myriam Rodrigues",
"Monica Kehoe",
"Joseph Ndunguru",
"Fred Tairo",
"Peter Sseruwagi",
"Charles Kayuki",
"Deogratius Mark",
"Joel Erasto",
"Hilda Bachwenkizi",
"Titus Alicai",
"Geoffrey Okao-Okuja",
"Phillip Abridrabo",
"Emmanuel Ogwok",
"John Francis Osingada",
"Jimmy Akono",
"Elijah Ateka",
"Brenda Muga",
"Samuel Kiarie"
],
"abstract": "Crop losses due to viral diseases and pests are major constraints on food security and income for millions of households in sub-Saharan Africa (SSA). Such losses can be reduced if plant diseases and pests are correctly diagnosed and identified early. Currently, accurate diagnosis for definitive identification of plant viruses and their vectors in SSA mostly relies on standard PCR and next generation sequencing technologies (NGS). However, it can take up to 6 months before results generated using these approaches are available. The long time taken to detect or identify viruses impedes quick, within-season decision-making necessary for early action, crop protection advice and disease control measures by farmers. This ultimately compounds the magnitude of crop losses and food shortages suffered by farmers. The MinION portable pocket DNA sequencer was used, to our knowledge globally for the first time, to sequence whole plant virus genomes. We used this technology to identify the begomoviruses causing the devastating cassava mosaic virus, which is ravaging smallholder farmers’ crops in sub-Saharan Africa.",
"keywords": [
"cassava",
"uganda",
"kenya",
"tanzania",
"nanopore",
"minion",
"SDG2"
],
"content": "Introduction\n\nThe United Nations has listed Zero Hunger as one of the 17 global sustainable development goals to end extreme poverty by 2030. Plant viruses are a major constraint to crop production globally, causing an estimated $30 billion in damage1 and leaving millions of people food-insecure2. In Africa, agriculture employs up to 50% of the workforce, yet only contributes 15% to the GDP on average3, suggesting that there is low productivity and limited value addition. This can be addressed through continued innovation in the fields of science and technology, as suggested in the Science Agenda for Agriculture in Africa (S3A)4. Sustainable management of plant viruses and their associated vectors must include efficient diagnostics for surveillance, detection and identification to inform disease management, including the development and strategic deployment of virus resistant varieties. To date, researchers have been utilizing conventional methods such as; PCR, qPCR, high-throughput sequencing (RNA-Seq, DNA-Seq) and Sanger sequencing for pathogen identification. However, these methods are both costly and time consuming, delaying timely control actions. The emergence of new tools for real-time diagnostics, such as the Oxford Nanopore MinION, have recently proven useful for the early detection of Ebola5 and Zika6,7, even in poorly resourced laboratories. For the first time globally, the MinION portable pocket DNA sequencer was used to sequence whole plant virus genomes. We used this technology to identify the begomoviruses causing the devastating cassava mosaic virus, which is ravaging smallholder farmers’ crops in sub-Saharan Africa. Cassava, a carbohydrate crop from which tapioca originates, is a major source of calories for over 800 million people worldwide. With this technology, farmers struggling with diseased crops can take immediate, restorative action to improve their livelihoods based on information about the health of their plants, generated using a portable, real-time DNA sequencing device.\n\nPortable DNA sequencing technology has great potential to reduce the risk of community crop failure and help improve livelihoods of millions of people, especially in low resourced communities. Plant diseases are a major cause of low crop productivity and viruses such as tobacco mosaic virus, tomato mosaic, tomato spotted wilt, potato leaf roll, Potato virus X and Y in potato, papaya mosaic, citrus tristeza, chilli leaf curl, and banana bunchy top have been implicated. In particular, cassava viruses are among the world’s greatest risk to food insecurity. Losses caused by cassava mosaic disease (CMD) and cassava brown streak disease are estimated at US $2-3 billion annually8.\n\n\nResults and Discussion\n\nWe visited smallholder farmers in Tanzania, Uganda and Kenya (Table 1) who are suffering yield shortages due to cassava virus infections. We utilized the MinION to test infected material and farmers were informed within 48 hours of the specific strain of the virus that was infecting their cassava, and a resistant cassava variety was deployed. The advantages of adopting this technology far outweigh the challenges (Table 2). Cassava mosaic begomoviruses were in high enough concentration that reads of whole genomes were obtained without an enrichment step (Table 1). As expected, the viral reads increased with the severity of the symptoms observed (Table 1). We detected a dual infection for a leaf sample with the severity score of 5 in Uganda. In addition, one asymptomatic plant in Tanzania had one viral read detected. The shortest time to obtain a viral read was 15 s (severity score 5) and the longest was 4 h 11 m 15 s (severity score 1).\n\nDisease severity described by Legg et al.2, where 1 is healthy and 5 shows severe symptoms of the disease including leaf distortion and stunting of the plant (Figure 1).\n\nJKUAT, Jomo Kenyatta University of Agriculture and Technology.\n\nMinION sequencing is superior to traditional methods of PCR identification, given its generation of whole genome sequences which enable the identification of the plant virus strain even if it becomes mutated or divergent, as it is not biased using primers that rely on known virus sequences. With regards to cassava, there are three major advantages of this technology. Firstly, improved diagnostics are required and real-time whole-genome sequencing will help develop diagnostic primers that are up-to-date. Secondly, this technology will assist with the development of resistant cassava varieties and will allow breeders to immediately test the varieties they are developing against different viral strains. Lastly, it ensures the delivery of the correct healthy uninfected planting material to farmers. In addition, we could detect virus in a plant before it showed symptoms (Table 1). Utilizing traditional PCR methods, three samples collected from farmer 1’s field in Tanzania tested positive for EACMVs and none were positive for ACMV. The asymptomatic sample from Mikocheni Agricultural Research Institute (MARI) tested negative for both ACMV and EACMVs. There were eight fresh cassava leaf samples from Uganda found to be dually infected with ACMV and EACMV-UG using conventional PCR primers for ACMV and EACMV-UG. The primers used in this PCR yield products of 1000 bp and 1500 bp for ACMV and EACMV-UG, respectively. In total, 12 Kenyan samples were tested and all but two (barcode 2 and 3) were found to be positive using conventional PCR (Table 1). Further studies are needed to verify our results regarding the sensitivity of the protocol for early detection of CMD in cassava, but these results are very promising for ensuring farmers receive clean planting material through the early detection of viral infection.\n\nNanopore sequencing technology has wide applications globally, but in East Africa these include: (a) crop improvement by screening for virus resistant germplasm and genetic diversity during breeding; (b) indexing of cassava planting materials for virus presence or absence to ensure that only clean materials in multiplication fields are distributed to farmers; (c) detection and identification of alternative plant species for cassava-infecting begomoviruses, so that farmers are advised to remove and/or grow crops away from such plants as a management strategy; and (d) virus and biodiversity studies.\n\n\nMethods\n\nIn Tanzania, three cassava mosaic disease (CMD) symptomatic cassava leaf samples (Figure 1, Table 1) were collected from the smallholder cassava farmer 1’s field in Bagamoyo. Disease severity was assessed as described by Legg et al.2, where 1 is healthy and 5 shows severe symptoms of the disease, including leaf distortion and stunting of the plant. One more asymptomatic leaf sample was collected at MARI, Dar es Salaam. In total, seven CMD symptomatic plants were collected from farmer 2’s farm in Wakiso district in Uganda. Both Tanzanian and Ugandan samples were collected in September 2017. A total of 12 samples from Kenya were collected in February 2018 from various sources (Table 1). High-quality DNA was isolated using the cetyl trimethylammonium bromide method9. Each DNA sample (Table 1) was quantified and the purity checked using a NanoDrop 2000c UV–vis Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) was used to check the purity and quantity of DNA for each sample, and results were recorded in Table 1.\n\nIn Tanzania and Uganda, the Rapid Barcoding kit SQK-RBK001 and MinION 9.4.1 flow cells (Oxford Nanopore) were used to process genomic DNA extracted using a standard CTAB method9. We utilized the Rapid Barcoding kit SQK-RBK004 with 9.4.1 flow cells in Kenya. DNA was diluted to 700 ng as specified in the library protocol. The SQK-RBK001 (Sept 2017) and/or the SQK-RBK004 (Feb 2018) protocols were performed as described by the manufacturer. In Tanzania and Uganda, the MinION was run for 24 h instead of the recommended 48 h, and in Kenya we had a total run time of approximately 17 h due to power interruptions.\n\nIn Tanzania and Uganda, Albacore 2.0.2 was used for base calling. In Kenya, Albacore 2.1.10 was used and the scripts were modified to reflect the newest rapid barcoding kit RBK004. Fastq files were imported into Geneious10 and a local blast database of all known cassava mosaic begomovirus whole genomes were downloaded from GenBank (program set to megablast, maximum hits=1, scoring (match mismatch 1-2, max e-value 1e-1, word size 28, gap cost=linear) and a local blast was performed on each of the reads generated using the Nanopore device. There are other bioinformatics pipelines that are used for Nanopore data, but the focus of our study was to use a simple, efficient analyses pipeline, which was a local blast database and a well-curated reference dataset11. To ensure we did not miss other begomoviruses or pathogens, we confirmed these in-field results by performing a post diagnostic blast of reads on the Nimbus Cloud at the Pawsey Supercomputing Center (Kensington, Australia) with blast 2.2.31 against the full NCBI nucleotide database to confirm results. For reference, the specific database used was {$ blastcmd –db nt/nt –info} Database: Nucleotide collection (nt) 46,853,753 sequences; 170,830,796,758 total bases Date: Feb 20, 2018 5:04 PM. The data were processed into a blast archive using a blast script with the following parameters (Script attached) {$blastn -query \"$file\" -db /mnt/nucdb/nt/nt -outfmt 11 -culling_limit 10 -out \"out.$file.asn\" -num_threads 17 } then converted into XML (for loading into Geneious) and HTML for viewing where possible.\n\nTraditional PCR was used to verify our DNA sequencing results. In Tanzania and Kenya, two primer pairs: EAB 555F/EAB 555F12 and JSP001/JSP00213, which amplify 556 bp and 774 bp, respectively, were used to detect East African CMVs (EACMVs) and African CMVs (ACMVs), respectively. Specifically, high quality DNA were isolated using CTAB method as described by Lodhi et al.9. Electrophoresis and spectrophotometric measurements were used to check the quality and quantity of DNA. PCR reactions were carried out in a 25 µl volume, with 12.5 µl (1X final concentration) of One Taq PCR 2X Master Mix (New England Biolabs), One Taq DNA Polymerase in an optimized buffer with 1.5 mM Mg2+ and 0.2 mM each dNTPs in 1X final concentration. A total of 0.2 µM (0.5 µl of each) of primer sets were used in the reaction and nuclease-free water were added to 25µl final reaction volume. The template used was less than 500 ng per reaction. PCR was carried out in 2720 Thermal Cycler (Applied Biosystems) with the following programme: 1 cycle of 3 min at 94°C, then 30 cycles at 94°C for 45 s, 56°C for 45 s, 72°C for 1 min, and a final cycle at 72°C for 7 min. Electrophoresis of the PCR product was run on 1% agarose gel stained in ethidium bromide in 1X TAE buffer in a submarine gel electrophoresis unit and visualized using a BioDoc-It 210 Imaging System m-20V Transilluminator (Thomas Scientific).\n\nIn Uganda, the presence of ACMV and EACMV in each sample was detected using a pair of specific primers for ACMV, ACMV-AL1/F and ACMV-ARO/R, and primers specific for EACMV-UG2, UV-AL1/F and ACMV-CP/R314. Specifically, the presence of ACMV and EACMV in each sample was detected using a pair of specific primers for ACMV, ACMV-AL1/F (5’-GCC GGA ATC CCT AAC ATT ATC -3’) and ACMV-ARO/R (5’-GCT CGT ATG TAT CCT CTA AGG CCT G -3’) and specific for EACMV-UG2, UV-AL1/F (5’-TGT CTT CTG GGA CTT GTG TG -3’) and ACMV-CP/R3 (5’- TGC CTC CTG ATG ATT ATA TGT C-3’) described by Zhou et al.14. The primers amplify about 1000 bp for ACMV and 1500 bp for EACMV-UG2 of the Coat Protein and AV2 gene sequences of the ACMV and EACMV genomes, respectively. The PCR reaction was set up using GoTaq® Green Master Mix (Promega, Madison USA). Each of the 25 µl PCR mix contained 12.0 µl of 2X GoTaq® Green Master Mix [containing GoTaq® DNA polymerase, 2X Green GoTaq® Reaction Buffer (pH 8.5), 400 µM dATP, 400 µM dGTP, 400 µM dCTP, 400 µM dTTP and 3 mM MgCl2], 1.0 µl of forward primer, 1.0 µl of reverse primer, 10.0 µl of nuclease-free water (Amressco, Ohio, USA) and 1.0 µl of DNA. PCR was performed using a Biometra professional thermocycler (Biometra, Gottingen, Germany) programmed as follows: 94 °C for 2 min for initial denaturation followed by 30 cycles of 94 °C for 1 min, 60 °C for 1.5 min, 72°C for 2 min and 72°C for 10 min for denaturation, annealing, extension and final extension, respectively. PCR amplicons were separated by electrophoresis in a 1× Tris-acetate-EDTA (TAE) buffer in a 1.2% agarose gel, stained with ethidium bromide (0.1 mg/ml) and visualized using a U:Genius3 (Syngene, Cambridge, UK) gel documentation system.\n\nThe image in Figure 1 was captured in Kiromo-Kitonga Bagamoyo, Tanzania by J.N. using a Samsung s8 smartphone.\n\nAn earlier version of this article can be found on bioRxiv (DOI: https://doi.org/10.1101/314526).\n\n\nData availability\n\nRaw data for this study are available on figshare: DOI: https://doi.org/10.6084/m9.figshare.666740911.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was part-funded by the Crawford Fund, Australia (grant number WA-803-2017).\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\nFunding for the Kenyan trip was provided by the Crawford Fund. We also thank the participants from the University of Eldoret who assisted in the preparation of libraries for the Kenyan samples.\n\n\nReferences\n\nSubramanya Sastry K, Zitter TA: Management of Virus and Viroid Diseases of Crops in the Tropics. In Plant Virus and Viroid Diseases in the Tropics Vol. 2 Epidemiology and Management. 2014; 149–480. Publisher Full Text\n\nLegg JP, Thresh JM: Cassava mosaic virus disease in East Africa: a dynamic disease in a changing environment. Virus Res. 2000; 71(1–2): 135–149. PubMed Abstract | Publisher Full Text\n\nAFDB: African Development Bank. Annual Report. 2016. Reference Source\n\nFARA: Science agenda for agriculture in Africa (S3A): “Connecting Science” to transform agriculture in Africa. Forum for Agricultural Research in Africa (FARA), Accra, Ghana. 2014. Reference Source\n\nQuick J, Loman NJ, Duraffour S, et al.: Real-time, portable genome sequencing for Ebola surveillance. Nature. 2016; 530(7589): 228–232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFaria NR, Sabino EC, Nunes MR, et al.: Mobile real-time surveillance of Zika virus in Brazil. Genome Med. 2016; 8(1): 97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuick J, Grubaugh ND, Pullan ST, et al.: Multiplex PCR method for MinION and Illumina sequencing of Zika and other virus genomes directly from clinical samples. Nat Protoc. 2017; 12(6): 1261–1276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScholthof KB, Adkins S, Czosnek H, et al.: Top 10 plant viruses in molecular plant pathology. Mol Plant Pathol. 2011; 12(9): 938–954. PubMed Abstract | Publisher Full Text\n\nLodhi MA, Ye GN, Weeden NF, et al.: A simple and efficient method for DNA extraction from grapevine cultivars and Vitis species Plant Mol Biol Rep. 1994; 12(1): 6–13. Publisher Full Text\n\nKearse M, Moir R, Wilson A, et al.: Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics. 2012; 28(12): 1647–1649. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoykin L: Nanopore sequencing of cassava from Tanzania, Uganda and Kenya. figshare. Fileset. 2018. Data Source\n\nFondong VN, Pita JS, Rey ME, et al.: Evidence of synergism between African cassava mosaic virus and a new double-recombinant geminivirus infecting cassava in Cameroon. J Gen Virol. 2000; 81(Pt 1): 287–297. PubMed Abstract | Publisher Full Text\n\nPita JS, Fondong VN, Sangaré A, et al.: Genomic and biological diversity of the African cassava geminiviruses. Euphytica. 2001; 120: 115–125. Publisher Full Text\n\nZhou X, Liu Y, Calvert L, et al.: Evidence that DNA-A of a geminivirus associated with severe cassava mosaic disease in Uganda has arisen by interspecific recombination. J Gen Virol. 1997; 78(Pt 8): 2101–2111. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "36183",
"date": "26 Jul 2018",
"name": "Ruth E Timme",
"expertise": [
"Reviewer Expertise I'm a Research Microbiologist for the US Food and Drug Administration. My research focuses on the validation of phylogenetic tools for molecular surveillance. I also manage a genome surveillance network for foodborne pathogens called the GenomeTrakr",
"that collects and publishes WGS data from food and environmental isolates."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nBoykin and co-authors present results from a pilot study exploring the use of MinION technology to detect a plant viral pathogen in real-time. Their data shows a huge advantage of using this hand-held sequence technology over the standard PCR methods. The authors propose implementing a MinION quality-control step on the plants before they are distributed, so that CMD-infected plants can be removed from the supply chain before they reach local farmers. This report is short, but its impact appears to be far-reaching. Most of my comments are minor editorial suggestions, but overall the writing and readability is excellent.\n\nMinor revisions: Public availability of the DNA sequence data. While the authors technically made the sequence files public by posting to FigShare, the standard repository for DNA sequence data is the INSDC (NCBI/EBI/DDBJ). I highly urge the authors to create a BioProject at NCBI or EBI that houses the raw and assembled sequences (fasta files) for this effort so that other researchers in this area can easily build off this important work.\n\nEditorial comments: Results and Discussion “We utilized the MinION to test infected material and farmers were informed within 48 hours of the specific strain of the virus that was infecting their cassava, and a resistant cassava variety was deployed.” Consider converting to two sentences. One about using the MinION and the second to cover the response. Were resistant cassava plants really deployed within 48 hrs? wow.\n\n“MinION sequencing is superior to traditional methods of PCR identification, given its generation of whole genome sequences which enable the identification of the plant virus strain even if it becomes mutated or divergent, as it is not biased using primers that rely on known virus sequences.” Consider a minor re-write: “In general MinION sequencing is superior to traditional PCR methods of identification because the virus can be detected even when the PCR primers don’t work, and 2) entire viral genome sequence is generated enabling the identification of the specific viral strain, along with other molecular information, which allows for a much higher resolution of surveillance.\n\n“In addition, we could detect virus in a plant before it showed symptoms (Table 1).” Change to present tense to match the rest of the paragraph?\n\n“Utilizing traditional PCR methods, three samples collected from farmer 1’s field in Tanzania tested positive for EACMVs and none were positive for ACMV.” Define EACMV and ACMV before abbreviation.\n\nMethods:\n\n“In Tanzania, three cassava mosaic disease (CMD) symptomatic cassava leaf samples (Figure 1, Table 1) were collected from the smallholder cassava farmer 1’s field in Bagamoyo.” CMD already defined in Intro.\n\n“In Tanzania and Kenya, two primer pairs: EAB 555F/EAB 555F12 and JSP001/JSP00213, which amplify 556 bp and 774 bp, respectively, were used to detect East African CMVs (EACMVs) and African CMVs (ACMVs), respectively.” Use the abbreviations here after adding the full names to the Results.",
"responses": []
},
{
"id": "36979",
"date": "24 Aug 2018",
"name": "Alfonso Benítez-Páez",
"expertise": [
"Reviewer Expertise My expertise areas are the microbial genomics",
"nanopore DNA sequencing technology",
"molecular biology",
"and the massive DNA data analysis."
],
"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 article Boykin et al present a pilot study aiming the application of real-time DNA sequencing for the detection of ACMV and EACMV in cassava plants in multiple crops of African East countries. This study represents the successful approaching of the most valuable feature of the MinION nanopore sequencing platform, its portability. At the same time, the authors made the maximum use of the singularities of the system by the translational application of their results, thus preventing the spread of plant viruses and to improve the crop efficiency by timely advising of farmers. This last exercise really highlights the value of such technology, particularly in the epidemiological surveillance and control of pathogens.\nNotwithstanding, I have some minor concerns that if addressed they would constitute an added value to the approach described.\n1) I strongly recommend that authors store and make publicly available the genetic information retrieved from different sequencing runs to a specialized repository such as ENA or GenBank.\n2) It would be very informative for future studies in this field that Table 1 contains additional information the average or median values of the sequence identity derived from the comparison between nanopore reads and reference sequences. In a similar way, they should declare the level of relationship between ACMV and EACMV, in terms of genome-wide nucleotide identity, in order to disclose any potential misidentification given the high error rate of nanopore-derived DNA reads.\n3) The authors must be aware that there is a strong effect derived from sequencing kits used, different from Uganda/Tanzania and Kenya. They should make some correlations between the CMD severity scoring and DNA reads retrieved independently accordingly to the kits used.\n4) In the same line of thoughts than above, it would be very elegant that authors will estimate the maximum time for expecting a viral DNA read just for setting a threshold and optimize the sequencing time. I have noticed that for CMD severity = 1, it took maximum 4h to retrieve viral DNA reads, using the sequencing kit SQK-RBK001. Another different history was the utilization of SQK-RBK004 in Kenya, where apparently there is not a correlation between CMD severity and viral DNA reads retrieved. In the last cases, the apparently not symptomatic plants were detected as positive in less than one hour. The setting of a time threshold for a proper detection (getting enough number of reads to estimate reliable identification) would be useful to speed up the farmers' advising and consequently the reduction of risks for the spread.",
"responses": [
{
"c_id": "4227",
"date": "13 Nov 2018",
"name": "Edwin karoney",
"role": "Reader Comment",
"response": "Great work! Outstanding technology!"
}
]
}
] | 1
|
https://f1000research.com/articles/7-1101
|
https://f1000research.com/articles/7-665/v1
|
25 May 18
|
{
"type": "Research Article",
"title": "Phenotype profiling of white-nose syndrome pathogen Pseudogymnoascus destructans and closely-related Pseudogymnoascus pannorum reveals metabolic differences underlying fungal lifestyles",
"authors": [
"Vishnu Chaturvedi",
"Holland DeFiglio",
"Sudha Chaturvedi",
"Holland DeFiglio",
"Sudha Chaturvedi"
],
"abstract": "Background: Pseudogymnoascus destructans, a psychrophile, causes bat white-nose syndrome (WNS). Pseudogymnoascus pannorum, a closely related fungus, causes human and canine diseases rarely. Both pathogens were reported from the same mines and caves in the United States, but only P. destructans caused WNS. Earlier genome comparisons revealed that P. pannorum contained more deduced proteins with ascribed enzymatic functions than P. destructans. Methods: We performed metabolic profiling with Biolog PM microarray plates to confirm in silico gene predictions. Results: P. pannorum utilized 78 of 190 carbon sources (41%), and 41 of 91 nitrogen sources (43%) tested. P. destructans used 23 carbon compounds (12%) and 23 nitrogen compounds (24%). P. destructans exhibited more robust growth on the phosphorous sources and nutrient supplements (83% and 15%, respectively) compared to P. pannorum (27% and 1%, respectively.). P. pannorum exhibited higher tolerance to osmolytes, pH extremes, and a variety of chemical compounds than P. destructans. Conclusions: An abundance of carbohydrate degradation pathways combined with robust stress tolerance provided clues for the soil distribution of P. pannorum. The limited metabolic profile of P. destructans validated in silico predictions of far fewer proteins and enzymes. P. destructans ability to catabolize diverse phosphorous and nutrient supplements might be critical in the colonization and invasion of bat tissues. The present study of 1,047 different metabolic activities provides a framework for future gene-function investigations of the unique biology of the psychrophilic fungi.",
"keywords": [
"Psychrophilic fungi",
"phenotype microarray",
"metabolism",
"catabolism",
"gene function"
],
"content": "Introduction\n\nPseudogymnoascus destructans causes white-nose syndrome (WNS), a disseminated disease afflicting hibernating bats in North America since 20061–3. WNS is linked to mass mortality and now afflicts bats over large geographic areas in the United States and Canada. P. destructans’ pathogenic mechanisms remain mysterious especially as no other human or animal fungal pathogen expresses virulence attributes at such low temperatures. Pseudogymnoascus pannorum, a closely related fungus, is widely distributed in the soil and substrates of caves and mines in North America3. P. pannorum grows both at psychrophilic and mesophilic temperature ranges and causes human and canine diseases rarely4. However, P. pannorum does not cause any disease in hibernating bats. These facts raise the exciting possibilities that P. destructans is more specialized for the pathogenic lifestyle on bats while P. pannorum successfully colonizes a broader range of substrates in nature.\n\nEnvironmental studies on the psychrophilic and psychrotolerant fungi documented the versatility of Pseudogymnoascus (Geomyces) pannorum for the utilization of complex carbohydrates and keratin-enriched substrates, and tolerance to high salt5–7. Additional laboratory studies demonstrated extensive saprotrophic enzymatic activities that would allow resource capture by the non-pathogenic Pseudogymnoascus species vis-a-vis P. destructans8,9. P. destructans is known to secrete proteolytic, lipolytic, and keratinolytic exoenzymes, and possesses specialized catabolic activities that contribute to its growth and survival in the nutrient-poor caves and mines2,10.\n\nAlthough their draft genomes are similar in size (~30 Mb), there are numerous repeats and far fewer proteins and enzymes in P. destructans (2,052 proteins) than in P. pannorum (2,734 proteins)11. In the present study, we report the results of extensive Biolog Phenotype Microarray metabolic profiling to confirm in silico gene predictions, and find clues for the different lifestyles of these psychrophilic fungi.\n\n\nMethods\n\nThe metabolic analysis was conducted using P. destructans (M1379) and P. pannorum (M1372)11. The PM1-10 and PM21, 23–25 phenotype microarray plates were procured from Biolog, Hayward, CA. The fungal spores were harvested in sterile water from 3 - 5-week-old, heavily sporulating culture on potato dextrose agar (PDA) flasks at 15°C. In preliminary experiments, spore counts and viability were determined on agar plates using a hemocytometer and colony forming units (CFU). For the final tests, the spores were harvested, washed once in sterile water by centrifugation, and the suspension adjusted to an OD600 0.2 (transmittance = 62%). This suspension equated to between 550 and 950 spores per well via hemocytometer count, and 250–500 spores per well by CFU.\n\nThe PM plates were inoculated per Biolog protocol and incubated at 15°C12,13. The presence or absence of growth was measured by OD600 on day 10 for P. destructans, and day 7 for P. pannorum. Negative control wells were growth positive for both P. destructans and P. pannorum. Therefore, the corresponding negative control well reading from each experiment were averaged together and used to normalize the OD values averages for each test compound. The phenotypic assay was repeated once, and the average of two readings used in the subsequent analysis.\n\n\nResults\n\nNearly 1,047 different metabolic activities were analyzed for each test fungus (Figure 1 and Datasets 1–414). P. pannorum metabolized far more carbon and nitrogen compounds; P. destructans exhibited prominent activity on phosphorous sources and nutrient supplements (Figure 2). P. pannorum utilized 78 of 190 carbon sources (41%), and 41 of 91 nitrogen sources (43%) tested. P. destructans used 23 carbon compounds (12%) and 23 nitrogen compounds (24%). P. destructans exhibited more robust growth on the phosphorous sources and nutrient supplements (83% and 15%, respectively) compared to P. pannorum (27% and 1%, respectively.). P. pannorum metabolized nearly all carbon intermediates in the major fungal metabolic cycles13 (Figure 3). P. destructans utilized only a few simple sugars in glycolysis with no activity on a range of carbon intermediates. P. pannorum used a wider variety of nitrogen sources including amino acids, amino bases, and alkanes while P. destructans had a preference for the simple N sources and dipeptides13 (Figure 4). Most phosphorous sources tested were utilized by P. destructans while P. pannorum only grew on few phosphosugars and phosphorylated nucleosides (Figure 5). Both fungi did not utilize sulfur intermediates (Datasets 1–414). Fifteen of ninety-five nutrient supplements supported good growth of P. destructans while P. pannorum grew only on D-Pantothenic acid. P. pannorum grew at very high salt concentrations and extreme acidic and basic pH ranges while P. destructans was sensitive to high salt and basic pH (Figure 6). P. pannorum showed extreme tolerance to 96 xenobiotics in PM21, PM23 - PM25 plates in contrast to severe sensitivity observed in P. destructans (Figure 1).\n\nGreen squares include substrates with positive reactions; red boxes denote lack of metabolic activity in some plates.\n\n\nDiscussion\n\nMetabolic profiles of P. destructans and P. pannorum validated in silico predictions about the notable differences in the number of protein-encoding genes in their genomes11. P. destructans contained enzymes and catabolic pathways that support fungal growth on a limited range of substrates of non-plant origin and showed high sensitivity to stress. P. pannorum was remarkably adapted for the nutrient poor environments of the caves and mines (‘extremophile’) with oligotrophic metabolism, osmotolerance, xerotolerance, and xenobiotic tolerance.\n\nThe findings in the present study confirm and expand on results from other reports on P. destructans’ adaptation and persistence in the North American caves and mines in the face of possible competitive interactions with the native fungal species8–10. Both Raudabaugh and Miller (2013) and Reynolds and Barton (2014) used a variety of biochemical tests to probe the metabolic activities in a collection of Pseudogymnoascus species isolates9,10. The authors of the former study surmised the suitability of P. destructans as a saprobe in the affected caves and mines in limited biotic competition (‘resource island’)10. Reynolds and Barton (2014) found a reduced saprotrophic ability in P. destructans isolates vis-à-vis P. pannorum and other Pseduogymnoascus species, which suggested ‘co-evolution with the host’9. Wilson et al. (2017) performed a variety of tests including Biolog FF Microplate with 95 different substrates, and found limited saprotrophic ability in P. destructans in comparison to other Pseudogymnoascus species8.\n\nFurther Phenotype Microarray profiling of P. destructans and P. pannorum would be crucial to fill-in current gaps in their genome sequences, define gene functions, and elucidate pathophysiological attributes11,15,16. We and others hope to accomplish these milestones with the recent availability of a high-quality P. destructans genome and data pipelines to automate Biolog analysis15,17,18.\n\n\nData availability\n\nDatasets 1–4: Excel sheets with OD600 values for all Biolog plates tested in this study. DOI, 10.5256/f1000research.15067.d20467914",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the National Science Foundation (Award Number 1203528).\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\nBlehert DS, Hicks AC, Behr M, et al.: Bat white-nose syndrome: an emerging fungal pathogen? Science. 2009; 323(5911): 227. PubMed Abstract | Publisher Full Text\n\nChaturvedi V, Springer DJ, Behr MJ, et al.: Morphological and molecular characterizations of psychrophilic fungus Geomyces destructans from New York bats with White Nose Syndrome (WNS). PLoS One. 2010; 5(5): e10783. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinnis AM, Lindner DL: Phylogenetic evaluation of Geomyces and allies reveals no close relatives of Pseudogymnoascus destructans, comb. nov., in bat hibernacula of eastern North America. Fungal Biol. 2013; 117(9): 638–49. PubMed Abstract | Publisher Full Text\n\nChristen-Zaech S, Patel S, Mancini AJ: Recurrent cutaneous Geomyces pannorum infection in three brothers with ichthyosis. J Am Acad Dermatol. 2008; 58(5 Suppl 1): S112–3. PubMed Abstract | Publisher Full Text\n\nMarshall WA: Aerial Transport of Keratinaceous Substrate and Distribution of the Fungus Geomyces pannorum in Antarctic Soils. Microb Ecol. 1998; 36(2): 212–9. PubMed Abstract | Publisher Full Text\n\nFenice M, Selbmann L, Zucconi L, et al.: Production of extracellular enzymes by Antarctic fungal strains. Polar Biol. 1997; 17(3): 275–80. Publisher Full Text\n\nKochkina GA, Ivanushkina NE, Akimov VN, et al.: [Halo- and psychrotolerant Geomyces fungi from arctic cryopegs and marine deposits]. Microbiology. 2007; 76(1): 31–8. Publisher Full Text\n\nWilson MB, Held BW, Freiborg AH, et al.: Resource capture and competitive ability of non-pathogenic Pseudogymnoascus spp. and P. destructans, the cause of white-nose syndrome in bats. PLoS One. 2017; 12(6): e0178968. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReynolds HT, Barton HA: Comparison of the white-nose syndrome agent Pseudogymnoascus destructans to cave-dwelling relatives suggests reduced saprotrophic enzyme activity. PLoS One. 2014; 9(1): e86437. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRaudabaugh DB, Miller AN: Nutritional capability of and substrate suitability for Pseudogymnoascus destructans, the causal agent of bat white-nose syndrome. PLoS One. 2013; 8(10): e78300. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChibucos MC, Crabtree J, Nagaraj S, et al.: Draft Genome Sequences of Human Pathogenic Fungus Geomyces pannorum Sensu Lato and Bat White Nose Syndrome Pathogen Geomyces (Pseudogymnoascus) destructans. Genome Announc. 2013; 1(6): pii: e01045-13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBochner BR, Gadzinski P, Panomitros E: Phenotype microarrays for high-throughput phenotypic testing and assay of gene function. Genome Res. 2001; 11(7): 1246–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNai C, Wong HY, Pannenbecker A, et al.: Nutritional physiology of a rock-inhabiting, model microcolonial fungus from an ancestral lineage of the Chaetothyriales (Ascomycetes). Fungal Genet Biol. 2013; 56: 54–66. PubMed Abstract | Publisher Full Text\n\nChaturvedi V, DeFiglio H, Chaturvedi S: Dataset 1 in: Phenotype profiling of white-nose syndrome pathogen Pseudogymnoascus destructans and closely-related Pseudogymnoascus pannorum reveals metabolic differences underlying fungal lifestyles. F1000Research. 2018. Data Source\n\nCuevas DA, Garza D, Sanchez SE, et al.: Elucidating genomic gaps using phenotypic profiles [version 2; referees: 1 approved, 1 approved with reservations]. 2016; 3: 210. Publisher Full Text\n\nMackie AM, Hassan KA, Paulsen IT, et al.: Biolog Phenotype Microarrays for phenotypic characterization of microbial cells. Methods Mol Biol. 2014; 1096: 123–30. PubMed Abstract | Publisher Full Text\n\nVehkala M, Shubin M, Connor TR, et al.: Novel R pipeline for analyzing Biolog Phenotypic MicroArray data. PLoS One. 2015; 10(3): e0118392. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDrees KP, Palmer JM, Sebra R, et al.: Use of Multiple Sequencing Technologies To Produce a High-Quality Genome of the Fungus Pseudogymnoascus destructans, the Causative Agent of Bat White-Nose Syndrome. Genome Announc. 2016; 4(3): pii: e00445-16. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "34411",
"date": "07 Jun 2018",
"name": "Christine Salomon",
"expertise": [
"Reviewer Expertise Microbial ecology",
"microbial natural products chemistry",
"fungal and bacterial infectious disease",
"biological control"
],
"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 relatively straightforward study focused on comparing the nutrient utilization capacity of the fungal bat pathogen Pseudogymnoascus destructans versus the closely related species P. pannorum using the well established Biolog phenotype system. In general, the results suggest that P. pannorum can more readily utilize most carbon and nitrogen sources compared to P. destructans under the experimental conditions tested (15 deg C, and 7 and 10 days, respectively). The bat pathogen was also more sensitive to pH extremes and less tolerant to high salt.\n\nThe authors conclude that these results validate their previous whole genome studies which compared the predicted protein numbers between these two species. In general, the results also support previous metabolic capacity studies of P. destructans and other non-pathogenic Pseudogymnoascus species.\nThere are some important questions that should be addressed, and additional details that would improve this manuscript:\nHow was the incubation time of 10 and 7 days for each species determined? Presumably by comparison of equivalent growth in the control wells, but this detail should be provided. If this time is increased, does the utilization capacity of P. destructans eventually catch up?\n\nMore details should be provided about the cut-off value determination for growth versus no growth. Also, there should be some analysis of the range of results, versus simply using the average of the two readings. For example, what was the standard deviation for replicates?\n\nIt would be helpful to have a map of the nutrient sources, xenobiotics, etc. in the supplementary data to accompany the OD data (the numerical data alone is impossible to interpret without any other identifying information)--presumably Biolog provides this as a document.\n\nIt’s not clear how the heat map values were generated, if the starting spore inoculum had an OD of 0.2. (since the lowest value on the heat map is “0” OD). Presumably, if no growth occurred under a given condition, it would remain at the starting OD? Also, is 1.0 the highest OD obtained or was the data scaled to 0-1.0?\n\nThe heat map figure for the “nutrient supplements” is missing (also, I’m not sure what compounds this category encompasses, so some information about this would also be helpful, perhaps even just referring to the plate map in the supplementary data if that is added).\n\nFor Figure 6, the solid and dotted green and red lines seem to indicate relative growth, but the numerical cutoffs should be provided in the methods (or figure legend). For example, the growth in the well for 2% NaCl looks (labeled with a solid green line above) looks similar to the well for pH 9 with a dotted green line, but presumably are numerically different.\n\nPart of the justification for doing this work is stated as confirming the in silico gene predictions (from a previous publication by the authors). However, it’s not entirely clear that the just comparing the overall numbers of predicted proteins is actually correlated to the overall number of nutrient sources that can be utilized. This seems likely to be true, but the two studies don’t necessarily test/confirm this connection. The reference cited is a short report on the overall sequencing of the P. destructans and P. pannorum genomes, and prediction of encoded proteins, but no significant functional analysis. It might be more relevant to include references that include more functional data on metabolic and enzymatic capacities.\n\nOverall, this work adds important information about the competitive ability and metabolic specificity of P. destructans and could provide additional insight into fungal life history strategies and potential ways to control or mitigate white nose syndrome in bats. Some additional details (highlighted above) would provide critical information that would allow others to replicate or expand on this work.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3815",
"date": "17 Jul 2018",
"name": "Vishnu Chaturvedi",
"role": "Author Response",
"response": "Thanks very much for your insightful comments on the manuscript. We have modified the manuscript given your suggestions. How was the incubation time of 10 and 7 days for each species determined? Presumably by comparison of equivalent growth in the control wells, but this detail should be provided. If this time is increased, does the utilization capacity of P. destructans eventually catch up? In preliminary experiments, both fungi grew at different growth rates and comparable growth was observed after day 7 for P. pannorum and day 10 for P. destructans (details not shown). Further incubation of the plates beyond the observation period did not change the observed growth pattern. More details should be provided about the cut-off value determination for growth versus no growth. Also, there should be some analysis of the range of results, versus simply using the average of the two readings. For example, what was the standard deviation for replicates? The PM plates were inoculated per Biolog protocol and incubated at 15° C. The presence or absence of growth was measured by OD600 on day 7 for P. pannorum and day 10 for P. destructans. The negative control wells were weakly growth positive for both P. destructans and P. pannorum. The observation was also reported for the Biolog PM plates in another study (Nai C, et al. Fungal Genet Biol. 2013; 56:54-66). Therefore, the corresponding negative control well reading from each experiment were averaged together and used to normalize the OD values averages for each test compound. For the heat map visualization, the negative control reading was assigned a score 0.0 and the positive growth scored on a 0.0 -1.0 scale. The phenotypic assay was repeated once. The limited dataset precluded any quantitative statistical analyses. It would be helpful to have a map of the nutrient sources, xenobiotics, etc. in the supplementary data to accompany the OD data (the numerical data alone is impossible to interpret without any other identifying information)--presumably Biolog provides this as a document. The Biolog maps for all test compound are now included as a supplementary file. It’s not clear how the heat map values were generated, if the starting spore inoculum had an OD of 0.2. (since the lowest value on the heat map is “0” OD). Presumably, if no growth occurred under a given condition, it would remain at the starting OD? Also, is 1.0 the highest OD obtained or was the data scaled to 0-1.0? Please see details provided earlier about the heat map visualization. The heat map figure for the “nutrient supplements” is missing (also, I’m not sure what compounds this category encompasses, so some information about this would also be helpful, perhaps even just referring to the plate map in the supplementary data if that is added). The nutrient map from Biolog was uploaded as a supplementary file. For Figure 6, the solid and dotted green and red lines seem to indicate relative growth, but the numerical cutoffs should be provided in the methods (or figure legend). For example, the growth in the well for 2% NaCl looks (labeled with a solid green line above) looks similar to the well for pH 9 with a dotted green line, but presumably are numerically different. The legend for figure 6 (now figure 5) was modified for clarity; the summary numerical cutoff values were included in the revised text. Part of the justification for doing this work is stated as confirming the in silico gene predictions (from a previous publication by the authors). However, it’s not entirely clear that the just comparing the overall numbers of predicted proteins is actually correlated to the overall number of nutrient sources that can be utilized. This seems likely to be true, but the two studies don’t necessarily test/confirm this connection. The reference cited is a short report on the overall sequencing of the P. destructans and P. pannorum genomes, and prediction of encoded proteins, but no significant functional analysis. It might be more relevant to include references that include more functional data on metabolic and enzymatic capacities. Instead of ‘confirm,’ we switched to ‘assess correlations with in silico gene predictions.’ The two studies provide initial information about the genomes and metabolic pathways. The findings indicate a trend towards unique genomic and phenomic attributes in two psychrophilic fungi with different lifestyles. We agree with the reviewer that there is need to carry out more detailed functional analysis to identify unique genes, enzymes, and other proteins that differentiate these two psychrophiles. Such studies, necessitating considerable resources, are planned for the future."
}
]
},
{
"id": "34574",
"date": "15 Jun 2018",
"name": "Christopher T Cornelison",
"expertise": [
"Reviewer Expertise Emerging fungal pathogens",
"microbial control",
"applied microbiology."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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, \"Phenotype profiling of white-nose syndrome pathogen Pseudogymnoascus destructans and closely-related Pseudogymnoascus pannorum reveals metabolic differences underlying fungal lifestyles\", describes the comparative analysis of metabolic profiling of 2 closely related fungal pathogens with vastly different hosts and virulence. The manuscript utilizes the well known BioLog system to accomplish the comparison. In general the the study is well designed and executed and the manuscript well written. The conclusions are not surprising considering previous publications regarding the genomics of Pd and its loss of carbon utilization related gene content. Accordingly the impact of the findings on the field are modest and the sophistication of the analysis is simplistic. Regardless the manuscript does support previous findings and although the methods are limited in scope they are sound and well vetted. Accordingly it is my recommendation that the manuscript is acceptable as it is.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3816",
"date": "17 Jul 2018",
"name": "Vishnu Chaturvedi",
"role": "Author Response",
"response": "We appreciate very much the encouraging comments of the reviewer and the approval of the manuscript as it is."
}
]
},
{
"id": "35120",
"date": "29 Jun 2018",
"name": "Flavia Pinzari",
"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 is focussed on a comparison between two Pseudogymnoascus fungal species, which belong to different species but have a partly overlapped ecological niche, based on the use of metabolic profiling with Biolog Phenotype Microarray commercial multiwell plates. These plates are used with redox dyes to evaluate substrate use. With fungi this colorimetric approach is very complicated, and as the authors actually did, the fungal growth in the wells is usually measured as a change in optical density.\nThe interesting aspects and merits of this work are the following:\nThis is a brilliant use of this technique, since the comparison between a pathogenic/parasitic species and a mainly saprophytic one can really highlight important clues on the nutritional requirements for the pathogenic organism to spread and develop. The two compared species P. destructans and the closely related species P.pannorum are truly interesting from different points of view. They live in caves, they live at low temperatures, they are very close but behave differently, they attack mammals, etc.\nThe main criticisms regard:\n\nThe description of the methods (very poor: it is difficult to understand the procedure followed both in data production and analysis), and the way data are presented. Figure 1 is really useless. Figure 3 is informative, but little or nothing is reported on data analysis in the methodological section. Apparently, the authors harvested the two fungi for inoculum preparation in two different moments (different sporification time: 3 to 5 weeks, it can be very different). They also chose two different incubation times for comparing the catabolism of the two fungi. These choices should be discussed and justified. The two fungi have different development times. This can be the main reason for the differences observed. A better description and motivation of the chosen approach would make the work stronger and clearer. Instead of single time-point comparison the authors could have used empirical models and regression splines that allow extrapolation of curve parameters of biological interest, namely, lag time, maximum rate of increase and maximum absorbance. Curve integration and the resulting area under the curve condenses these three parameters into a single estimate that can also be used to compare kinetics across substrates and samples. Curve parameters offer the main advantage of being independent of incubation time, while also accounting for potential differential rates of colour development across substrates and plates. An example of this approach is given in the following work: Canfora et al (2017)1, and theory is reported in the references therein. The authors tested only one strain for each species (I agree that PM plates are very expensive, but drawing a result using only one strains in the comparison is a limitation in case of intraspecific variability, even if they used two replicates) The lowest value on the heat map is an OD of “zero”. How was data scaling performed? What the red and green lines stand for in Figure 3? Legends are needed. References list is lacking of some important elements. One is the following: Atanasova L, Druzhinina IS2. The authors reported that predicted enzymes are related to the number of carbon sources that can be utilized by the two fungal species however, a better definition of the kind of correlations observed between genotype and phenotype is needed to better understand this connection.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3818",
"date": "17 Jul 2018",
"name": "Vishnu Chaturvedi",
"role": "Author Response",
"response": "We appreciate the thoughtful comments of the reviewer. The main criticisms regard: The description of the methods (very poor: it is difficult to understand the procedure followed both in data production and analysis), and the way data are presented. We have added required details in the methods. Figure 1 is really useless. Figure 3 is informative, but little or nothing is reported on data analysis in the methodological section. Figure 1 provided a bird’s-eye view of the scope of work and different trends noticed in this comparative study. We appreciate that someone knowledgeable about Biolog typing might find this information redundant and therefore, figure 1 was removed from the revised version. We expanded the legend for figure 3 (now figure 2). Apparently, the authors harvested the two fungi for inoculum preparation in two different moments (different sporification time: 3 to 5 weeks, it can be very different). They also chose two different incubation times for comparing the catabolism of the two fungi. These choices should be discussed and justified. The two fungi have different development times. This can be the main reason for the differences observed. A better description and motivation of the chosen approach would make the work stronger and clearer. The reviewer’s criticism is valid similar to the reviewer 1. We have added details in the revised version to ally concern that “differences observed were due to different development times”. Instead of single time-point comparison the authors could have used empirical models and regression splines that allow extrapolation of curve parameters of biological interest, namely, lag time, maximum rate of increase and maximum absorbance. Curve integration and the resulting area under the curve condenses these three parameters into a single estimate that can also be used to compare kinetics across substrates and samples. Curve parameters offer the main advantage of being independent of incubation time, while also accounting for potential differential rates of colour development across substrates and plates. An example of this approach is given in the following work: Canfora et al (2017)1, and theory is reported in the references therein. We have used the biolog phenotype plates as per the manufacturer’s instructions. There are many publications on Biolog profiling that use single end-point reading for trend analysis. The reviewer refers to a more elaborate setup, and we agree that studies similar to Canofora et al. (2017) provide more accurate and dynamic data for building predictable models. This reference was included in the bibliography and a sentence added in the revised text about the limitation of this qualitative study, and the scope for more dynamic data generation. We shall pursue future opportunities to conduct such a study with the two fungi. The authors tested only one strain for each species (I agree that PM plates are very expensive, but drawing a result using only one strains in the comparison is a limitation in case of intraspecific variability, even if they used two replicates) We believe that a trend analysis is possible and meaningful with one well-defined strain representative from each species. This is also the basis for whole genome sequencing of most fungi. The lowest value on the heat map is an OD of “zero”. How was data scaling performed? We have added more details to describe the heat map. What the red and green lines stand for in Figure 3? Legends are needed We have modified the legends for the figure 6 (now figure 5). The red and green lines were deleted. References list is lacking of some important elements. One is the following: Atanasova L,Druzhinina IS2. We have included the article by Druzhinina et al. The authors reported that predicted enzymes are related to the number of carbon sources that can be utilized by the two fungal species however, a better definition of the kind of correlations observed between genotype and phenotype is needed to better understand this connection. We modified the text to suggest that the trend analysis of Biolog data indicated correlation with the in silico analysis of genome sequencing data."
}
]
}
] | 1
|
https://f1000research.com/articles/7-665
|
https://f1000research.com/articles/7-407/v1
|
28 Mar 18
|
{
"type": "Systematic Review",
"title": "Predictive physiological anticipation preceding seemingly unpredictable stimuli: An update of Mossbridge et al’s meta-analysis",
"authors": [
"Michael Duggan",
"Patrizio E. Tressoldi",
"Michael Duggan"
],
"abstract": "Background: This is an update of the Mossbridge et al’s meta-analysis related to the physiological anticipation preceding seemingly unpredictable stimuli. The overall effect size observed was 0.21; 95% Confidence Intervals: 0.13 - 0.29 Methods: Eighteen new peer and non-peer reviewed studies completed from January 2008 to October 2017 were retrieved describing a total of 26 experiments and 34 associated effect sizes. Results: The overall weighted effect size, estimated with a frequentist multilevel random model, was: 0.29; 95% Confidence Intervals: 0.19-0.38; the overall weighted effect size, estimated with a multilevel Bayesian model, was: 0.29; 95% Credible Intervals: 0.18-0.39. Effect sizes of peer reviewed studies were slightly higher: 0.38; Confidence Intervals: 0.27-0.48 than non-peer reviewed articles: 0.22; Confidence Intervals: 0.05-0.39. The statistical estimation of the publication bias by using the Copas model suggest that the main findings are not contaminated by publication bias. Conclusions: In summary, with this update, the main findings reported in Mossbridge et al’s meta-analysis, are confirmed.",
"keywords": [
"pre-stimulus activity",
"anticipatory physiology",
"temporal processing",
"psychophysiology",
"presentiment"
],
"content": "Introduction\n\nThe human ability to predict future events has been crucial in our evolutionary development and proliferation over epochs of time, both from a species perspective, but also, on an individual level. Our day-to-day survival is predicated on a successful marriage of experience (e.g., memory) and sensory processing (e.g., perceptual cues); for example, on a very humid heavily overcast night, our perceptions and memories inform us that a thunder storm is possible and it might be intelligent to find shelter. Such behaviour is highly adaptive as it fosters survival based strategies and is perfectly explicable in terms of current theories of biological causality. Now imagine if such prognosticating ability was possible without any sensory or other inferential cues. Such seemingly inexplicable ability would definitely hold survival advantage, if they existed. For millennia people have been reporting strange feelings of foreboding that later transpired to have significance. Over the last 36 years these phenomena have been scrutinized in the laboratory in which a subject’s physiology is monitored before a randomly presented stimulus that is designed to evoke a significant post-stimulus response. Disturbingly, moments before the stimulus is presented there are murmurings of activity, as if the body is predicting moments ahead of time. This effect is termed presentiment, or more recently, Predictive Anticipatory Activity (Mossbridge et al., 2014). By 2012 a good number of these studies had been completed and it was deemed worthwhile to conduct a meta-analysis of the extant literature at the time. Mossbridge, Tressoldi and Utts located 42 studies published from 1978 to 2010, testing the presentiment hypothesis, out of which 26 enabled a true comparison between pre and post-stimulus epochs (Mossbridge et al., 2012), that is the pre-stimulus physiological responses mirrored even if to a lesser degree, the post-stimulus responses.\n\nHere two paradigms were used: either a randomly ordered presentation of arousing vs. neutral stimuli or guessing tasks in which the stimulus is the feedback about the participant’s guess (correct vs. incorrect). In both of these approaches it is difficult to envision mundane strategies that might explain the anomalous pre-stimulus effects observed, and indeed, Mossbridge et al, went to significant lengths in refuting the leading candidate – expectancy effects, both in the 2012 meta-analysis and in post-review exchanges with sceptical psychologists and physiologists. Regardless of the paradigm, a broad range of physiological measures were employed from skin conductance, heart rate, blood volume, respiration, electroencephalographic (EEG) activity, pupil dilation, blink rate, and/or blood oxygenation level dependent (BOLD) responses. These are recorded throughout the session, with a pre-determined anticipatory period of between 4 to 10 seconds, in which the any pre-stimulus effect is captured. The presentiment hypothesis calls for a difference between arousing and neutral pre-stimulus responses and this is calculated across sessions. Mossbridge et al. found substantive evidence in favour of a presentiment effect concatenated to over 6 sigma – extreme statistical significance. Additionally, they also found evidence of presentiment effects from mainstream research programs – something that is becoming increasingly important as these effects become more widely known.\n\nBecause of the high profile nature of Mossbridge et al, (over 93,000 views as of January 2018) there has been a good number of replications in the few years since publication. We located an additional 26 studies describing 34 effect sizes from a dozen laboratories. The most striking aspect of this fresh database is the sheer variation in experimental approaches as researchers seek to tackle more process oriented questions rather than continuing the proof-oriented work found in the earlier meta-analysis. Because expectancy effects have been forwarded to explain at least some of the presentiment effect, it is noteworthy that several experiments in this fresh cohort of studies tackle this head on by only analysing the first trial of a run. These single-trial presentiment studies are expectancy free and are becoming more dominant in this research domain. Another interesting question that is probed in these new studies is the idea of utilizing pre-stimulus physiological activity to predict future events. This provides a second objective measure of the validity of the presentiment effect. There are several studies that utilize this approach and they are discussed later on. Additionally, we also found increasing evidence of presentiment research piggybacking onto mainstream psychology programs, even informing aspects of the conventional research. Also of note we found several PhD theses describing presentiment research and a greater geographical spread than in 2012, both evidence of the increasing attention such research is garnering. Lastly, we found increasing dialogue between presentiment researchers and physicists interested in retrocausality – the idea that effects can precede their cause. This is witnessed in the recent AAAS retrocausality symposium in which several researchers participated and in which some of those papers made their way into this meta-analysis (Sheehan, 2017).\n\n\nMethods\n\nThe whole procedure followed both the APA Meta-Analysis Reporting Standards (APA Publications and Communications Board Working Group on Journal Article Reporting Standards, 2008), the Preferred Reporting Items for Systematic reviews and Meta-Analyses for Protocols 2015 (Moher et al., 2015) and the reporting standards for literature searches and report inclusion (Atkinson et al., 2015). A completed PRISMA checklist can be found in Supplementary File 1.\n\nStudy inclusion criteria were the analysis of both psychophysiological or neurophysiological signals before the random presentation of whichever type of stimulus, e.g. pictures, sounds etc. Randomization could be performed by using pseudo-random algorithms e.g. like those implemented in MatLab or E-Prime® or true random sources of random digits, e.g. TrueRNG.\n\nIt is important to point out that these eligibility criteria are different from those used by Mossbridge et al. Those authors selected only studies were the anticipatory signals mirrored the post-stimulus ones. Differently we included all studies that used anticipatory signals to predict future events independently of the presence of post-stimulus physiological signals. For example, some authors, e.g. Mossbridge (2015) used heart rate variability to predict winning i.e. $4, versus losing outcomes. Our inclusion criteria are consequently more comprehensive than those used by Mossbridge et al.\n\nBoth co-authors who are experts in this type of investigations, searched for studies through Google Scholar and PubMed by using the keywords: “presentiment” OR “anticipation” OR “precognition”. Furthermore, we emailed a request of the data of completed studies to all authors we knew were involved in this type of investigations. Even if Mossbridge et al. included all studies available up to 2010, we also searched studies that could have been missed in that meta-analysis. We searched all completed studies, both peer reviewed and non-peer reviewed, e.g. Ph.D dissertations, from January 2008 to October 2017.\n\nStudy selection is illustrated in the flow-diagram presented in Figure 1\n\nExcluded records were studies were the psychophysiological variables were analysed only after and not before the stimuli presentations (Jin et al., 2013) and with an unusual procedure (Tressoldi et al., 2015), i.e. using heart rate feedback to inform a voluntary decision to predict random positive or negative events.\n\nRecords excluded after the screening were studies where authors did not agree to share their data for different reasons (Baumgart et al., 2017; Modestino et al., 2011). Excluded studies revealed either statistically significant or trending evidence in support of the anticipation effect in most cases, thus reducing the concerns surrounding biased removal.\n\nThe references of the included studies are reported in Supplementary File 2.\n\nThe two co-authors agreed on the following coding variables: Authors; year of publication; participant selection: yes = selected according to specific criteria; no = selected without specific criteria; number of participants; number of trials; stimuli type; type of randomisation: pseudo or true random; psychophysiological signals, e.g. EEG, Heart Rate, etc.; anticipatory period; type of statistics; value of statistics and independently extracted them from the eligible studies. After the comparison, they discussed how to solve the inter-coder’ differences.\n\nOn the database we have added a note for each effect size, describing where we extracted the corresponding statistics in the original papers. The database along with all 18 papers are available from Tressoldi (2017). A summary of the selected studies along with their corresponding effect sizes, variance and standard error, is reported on Table S1 in the Supplementary File 3.\n\nApart from the overall effect, we chose to compare the following moderator variables, peer review (PeerRev, yes vs no) as a control of study quality. Given the low number of studies no further moderator analyses were carried out.\n\nThe standardized effect size d of each dependent variable, was estimated from the descriptive statistics (means, standard deviation and number of participants) when available. In all other cases, it was estimated by using the available summary statistics, i.e. paired t-test; Stouffer’s Z; etc. by using Lakens’ software (Lakens, 2013) and the function escalc () of the R package metaphor (Viechtbauer, 2017).\n\nAll effect sizes were then converted into the Hedges’ g and the corresponding variance by using the formulae suggested by Borenstein et al. (2009) estimating an average correlation of 0.5 between the dependent variables.\n\nGiven our choice of keeping (not averaging) all effect sizes when multiple dependent variables were analysed, we estimated the overall random model weighted effect size by using the robumeta package (Fischer et al., 2017) which implement a Robust Variance Estimation method when there are dependent effect sizes (Tanner-Smith & Tipton, 2014).\n\nIn order to control the reliability of the results, a second analysis was carried out by using a multilevel approach as suggested by (Assink & Wibbelink, 2016) implemented with the metafor package (Viechtbauer, 2010) and reported in the Table S2 in the Supplementary File 3.\n\nThe Bayesian meta-analysis was implemented with the brms package (Bürkner, 2017).\n\nA copy of the syntax is available here: https://doi.org/10.6084/m9.figshare.5661070.v1 (Tressoldi, 2017)\n\nEven if with our search activity we are quite sure to have reduced to a minimum the problem of publication bias, we performed a statistical estimation by using the Copas selection model which is recommended by Jin et al. (2015).\n\n\nResults\n\nStudies: Peer review papers: 8; Non -Peer review papers:10. Number of experiments: 26 contributed by 13 authors. Number of effect sizes: 34. Average number of participants: 97.5. Average anticipatory period: 3.5 seconds. Four studies were preregistered (see database).\n\nThe group analyses for males and females reported in three papers (Mossbridge, 2014; Mossbridge, 2015; Singh, 2009), were considered independent effect sizes.\n\nThe forest plot is presented in Figure 2. The summary of the frequentist multilevel random model analysis is presented in Table 1 compared with the results obtained by Mossbridge et al., whereas the summary of the Bayesian multilevel random model meta-analysis is presented in Table 2.\n\nn= number of experiments; ES= estimated effect size with corresponding 95% confidence intervals, p values; I2: effect sizes heterogeneity; τ2: effect size variance heterogeneity.\n\nRhat = ratio of the average variance of samples within each chain to the variance of the pooled samples across chains. CI – Credible Intervals.\n\nSensitivity analysis of the overall effect size, didn’t reveal any change from Rho 0 to Rho 1, suggesting that the degree of correlations among the dependent effect sizes don’t affect its magnitude.\n\nAnother “sensitivity analysis” was carried out excluding the Mossbridge and the Tressoldi studies in order to control whether different authors could obtain similar results. The main results of this analysis by using the same frequentist multilevel random model, is reported in Table 3.\n\nI2 = percentage of variation across studies that due to heterogeneity; τ2 = Tau2, variance of the true effect sizes. CI – Confidence Interval.\n\nBoth the frequentist and the Bayesian analyses support the evidence of an overall main effect of approximately .29, and a small difference between the peer and non-peer reviewed studies. These findings will be commented further in the discussion of the comparison with Mossbridge et al.\n\nThe search method used and the small number of people interested in this research field, guarantee that from an empirical point of view, any publication bias is almost absent.\n\nUnfortunately, there is no consensus about what tests are statistically more valid (Carter et al., 2017).\n\nAll the traditional tests, like the Fail-Safe, the Trim-and-Fill, the Funnel Plot have been criticized for their limitations (Jin et al., 2015; Rothstein, 2008). We hence applied the Copas selection model which is recommended by Jin et al. (2015).\n\nThe Copas selection model was implemented using the metasens package (Schwarzer et al., 2016). The results are presented in the Table 4. With this statistic, it emerges that there is no apparent statistical publication bias.\n\n\nDiscussion\n\nThis update of the Mossbridge et al. (2012) meta-analysis related to the so called predictive anticipatory activity (PAA) responses to future random stimuli, covers the years 2008- October 2017. Overall, we found 18 new studies describing a total of 34 effect sizes. Differently from the statistical approach of Mossbridge et al., in this meta-analysis we used a frequentist and a Bayesian multilevel model which allows an analysis of all effect sizes reported within a single study instead of averaging them.\n\nBoth the frequentist and the Bayesian analyses converged on similar results, making our findings quite robust. The overall effect size 0.29, 95% CI = 0.18 - 0.39, overlaps to that reported in the original paper: 0.21, 95% CI = 0.13–0.29, even if the heterogeneity is substantially higher: I2= 80.5 vs 27.4.\n\nThe high level of heterogeneity is expected considering the varieties of experimental protocols and the diversity of dependent variables, from heart rate to pupil dilation.\n\nFurthermore, we did not find substantial differences between peer and not-peer reviewed papers as in the original paper.\n\nWe found very interesting evidence of presentiment distilled from the conventional post-stimulus psychological research of Jolij and Bierman, who have performed a long series of experiments using a face detection paradigm. Additionally, the work of Kittenis found prestimulus effects from a conventional research program and pre-registered single-trial work of Mossbridge represent an important conceptual replication in countering both the use of questionable research practices and expectancy effects arguments.\n\nA promising development of this line of research is the development of paradigms that use software in real-time to predict meaningful future outcomes before they occur, e.g. (Franklin et al., 2014)\n\n\nConclusion\n\nThis update confirms the main results reported in Mossbridge et al. (2012) original meta-analysis and gives further support to the hypothesis of predictive physiological anticipation of future random events.\n\nThe limitations of the present meta-analysis are similar to most meta-analyses which include non pre-registered studies that cannot be controlled for the degree of freedoms in the methodology and data analysis in the course of their implementations, making them prone, for example, to the so-called “questionable research practices” (John et al., 2012).\n\nThe solution is that of prospective meta-analyses (Watt & Kennedy, 2017), based on preregistered studies where the methods and data analyses have been declared and made public beforehand.\n\n\nData availability\n\nUnderlying data for this meta-analysis is available from FigShare: https://doi.org/10.6084/m9.figshare.5661070.v1 (Tressoldi, 2017) under a CC BY 4.0 licence",
"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\nSupplementary material\n\nSupplementary File 1 – Completed PRISMA checklist.\n\nClick here to access the data.\n\nSupplementary File 2 – List of references used in this analysis.\n\nClick here to access the data.\n\nSupplementary File 3 – contains Table S1: Summary of the selected studies along with their corresponding effect sizes, variance and standard error. Table S2: results obtained with the multilevel approach suggested by Assink & Wibbelink, 2016.\n\nClick here to access the data.\n\n\nReferences\n\nAPA Publications and Communications Board Working Group on Journal Article Reporting Standards.Reporting standards for research in psychology: why do we need them? What might they be? Am Psychol. 2008; 63(9): 839–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAssink M, Wibbelink CJ: Fitting three-level meta-analytic models in R: A step-by-step tutorial. The Quantitative Methods for Psychology. 2016; 12(3): 154–174. Publisher Full Text\n\nAtkinson KM, Koenka AC, Sanchez CE, et al.: Reporting standards for literature searches and report inclusion criteria: making research syntheses more transparent and easy to replicate. Res Synth Methods. 2015; 6(1): 87–95. PubMed Abstract | Publisher Full Text\n\nBaumgart SL, Franklin MS, Jimbo HK, et al.: Prediction of truly random future events using analysis of prestimulus electroencephalographic data. AIP Conference Proceedings. AIP Publishing. 2017; 1841(1). Publisher Full Text\n\nBorenstein M, Hedges LV, Higgins JP, et al.: Introduction to Meta-Analysis. Chichester, UK: John Wiley & Sons, Ltd. 2009. Publisher Full Text\n\nBürkner PC: brms: An R Package for Bayesian Multilevel Models Using Stan. J Stat Softw. 2017; 80(1): 1–28. Publisher Full Text\n\nCarter EC, Schönbrodt FD, Gervais W, et al.: Correcting for bias in psychology: A comparison of meta-analytic methods. 2017. Publisher Full Text\n\nFischer Z, Tipton E, Zhipeng H: Package robumeta. Retrieved 3 August 2017, 2017. Reference Source\n\nFranklin MS, Baumgart SL, Schooler JW, et al.: Future directions in precognition research: more research can bridge the gap between skeptics and proponents. Front Psychol. 2014; 5: 907. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin Y, Yan K, Zhang Y, et al.: Gender differences in detecting unanticipated stimuli: an ERP study. Neurosci Lett. 2013; 538: 38–42. PubMed Abstract | Publisher Full Text\n\nJin ZC, Zhou XH, He J: Statistical methods for dealing with publication bias in meta-analysis. Stat Med. 2015; 34(2): 343–360. PubMed Abstract | Publisher Full Text\n\nJohn LK, Loewenstein G, Prelec D: Measuring the Prevalence of Questionable Research Practices With Incentives for Truth Telling. Psychol Sci. 2012; 23(5): 524–532. PubMed Abstract | Publisher Full Text\n\nLakens D: Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol. 2013; 4: 863. PubMed Abstract | Publisher Full Text | Free Full Text\n\nModestino EJ, Kelly EF, Ross Dunseath WJ, et al.: Anomalous physiological responses to local and remote emotive stimulation. Parapsychological Association Annual Convention, Curitiba. 2011.\n\nMoher D, Stewart L, Clarke M, et al.: Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015; 4(1): 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMossbridge JA: Single-trial presentiment experiment. KPU Registry ID N 1005. 2014. Reference Source\n\nMossbridge JA: Single-trial confirmatory presentiment experiment. KPU Registry ID NV 1018. 2015. Reference Source\n\nMossbridge JA, Tressoldi P, Utts J, et al.: Predicting the unpredictable: Critical analysis and practical implications of predictive anticipatory activity. Front Hum Neurosci. 2014; 8: 146. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMossbridge J, Tressoldi P, Utts J: Predictive physiological anticipation preceding seemingly unpredictable stimuli: A meta-analysis. Front Psychol. 2012; 3: 390. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRothstein HR: Publication bias as a threat to the validity of meta-analytic results. J Exp Criminol. 2008; 4(1): 61–81. Publisher Full Text\n\nSchwarzer G, Carpenter J, Rücker G: Package metasens. 2016. Reference Source\n\nSheehan DP: Preface and Acknowledgements: Quantum Retrocausation III. 2017; 1841(1): 10001. Publisher Full Text\n\nSingh PK: Personality correlates to electrophysiological measures of prestimulus response. Dissertation presented to Institute of Transpersonal Psychology, Palo Alto, California. 2009. Reference Source\n\nTanner-Smith EE, Tipton E: Robust variance estimation with dependent effect sizes: practical considerations including a software tutorial in Stata and spss. Res Synth Methods. 2014; 5(1): 13–30. PubMed Abstract | Publisher Full Text\n\nTressoldi PE, Martinelli M, Torre J, et al.: CardioAlert: A Heart Rate based decision support system for improving choices related to negative or positive future events. 2015. Publisher Full Text\n\nTressoldi P: Mossbridge's et al. 2012 meta-analysis update. Figshare. 2017. Publisher Full Text\n\nViechtbauer W: Conducting Meta-Analyses in R with the metafor Package. J Stat Softw. 2010; 36(3). Publisher Full Text\n\nViechtbauer W: The metafor Package. Retrieved 3 August 2017. Reference Source\n\nWatt CA, Kennedy JE: Options for Prospective Meta-Analysis and Introduction of Registration-Based Prospective Meta-Analysis. Front Psychol. 2017; 7: 2030. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "32577",
"date": "10 Apr 2018",
"name": "David Vernon",
"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\nIntroduction P.2, Line 21: Not sure I would agree with ‘body predicting moments ahead of time’ as this suggests understanding – try ‘reacting ahead of time’ or simply ‘physiological changes ahead…’ P.2: Para 2: the authors note that two paradigms were used, presentation of arousing/neutral stimuli or guessing tasks. Were any clear differences in PAA effects reported between these tasks? Also, given the ‘broad range of physiological measures’ used to assess such changes were there any key differences here? P.2, Para 2, final sentence: the ‘evidence from mainstream research’ – what specifically does this refer to? Behavioural effects? Ie changes in accuracy and/or response times and if so could do with a clear reference. P. 2, Para 3, line 9: ‘forwarded’ doesn’t make sense. Do you mean ‘proposed as a potential framework/theory’? P. 2, Para 3: Not sure I’d agree that using physiological markers to ‘predict’ future events is a ‘second objective’ measure. It is simply another way to view the same procedure. P. 2, Para 3: the vague references to ‘presentiment piggybacking onto mainstream research’ needs clarifying and supporting with references.\n\nMethods P.2, Para 1: need to identify the acronym ‘PRISMA’ after it is outlined.\n\nP. 2, Para 3, line 3: change ‘were’ to ‘where’ ……………….., line 4: change ‘Differently’ to ‘In addition,’\nAlso, what is the rationale for utilising a distinct eligibility criterion? It seems that prior research focused on testing for a pre-stim signal that would match the post-stim presentation. By not using this method you open yourself up to the criticism of widening the scope and also of looking for ‘any physiological change’ as opposed to one that would be specifically linked to the presentation of the target. The authors claim this is ‘more comprehensive’ but it could just as easily be seen as less conservative.\nP.3, line 4: change to ‘this type of investigation’ Line 8: change ‘investigations’ to ‘research’. Line 8: The point about studies possibly ‘missed’ by Mossbridge et al is not clear. What makes you think any studies were ‘missed’ and why did you then include the same time period – ie from 2008 to 2010 – if you are ‘adding’ to the data it would make sense to begin your inclusion time from 2010 unless you have evidence that some studies were ‘missed’?\n\nP.3, Para 4: line 1: change ‘were studies were’ to ‘were studies where’ P.3 – is it possible to say a bit more about why some authors did not agree to share their data – looks distinctly odd. P.4, Para 6: sentence referring to ‘Assink’ doesn’t make sense – unless you move the ref out of parenthesis and into the sentence. P.4: Change ‘The Bayesian’ to ‘A Bayesian’. And pull the sentence with syntax to the same paragraph. P. 4: Change: ‘Even if with our search activity we are quite….’ To ‘The robust search is likely to have reduced the probability of a publication bias occurring. Nevertheless, to test this a statistical estimation was conducted using the Copas selection model, as recommended by Jin et al’\n\nResults Keep tense to past ie peer reviewed not review.\nIt doesn’t make sense to compare data from the current review to Mossbridge et al ‘if’ both sets of data contain the same studies – as this would lead to obvious similarities etc. To an extent this seems to be addressed by the data in Table 3 but not made clearly – ie why not simply state that when X studies were excluded due to Y reasons the overall effect was still significant?\n\nI don’t see the moderation results for PeerRev reported here?\n\nThe reported ‘small difference between the peer reviewed and non-peer reviewed’ is vague and unhelpful. State clearly what was found – ie, are they ‘significantly different’ if not then they are not ‘different’ in any meaningful way.\n\nUnder ‘Publication bias’ I think para 2, 3 and 4 (which appears on P.6) should be joined as one single paragraph.\n\nDiscussion This is rather poor and reads like a list of points. There needs to be some discussion here not simply a repetition of the data. Ie – given this effect size how would the authors attempt to account for it? what are the implications of such a finding? Is there any scope for teasing out of the data any factors that may/may not influence the outcome – e.g., a possible relationship between the PAA and the various DV measures used?\n\nThe point relating to the work of Jolij and Bierman is again vague and unclear. What evidence precisely are you referring to here and how/why is this similar to the ‘psychological research’ [and what does this refer to?] What is the ‘conventional research program’ of Kittenis? How, exactly, does the single trial work of Mossbridge counter QRP?\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? No\n\nIs the statistical analysis and its interpretation appropriate? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "3821",
"date": "17 Jul 2018",
"name": "Patrizio Tressoldi",
"role": "Author Response",
"response": "Thank you for your detailed and constructive comments. Here it follows our replies to your main comments. Introduction P.2, Line 21: Not sure I would agree with ‘body predicting moments ahead of time’ as this suggests understanding – try ‘reacting ahead of time’ or simply ‘physiological changes ahead…’ P.2, Line 21: Not sure I would agree with ‘body predicting moments ahead of time’ as this suggests understanding – try ‘reacting ahead of time’ or simply ‘physiological changes ahead…’ Reply: we changed with “‘physiological changes ahead of time”. P.2: Para 2: the authors note that two paradigms were used, presentation of arousing/neutral stimuli or guessing tasks. Were any clear differences in PAA effects reported between these tasks? Reply: No Also, given the ‘broad range of physiological measures’ used to assess such changes were there any key differences here? Reply: No P.2, Para 2, final sentence: the ‘evidence from mainstream research’ – what specifically does this refer to? Behavioural effects? Ie changes in accuracy and/or response times and if so could do with a clear reference. Reply: Added reference P. 2, Para 3, line 9: ‘forwarded’ doesn’t make sense. Do you mean ‘proposed as a potential framework/theory’? Reply: replaced with \"proposed as a potential mechanism\". P. 2, Para 3: Not sure I’d agree that using physiological markers to ‘predict’ future events is a ‘second objective’ measure. It is simply another way to view the same procedure. Reply: changed as “another way..” P. 2, Para 3: the vague references to ‘presentiment piggybacking onto mainstream research’ needs clarifying and supporting with references. Reply: deleted this paragraph Methods P.2, Para 1: need to identify the acronym ‘PRISMA’ after it is outlined. Reply: added P. 2, Para 3, line 3: change ‘were’ to ‘where’ ……………….., line 4: change ‘Differently’ to ‘In addition,’ Reply: changed accordingly Also, what is the rationale for utilising a distinct eligibility criterion? It seems that prior research focused on testing for a pre-stim signal that would match the post-stim presentation. By not using this method you open yourself up to the criticism of widening the scope and also of looking for ‘any physiological change’ as opposed to one that would be specifically linked to the presentation of the target. The authors claim this is ‘more comprehensive’ but it could just as easily be seen as less conservative. Reply: We prefer the term more comprehensive because some experimental designs, e.g. hit guessing, don’t allow a post-stimulus physiological measure. However, all experimental designs tied the differential anticipatory physiological activity to two different outcomes, e.g. hits or misses. P.3, line 4: change to ‘this type of investigation’ Line 8: change ‘investigations’ to ‘research’. Reply: fixed. Line 8: The point about studies possibly ‘missed’ by Mossbridge et al is not clear. What makes you think any studies were ‘missed’ and why did you then include the same time period – ie from 2008 to 2010 – if you are ‘adding’ to the data it would make sense to begin your inclusion time from 2010 unless you have evidence that some studies were ‘missed’? Reply: after Mossbridge et al publication, we discovered that Singh, P.K. (2009), was missed. P.3, Para 4: line 1: change ‘were studies were’ to ‘were studies where’ Reply: fixed. P.3 – is it possible to say a bit more about why some authors did not agree to share their data – looks distinctly odd. Reply: the reasons for such decisions are confidential. P.4, Para 6: sentence referring to ‘Assink’ doesn’t make sense – unless you move the ref out of parenthesis and into the sentence. Reply: fixed P.4: Change ‘The Bayesian’ to ‘A Bayesian’. And pull the sentence with syntax to the same paragraph. Reply: fixed P. 4: Change: ‘Even if with our search activity we are quite….’ To ‘The robust search is likely to have reduced the probability of a publication bias occurring. Nevertheless, to test this a statistical estimation was conducted using the Copas selection model, as recommended by Jin et al’ Reply: fixed Results Keep tense to past ie peer reviewed not review. Reply: fixed It doesn’t make sense to compare data from the current review to Mossbridge et al ‘if’ both sets of data contain the same studies – as this would lead to obvious similarities etc. To an extent this seems to be addressed by the data in Table 3 but not made clearly – ie why not simply state that when X studies were excluded due to Y reasons the overall effect was still significant? Reply: we clarified that the Mossbridge and Tressoldi's studies were those included in this update. I don’t see the moderation results for PeerRev reported here? Reply: we think this analysis redundant given the data reported on Tables 1 and 2 The reported ‘small difference between the peer reviewed and non-peer reviewed’ is vague and unhelpful. State clearly what was found – ie, are they ‘significantly different’ if not then they are not ‘different’ in any meaningful way. Reply: we clarified that the means are different, but their precision estimate, i.e. confidence intervals, overlap. Under ‘Publication bias’ I think para 2, 3 and 4 (which appears on P.6) should be joined as one single paragraph. Reply: fixed Discussion This is rather poor and reads like a list of points. There needs to be some discussion here not simply a repetition of the data. Ie – given this effect size how would the authors attempt to account for it? what are the implications of such a finding? Is there any scope for teasing out of the data any factors that may/may not influence the outcome – e.g., a possible relationship between the PAA and the various DV measures used? Reply: we changed the discussion and the conclusion to include our evaluation of the status of art and the future of this phenomenon. What is the ‘conventional research program’ of Kittenis? Reply: omitted How, exactly, does the single trial work of Mossbridge counter QRP? Reply: we wrote “pre-registered single-trial work”. Preregistration of data analyses constraints the use of post-hoc data analysis flexibility."
}
]
},
{
"id": "35197",
"date": "05 Jul 2018",
"name": "Stephen Baumgart",
"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\nAddressing Major Criticisms\nThis is a controversial topic and careful consideration of objections is needed in a meta-analysis. Presentiment or Predictive Physiological Anticipation Studies (PAA) are typically criticized on these grounds (see, for example, Wagenmakers, Wetzels, Borsboom, Kievit, van der Maas, 2015):\nPhysical impossibility File-drawer effect Biases due to multiple comparisons or p-hacking\nAs for the first criticism, discussion of physical plausibility is beyond the scope of this meta-analysis and is left to the discretion of the authors. Nevertheless, apparent violations of our intuitions of time are found at the quantum level, such as the Wheeler Delayed-Choice experiment. It is not impossible that such effects may scale up to a macroscopic level in a not-yet-understood emergent process. I think the final two sentences of the introduction satisfy considerations of the physical impossibility objection and no changes are needed.\nThough file-drawer effects are frequently cited as a serious concern, the results section adequately discusses this issue. However, expert review is needed for this area (my response to \"Is the statistical analysis and its interpretation appropriate? \" should really be a combination of \"Partly\" and \"A qualified statistician is needed\".) I agree with the first sentence of the \"Publication Bias\" subsection that publication bias is not that serious of a concern because of the limited number of researchers and available funding.\nBy far the most serious concern is the third, that of multiple comparisons or p-hacking, which I do not believe is adequately addressed by either the discussion or conclusion sections. Two sentences in the conclusion are not sufficient to address this serious concern. I have included recommendations later in this review. I am aware the authors already know the following but by doing multiple analyses and only reporting a sub-sample of them a believer or supporter of a hypothesis could bias effect sizes up while a skeptic or opponent could bias effect sizes down (and none of these biases are necessarily intentional or even conscious).\nIn the context of PAA, serious sources of p-hacking concern are establishing baselines for electrophysiological data, deciding time regions for analysis, and methodologies for rejecting bad data and artifacts. For some physiological measurements, the problem is even worse. In Electroencephalography (EEG) studies, for example, a researcher could either study event-related potentials (ERPs), the spectral power densities of various oscillations, or the phases of such oscillations, or a host of other possible analyses. Considering oscillations, the frequency range of an analysis can also be freely selected. Additionally, a researcher could select different bandpass filters to use or even which section of the head is included in the analysis. This is in addition to the concerns with artifact rejection, time region, and baselining already discussed. With so many free parameters, a non-preplanned study is practically useless as hard evidence for an effect unless the statistical significance of the effect is high enough that it becomes implausible that the effect in question can be generated by tweaking free parameters. Even if the statistical significance is high, the effect size is still untrustworthy because an analyst could be tweaking parameters in an effort to improve the analysis or fix problems but is only homing in on statistical fluctuations. These concerns are one reason why I refused to include exploratory EEG research from my own lab in this meta-analysis.\nThe solution to the multiple analysis problem is to separate research into exploratory studies where adjustments can be made in analysis and pre-planned confirmatory studies. Some of the studies included in the meta-analysis are pre-planned confirmatory studies, which should be considered the only truly reliable results for estimates of effect size due to the concerns laid out in this review (even for confirmatory studies, mistakes by researchers could distort effect sizes but these mistakes may average out in the long run).\nMy recommended solutions for this paper are:\nMore discussion of the risks of p-hacking in biasing results in the discussion section\n\nSeparated analyses of pre-registered confirmatory studies and exploratory studies and discussion comparing the two\n\nFor exploratory studies in the study tables, include the experimenter expectation of whether the hypothesis will be verified (such as in Galak, LeBoeuf, Nelson, & Simmons, 2012)\n\nShow whether multiple comparison corrections were made for exploratory studies\nExploratory studies are necessary for advancing the field. But a meta-analysis should not include them without major caveats due to potential distortions of the effect size.\nI am aware the extra attention given to p-hacking risks in this research is not precedented by other fields but the small effect sizes and the major implications to our understanding of physics, psychology, and neuroscience PAA research engenders may justify additional caution be used. My colleagues and I discuss this further in Schooler, Baumgart, & Franklin, 2018.\nOther Comments\n“The presentiment hypothesis calls for a difference between arousing and neural pre-stimulus response and this is calculated across sessions” is not always true. For example, the hypothesis could also cover the difference between two different types of arousing stimulus (for example, auditory versus visual stimulus or two different types of visual stimulus).\n\nFurther discussion should be included for the observations mentioned of the second-to-last paragraph of the discussion; otherwise, it may be unclear why these studies are interesting as the paper asserts.\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? Partly",
"responses": [
{
"c_id": "3820",
"date": "17 Jul 2018",
"name": "Patrizio Tressoldi",
"role": "Author Response",
"response": "Thank you for your detailed and constructive comments. Here it follows our replies to your main comments. Though file-drawer effects are frequently cited as a serious concern, the results section adequately discusses this issue. However, expert review is needed for this area (my response to \"Is the statistical analysis and its interpretation appropriate? \" should really be a combination of \"Partly\" and \"A qualified statistician is needed\".) I agree with the first sentence of the \"Publication Bias\" subsection that publication bias is not that serious of a concern because of the limited number of researchers and available funding. Reply: we think we have a sufficient expertise in dealing with this problem. Furthermore we consulted with R.C.M. van Aert who is an expert on this topic. My recommended solutions for this paper are: More discussion of the risks of p-hacking in biasing results in the discussion section Separated analyses of pre-registered confirmatory studies and exploratory studies and discussion comparing the two Reply: we have added a direct comparison between preregistered and no-preregistered studies, see Table 4 and the paragraph “Preregistered vs No-preregistered studies” For exploratory studies in the study tables, include the experimenter expectation of whether the hypothesis will be verified (such as in Galak, LeBoeuf, Nelson, & Simmons, 2012) Reply: Unfortunately no one study checked this moderating variable, but our sensitivity analysis reported in Table 3, suggests that the experimenter expectation did not affect considerably the overall results. Show whether multiple comparison corrections were made for exploratory studies Reply: our choice to use multivariate analyses, partly reduce the impact of this procedure. “The presentiment hypothesis calls for a difference between arousing and neural pre-stimulus response and this is calculated across sessions” is not always true. For example, the hypothesis could also cover the difference between two different types of arousing stimulus (for example, auditory versus visual stimulus or two different types of visual stimulus). Reply: revised as “The presentiment hypothesis calls for a difference between the pre-stimulus responses of the two stimulus categories..” Further discussion should be included for the observations mentioned of the second-to-last paragraph of the discussion; otherwise, it may be unclear why these studies are interesting as the paper asserts. Reply: we expanded our conclusion as suggested."
}
]
}
] | 1
|
https://f1000research.com/articles/7-407
|
https://f1000research.com/articles/7-1096/v1
|
17 Jul 18
|
{
"type": "Software Tool Article",
"title": "Epiviz Web Components: reusable and extensible component library to visualize functional genomic datasets",
"authors": [
"Jayaram Kancherla",
"Alexander Zhang",
"Brian Gottfried",
"Hector Corrada Bravo",
"Jayaram Kancherla",
"Alexander Zhang",
"Brian Gottfried"
],
"abstract": "Interactive and integrative data visualization tools and libraries are integral to exploration and analysis of genomic data. Web based genome browsers allow integrative data exploration of a large number of data sets for a specific region in the genome. Currently available web-based genome browsers are developed for specific use cases and datasets, therefore integration and extensibility of the visualizations and the underlying libraries from these tools is a challenging task. Genomic data visualization and software libraries that enable bioinformatic researchers and developers to implement customized genomic data viewers and data analyses for their application are much needed. Using recent advances in core web platform APIs and technologies including Web Components, we developed the Epiviz Component Library, a reusable and extensible data visualization library and application framework for genomic data. Epiviz Components can be integrated with most JavaScript libraries and frameworks designed for HTML. To demonstrate the ease of integration with other frameworks, we developed an R/Bioconductor epivizrChart package, that provides interactive, shareable and reproducible visualizations of genomic data objects in R, Shiny and also create standalone HTML documents. The component library is modular by design, reusable and natively extensible and therefore simplifies the process of managing and developing bioinformatic applications.",
"keywords": [
"genomics",
"visualization",
"epigenetics",
"bioinformatics",
"web components"
],
"content": "Introduction\n\nThe complex and diverse genomic data sets require flexible software libraries and tools to perform integrative data exploration and analyses. Web-based genome browsers and genomic data visualization tools like the UCSC Genome Browser1 and the Integrated Genomics Viewer2 are developed for specific use cases i.e., integrative data exploration of a large number of datasets for a region in the genome. Genomic exploration of data on these platforms is usually track-based, where the data is aligned to a reference genome and visualized as a line track. Since these tools are developed for specific use cases, integration and extensibility of these visualizations and libraries is a challenging task.\n\nThe Web as a platform has been used to serve static HTML documents traditionally. The implementation of HTML5 and the newer APIs made the Web more of a platform that supports rich and dynamic web applications. But HTML is still restrictive and limited to the tags/elements defined as part of the markup language and is not extensible. Various existing frameworks like Vue.js and React have introduced modular components, but components built for one framework do not work with another framework. Newer web platform APIs and technologies like Web Components introduced a standards-based component model that allows developers to create custom HTML elements that are natively extensible and reusable. Custom components work across modern web browsers and can be used along with most JavaScript libraries or frameworks designed for HTML. Web Components provide the ability to natively extend, import and encapsulate HTML elements. This makes the process of creating and managing web applications easier and a much smoother process. These components are modular, making the code cleaner and less expensive to maintain compared to JavaScript libraries and frameworks like BioJs3.\n\nWe present the Epiviz Component Library, an open source reusable and extensible data visualization library and application framework for functional genomic data. Building upon the Web Component framework, we developed various HTML elements/tags as part of our design as shown in Figure 1. The visualization components (epiviz-charts) are the core of the library and render extensible and interactive track and feature-based charts. In addition to the chart library, we developed components for creating interactive genomic applications for different use cases and datasets. These include app components (epiviz-navigation and epiviz-environment) to coordinate interactions (linking data across visualizations to implement brushing and events) and manage layouts, datasource components including epiviz-data-source to manage data requests from a web server or WebSocket backend and epiviz-workspace for handling user authentication and to create shareable and reproducible visual analytical workspaces. The design of the component library is based on visualizations and features of the Epiviz4 web application for visual exploration and analysis of functional genomic data.\n\nThe epiviz web components architecture is organized into three categories: 1) visualization components is a library of extensible and interactive D3JS based chart components specifically designed for genomic data; 2) app components are responsible for managing the layouts, events arising from genomic coordinate navigation, linking data across visualization components to implement brushing and coordinating data requests across multiple charts; 3) datasource components manage requests to web server or WebSocket connections using the epiviz-data-source component. epiviz-workspace handles user authentication and saves the state of the app to a google firebase instance allowing users to create shareable and reproducible visual analytics workspaces.\n\nBioconductor5 is an open source community that develops bioinformatics software tools and pipelines. Ease of developing integrative analyses and a framework for interactive visualizations is one of the core infrastructure needs of the Bioconductor community. Since the web components introduced in this paper can be easily embedded or integrated with any web-based application, the library reduces the effort to visualize and create applications for genomic datasets encapsulated in Bioconductor infrastructure data representations. We developed an R/Bioconductor package, epivizrChart6 to visualize genomic data objects within HTML documents created using RMarkdown. We also integrated our components with Shiny7, a web application framework for R to interactively visualize functional genomic data.\n\n\nMethods\n\nVisualization components. epiviz-chart components are a collection of reusable and extensible data visualizations specifically designed for genomic data. The library provides multiple data visualizations for both location (visualizing data along the genome genes tracks (epiviz-genes-track) or line tracks (epiviz-line-track)) and feature based data (visualizing quantitative measurements like gene expression with scatterplots (epiviz-scatter-plot) and heatmaps (epiviz-heatmap-plot)). We use D3.js8 (version: 3.5.17) JavaScript library to render customizable and interactive charts.\n\nAn epiviz-chart component requires two attributes to render a visualization on the page 1) data attribute - a JSON (JavaScript Object Notation) representation of genomic data. 2) dimensions (or columns) from the data attribute to visualize. Figure 2 demonstrates the ease of embedding or adding an epiviz-chart to a HTML document or web application.\n\nEpiviz components can be inserted in any HTML page using tags defined by the component library (e.g., epiviz-json-scatter-plot in this example). Data is supplied to the chart via the json-data attribute of the HTML tag. In this example, we show a sample JSON object representing genomic data. In this figure, we are only showing the first 5 data points although the plot renders more visual objects. When used in conjunction with epiviz-data-source components, data can be queried from a web server or via a WebSocket connection through a corresponding assignment of the json-data attribute. Adding the epiviz element to the HTML page renders the interactive scatter plot.\n\nepiviz-chart components are reactive components that render the visualization only after the json-data attribute is initialized on the element. Any change to the json-data attribute triggers an event to revisualize the chart. Visualizations are extensible and easily customizable to define various settings and colors. To demonstrate the extensibility of the components, we created a component epiviz-genes-table extending epiviz-genes-track and displays a table of all the genes in the current genomic region (Supplementary File 1).\n\nIn addition to visualizing data, chart elements can also perform client-side operations on data sets/measurements. For example, if an epiviz-line-track is visualizing methylation data from multiple samples (tumor and normal), samples can be aggregated using a metric (mean, min, max, etc.) to visualize the difference in methylation between normal and tumor samples. Similarly, epiviz-heatmap-plot interactively and dynamically clusters data and renders the clustered dendrogram. Settings are available to change clustering type and the distance metrics.\n\nChart components provides performance optimizations for visualizing large amounts of data by precomputing and grouping overlapping data points to a single visual object on the chart. This minimizes the number of overlapping data points to visualize and reduce rendering time of charts.\n\nData model\n\nThe json-data attribute on an epiviz element is a JSON object that represents genomic data in a columnar format as shown in Figure 2. The required keys in the JSON are chr, start, end and data columns to visualize. Developers can also extend the epiviz-data-behavior element to implement custom data parsers and formats.\n\nLinked data selection/brushing\n\nChart components implement a linked data selection/highlighting (brushing) feature, to provide a quick overview and visually link the highlighted genomic region across all visualizations and datasets. The linking happens on the client side by finding positions that overlap with the highlighted region. In feature-based visualizations, for example in scatter plots and heatmaps, the visual objects on the chart are aggregated and mapped to multiple data objects across genomic regions. This mapping allows for implementing brushing and propagating events to other charts when using plots. In track-based visualization, events for brushing and selection are propagated based on the region (chr, start and end) in the chart.\n\nAnother essential part of the epiviz design is that data and plots are separated. Users can visualize multiple charts from the same data object without having to replicate the data. This way data queries are made by the data object and not per chart, which leads to a more responsive design of the system.\n\nepiviz-chart components are simple user interface (UI) elements. They cannot make data requests or can directly interact with other epiviz-* elements on the page. Chart elements create hover events that propagate up the document object model (DOM) hierarchy. To build interactive web applications or to coordinate interactions by linking data across charts, implement brushing and manage data requests across chart elements, we encapsulate charts inside app components.\n\nApp components. epiviz-app components are abstract components that 1) Manage layouts of multiple visualizations, 2) Coordinate interactions across charts by genomic position to implement brushing, and 3) Manage data requests.\n\nThere are two different types of epiviz-app elements -\n\nepiviz-environment elements are not linked to a specific genomic region. If a genomic region (chr, start & end attributes) is not initialized on the element, charts visualize the entire data set genome-wide. This helps identify patterns or interesting regions in the dataset and then investigate specific regions of interest.\n\nepiviz-navigation is a specific instance of epiviz-environment with genomic region linked to the element using the chr, start and end attributes. Navigation elements provide UI functionality to search for a gene/microarray probe (since we serve data from the Gene Expression Barcode project9) or update the location to a specific region of interest. Figure 3 (bottom) shows a navigation element with various charts when expanded. The top header bar contains functionality to navigate left/right and zoom in/out around the current genomic location. Navigation elements implement the usual genome browser interactions (pan, zoom, location input and gene name search). The chromosome location text box identifies the current location of the navigation element and can be updated to change the genomic region. Hovering over the chromosome location sends a brushing event to highlight this region across other charts encapsulated within the component. Navigation elements can be collapsed (as shown in the top panel) to allow users to flexibly focus on specific genomic regions of interest while providing an overview of other regions of interest. When collapsed, navigation components show an ideogram of the corresponding chromosome with an indication of the specific genomic region encompassed within the components (yellow rectangle). No data requests are made from charts within collapsed navigation components.\n\nIn this workspace, we explore data from the Epigenome Roadmap project in two genomic regions simultaneously (Epiviz Navigation components) along with a genome-wide scatterplot of gene expression (top left). The environment element is not constrained to a specific genomic region, and hence charts included within them visualize entire datasets. In this example, the scatter plot in the top left shows RNA-seq data for esophagus and colon tissues across the entire genome. EpivizNavigation components, on the other hand, are constrained to specific genomic regions. Given genomic regions or genes of interest in the dataset to further investigate, multiple navigation elements, each corresponding to distinct genomic regions can be added to the workspace. In this example, the navigation element at the bottom of the page visualizes (in order from top to bottom): 1) a genes track showing gene location span and strand, 2) a stacked-blocks track of ChIP-seq peaks in esophagus and colon across two different histone markers (H3K27me3 and H3k9me3), and 3) a line track that visualizes the fold change signal data for the same ChIP-seq data. The line track shows that the region around the gene “ATP6V1C1” shows a peak for H3K27me3 in Esophagus but not in Colon. The stacked blocks track compares the peak regions with other histone markers (H3K9me3). We can also investigate this region further by exploring methylation and gene expression data from these tissues by adding a navigation element (top right). The component library provides and interactive and integrative environment for genomic data exploration. This example workspace can be accessed at http://epiviz2.cbcb.umd.edu/#/epiviz-C7O4UmIb.\n\nApp elements coordinate events across charts, i.e., when a chart element is highlighted, an event is propagated to all other charts in the workspace (including those visualizing genome-wide data). App elements also manage layouts for positioning and resizing chart elements. The default grid layout splits the available width into six equal columns. When charts are added to a workspace, track-based charts extend across all the six columns but plot-based chart elements only span across two columns. App components have the functionality to navigate the genome and add new visualizations. Adding a new visualization opens a measurement browser, a UI interface that allows filtering and selection of measurements across different data sets.\n\nApp components can also detect if the application or page has an active web server or WebSocket connection initialized using the datasource components. If the page has no active datasource component, interactive features that generate data requests (for example – navigating to a new genomic region or adding new charts) are disabled.\n\nDatasource components. epiviz-data-source component provides functionality for the epiviz app components to interact with an active web server or a WebSocket connection. Datasource components require the API endpoint (provider-url) attribute where the web server or WebSocket is located and the provider-type attribute that specifies if it’s a web server or a WebSocket connection. When the user interacts with epiviz components, for example, adding a new visualization or navigating to a new genomic region, these interactions generate data requests that are eventually propagated and managed by the datasource elements.\n\nWebServer data provider\n\nWe developed a Python Flask (version 0.12.4) based data provider that queries genomic data stored in MySQL database and responds to data requests. The data provider enables summarization where we bin small regions together and average the value for the measurements. We see a significant improvement in draw times of charts by summarizing data as discussed in the Benchmarks section of this paper. We also implemented data import functions for commonly used Bioconductor datatypes like GenomicRanges, SummarizedExperiment, etc., in our R/Bioconductor epivizrData package.\n\nWebSocket data provider\n\nThe JavaScript data types that manage genomic data in epiviz components are designed similarly to Bioconductor data types. This enables easy integration and visualization of Bioconductor data objects using the visualization components. The R/Bioconductor epivizrChart package is an API to interactively visualize Bioconductor data objects. We discuss more about epivizrChart package in the Use case section of the paper.\n\nWorkspace component. epiviz-workspace component is built upon the Google Firebase infrastructure to manage user authentication, create shareable and reproducible visual analysis workspaces. Workspace components are easily reconfigurable and allow developers to customize this component to their firebase instance.\n\nWeb components are a set of standardized browser APIs still being implemented across various browsers. Web components implement the Shadow DOM feature, wherein the element defined by the component is rendered separately from the rest of the HTML document avoiding namespace collisions and is isolated to keep element styling and access private to the element. Web Components are natively supported in Chrome and Safari and are still in development in Mozilla Firefox and Microsoft Edge browser.\n\nEpiviz Components are developed using the Google Polymer library. For browsers without native web component support, the Google Polymer library provides polyfill that helps developers use components seamlessly with little performance overhead. It uses a dynamic loader to lazy load polyfill libraries for missing implementations. Documentation on attributes and methods in epiviz components is available from GitHub. The component library can visualize data by adding the chart tag to a HTML page with the data attribute as shown in Figure 2.\n\nThe epivizrChart package requires R version 3.4.0 or higher and packages from Bioconductor version 3.6 or higher. The memory requirements for using the epivizrChart package depends on the size of the dataset. However, for most use cases, a standard laptop will handle most applications visualizing data using the component library and the epivizrChart package. To visualize a Bioconductor data object, supply the supported object to the epivizChart() function.\n\n\nUse cases\n\nEpiviz2 is an interactive and integrative genome browser that sends requests to a Python Flask data provider and a MySQL database. Epiviz2 allows users to interactively explore and simultaneously visualize datasets across multiple genomic regions, a feature not available in most current genome browsers. The real advantage of the genome browser lies in the ability to visualize data from multiple regions of the genome or the entire dataset to identify genomic regions of interesting patterns or outliers. Users can then further explore and visualize annotations or measurements from other datasets in these regions to gain insights. Figure 3 illustrates this workflow of exploratory data analysis. The gene expression scatter plot is encapsulated inside the environment element and visualizes the entire dataset, whereas the navigation elements are linked to a specific genomic region. We also implemented a color by region for genome-wide scatter-plots, where visual objects in the scatter plot will be colored with a different color specific to each of the genomic regions shown in navigation elements. Our instance of the Epiviz2 application is hosted here.\n\nThe Epiviz2 instance we host at the University of Maryland contains data from the NIH Roadmap Epigenomics10 project. The NIH Roadmap Epigenomics Mapping Consortium leverages next-generation sequencing technologies to map DNA methylation, histone modifications, chromatin accessibility and small RNA transcripts in tissues selected to represent the normal counterparts of tissues and organ systems frequently involved in human disease. Our instance of the roadmap database contains DNA methylation, RNA seq, and histone modification (for markers: h3k9ac, h3k9me3, h3k27ac, h3k27me3) fold change and peak data for seven different tissue types – Breast Myoepithelial cells, Brain Hippocampus Middle, Lung, Liver, Sigmoid Colon, Pancreas and Esophagus. The corresponding data files are downloaded from Bioconductor’s AnnotationHub repository and imported into the MySQL database using the functions available in the epivizrData package.\n\nThe Bioconductor open source software community creates bioinformatics workflows and pipelines to analyze and visualize genomic data sets. To support interactive visualization of Bioconductor data objects, we developed an R/Bioconductor package epivizrChart, an API package to programmatically create and visualize genomic datasets using epiviz components without having to import data into a MySQL database. epivizrChart demonstrates the ease of integration with existing frameworks and can create interactive web pages or RMarkdown documents as shown in Figure 4. Integrating with a statistical and powerful state-of-the-art bioinformatics data analysis platform allows users to quickly explore, analyze and visualize genomic datasets with various packages available through Bioconductor.\n\nThis figure is part of an RMarkdown document and demonstrates the ease of integrating the visualizations from the Epiviz component library with existing frameworks. The epivizChart function infers the chart type based on the data object parameter. “Homo.sapiens” from the top panel is a UCSC Gene annotation object for human hg19 reference genome and is visualized by epivizChart as a genes track. tcga_colon_curves is a sample dataset from The Cancer Genome Atlas for colon tissue. This is a GRanges object and is visualized as a Line Track. The epivizrChart package can also programmatically create navigation elements. This enables interactions and brushing across the charts as shown in the bottom panel around the gene “PHLDB1”. A vignette describing more examples and use cases is available in the package either on the GitHub repository or through Bioconductor (http://bioconductor.org/packages/release/bioc/html/epivizrChart.html).\n\nonline vs offline. Using the epivizrChart package in an online mode creates an active WebSocket server and allows interactions between the components and the R-session. In online mode, components make data requests using the WebSocket connection. In offline mode, data is attached to the components and a standalone HTML page is generated. This allows researchers to create interactive, shareable and reproducible visualization documents.\n\nIntegration with Shiny. Shiny is a web application framework to create standalone web applications on a webpage or in an RMarkdown document. Since Shiny supports HTML, epiviz components can be embedded or integrated in Shiny applications or dashboards to interactively visualize genomic data. The vignette IntegrationWithShiny.Rmd in the epivizrChart package demonstrates 1) a simple application that integrates Shiny to visualize R/Bioconductor data objects using epiviz components 2) interactions with non-epiviz components in Shiny as shown in Figure 5.\n\nIn this Shiny application, we explore gene expression from the Gene Expression Barcode Project11 for colon, lung and breast tissues for tumor and normal samples as a heatmap. We visualize annotation tracks for the genes and position of CpG islands in the current region. We also integrated IGV with epiviz components and the igv track (bottom right) displays the gene position and the aligned illumina reads for HG01879 sample from 1000 genomes12 project. The IGV track queries the file directly to get data and visualize the reads. We also have a genomic location text box (top left) that is a non-epiviz component and can be used to interact with epiviz components within the Shiny application. Changing the location, updates the genomic region in the navigation element and all charts.\n\n\nBenchmarks\n\nWe use the google chrome headless puppeteer tool to measure request times and chart draw times to compare our web component implementation of the Epiviz2 application (with Python-MySQL backend) to the current Epiviz4 application (with PHP-MySQL backend). We compare the times by varying the genomic region on the scatter plot component (epiviz-scatter-plot) across two different backend implementations: 1) Summarized responses (current implementation), where we bin the genomic region into 2000 intervals and average the data values for the measurement within each interval and, 2) Unsummarized responses (previous epiviz implementation), where the entire dataset for the region is sent back to the UI. When visualizing large genomic regions, data points tend to overlap on scatter plots and other visualizations because of pixel and chart size limitations on the page. Summarizing reduces the draw times in rendering charts because of fewer overlapping points as shown in Figure 6. However, the response times for data requests have not changed significantly because the computation time for summarization is usually similar to the time taken to transfer the entire dataset in the unscaled implementation. The scripts for the benchmarks are available in the epiviz-chart GitHub repository. The benchmark scripts can also save a screenshot of the page rendered to make sure that the page is completely loaded and rendered.\n\nHere we compare average data request and data rendering time for continuous data along the genome to study the effect of summarizing data on the data backend across 10 runs. Lines for ‘unsummarized (u)’ correspond to the previous Epiviz implementation where all data within a genomic region is returned by the php-backend to the web browser client. Lines for ‘summarized (s)’ correspond to our new implementation of the python-data backend, where data summarization within genomic regions is performed in the backend. The left panel shows the mean draw times between these scenarios where we see a significant improvement in the draw times when the data is summarized in the backend. The bar plot in the right panel shows the total http time and is separated to show mean latency times and data transfer times. The number of bytes transferred for summarized and unsummarized backends is also displayed. The error bars represent one standard deviation away from the mean draw time in the left panel and mean http time in the right panel. We observe that the total http request time (summarization plus data transfer) is comparable to transfer time for the larger unsummarized data scenarios.\n\n\nDiscussion\n\nThe component library is an extension to our Epiviz web application for visualizing functional genomic data sets. The component library is our solution to creating reusable and extensible visualization elements that work with any modern web browsers. The value of a data visualization library depends on its usability and easy integration with existing web frameworks. Epiviz components can be integrated with any framework that supports HTML.\n\nThe Web has now become the platform for application development and the demand for modular, extensible and reusable frameworks like web components is on the rise. Since epiviz components are modular, we believe it simplifies the process of developing and managing genomic web applications. We also welcome developers to contribute to and extend our component library.\n\n\nConclusion\n\nTo our knowledge, the Epiviz component library is the first genomic data visualization library based on web components. The library provides an easy and efficient way for bioinformatics developers to add interactive data visualization features to their web applications or datasets with minimal programming experience. It is cross-platform, modular and runs on any modern web browser. We introduced our Epiviz2 web application to demonstrate the features and interactions that can be developed using the component library. We also showed the ease of integration with other frameworks by the R/Bioconductor epivizrChart package, that provides interactive, reproducible visualizations of data objects in R and also create interactive standalone HTML documents.\n\n\nFuture work\n\nOne of the advantages of web components is that HTML is now more readable. With a more declarative implementation, elements can be self-descriptive. We would like to implement a visualization grammar13 similar to ggvis as attributes/properties on the epiviz elements. We plan to further develop the library to extend our current set of visualizations and support various genomic data types including those implemented in Metaviz14 an interactive and statistical metagenomic data browser. We plan to implement canvas-based rendering of charts to scale and significantly reduce draw times especially when rendering large datasets.\n\n\nData availability\n\nThe Datasets used for the use case describing the Epiviz Application come from the NIH Roadmap Epigenomics Project. The data files are downloaded from Bioconductor’s AnnotationHub repository and imported into the MySQL database using the functions available in the epivizrData package. For the epivizrChart package, the datasets used are included as part of the package. The vignettes describing the use cases are also available on GitHub or through Bioconductor.\n\n\nSoftware availability\n\nEpiviz component library is open sourced and is available on GitHub. The collection of components discussed in this article are available at:\n\nepiviz charts - http://github.com/epiviz/epiviz-chart\n\nepiviz data - http://github.com/epiviz/epiviz-data-source\n\nepiviz workspace - http://github.com/epiviz/epiviz-workspace\n\nepiviz app - http://github.com/epiviz/epiviz-app\n\nThe scripts for benchmarks are available in the epiviz-charts repository. R/Bioconductor epivizrChart package is available either through Bioconductor (http://bioconductor.org/packages/release/bioc/html/epivizrChart.html) or GitHub (http://github.com/epiviz/epivizrChart). Both the respositories also contain the vignettes described in Figure 4 and Figure 5. The Python Flask API data provider is available at http://github.com/epiviz/epiviz-data-provider. Documentation is available at http://epiviz.github.io.\n\nArchived source code at the time of publication – https://doi.org/10.5281/zenodo.129999015\n\nSoftware license: MIT License.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the grant funded by the National Institutes of Health [R01GM114267].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Lan Tran for his early work on the development of the component library.\n\n\nSupplementary material\n\nSupplementary File 1 – Document containing extending-genes-track contains code that extends an existing epiviz chart component (epiviz-genes-track) and displays a table of genes and their genomic positions.\n\nClick here to access the data.\n\n\nReferences\n\nKent WJ, Sugnet CW, Furey TS, et al.: The Human Genome Browser at UCSC. Genome Res. 2002; 12(6): 996–1006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson JT, Thorvaldsdóttir H, Winckler W, et al.: Integrative genomics viewer. Nat Biotechnol. 2011; 29(1): 24–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGómez J, García LJ, Salazar GA, et al.: BioJS: an open source JavaScript framework for biological data visualization. Bioinformatics. 2013; 29(8): 1103–1104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChelaru F, Smith L, Goldstein N, et al.: Epiviz: interactive visual analytics for functional genomics data. Nat Methods. 2014; 11(9): 938–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGottfried B, Kancherla J, Corrada Bravo H: epivizrChart: R interface to epiviz web components. R package version 1.2.0. 2018. Publisher Full Text\n\nChang W, Cheng J, Allaire JJ, et al.: shiny: Web Application Framework for R. R package version 1.1.0. 2018. Reference Source\n\nBostock M, Ogievetsky V, Heer J: D³: Data-Driven Documents. IEEE Trans Vis Comput Graph. 2011; 17(12): 2301–2309. PubMed Abstract | Publisher Full Text\n\nMcCall MN, Jaffee HA, Zelisko SJ, et al.: The Gene Expression Barcode 3.0: improved data processing and mining tools. Nucleic Acids Res. 2014; 42(Database issue): D938–D943. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoadmap Epigenomics Consortium, Kundaje A, Meuleman W, et al.: Integrative analysis of 111 reference human epigenomes. Nature. 2015; 518(7539): 317–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTimp W, Bravo HC, McDonald OG, et al.: Large hypomethylated blocks as a universal defining epigenetic alteration in human solid tumors. Genome Med. 2014; 6(8): 61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSudmant PH, Rausch T, Gardner EJ, et al.: An integrated map of structural variation in 2,504 human genomes. Nature. 2015; 526(7571): 75–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWickham H: ggplot2. Media. 35, Springer New York, 2009. Reference Source\n\nWagner J, Chelaru F, Kancherla J, et al.: Metaviz: interactive statistical and visual analysis of metagenomic data. Nucleic Acids Res. 2018; 46(6): 2777–2787. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKancherla J, Zhang AY, Bradford A, et al.: epiviz/epiviz-chart: epiviz-chart. 2018. Data Source"
}
|
[
{
"id": "36143",
"date": "03 Aug 2018",
"name": "Bianca H. Habermann",
"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\nKancherla et al describe a set of re-usable web-components for the Epiviz tool, which are designed for visualizing genomic/epigenetic data in an integrative and interactive way. The components include the visualization components, app components and datasource components, are designed in a modular way and can be integrated with JavaScript libraries and other HTML frameworks. Moreover, there is an R-version available, which can be used with R / Shiny and to produce stand-alone HTML documents.\n\nThe general concept and design of the Epiviz web components is very good. With growing amounts of -omics data, we need easy-to-use, re-usable and extensible code for interactive data visualization, since visualizing large-scale data enhances our understanding of inherent trends and features of the datasets under analysis. We therefore have a growing need for easy-to-use and re-usable software for their visualization; re-useable and extensible tools like Epiviz are highly sought-after.\n\nThe manuscript is in general well written and the software has a nice look and feel. Though, there are a couple of points that should be addressed prior to indexing.\n\nWhile I find the description of the tools quite good and explicit, the description of the used biological examples, as well as the plots, is in my view too sparse. In fact, visualization only makes sense to the reader, if he/she understands, what is plotted and how to interpret the plots. With the manuscript at hand, as well as the user manual, this is not made easy. For instance, there is a general lack of axis labels on the scatter plots. E.g. Figure 2 and 3 show scatter plots, however from different data sources (peak positions/height or differential detection(?) on the genome, RNA-seq data (?)). What are the numbers on Figure 3 (ie the scatter plot of the RNA-seq data)? Are these rpkms? Log2 fold change? As the plots are quite useful and might be used directly for publication, this is a feature that should be implemented. Also, in the same Figure/example (methylation workspace following the given link of the legend to Figure 3), how does the RNA-seq data relate to the chip-seq/methylation data? Is there a possibility to see the identity of the genes, when hovering over the dots in the scatter plot? At least for the pre-selected example shown on the web-page, there is no brushing over from the zoomed-chromosome plot to the scatter plot and vice versa. However, this feature seems to work, when changing the chromosomal region. Is this due to the fact that the genes in the selected region are not present in the RNA-seq data? If so, the authors should think about changing the selected region for their demonstration. Also, what is the relationship of genes in the scatter plot that are highlighted together, when one is selected? Do they share the same enhancer/methylation peaks? It might be useful to provide information on the genes selected in the RNA-seq plot also via an info box or pop-up, which shows the name of the gene(s) and its(their) associated differential expression values. Is there the possibility to e.g. change the region of the chromosome-zoom/the chromosome view, when selecting dots in the scatter plot of the RNA-seq data? This seems currently not possible, however would be desirable. In the same example, the brushing of the chromosome view vs the zoomed-in chart does not work. When changing the genomic region, there is no highlighting any more in the chromosome view, so only the pre-selected region seems to work. This needs to be fixed. Finally, in the same plot, changing the chromosome for zoomed-in visualization has no effect on the whole-chromosome view, which I assume should also be updated. In my opinion, the whole-chromosome view is in its present form not really useful and could be omitted. In the same workspace (methylation workspace), there is a scatter plot at the bottom of the page; what do the numbers in the scatter plot refer to? See data labels problem already discussed above. On the binning of data values, the authors describe that the genomic region is binned into 2000 intervals. To which overall length does this refer to? Is always a fixed length chosen or can the user determine the length? Binning e.g. 100000 bp would then give quite different results than e.g. 1000000 bp.\n\nThere are also some errors in the manuscript, more specifically in the description of Figure 3. It currently states: “In this example, the navigation element at the bottom of the page visualizes (in order from top to bottom): 1) a genes track showing … “ Instead, it should read, I assume, as the genes track is at the bottom of the figure: “In this example, the navigation element at the bottom of the page visualizes (in order from bottom to top): 1) a genes track showing … “\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
},
{
"id": "36142",
"date": "08 Aug 2018",
"name": "Ian Holmes",
"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 library for integrative genomic data exploration, using modern web technologies, particularly HTML Web Components, with solid R integration. The interactive figures in the paper are very interesting, and show how the library can be used to build basic dashboards or RMarkdown pages with interactive charts and genome views. The dashboards look a little spartan, and a bit cryptic at times, but are a great illustration of the potential for this kind of thing and could develop into something really powerful. Plus, with the Bioconductor integration, they offer a lot of ready-to-go useful data for use cases around human genomics. All round a good contribution, although I think marred (very slightly) by some opening text in the paper which presents the software as being very general-purpose and detached from specific use cases - a style of presentation which I view as a mistake that risks detracting from a reader’s clear understanding of what the software can actually do. Modularity is a virtue, but developers often overestimate readers' level of interest in it.\nFor example, according to the authors, previous genome browsers have all been based on “specific use cases”, whereas Epiviz Web Components are more amenable to “integration and extensibility”. On closer reading, it turns out that the primary use case (the NIH Roadmap Epigenomics project) is just a bit further down in the paper, as is the platform (R/Bioconductor) currently required to load any useful data or the documentation vignettes. So I think the generality of the tool is possibly a bit hyperbolically described, in a way that risks obscuring the actual current uses (and software dependencies) of the tool.\nThe authors claim that, in Epiviz Web Components, they have developed the first genome browser that uses Web Components, which form a collection of browser features and APIs that have emerged from JavaScript libraries like React. Due to the fast-changing nature of the web, technologies such as this emerge with rapidity and regularity, and I find it plausible that this is the first genome browser to use these tools. They should be useful and it’s interesting to see them described in practice. The Epiviz Web Components tool is a useful addition to the bioinformatics visualization ecosystem, bridging R and the modern web.\nWith that said, I think perhaps that web bioinformatics software - and its associated publications - should be measured on several axes. Compliance with emerging web standards (such as Web Components) is certainly one such axis, but compliance with established bioinformatics standards would be another axis (how many formats/database schemas does the software support? how many other resources does it integrate with? what choices were made about which data sources to support, and which to omit? are those choices explained in the text? is there a guiding philosophy that can help inform readers who might be considering using this software?)\nIn terms of bioinformatics compatibility, bioconductor’s AnnotationHub seems intended to be the primary way of loading data into the browser, though it isn’t quite presented this way. Bioconductor is mentioned in the abstract as an example that was developed “to demonstrate the ease of integration with other frameworks”; I think it would be more accurate to say that Bioconductor is the only framework that this tool supports, and while the tool was written in such a way as to be hypothetically platform-independent, that hypothesis has not yet been seriously tested.\n\nEpiviz Web Components does have its own JSON formats for data, which is promising in terms of backing up the claim that data import would be straightforward, but as far as I can tell there are no tools to import common file formats (FASTA, GFF, etc) and it’s not clear whether there are any plans for developing such import tools, or if the Bioconductor dependency will in practice be permanent.\nThe paper’s interactive figures (Figures 3 & 4) are intriguing. I unfortunately had some problems installing the R from source (several dependencies of epivizrChart failed to build on my Mac Pro running macOS Sierra 10.12.6, R version 3.3.3), so I have not tried them on my local machine, only the JavaScript running in the web browser client. I found the captions to the Figures hard to follow: for example, I found myself wondering if there was a way for the user in the web browser to figure out how the components in Figure 3 (the genes track, stacked-blocks track, and line track) are linked, to use the same underlying data sets? In other words, does that linking only happen via configuration files and other back-end interfaces, or is it exposed to the end user? How were the markers H3K27me3 and H3k9me3 selected, why esophagus & colon, why drill into methylation and expression, what’s the biomedical back-story here? And just in terms of how to use the app to explore these cases as described, there could be more exposition. There are quite a lot of interactive buttons and menus and nested windows, so I was looking around for explanations of all that, passing by the documentation at https://epiviz.github.io/epiviz-chart/ (which is apparently oriented toward developers, not end-users, and is pretty minimal) and ending up at the https://epiviz.github.io/ video tutorial (“Epiviz quick tour”). I couldn’t find any thorough written description of the Epiviz web app, only that 3-minute YouTube video. I’m not very confident in my understanding of the totality of its capabilities after watching the video and looking at Figure 3. The potential to build pages out of RMarkdown, as shown in Figure 4, is pretty exciting.\nIn summary: I would suggest improving the article by adding more extensive descriptions of the biological use cases, and how they might be investigated using the tool. Make another video if you insist, but what I’d really be looking for is clearly-written tutorial text with figures. This will complement and enhance the current concise description of the software, and offer an alternative approach for a broader class of readers.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? No\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "36145",
"date": "14 Aug 2018",
"name": "Zhaohui Steve Qin",
"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, Kancherla and colleagues present the Epiviz Component Library, an open source reusable and extensible data visualization library and application framework for functional genomic data. According to the authors, the Epiviz component library is the first genomic data visualization library based on web components. The library provides an easy and efficient way for bioinformatics developers to add interactive data visualization features to their web applications or datasets with minimal programming experience.\n\nOverall, I found the tools describe in the paper very useful and promising. It makes complex visualization of multiple types of omics data easy and convenient. The new infrastructure is built upon the established epiviz tool that the group has developed earlier. This group is highly experienced in genomics data visualization and are developing state-of-art infrastructure using the latest technologies. The utilities of the tools great outweigh their limitations and shortcomings. I only have a few minor points.\n\nFigure 3. In this use case, I don’t quite understand the connection between the gene expression scatter plot and the other panel. For me, the figure is more like the demonstration of all types of plots it can produce. It is not obvious what biological insight can be gleaned from the plots. It is unclear what kind of biological question the user is trying to ask when making these plots. To be fair, this is always difficult. In most cases, discovery is made unintentionally, out of luck. Hence just demonstrating all the plotting capabilities is probably okay. If so, may be a systematic catalog of all the plots that can be produced will be helpful.\n\nFigure 6 is very informative and interesting. I noticed that the numbers of bytes transferred for summarized and unsummarized backends are exactly the same for 10K, 100K and 1M, but dramatically different for 10M, 100M and chr. Why is that?\n\nIt would be very helpful to the general audience if the authors can articulate in more details, and intuitively, the benefits and advantages of web components-based visualization library for genomics data, compared to the current technologies.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1096
|
https://f1000research.com/articles/6-2185/v1
|
28 Dec 17
|
{
"type": "Opinion Article",
"title": "A proposal to change the name of the NBPF/DUF1220 domain to the Olduvai domain",
"authors": [
"James M. Sikela",
"Frans van Roy"
],
"abstract": "We are jointly proposing a new name for a protein domain of approximately 65 amino acids that has been previously termed NBPF or DUF1220. Our two labs independently reported the initial studies of this domain, which is encoded almost entirely within a single gene family. The name Neuroblastoma Breakpoint Family (NBPF) was applied to this gene family when the first identified member of the family was found to be interrupted in an individual with neuroblastoma. Prior to this discovery, the PFAM database had termed the domain DUF1220, denoting it as one of many protein domains of unknown function. It has been PFAM’s intention to use “DUF” nomenclature to serve only as a temporary placeholder until more appropriate names are proposed based on research findings. We believe that additional studies of this domain, primarily from our laboratories over the past 10 years, have resulted in furthering our understanding of these sequences to the point where proposing a new name for this domain is warranted. Because of considerable data linking the domain to human-specific evolution, brain expansion and cognition, we believe a name reflecting these findings would be appropriate. With this in mind, we have chosen to name the domain (and the repeat that encodes it) Olduvai. The gene family will remain as NBPF for now. The primary domain subtypes will retain their previously assigned names (e.g. CON1-3; HLS1-3), and the three-domain block that expanded dramatically in the human lineage will be termed the Olduvai triplet. The new name refers to Olduvai Gorge, which is a site in East Africa that has been the source of major anthropological discoveries in the early-mid 1900’s. We also chose the name as a tribute to the scientists who made important contributions to the early studies of human origins and our African genesis.",
"keywords": [
"DUF1220",
"NBPF",
"protein domain",
"human brain evolution",
"copy number",
"gene duplication",
"genome evolution",
"Olduvai Gorge"
],
"content": "\n\nProtein domains are portable units within proteins that can serve important biological functions. They have been implicated in a broad range of key biological phenomena, including development, disease and evolution1,2. Here we jointly propose a new name for a protein domain of approximately 65 amino acids that has been previously termed NBPF3 or DUF12204. Our two labs independently reported the initial studies of this domain, which is encoded almost entirely within a single gene family. The name Neuroblastoma Breakpoint Family (NBPF) was applied to this gene family when the first identified member of the family was found to be interrupted in an individual with neuroblastoma3,5. Prior to this discovery, the PFAM database had termed the domain DUF1220, denoting it as one of many protein domains of unknown function6. It has been PFAM’s intention to use “DUF” nomenclature to serve only as a temporary placeholder until more appropriate names are proposed based on research findings. We believe that additional studies of this domain, primarily from our laboratories over the past 10 years, have resulted in furthering our understanding of these sequences to the point where proposing a new name for this domain is warranted.\n\nKey findings relevant to assigning a name to the domain are as follows:\n\n1. The domain has been repeatedly linked with human-specific evolution. The haploid human genome is estimated to encode approximately 300 copies of the NBPF/DUF1220 domain, while the copy number for other species is substantially lower: great apes 90–120, monkeys 30–40, and all other mammalian species 1–97,8. The increase in humans (at least 165 additional human copies) represents the largest human lineage-specific copy number increase of any coding region in the genome. These findings, involving the copy number of a protein coding domain, provide strong support for an involvement in human-specific evolution.\n\n2. The domain has been linked with human brain evolution and cognitive function. Over the past 10 years we have published several papers on NBPF/DUF1220 protein domains and the NBPF gene family. These have implicated the copy number of the domain in human brain evolution7,9–11, brain size-related phenotypes7,9–11, brain disorders (autism/schizophrenia/micro- and macrocephaly)9,12–14, and measures of cognitive function13,15. Also, our finding of a robust linear association between NBPF/DUF1220 copy number and brain size across primate species was confirmed by an independent study16.\n\nGiven the above research findings, a new name for this protein domain that is related to human-specific evolution would be appropriate. We believe a name that would do this is “Olduvai” (ohl’-du-vi) (when necessary it can be abbreviated as “ODV”). This name refers to Olduvai Gorge, which is located in the rift valley of Eastern Africa. Olduvai has been the site of key paleoanthropological discoveries related to human origins and has been called “the Cradle of Mankind”, and “the Grand Canyon of Human Evolution”17. Deposits at the gorge are estimated to cover a time span from 2.1 million to 15,000 years ago, and the fossil remains that have been identified there are thought to represent more than 60 hominins (members of the human lineage). These findings are believed to constitute the most continuous known record of human evolution over the past 2 million years, and the longest known archaeological record of the development of stone-tool industries. Olduvai Gorge was designated part of a UNESCO World Heritage site in 197918. Just as the protein domain appears to be important to human-specific evolution, so too, Olduvai Gorge has provided key insights into human’s evolutionary origin. We believe both are central to the story of what made us human.\n\nFinally, we have also chosen this name because it reflects an appreciation for the important contributions of the scientists who made major anthropological discoveries in Africa in the early/mid-20th century that stimulated further research into human origins and our African genesis.\n\nWhile we believe that the domain and repeat should now be called “Olduvai”, we also propose that, for now, the gene family name should remain NBPF. In summary, the NBPF domain and DUF1220 domain will be termed the Olduvai domain, and the NBPF repeat and DUF1220 repeat will be termed the Olduvai repeat. The primary gene family that encodes these sequences will continue to be called NBPF, and the primary domain subtypes will retain established nomenclature (CON1-3, HLS1-3)8. The three-domain block, composed of HLS1, HLS2 and HLS3 subtypes, that is tandemly repeated within several human NBPF genes and responsible for the great majority of additional human copies of the domain, will be called the Olduvai triplet.\n\nThe Olduvai domain hyper-amplification in the human lineage was one of the most extreme and rapid copy number expansions in the human genome, and we look forward to additional studies that may provide further insights into the role this protein domain family plays in human disease and evolution.",
"appendix": "Competing interests\n\n\n\nJMS is founder of GATC Science, LLC, a biotech company involved in genomic research.\n\n\nGrant information\n\nJMS is funded by NIH R01 grant MH108684. FVR is supported by the Foundation Against Cancer – Belgium, the Research Foundation – Flanders (FWO-Vlaanderen), and the Queen Elisabeth Medical Foundation (G.S.K.E.), Belgium.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank Jonathon Davis, Veronica Searles Quick, Ilea Heft, Laura Dumas, Vanessa Andries and Karl Vandepoele for constructive discussions.\n\n\nReferences\n\nMark M, Rijli FM, Chambon P: Homeobox genes in embryogenesis and pathogenesis. Pediatr Res. 1997; 42(4): 421–429. PubMed Abstract | Publisher Full Text\n\nLi W-H: Molecular Evolution. Sinauer Associates, Sunderland, Massachusetts; 1997.\n\nVandepoele K, Van Roy N, Staes K, et al.: A novel gene family NBPF: Intricate structure generated by gene duplications during primate evolution. Mol Biol Evol. 2005; 22(11): 2265–2274. PubMed Abstract | Publisher Full Text\n\nPopesco MC, Maclaren EJ, Hopkins J, et al.: Human lineage-specific amplification, selection, and neuronal expression of DUF1220 domains. Science. 2006; 313(5791): 1304–1307. PubMed Abstract | Publisher Full Text\n\nVandepoele K, Andries V, Van Roy N, et al.: A constitutional translocation t(1;17)(p36.2;q11.2) in a neuroblastoma patient disrupts the human NBPF1 and ACCN1 genes. Bielinsky A-K, ed. PLoS One. 2008; 3(5): e2207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBateman A, Coin L, Durbin R, et al.: The Pfam protein families database. Nucleic Acids Res. 2004; 32(Database issue): D138–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDumas L, Sikela JM: DUF1220 domains, cognitive disease, and human brain evolution. Cold Spring Harb Symp Quant Biol. 2009; 74: 375–382. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Bleness MS, Dickens CM, Dumas LJ, et al.: Evolutionary history and genome organization of DUF1220 protein domains. G3 (Bethesda, Md). Genes/Genomes/Genetics. 2012; 2(9): 977–986. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDumas LJ, O’Bleness MS, Davis JM, et al.: DUF1220-domain copy number implicated in human brain-size pathology and evolution. Am J Hum Genet. 2012; 91(3): 444–454. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeeney JG, Dumas L, Sikela JM: The case for DUF1220 domain dosage as a primary contributor to anthropoid brain expansion. Front Hum Neurosci. 2014; 8: 427. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeeney JG, Davis JM, Siegenthaler J, et al.: DUF1220 protein domains drive proliferation in human neural stem cells and are associated with increased cortical volume in anthropoid primates. Brain Struct Funct. 2015; 220(5): 3053–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis JM, Searles VB, Anderson N, et al.: DUF1220 dosage is linearly associated with increasing severity of the three primary symptoms of autism. PLoS Genet. 2014; 10(3): e1004241. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis JM, Searles Quick VB, Sikela JM: Replicated linear association between DUF1220 copy number and severity of social impairment in autism. Hum Genet. 2015; 134(6): 569–575. PubMed Abstract | Publisher Full Text\n\nSearles Quick V, Davis JM, Olincy A, et al.: DUF1220 copy number is associated with schizophrenia risk and severity: Implications for understanding autism and schizophrenia as related diseases. Transl Psychiatry. 2015; 5(12): e697. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis JM, Searles VB, Anderson N, et al.: DUF1220 copy number is linearly associated with increased cognitive function as measured by total IQ and mathematical aptitude scores. Hum Genet. 2015; 134(1): 67–75. PubMed Abstract | Publisher Full Text\n\nZimmer F, Montgomery SH: Phylogenetic analysis supports a link between DUF1220 domain number and primate brain expansion. Genome Biol Evol. 2015; 7(8): 2083–2088. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArdrey R: African Genesis. Collins, London, UK; 1961. Reference Source\n\nOlduvai Gorge. In: Encyclopedia Britannica. Reference Source"
}
|
[
{
"id": "29376",
"date": "02 Jan 2018",
"name": "Alex Bateman",
"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 a straightforward proposal to rename the NBPF/DUF1220 domain to the Olduvai domain. This is welcome because the domain has been referred to by two different names in the past. The authors are the major players in the field, who have used the divergent nomenclature and thus will be able to ensure that the proposed naming does become the standard. The authors had already contacted me to ask about having the domain renamed within the Pfam database, which has now been done (although not visible on the public website yet). Perhaps calling this paper a proposal is a bit weak. The authors should consider renaming the article to “Changing the name of the NBPF/DUF1220 domain to the Olduvai domain”.\nThere are no agreed upon standard naming schemes for protein domains or families. To date the most popular schemes are based on naming after one or more proteins that contain the domain, sometimes based upon the initial letters of these members. A few domains are named after scientists. I don’t recall any domains or families named after places. However, the renaming makes sense within the context of the evolution of this domain and human evolution.\nI think it would be worth mentioning the reasons that the NBPF nomenclature is retained for the gene family. I understand this is to fit with the HGNC guidelines.\nMinor points\nPlease change PFAM to Pfam throughout the text.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "29374",
"date": "17 Jan 2018",
"name": "Christian J.A. Sigrist",
"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 proposes to change the name of a protein domain from NBPF (Neuroblastoma Breakpoint Family) to Olduvai, arguing that this new name illustrates the fact that the domain copy number has been linked with human evolution. Although, it is in principle not recommended to multiply the names for a domain, the fact that the change is submitted by two major players in the field makes the proposal acceptable as it should rapidly spread in future works on this promising domain. In addition, the limited distribution of the domain should facilitate its diffusion. However, it should be noted that although having a very limited distribution found in roughly two protein families, i.e. NBPF and myomegalin (PDE4DIP), neither of the current names, NBPF or DUF1220, is commonly used by the myomegalin community. In this respect, I think that a paragraph presenting the myomegalin family and discussing the presence of a single Olduvai “repeat” amongst its members should be added prior to indexing.\nThe previous name, NBPF, was derived from a breakpoint in the first member identified and was also applied to the protein family. In the future, the use of the new name Olduvai only for the repeat will avoid the confusion between NBPF designating both the protein family and the repeat.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "29429",
"date": "18 Jan 2018",
"name": "Elspeth A. Bruford",
"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 brief opinion article describes the authors' proposal to rename a domain found in a discrete group of proteins, which was originally named DUF1220 (domain of unknown function 1220) and subsequently reassigned as NBPF. The NBPF nomenclature referenced the fact that this domain is found in proteins encoded by a family of genes initially reported as the \"Neuroblastoma BreakPoint Family\"; however, as human gene nomenclature tries to move away from directly referencing specific phenotypes and conditions, the reference to neuroblastoma has been removed from the names for the NBPF genes, and they are now simply the \"NBPF family\".\nInterestingly, as well as this family the only other protein this domain is found in is the large \"myomegalin\" protein in mammals, encoded by the PDE4DIP gene. I think the fact that this is not mentioned anywhere in the article is perhaps an oversight on the part of the authors which could be remedied.\nIt should be noted that the authors have already contacted PFAM to request this change and HGNC were also consulted and agreed this update would not create problems from a gene naming perspective. As we said then, we think it would be more informative for the new domain name to reflect functional information as opposed to something more esoteric, but the authors are not advocating any parallel change in the NBPF gene family nomenclature which is quite widely used for this family. A parallel change to the gene names would perhaps be more difficult to justify. Therefore we told the authors that we see no issue with their proposal.\nImportantly, this article represents the collaborative effort of two independent groups, who easily could be viewed as \"rivals\". Therefore this article effectively shows that such collaborations are possible and beneficial to the scientific community at large; this kind of activity should be encouraged wherever possible and I applaud the authors for making the effort to work productively together.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/6-2185
|
https://f1000research.com/articles/7-1088/v1
|
16 Jul 18
|
{
"type": "Correspondence",
"title": "PDZD8 is not the ‘functional ortholog’ of Mmm1, it is a paralog",
"authors": [
"Jeremy G. Wideman",
"Dario L. Balacco",
"Tim Fieblinger",
"Thomas A. Richards",
"Dario L. Balacco",
"Tim Fieblinger",
"Thomas A. Richards"
],
"abstract": "Authors of a recent paper demonstrate that, like ERMES (ER-mitochondria encounter structure) in fungal cells, PDZD8 (PDZ domain containing 8) tethers mitochondria to the ER in mammalian cells. However, identifying PDZD8 as a “functional ortholog” of yeast Mmm1 (maintenance of mitochondrial morphology protein 1) is at odds with the phylogenetic data. PDZD8 and Mmm1 are paralogs, not orthologs, which affects the interpretation of the data with respect to the evolution of ER-mitochondria tethering. Our phylogenetic analyses show that PDZD8 co-occurs with ERMES components in lineages closely related to animals solidifying its identity as a paralog of Mmm1. Additionally, we identify two related paralogs, one specific to flagellated fungi, and one present only in unicellular relatives of animals. These results point to a complex evolutionary history of ER-mitochondria tethering involving multiple gene gains and losses in the lineage leading to animals and fungi.",
"keywords": [
"Pdz8",
"ERMES",
"paralog",
"ortholog",
"evolution"
],
"content": "\n\nHirabayashi et al.1 show that PDZD8 (PDZ domain containing 8) is an SMP domain-containing protein involved in ER-mitochondria tethering and regulation of Ca2+ dynamics in mammalian neurons. We do not dispute the authors’ interesting results with respect to PDZD8 function in mammalian cells. However, claims made by the authors regarding the evolutionary relationship of this gene with fungal homologs represents a misuse of the term ortholog, where the term homolog or, more appropriately, paralog is correct. This misuse has consequences for interpreting data in the paper, including in the domain-swapping experiments conducted as part of the complementation assay. The misclassification affects how their data should be interpreted and results in confused explanations of how ER-mitochondria contact sites evolved in animals and fungi. Approaching the data with correct terminology alleviates these problems and illuminates interesting possibilities about trait evolution in fungi and animals.\n\nOrthology and paralogy are special cases of homology (Figure 1). Orthologous genes arise by speciation events. Mouse α-haemoglobin is orthologous to human α-haemoglobin. Paralogous genes arise by gene duplication events. Duplication of a globin gene in an ancestor of vertebrates gave rise to two haemoglobin families in which α- and β-haemoglobin subsequently evolved. Therefore, α- and β-haemoglobin, regardless of which organisms they appear in, are paralogs. In this simple case, the orthologous proteins perform the same function in different organisms (i.e. they are isofunctional orthologs2). However, orthologs can diverge and perform different functions in different lineages (heterofunctional orthologs). An example of heterofunctional orthologs is animal Miro (mitochondrial rho GTPase), which, in yeast, is called Gem1 (GTPase EF-hand protein of Mitochondria) and has been suggested to interact with ERMES (ER-mitochondria encounter structure) in fungi3,4. Miro is important for microtubule-dependent mitochondrial motility in animals (reviewed by Reis et al.5). However, although a Miro ortholog is present in all eukaryotic lineages, it does not function in microtubule-dependent mitochondrial motility in yeast, Neurospora, Arabidopsis, or Dictyostelium5–7. Thus, functionality alone does not imply orthology.\n\nOrthologs are a consequence of speciation, whereas paralogs are a consequence of gene duplication. Human α- and β-haemoglobin share 43% identity whereas Human α-haemoglobin and Mouse α-haemoglobin share 87% identity. When performing phylogenetic analyses, the orthologous α-haemoglobin subunits from different animals branch together separate from their paralogs, the β-haemoglobin subunits. Taken together, all haemoglobin subunits are homologs.\n\nAny SMP domain is homologous to any other SMP domain because they have shared ancestry, but only those arising via speciation are orthologs. To demonstrate that PDZD8 and Mmm1 (maintenance of mitochondrial morphology protein 1) are “functional orthologs”, Hirabayashi et al.1 swap the SMP domain from PDZD8 into the Mmm1 protein in Saccharomyces cerevisiae thereby rescuing the defects imparted by loss of Mmm1. But, they also successfully rescue Mmm1 defects by swapping the SMP domain from its paralog, S. cerevisiae Mdm12 (mitochondrial distribution and morphology protein 12), into the S. cerevisiae Mmm1 protein. The domain swapping experiments should not be interpreted to mean that the proteins (or even the domains) carry out the same function. Instead, these experiments suggest that the paralogous SMP domains from PDZD8, Mmm1, and Mdm12 are biochemically isofunctional in this specific scenario (i.e. the SMP domains can carry out similar functions when they are placed very specifically into the S. cerevisiae Mmm1 protein). However, it must be stressed that this does not mean that the proteins themselves are isofunctional. The fact that mammalian PDZD8 and similar proteins from other animals have long C-terminal extensions containing accessory domains (e.g. a PDZ domain and cysteine-rich C1 domains), while Mmm1 and Mdm12 do not, suggests these proteins have different or additional mechanisms of function. Thus, we cannot say that the full proteins are isofunctional homologs and especially not isofunctional orthologs.\n\nThe fact that PDZD8 is a homolog of SMP domain-containing proteins like Mmm1 has been demonstrated previously8–11; but it has also been demonstrated that PDZD8 is not an ortholog of ERMES components (Figure 2, also see Wideman et al.11). This can be seen very clearly in the SMP-domain proteins of Capsaspora owczarzaki, which include orthologs of all the ER-mitochondrial contact site SMP proteins (PDZD8, Mmm1, Mdm12, Mdm34). This organism is closely related to animals12, but still retains both a complete ERMES complex and PDZD8 and represents a future model for investigating their differential functions. Interpreting the work of Hirabayashi et al.1 from a comparative perspective demonstrates that mitochondria-ER tethering is a function that is conserved deep within this gene family predating the duplication that gave rise to PDZD8 and Mmm1, and indeed, all known ERMES paralogs. This is important because it allows us to phrase the next questions, ‘how have different paralogs expanded, or changed, in function? (is there evidence of heterofunctionality?)’ and ‘why have some evolutionary lineages maintained multiple paralogs while others have tolerated loss?’. When uninformative terms like ‘functional ortholog’ are used, especially in exceptional papers like Hirabayashi et al.1, biologists interested in explaining functional differences between organisms are being misled. We stress that only comparative evolutionary methods can identify the starting data for phrasing the above questions and functional data cannot be used when inferring orthology or paralogy. Once questions of orthology and paralogy have been resolved, questions of functional conservation and divergence can be addressed at the bench.\n\nPDZD8 groups to the exclusion of Mmm1, Mdm12, Mdm34, Nvj2 and two unnamed paralogs. SMP domain-containing proteins were gathered from diverse opisthokonts (i.e. animals, fungi, and closely related protists) and their sister species Thecamonas trahens (sequences were obtained from public databases (Joint Genome Institute and NCBI) as well as from recently available genomes and transcriptomes13,14. The SMP domains were aligned and subjected to phylogenetic reconstruction using RaxML v8.2 (100 pseudoreplicates using the LG model) and MrBayes v3.2 (1 million generations using the WAG model) as in 11. Sequences and alignments are available at https://github.com/mbzdlb/PDZD8. Six strongly supported paralogs, including PDZD8, Nvj2, Mdm12, Mdm34, and two unnamed paralogs comprising sequences from flagellated fungi and opisthokont protists, were recovered. Fungal Mmm1 is recovered in a strongly supported clade whereas some proteins designated as Mmm1 previously11 do not. However, sequence inspection identified motifs outside the SMP domains present in both fungal and non-fungal Mmm1s, strongly suggesting that proteins designated as Mmm1 are orthologous. Mmm1 proteins lacking these motifs may represent truncated or mispredicted proteins. Similarly, some predicted Pdz8 proteins lack Pdz domains and C-terminal extensions. Human PDZD8 is considered paralogous to Mmm1 because it groups separately and can be found in organisms that already contain Mmm1-like proteins (e.g. Capsaspora owczarzaki). Motifs were visualized using WebLogo15. One asterisk indicates the presence of Mmm1-specific short motif, two asterisks indicate both the short and the transmembrane motifs are present. Support values are as iconized in inset key (MrBayes/RAxML).\n\n\nData availability\n\nAll sequences were downloaded from publicly available databases (e.g. Joint Genome Institute or NCBI) or obtained from published transcriptomic and genomic data13,14. The sequences used to generate Figure 2 are downloadable at https://github.com/mbzdlb/PDZD8.",
"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\nSupplementary material\n\nSupplementary File 1. List of organisms and accession numbers (with database) used to generate Figure 2. The corresponding sequences can be found at https://github.com/mbzdlb/PDZD8 or from the indicated online database.\n\nClick here to access the data.\n\n\nReferences\n\nHirabayashi Y, Kwon SK, Paek H, et al.: ER-mitochondria tethering by PDZD8 regulates Ca2+ dynamics in mammalian neurons. Science. 2017; 358(6363): 623–630. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJensen RA: Orthologs and paralogs - we need to get it right. Genome Biol. 2001; 2(8): INTERACTIONS1002. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKornmann B, Osman C, Walter P: The conserved GTPase Gem1 regulates endoplasmic reticulum-mitochondria connections. Proc Natl Acad Sci U S A. 2011; 108(34): 14151–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStroud DA, Oeljeklaus S, Wiese S, et al.: Composition and topology of the endoplasmic reticulum-mitochondria encounter structure. J Mol Biol. 2011; 413(4): 743–50. PubMed Abstract | Publisher Full Text\n\nReis K, Fransson Å, Aspenström P: The Miro GTPases: at the heart of the mitochondrial transport machinery. FEBS Lett. 2009; 583(9): 1391–1398. PubMed Abstract | Publisher Full Text\n\nWideman JG, Lackey SW, Srayko MA, et al.: Analysis of mutations in Neurospora crassa ERMES components reveals specific functions related to β-barrel protein assembly and maintenance of mitochondrial morphology. PLoS One. 2013; 8(8): e71837. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVlahou G, Eliáš M, von Kleist-Retzow JC, et al.: The Ras related GTPase Miro is not required for mitochondrial transport in Dictyostelium discoideum. Eur J Cell Biol. 2011; 90(4): 342–355. PubMed Abstract | Publisher Full Text\n\nLee I, Hong W: Diverse membrane-associated proteins contain a novel SMP domain. FASEB J. 2006; 20(2): 202–206. PubMed Abstract | Publisher Full Text\n\nKopec KO, Alva V, Lupas AN: Homology of SMP domains to the TULIP superfamily of lipid-binding proteins provides a structural basis for lipid exchange between ER and mitochondria. Bioinformatics. 2010; 26(16): 1927–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlva V, Lupas AN: The TULIP superfamily of eukaryotic lipid-binding proteins as a mediator of lipid sensing and transport. Biochim Biophys Acta. 2016; 1861(8 Pt B): 913–923. PubMed Abstract | Publisher Full Text\n\nWideman JG, Gawryluk RM, Gray MW, et al.: The ancient and widespread nature of the ER-mitochondria encounter structure. Mol Biol Evol. 2013; 30(9): 2044–9. PubMed Abstract | Publisher Full Text\n\nFerrer-Bonet M, Ruiz-Trillo I: Capsaspora owczarzaki. Curr Biol. 2017; 27(17): R829–R830. PubMed Abstract | Publisher Full Text\n\nTorruella G, de Mendoza A, Grau-Bové X, et al.: Phylogenomics Reveals Convergent Evolution of Lifestyles in Close Relatives of Animals and Fungi. Curr Biol. 2015; 25(18): 2404–2410. PubMed Abstract | Publisher Full Text\n\nGrau-Bové X, Torruella G, Donachie S, et al.: Dynamics of genomic innovation in the unicellular ancestry of animals. eLife. 2017; 6: pii: e26036. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrooks GE, Hon G, Chandonia JM, et al.: WebLogo: a sequence logo generator. Genome Res. 2004; 14(6): 1188–90. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "36417",
"date": "30 Jul 2018",
"name": "Tim P. Levine",
"expertise": [
"Reviewer Expertise Lipid traffic",
"membrane contact sites",
"structural bioinformatics using HHsearch"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 correspondence, Wideman and colleagues show that PDZD8 (also called PDZ-8 or PDZK8) is not an ortholog of the ERMES subunit Mmm1p. This conclusion by Wideman et al. is well founded, and the underlying argument is well explained in the text here.\n\nThe same point of non-orthology has been said before in Wideman et al., 2015, but was then missed in the first paper that focussed on the function of PDZD8 (Hirabayashi et al., 20171). There, the SMP domain from mammalian PDZD8 substituted for the function of the Mmm1p SMP domain in yeast, which was used to support the view that “The SMP domain of PDZD8 is functionally orthologous to the SMP domain found in yeast Mmm1.” However, a definition of orthologue that I accept is “Homologous sequences are orthologous if they are descended from the same ancestral sequence separated by a speciation event” [https://en.wikipedia.org/wiki/Homology_(biology)]. By that definition, I agree with Wideman et al. that bioinformatic studies indicate PDZD8 and all 3 ERMES subunits with SMP domains, including Mmm1p, are paralogs, not orthologs.\nOn that basis the function of PDZD8 may have evolved away from that of ERMES at some time, and might also have converged towards it, but we cannot tell by just showing that the activity of one’s SMP domain (most likely to transfer lipid with fairly low specificity) can be substituted by the other’s SMP domain.\n\nThere are several points where the current article might be marginally improved by revision:\n\nImprove text and diagram describing domains within PDZD8: apart from the SMP and PDZ domains, PDZD8 is described as containing a C1 domain e.g. NCBI human PDZD8 (1154aa): 841-883. Here that domain is missing from the diagram. In addition, a C2 domain is shown in the diagram in an incorrect way, because the C2 domain sequence in PDZD8 is split. ~65 aa lies between the SMP and PDZ domains (320-372), then the rest (695-800) is before the C1 domain. See Figure 1B in Wong et al. 20172.\n\nThe text would benefit from a brief analysis of likely branching of the SMP protein family, based on supplementary data from the 2015 paper, possibly plus extra work to place the different Mmm1 groupings and the two new designations (X and Y).\n\nThe text should explcitily state that the Mmm1s in C. owczarzaki and other holozoa are variant, rooting onto the main tree separately from fungal Mmm1s, despite the shared presence of the 2 accompanying motifs.\n\nThis raises another point. The diagram adds to cladistic analysis of SMP domain sequences alone by looking for accompanying motifs. How can using accompanying motifs be justified, when accompanying domains are excluded from the orthology/paralogy argument? As SMP sequences can be inherited separately from other elements, I do not fully grasp that C. owczarzaki-1867 is included in Mmm1, while C. owczarzaki-5382 is excluded from Pdz8.\n\nUnderstanding the figure might be improved by more information in the legend:\nSay what “Fungi” refers to each of the three times it appears in the diagram. Is it always the same set of fungal sequences or a different group in each case? How many? Say that the figure is not the same as appears in ref 11 or its supplementary data, but based on the same data set. Say that Tcbs have been excluded For “closely related protists” (line 3) say if every protist has been included or a selection has been made (and how) Identify falgellated fungi (compared to non-flagellated) Add description of one/two asterisks to key, not at bottom of long legend.\n\nData set on Git Hub: I note that some sequences are incomplete. This should be made explicit.\n\nIs the rationale for commenting on the previous publication clearly described? Yes\n\nAre any opinions stated well-argued, clear and cogent? Yes\n\nAre arguments sufficiently supported by evidence from the published literature or by new data and results? Yes\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "36318",
"date": "31 Jul 2018",
"name": "Vikram Alva",
"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 correspondence, Wideman et al. comment on a recent seminal study by Hirabayashi et al., which describes the biochemical characterization of the SMP domain-containing protein PDZD8, as an ER-mitochondria tether important for the regulation of Ca2+ dynamics in mammalian neurons. Although the functional data presented by Hirabayashi et al. is sound and of great interest, their interpretation of the homologous relationship between PDZD8 and the yeast SMP domain-containing protein Mmm1 is incorrect. Mmm1 is part of the ERMES complex, which tethers the ER to mitochondria in yeast and is important for the exchange of phospholipids between these organelles. This complex contains two further proteins of the SMP family, Mdm12 and Mdm34. Wideman et al.1 and we2 have previously shown that some eukaryotes contain all three of these ERMES proteins as well as PDZD8, indicating that they are paralogs. Hirabayashi et al., however, misinterpret PDZD8 to be a functional ortholog of yeast Mmm1, and substantiate this relationship experimentally by substituting the SMP domain of yeast Mmm1 with that of PDZD8 to rescue the phenotypic defects caused by the deletion of Mmm1. Here, Wideman et al. present convincing data to show that PDZD8 is a paralog of Mmm1 and not its ortholog. The arguments presented by Wideman et al. are sound and will help to clear the confusion regarding the evolutionary relationship between these proteins.\nMinor comment: The human PDZD8 protein contains a C1 domain as well as a coiled-coil segment at its C-terminal end. Do other orthologs also contain these domains? If so, the domain composition of PDZD8 shown in Fig. 2 should be updated to show this. Also, HHpred searches suggest that the PDZ domain in human PDZD8 is inserted into the C2 domain; is this the case for other orthologs too?\n\nIs the rationale for commenting on the previous publication clearly described? Yes\n\nAre any opinions stated well-argued, clear and cogent? Yes\n\nAre arguments sufficiently supported by evidence from the published literature or by new data and results? Yes\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1088
|
https://f1000research.com/articles/7-1087/v1
|
16 Jul 18
|
{
"type": "Opinion Article",
"title": "Shaping the Future of Research: A perspective from early career researchers in Vancouver, Canada",
"authors": [
"Peter G.K. Clark",
"James M. McCoy",
"Jenny H.L. Chik",
"Azadeh HajiHosseini",
"Manuel Lasalle",
"Brianne A. Kent",
"Stefanie L. Vogt",
"Peter G.K. Clark",
"James M. McCoy",
"Jenny H.L. Chik",
"Azadeh HajiHosseini",
"Manuel Lasalle",
"Brianne A. Kent"
],
"abstract": "Future of Research is an organization dedicated to championing, engaging, and empowering early career researchers (ECRs). The organization was founded in 2014 and has since inspired other groups to advocate for a more equitable and sustainable research enterprise. Here we report the findings of the Future of Research Vancouver Symposium. The goals of the Vancouver symposium were to ascertain the perspective of ECRs in Canada and to outline pathways to a sustainable future for Canadian research. The symposium had two sessions. The first session was a series of talks that were intended to prepare attendees with an informed understanding of several perspectives in the science enterprise, with a particular focus on the Canadian system. The second session was a series of interactive workshops to identify the greatest challenges facing ECRs in Canada and to propose solutions. The results of the workshops illuminated three main themes for the challenges facing Canadian ECRs: funding, mentorship, and the divide between academia and other sectors. These themes are similar to those discussed at the Future of Research symposiums in the United States, emphasizing that these issues are not isolated to Canada; however, Canadian policies are trailing behind the progress being made in other countries.",
"keywords": [
"training",
"funding",
"mentoring",
"junior scientists",
"career development",
"science policy",
"research enterprise"
],
"content": "Executive summary\n\nThe Future of Research Vancouver Symposium (FoRV2017) was held on February 20th, 2017, following increasing concern by members of the local academic community that the voices of junior researchers were not being considered in discussions around the future of funding and training structures in Canadian research.\n\nIn laboratories and offices across Canada today, the majority of research is undertaken by early career researchers (ECRs), namely graduate students and postdoctoral fellows (PDFs). ECRs design and execute experiments, collect and analyze data, write papers, and are often solely responsible for supervising more junior team members. This multifaceted contribution suggests that ECRs play a core role in Canada’s science, technology and health sectors.\n\nHowever, the Canadian research landscape now presents significant challenges to ECRs:\n\nThe number of PhDs awarded annually by Canadian universities are growing, as are the length of PDFs’ tenures. However, the number of junior faculty positions available at Canadian universities has shrunk.\n\nCanadian ECRs are often advised to seek alternative career paths, but report high levels of dissatisfaction with the career development and professional training available to them.\n\nA lack of “staff scientist” or stable mid-career options make academic employment undesirable to many trainees, and result in lab management problems including institutional knowledge loss, and a dearth of supervision and support.\n\nWages for Canadian ECRs are not internationally competitive, which is exacerbated in BC by Vancouver’s high cost of living, and many ECRs do not receive basic employment benefits available to other working residents.\n\nECRs report high levels of symptoms of poor mental health.\n\nRecent announcements regarding increases in the Canadian research councils’ budgets offer promise for positive change, but few details have been offered on actual plans to improve opportunities for ECRs.\n\nTo effect change, junior researchers must identify the multifaceted challenges they face, and confront the role that academia, government, and industry can play in addressing them. As such, the opening session of FoRV2017 consisted of talks and panel discussions from local members of the scientific community, including industry and academic leaders, who have been vocal regarding ECR issues and the sustainability of Canadian science. This panel was followed by workshops aimed at discussing the issues that had been raised, and prompting potential solutions from attendees. Workshops focussed on 4 core topics: (1) how trainees could be better prepared for careers in science, (2) how sustainable and secure career pathways could be created for ECRs, (3) how funding of research in Canada could be structured to balance basic research, knowledge translation, and training of ECRs, and (4) how scientists and institutions could be better incentivised for behaviours that support the future of Canadian science.\n\nBased on the responses from attendees, and further literature review and discussion among organisers, we endorse the following recommendations:\n\n1. Improve ECR-targeted funding, including grants which provide operating costs for ECRs and support the transition from a PDF to junior faculty position, and recognising ECRs contributions to grants awarded to their supervising professors.\n\n2. Develop guidelines for mentorship and training, such as professional development programs, and tools to help supervising professors provide high-quality mentorship, including incentivising them to allow ECRs to seek training outside the supervisor-ECR relationship.\n\n3. Bridge gaps between academia and alternative career paths, such as through partnered research with private industry, and internships with non-academic groups.\n\nIf the future of Canadian research lies in its junior researchers, then strategies must be laid out for how universities, government, and the knowledge-intensive industries can be better at nurturing our ECRs. Recent grassroots campaigns, such as #SupportTheReport to encourage the Canadian government to take up the recommendations of the Fundamental Science Review, have shown that effective change is possible. Canadian ECRs must be ready to stake their claims in their future, and we hope that meetings such as FoRV2017 are only the beginning.\n\n\nIntroduction\n\nIn Canada, a large proportion of science, engineering, and social research is carried out by early career researchers (ECRs), a term that encompasses graduate students, postdoctoral fellows (PDFs), and principal investigators (PIs) who have held their independent position for less than five years. ECRs represent not only the core workforce of most academic institutions, but the next generation of scientific leaders. However, there has been growing concern that current funding and training structures almost entirely ignore the best interests of ECRs.\n\nBetween 2002 and 2011, the number of PhDs awarded annually by Canadian universities increased by 68% to over 60001. However, this has not been met with an accompanying increase in the number of faculty positions. Between 2010 and 2017, the number of assistant professorships in Canada, representing potential faculty positions that these PhD graduates could hope to one day occupy, declined by 15.9%2. Overall, about 16% of PhD holders in the Canadian labour force are employed in tenure-track positions1. Therefore, the graduate training environment should reflect the reality that most students will need to find employment outside of academia. As a way to increase employment options for scientists, mid-career positions including staff scientists and group leaders have been proposed3. Staff scientists could offer many additional benefits to labs, including providing on-hand supervision for trainees, and acting as reservoirs of institutional knowledge. However, the current grant-by-grant nature of most research funding makes such permanent positions difficult to implement for all but the largest groups.\n\nCanadian PhDs do find high levels of employment outside of academia1, but many graduates are concerned about a lack of career training and exposure to alternative career paths for graduate students and PDFs, collectively termed “trainees”. For example, among PhD students in Canada, only 51% are satisfied with the available advice and workshops concerning academic careers, and only 40% are satisfied with the resources for non-academic careers1. Around 70% of Canadian PDFs report being satisfied with the resources and supervision available to them from PIs, but only ~40% are satisfied with opportunities for career development and professional training4. Postdoctoral fellowships are ostensibly trainee positions, thus it is particularly concerning that PDFs are so dissatisfied with the training opportunities available to them. Some of the areas where Canadian PhD holders feel their professional development has been lacking include training in communication with non-specialist audiences, business and financial management, and the ability to work in a team1. Given that skills like team management are core to running a lab as a professor, this is a significant weakness for Canadian science.\n\nCanadian PDFs also report high levels of dissatisfaction with their salaries4. In 2016, around half of Canadian PDFs were earning less than $45,000 CAD per year; if such a PDF were a sole earner with dependents, this salary may place them below the Canadian Low Income Cut-Off4. In the same period, salary of UK PDFs increased to ~CAD$60,0004, while recent US Department of Labor rulings have led to a dramatic increase in the salary of some US PDFs to ~CAD $59,500. Moreover, other countries offer PDF salary increments based on experience, while Canadian PDFs will often see no salary increase even after a 5-year appointment, which is the maximum appointment length for a PDF at the University of British Columbia, Vancouver. The relatively low salaries may make Canada a less desirable destination for those seeking postdoctoral positions; indeed, the proportion of non-Canadian citizens holding postdoctoral positions in Canada declined from 56% in 2009 to 42% in 20164. British Columbia and Vancouver are particularly unattractive for prospective ECRs, as costs of living here are among the highest anywhere in North America5. Geographical cost of living adjustments to ECR salaries and fellowships have been suggested by other groups around the world to make working in cities like Vancouver possible3, but have been largely ignored by government and academic groups.\n\nAnother cause for dissatisfaction among Canadian PDFs concerns employment benefits. Although the majority of Canadian PDFs have access to basic provincial health care, only around half have access to extended health benefits that cover the cost of prescription medication and dental and vision care4. Furthermore, while all Canadian PDFs pay federal and provincial income tax, whether supported by fellowships or by their supervisor’s grant funding, only those classified as “employees” are eligible for government benefits available to other Canadian employees, such as the Canada Pension Plan or Employment Insurance (EI). Since federally funded parental leave is available only to those employees who contribute to EI, 38% of Canadian PDFs report being ineligible for paid parental leave and a further 38% are unsure of their eligibility4. It should be noted that some fellowship-funded PDFs do receive paid parental leave but the amount and duration is not regulated, and thus varies depending on the specific source of funding. The lack of paid parental leave and unaffordability of child care is particularly problematic because the average age of Canadian PDFs has now increased to 34 years4, meaning that the most productive years for a scientist now increasingly overlap with peak child-rearing years. These issues may also contribute to attrition of female researchers. While women make up 48% of PDFs across all fields4, they represent only 40% of Canadian faculty members2; however, this does represent an improvement from 36.6% five years ago.\n\nFor many years, Canadian ECRs who looked to apply for grants as new PIs have faced science funding budgets that did not keep pace with inflation6. Between 2008 and 2015, the number of grants awarded by the Canadian Institutes of Health Research (CIHR) to early career PIs declined by 38%7. Starting in 2014, CIHR diverted up to 45% of its funding for investigator-initiated research to the new Foundation Grant scheme, designed to provide longer term support for leaders in health research8. Foundation Grant funding has been heavily biased toward senior researchers; in the first Foundation competition, only 5% of funds were awarded to ECRs7. In 2017, the Foundation Grant eligibility rules were revised and ECRs became ineligible for funding via this mechanism9. While an equalization mechanism exists within the CIHR Project Grant scheme to ensure that the proportion of grants awarded to ECRs matches the proportion of applicants who are ECRs10, these grants are smaller in value: the average size of Foundation Grants is $2.6 million CAD, while Project Grants are worth CAD$720 thousand on average. More still needs to be done to ensure that Canadian ECRs have access to the funding that they will use to build their research programs.\n\nAlong with the dissatisfaction with current funding and training structures, there is growing awareness that ECRs worldwide experience high levels of depression and anxiety11. A recent study showed that graduate students across the globe experience symptoms of depression at six times the rate of the general population, sparking concerns of a mental health crisis12. More worryingly, ~75% of Canadian PDFs report symptoms of mental illness4. These factors can only be exacerbated by concerns over job security, and feelings of low status and underappreciation in the university hierarchy. These mental health concerns further underscore the need for extended health benefits (including counselling/therapy) among Canadian PDFs.\n\nDespite these important concerns, there is hope for the future. The Fundamental Science Review, commissioned in 2016 to evaluate the effectiveness of Canada’s investment in research, recommended a CAD$1.3 billion increase in base funding of the three federal research granting agencies, the CIHR, the Natural Sciences and Engineering Research Council (NSERC), and the Social Sciences and Humanities Research Council (SSHRC), just to bring Canada on par with peer nations6. Increased investment in ECRs was recommended, but few details were provided. The Fundamental Science Review helped galvanise scientists to campaign federal Members of Parliament for better research funding. While the recent 2018 Federal Budget did not meet all expectations, it did increase fundamental research funding by $925 million over 5 years, with $275 million set for international, inter-disciplinary, and high-impact research13. Furthermore, it proposes $210 million over 5 years for Canada Research Chair (CRC) appointments specifically for ECRs. However, while this provides salary support it does not provide operating funds, a major issue as nearly half of the CIHR-funded tier-2 Canada Research Chairs held no CIHR operating grants in 20157. Thus, while significant challenges still face Canadian ECRs, there is good reason to believe that strong advocacy by grassroots movements can reverse the decline in public funding of Canadian science.\n\nWhile the specifics of funding bodies and university structure may differ, similar concerns are growing in many developed nations with long-standing publicly funded research programs. In 2014, in Boston, MA, a group of postdoctoral researchers founded Future of Research, an organization dedicated to championing, engaging, and empowering ECRs. Driven by what they perceived as a severe lack of investment in the next generation of scientists, they convened a symposium to learn what ECRs in the US were most concerned about. The ultimate goal of the Future of Research was to develop a series of recommendations for universities and governments, which could act as a roadmap to move toward a more equitable and sustainable research enterprise.\n\nThe outcomes of the Future of Research Boston Symposium were later published3. Broadly speaking, their recommendations were: (1) to promote discussions between ECRs and stakeholders on how the scientific enterprise could be reformed, (2) to increase transparency on career outcomes, and expectations of the balance between training and research in PDF appointments, and (3) to increase investment in ECRs, with more grants to PDFs to provide financial independence from PIs, and accountability for the quality of training received during a postdoctoral appointment. Since the symposium, Future of Research has played an essential role in advocating for increased pay for US PDFs, in response to changes in overtime payment thresholds set by the Department of Labor14.\n\nFuture of Research inspired other groups of PDFs and graduate students to form their own local organisations. Here we report the findings of the Future of Research Vancouver Symposium, which was organised by PDFs in Vancouver, BC, to discuss how to create a sustainable future for Canadian research.\n\n\nWorkshop overview\n\nFuture of Research Vancouver 2017 (FoRV2017) was organised by PDFs from the University of British Columbia (UBC) and Simon Fraser University (SFU). The event was hosted at the Vancouver Campus of SFU, on the afternoon of February 20th 2017, with 136 participants in attendance. The event was financed through sponsorship from community partners, including universities (SFU, UBC), associations (UBC Postdoctoral Association), biotechnology companies (STEMCELL Technologies, Zymeworks Inc., AbCellera Biologics Inc., Mesentech Inc.), funding agencies (Michael Smith Foundation for Health Research, Genome British Columbia), research infrastructure programs (WestGrid), and research centres (Centre for Blood Research).\n\nThe first session of the workshop was focused on preparing attendees with an informed understanding of the current state of the scientific enterprise, with a focus on the Canadian system. A diverse range of speakers was sought to cover perspectives from many sectors of the research ecosystem to best frame the issues at hand. The session began with an introduction by Dr. James McCoy, a PDF at UBC and a member of the organising committee, to define some of the key challenges currently faced by ECRs. This was followed by keynote addresses by Dr. Santa Ono, President and Vice-Chancellor of UBC, and Dr. Terry Thomas, Senior Vice President (Research and Development) of STEMCELL Technologies, to convey perspectives on Canadian research from two of its largest sectors, namely academia and industry. This was followed by a panel discussion, composed of: Dr. Terry Thomas; Dr. Liisa Galea, a Professor at UBC and strong proponent of women in science; Dr. Lara Boyd, an Associate Professor at UBC and the representative of her institution to CIHR; and Dr. Eric Hsu, the lead for Data Science at Istuary Innovation group and a representative for alternative career paths for scientists.\n\nThe second session of the workshop was centred on discussions of the issues facing ECRs in Canada and possible solutions. The attendees self-selected one of four individual Breakout Sessions, each with a particular theme similar to those used at the FoR Boston event3. The themes for the four sessions were:\n\nSession 1. How can trainees be better prepared for careers in science in 2017?\n\nSession 2. How should the supply of postdocs and graduate students be matched to demand to create sustainable, secure career pathways for young researchers?\n\nSession 3. How can the funding of science research in Canada be structured to balance and promote basic research, knowledge translation, and training of the next generation of scientists?\n\nSession 4. How can the current system of incentives be fixed so that scientists and institutions are rewarded for the behaviours that are believed to support good science?\n\nBreakout Sessions followed a similar procedure to that used at FoR Boston3,15. Briefly, each Breakout Session began with a short introduction to the topic of discussion and relevant background information by moderators. The participants were then given time to brainstorm “problems” related to the topic and record their ideas on sticky notes. Subsequently, there was a group discussion about the problems and participants attempted to group the sticky notes to identify broader trends. A similar procedure was then followed to identify possible solutions. The problems and solutions recorded on the sticky notes in the Breakout Sessions represent the primary data for this article. Before leaving the Sessions, participants were asked to complete an Exit Survey, which included questions regarding demographic data, opinions about the outcomes of the Breakout Session, and feedback towards future events. The workshop concluded with a reception that allowed participants to continue the discussions in an informal setting and network with the event sponsors.\n\n\nResults\n\nThe Exit Survey results provided insight into the range of participants involved in the discussions. Diversity was desired to ensure the discussions incorporated inputs, views and opinions from a variety of sources, to ensure representation of the wider community of ECRs. The Exit Survey supplied to participants and the responses are available online (Supplementary File 1). From the 65 responses, the average age of attendees was 33 years old, with only Session 3 showing a notable digression from this average. The gender ratio was in favour of female participants (31:20), although there were a notable number of non-responders to this question. Attendees represented a variety of positions; however, the majority of attendees were either graduate students (31) or PDFs (22).\n\nM = Male; F = Female; NR = No Response; PDF = Postdoctoral Fellow; GS = Graduate Student; US = University Staff; I = Industry; F = Funding body; RS = Research Scientist; P = Public; U = Unemployed.\n\nProblems and solutions identified in Breakout Sessions. The problems and solutions proposed by participants in each Breakout Session were recorded as described in Workshop Overview (for raw outputs see Supplementary File 2). As many of the data points showed overlap or were related, the data were summarized into sets of key problems and solutions, which are presented in Table 2–Table 5. Note that the solutions supplied were not tied to specific problems that had been raised, and the grouping of this information into general themes has been performed ex post.\n\nRating of outputs identified in Breakout Sessions. After the discussion of possible problems and solutions described above, each Breakout Session concluded with an Exit Survey in which participants were asked to choose the most significant problem and most impactful solution identified in their Session. These answers are summarized in Table 6–Table 9; all responses (raw data) can be found in the Supporting Information (Supplementary File 1).\n\nAttendees were asked to evaluate the process and outcomes of FoRV2017 and their specific Breakout Session in the Exit Survey. The data from the Exit Survey were analysed to give the average responses reported below (Table 10). Questions requested a ranking between 1 and 5, with 5 representing complete agreement (Supplementary File 1). The lowest importance (3.3/5) was ascribed to the questions regarding whether a consensus was reached and whether it was important to reach consensus, showing that the majority of attendees believed that a general agreement on the discussed problems and solutions was neither necessary nor attained in order to achieve positive change for ECRs. In contrast, the importance for diversity - in the meeting, the speakers, the attendees and the discussions - was highly scored (4.1/5). Finally, attendees rated FoRV2017 highly, by giving a 4.2/5 score on whether they would recommend the event to friends and colleagues.\n\nScores are out of a total of 5.\n\n*Scores were on a scale of 1–5, with 5 being in the affirmative. For the question on reaching a consensus, 5 represented diverse views and 1 represented a consensus.\n\n\nDiscussion\n\nThere have been major changes to science funding in Canada and scientific research globally in the past decade, yet ECRs, particularly graduate students and PDFs, have had few opportunities to voice their opinions on these matters. FoRV2017 aimed to fill this gap. Although the Breakout Sessions focused on different topics, the identified problems and solutions showed significant overlap. Three major challenges facing ECRs in Canada that were brought up repeatedly relate to funding, mentorship, and the existing divide between academia and other sectors (particularly industry). We discuss each of these issues and potential solutions in more detail below.\n\nShortage of funding throughout the Canadian science ecosystem. According to FoRV2017 participants, the underfunding of science is one of the largest issues facing Canadian research (Table 3, Table 4, and Table 5). This problem was also exposed by the Fundamental Science Review6 and felt across all facets of research, with calls for increased funding for both short-term and longer-term projects, fundamental research, and interdisciplinary research (Table 4).\n\nThe science funding increase announced in the Federal Budget 201816 may make inroads to correct some of these issues. Although there was no change in federal science funding in Budget 201717, Budget 2018 responded to many of the recommendations of the Fundamental Science Review6, including: increases in base funding for granting councils and the Canadian Foundation for Innovation; targeted funds for ECRs through new CRC allocations; targeted funds for international, interdisciplinary, fast-breaking and high-risk research; and support to develop better equity and diversity outcomes16. However, it is unlikely that these changes will fully resolve the concerns of FoRV2017 participants regarding research funding in Canada. For example, the increase in unfettered funding for investigator-initiated research in Budget 2018 amounts to a real dollar increase of 14% over five years13; however, the available federal funding per researcher decreased by 35% between 2007 to 20156. Therefore, further funding increases will be required to restore the resources available to each researcher to the level available a decade ago.\n\nFoRV2017 participants recommended direct engagement and lobbying of the government as a key action needed to encourage funding increases (Table 2 and Table 4). Evidence that this approach can be successful comes from the recent #SupportTheReport campaign, which involved direct petitioning of parliamentarians by scientists and the public to support calls for increased science funding18. In addition, FoRV2017 participants suggested that alternative sources of funding should be further explored (Table 4). For example, industry could play a larger role in funding research (discussed in more detail below). In addition, public donors could also play a larger role; for example, universities could help researchers pursue crowd-funding mechanisms (Table 4). Given the diversity of existing funding sources, FoRV2017 participants recommended that tools should be established to help scientists identify and locate these sources (Table 4).\n\nInsufficient ECR-targeted funding. In addition to the overall shortage of research funding in Canada, numerous FoRV2017 participants voiced concern that ECRs have additional difficulties in accessing these funds (Table 4, Table 5, and Table 8). Participants believed that an overemphasis on track record (as opposed to the proposed research) in grant applications makes it difficult for new PIs to compete for funding against established researchers (Table 4 and Table 5). Participants put forward many proposals of how granting agencies could level the playing field for ECRs, such as evaluating proposals independent of CVs or conducting double-blind reviews to reduce bias in favour of established researchers (Table 4). In order to help ECRs build their track records, it was also suggested that trainees should be recognised for their contributions to grant applications, such as by allowing them to be listed as co-investigators (Table 2). However, the intervention that numerous participants rated as the most potentially impactful is the establishment of dedicated funding for ECRs (Table 8). All three of the federal granting agencies have recognised and responded to this issue with controls established on ECR success rates or dedicated funds in at least one of their programs (including the CIHR Project Grant19, the NSERC Discovery Grant20, and the SSHRC Insight Development Grant21); whether this will be sufficient will be evidenced in time.\n\nIt is uncertain whether the funding plan laid out in Budget 2018 will substantially improve the situation for ECRs. First, graduate scholarships and postdoctoral fellowships, which the FoR Boston recommendations favour as a means to give more independence to ECRs3, received no funding increases in Budget 201813. On the other hand, Budget 2018 includes $210 million CAD over five years to fund new CRCs, which could support the salaries of up to 250 early career PIs13. However, since CRC funds cannot be used toward operating expenses, ECRs hired with this salary support may continue to have difficulty acquiring sufficient funding for their research.\n\nAnother major area of concern among FoRV2017 participants was the poor quality of mentorship that ECRs receive. Indeed, around half of the participants in Breakout Session 1 listed mentorship as their most significant concern (Table 6). There were several areas in which mentorship for ECRs was considered to be lacking. First, participants were concerned that there is little incentive for supervisors to invest time and effort in mentoring their graduate students and PDFs, since evaluation of PIs tends to be heavily weighted toward research output (i.e., publications and grants) rather than mentoring outcomes for their trainees (Table 5). Second, some participants believed that there is a lack of training available to PIs regarding how to train and mentor their trainees; the effect of this lack of experience among PIs can range from ineffective mentorship to unprofessional behaviour and misconduct (Table 2). Finally, participants expressed concern that trainees who intend to work outside academia are receiving especially poor mentorship, since many PIs place little value in or have little experience in work outside of academia (Table 2).\n\nParticipants of FoRV2017 put forward several proposed solutions for the mentorship problem in academia. Participants suggested that PI training in mentorship skills could be directly offered by universities, or that funds could be provided to PIs who wished to pursue mentorship training on their own (Table 2). In addition, professional development programs, run by the institution, could be made mandatory for graduate students or postdocs (Table 2); this policy would remove the ability of supervisors to dissuade their trainees from participating in external training and mentorship programs. Participants also suggested that trainees take mentorship into their own hands to a greater extent, and seek out mentors external to their supervisor and formal supervisory committee (Table 2).\n\nConcerns around mentorship and training of ECRs have also been raised by numerous organisations in recent years. In BC, the UBC Graduate Student Society published a white paper on supervisory excellence, addressing what many graduate students perceived as a dire lack of mentorship in Masters and PhD programs22. Several recommendations in this paper, including mentorship programs for supervising professors and setting measurable minimum standards for supervision by universities, closely match suggestions made by FoRV2017 participants. This paper has been met positively by UBC administration and the Faculty of Graduate Studies, prompting such measures as making seats available for students on the Faculty of Graduate Studies Committee, but actual change has yet to be demonstrated.\n\nThe FoR Boston group has recommended that supervisors, departments, and institutions be held accountable for providing a good training experience for postdocs by requiring direct feedback from postdocs to granting agencies3. They also suggested that anonymized, aggregated feedback on the training environment in individual departments could be made publicly available, to assist prospective graduate students and postdocs3. The participants of FoR Chicago similarly recommended that institutions and granting agencies conduct performance evaluations of mentorship practices by individual PIs23. Much like the participants of FoRV2017, the attendees of the Chicago symposium also expressed a wish for improved training in mentorship skills for PIs23. Clearly, a need for improved mentorship in academia is a common theme in universities throughout North America.\n\nThe historical separation between academia and other sectors contributes to the issues faced by ECRs today. In recent years, the need to bridge this divide with industry24, decision-makers25 and the public26 has become evident. The need to increase crosstalk, collaboration and engagement specifically with industry was highlighted by many FoRV2017 attendees across different Breakout Sessions (Table 2, Table 3, and Table 4). Such interaction would likely be mutually beneficial to both trainees and the industries in which they intend to seek employment.\n\nAlternative career paths. Bridges between academia and other sectors will assist in exposing trainees to alternative career paths. In every field of study, Canadian PhD holders are more likely to be employed in a non-academic sector than as a full-time university professor1; however, many ForV2017 participants reported a lack of knowledge about what alternative career opportunities were available (Table 2 and Table 3). To address this issue, some participants expressed a desire to gain experience working in non-academic positions, such as through internships incorporated into the PhD program (Table 2 and Table 3). Of note, graduate students and postdocs in Canada can already complete internships in industry through the Accelerate program run by the non-profit organisation Mitacs27. It is unclear whether some attendees might have been unaware of this program or whether demand for this type of internship exceeds capacity in the Accelerate program. Participants also suggested several ways that academic institutions could provide greater exposure to industry to their trainees, such as recruitment fairs and career panels (Table 2).\n\nIn addition to being unaware of alternative careers paths, attendees also reported feeling unprepared for these careers (Table 2). Participants suggested several ways in which academic institutions could improve training for non-academic careers, including career-tailored professional development programs (Table 2) and directly involving industry in designing and implementing career training programs for graduate students and postdocs (Table 3). Participants also recommended that professional development programs place a greater emphasis on recognizing and marketing the transferrable skills that trainees gain during their academic training (Table 2). Concerns about poor training for careers outside of academia have also been raised in FoR symposia in the US. The FoR Chicago group, for example, recommended that professional development opportunities for graduate students and postdocs be expanded, through either curricular or extracurricular programs, and that “professional PhD” programs, featuring training in other areas such as business or law, be offered instead of or in addition to traditional PhD programs23.\n\nFoRV2017 attendees also felt that wider stakeholder engagement in science should be pursued. Building public support for science will be crucial to convincing policy-makers to enact many of the changes recommended in FoRV2017. In particular, funding increases suggested by FoRV2017 participants will require the direct engagement and lobbying of government; however, participants recognized that public support will also be required to influence policy makers (Table 5). In addition to direct appeals to the public, it is also important that ECRs present a united front to stakeholders such as granting agencies to effectively advocate for changes that will benefit ECRs3. Recent campaigns for increased research funding in Canada have led to the creation or increased the activity of ECR advocacy groups such as the Association of Canadian Early Career Health Researchers (ACECHR) and the Science & Policy Exchange; continued activity of these and other groups in the future will be required for full implementation of the recommendations made by FoRV2017 participants.\n\n\nConclusion\n\nECRs in Canada face many of the same challenges experienced by ECRs in the US. FoRV2017 participants highlighted insufficient funding, mentorship, and connection with industry as the most critical factors contributing to the hurdles faced by Canadian ECRs. To make the research enterprise in Canada more equitable and sustainable, FoRV2017 participants recommend more funding directed to supporting ECRs (particularly research operating costs and career transition support), increased incentives and training for high quality mentorship, and strengthened partnerships with private industry in academic training and research initiatives. Following the Fundamental Science Review and the announcement of the 2018 federal budget, it is clear that progress is being made to support Canadian research; however, more is needed to support ECRs to retain and attract top talent towards building a more sustainable scientific enterprise in Canada.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors wish to thank our generous sponsors without whom the symposium would not have been possible: Office for Graduate Students and Postdoctoral Fellows, SFU; Postdoctoral Fellows Office, UBC; Zymeworks Inc.; STEMCELL Technologies Inc.; AbCellera Biologics Inc.; UBC Postdoctoral Association; the Michael Smith Foundation for Health Research; WestGrid; Genome British Columbia; Mesentech Inc.; the Centre for Blood Research. We would like to thank Dr Santa Ono, Dr Terry Thomas, Dr Liisa Galea, Dr Lara Boyd and Dr Eric Hsu for speaking at the symposium. We would also like to thank the moderators and volunteers who helped plan the symposium and/or run the breakout sessions: Dr Kaarina Kowalec, Dr Eva Kwoll, Dr Jess Inskip, Dr Mariya Cherkasova and Matthew MacLennan. We would also like to thank Dr Gary McDowell from Future of Research and Dr Lauren Drogos from Future of Research Canada for their advice.\n\n\nSupplementary material\n\nSupplementary File 1: Exit Survey – text of the Exit Survey administered to participants, with quantitative scoring data, summarized qualitative answers, and raw qualitative answers.\n\nClick here to access the data.\n\nSupplementary File 2: Output from Breakout Sessions.\n\nClick here to access the data.\n\n\nReferences\n\nEdge J, Munro D: Inside and Outside the Academy: Valuing and Preparing PhDs for Careers. 2015. Reference Source\n\nShen A: Data on number of professors and their salaries released after five-year hiatus | University Affairs. University Affairs. 2017; (Accessed: 3rd May 2018). Reference Source\n\nMcDowell GS, Gunsalus KT, MacKellar DC, et al.: Shaping the Future of Research: a perspective from junior scientists [version 2; referees: 2 approved]. F1000Res. 2015; 3: 291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJadavji NM, Adi MN, Corkery TC, et al.: The 2016 Canadian National Postdoctoral Survey Report. National Postdoctoral Survey Committee Chair. 2016. Reference Source\n\nCost of Living in Vancouver. 2018; (Accessed: 6th June 2018). Reference Source\n\nNaylor D, Birgeneau RJ, Crago M, et al.: INVESTING IN CANADA’S FUTURE Strengthening the Foundations of Canadian Research. 2017. Reference Source\n\nThe Association of Canadian Early Career Health Researchers: Early Career Investigators (ECIs) in health research: final report of a cross-Canada survey. 2016. Reference Source\n\nCanadian Institutes & of Health Research: Foundation Grant: Overview. (Accessed: 8th May 2018). Reference Source\n\nCanadian Institutes & of Health Research: Foundation Grant: Eligibility - CIHR. 2017; (Accessed: 3rd May 2018). Reference Source\n\nCanadian Institutes of Health Research: Clarifying the Fall 2017 Project Grant funding results. (Accessed: 8th May 2018). Reference Source\n\nTime to talk about why so many postgrads have poor mental health. Nature. 2018; 556(7699): 5. PubMed Abstract | Publisher Full Text\n\nEvans TM, Bira L, Gastelum JB, et al.: Evidence for a mental health crisis in graduate education. Nat Biotechnol. 2018; 36(3): 282–284. PubMed Abstract | Publisher Full Text\n\nThe Association of Canadian Early Career Health Researchers: Budget 2018 Report Card: A Good Start. 2018; (Accessed: 3rd May 2018). Reference Source\n\nBankston A, McDowell GS: Monitoring the compliance of the academic enterprise with the Fair Labor Standards Act [version 2; referees: 3 approved]. F1000Res. 2017; 5: 2690. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMazzilli SA, Gunsalus KTW, McDowell GS, et al.: Logistics of Organizing the FOR Symposium. The Winnower. 2014; 5: e141697.77958. Publisher Full Text\n\nMinister of Finance: Budget 2018: Equality and Growth for a Strong Middle Class. 2018. Reference Source\n\nKondro W: Research stays frozen in Canadian budget. Science. 2017. Publisher Full Text\n\nSupport the report. (Accessed: 3rd May 2018). Reference Source\n\nCanadian Institutes of Health Research: Message from the Acting President of CIHR - CIHR. 2017; (Accessed: 3rd May 2018). Reference Source\n\nNatural Sciences and Engineering Research Council of Canada: Discovery Grants - Applicant Categories. (Accessed: July 12, 2018). Reference Source\n\nSocial Sciences and Humanities Research Council: Insight Development Grants. (Accessed: 3rd May 2018). Reference Source\n\nMarshall N, Ringsred A, Kirlin A, et al.: SUPERVISORY EXCELLENCE: A GRADUATE STUDENT PERSPECTIVE. 2017. Reference Source\n\nDolan KT, Pierre JF, Heckler EJ: Revitalizing biomedical research: recommendations from the Future of Research Chicago Symposium [version 1; referees: 2 approved with reservations]. F1000Res. 2016; 5: 1548. Publisher Full Text\n\nAppleyard D: Mind the gap: A bridge between industry and academia. Renew Energy Focus. 2017; 18: 36–38. Publisher Full Text\n\nvon Winterfeldt D: Bridging the gap between science and decision making. Proc Natl Acad Sci U S A. 2013; 110 Suppl 3: 14055–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBinoy VV: Introduction: When Science Meets the Public—Bridging the Gap. In Bridging the Communication Gap in Science and Technology. Springer Singapore, 2017; 1–9. Publisher Full Text\n\nMitacs: Accelerate | Mitacs. (Accessed: 3rd May 2018). Reference Source"
}
|
[
{
"id": "36084",
"date": "08 Aug 2018",
"name": "Michael Hendricks",
"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 presents a snapshot of career and funding-related concerns for early-career researchers in Canada, focused on a particular event held to discuss these issues. It is important that the motivations for and outcomes of this event be broadly communicated. However, as it stands the piece is unfocused and potentially confusing for its intended audience. In fact, it is not clear who the intended audience is, or what it hopes to accomplish. I think these are issues that can be addressed with reorganization of the content and bringing more focus to the points and arguments made.\n1. In Canada, there seems to be considerable confusion depending on the government agency, funder, or day of the week whether the term early-career researcher (ECR) refers to graduate students and postdocs (trainees), early-career faculty/PIs, or both. “Trainees” is itself a problematic term—designating the people who actually conduct the majority of federally-funded academic research “trainees” (particularly postdocs, who are by any measure professional scientists) is part of the justification for many of the labor conditions at issue in the piece.\nWhile trainees and junior PIs share many of the same concerns about overall support for academic research, they also have distinct policy interests and career stage-specific needs and incentives. From the point of view of the authors, perhaps the biggest concern of lumping them together is that, in many cases, the interests and incentives of early-career PIs align much more with the interests of established PIs and institutions—particularly with respect to management/labor relations with graduate students and postdocs.\n\nBecause this article is focused on graduate students and postdocs, and because it is presumably in part addressed to a policy audience who may be less-familiar with these distinctions and the details of academic career progression, I think a clearer distinction needs to be made. It would be most effective to focus on the concerns of trainees and transitions out of trainee status (outside and within academia), and just leave early-career PIs out of it.\n2. The authors need to do a much better job at defining terms and concepts and the current “lay of the land.” Specifically, it should not assume familiarity with the Naylor Report and surrounding advocacy efforts. How are trainees supported now—what is the relative proportion of support from fellowships, grants to their PI, institutional funds? What are the advantages of fellowship funding over support from a grant? How even is institutional support across Canada or across programs within universities? How would more fellowships benefit trainees?\nPolicy makers often do not know that the great majority of trainees are supported on grants. This makes the section on insufficient ECR funding particularly confusing, as it seems to argue against itself by pointing out that all three granting councils have policies to support junior PIs in place. As far as I know, there is no longer a broad sense that junior PIs are in crisis with respect to funding opportunities compared to more established PIs within the main granting programs of the councils—or that this was ever the perception at NSERC or SSHRC. The authors should be careful not to generalize from specific issues that arose in the context of the CIHR reforms to the Canadian research enterprise as a whole, or at least make it clear when they are talking specifically about health-related research. In any case, the relationship of ECI PI grants to funding to support trainees is not clear. Again, this is a major weakness of lumping ECI PIs and trainees together.\n3. Related to the above, if the goal is to influence policy, it needs to take a more opinionated tone, particularly in the Discussion. There are long lists of problems and solutions that read like a brainstorming session with very little synthesis or prioritization. How many of these problems are in effect the same problem (insufficient overall funding available through councils, not enough academic positions, inadequate mentoring or career resources)? What are the most urgent problems? What do you believe are the most practical and effective solutions? Which problems can be approached without additional money in the system, or cannot be addressed through funding policy (perhaps some of the issues around academic culture and mentorship)? Which can only be handled by institutions, which by funders, which by government? What three (or one or five) things should a grants council president or MP or university VPR come away from this piece understanding about what trainees need and want?\n4. The piece (and event) are strongly focused on biomedical science. This focus should be made clear up front. The issues faced by students and postdocs (funding, job markets, mentorship) are similar in some ways but also vary substantially across fields, even within CIHR, and are radically different for those who fall under the NSERC or SSHRC umbrellas.\n5. There are missed opportunities to frame the advocacy in terms of the purpose of research funding and the broader interests of the Canadian public, as represented by federal funding agencies. For example, an argument for shifting trainee support away from operating grants and toward more emphasis on fellowships could be made based on the goals of these programs. CIHR operating grants are awarded to conduct specific research on behalf of Canadians, there is no training mandate. In contrast, the purpose of CIHR fellowships is to create “scientific, professional, or organizational leaders within and beyond the health research enterprise.” This seems much more in line with training as envisioned in this paper. It is compatible with the diversification of training (e.g. to include other kinds of experience and professional training) while operating grants as they are currently constructed are not. This could be seen as an argument for more fellowships, or to lobby funders to specifically evaluate on mentorship and training in operating grants.\n\n6. Another example of this would be to make the case for why it is of benefit to Canada to have so many PhDs in the non-academic workforce—these arguments are articulated in the FSR and elsewhere. Right now, the piece can be read has emphasizing the needs of trainees, and it makes it much easier to be on your side if it is made equally clear how and why Canada needs you, your training, and your degrees. Should we be producing fewer PhDs? It would be one easy conclusion to draw from many of the arguments presented here, but it is a weak case for investing in better training.\nThe Canadian ECR community showed leading up to Budget 2018 that they can be forceful and effective advocates for good science policy. It was a major disappointment, then, that their specific concerns and trainee-related FSR recommendations were for the most part ignored. Continuing to advocate is essential, and I applaud all the effort that has gone into this. I think this piece will be a valuable snapshot of current conditions for trainees and should serve as a reference for institutions and policy-makers. However, to be most effective as a reference it needs more clarity and organization.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "36086",
"date": "13 Aug 2018",
"name": "Holly Witteman",
"expertise": [
"Reviewer Expertise My areas of expertise relevant to this manuscript are: Canadian research/funding policy",
"survey and workshop methods. (My primary research is interdisciplinary and is best described as the intersections of health informatics",
"health education/communication/decision making",
"user-centered design and human-computer interaction.)"
],
"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 describes the outcomes of a workshop conducted with n=136 participants, mostly postdoctoral fellows and graduate students, in Vancouver, Canada to discuss issues facing trainees. This event was modeled on a previous meeting held in Boston. The authors present a summary of participants’ views regarding challenges faced by trainees and some potential solutions to address these. This manuscript offers a well-written account of issues facing trainees in Canada, and has potential policy implications. However, I have quite a number of concerns about this manuscript, which I detail below. I hope these are helpful to the authors to improve their manuscript and achieve the impact that they hope to have in Canadian science and research policy.\nReason for Not Approved\nNote: I moved this point up to the top of my review after reflecting on the likely reaction of the authors and what they might first want to know from this review.\nResearch with human participants (a.k.a. \"human subjects research\") requires oversight by an research ethics board to be publishable in a reputable peer-reviewed journal. Was the survey approved by the relevant research ethics board (in this case, SFU, since that’s where the data collection occurred)? Did all survey respondents provide informed consent to have their responses analysed and published? If the answer to either of these questions is no, this is problematic. This lack of information about ethical approval gave me great pause when reviewing this manuscript. Current practice in many other journals for a manuscript describing data from human participants without a statement about ethical approval is to send it back before it even gets to the editor. This is the reason I selected Not Approved. I would prefer not to have to select this, especially as this is a trainee-led manuscript, but this is a required element for human subjects research to merit publication in a peer-reviewed journal. I can't approve this manuscript without this. I'm sorry.\n\nOverall\n\n1. Having been extensively involved in policy discussions about Canadian health research funding and funding policy, I began this review by reading the manuscript twice—once with my existing base of knowledge, and a second time, attempting to put myself in the place of someone who is less familiar with the past few years of changes and events in Canadian health research policy. My experience as a reader was significantly worse the second time. To enable this manuscript to contribute to the global literature on science/research policy, it would be preferable if the authors were to more carefully explain terminology and particularities of the Canadian research funding landscape. For example, a reader from outside Canada may not be familiar with #SupportTheReport or the Fundamental Science Review. Merely providing a citation puts the onus on the reader to understand things that the authors should make clear and understandable.\n\n2. There is often confusion about the definition of an early career researcher (ECR), particularly among policymakers. As it stands, this manuscript adds to the confusion. In the opening sentence, the authors refer to ECRs as a group that includes graduate students, postdoctoral fellows, and faculty within their first five years of appointment. This muddles the issues facing these groups. The issues faced by scientists who are still completing their training are considerably different from the issues faced by new faculty. Even within trainees, there are key differences between problems faced by graduate students versus postdoctoral fellows in Canada (for example, as the authors note: access to extended health benefits, employment insurance.) The workshop was organized by postdoctoral fellows and seems to have focused almost entirely on issues facing postdoctoral fellows and graduate students. The parent organization was postdoc-organized. ECR participants at the workshop seem to have been largely, or perhaps entirely (this is not clear due to missing data in Table 1), postdoctoral fellows or graduate students. Thus, aside from some introductory paragraphs about new faculty, which read as distinctly out of place within the rest of the manuscript, there is no apparent representation of new faculty’s views about any of these issues. To improve the coherence and clarity of the manuscript, I recommend that the authors (a) clearly define ECRs as graduate students and trainees, (b) put the definition in both the abstract and in the main text, and (c) remove all sections focusing on new faculty.\n\n3. What fields are covered in this manuscript and/or by the participants? The authors appear to be largely in biomedicine, with perhaps some representation in natural sciences? Many of the references and terminology hint to the majority of people involved in this work being in biomedicine, or, as termed at the Canadian Institutes of Health Research, “pillar 1” or “theme 1”. The parent organization, Future of Research, also focuses on biomedicine. The other pillars of health research (2: clinical research, 3: health services research, 4: social, cultural, environmental, and population health research) do not seem to have been represented. This is suggested, for example, by the focus on industry as the logical non-academic path for employment, which is not the case in these other areas. In addition, were researchers from other fields (e.g., engineering, social sciences, humanities) represented? The disciplinary backgrounds of the people involved should be explicitly stated and it should be clarified for the reader whether the findings and views presented in the paper reflect the entirety of the Canadian research landscape or are focused on one or more particular segments.\n\n4. Is this a policy brief with an appended report, a research paper, or an opinion/commentary paper (the category noted on the manuscript)? It presents aspects of all three of these. It is not possible for one manuscript to be all of these things. I would encourage the authors to pick one and do it well rather than trying to make this manuscript try to be all three simultaneously. The guidance offered by the F1000 website states quite clearly that, “Opinion articles must focus on previously published literature and not include new research and data.”\n\nAbstract\n\n5. The authors state, “however, Canadian policies are trailing behind the progress being made in other countries.” As far as I can tell, this is never again discussed nor referred to, and no citations are provided to support this statement.\n\nExecutive Summary\n\n6. The first major recommendation by the authors is to, “1. Improve ECR-targeted funding, including grants which provide operating costs for ECRs and support the transition from a PDF to junior faculty position, and recognising ECRs contributions to grants awarded to their supervising professors.” I have three concerns about this. First, experiences in other countries suggest that transitional grants may simply prolong one’s period of career uncertainty. In other words, rather than years of uncertainty ending at the point at which a scientist who wished to remain in academic science either secures a position or doesn’t, such transitional grants move the “up or out” point to several years into one’s faculty position. Second, are the authors concerned that such an option would impact negatively on start-up funds committed by institutions? Faculty start-up funds are already lower in Canada than, for example, in the US. Such a strategy risks further reducing institutions’ commitment to new faculty by expecting them to use their transitional grants in the place of providing start-up funds. Third and finally, can the authors please clarify what they mean by “recognising ECRs contributions to grants awarded to their supervising professors”? I think they are referring to the situation in which a trainee can either be listed on a grant as a co-investigator or be paid off the grant, but not both. However, this is not clear and should be stated explicitly.\n\nIntroduction\n\n7. The opening sentence refers to, “science, engineering, and social research,” in Canada. This is an unusual collection of terms. Where is health research in this list? What happened to the humanities? More importantly, this broad opening leads me as a reader to expect that this paper will address all research areas, when in fact, as noted above in comment 3, this does not appear to be the case. It would help readers orient themselves if the authors were to avoid inviting this confusion in the opening line of the manuscript.\n\n8. The authors state in their first paragraph, “However, there has been growing concern that current funding and training structures almost entirely ignore the best interests of ECRs.” I am not aware of any longitudinal analyses of such concerns. Can the authors please provide a citation to support the assertion that such concerns are growing? It would also be helpful to clarify what they define as “the best interests of ECRs.”\n\n9. In my view, the Introduction would be stronger overall with the first paragraph completely removed. The second paragraph opens in a more compelling way and more swiftly leads the reader to an understanding of the problems addressed in the manuscript. If the authors were to do this, the concerns I raise above in comments 7 and 8 would no longer exist.\n\nWorkshop Overview\n\n10. Given that the rest of the manuscript focuses on results from the second workshop session, it may be useful to have a summary of the first session beyond the list of people who presented. What did they present? Such a summary may or may not be possible at this date, depending on whether or not the session was recorded, but if it is possible, it would be a good addition to the paper.\n\n11. To preface this next set of comments, I will note that my own training and current research combine social science methods with other fields. Also, I will note that this is the third manuscript I have reviewed this year from biomedical/natural scientists or clinicians demonstrating a lack of awareness of best practices in social sciences, including workshop and survey methods. It is too late now, but I hope that in future, the authors will involve colleagues in social sciences to help improve their workshop/focus group methods, survey methods, and analyses of data from these. Considering the data they already have and the things they already did and can’t undo, this manuscript might be improved in the following ways: a. More detail is needed about the event. Who was invited to the event? How were invitations distributed? Were the organizers hoping to attract specific groups of people? How was this phrased in promotion materials? b. More detail is also needed about the survey process. The authors state, “Before leaving the Sessions, participants were asked to complete an Exit Survey, which included questions regarding demographic data, opinions about the outcomes of the Breakout Session, and feedback towards future events.” Does that mean that participants were asked to complete the survey on paper right there, or was this a reminder issued to participants to please complete the online survey that would be sent to them later? Was the survey pre-tested in any way? Was it paper or online? The supplementary materials suggest it was a paper survey but it should be explicitly stated. c. Can the authors please better explain how they grouped ideas? They state, “Note that the solutions supplied were not tied to specific problems that had been raised, and the grouping of this information into general themes has been performed ex post.” How was this done? By whom? Did the person/people doing this grouping have training and experience in qualitative research? Were two independent analysts involved (if so: what was their kappa) or did one person make these decisions? What qualitative methods were used?\nResults\n\n12. Like the Methods, the Results section is missing key data we would expect from this type of work. The authors should consult the Equator Network (equator-network.org) and consider the available reporting guidelines. Although I don’t believe there exists a reporting guideline for a policy workshop such as this, the authors might consult and consider which elements of STROBE (cohort study) and COREQ (focus group) apply to their work. If their survey was conducted online, I would suggest they also consult CHERRIES.\n\n13. What was the survey response rate (number of completes/136)? It looks like the response rate was, at most, about 50%. That’s lower than we might typically expect for an event-linked evaluation survey. How did this happen? Did people leave without completing the survey? Did they not respond to requests to complete it online?\n\n14. Several comments about Table 1: a. In the questionnaire provided in the supplementary file, the categories of positions are: Postdoc, Graduate/PhD student, Faculty, Other (please specify). These same categories (and subcategories as appropriate) should be used in this table. If no participants belonged to a given category (e.g., Faculty) then this should be shown in Table 1 with zeroes. b. Particularly with n<100, it is inappropriate to report mean age to one decimal point. Even if n>=100, it isn’t useful precision. There isn’t a meaningful difference between 32 years old and 32.2 years old. Please report mean age as integers. c. Again re: reporting of age, for these kinds of descriptive statistics, one would typically report a measure of central tendency (usually either sample mean or median, depending on distribution) as well as a measure of dispersion (sample standard deviation if mean is used, interquartile range if median is used, sometimes full range.) If the authors are unsure which is most appropriate for their data, they may wish to consult with the statistical consulting service at their institution. These services often provide free or inexpensive consulting to students and trainees.\n\n15. Several comments about Tables 2-5: a. Tables provide grouped listings of perceived problems and suggested solutions. I would suggest naming them this way rather than problems and solutions, to be clearer. b. Many of these points are opaque. For example, what do the following points mean? “Getting the science to the public” (What science? Which public? How would one define having gotten science to the public?) “Short term pilot project grants with no expectation of returns” (Surely this is not suggesting funding with no anticipated returns whatsoever?) “Quality of the peer-review system” (This is so vague it’s meaningless. Can this be better explained?) “Professional Development (PD) topics too centralized” (What does this mean?) “Lack of extracurricular training” (What is this?) “Pressure for sycophancy to assist career” (What does this mean?) This is not an exhaustive list. There are many other points whose meanings are unclear. c. Some of the points display a lack of awareness about the research landscape. For example, one such comment is, “Engage reviewers from outside academia for relevant expertise on knowledge translation/knowledge mobilization.” While, in my view, it’s good to have knowledge users and knowledge brokers from outside academia reviewing relevant grants, the authors might consider consulting the Knowledge Translation panel from the last cycle of CIHR Project grants (KTR in the list available here: cihr.ca/e/50845.html), the reports from KT Canada, or the list of people responsible for KT at each province’s SPOR SUPPORT unit and perhaps reflect on whether they are aware of academic expertise in this area. Other such examples include, “Mandated staff composition of labs and groups (by funding agencies).” This is outside the purview of funding agencies. d. Some of the points contradict each other. For example, directly one after the other are the suggested solutions, “Increased funding for translational and applied research,” and, “Funding devoted to basic research.” e. Some of the points appear to be poorly classified. For example, “Convince public that research has value even without immediate application,” is listed under Problems while, “Convey to the public that all science matters, not just ‘hot topics’,” is listed under Solutions. f. A number of the points repeat themselves across tables and some categories lack meaning (e.g., “Assorted.”) See comment 18 below for a potential suggestion of how to address this. g. While it’s lovely to have the raw data supplied (thank you), the data themselves are even more difficult to interpret than the points in the tables.\n\n16. The workshop included trainees in Vancouver from the University of British Columbia and Simon Fraser University. If I am understanding correctly, there were no participants from other academic institutions. Is that correct? This is completely understandable given the nature of the event and the geography of Canada. However, it means that some of the ideas come across as under-informed. For example, Canadian institutions already offer mentorship training to PIs. Université Laval offers such training and, while it is not mandatory, it is strongly recommended. (To illustrate: I have taken it, as has every professor in my unit.)\n\n17. Related to the above, some aspects of the report read as national; for example, statistics about Canadian funding, while others read as local; for example, references to, “Vancouver’s high cost of living.” The authors might wish to reflect on whether they wish to add more national context (e.g., “high cost of living in cities like Vancouver and Toronto”) or focus even more on issues that are specific to trainees at institutions in Vancouver.\n\n18. Similar to the above, and related to point 15c above, one of the suggestions was, “Training for KT in grant proposals.” Such training exists; for example, trainees may wish to explore the training programs available through the Li Ka Shing Knowledge Institute at St. Michael’s Hospital, Toronto (knowledgetranslation.net), KT Canada, or other organizations. Given they are in BC, they should contact their provincial Strategy for Patient-Oriented Research (SPOR) Support for People and Patient-Oriented Research and Trials (SUPPORT) unit (bcsupportunit.ca). BC’s KT component is led by Dr. Linda Li at UBC. They may also wish to consult the list of resources here cihr.ca/e/49443.html. This is but one example of many in which the authors present perceived problems and suggested solutions with no acknowledgement that such solutions already exist and, indeed, have existed for quite some time. (To illustrate: I attended a KT Canada Summer Institute as a PhD student in 2008.) Others include things like reviewer training, which is already underway at CIHR, or networking opportunities, which raises the question of whether the participants have made use of the existing networking opportunities available to them? It took me less than 60 seconds to search for ‘biomedicine networking event Vancouver’ and find three recent such events. Of course, it is not reasonable to expect trainees (nor anyone) to know every opportunity that’s out there for them, but this manuscript would be improved by better mapping the perceived problems and suggested solutions onto current opportunities to identify true gaps. Right now, unfortunately, the results come across as a very long, semi-organized laundry list of perceived problems and suggestion solutions that range from very insightful to woefully under-informed. This is a shame, as the authors and their colleagues clearly put a great deal of time into understanding the context of research funding, organizing the workshop, and preparing this report. To help this manuscript better capitalize on that effort and the expertise that the authors and workshop participants bring re: their experiences as trainees, the authors might consider identifying a few key issues and solutions and contextualizing them in the broader literature. For example, the structure of postdoctoral training programs may be present as a key theme. This could be put in the context of previous literature such as: PMIDs 27543634, 25673353, 25771193, etc. This would also help reduce the large number of repetitive tables. The authors might consider consulting the chart in the paper published by Future of Research as an example of how taking the time to synthesize can help to better present data from a workshop such as this. Participants’ ratings in Tables 6-9 might help identify the perceived problems and suggested solutions that are most important and worth highlighting. Grouping and identifying key issues and solutions is particularly defensible given the authors’ own statement in the Discussion that, across breakout sessions, “solutions showed significant overlap.”\n\n19. There is quite an emphasis on mentorship in the results and discussion. I wonder if the trainees who participated in the workshop have realistic expectations regarding what mentorship is and should be. Even the best-trained, most well-meaning PI may have trouble mentoring trainees seeking positions in industry, simply out of lack of experience in industry and a weak or nonexistent industry network. If the authors choose to continue discussing mentorship as one of their key issues, the manuscript might be improved by drawing more on the relevant literature. I recommend the authors consult work by Dr. Sharon Straus, who has done a great deal of work studying mentorship in academic medicine. Her book written with Dr. David Sackett is comprehensive, and the authors may wish to consult papers such as this systematic review of practices of good mentors and good mentees (Sanbunjack et al1. Of note, the authors of the systematic review found that good mentees should be in the ‘driver’s seat’, and should be respectful, organized, and committed. The point about mentees being in the driver’s seat is particularly salient here, because if trainees wish to be better equipped to obtain a job in industry, that is a goal they should be driving towards. Reading some of the comments about mentorship, it is not evident that this is well-understood by workshop participants. It is reasonable to expect one's supervisor to support one's career progression. All supervisors should do this. However, in the same way that you wouldn't reasonably expect your hockey coach to help you prepare to play water polo, aiming for a career in industry will require a broader mentorship team than just your PI.\n\nDiscussion\n\n20. The authors seemed to choose certain issues to highlight in the Discussion, for reasons that are not entirely clear. If they choose to follow my suggestion in comment 18 above, their selection of which issues to highlight will be driven by the data from exit surveys. Alternatively, if they choose to keep the current format, they should give the reader more of an indication about why they are focusing on these specific issues.\nMinor comments\n\n21. The authors open a paragraph in the Introduction, “Despite these important concerns, there is hope for the future.” Although this is categorized as an opinion article, this level of editorializing is probably unnecessary. However, I leave it to the authors to determine what kind of article they wish to write.\n\n22. In paragraph 8 of the Introduction, the authors state, “Furthermore, it proposes $210 million over 5 years for Canada Research Chair (CRC) appointments specifically for ECRs.” This did not align with my interpretation of the budget, so I went back and re-read page 89 of the full budget document. As I recalled, the budget signaled a desire for these CRCs to be allocated to more junior researchers, however, this is not the same as the authors’ definitions of ECRs. The Government of Canada may choose to use overly broad and vague terms like “early career researcher” to refer to the recipients of Tier 2 Canada Research Chairs, meaning people with faculty appointments and who are up to 10 years past the receipt of their PhD plus extensions for leaves. However, the authors should strive to be clearer. Furthermore, the Government of Canada “expects” the granting councils to target new funding to early career researchers, which means this will likely occur, but this is a little more nuanced than the authors’ statement that these funds are “specifically for” ECRs.\n\n23. At the beginning of the Results, the authors note, “From the 65 responses, the average age of attendees was 33 years old, with only Session 3 showing a notable digression from this average.” Unless the authors believe that the higher mean age is meaningful for some reason and may explain differences between this session and others, I’m not sure this adds all that much to the paper. Also, it looks to me that this is likely because of the preponderance of non-trainee participants in this session (e.g., all the university staff at the workshop attended this session, constituting a full quarter of the people in the session.)\n\n24. Were the authors among the n=136 participants? Were authors’ views included in the data presented? If the answer to either of these is yes, this should be explicitly stated.\n25. In the Discussion, the authors state that, “Recent campaigns for increased research funding in Canada have led to the creation or increased the activity of ECR advocacy groups such as the Association of Canadian Early Career Health Researchers (ACECHR) …” ACECHR was not formed to campaign for increased research funding, though the organization has certainly worked on this as one of its more recent activities. The organization was formed in reaction to changes to the grant program structure and grant review mechanisms at the Canadian Institutes of Health Research (CIHR) in 2014. Modeling by Dr. Hendricks suggested that new investigators in their first five years as faculty were poised to lose approximately $30M in grant funding in the new system, and indeed, the first new cycles suggested that without the stopgap measures put in place at the CIHR in response to ACECHR advocacy, that size of gap would have occurred. As the person who wrote the first draft of many of ACECHR's documents and now knows all the train numbers and schedules between Quebec City and Ottawa by heart, I can assure you that ACECHR's participation in advocating for sustainable federal research funding was neither the reason for ACECHR's creation nor was it an increase in ACECHR's activity. Joining in with a number of other organizations, with practically the whole Canadian research community on side, was far less work than the activities ACECHR did alone. Please correct this.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? No\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1087
|
https://f1000research.com/articles/7-697/v1
|
04 Jun 18
|
{
"type": "Research Article",
"title": "TMD symptoms and vertical mandibular symmetry in young adult orthodontic patients in North Sumatra, Indonesia: a cross-sectional study",
"authors": [
"Ervina Sofyanti",
"Trelia Boel",
"Benny Soegiharto",
"Elza I. Auerkari",
"Ervina Sofyanti",
"Trelia Boel",
"Benny Soegiharto"
],
"abstract": "Background: Temporomandibular joint disorder (TMD) includes symptoms of pain and dysfunction in the muscles of mastication and the temporomandibular joint. Differences in vertical condylar height, observed in the assessment of mandibular asymmetry, is a structural alteration that represents a risk factor for TMD. The study aimed to evaluate the association between TMD symptoms and vertical mandibular symmetry in young adult orthodontic patients in North Sumatra, Indonesia. Methods: The cross-sectional study included 18-25-year-old (mean ± SD, 21.9 ± 2.0 years) old orthodontic patients admitted to the Dental Hospital of Universitas Sumatera Utara, Medan, between June 2016 and March 2017. Vertical mandibular asymmetry was assessed from all 106 subjects using Kjellberg’s technique from pre-treatment panoramic radiographs. The TMD symptoms were assessed by structural interviews using modified questionnaires based on Temporomandibular Disorder Diagnostic Index and Fonseca’s Anamnestic Index. Results: Of the 106 subjects, 26 (24.5% of the total) with vertical mandibular symmetry and 39 (36.8%) with vertical mandibular asymmetry were positive for TMD symptoms. By contrast, 17 patients (16.0% of the total) with vertical condylar symmetry and 24 patients (22.6%) with vertical mandibular asymmetry were regarded negative for TMD symptoms. There was no significant difference (p=0.520) in TMD symptoms based on vertical mandibular symmetry. Conclusion: The results from this studied Sumatran population indicate that there are common TMD symptoms in young adult orthodontic patients, but there is no significant association between vertical mandibular asymmetry and TMD symptoms. Further study on the development of TMD, mandibular asymmetry and treatment planning for growing patients is suggested, using longitudinal and transitional approaches.",
"keywords": [
"vertical mandibular asymmetry",
"temporomandibular disorder"
],
"content": "Introduction\n\nThe goal of orthodontics for young patients is to provide a functional occlusion to give harmony in the dental arrangement, the anatomy of temporomandibular joints, and the activity of the masticatory muscles in later adulthood1. The assessment of symmetry is important in comprehensive orthodontic treatment, as well as in malocclusions and dental evaluation; it is related to this aforementioned goal of orthodontic treatment, especially the functional and aesthetic evaluation of the craniofacial region2,3. The asymmetries in human facial structures affect the skeleton, muscles and corresponding attached facial tissues. The prevalence of mandibular asymmetry is highest when compared to asymmetry of cranial base and maxillary arch in the human skull4. In orthodontic assessment, it is also important to consider whether the development of mandibular asymmetry could affect jaw, head and even shoulder movement, creating problems that should not occur in healthy subjects5–7.\n\nAccording to a study on 8–30-year-old subjects in Jakarta, Indonesia, based on questionnaires and posteroanterior radiography, the main risk factor of mandibular asymmetry is temporomandibular joint disorder (TMD)8. Asymmetry in the vertical dimension, based on posteroanterior radiography, was significantly correlated with temporomandibular joint internal derangement in a study on 187 Japanese subjects with pre-orthodontic mandibular asymmetry and a mean age of 23.9 years9.\n\nThe etiology, diagnosis and management of mandibular asymmetries focuses particularly on developmental asymmetries. Increasing mandibular asymmetry with bilateral asymmetry of morphological traits causes malfunctions in developmental homeostasis, associated with environmental and/or genetic stresses. The development of morphological asymmetry may serve as a risk factor for disorders of developmental origin if these stresses are involved. The biopsychosocial model is hypothesized to be the most accepted theory for developmental asymmetry and complexity of TMD5,10. Studies concerning TMD and the relevance of orthodontic treatment suggest that the achievement of a balance in the dynamic occlusion is necessarily related to the development of mandibular symmetry as a part of the successful management of TMD2,10–13.\n\nPanoramic radiography is commonly used to assess the extent of mandibular asymmetry, as bilateral information is provided in routine dental practice. The asymmetry indices of mandibular height based on the ratio of condylar height (CH) and ramus height (RH) asymmetry, according to Habet’s method and Kjellberg’s technique, correlated significantly between TMD and non-TMD patients14,15. However, Kjellberg’s technique is easier in terms of identifying the points and measurements and compares both sides because the measurement of CH from the highest point of condylar head to the mandibular notch; this differs from Habet’s method, which uses the distance from the highest point of the condyle to the most lateral point of the condyle16. A previous study of 100 patients with TMD in the Seoul National University Dental Hospital between 2009 and 2011 found that asymmetry resulting in more than a 4.37% difference between mandibular heights may increase the risk of TMD and was positively correlated with the incidence of arthritic change in the temporomandibular joint of patients with TMD, although this does not necessarily indicate a direct cause-and-effect relationship7. By contrast, there was no statistically significant difference found between the severity of signs and symptoms of TMD based on vertical mandibular asymmetry, assessed using Habets' method and Kjellberg's techniques in 12–65-year-old patients17. Since TMD and mandibular asymmetry are complex issues that cover a large variety of symptoms, this study aims to analyze the association between TMD symptoms and vertical mandibular asymmetry measured using Kjellberg’s technique in young adults that sought orthodontics treatment at the Dental Hospital in Universitas Sumatera Utara, Medan, Indonesia.\n\n\nMethods\n\nThis cross-sectional study was conducted at the Dental Hospital of the Faculty of Dentistry, Universitas Sumatera Utara between June 2016 and March 2017. The Health Research Ethical Committee of the Medical Faculty, Universitas Sumatera Utara (100/DATE/KEPK FK USU-RSUP HAM/2017) approved the study. All 106 subjects were 18–25-year-old patients that attended the orthodontic clinic for a consultation, with following eligibility criteria: no previous orthodontic or occlusal adjustment treatment, no history of traumatic facial injury or congenital disease. Patients who attended the orthodontic clinic had been informed that if they provided written informed consent, they would be included in a survey. In compliance with the Declaration of Helsinki, the consenting participants were asked to fill in the questionnaire on the Temporomandibular Disorder Diagnostics Index (TMD-DI; Table 1) at the Orthodontics clinic, Dental hospital Faculty of Dentistry, Universitas Sumatera Utara. The assessment of TMD symptoms were based on TMD-DI with categories of TMD-positive or TMD-negative18. The assessment of stress (Table 2) was using questions for a modified Fonseca’s Anamnestic Index, related to bruxism, joint noise and nervousness19.\n\n*Modification to questionnaires to assess stress.\n\nSubjects were referred to take panoramic radiography with exposure parameters (80 kV, 15 mA, 12 seconds) in the Pramita Clinic and laboratory, Medan, North Sumatera, Indonesia. Figure 1 shows the classification of vertical mandibular height symmetry. A percentage symmetry of 93.7% or lower was defined as vertical mandibular asymmetry based on Kjellberg’s technique17. Figure 2 showed the measurement of vertical mandibular symmetry manually on tracing paper. Figure 1 showed the points in measuring the vertical mandibular symmetry as follows: CH is defined as the distance from CO (the highest point of the condylar head) to the mandibular notch (the deepest point between the coronoid process and the condylar process). RH is the distance from CO to the gonion. In order to obtain vertical mandibular symmetry based on the ratio of condylar and ramus height (Kjellberg's technique), the numerator should be smaller than the value resulting from the division of CH and RH/MH regardless of whether it corresponds to the right or left joint. The formula is as follows:\n\nKjellberg symmetry index = (CHRHA)(CHRHB)\n\nTo determine the random error, inter-rater (T.B. and E.S.) and intra-rater (E.S.) measurements of variables in this study were randomly selected from 20 panoramic radiographs. Finally, this study used intra-rater measurement as reference data for assessing vertical mandibular symmetry, which repeated the measurements 1 week after the first examination, while blinded to the initial values. The validity and reliability, measured using Cohen’s κ, showed moderate agreement for inter-rater measurements (κ=0.538) whilst intra-rater measurements (κ=0.674). Cronbach’s alpha analysis was used to provide reliability measurements of questionnaires in analysing items and total scores in the modified Fonseca’s Anamnestic Index (p>0.05). However, the point related to clenching or grinding was omitted as it failed to show the validity and reliability of criteria (p=0.023). Then any information regarding FAI was only collected for additional information and the TMD-DI used as early screening for analyzing the TMD symptom20,21. Significance of association between TMD symptoms and vertical mandibular symmetry (or asymmetry) was evaluated using a chi-squared test, with assumed significance at p < 0.05. All statistical analyses were performed using SPSS, version 18.0 (SPSS, Inc., Chicago, IL, USA).\n\n\nResults\n\nFrom 106 young adult orthodontic patients (mean ± SD, 21.9 ± 2.0 years old), TMD symptoms were present in 24.53% (n=26) of patients with vertical mandibular symmetry and in 36.79% (n=39) with vertical mandibular asymmetry. On the other hand, 16.04% (n=17) of patients with vertical mandibular symmetry and 22.64% (n=24) with vertical mandibular asymmetry had no TMD symptoms (Table 3). There was no significant difference (p=0.520) in the occurrence of TMD symptoms based on vertical mandibular symmetry (Table 3).\n\n\nDiscussion\n\nThe most frequent TMD symptoms include joint noises, followed by reduced mandibular mobility, muscular pain and joint pain. TMJ status is an important factor to consider in orthodontic diagnosis because related to imbalance occlusion and the development of mandibular asymmetry3,7. A previous study suggested that MRI or arthrography could be used as a valuable radiographic assessment in analyzing condylar hyperplasia or discus displacement in mandibular asymmetry and TMJ24. The assessment of posteroanterior cephalometric variables could be used as a key factor for evaluating the presence of unilateral TMD25. TMD signs and symptoms with multifactorial etiologies have been reported as a risk factor in patients with mandibular asymmetry that had menton deviations in Indonesia based on postero-anterior radiography8,20. In the early detection of mandibular asymmetry related to TMJ disharmony, panoramic radiography is routinely used in the clinic for orthodontic purposes, compared to bilateral tomography of TMJ and postero-anterior radiography. This technique allows a bilateral view and adequate information on vertical and horizontal measurements as early diagnostic evaluation of mandibular asymmetry because it focus mainly on intercondylar asymmetries and gonial angle measurements26–30. Previous studies about panoramic radiography reported that horizontal measurements of anatomic landmarks in the panoramic radiograph tend to be particularly unreliable because of the nonlinear variation in magnification at different object depths, whereas vertical and angular measurements are acceptable, provided the patient’s head is positioned properly26,27,30,31. This study used Kjellberg’s technique because it is easier to identify the condylar height using this technique than Habet’s method. Habet’s method is more complicated when making reference points of the most lateral point of the condyle due to variation in the condylar anatomy16,17,31.\n\nEarly detection of TMD in malocclusion, especially related to mandibular asymmetry before establishing orthodontic therapy, is mandatory for interdisciplinary approaches for any dentofacial treatment nowadays2,11,32,33. Some questionnaires can be used as a tool to achieve early detection of TMD. Fonseca’s Anamnestic Index has frequently been used to classify individuals according to TMD severity category, from no TMD to mild, moderate and severe TMD, to screen TMDs in Brazilian women with regards to anxiety as a stress factor19. The TMD-DI was developed by Himawan et al. in 2006, has been applied in the study the characteristics of TMD and other risk factors in the Indonesian society8,18,20,21. In our study, we modify the questions regarding anxiety as stress factor to detect the severity of TMD symptom. In validity and reliability analysis, There was a question regarding clenching and grinding habit was eliminated due to no significant difference in validity and reliability analysis, so this study executed to analyze the TMD symptom based on FAI data. Eventhough this study only used the TMD-DI, there were a higher prevalence of TMDs in both of symmetry and asymmetry vertical mandibular of these young adult orthodontic patients (mean age ± SD, 21.9 ± 2.0 years old). Based on the aforementioned goal of orthodontic treatment, the clinician should be aware of TMD symptoms in orthodontic treatment related to functional efficiency. However, the differences in pain threshold might be a distraction factor while answering the questions to assess TMD symptom. Then, the proper clinical examination of the temporomandibular joint should be considered in orthodontic patients.\n\nFundamentally, orthodontic treatment should create a balanced and stomatognathic system, especially the temporomandibular joint. One element of this balance is craniofacial symmetry, which is frequently subject to discussion between clinicians and is the subject of multiple different studies in the last decades2,11,12. Although perfect craniofacial symmetry does not exist in nature, gross abnormalities in symmetry are considered as a major cause of non-dental pain in the orofacial pain region10,12,34. The distribution of TMD symptoms is higher than that of non-TMD symptoms in orthodontic patients with and without vertical mandibular symmetry. However, there was no significant difference (p=0.502) in TMD symptoms based on the presence or absence of vertical mandibular symmetry (Table 3) since TMD the etiology of TMD is multidimensional35 and asymmetry of condylar width, height and length as common features in TMD based on 3D-computed tomography36,37. Indonesia, as a developing country, still uses panoramic digital radiography as initial evidence for planning early orthodontic intervention and avoiding the progression of asymmetries38.\n\nAccording to McNamara, orthodontic treatment performed during adolescence does not alter TMD risk, as TMD with mandibular asymmetry may increase with age, with no evidence originating during orthodontic treatment33. This condition is due to the asymmetrical function and activity of the jaws, and the different development of the right and left sides of the mandible. The morphology of the condyle on the deviated side differs from the non-deviated side in mandibular asymmetry, indicating the association between asymmetrical jaw function and joint remodeling, based on 3D-cone beam computerized tomography36,37. The present cross-sectional study concerning mandibular symmetry and TMD in young adult orthodontic patients in North Sumatra indicates that asymmetry has been an adaptive response to functional demands because the mandible adapts to mandibular deviations. The modelling of condyle and glenoid fossae, as well as higher appositional growth in the gonion region during jaw function, will influence skeletal and dental pattern in later adulthood.\n\nIt is vital for any clinician who is involved in altering the patient’s dentofacial appearance and stomatognatic function to consider the mandibular symmetry and TMJ function, whether through orthodontics, facial growth modification, corrective jaw surgery or any aesthetic dentistry. In the future, although the result in Table 3 showed a non-significant correlation, the TMD-DI as early screening for TMD might require panoramic radiography with postero-anterior radiography or 3D-cone beam computerized tomography to analyze the complexity of development of TMD and mandibular asymmetry. In this study, the orthodontic patients presented with complex stomatognati problems, such as missing posterior teeth which regardless the missing duration. This condition could affect the development of TMD symptoms and vertical mandibular asymmetry; this matches the study by Halicioglu et al., which that reported a slight difference in the vertical mandibular symmetry index was found in patients with early unilateral mandibular first molar extractions39.\n\nMandibular asymmetry and TMD are two common features associated with increased bilateral asymmetry in morphological traits which might involve environmental and/or genetic stresses as etiologies in breakdown in developmental homeostasis. The etiopathogenesis of TMD, which is a common feature in mandibular asymmetry, is poorly understood, because the complexity of biomechanical, neuromuscular, bio-psychosocial and biological factors has contributed to this disorder6,35. Clinicians should note that the complexity of dentofacial variation in orthodontic patients indicates in part why most treatment approaches for malocclusions with TMD are directed to the symptoms rather than to etiology. However, a combination of questionnaires (as diagnostic indexes) and radiography analysis indicates that susceptibility to fluctuating asymmetry is increasing. In the future, some translational approaches with the identification of molecular regulators of cell proliferation in the condylar cartilage, coupled with these phenomena, might carry this finding into the clinical setting. Expanding the fields of phenomics and genomic medicine to understand why asymmetric function occurs is required to achieve personalized orthodontic treatment in young orthodontic patients40. Stress might also have a role in the appearance of developmental disorders and required comprehensive diagnostic tools11,19,35.\n\n\nConclusions\n\nTMD symptoms appear common in the studied young adult orthodontic patients from North Sumatra, but no significant association was observed between vertical mandibular asymmetry and symptoms of TMD. Further study on the development of TMD, mandibular asymmetry and treatment planning for young patients is suggested, using longitudinal and transitional approaches.\n\n\nData availability\n\nDataset 1. All radiographic images taken of the patients. Answers to the original Indonesian language questionnaire are also present. DOI: 10.5256/f1000research.14522.d20535922.\n\nDataset 2. Vertical mandibular symmetry measurements using Kjellberg’s technique, alongside responses to each questionnaire. A key is present in the “Questionnaires” sheet. DOI: 10.5256/f1000research.14522.d20536023.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors wish to thank all the participants in the study, and Derek Diong and Tommy Kwan for their technical assistance in organizing the samples of the study, as well as Dr Putri Eyanoer of Faculty of Medicine, Universitas Sumatera Utara, for the statistical analysis.\n\n\nReferences\n\nIngervall B: Functionally optimal occlusion: the goal of orthodontic treatment. Am J Orthod. 1976; 70(1): 81–90. PubMed Abstract | Publisher Full Text\n\nConti AC, Oltramari PV, Navarro Rde L, et al.: Examination of temporomandibular disorders in the orthodontic patient: a clinical guide. J Appl Oral Sci. 2007; 15(1): 77–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgrawal M, Agrawal JA, Nanjannawar L, et al.: Dentofacial Asymmetries: Challenging Diagnosis and Treatment Planning. J Int Oral Health. 2015; 7(7): 128–31. PubMed Abstract | Free Full Text\n\nSevert TR, Proffit WR: The prevalence of facial asymmetry in the dentofacial deformities population at the University of North Carolina. Int J Adult Orthodon Orthognath Surg. 1997; 12(3): 171–176. PubMed Abstract\n\nNaugler CT, Ludman MD: Fluctuating asymmetry and disorders of developmental origin. Am J Med Genet. 1996; 66(1): 15–20. PubMed Abstract | Publisher Full Text\n\nDong Y, Wang XM, Wang MQ, et al.: Asymmetric muscle function in patients with developmental mandibular asymmetry. J Oral Rehabil. 2008; 35(1): 27–36. PubMed Abstract | Publisher Full Text\n\nNoh JY, Lee JY: Relationship between Mandibular Asymmetry and Temporomandibular Disorders. J Oral Med Pain. 2014; 39(3): 100–106. Publisher Full Text\n\nPurbiati M, Purwarnegara MK, Kushandy L, et al.: Prediction of Mandibulofacial Asymmetry using Risk Factor Index and Model of Dentocraniofacial Morphological Pattern. J Int Dent Med Res. 2016; 9(3): 195–201. Reference Source\n\nBuranastidporn B, Hisano M, Soma K: Temporomandibular joint internal derangement in mandibular asymmetry. What is the relationship? Eur J Orthod. 2006; 28(1): 83–88. PubMed Abstract | Publisher Full Text\n\nBhat S: Etiology of temporomandibular disorders: the journey so far. Int Dent SA. 2010; 12(4): 88–92. Reference Source\n\nBourzgui F, Aghoutan H, Diouny S: Craniomandibular disorders and mandibular reference position in orthodontic treatment. Int J Dent. 2013; 2013: 890942. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSharma S, Gupta DS, Pal US, et al.: Etiological factors of temporomandibular joint disorders. Natl J Maxillofac Surg. 2011; 2(2): 116–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSugiaman DH, Himawan LS, Fardaniah S: Relationship of Occlusal Schemes with the occurrence of Temporomandibular disorders. J Dent Indones. 2011; 18(3): 63–67. Publisher Full Text\n\nHabets LL, Bezuur JN, Naeiji M, et al.: The Orthopantomogram, an aid in diagnosis of temporomandibular joint problems. II. The vertical symmetry. J Oral Rehabil. 1988; 15(5): 465–71. PubMed Abstract | Publisher Full Text\n\nKjellberg H, Ekestubbe A, Kiliaridis S, et al.: Condylar height on panoramic radiographs. A methodologic study with a clinical application. Acta Odontol Scand. 1994; 52(1): 43–50. PubMed Abstract | Publisher Full Text\n\nFuentes R, Engelke W, Bustos L, et al.: Reliability of Two Techniques for Measuring Condylar Asymmetry with X-Rays. Int J Morphol. 2011; 29(3): 694–701. Publisher Full Text\n\nIturriaga V, Navarro P, Cantin M, et al.: Prevalence of Vertical Condilar Asymmetry of the Temporomandibular Joint in Patients with Signs and Symptoms of Temporomandibular Disorders. Int J Morphol. 2012; 30(1): 315–321. Publisher Full Text\n\nHimawan LS, Kusdhany L, Ismail I: Diagnostic index for temporomandibula disorder in Indonesia. Thai J Oral Maxillofac Surg. 2006; 20(2): 104–110.\n\nCampos JA, Carrascosa AC, Bonafé FS, et al.: Severity of temporomandibular disorders in women: validity and reliability of the Fonseca Anamnestic Index. Braz Oral Res. 2014; 28(1): 16–21. PubMed Abstract | Publisher Full Text\n\nBoel T, Sofyanti E, Sufarnap E: Analyzing Menton Deviation in Posteroanterior Cephalogram in Early Detection of Temporomandibular Disorder. Int J Dent. 2017; 2017: 5604068. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTanti I, Himawan LS, Kusdhany L: Development of Questionnaire to Determine the Etiology of Temporomandibular Disorders. Int J Clin Prev Dent. 2014; 10(2): 103–108. Publisher Full Text\n\nSofyanti E, Boel T, Soegiharto B, et al.: Dataset 1 in: Analyzing of TMD symptom and Vertical Mandibular Symmetry in Young Adult Orthodontic Patients at Dental Hospital Universitas Sumatera Utara. F1000Research. 2018. Data Source\n\nSofyanti E, Boel T, Soegiharto B, et al.: Dataset 2 in: Analyzing of TMD symptom and Vertical Mandibular Symmetry in Young Adult Orthodontic Patients at Dental Hospital Universitas Sumatera Utara. F1000Research. 2018. Data Source\n\nWestesson PL, Tallents RH, Katzberg RW, et al.: Radiographic assessment of asymmetry of the mandible. AJNR Am J Neuroradiol. 1994; 15(5): 991–999. PubMed Abstract\n\nAlmăşan OC, Băciuţ M, Hedeşiu M, et al.: Posteroanterior cephalometric changes in subjects with temporomandibular joint disorders. Dentomaxillofac Radiol. 2013; 42(1): 20120039. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKambylafkas P, Murdock E, Gilda E, et al.: Validity of panoramic radiographs for measuring mandibular asymmetry. Angle Orthod. 2006; 76(3): 388–393. PubMed Abstract\n\nVan Elslande DC, Russett SJ, Major PW, et al.: Mandibular asymmetry diagnosis with panoramic imaging. Am J Orthod Dentofacial Orthop. 2008; 134(2): 183–192. PubMed Abstract | Publisher Full Text\n\nHazan-Molina H, Molina-Hazan V, Schendel SA, et al.: Reliability of panoramic radiographs for the assessment of mandibular elongation after distraction osteogenesis procedures. Orthod Craniofacial Res. 2011; 14(1): 25–32. PubMed Abstract | Publisher Full Text\n\nSilvestrini-Biavati F, Ugolini A, Laffi N, et al.: Early diagnostic evaluation of mandibular symmetry using orthopantomogram. Indian J Dent Res. 2014; 25(2): 154–9. PubMed Abstract | Publisher Full Text\n\nAkcam MO, Altiok T, Ozdiler E: Panoramic radiographs: a tool for investigating skeletal pattern. Am J Orthod Dentofac Orthop. 2003; 123(2): 175–181. PubMed Abstract | Publisher Full Text\n\nHirpara N, Jain S, Hirpara VS, et al.: Comparative Assessment of Vertical Facial Asymmetry Using Posteroanterior Cephalogram and Orthopantomogram. J Biomed Sci. 2017; 6(1). Publisher Full Text\n\nAlmăşan OC, Băciuţ M, Almăşan HA, et al.: Skeletal pattern in subjects with temporomandibular joint disorders. Arch Med Sci. 2013; 9(1): 118–126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcNamara JA Jr: Orthodontic treatment and temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997; 83(1): 107–117. PubMed Abstract | Publisher Full Text\n\nChia MSY, Naini FB, Gill DS: The Aetiology, Diagnosis and Management of Mandibular Asymmetry. Orthod Updat. 2008; 44–52. Reference Source\n\nChisnoiu AM, Picos AM, Popa S, et al.: Factors involved in the etiology of temporomandibular disorders - a literature review. Clujul Med. 2015; 88(4): 473–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYáñez-Vico RM, Iglesias-Linares A, Torres-Lagares D, et al.: Association between condylar asymmetry and temporo-mandibular disorders using 3D-CT. Med Oral Patol Oral Cir Bucal. 2012; 17(5): e852–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin H, Zhu P, Lin Y, et al.: Mandibular asymmetry: a three-dimensional quantification of bilateral condyles. Head Face Med. 2013; 9: 42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLemos AD, Katz CR, Heimer MV, et al.: Mandibular asymmetry: A proposal of radiographic analysis with public domain software. Dental Press J Orthod. 2014; 19(3): 52–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHalicioglu K, Celikoglu M, Buyuk SK, et al.: Effects of early unilateral mandibular first molar extraction on condylar and ramal vertical asymmetry. Eur J Dent. 2014; 8(2): 178–183. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoreno Uribe LM, Miller SF: Genetics of the dentofacial variation in human malocclusion. Orthod Craniofac Res. 2015; 18 Suppl 1: 91–99. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "34688",
"date": "11 Jun 2018",
"name": "Ida Bagus Narmada",
"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\nOverall your paper is very good and well done, but here are some minor suggestions:\nCould you please state more about the Kjellberg’s technique, it is not well described in the methods section of your paper? Could you please inform about the study design (descriptive or analytic study)? Could you please inform about the sampling method and how to determine the sample size? You only mention the inclusion criteria, what about the exclusion criteria? Please mention it. Please mention and inform who examined the patient? How to examine the patient? Please explain. How about the competency of the examiner? How many people? How to calibrate each examiner’s perception. Please mention also who is in charge for radiographic image interpretation?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3758",
"date": "16 Jul 2018",
"name": "Ervina Sofyanti",
"role": "Author Response",
"response": "Dear Dr. Narmada, Thank you very much for your kind assistance in reviewing our manuscript. According to your some minor suggestions, we have revised within material and method section.This is descriptive analytic study with cross-sectional approach and selection of sample with simple random sampling. In radiography analysis, we did inter-rater and intra-rater examiner.TB : the consultant in dentomaxillofacial radiograph. ES did the radiographic image interpretation under supervised of TB. We traced the radiography manually according Kjellberg's technique as followed : CH is defined as the distance from CO (the highest point of the condylar head) to the mandibular notch (the deepest point between the coronoid process and the condylar process). RH is the distance from CO to the GO’ (the reflection of subdivision tangen from ramus and corpus mandibular to the ramus borderline). In order to obtain vertical mandibular symmetry based on the ratio of condylar and ramus height (Kjellberg’s technique)."
}
]
},
{
"id": "34685",
"date": "12 Jun 2018",
"name": "Farhad B. Naini",
"expertise": [
"Reviewer Expertise FN: Facial aesthetic analysis",
"orthognathic and craniofacial surgery",
"complex orthodontics. AM: TMJ surgery",
"including TMJ replacements",
"orthognathic surgery",
"surgery for obstructive sleep apnoea",
"dynamic and static facial reanimation",
"full range of surgery for facial deformity and facial aesthetic surgery"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 Sofyanti et al. regarding the potential correlation between symptoms of temporomandibular joint dysfunction/disorder (TMD) and vertical mandibular symmetry is timely. 106 young adults referred for orthodontic treatment were questioned formally about TMD symptoms, and their responses correlated with objective evaluation of their orthopantomographs (OPTs). Of the 63 patients confirmed as having some degree of vertical mandibular asymmetry, and presumably requiring orthodontic treatment, 39 presented with symptoms of TMD, and 24 did not have any symptoms of TMD. The conclusion of the study was that there does not appear to be a significant association between vertical mandibular asymmetry and TMD symptoms.\nThe conclusion of this investigation is not surprising. All else being equal, mandibular asymmetries without functional problems would logically not lead to any greater likelihood of TMD symptoms than the rest of the population. Though TMD may be multifactorial in origin, psychosocial stress appears to be a common thread. Patients with excellent dental occlusions may develop TMD, particularly during stressful periods in their lives, whereas many patients with quite complex malocclusions may never develop TMD. Nevertheless, malocclusions linked to parafunctional activities or leading to functional occlusal problems may demonstrate a higher preponderance to TMD. As such, it would be very interesting to repeat this study looking at the effects of mandibular lateral displacements (functional shifts) and any potential correlations with TMD. True skeletal asymmetries without lateral mandibular displacements are less likely to lead to TMD, but lateral displacements and associated functional problems may be of greater concern. Additionally, It would be potentially interesting to evaluate the prevalence of TMD comparing patients with predominantly vertical mandibular asymmetries (hemimandibular hyperplasia) and those with predominantly horizontal mandibular asymmetries (hemimandibular elongation).\nAnother interesting question is the correlation of the variety of TMD symptoms in relation to specific malocclusions, whether dental or skeletal. Different TMD symptoms point to different aetiological factors. For example, with TM joint noises, clicking tends to be a sign of disc displacement/internal derangement, whereas crepitus may be due to arthritic-type changes. In the same way, myogenic pain, or trismus, may have diverse aetiology. It is worth exploring the potential links of dentoskeletal malocclusions not only to TMD as an all-encompassing diagnosis, but to the different symptoms comprising the diagnosis of TMD.\nIt should also be noted that there is improved clinical awareness, and enhanced imaging availability, with t1 or t2-weighted magnetic resonance imaging (MRI) used to establish more accurately the aetiology and nature of a dysfunctioning joint. Therefore, in cases where asymmetry directly results from previous trauma, or with developmental, congenital, and hypermobility disorders, it is disc position that seems to be the determining factor in the overall stability of the mechanics of the joint. Disruption, or disc displacement, leads to various degrees of arthrogenic and neurogenic symptoms of TMD, as opposed to myogenic TMD that seems to be a common symptom amongst the anxious and stressed TMD sufferer.\nThe authors should be congratulated on a thought-provoking article.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3757",
"date": "16 Jul 2018",
"name": "Ervina Sofyanti",
"role": "Author Response",
"response": "Dear Dr Farhad and Dr. Messiha,Thank you very much for your kind assistance in reviewing our manuscript and providing us with valuable advices for the manuscript itself as well as for our future work. To respond with the reviews that have been provided, please allow us to comment as follows: 1. “It would be very interesting to repeat this study looking at the effects of mandibular lateral displacements (functional shifts) and any potential correlations with TMD. True skeletal asymmetries without lateral mandibular displacements are less likely to lead to TMD, but lateral displacements and associated functional problems may be of greater concern”.Thank you very much for this valuable insight. The authors have noted the concern of the presence of lateral displacement to the occurrence of TMD for our future work. This was not included in this project as this was just an initial pilot project.2. \"It would be potentially interesting to evaluate the prevalence of TMD comparing patients with predominantly vertical mandibular asymmetries (hemimandibular hyperplasia) and those with predominantly horizontal mandibular asymmetries (hemimandibular elongation)\".We thank you again for this valuable insight and definitely will be considered in our future work.3. \"Therefore, in cases where asymmetry directly results from previous trauma, or with developmental, congenital, and hypermobility disorders, it is disc position that seems to be the determining factor in the overall stability of the mechanics of the joint\".In this study, the possible influence of any history of traumatic facial injury as well as congenital disease have been eliminated. And with regards to the position of the disc, this was not considered in this study but will surely be considered for our future work. The possible use of CBCT may perhaps assist our future work in determining disc position."
}
]
},
{
"id": "34686",
"date": "18 Jun 2018",
"name": "Ashok Karad",
"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\nTemporomandibular disorders (TMD) in many patients constitute one of the most frequent causes of non-dental pain in the orofacial region. It is universally accepted that the etiology of TMD is multifactorial, including the structural factors such as mandibular asymmetry. However, a mild degree of asymmetry in the craniofacial region is common in humans, including in individuals with a normal facial appearance. These patients often present with complain of impaired facial appearance and occlusal malfunction, which underlines the importance of this subject. The literature shows that the Temporomandibular joint internal derangement is also associated with mandibular asymmetry. Nevertheless, how skeletal discrepancies in mandibular asymmetry relate to TMJ disorders is still debatable.\n\nThis article by Ervina Sofyanti et al is a great step towards evaluating the association between TMD symptoms and vertical mandibular asymmetry in 106 young adult orthodontic patients in North Sumatra, Indonesia. The results of this study show that there is no significant correlation between vertical mandibular asymmetry and TMD symptoms.\n\nFor better clarity and understanding, especially for the readers who are new to this subject, it would be appropriate if the authors could provide more details on Kjellberg's technique in the 'Methods' section. To enhance it further, it should include the exclusion criteria as well.\nModern dental practice is focused on individualized diagnosis and treatment planning. Considering extreme variations in morphological and functional characteristics of an individual patient, it would be interesting to design a study involving these factors to fine-tune the results. This is because presence or absence of functional disturbances determines the discrepancy at the joint level in sagittal, vertical and transverse planes. The authors have done well in this interesting research work.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3759",
"date": "25 Jun 2018",
"name": "Ervina Sofyanti",
"role": "Author Response",
"response": "Dear Dr. Karad, Thank you very much for your kind assistance in reviewing our manuscript and valuable insight for our future work. As Dr. Narmada's comment also about the Kjellberg's technique, we will revise and update the material and method section."
}
]
}
] | 1
|
https://f1000research.com/articles/7-697
|
https://f1000research.com/articles/7-1085/v1
|
16 Jul 18
|
{
"type": "Case Report",
"title": "Case Report: Cutaneous granular cell tumors",
"authors": [
"Fatemeh Montazer",
"Armaghan Kazeminejad",
"Ghasem Rahmatpour Rokni",
"Sepideh Tayebi",
"Fatemeh Montazer",
"Ghasem Rahmatpour Rokni",
"Sepideh Tayebi"
],
"abstract": "Granular cell tumors are uncommon tumors in the deep soft tissue of the extremities, especially those with intramuscular origin, with a good prognosis after surgical resection. We present a case study of a 30 year old man with a skin lesion on his shoulder, which was grown in size over the course of 2 months. Complete tumor excision was done and histopathological findings revealed a marked hyperplasia epidermis with pseudoepitheliomatous pattern. The pathologic report was compatible with a granular cell tumor.",
"keywords": [
"Granular cell tumor",
"skin",
"atypical type"
],
"content": "Introduction\n\nGranular cell tumor (GrCT) is a benign tumor of the nerve sheath1, which more commonly occurs in the tongue, breast, skin and subcutis. It can affect the dermis, subcutis or submucosa. Granular cell tumors are uncommon tumors and most of them have a good prognosis after surgical resection, however, around 0.5–2% of these tumors may be malignant, which have a poor prognosis due to local recurrence and distant metastasis1–3.\n\nGranular cell tumors are rare in the deep soft tissues of the extremities, especially those of intramuscular origin2–5. Although this type of tumor is more common in the 4th to 6th decades of life, one study found that GrCT occurs more commonly between 30–40 years of age1. In this case report, we present a 30 year old man with a skin lesion, which was diagnosed as a granular cell tumor.\n\n\nCase report\n\nA 30 year old man was referred to the pathology department of the Imam Hospital, Sari, Iran in August 2017, and presented with a skin lesion on his right shoulder (Figure 1). He had no pain or trauma, and no significant past medical history. The patient had the skin lesion for 6 months previous to presentation, which had grown slowly in size over the course of previous 2 months. The lesion size was approximately 2 cm in diameter with a verrucous appearance. On physical examination, a hard, fixed and non-tender skin mass was palpable on his right shoulder.\n\nComplete tumor excision was performed and histopathological findings revealed a marked hyperplasia epidermis with pseudoepitheliomatous pattern (Figure 2A). The dermis showed ill-defined and diffuse proliferation of large round to oval cells, with brightly eosinophilic granular cytoplasm (Figure 2B). Mitotic activity was rare. Atypia and necrosis was not seen. The pathologic report was compatible with a granular cell tumor.\n\nA - Hematoxylin and eosin (H & E) staining of tissue sample under ×40 magnification showing pseudoepitheliomatous hyperplasia of epidermis. B - H & E staining of tissue sample under ×400 magnification showing large granular cell.\n\nImmunohistochemical (IHC) staining was carried out for this patient in the pathology department of Imam Khomeini hospital. All samples were fixed in 10% buffered formalin and embedded in paraffin. Sections were cut 4 μM thick from wax blocks, mounted on to 3-Aminopropyltriethoxysilane (APES)-coated glass slides. Slides were deparaffinized in xylene twice for 10 minutes, rehydrated through graded ethanol to distilled water before incubation for 15 minutes with 3% hydrogen peroxidase-methanol to inhibit endogenous peroxidase activity, and heated in 0.01 M citrate buffer (pH 6.0) in a microwave oven for 5 minutes at 100°C; after boiling for antigen retrieval. Then the slides were taken out of microwave oven and cooled to room temperature for 30 minutes. After incubating for 15 minutes in a blocking solution containing 10% normal goat serum in PBS, sections were incubated at 4°C overnight in a humidified chamber with CD68, S100, neuron specific enolase (NSE), vimentin, Ki67, desmin and SMA antibody6. The prepared stained slides were read using Olympus CX31 microscope.\n\nPeriodic acid–Schiff stain (PAS) was positive in the suspected tumor cells, and IHC results showed, CD68 (Manufacturer No. Mob167, species: mouse, clone ID No: kp1, concentration: 1:100; CellPath Ltd, UK), S100 (Manufacturer no. Z 0311, species: rabbit, clone ID No: polyclonal, concentration: 1:500; Agilent, Santa Clara, CA, USA), NSE (Manufacturer no. RP 054, species: rabbit, concentration: 1:50; CellPath, UK), and vimentin (Manufacturer no. Mob 090, species: mouse, clone ID No: v9, concentration: 1:50; CellPath, UK), were strongly positive (dark brown staining +3) in suspected cells (Figure 3-A, B, C) and Ki67 (Manufacturer No. DB D-125, clone: C16-I, species: rabbit, concentration: 1:200; DB Biotech, Kosice, Slovakia) was positive in 3% of tumor cells. Desmin and SMA were negative (Figure 4- A, B). Finally, the granular cell tumor was confirmed and the patient has been followed for 1 year after the surgery and no recurrence has been reported.\n\nImmunohistochemistry study positive staining of tissue samples for NSE (A), CD68 (B), S100 (C).\n\nImmunohistochemistry negative staining of tissue samples for SMA (A) and Desmin (B).\n\n\nDiscussion\n\nGrCT was first described in 1926 as a myoblastoma which arises from the muscle in the tongue. Apart from the tongue, the skin and soft tissues are other common locations for GrCTs1. In 1935, Feyrter described the tumor as a granular cell neuroma because he hypothesized that the tumors were neural in origin. Fust and Custer named the tumor as granular cell neurofibroma in 1948. Finally in 1962, Fisher and Wechsler named the tumors as granular cell schwannomas, because Schwann cells was their most probable origin. Nowadays the name adopted by WHO is granular cell tumor7. GrCT usually presents as a solitary and small nodule, as a painless mass. It is most common in women aged 30–60 years old1,8. The presented case had a painless mass and as a 30 year old male, he did not conform to epidemiological evidence on the most common sex9. Furthermore, based on the clinical findings, the dermatologist diagnosed this lesion as dermatofibroma and keratoacanthoma and no differential diagnosis had been reported. However, the pathology report showed different results and identified it as granular cell tumor.\n\nTo our knowledge, three similar cases of GrCT with distinguished dermatofibroma-like morphology have been described in the literature. However, all these cases were presented as atypical GrCT. One was a 60 year old woman with a nodule on the back10, the second was a 48 year old man with a lesion in the pubic area11 and the third case was a 62 year old woman with a tumor on her back under the right scapula12.\n\nAccording to 6 histological criteria, GrCT can be classified as benign, atypical or malignant. The criteria are necrosis, spindling, vesicular nuclei with large nucleoli, increased mitotic activity (>2 mitoses/10 high-power fields), high nuclear to cytoplasmic ratio, and pleomorphism1. Tumors with 1 or 2 of these criteria can be classified as atypical. Another classification system states the only difference between benign GrCT and GrCT-uncertain malignant potential is the presence of necrosis and/or mitoses13. As to the former classification system, there are some cases of histological mild atypic GrCTs, which presented a malignant clinical course such as local recurrence, rapid recent growth, and large tumor diameter14.\n\nGrCTs are relatively uncommon and benign in most of the cases1,15. Malignant and atypical GrCTs account for only a small percentage of cases1. The most common immunological marker presented by GrCT is S100 protein. In our case, we excluded possible diagnosis of granular cell dermatofibroma (S100-protein negative) and malignant peripheral neural sheath tumor (weak S-100 expression). CD68, CD57, and NSE may be positive in GrCT cases1.We can evaluate malignant potential in GrCTs by the means of Ki-67 proliferation index. If the index is greater than 10% in a specific case, the malignant potential is higher in that case, although not all malignant GrCTs have a high Ki-67 index. In addition to immunological markers mentioned above, it is shown that 68% of GrCTs express p53 in over 50% of tumor cell nuclei1.\n\nIn conclusion, we presented a case of GrCT with the dermatofibroma-like morphology fulfilling criteria of benign GrCT and immunhistochemical positivity of S100, CD68, and NSE. The necessity for S-100 staining to differentiate granular cell tumor with dermatofibroma from dermatofibroma-like GrCT is highly recommended.\n\n\nInformed consent\n\nWritten informed consent for the publication of the patient’s clinical details and images was obtained from the patient.\n\n\nData availability\n\nDataset 1: Raw microscope images 10.5256/f1000research.13015.d20701016",
"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\nFanburg-Smith JC, Meis-Kindblom JM, Fante R, et al.: Malignant granular cell tumor of soft tissue: diagnostic criteria and clinicopathologic correlation. Am J Surg Pathol. 1998; 22(7): 779–94. PubMed Abstract | Publisher Full Text\n\nElkousy H, Harrelson J, Dodd L, et al.: Granular cell tumors of the extremities. Clin Orthop Relat Res. 2000; (380): 191–8. PubMed Abstract | Publisher Full Text\n\nEnginger FM, Weiss SW: Granular cell tumor. Soft Tissue Tumors. 4 ed: St Louis, MO: CV Mosby. 2004; 1178–87.\n\nBlacksin MF, White LM, Hameed M, et al.: Granular cell tumor of the extremity: magnetic resonance imaging characteristics with pathologic correlation. Skeletal Radiol. 2005; 34(10): 625–31. PubMed Abstract | Publisher Full Text\n\nThacker MM, Humble SD, Mounasamy V, et al.: Case report. Granular cell tumors of extremities: comparison of benign and malignant variants. Clin Orthop Relat Res. 2007; 455: 267–73. PubMed Abstract | Publisher Full Text\n\nDabbs DJ: Diagnostic Immunohistochemistry. Theranostic and Genomic Applications, Expert Consult: Online and Print. 4th, editor: Saunders. 2014; 960. Reference Source\n\nFletcher CDM, Bridge JA, Hogendoorn P, et al.: WHO Classification of Tumours of Soft Tissue and Bone. 4 ed: Lyon, France: IARC. 2013; 5. : 178. Reference Source\n\nTsuchida T, Okada K, Itoi E, et al.: Intramuscular malignant granular cell tumor. Skeletal Radiol. 1997; 26(2): 116–21. PubMed Abstract | Publisher Full Text\n\nCalonje JE, Brenn T, Lazar A, et al.: McKee's Pathology of the Skin. 4th, editor: Saunders; 2012; 1906. Reference Source\n\nCheng SD, Usmani AS, DeYoung BR, et al.: Dermatofibroma-like granular cell tumor. J Cutan Pathol. 2001; 28(1): 49–52. PubMed Abstract | Publisher Full Text\n\nHong SB, Yang MH, Lee MH, et al.: Dermatofibroma-like atypical granular cell tumour. Acta Derm Venereol. 2005; 85(2): 179–80. PubMed Abstract | Publisher Full Text\n\nSoukup J, Hadzi-Nikolov D, Ryska A: Dermatofibroma-like granular cell tumour: a potential diagnostic pitfall. Pol J Pathol. 2016; 67(3): 291–4. PubMed Abstract | Publisher Full Text\n\nNasser H, Ahmed Y, Szpunar SM, et al.: Malignant granular cell tumor: a look into the diagnostic criteria. Pathol Res Pract. 2011; 207(3): 164–8. PubMed Abstract | Publisher Full Text\n\nEnginger FM, Weiss SW: Soft tissue tumors. 2 ed: Louis: Mosby; 1988; 757–67.\n\nWeiss SW, Goldblum JR, Folpe AL: Enzinger and Weiss’s Soft Tissue Tumors. 5 ed: New York, NY: Mosby Elsevier; 2013; 878–80.\n\nMontazer F, Kazeminezhad A, Rokni GR, et al.: Dataset 1 in: Case Report: Cutaneous granular cell tumors. F1000Research. 2018. Data Source"
}
|
[
{
"id": "38473",
"date": "01 Oct 2018",
"name": "Nikoo Mozafari",
"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\nA case of granular cell tumor, with usual and typical presentation has been described. The manuscript needs English editing.\nIntroduction: The importance or novelty of this case has not been clarified.\nCase report: Why was patient referred to pathology department? Explaining method of IHC, is unnecessary. Figure 4a and figure 4b are identical, but authors have mentioned that these are 2 different slides stained with 2 markers. The quality of photos are not acceptable, photos with higher resolution are required.\n\nDiscussion: More explanation about “granular cell dermatofibroma” and methods of its differentiation from “granular cell tumor” seems necessary. A lot of IHC markers (such as CD68, CD57, and NSE) have been used but no data on their utilities have been given. How can these markers be helpful in differentiating “ granular cell tumor” from other entities? This term seems to be wrong: “granular cell tumor with dermatofibroma”\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
},
{
"id": "47150",
"date": "30 May 2019",
"name": "Jeffrey M Farma",
"expertise": [
"Reviewer Expertise Surgical oncology",
"DFSP",
"granular cell tumors",
"melanoma",
"sarcoma",
"squamous cell carcinoma",
"merkel cell carcinoma",
"colon cancer",
"rectal cancer",
"gastric cancer",
"neuroendocrine tumors."
],
"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 on Cutaneous granular cell tumors.\n\nThey should remove that they occur on the extremities as their case is actually on the truncal region. Why did they include how the slides were processed? This is useless information. The only novel part of this is the dermatofibroma-like morphology but overall I do not feel this case report adds much to the literature. What margins did they take on the excision? How long was the followup on the patient?\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1085
|
https://f1000research.com/articles/7-1054/v1
|
11 Jul 18
|
{
"type": "Opinion Article",
"title": "Time to really share real-world data?",
"authors": [
"Sophie Graham",
"Laura McDonald",
"Radek Wasiak",
"Michael Lees",
"Sreeram Ramagopalan",
"Laura McDonald",
"Radek Wasiak",
"Michael Lees",
"Sreeram Ramagopalan"
],
"abstract": "Data other than that from clinical trials are important for healthcare decision making. However, unlike the vocal calls seen for more open access to trial data, there are limited efforts being made to ensure that agencies that collect real-world data (RWD) share this, despite its importance. There are many RWD sources across the world that could be readily exploited for research once shared. There are policy and privacy questions that need to be tackled, but opening up and sharing RWD offers remarkable potential for improvements in care for individuals and more effective use of limited healthcare resources. Open science should become the standard for RWD as well as clinical trials, especially those that have a high likelihood to influence practice.",
"keywords": [
"real-world data",
"sharing",
"clinical data"
],
"content": "Introduction: Real-world data and its importance\n\nReal-world data (RWD) are data collected in the course of routine health care delivery or otherwise generated without constraints, as in the ‘real-world’1,2. These data are used to understand disease epidemiology, patterns of care and patient need as well as provide valuable insight into treatment effectiveness and safety in day to day clinical practice.\n\nDue to their potential to provide insight into questions not effectively addressed from clinical trials, RWD are increasingly influencing health care decision making, including regulatory assessment, clinical practice and policy. For example, the 21st Century Cures Act in the United States has required the Food and Drugs Administration (FDA) to draw up guidelines for the role of RWD in drug approvals (https://healthpolicy.duke.edu/sites/default/files/atoms/files/rwe_white_paper_2017.09.06.pdf), and in the United Kingdom (UK) the Academy of Medical Sciences and the Association of the British Pharmaceutical Industry have recently prioritised supporting the inclusion of RWD in regulatory and health technology assessment (HTA) processes (https://acmedsci.ac.uk/more/news/next-steps-for-using-real-world-evidence). These data are also influencing policy; for example, in the UK the controversial decision to increase hospital staffing levels on the weekend, was driven by analysis from inpatient National Health Service (NHS) hospital data from 2009–10 showing a significantly higher risk of mortality in the 30-day follow up period from patients admitted on the weekend, compared to those admitted midweek3.\n\n\nWhy share data?\n\nThe benefits of sharing data are evidenced with the increasing transparency of clinical trial data. Access to clinical trial data has allowed the validation of initially reported trial results4, but perhaps more importantly openly available data has also been used to answer new questions about disease. For example, the Dialogue for Reverse Engineering Assessment and Methods (DREAM) challenge aimed at identifying a prognostic model for overall survival in patients with metastatic prostate cancer using the raw data from the comparator arms of four clinical trials led to the identification of aspartate aminotransferase as an important prognostic factor for overall survival5. Similarly, the SPRINT competition challenged entrees to re-analyse data from a randomized trial of blood-pressure control and this led to the development of a new decision-making tool for clinicians to determine whether patients should receive intensive hypertensive treatment or not based on patient characteristics6\n\nIn genetics, the sharing of data have been a major enabling factor in the identification of disease associated genes. Genome-wide association studies often require data from tens of thousands of patients to have the statistical power to detect variants implicated in disease7. Initiatives such as the UK Biobank8 have enabled richly genotyped and phenotyped data on many thousands of patients to be available to the international scientific community. In doing so, these data have been used to better understand the genetic architecture underlying a number of complex diseases including Alzheimer’s, major depression and atrial fibrillation, among many others8–10. Given the often prohibitively high time and cost burden associated with establishing a cohort needed to unpick the genetic underpinnings of disease, the model employed by the UK Biobank exemplifies the benefits of data sharing not only by enabling reproducibility, but critically in the acceleration and advancement of scientific research.\n\nA number of European clinical databases provide access to RWD. For example, in the UK, the Clinical Practice Research Datalink (CPRD), an electronic medical record database that covers general practitioner encounters, has been available to researchers and has led to the generation of hundreds of peer-reviewed publications, with notable contributions including validation of the safety of the measles, mumps and rubella vaccine1. Further, CPRD data are being increasingly used in English national clinical guidelines and guidances11. The Nordic countries also have a long tradition of collecting patient medical information in the form of national population-based and prescription registries. The high quality of the data recorded and the good coverage of the sources has contributed to a vast number of pharmacoepidemiological insights12,13, for example in realising risks associated with serotonin reuptake inhibitors in pregnancy14.\n\nCombining accessible RWD sets is of particular importance in Post-Authorisation Safety Studies (PASS), which are studies carried out after a medicine has been approved for use to obtain further information on its safety in the real-world. In such studies, large patient numbers are required to increase power to detect true rare adverse outcomes associated with a particular treatment. To meet the FDA’s Post Marketing Requirements for a post-marketing safety surveillance system, the Mini-Sentinel Initiative was launched in 2009, and was eventually expanded to the full Sentinel Initiative in February 2016. The initiative includes electronic data from different healthcare data holders that is automatically collected on an ongoing basis and merged into a common data format, which the FDA can query at any point to quickly and securely monitor drug safety issues. It now includes over 300 million person years of unduplicated data from over 17 different data partnerships. The Mini-Sentinel Initiative (pilot project) identified a 1.5 (95% CI, 0.2–3.2) increased risk of intussusception per 100,000 infants administered the rotavirus vaccination15, highlighting the power of sharing data to identify adverse events that occur at very low rates.\n\n\nLack of data sharing has consequences\n\nThere is a considerable lack of data sharing in observational research. This was shown in a recent review of 237 observational studies published in the BMJ from 1st January 2015 to 31st August 2017. This work found that 63% of studies reviewed during this period did not share the raw data upon which the analyses were conducted16. There are also important examples of large medical record data databases that do not readily share data with external researchers. These include data from the Spanish regional general practitioner database Base De Datos Para La Investigacion Farmacoepedemiologica en Atencion Primaria (BIFAP) and the Secure Anonymised Information Linkage (SAIL) databank, which includes primary and secondary care data from the Welsh population.\n\nLack of data sharing is potentially not without consequence. For example, the Infections in Oxfordshire Research Database (IORD) showed that any apparent aforementioned UK ‘weekend effect’ on mortality arises from patient-level differences at admission3,17,18 Data from IORD has been available for many years and if this was openly accessible for researchers to analyse the negative impacts of policy changes could perhaps have been prevented. Rosiglitazone, a glitazone used in the treatment of type II diabetes, was initially approved by the European Medicines Agency in 1999, despite there being limited evidence to support during the approval process. In July 2010, the FDA convened an expert panel to discuss removing rosiglitazone from the market because of arising evidence that rosiglitazone was associated with increasing risk of myocardial infarction. The original evidence was based on the RECORD trial, which was an unblinded, open label trial19. It then took two additional clinical trials and over 10 years of post-marketing surveillance to detect an increased risk in myocardial infarction20. Had a real-world PASS been carried out immediately after market approval utilising a number of different data sources along the lines of the aforementioned Sentinel initiative, unwarranted clinical outcomes could have been robustly detected much earlier.\n\n\nBarriers to data sharing; combating privacy concerns\n\nA fundamental barrier associated with the sharing of RWD are the concerns over patient confidentiality. This was illustrated when the UK governmental Care.data programme, which aimed to create a national combined primary and secondary care database, was halted over general public concerns for patient privacy (https://blogs.biomedcentral.com/on-medicine/2016/08/19/need-care-closure-care-data/). Stricter regulatory frameworks introduced as part of the General Data Protection Regulation (GDPR) within the EU aim to tackle many privacy issues involved in the storage, processing and management of data in a digital era. New regulations will not only tighten organisational laws around the handling of data, but will also work to increase the rights of individuals, giving them more control over their personal information including the right to transparency, access and erasure. Importantly, despite privacy concerns, studies demonstrate that patients are generally willing to share their healthcare data in the context of contributing to public health as long as the potential benefits are appropriately communicated20,21\n\n\nConclusion\n\nEven though there are significant potential benefits that improved access to RWD could provide patients, the progress to increase data sharing is slow. There are policy and privacy questions that need to be tackled22, but opening up and sharing healthcare data offers remarkable potential for improvements in care for individuals as well as potential for more effective use of limited healthcare resources. In some instances, for example with the National Quality Registries in Sweden, this potential is already being realised. It is hoped that sharing of RWD becomes standard practice, especially in those data-sets that have a high likelihood of influencing public health and clinical practice.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Competing interests\n\n\n\nSG and RW are employees of Evidera. LM, ML and SR are employees of Bristol-Myers Squibb.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMcDonald L, Lambrelli D, Wasiak R, et al.: Real-world data in the United Kingdom: opportunities and challenges. BMC Med. 2016; 14(1): 97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJarow JP, LaVange L, Woodcock J: Multidimensional Evidence Generation and FDA Regulatory Decision Making: Defining and Using \"Real-World\" Data. JAMA. 2017; 318(8): 703–704. PubMed Abstract | Publisher Full Text\n\nFreemantle N, Richardson M, Wood J, et al.: Weekend hospitalization and additional risk of death: An analysis of inpatient data. J R Soc Med. 2012; 105(2): 74–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEbrahim S, Sohani ZN, Montoya L, et al.: Reanalyses of Randomized Clinical Trial Data. JAMA. 2014; 312(10): 1024–1032. PubMed Abstract | Publisher Full Text\n\nGuinney J, Wang T, Laajala TD, et al.: Prediction of overall survival for patients with metastatic castration-resistant prostate cancer: development of a prognostic model through a crowdsourced challenge with open clinical trial data. Lancet Oncol. 2017; 18(1): 132–142. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLedford H: Open-data contest unearths scientific gems — and controversy. Nat News. 2017; 543(7645): 299. PubMed Abstract | Publisher Full Text\n\nVisscher PM, Wray NR, Zhang Q, et al.: 10 Years of GWAS Discovery: Biology, Function, and Translation. Am J Hum Genet. 2017; 101(1): 5–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWray NR, Ripke S, Mattheisen M, et al.: Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018; 50(5): 668–681. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarioni RE, Harris SE, Zhang Q, et al.: GWAS on family history of Alzheimer’s disease. Transl. Psychiatry 2018; 8(1): 99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNielsen JB, Fritsche LG, Zhou W, et al.: Genome-wide Study of Atrial Fibrillation Identifies Seven Risk Loci and Highlights Biological Pathways and Regulatory Elements Involved in Cardiac Development. Am J Hum Genet. 2018; 102(1): 103–115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOyinlola JO, Campbell J, Kousoulis AA: Is real world evidence influencing practice? A systematic review of CPRD research in NICE guidances. BMC Health Serv Res. 2016; 16: 299. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReigstad MM, Larsen IK, Myklebust TÅ, et al.: The Nordic Health Registries: an important part of modern medical research. Hum Reprod. 2016; 31(1): 216–217. PubMed Abstract | Publisher Full Text\n\nFuru K, Wettermark B, Andersen M, et al.: The Nordic countries as a cohort for pharmacoepidemiological research. Basic Clin Pharmacol Toxicol. 2010; 106(2): 86–94. PubMed Abstract | Publisher Full Text\n\nKieler H, Artama M, Engeland A, et al.: Selective serotonin reuptake inhibitors during pregnancy and risk of persistent pulmonary hypertension in the newborn: population based cohort study from the five Nordic countries. BMJ. 2012; 344: d8012. PubMed Abstract | Publisher Full Text\n\nYih WK, Lieu TA, Kulldorff M, et al.: Intussusception risk after rotavirus vaccination in U.S. infants. N Engl J Med. 2014; 370(6): 503–512. PubMed Abstract | Publisher Full Text\n\nMcDonald L, Schultze A, Simpson A, et al.: Lack of data sharing in observational studies. BMJ. 2017; 359: j4866. PubMed Abstract | Publisher Full Text\n\nWalker AS, Mason A, Quan TP, et al.: Mortality risks associated with emergency admissions during weekends and public holidays: an analysis of electronic health records. Lancet. 2017; 390(10089): 62–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcDonald L, Ramagopalan SV, Lees M: Real-world data really matter. CMAJ. 2017; 189(41): E1293. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHome PD, Pocock SJ, Beck-Nielsen H, et al.: Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. The Lancet. 2009; 373(9681): 2125–2135. PubMed Abstract | Publisher Full Text\n\nMello MM, Lieou V, Goodman SN: Clinical Trial Participants’ Views of the Risks and Benefits of Data Sharing. N Engl J Med. 2018; 378(23): 2202–2211. PubMed Abstract | Publisher Full Text\n\nWeitzman ER, Kelemen S, Kaci L, et al.: Willingness to share personal health record data for care improvement and public health: a survey of experienced personal health record users. BMC Med Inform Decis Mak. 2012; 12: 39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarren E: Strengthening Research through Data Sharing. N Engl J Med. 2016; 375(5): 401–403. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "35957",
"date": "18 Jul 2018",
"name": "Sajan Khosla",
"expertise": [
"Reviewer Expertise Real World Evidence",
"Health Informatics",
"Data science"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 would suggest adding commentary on the reproducibility of RWD studies by better sharing of code and data. Those practitioners of RWD are all aware of the intricate nuances that lead to variable definitions and sharing of these may enable better reproducibility. Perhaps the better use of technology to better develop transparency and reproducibility would be beneficial in these areas, leading the clinical trial setting1.\nThe importance of RWD is a combination of the potential regulatory use within the context of the 21st Century Cures act, being able to reinforce clinical decision making with information at the patient bedside and to also support policy making decisions. Having been part of the team on the weekend mortality paper, this analysis was less influential but more supportive to larger initiatives at hand. I would also support that with the outputs from international disease registries in oncology which highlight how poor/late presentation of cancer within the UK system leads to the significant gap in cancer outcomes and led to better public health initiatives around recognising symptoms of cancer2.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "36415",
"date": "26 Jul 2018",
"name": "David Madigan",
"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 in interesting short opinion piece about RWD. I worry that it exhibits the kind of unbridled enthusiasm that I fear ultimately undermines efforts to appropriately incorporate RWD into healthcare decision making. There are real limitations to RWD studies and the authors should acknowledge these. See, for example, many of Ioannidis’s papers or Madigan, D., et al. (2013)1.\n\nThe CPRD is not a particularly good example since the data are available to researchers under severe restrictions, e.g., “A dataset may only be used for the study for which it has been approved.”\n\nAs I understand it, Sentinel queries require multi-week/month turnaround times.\n\nThe notion that an observational study could have uncovered the rosiglitazone problem early is speculative.\n\nThe section on privacy concerns pays lip service to some really complex issues. These data are de-identified in superficial ways and identification of specific individuals is entirely possible. The implications for health insurance, employment, etc. are highly complex.\n\nThe paper should discuss common data models. Shared data with idiosyncratic data models are of somewhat limited value. In a similar vein, the article should discuss OHDSI.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "38761",
"date": "15 Oct 2018",
"name": "Laurie A. Tomlinson",
"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\nShould real-world data be more freely available? Of course it should, it is impossible to disagree. But, \"I think you'll find it's a bit more complicated than that\".\nUse of observational data for drug safety (among other things) is crucially important. However, many things are critical to make this happen. High quality research training with access to excellent statisticians. A research system that doesn't depend on impact factor and no benefit for getting an 'interesting' result. Quality peer-review. Lots of funding to do important work of this kind. An educated audience capable of critical analysis of complex study designs. Brexit not gutting the MHRA. Close links with the EMA. Affordable access for academics to access the data. And yes open data would help but who has the time to rerun the analyses of other groups who have proven themselves trustworthy when there are probably more important things to do? There are also risks. Anti-vaxxers on a mission to identify spurious drug-effects. And vitally loss of public trust. Yes there are surveys showing public support for access to their data but we have already made a mess of this with care.data and can we risk that again? Especially when pharma or profit-driven companies are promoting it? In a world ever more driven by conspiracy theories and paranoia this seems a big step at the moment.\nYes real-world data should be more open. But at the moment it is only one of many important battles to fight.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1054
|
https://f1000research.com/articles/7-1043/v1
|
10 Jul 18
|
{
"type": "Review",
"title": "GSK-3β, a double-edged sword in Nrf2 regulation: Implications for neurological dysfunction and disease",
"authors": [
"Megan Culbreth",
"Michael Aschner",
"Megan Culbreth"
],
"abstract": "In the past decade, it has become evident that glycogen synthase kinase 3β (GSK-3β) modulates the nuclear factor erythroid 2-related factor 2 (Nrf2) oxidative stress response. GSK-3β functions as an inhibitor, both directly in the activation and indirectly in the post-induction of Nrf2. The incidence of oxidative stress in neurological dysfunction and disease has made this signaling pathway an attractive therapeutic target. There is minimal evidence, however, to support a distinctive function for GSK-3β mediated Nrf2 inhibition in nervous system decline, apart from the typical oxidative stress response. In both Alzheimer’s disease and brain ischemia, this pathway has been explored for potential benefits on disease etiology and advancement. Presently, it is unclear whether GSK-3β mediated Nrf2 inhibition markedly influences these disease states. Furthermore, the potential that each has unique function in neurodegenerative decline is unsubstantiated.",
"keywords": [
"GSK-3β",
"Nrf2",
"oxidative stress"
],
"content": "Abbreviations\n\nAlzheimer’s disease (AD), β-amyloid (Aβ), glycogen synthase kinase 3β (GSK-3β), middle cerebral artery occlusion and reperfusion (MCAO/R), nuclear factor erythroid 2-related factor 2 (Nrf2).\n\n\nIntroduction\n\nGlycogen synthase kinase 3β (GSK-3β) is a complex signaling molecule, involved in a variety of cellular processes1. In the past decade, GSK-3β has been extensively investigated, particularly in regard to its role in neurological dysfunction and disease2,3. It is now evident GSK-3β participates in the cellular response to oxidative stress, a hallmark of several nervous system disorders. GSK-3β modulates this response through its interaction with nuclear factor erythroid 2-related factor 2 (Nrf2)4,5. Although the GSK-3β-Nrf2 association is well established, it has yet to be determined whether the dual regulatory nature of this interaction impacts disease advancement and potential therapies.\n\nGSK-3β phosphorylates Nrf2, which results in its nuclear exclusion6 and degradation4,7,8. Interestingly, this was demonstrated to occur in a kelch-like ECH-associated protein 1 (Keap1)-independent manner4; however, active GSK-3β was induced by hydrogen peroxide6. Thus GSK-3β regulated Nrf2 activity possibly is oxidant-sensitive as well. GSK-3β has also been observed to indirectly modulate Nrf2 via Fyn phosphorylation. Fyn translocates to the nucleus resultant to GSK-3β phosphorylation, and in turn phosphorylates Nrf2, which stimulates its nuclear export5. It must be noted that none of the above-mentioned reports were conducted in neuronal models. Moreover, observed effects were in transiently transfected cells, and not on endogenous protein.\n\nDespite limited evidence that GSK-3β regulates Nrf2 in a similar manner in the nervous system, this signaling pathway has emerged as an attractive therapeutic target for neurodegenerative diseases. Oxidative stress is a well-established feature of several cognitive disorders, including Alzheimer’s disease (AD) and brain ischemia. However, a possible mechanism by which neurological dysfunction and disease modify GSK-3β mediated Nrf2 activity is unknown. Furthermore, the potential that this signaling axis contributes to disease etiology and progression has not been extensively explored.\n\nHere, current research in AD and brain ischemia is highlighted with particular focus on potentially unique mechanisms by which these disorders modulate GSK-3β regulated Nrf2 activity.\n\n\nAlzheimer’s disease\n\nGSK-3β is implicated in the advancement and potentially the etiology of AD. Extensive evidence corroborates its function in the learning and memory deficits observed in AD patients. Mechanistically, GSK-3β is linked to tau hyperphosphorylation and β-amyloid (Aβ) deposition, significant markers of AD pathogenesis. Furthermore, it modifies the oxidative stress response resultant to AD progression3.\n\nRecent reports have started to build connections between GSK-3β and Nrf2 in AD pathology and propose prospective therapies. GSK-3β suppression in a mouse model of AD was found to increase nuclear Nrf2 and total glutathione-S transferase (GST), an Nrf2 transcriptional target, in cortex. Moreover, reduced oxidative stress biomarkers and tau phosphorylation were observed9. GSK-3β activation was also required for Aβ-induced axonal transport deficits detected in primary hippocampal neurons. Tau presence was necessary for this outcome, but intriguingly was Fyn-independent10. Fyn phosphorylates tau11 and Nrf2. The latter requires GSK-3β activation5; however, GSK-3β directly phosphorylates Nrf2 as well6. Irrespectively, Nrf2 is excluded from the nucleus as a result5,6. Potential associations between GSK-3β, Fyn, and Nrf2 were not demonstrated in this study10. Furthermore, the Fyn-independent nature of Aβ and tau in AD could underscore distinct mechanisms by which GSK-3β activation mediates disease advancement.\n\nA new AD therapy, which simultaneously inhibits GSK-3β and induces Nrf2 has been proposed. Interestingly, therapeutic Nrf2 activation was found to be GSK-3β inhibition-independent in the in vitro AD model. However, the potentially distinct beneficial properties of GSK-3β inhibition and Nrf2 induction were not examined. Tested compounds exhibited enhanced neuroprotection and reduced oxidative stress12. Similar to the reports mentioned above, this study did not connect GSK-3β inhibited Nrf2 activation to GSK-3β activated tau hyperphosphorylation and Aβ deposition.\n\nCurrent research on AD pathology has only just begun to elucidate a possible association between GSK-3β activation and the Nrf2-mediated oxidative stress response. Existing data does not, however, clearly demonstrate that GSK-3β activation represses Nrf2 function in AD models. Theoretically, observed oxidative stress in AD could occur entirely independent of GSK-3β. Moreover, other stress response factors apart from Nrf2 might be involved.\n\n\nIschemia\n\nGSK-3β has also emerged as a target in the treatment of brain ischemia. However, distinct from AD, recent studies have focused on GSK-3β regulation of Nrf2 in this disease state. Also unlike AD, an altered oxidative environment is the primary feature of brain ischemia, and not simply a secondary factor in the disease.\n\nExisting evidence indicates GSK-3β inhibition enhances Nrf2 activity in the rat cerebral cortex following middle cerebral artery occlusion and reperfusion (MCAO/R). Interestingly, however, this brain ischemia model did not exhibit increased active GSK-3β and decreased nuclear Nrf2 simply as a consequence of MCAO/R. Furthermore, Nrf2 bound at the antioxidant response element (ARE) and subsequent target gene expression was only enhanced following MCAO/R in the presence of GSK-3β inhibitors13. These data potentially imply GSK-3β regulation of Nrf2 is not directly involved in brain ischemia. Moreover, this study did not demonstrate improvement in the rat cerebral cortex following MCAO/R resultant to GSK-3β inhibition and Nrf2 activation.\n\nAnother report, however, did validate a GSK-3β inhibitor in an acute ischemia rat model, displaying reduced infarct volume, brain edema, and neurological deficit. These improvements were correlated with increased total Nrf2 protein and heme-oxygenase. Curiously, however, recovery was only examined 24 hours post-MCAO14, whereas the article mentioned above showed GSK-3β inhibition effects as early as 6 hours post-MCAO13. The timescale by which GSK-3β inhibitors exhibit therapeutic properties in brain ischemia must be further examined.\n\nPresent knowledge does not indicate brain ischemia directly alters GSK-3β regulation of Nrf2. Active GSK-3β is not enhanced in ischemia models, and Nrf2 downregulated as a consequence. Intriguingly, however, GSK-3β inhibitors do render improvement in these models, although evidence that Nrf2 is involved in this response is minimal. Possibly, GSK-3β inhibitors act on another signaling pathway in which GSK-3β is involved.\n\n\nConclusion\n\nThe current body of evidence available to support a primary role for GSK-3β regulation of Nrf2 in neurological dysfunction and disease is incomplete. There is little to substantiate that this signaling axis is modulated differently under nervous system distress relative to a typical endogenous oxidative stress response. Moreover, the potentially distinct functions of GSK-3β and Nrf2 in cognitive disorders have not been linked to their possible combinatorial roles.\n\nPolymorphisms in both the GSK-3β15 and Nrf216 gene promoter sequences have been associated with the onset of neurodegenerative diseases. There is no data, however, to validate that these polymorphisms impact the function of the GSK-3β and Nrf2 signaling pathway.\n\nFuture studies should focus on the distinctive functions for GSK-3β and Nrf2 in neurological dysfunction and disease. Based on these unique roles, potential links to their effects on each other should be formulated. Without these efforts, there is minimal evidence to substantiate a causative function for this signaling axis on cognitive disorder apart from an oxidative stress response.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nMA has been supported in part by the National Institutes of Health (grant numbers NIEHS R01ES07331, NIEHS R01ES10563 and NIEHS R01ES020852).\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\nGrimes CA, Jope RS: The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol. 2001; 65(4): 391–426. PubMed Abstract | Publisher Full Text\n\nGolpich M, Amini E, Hemmati F, et al.: Glycogen synthase kinase-3 beta (GSK-3β) signaling: Implications for Parkinson's disease. Pharmacol Res. 2015; 97: 16–26. PubMed Abstract | Publisher Full Text\n\nHooper C, Killick R, Lovestone S: The GSK3 hypothesis of Alzheimer's disease. J Neurochem. 2008; 104(6): 1433–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRada P, Rojo AI, Chowdhry S, et al.: SCF/{beta}-TrCP promotes glycogen synthase kinase 3-dependent degradation of the Nrf2 transcription factor in a Keap1-independent manner. Mol Cell Biol. 2011; 31(6): 1121–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain AK, Jaiswal AK: GSK-3beta acts upstream of Fyn kinase in regulation of nuclear export and degradation of NF-E2 related factor 2. J Biol Chem. 2007; 282(22): 16502–10. PubMed Abstract | Publisher Full Text\n\nSalazar M, Rojo AI, Velasco D, et al.: Glycogen synthase kinase-3beta inhibits the xenobiotic and antioxidant cell response by direct phosphorylation and nuclear exclusion of the transcription factor Nrf2. J Biol Chem. 2006; 281(21): 14841–51. PubMed Abstract | Publisher Full Text\n\nRada P, Rojo AI, Evrard-Todeschi N, et al.: Structural and functional characterization of Nrf2 degradation by the glycogen synthase kinase 3/β-TrCP axis. Mol Cell Biol. 2012; 32(17): 3486–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChowdhry S, Zhang Y, McMahon M, et al.: Nrf2 is controlled by two distinct β-TrCP recognition motifs in its Neh6 domain, one of which can be modulated by GSK-3 activity. Oncogene. 2013; 32(32): 3765–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarr SA, Ripley JL, Sultana R, et al.: Antisense oligonucleotide against GSK-3β in brain of SAMP8 mice improves learning and memory and decreases oxidative stress: Involvement of transcription factor Nrf2 and implications for Alzheimer disease. Free Radic Biol Med. 2014; 67: 387–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVossel KA, Xu JC, Fomenko V, et al.: Tau reduction prevents Aβ-induced axonal transport deficits by blocking activation of GSK3β. J Cell Biol. 2015; 209(3): 419–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee G, Thangavel R, Sharma VM, et al.: Phosphorylation of tau by fyn: implications for Alzheimer's disease. J Neurosci. 2004; 24(9): 2304–12. PubMed Abstract | Publisher Full Text\n\nGameiro I, Michalska P, Tenti G, et al.: Discovery of the first dual GSK3β inhibitor/Nrf2 inducer. A new multitarget therapeutic strategy for Alzheimer's disease. Sci Rep. 2017; 7: 45701. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen X, Liu Y, Zhu J, et al.: GSK-3β downregulates Nrf2 in cultured cortical neurons and in a rat model of cerebral ischemia-reperfusion. Sci Rep. 2016; 6: 20196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPang T, Wang YJ, Gao YX, et al.: A novel GSK-3β inhibitor YQ138 prevents neuronal injury induced by glutamate and brain ischemia through activation of the Nrf2 signaling pathway. Acta Pharmacol Sin. 2016; 37(6): 741–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMateo I, Infante J, Llorca J, et al.: Association between glycogen synthase kinase-3beta genetic polymorphism and late-onset Alzheimer's disease. Dement Geriatr Cogn Disord. 2006; 21(4): 228–32. PubMed Abstract | Publisher Full Text\n\nvon Otter M, Bergström P, Quattrone A, et al.: Genetic associations of Nrf2-encoding NFE2L2 variants with Parkinson's disease - a multicenter study. BMC Med Genet. 2014; 15: 131. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "35943",
"date": "13 Jul 2018",
"name": "Masatake Fujimura",
"expertise": [
"Reviewer Expertise Neurotoxicology",
"Neuropharmacology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 am very interested in this report. This article was well written without any over discussion. Although the roles of GSK-3beta on neurological disease were well investigated for long time, that was not fully resolved at this time. Authors indicated the possibility that GSK-3beta modified the state of neurological disease through the regulation of Nfr2. In this regard, however, authors only mentioned Alzheimer's disease and ischemia among neurological diseases. I hope that they discussed the roles of GSK-3beta through the regulation of Nfr2 in other neurological disease.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
},
{
"id": "35941",
"date": "19 Jul 2018",
"name": "Alexey Tinkov",
"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 provides interesting data on the role of GSK-3b-Nrf2 interplay in neurological diseases. Although the authors review the existing data in relation to Alzheimer's disease and ischemic stroke, the mechanism may play a significant role in other disorders. It is also of particular interest whether this mechanism may be involved in mercury neurotoxicity, especially in view of the earlier study by the authors (1).\nOnly minor changes may be recommended: It is recommended to add a reference to the sentence in the Introduction “Oxidative stress is a well-established feature of several cognitive disorders, including Alzheimer’s disease (AD) and brain ischemia” I suppose that the chemical nature of the compound family should be mentioned (2,4-dihydropyrano[2,3-c]pyrazoles)\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1043
|
https://f1000research.com/articles/7-8/v2
|
24 Jan 18
|
{
"type": "Method Article",
"title": "netSmooth: Network-smoothing based imputation for single cell RNA-seq",
"authors": [
"Jonathan Ronen",
"Altuna Akalin",
"Jonathan Ronen"
],
"abstract": "Single cell RNA-seq (scRNA-seq) experiments suffer from a range of characteristic technical biases, such as dropouts (zero or near zero counts) and high variance. Current analysis methods rely on imputing missing values by various means of local averaging or regression, often amplifying biases inherent in the data. We present netSmooth, a network-diffusion based method that uses priors for the covariance structure of gene expression profiles on scRNA-seq experiments in order to smooth expression values. We demonstrate that netSmooth improves clustering results of scRNA-seq experiments from distinct cell populations, time-course experiments, and cancer genomics. We provide an R package for our method, available at: https://github.com/BIMSBbioinfo/netSmooth.",
"keywords": [
"scRNA-seq",
"single-cell",
"genomics",
"imputation",
"networks"
],
"content": "Introduction\n\nSingle cell RNA sequencing (scRNA-seq) enables profiling of single cells’ transcriptomes at unprecedented throughput and resolution. It has enabled previously impractical, studies of cell type heterogeneity, differentiation, and developmental trajectories1. However, the adaptation of RNA sequencing techniques from bulk samples to single cells did not progress without challenges. Typically, only a fraction of a cells transcriptome may be captured by the experiment, leading to so called \"drop-out\" events where a gene gets a false 0 (or near 0) count in some cell. The dropout rate is related to the population level expression of a gene leading to many false zero counts for lowly expressed genes, and artificially low counts for highly expressed ones2. Furthermore, the drop-out rate could be related to the biology of the cell type, as some cell types transcribe fewer genes than others, which will appear as drop-out events2. When summed over many samples, transcript counts from single cells resemble those of bulk experiments3, but across individual cells there is significant variation. This makes analysis more difficult than in bulk RNA sequencing experiments.\n\nComputational methods designed to deal with these issues treat dropout events as missing data points, whose values may be imputed based on non-missing data points (observed measurements). The proportion of 0 counts per gene, a proxy for its technical dropout rate, is a function of the population-wise mean expression of that gene2,4. This observation has led researchers to treat 0 counts as dropout candidates to be imputed.\n\nCIDR5 attempts to impute missing values based on the predicted mean expression of a gene, given its empirical dropout rate (0-count). scImpute6 estimates dropout likelihoods per gene and per sample, and assigns each gene in each sample a status as a dropout candidate. Genes might be considered likely dropouts even with nonzero expression, and 0-count genes might not be considered likely dropouts, based on their population-wide expression distributions. It then uses a regularized linear model to predict the expression of dropout genes based on the expression of likely non-dropouts in all other cells. MAGIC7 performs local averaging after building a topological graph of the data, updating the expression value of all genes in all cells to their local neighborhood average.\n\nAll of the methods mentioned above use measured information in the data in order to impute the missing information within the same data. As such, they amplify whatever biases are present in a dataset; similar cells pre-imputation will become more similar after imputation, as expression profiles of non-dropout genes will drive similarities in imputed dropped-out genes. Further, all methods except MAGIC only impute unobserved expression events (0s or near 0s), while the dropout phenomenon actually affects the whole transcriptome. Hence, imputation methods for scRNAseq should also adjust non-0 expression measurements in order to recover the true signal.\n\nWe present a method, called netSmooth, that uses prior knowledge to temper noisy experimental data. RNA sequencing experiments produce counts data as a proxy for gene activity, which is not known a-priori, especially for experiments profiling unknown cell types. However, decades of molecular biology research have taught us much about the principles of gene interaction. Interacting genes are likely to be co-expressed in cells8,9, and as such, protein-protein interaction (PPI) databases10,11 describe genes’ propensity for co-expression. We developed a graph-diffusion method on PPI networks for smoothing of gene expression values. Each node in the graph (a gene) has an associated gene expression value, and the diffusion presents a weighted averaging of gene expression values among adjacent nodes in the graph, within each cell. This is done iteratively until convergence, strengthening co-expression patterns which are expected to be present. Incorporation of prior data from countless experiments in the preprocessing of scRNA-seq experiments improves resistance to noise and dropouts. Similar network based approaches have been used to extract meaningful information from sparse mutational profiles12,13, and indirectly on gene expression data by diffusing test statistics on the network to discover regulated gene candidates14. We propose diffusion of gene expression values directly on the network as a method for data denoising and imputation. Furthermore, the parameters of this proposed method could be optimized using clustering robustness metrics. We applied our method to a variety of single cell experiments and compared its performance to other selected imputation methods scImpute and MAGIC. These methods represent the latest and divergent ways of imputing the scRNA-seq data.\n\nWe also made available an R package providing the necessary functionality to use our method on other data. It is available on GitHub: https://github.com/BIMSBbioinfo/netSmooth.\n\n\nResults\n\nThe intuition behind the netSmooth algorithm is that gene networks encoding co-expression patterns can be used to smooth scRNA-seq data, pushing its coexpression patterns in a biologically meaningful direction. We demonstrate this using protein-protein interaction networks, which are predictive of coexpression9. We produced a PPI graph of high-confidence interactions based on the PPI database STRING10.\n\nThere are 2 inputs to the method: (1) a gene expression matrix, N genes by M cells, and (2) a graph where genes are nodes, and edges indicate genes which are expected to be co-expressed. The edges may be weighed, indicating the strength or direction of a relationship; an edge weight of 2 indicates stronger expected co-expression than an edge weight of 1, and an edge weight of −1 indicates negative expected co-expression, such as one gene being a repressor for another. The expression profile of each cell is then projected onto the graph, and a diffusion process is used to smooth the expression values, within each sample, of adjacent genes in the graph (Figure 1). In this way, post-smoothing values of genes represent an estimate of activity levels based on reads aligned to that gene, as well as those aligned to its neighbors in the graph. Thus, a gene with a low read count (possible technical drop-out), whose neighbors in the graph are highly expressed, will get a higher value post smoothing. The rate at which expression values of genes diffuse to their neighbors is degree-normalized, so that genes with many edges will affect their neighbors less than genes with more specific interactions. The diffusion is done using a \"random walks with restarts\" (RWR) process13, where a conceptual random walker starts in some node in the graph, and at each iteration moves to a neighboring node with a probability determined by the edge weight between the nodes, or, with some probability, restarts the walk from the original node. The network-smoothed value is the stationary distribution of this process. The RWR process has one free parameter, the restart rate. A low value for the restart rate allows diffusion to reach further in the graph; a high restart rate will lead to more local diffusions. For more details see the Methods section.\n\nThe expression profile of each sample is projected onto the network, where a diffusion process allows genes’ expression values to be smoothed by their neighbors’. This is done for each cell independently of others. The end result is a network smoothed gene expression matrix.\n\nWe first assess netSmooth on a dataset of 1645 mouse hematopoietic stem/progenitor cells (HSPCs) assayed using flow cytometry as well as scRNA-seq15. The cells are FACS-sorted into 12 common HSPC phenotypes. This presents an atlas of the hematopoiesis process at a single cell resolution, showing the differentiation paths taken by E-SLAM HSCs as they differentiate to E, GM, and L progenitors. The authors of this study demonstrate that upon clustering the data, some clusters corresponds to cell types. However, the clusters are not noise free and do not fully recapitulate cell type identity. We obtained clusterings of the cells from the normalized counts, as well as after application of netSmooth, MAGIC7, and scImpute6, using a robust clustering procedure based on the clusterExperiment R package16 (See Methods). After clustering, we used the edgeR-QLF test17 to identify genes that are differentially expressed in any of the discovered clusters. Figure 2a,b shows that after network-smoothing, we are able to identify clusters with a more pronounced differential expression profile. Further, many more of the genes identified as differentially expressed between the clusters (without smoothing) seem to have low and uninformative expression values overall. MAGIC and scImpute also improve this pattern (Figure 2c,d). MAGIC seems to do the strongest transformation to the data, as seen in lower dimension embeddings (Figure S2, Figure S3).\n\nA) raw (no imputation), B) after application of netSmooth, C) missing values imputed using MAGIC D) missing values imputed using scImpute.\n\nAs this dataset has cells with labels independent of the RNAseq (FACS-sorted phenotypes), it presents us with an opportunity to compare the gene expression levels (as measured by RNAseq), to a meaningful phenotypic variable, i.e. the cell type. The cell type discrimination of a clustering result is compared using a cluster purity metric and and the adjusted mutual information (AMI). The cluster purity measures how cell-type specific clusters are by comparing homogeneity of the external labels (FACS-defined cell types), within clusters provided by scRNA-seq data. AMI is a chance-adjusted information-theoretic measure of agreement between two labellings. This method accounts for artificially high mutual information between external labels and clusters when there are high number of clusters (See Methods for details on metrics). We also measured number of cells in robust clusters as quantitative metric. The robust clustering procedure allows cells to be omitted (not be assigned to a cluster) if they cannot be placed in a cluster across multiple clustering methods and/or parameters (See Methods). Only MAGIC is able to increase the proportion of cells in this dataset which fall into robust clusters (Figure 3a), but only netSmooth leads to more biologically meaningful clusters, in terms of purity and AMI (Figures 3b,c), demonstrating that netSmooth can assist in cell type identification, and outperformed both MAGIC and scImpute in this task. The higher clusterability following application of MAGIC than netSmooth, might indicate that MAGIC was overzealous in its transformation, squeezing more cells into the same space. This might lead to more robust clusters, but less reliable cell type identification.\n\nA) The proportion of cells which were assigned to robust clusters. B) cluster purity (proportion of dominant cell type) for the robust clusters. netSmooth produces the most pure clusters in terms of cell types. C) AMI of the clustering results obtained after application of each of the methods. Only netSmooth increases the AMI between the clustering and the cell types. The online version of this figure is interactive.\n\nNext, we test netSmooth on 269 isolated cells from mouse embryos at different stages of pre-implantation development between oocyte and blastocyst, as well as 5 liver cells and 10 fibroblast cells18. The authors of this study demonstrated that lower dimension embeddings capture much of the developmental trajectory (Figure 4a, Figure S4a, Figure S5a). We then applied netSmooth, MAGIC, and scImpute. Figure 4b shows the principal component analysis of netSmooth-processed data, and Figures 4c and 4d show the PCA plot following application of MAGIC and scImpute, respectively. netSmooth and scImpute preserve most of the variance structure of the data, while MAGIC seems to push the data onto a completely different manifold (Figure 4, Figure S5). We used the robust clustering procedure to obtain clusters, and computed the cluster purity and AMI metrics. netSmooth enabled the clustering procedure to place more of the samples into robust clusters (Figure 5a), and as in the hematopoiesis case, netSmooth is able to assist in identifying the developmental stage or tissue that cells belong to better than the other methods, as evidenced by the higher cluster purities (Figure 5b) and AMI (Figure 5c). Although MAGIC and scImpute reduce the 0-count genes further than netSmooth (Figure S1), they do not add as much clarity to the developmental stage signal inherent in the data. This shows that imputing missing counts based on data from the same experiment is not as powerful as including priors in the quasi-imputation process netSmooth does.\n\n2D PCA plots of the embryonic development dataset A) no preprocessing, B) after application of netSmooth, C) after imputing missing values with scImpute, and D) after application of MAGIC. The online version of this figure is interactive.\n\nA) The proportion of cells which were assigned to robust clusters. All three methods lead to better clusterability, with MAGIC having the strongest effect. B) cluster purity (proportion of dominant cell type) for the robust clusters. netSmooth produces the most pure clusters in terms of cell types. C) Adjusted mutual information of clusterings and cell types. Only netSmooth increases the AMI over the non-preprocessed data. The online version of this figure is interactive.\n\nFinally, we demonstrate applicability of netSmooth to cancer research. Patel et al. generated scRNA-seq data of 800 cells from 5 glioblastoma tumors and 2 cell lines19. Lower dimension embedding plots show that cells from different tumors or cell lines generally group together, but some are not wholly distinguishable from other tumors (Figure 6a, Figure S4a, Figure S5a). Further, the two cell lines group closer to each other than the other patient samples. After applying netSmooth to the data, tumors become easier to distinguish in a lower dimensional embedding (Figure 6b), indicating that netSmooth improves assignment of each cell to its tumor, cell line, or clone of origin. Again, scImpute also leads to similar reduced dimension embedding (Figure 6d), while MAGIC distorted the data more than the other methods (Figure 6c). We used the robust clustering procedure before and after netSmooth, MAGIC, and scImpute. Only MAGIC increase the clusterabitliy of the data (Figure 7a), but netSmooth leads to the most pure clusters, in terms of tumor or cell line of origin (Figure 7b, Figure 7c).\n\nt-SNE plots of the glioblastoma dataset A) no preprocessing, B) after application of netSmooth, C), using MAGIC, and D) after application of scImpute. The online version of this figure is interactive.\n\nA) The proportion of cells which were assigned to robust clusters. netSmooth, MAGIC, and scImpute all increased the proportion of cells that are assigned to robust clusters, with MAGIC leading, netSmooth in second place, and scImpute in third. B) cluster purity (proportion of dominant cell type) for the robust clusters. netSmooth produces the most pure clusters in terms of tumor or cell line of origin. C) AMI of the clustering results obtained after application of each of the methods. The online version of this figure is interactive.\n\nTumor or cell line of origin is an imperfect proxy for phenotypical variation in cancer cells, because some cells cluster by cell type rather than tumor of origin, demonstrating the heterogeneity in these glioblastoma tumors and similarities across origins19. Nevertheless, we chose to compute cluster purity based on the cell origin rather than other labels which might be assigned to them, as it is the only ground truth variable that is independent of the RNAseq experiment. Further, cells do group by origin (Figure 6, Figure S6), and identification of origin is an interesting question in its own right in the field of cancer genomics, particularly for heterogeneous tumors such as these.\n\nNext, we set out to ensure that the results are not an artifact of the network structure, i.e. that the actual links between genes that we used in the network are important. We expect netSmooth not to perform well when using networks with similar characteristics, but where edges do not represent real interactions. To that effect, we constructed 20 random networks by keeping the same graph structure of the real PPI graph, but shuffling the gene names.\n\nThus, these random networks share all the characteristics of the real network (degree distribution, community structure), except for the true identity of the nodes. We then used those networks as inputs to netSmooth and ran the benchmarks as before on the hematopoiesis dataset. Using random networks as an input to netSmooth gives cluster purities distributed around a mode given by the cluster purities of the raw data, while the cluster purities given from using the real PPI network lie at the extreme edge of the distribution (Figure 8a). Further, most random networks result in fewer samples belonging to robust clusters (Figure 8b). These results demonstrate that it is indeed the information contained in the PPI graph enables netSmooth to transform the gene expression matrix in a more biologically coherent direction, and that the transformation we see can not be explained simply by the network structure.\n\nA) The median cluster purity achieved with the random networks. The real network outperforms the random ones, which result in cluster purities distributed around the purity given without using netSmooth. B) The proportion of samples assigned to robust clusters using the random networks as well as the real one. While all networks result in fewer samples robustly clustered (in the hematopoiesis dataset), the real network outperforms most random networks. The online version of this figure is interactive.\n\nIn addition to using an unweighed (where all edge weights are 1), undirected (where all edge weights are positive) network from string-db, we constructed other gene networks and used them as inputs to netSmooth. We created a directed gene network from only those edges in string-db which are marked as activating or inhibitingi. We set the edge weights of the activating interactions to +1, and −1 for the inhibiting interactions, allowing gene expression values to be adjusted downwards for genes whose known antagonists are highly expressed. After smoothing, we set all negative smoothed expression values to 0. We also constructed a gene network from string-db using only genes that are known to demonstrate cell-type specific expression. In order to obtain a list of genes with such cell-type specific expression patterns from the Expression Atlas20, we used only the genes which show a cell-type specific expression with a mean TPM of at least 1 in some cell type, and used the subset of string-db network containing those genes as an input to netSmooth. Both of those modified graphs perform similarly to the undirected graph from string-db (Figure 9, Figure S8a, S8b), demonstrating that netSmooth is able to use priors from different types of experiments in order to improve clustering of scRNA-seq.\n\nRaw refers to no smoothing, non-directional is the same as the results shown in previous sections. Directional refers to a gene network where inhibitory relationships have negative edge weights, and cell-type specific refers to a gene network of only genes which are known to have cell-type specific expression patterns. The online version of this figure is interactive.\n\nWe also considered other sources for the gene network. We constructed a gene network from HumanNet11, a functional gene network where edges denote interactions between two genes. We constructed a smoothing graph by taking all edges from HumanNet, and producing a graph where all edge weights are set to 1. We then used this graph as an input to netSmooth on the glioblastoma dataset. It performs similarly to the network from string-db (Figure 10, Figure S8c), demonstrating that other sources for gene interactions may also be used by netSmooth to improve clustering results of scRNA-seq.\n\nRaw refers to no smoothing, string-db is the same as the results shown in previous sections, and HumanNet refers to a gene network constructed from the HumanNet database. The online version of this figure is interactive.\n\nThe netSmooth algorithm, given a gene network, has one free parameter - the restart rate of the random walker, (1 − α). Alternatively, α is the complement of the restart rate. An α = 0 indicates a perfect restart rate and consequently no smoothing; an α = 1 corresponds to a random walk without restarts. Intermediate values for α result in increasing levels of smoothing; the value of α determines how far random walks will go on the graph before restarting, or how far along the network a gene’s influence is allowed to reach (See Methods). It is tempting to optimize α with respect to the variable the experiment sets out to measure, e.g. cluster purity. For instance, in the embryonic development dataset, we would choose α = 0.7 as the value that produces the highest cluster purity (Figure 11b). However, in many experiments the identity of the samples is not known a-priori. Therefore, we propose a data driven workflow to pick a sensible value for α.\n\nα = 0 is equivalent to not using netSmooth at all. The procedure is robust to alpha, that is, most values of alpha produce more robust clusters. A) HSPCs, B) embryonic cells, C) glioblastomas. The online version of this figure is interactive.\n\nOne such data-driven statistic is the proportion of samples assigned to robust clusters; following application of netSmooth, the robust clustering procedure is able to assign more samples to statistically robust clusters. For all three datasets, picking the α that gives the highest proportion of cells in robust clusters, also gives the clusters with the highest purity index (Figure 12). Importantly, this metric is entirely data-driven and does not require external labels, making it feasible for any scRNA-seq study. The results in the previous sections all use the value of α picked to optimize proportion in robust clusters.\n\nA) hematopoietic stem/progenitor cells B) embryonic cells, C) glioblastomas. The online version of this figure is interactive.\n\n\nDiscussion\n\nSingle cell RNA sequencing technology provides whole-genome transcriptional profiles at unprecedented throughput and resolution. However, high variance and dropout events that happen in all current scRNA-seq platforms complicate the interpretation of the data. Methods that treat 0 counts as missing values and impute them based on nonzero values in the data may amplify biases in the data.\n\nWe presented netSmooth as a preprocessing step for scRNA-seq experiments, overcoming these challenges by the use of prior information derived from protein-protein interactions or other molecular interaction networks. We demonstrated that network smoothing assists in several standard analyses that are common in scRNA-seq studies. This procedure enhances cell type identification in hematopoiesis; it elucidates time series data and assists identification of the developmental stage of single cells. Finally, it is also applicable in cancer, improving identification of tumor of origin for glioblastomas. In addition, we showed that network smoothing parameter can be optimized by cluster robustness metrics, providing a workflow when there are no other external labels to distinguish cells. We demonstrated that netSmooth can use prior information from different sources in order to achieve this. We compared netSmooth with scImpute, a statistical genome-wide imputation method, and MAGIC, a genome-wide data smoothing algorithm, and demonstrated that while scImpute and MAGIC reduce the drop-out phenomenon more than netSmooth does, netSmooth outperforms them in amplifying the biological/technical variability ratio. netSmooth provides clusters that are more homogeneous and have higher adjusted mutual information (AMI) with respect to cell types. Although, in some cases data processed by MAGIC produces more robust clusters, the clusters returned after MAGIC processing do not have higher AMI or cluster purity. Higher robustness achieved by MAGIC processing might be due to the fact that the algorithm reinforces local structures too much in the data and producing artificially similar expression profiles between cells.\n\nFinally, netSmooth is a versatile algorithm that may be incorporated in any analysis pipeline for any experiment where the organism in question has a high quality PPI network available. Although not shown, the algorithm is applicable to any omics data set that can be constructed as a genes-by-samples matrix, such as proteomics, SNPs and copy number variation. In addition, most of the computational load of network smoothing can be done \"off-line\". As such it scales well with the number of cells, which is likely to increase in future scRNA-seq experiments.\n\n\nMethods and data\n\nThe hematopoiesis dataset15 was obtained from the Gene Expression Omnibus21. The embryonic18 and glioblastoma19 datasets were obtained from conquer22, a repository of uniformly processed scRNA-seq datasets. The datasets are available publicly, see Table 1.\n\nThe analysis for this paper was done using the companion netSmooth R-package, which is available online under Artistic-2.0 license: https://github.com/BIMSBbioinfo/netSmooth. The netSmooth R package was developed and tested under R version 3.4.2.\n\nArchived code at time of submission: https://doi.org/10.5281/zenodo.1119064.\n\nThe netSmooth algorithm takes a graph G = {V, E} where V = {genei} is the set of genes, and E = {(i → j)} is the set of edges between genes. The edge weights are degree-normalized, so that each gene’s outgoing edges’ weights sum to 1. We then define a process of random walk with restarts as in 13, on the PPI graph, where a conceptual random walker starts on a node in the graph (a gene/protein) and at each step walks to an adjacent node with the probability determined by the α times the edge weight. Further, at each step, there is a probability of (1 − α) that the walker restarts to its original node.\n\nMathematically, given a graph defined by an adjacency matrix A[MxM], where Ai j is the edge weight between gene i and gene j (and 0 for unconnected genes), and a vector f[Mx1], where fit is the probability that the walker is at node i at step t, the process is defined by\n\n\n\nThis process is convergent, and the stationary distribution is given by\n\n\n\nHence, the random walk with restarts process is a diffusion process defined on the PPI graph, or through the diffusion kernel (smoothing kernel)\n\n\n\nwhere (1 − α) is the restart probability, and A is the (column normalized) adjacency matrix of the PPI graph. Consequently, we define the network-smoothed expression profile\n\n\n\nwhere E[MxN] is the normalized count values of the M genes in the N cells.\n\nClustering analysis features prominently in scRNA-seq analyses; whether recapitulating known results or discovering new cell types, clustering cells by their gene expression profiles is commonly used to identify distinct populations. While some approaches directly take into account the zero-inflation of scRNA-seq data5, other studies use traditional methods18. There is no standard method for clustering single cell RNAseq data, as different studies produce data with different topologies, which respond differently to the various clustering algorithms.\n\nIn order to avoid optimizing different clustering routines for the different datasets we benchmark on, we have implemented a robust clustering routine based on clusterExperimentii16, a framework for robust clustering based on consensus clustering of clustering assignments obtained from different clustering algorithms, different parameters for these algorithms, and different views of the data. The different views are different reduced dimensionality projections of the data based on different techniques. Thus, no single clustering result will dominate the data, and only cluster structures which are robust to different analyses will prevail. The procedure we implemented using the framework is as follows:\n\n1. Perform different dimensionality reduction techniques on the data\n\n• PCA on the 500 most variable genes\n\n– with 5 components\n\n– with 15 components\n\n– with 50 components\n\n• Alternatively to PCA, t-SNE on the 500 most variable genes\n\n– with 2 dimensions\n\n– with 3 dimensions\n\n• Select the most variable genes\n\n– 100 most variable genes\n\n– 500 most variable genes\n\n– 1000 most variable genes\n\n2. On each reduced dimension view of the data, perform PAM clustering with K ranging from 5 to 10\n\n3. Calculate the co-clustering index for each pair of samples (the proportion of times the samples are clustered together, in the different clustering results based on the different reduced dimensions and clustering parameters above)\n\n4. Find a consensus clustering from the co-clustering matrix. This is done by constructing a dendrogram using average linkage, and traversing down the tree until a block with a self-similarity of at least 0.6, and a minimum size of 20 samples emerges. (instead of using cutree).\n\n5. Perform hierarchical clustering of the cluster medioids, with similarities based on expression of the 500 most variable genes\n\n6. Perform a DE analysis between clusters that are adjacent in the hierarchy from (5), and merge them if the proportion of genes that are found to be significantly differentially expressed between them (adjP < .05) is less than than 0.1.\n\nUsing only the 500 most variable genes insures the biological variation will dominate the technical variation, and enhances the reproducibility of t-SNE23.\n\nImportantly, samples that at step (4) don’t have a high enough affinity to any emerging cluster, will not be assigned to any cluster. The clustering is performed using the clusterExperiment::clusterSingle and clusterExperiment::clusterMany functions, the consensus clustering is obtained using the clusterExperiment::combineMany function, and the cluster merging (steps 5 and 6) using the clusterExperiment::makeDendrogram and clusterExperiment::mergeClusters functions. For more details, see 16.\n\nIn step (1) above, we cluster cells in a lower dimension embedding using either PCA24 or t-SNE25, in a dataset-dependent manner. Different single cell datasets respond better to different dimensionality reduction techniques which are better able to tease out the biological cluster structure of the data. In order to pick the right technique algorithmically, we compute the entropy in a 2D embedding. We obtained 2D embeddings from the 500 most variable genes using either PCA or t-SNE, binned them in a 20x20 grid, and computed the entropy using the discretize and entropy functions in the entropy R packageiii26. The entropy in the 2D embedding is a measure for the information captured by it. For the clustering procedure, we pick the embedding with the highest information content. For the hematopoiesis and glioblastoma datasets, this is t-SNE, while for the embryonic development dataset it is PCA (Table 2). This method may be used to pick any dimensionality reduction technique other than the ones mentioned here, which might be more suitable for other analyses.\n\nThe cluster purity metric displayed above refers to the proportion of the samples in a cluster which are of the dominant cell type in that cluster. The purity for cluster i is given by\n\n\n\nwhere Ci = {j|cellj ∈ clusteri}, labelj is the cell type of cellj, ni = |Ci| is the number of cells in cluster i, and\n\n\n\nis the dominant cell type in cluster Ci.\n\nIn addition to the cluster purity metric, we computed the Adjusted Mutual Information (AMI)27, an information theoretic measure of clustering accuracy which accounts for true positives (two cells of the same type in the same cluster) being caused by chance. The AMI between a clustering C and the true labels L is given by\n\n\n\nwhere MI(a, b) is the mutual information between labellings a and b, H(a) is entropy of clustering a, and E[·] denotes the expectation.\n\nWe do not compare the clusterings using the Rand index, as that measure penalizes for so-called false negatives (two cells of the same cell type but in different clusters), which is undesirable as cells from the same cell type might be rightly split into several clusters when a novel cell type is identified.\n\nThe PPI graph from which the diffusion kernel was derived was constructed using data from string-db10. For each pair of proteins, string-db provides a combined interaction score, which is a score indicating how confident we can be in the interaction between the proteins, given the different kinds of evidence string-db collates. We subset the links to only those above the 90th percentile of combined interaction scores, only keeping the 10% most confident interactions. For mouse that is 1,020,816 interactions among 17013 genes. For human, 852,722 interactions among 17467 genes.\n\nFor all the results presented in this paper, scImpute was run using the default parameters (drop_thre = 0.5). For MAGIC, we used values for the diffusion time parameter (T = {1, 2, 4, 8, 16}). Unlike netSmooth, for MAGIC the proportion of samples in robust clusters and the cluster purities were anti-correlated; thus we picked the one that gave the best cluster purities as the best MAGIC parameter. The chosen T values are given in Table 3. We used MAGIC version 0.14 and scImpute version 0.0.25.",
"appendix": "Author contributions\n\n\n\nAA conceptualized the project, AA and JR conceived of the algorithm together. All the analysis and software development was done by JR, who also wrote the initial draft of manuscript with input from AA. AA supervised the writing, software development and analysis. JR wrote R package with input and code review and contributions from AA.\n\n\nCompeting interests\n\n\n\nThe authors declare none.\n\n\nGrant information\n\nAA and JR are funded by core funding from Max Delbrück Center, part of Helmholtz Association.\n\n\nAcknowledgements\n\nWe would like to thank Vedran Franke, Bora Uyar and Brendan Osberg for valuable comments and input for the development of this manuscript.\n\n\nSupplementary material\n\nFigure S1. PCA plots of the HSPC dataset A) no preprocessing, B) after application of netSmooth, C), using scImpute, and D) after application of MAGIC. The online version of this figure is interactive.\n\nClick here to access the data.\n\nFigure S2. t-SNE plots of the HSPC dataset A) no preprocessing, B) after application of netSmooth, C), using scImpute, and D) after application of MAGIC. The online version of this figure is interactive\n\nClick here to access the data.\n\nFigure S3. Single cells from the embryonic development dataset were clustered using the robust clustering procedure, and the 500 most differentially expressed genes (by edgeR-QLF test adjusted P value) in any of the discovered clusters are shown in a heatmap, as well as cluster assignments and cell types. A) raw (no imputation), B) after application of netSmooth, C) missing values imputed using scImpute D) after application of MAGIC.\n\nClick here to access the data.\n\nFigure S4. t-SNE plots of the embvryonic development dataset A) no preprocessing, B) after application of netSmooth, C), using scImpute, and D) after application of MAGIC. The online version of this figure is interactive.\n\nClick here to access the data.\n\nFigure S5. The proportion of genes with 0 counts is a proxy for technical dropouts. A) no preprocessing, B) after application of netSmooth, C), using scImpute, and D) after application of MAGIC.\n\nClick here to access the data.\n\nFigure S6. Single cells from the glioblastoma dataset were clustered using the robust clustering procedure, and the 500 most differentially expressed genes (by edgeR-QLF test adjusted P value) in any of the discovered clusters are shown in a heatmap, as well as cluster assignments and cell types. A) raw (no imputation), B) after application of netSmooth, C) missing values imputed using scImpute D) after application of MAGIC.\n\nClick here to access the data.\n\nFigure S7. PCA plots of the glioblastoma dataset A) no preprocessing, B) after application of netSmooth, C), using scImpute, and D) after application of MAGIC. The online version of this figure is interactive.\n\nClick here to access the data.\n\nFigure S8. Cluster purity by smoothing parameter. A) for the hematopoiesis dataset with a directional (signed) graph, where inhibitory interactions have a negative edge weight. B) For the hematopoiesis dataset using a gene network with only genes that have a cell-type specific expression in any cell type. C) In the glioblastoma dataset using a gene network from HumanNet. The online version of this figure is interactive.\n\nClick here to access the data.\n\n\nFootnotes\n\ni Most interactions in string-db do not specify the direction, or nature of the interaction\n\nii Version 1.4.0, available from Bioconductor https://bioconductor.org/packages/release/bioc/html/clusterExperiment.html\n\niii Version 1.2.1, available from CRAN: https://cran.rproject.org/web/packages/entropy/index.html\n\n4 Available from GitHub: https://github.com/pkathail/magic.\n\n5 Available from GitHub: https://github.com/Vivianstats/scImpute.\n\n\nReferences\n\nWagner A, Regev A, Yosef N: Revealing the vectors of cellular identity with single-cell genomics. Nat Biotechnol. 2016; 34(11): 1145–1160. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKharchenko PV, Silberstein L, Scadden DT: Bayesian approach to single-cell differential expression analysis. Nat Methods. 2014; 11(7): 740–742. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu AR, Neff NF, Kalisky T, et al.: Quantitative assessment of single-cell RNA-sequencing methods. Nat Methods. 2014; 11(1): 41–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPierson E, Yau C: ZIFA: Dimensionality reduction for zero-inflated single-cell gene expression analysis. Genome Biol. 2015; 16: 241. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin P, Troup M, Ho JW: CIDR: Ultrafast and accurate clustering through imputation for single-cell RNA-seq data. Genome Biol. 2017; 18(1): 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi WV, Li JJ: scimpute: Accurate and robust imputation for single cell RNA-seq data. bioRxiv. 2017. Publisher Full Text\n\nvan Dijk D, Nainys J, Sharma R, et al.: Magic: A diffusion-based imputation method reveals gene-gene interactions in single-cell RNA-sequencing data. bioRxiv. 2017. Publisher Full Text\n\nBhardwaj N, Lu H: Correlation between gene expression profiles and protein-protein interactions within and across genomes. Bioinformatics. 2005; 21(11): 2730–2738. PubMed Abstract | Publisher Full Text\n\nFraser HB, Hirsh AE, Wall DP, et al.: Coevolution of gene expression among interacting proteins. Proc Natl Acad Sci U S A. 2004; 101(24): 9033–9038. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSzklarczyk D, Morris JH, Cook H, et al.: The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45(D1): D362–D368. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee I, Blom UM, Wang PI, et al.: Prioritizing candidate disease genes by network-based boosting of genome-wide association data. Genome Res. 2011; 21(7): 1109–1121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHofree M, Shen JP, Carter H, et al.: Network-based stratification of tumor mutations. Nat Methods. 2013; 10(11): 1108–1115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVandin F, Upfal E, Raphael BJ: Algorithms for detecting significantly mutated pathways in cancer. J Comput Biol. 2011; 18(3): 507–522. PubMed Abstract | Publisher Full Text\n\nDørum G, Snipen L, Solheim M, et al.: Smoothing gene expression data with network information improves consistency of regulated genes. Stat Appl Genet Mol Biol. 2011; 10(1): pii: /j/sagmb.2011.10.issue-1/sagmb.2011.10.1.1618/sagmb.2011.10.1.1618.xml. PubMed Abstract | Publisher Full Text\n\nNestorowa S, Hamey FK, Pijuan Sala B, et al.: A single-cell resolution map of mouse hematopoietic stem and progenitor cell differentiation. Blood. 2016; 128(8): e20–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPurdom E, Risso D: clusterExperiment: Compare Clusterings for Single-Cell Sequencing. R package version 1.2.0. 2017. Reference Source\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeng Q, Ramsköld D, Reinius B, et al.: Single-cell RNA-seq reveals dynamic, random monoallelic gene expression in mammalian cells. Science. 2014; 343(6167): 193–196. PubMed Abstract | Publisher Full Text\n\nPatel AP, Tirosh I, Trombetta JJ, et al.: Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science. 2014; 344(6190): 1396–1401. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetryszak R, Keays M, Tang YA, et al.: Expression Atlas update--an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 2016; 44(D1): D746–D752. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdgar R, Domrachev M, Lash AE: Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002; 30(1): 207–210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoneson C, Robinson MD: Bias, robustness and scalability in differential expression analysis of single-cell RNA-seq data. bioRxiv. 2017. Publisher Full Text\n\nMcCarthy DJ, Campbell KR, Lun AT, et al.: Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics. 2017; 33(8): 1179–1186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHastie T, Tibshirani R, Friedman J: The Elements of Statistical Learning. Springer Series in Statistics. Springer New York Inc., New York, NY USA, 2001. Reference Source\n\nvan der Maaten LJP, Hinton GE: Visualizing high-dimensional data using t-SNE. J Mach Learn Res. 2008; 9: 2579–2605. Reference Source\n\nHausser J, Strimmer K: entropy: Estimation of Entropy, Mutual Information and Related Quantities. 2014. Reference Source\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"
}
|
[
{
"id": "30157",
"date": "05 Feb 2018",
"name": "Fernando J. Calero-Nieto",
"expertise": [
"Reviewer Expertise Haematopoiesis",
"single-cell technologies"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors describe a method for imputing values to overcome the problem of technical dropouts in single-cell RNA-seq datasets. As stated by the authors, the problem is well known and caused by technical limitations that affect low and high expressed genes. The approach discussed in the manuscript uses prior knowledge about protein interactions in order to smooth the expression values between pairs of genes encoding interacting proteins, thus reducing the number of zero values and altering the expression values of detected genes in each cell and influencing clustering and visualisation results.\n\nThe proposed fundament is interesting and certainly worth exploring. However, there are a few considerations when using this type of approach: 1) the possible enhancement of known relationships to the detriment of the discovery of previously unknown ones; 2) Inferring dropouts using pre-known interactions could result in the overestimation of the expression of certain genes; 3) the gene relationships tend to be very cell-type specific so networks and PPIs should be different from cell type to cell type. In relation to the latter point, it is very interesting that the method is flexible enough to accept networks constructed from different sources.\n\nThis manuscript compares the netSmooth algorithm to two existing approaches: Magic and scImpute. Overall, netSmooth presents an approach to smooth scRNA-seq data, which may prove useful in noisy datasets affected heavily by dropouts. However, there are several aspects of the manuscript that we found unclear or feel warrant further discussion.\n\nAn important point is how the performance of these type of methods can be assessed. The authors decided to use a combination of clustered heatmaps, robustness and purity of the clustering, including a measurement of the correspondence between the 2 of them (AMI). Important downsides of this method are: a) external annotation of the datasets is required, which is not always available; b) robustness of the clustering seems to be strongly affected by the processing of the data, as the authors show in relation to MAGIC; c) the purity of the clustering could be strongly biased by the size of the clusters, since small clusters could have a greater chance to get a higher score. It would be interesting if the authors could comment on how their metrics are affected by the number of robust clusters identified. For example, could identifying more small clusters in the dataset have an effect in increasing the median cluster purity? And if so, is this a reliable measure for comparing between algorithms. It would be useful to include colour bars for the heatmaps. It should also be mentioned what scale the data is plotted using e.g. is it linear or log-transformed and is the scale comparable between all of the processed datasets? Additionally, some panels (for example 2C and 2D) have more clusters than there are colors shown in the cluster color key next to panel A, so the keys should be changed to match the data. Figures showing the cluster purity are quite confusing. From the legend and methods, we understood that each point on the boxplot represents the purity for one of the clusters displayed in the clustered heatmaps. Yet in figure 2A, for example, there are 4 robust clusters found in the raw data, but 8 points for the raw clusters in figure 2B. It seems either that there is an error in one of the plots, or that we have misunderstood the cluster purity metric in which case it needs to be more clearly explained. For continuity purposes, it would be better that either PCA or tSNE visualisations were shown for all dataset comparisons in the main figures instead of alternation between clustered heatmaps, PCA and tSNE. In the introduction section, the authors mention the imputation programme CIDR along with scImpute and MAGIC, yet only compare the performance of netSmooth to scImpute and MAGIC. The authors should either include benchmarking against CIDR on the three datasets, or discuss why this is not appropriate. When discussing applying netSmooth to the haematopoietic data, the authors state that “Figure 2a,b shows that after network-smoothing, we are able to identify clusters with a more pronounced differential expression profile. Further, many more of the genes identified as differentially expressed between the clusters (without smoothing) seem to have low and uninformative expression values overall.” However, from visual inspection of Figure 2 there appears to be very little difference in the expression levels of the differentially expressed genes in the two heatmaps, or in the number of genes with low expression levels. We are unsure exactly what the authors mean by “more pronounced differential expression profile” as it is hard to see a difference in the heatmaps. The authors state that “Only MAGIC is able to increase the proportion of cells in this dataset which fall into robust clusters (Figure 3a), but only netSmooth leads to more biologically meaningful clusters, in terms of purity and AMI (Figures 3b,c), demonstrating that netSmooth can assist in cell type identification, and outperformed both MAGIC and scImpute in this task.” The increase in AMI in 3B is marginal compared to the raw data, and the proportion in robust clusters is higher for raw data than for netSmooth. There is also no clear improvement in the visualizations of figures S1 and S2 between netSmooth and raw data. Combined with the heatmaps in figure 2, we didn’t feel that there was compelling evidence that netSmooth was useful in cell type identification, and therefore this statement should be toned down. In figure 4, it is hard to see how either netSmooth or scImpute offer an improved visualization compared to the raw data. This is backed up by very similar metric scores in Figure 5A and 5C between the raw, netSmooth and scImpute bars. Therefore the statement “Although MAGIC and scImpute reduce the 0-count genes further than netSmooth (Figure S1), they do not add as much clarity to the developmental stage signal inherent in the data.” appears to overstate how well netSmooth performs on this dataset in comparison to the other two algorithms. In several places the text references the wrong supplementary figures. For example, in the sentence “Although MAGIC and scImpute reduce the 0-count genes further than netSmooth (Figure S1)” the authors appear to be actually referring to Figure S5. In Figure S5, it should be clarified what is plotted. The legend needs to be changed to make it clear what this is showing. Is this the proportion of zero genes per cell in each dataset? Also, the data in this figure suggests that this method has a stronger effect on the expression of genes that are already expressed more than in the removal of zeros. The authors should comment on this in the main manuscript. When applying netSmooth to the tumor data, the authors assess the ability of their algorithm on the extent to which it separates cells from different samples, stating “it is also applicable in cancer, improving identification of tumor of origin for glioblastomas.” In fact, many researchers are actually interested in removing this effect in order to be able to compare similar cell types between different patients. Is it possible that netSmooth is actually enhancing “batch effect” in this dataset? It would be interesting to see whether netSmooth increases technical (rather than biological) batch effect in another dataset where a strong biological batch effect is not expected. When assessing the importance of the PPI network structure, the authors calculate clustering metrics for randomly permuted networks. Can the authors comment on the fact some random networks have better cluster purity than the real network? Also, why do the authors not show AMI for these random clusters when it is often used to support the success of netSmooth compared to other approaches (e.g. in the haematopoiesis dataset)? When discussing the parameter selection the authors state that “in the embryonic development dataset, we would choose alpha= 0.7 as the value that produces the highest cluster purity”. But in figure 11B it is actually alpha= 0.4 that has the highest cluster purity. It would be interesting if the authors commented in why there are at least 2 alpha values that get very similar maximum values for each dataset. Also, the figure would benefit from including alpha=0 values to compare with raw data.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3772",
"date": "10 Jul 2018",
"name": "Altuna Akalin",
"role": "Reader Comment",
"response": "We are thankful for these valuable comments by Drs. Hamey and Calero-Nieto. We responded to the comments and changed the text and figures wherever necessary. 1. An important point is how the performance of these type of methods can be assessed. The authors decided to use a combination of clustered heatmaps, robustness and purity of the clustering, including a measurement of the correspondence between the 2 of them (AMI). Important downsides of this method are: a) external annotation of the datasets is required, which is not always available; b) robustness of the clustering seems to be strongly affected by the processing of the data, as the authors show in relation to MAGIC; c) the purity of the clustering could be strongly biased by the size of the clusters, since small clusters could have a greater chance to get a higher score. It would be interesting if the authors could comment on how their metrics are affected by the number of robust clusters identified. For example, could identifying more small clusters in the dataset have an effect in increasing the median cluster purity? And if so, is this a reliable measure for comparing between algorithms.The reviewers raised the accurate point that the cluster purity metric may be biased towards clusterings with larger numbers of clusters. For instance, at the edge case, a clustering which assigns a unique cluster to each sample, will score 100% on cluster purity. In order to address this issue, we also computed the adjusted mutual information (AMI) for each clustering, which is a chance-adjusted metric which is not biased in the same way. We expand on this both in the results section under \"Network smoothing improves cell type identification from single-cell RNA-seq\", and in the methods section.2.It would be useful to include colour bars for the heatmaps. It should also be mentioned what scale the data is plotted using e.g. is it linear or log-transformed and is the scale comparable between all of the processed datasets? Additionally, some panels (for example 2C and 2D) have more clusters than there are colors shown in the cluster color key next to panel A, so the keys should be changed to match the data. The Reviewers note that the heatmap figures (Figures 2, S3 and S6) are not clear about what the value is in the heatmap, nor do they include a colorbar to gauge whether expression values in the different methods are comparable. We agree that this was a shortcoming and have amended those figures and legends to reflect what is plotted more accurately.3. Figures showing the cluster purity are quite confusing. From the legend and methods, we understood that each point on the boxplot represents the purity for one of the clusters displayed in the clustered heatmaps. Yet in figure 2A, for example, there are 4 robust clusters found in the raw data, but 8 points for the raw clusters in figure 2B. It seems either that there is an error in one of the plots, or that we have misunderstood the cluster purity metric in which case it needs to be more clearly explained. The reviewers correctly point out a mistake in the plot where figure 2A only shows 4 robust clusters, where there are in fact 8. We have re-created and corrected the error, and thank the reviewers for pointing out our mistake.4. For continuity purposes, it would be better that either PCA or tSNE visualisations were shown for all dataset comparisons in the main figures instead of alternation between clustered heatmaps, PCA and tSNE. The reviewers point out that we alternate between using PCA and t-SNE for scatter plots of the different datasets, and express a wish for consistency with the visialuzations throughout the paper. While we understand the desire for consistency in the visual information presented, we wish to stress that this was done on purpose. Different scRNAseq datasets respond differently to the different dimensionality reduction techniques, and it is standard practice in the community to try more than one and then pick the best one ad-hoc. We present as a part of the netSmooth R package a way to automate this step using the entropy of 2D embeddings. We expand on this in more detail in the Methods section, under \"Choice of dimensionality reduction technique inthe clustering procedure\".5. In the introduction section, the authors mention the imputation programme CIDR along with scImpute and MAGIC, yet only compare the performance of netSmooth to scImpute and MAGIC. The authors should either include benchmarking against CIDR on the three datasets, or discuss why this is not appropriate. The reviewers pointed out that while we included CIDR, a method for imputation and clustering of scRNAseq in our introduction, we did not benchmark it against our method. The reason for this omission was that CIDR uses a built-in clustering procedure as a part of the imputation workflow. We chose to compare to MAGIC and scImpute, which are agnostic to the clustering procedure, in order to have an apples-to-apples comparison of imputation methods using the same post-imputation analysis. However we agree with the reviewers that the omission may have been glaring, and have included benchmarks that were possible. We were not able to compare them on the cluster robustness metric, as CIDR assigns all samples to clusters, and does now have a notion of robust clusters.6. When discussing applying netSmooth to the haematopoietic data, the authors state that “Figure 2a,b shows that after network-smoothing, we are able to identify clusters with a more pronounced differential expression profile. Further, many more of the genes identified as differentially expressed between the clusters (without smoothing) seem to have low and uninformative expression values overall.” However, from visual inspection of Figure 2 there appears to be very little difference in the expression levels of the differentially expressed genes in the two heatmaps, or in the number of genes with low expression levels. We are unsure exactly what the authors mean by “more pronounced differential expression profile” as it is hard to see a difference in the heatmaps. We agree with the reviewers that the claim about the more pronounced differential expression pattern in the heatmap was unsubstantiated, and have accordingly changed the text to point out that the difference in the heatmaps is negligible7. The authors state that “Only MAGIC is able to increase the proportion of cells in this dataset which fall into robust clusters (Figure 3a), but only netSmooth leads to more biologically meaningful clusters, in terms of purity and AMI (Figures 3b,c), demonstrating that netSmooth can assist in cell type identification, and outperformed both MAGIC and scImpute in this task.” The increase in AMI in 3B is marginal compared to the raw data, and the proportion in robust clusters is higher for raw data than for netSmooth. There is also no clear improvement in the visualizations of figures S1 and S2 between netSmooth and raw data. Combined with the heatmaps in figure 2, we didn’t feel that there was compelling evidence that netSmooth was useful in cell type identification, and therefore this statement should be toned down. The reviewers point out that the difference in benchmark scores shown in Figure 3 represent only a modest improvement over the raw data, and that the statement about the improvement gained from applying netSmooth should be toned down. We've taken this advice and updated the text to make more modest claims.8. In figure 4, it is hard to see how either netSmooth or scImpute offer an improved visualization compared to the raw data. This is backed up by very similar metric scores in Figure 5A and 5C between the raw, netSmooth and scImpute bars. Therefore the statement “Although MAGIC and scImpute reduce the 0-count genes further than netSmooth (Figure S1), they do not add as much clarity to the developmental stage signal inherent in the data.” appears to overstate how well netSmooth performs on this dataset in comparison to the other two algorithms. The reviewers suggest that several of the statements about results, in Figures 2, S2, S3, and 4, over-state the performance of netSmooth relative to the other methods we compared it to. We have toned down several of the statements, and hope the reviewers will find the current text more acceptable.9. In several places the text references the wrong supplementary figures. For example, in the sentence “Although MAGIC and scImpute reduce the 0-count genes further than netSmooth (Figure S1)” the authors appear to be actually referring to Figure S5. Thank you. This is corrected 10. In Figure S5, it should be clarified what is plotted. The legend needs to be changed to make it clear what this is showing. Is this the proportion of zero genes per cell in each dataset? Also, the data in this figure suggests that this method has a stronger effect on the expression of genes that are already expressed more than in the removal of zeros. The authors should comment on this in the main manuscript.We have changed the legend in order to make this figure more clear. We also appreciate the note from the reviewers about netSmooth having a stronger effect on nonzero genes than the dropouts, compared with other imputation methods, and have added a short discussion of this, under the section \"Network smoothing improves capture of developmental expression patterns''.11. When applying netSmooth to the tumor data, the authors assess the ability of their algorithm on the extent to which it separates cells from different samples, stating “it is also applicable in cancer, improving identification of tumor of origin for glioblastomas.” In fact, many researchers are actually interested in removing this effect in order to be able to compare similar cell types between different patients. Is it possible that netSmooth is actually enhancing “batch effect” in this dataset? It would be interesting to see whether netSmooth increases technical (rather than biological) batch effect in another dataset where a strong biological batch effect is not expected. The reviewers note that in the Glioblastoma case, we benchmark the methods' ability to identify cells' tumor of origin, while researchers might in fact be interested in the opposite - removing this effect in order to compare cells between cell types. The reviewers are correct and Patel et. al. (the originators of the data) note that these heterogenous tumors consist of different cell types (Pro-neural, Neural, Mesenchymal, and Classical). Identifying such subsets across tumors is an interesting question for researchers of tumor heterogeneity, and we believe that netSmooth might in fact assist in identifying such groups, in this case by making clear which part of the expression signature is patient specific, and which one owed to cross-tumor signatures.12. When assessing the importance of the PPI network structure, the authors calculate clustering metrics for randomly permuted networks. Can the authors comment on the fact some random networks have better cluster purity than the real network? Also, why do the authors not show AMI for these random clusters when it is often used to support the success of netSmooth compared to other approaches (e.g. in the haematopoiesis dataset)? Reviewers point out that some random networks produces better cluster purity. Actually, figure 8 demonstrates that using the real network in netSmooth leads to cluster purity in the extreme edge of the distribution of random networks. Certainly, with enough random perturbations, a random network can be constructed that will outperform a real network. We demonstrate that in spite of this, the real network scores significantly above expectation in each of the metrics, which demonstrates that the real network holds useful information.The reviewers point out that when we run the benchmarks using randomized networks in order to demonstrate that the true network structure is important to the results, we only showed the cluster purity and proportion in robust clusters. We agree with the reviewers that the AMI metric also belongs in this context, and have added the AMI to that plot as well.13. When discussing the parameter selection the authors state that “in the embryonic development dataset, we would choose alpha= 0.7 as the value that produces the highest cluster purity”. But in figure 11B it is actually alpha= 0.4 that has the highest cluster purity. It would be interesting if the authors commented in why there are at least 2 alpha values that get very similar maximum values for each dataset. Also, the figure would benefit from including alpha=0 values to compare with raw data.The reviewers noted a mistake in the text referring the the value of alpha which results in the optimal cluster purity in figure 11. We have corrected the misprint."
}
]
},
{
"id": "30156",
"date": "14 Feb 2018",
"name": "Siddharth Dey",
"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 manuscript, the authors have developed a new computational method to reduce technical biases that result from dropout events in single-cell mRNA sequencing experiments, a problem that particularly affects genes that are expressed at low levels. While single-cell mRNA sequencing has revolutionized our understanding of several biological systems in the last few years, the relatively low efficiency of amplifying small quantities of mRNA from a single cell results in dropout events that bias downstream analysis. To reduce this technical bias, the authors use information from protein-protein interaction maps to smoothen transcript counts across the entire dataset. Several groups are working on imputation based methods in single-cell mRNA-seq, and this manuscript presents an exciting approach to reduce technical noise. It would be helpful if the authors could clarify and discuss the points below in greater detail:\n\n1. Does the smoothening process bias against genes or gene networks that are not well represented in the protein-protein interaction network? One of the striking features of sc mRNA-seq is that it can identify the expression of specific genes that were previously not associated with a particular cell-type. Would this be impacted by netSmooth and can the authors provide examples from the datasets they have analyzed that netSmooth still retains these observations?\n\n2. There are several sc mRNA-seq methods (for example, CEL-Seq, Smart-Seq etc.) that are currently used by different labs. These methods have different features, such as, full-length transcripts or 3’ end sequencing, and the possibility of employing unique molecule identifiers. How do the 3 computational methods compared in this manuscript work on different experimental techniques?\n\n3. For most of the example datasets used in this manuscript, scImpute shows very similar performance to netSmooth on all 3 metrics used to compare the methods. Can the authors discuss how these two methods, while using different approaches, achieve similar performance. Are there conditions/datasets where one method would perform better than the other?\n\n4. The proportion of cells in robust clusters seems to be very sensitive to the choice of the free parameter in netSmooth (Figure 12). Further, in contrast to a statement in the text (last paragraph on page 10), the value of the free parameter that gives the highest proportion of cells in robust clusters does not correspond to the highest median cluster purity in the glioblastoma dataset. This high sensitivity to alpha can potentially pose a challenge. Can the authors comment on this? Could the authors propose alternate strategies for picking the optimal alpha value.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3771",
"date": "10 Jul 2018",
"name": "Altuna Akalin",
"role": "Reader Comment",
"response": "We thank Dr. Dey for valuable comments. We tried to address them as demonstrated below and we are in the process of uploading a new version with changes. 1. Does the smoothening process bias against genes or gene networks that are not well represented in the protein-protein interaction network? One of the striking features of sc mRNA-seq is that it can identify the expression of specific genes that were previously not associated with a particular cell-type. Would this be impacted by netSmooth and can the authors provide examples from the datasets they have analyzed that netSmooth still retains these observations? Dr. Dey point out that we have demonstrated netSmooth using gene networks derived from all sorts of general context experiments, while scRNAseq experiments might reveal cell-type-specific gene interactions. We appreciate this comment, and have added a paragraph discussing this to the text. We would mostly like to underline that, until such context-specific networks may be constructed, we have demonstrated netSmooth's applicability using general context networks. 2. There are several sc mRNA-seq methods (for example, CEL-Seq, Smart-Seq etc.) that are currently used by different labs. These methods have different features, such as, full-length transcripts or 3’ end sequencing, and the possibility of employing unique molecule identifiers. How do the 3 computational methods compared in this manuscript work on different experimental techniques? Dr. Dey raised interesting questions about the performance of the different imputation methods we compared, coupled with different scRNAseq methods (Smart-Seq, CEL-Seq, etc.). While this is a highly relevant question, we feel it is beyond the scope of this study, as answering that question would require obtaining relevant datasets with appropriate ground-truth labels, and using each relevant technique 3. For most of the example datasets used in this manuscript, scImpute shows very similar performance to netSmooth on all 3 metrics used to compare the methods. Can the authors discuss how these two methods, while using different approaches, achieve similar performance. Are there conditions/datasets where one method would perform better than the other? We agree that there is a similar performance of scImpute and netSmooth, although netSmooth is slightly better in our metrics when looking at all the data sets. Our analysis shows that netSmooth affects the drop-out rate less than scImpute, while uncovering slightly more of the biological signal. This happens across the different overall drop-out rates in the 3 experiments we profiled, indicating that netSmooth can achieve better results with less obtrusive transformations of the data than the imputation methods, across a range of experimental conditions. 4. The proportion of cells in robust clusters seems to be very sensitive to the choice of the free parameter in netSmooth (Figure 12). Further, in contrast to a statement in the text (last paragraph on page 10), the value of the free parameter that gives the highest proportion of cells in robust clusters does not correspond to the highest median cluster purity in the glioblastoma dataset. This high sensitivity to alpha can potentially pose a challenge. We provide also a different way of picking alpha parameter in the R package, which is based on 2D entropy. This way we can pick alpha that optimizes the entropy in 2D PCA embedding."
}
]
}
] | 2
|
https://f1000research.com/articles/7-8
|
https://f1000research.com/articles/7-726/v1
|
12 Jun 18
|
{
"type": "Research Article",
"title": "Morphology and performance of polyvinyl chloride membrane modified with Pluronic F127",
"authors": [
"Nasrul Arahman",
"Afrilia Fahrina",
"Mukramah Yusuf Wahab",
"Umi Fathanah",
"Afrilia Fahrina",
"Mukramah Yusuf Wahab",
"Umi Fathanah"
],
"abstract": "Background: Attempts to modify the morphology of membrane for application in industrial separation are being undertaken by many researchers. The present study discusses the morphological modification of polyvinyl chloride (PVC) membrane by combining the hydrophilic surfactant Pluronic F127 (PF127) in a polymer solution to improve the performance of the membrane. Method: The membrane is formed using the non-solvent induced-phase separation (NIPS) method. PF127 is added to the membrane solution as a membrane modifying agent. The effects of the surfactant concentration in the dope solution on the permeability of pure water, solute rejection, hydrophilic characteristics, and membrane morphology are investigated. Results: Higher concentrations of PF127 had a significant effect on the permeability of pure water. The highest membrane permeation was 45.65 l/m2.hr.atm with the addition of 7% PF127 additive. Conclusion: PF127 is successfully proposed as a membrane pore-forming agent in this work; the blending of this additive in appropriate amounts in the polymer solution is adequate to improve the performance of the PVC membrane.",
"keywords": [
"polyvinyl chloride (PVC)",
"Pluronic F127",
"pore forming agent"
],
"content": "Introduction\n\nNowadays, separation of contaminant elements from drinking water using membrane technology is developing rapidly. Membrane separation technology has been adopted in many industries, owing to its numerous advantages compared with other common methods. One of the most widely applicable membrane separation technologies in industry is the use of a group of ultrafiltration (UF) membranes, particularly for the process of water purification1,2. In view of the requirements for application in the water treatment industry, the modification of UF membrane to generate high flux, improve the resistance to fouling and chemical substances, and provide good mechanical properties is being continuously improved3,4.\n\nPolyvinyl chloride (PVC), with the molecular formula shown in Figure 1, is a relatively cheap polymer with suitable chemical characteristics for use as a membrane material. Hydrophobic PVC polymers cause fouling of the membrane pores due to the adherence of organic molecules to the surface of the membrane. Numerous methods have been developed to improve fouling resistance. The most common method is improving the hydrophilicity of the membrane material to minimize the attachment of foulant molecules5,6.\n\nThe hydrophilic polymers that are frequently used as an additive are polyvinylpyrrolidone, polyethylene glycol, Brij, Tetronic, and Pluronic7,8. Of these, Pluronic is used as a surface modifying agent for many hydrophobic polymers. Raslan and Mohammad9 added Pluronic F127 to a polysulfone membrane. The resulting membrane is resistant to fouling and possesses good pore distribution9. Pluronic has also been used to improve the anti-fouling of cellulose acetate (CA) membranes10. The combination of CA polymer and Pluronic surfactant results in a membrane that is more resistant to fouling and has a more stable filtration profile.\n\nIn this study, pluronic was developed to improve the performance of PVC membranes. PF127 is a copolymer with two segments—hydrophilic and hydrophobic (Figure 2). The polyethylene oxide (PEO) segment of PF127 improves the hydrophilic characteristics of the membrane’s surface, while polypropylene oxide, which is hydrophobic, attaches closely to the matrix of the membrane9. The hydrophile–lipophile balance value of PF127 ranges from 18 to 2311. In this study, PF127 is used as an additive to improve the performance of a PVC membrane.\n\nx, ethylene oxide (EO) number, y, propylene oxide (PO) number.\n\n\nMethods\n\nPolyvinyl chloride (PVC) with an average molecule weight of 43,000 Da was obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). Dimethyl acetate (DMAc) solvent was obtained from Wako, Pure Industries Japan. Distilled water was produced in the laboratory. PF127 was obtained from BASF Co. (Ludwigshafen, Germany). Dextran with an average molecular weight of 10,000 Da, which was used for the rejection test, was bought from Sigma-Aldrich. All chemicals were used directly without previous treatment.\n\nThe wet inversion technique was adopted to prepare the membrane using water as a non-solvent coagulation media. PVC with a concentration of 15 wt% was dissolved in DMAc until it was homogeneous. The solution was stirred with a magnetic stirrer at 200 rpm. The homogeneous membrane solution was left for 24 hours at room temperature to completely discharge the air bubbles. The solution was then framed on a glass plate using an automatically adjustable applicator (YBA-3, Yoshimitsu, Japan) at a thickness of 0.450 µm. The glass plate containing membrane was dipped in a coagulation bath of distilled water. The de-mixing process between the DMAc solvent and non-solvent distilled water solidified the membrane and separated it from the glass plate. To improve the performance of the membrane, PF127 was added at concentrations of 1, 3, 5 and 7 wt%.\n\nMembrane morphology was observed using scanning electron microscopy (SEM) (Hitachi Co, S-800) with an accelerating voltage of 15 kV. To obtain a clean and dry sample, about 1 cm2 of the membrane sample was freeze-dried (Eyela FD-1000, Japan) for 24 hours. To ensure that the structure of the membrane was not damaged, the membrane sample was ruptured in liquid nitrogen. Next, the membrane sample was mounted on the metal module, followed by the coating process with Pt/Pd sputtering. The coated sample was inserted into the SEM apparatus, and the photo was captured at 5.0 kV. Three images was collected for each Three images each were collected for PVC membranes containing 0, 3, 5 and 7% P127.\n\nThe permeability of water and solute rejection were tested with the module of dead-end filtration (Advantec, UHP-43K, Japan). The transmembrane pressure was regulated at a pressure of 0.5 atm. The effective membrane surface area that passed by water was 0.023 m2. The testing of water permeability was conducted four times, and the average values was taken to determine the final permeability. The permeability coefficient of pure water was counted using Equation 1.\n\n\n\nWhere Lp = permeability coefficient (L/m2.jam.atm); V = permeate volume (L); A = membrane surface area (m2); and Δp = pressure change (atm).\n\nA dextran solution of 100 p.p.m. was prepared to analyze the rejection efficiency. Equation 2 was used to calculate the rejection value of the fabricated membrane.\n\n\n\nWhere R = rejection coefficient; Cp = permeate concentration; and Cf = concentration of feed.\n\nThe hydrophilicity of the surface of the membrane was measured using a water contact angle meter (Kyowa Kaiwenkagaku, Saitama, Japan, CA-A). The contact angle is the angle formed between the surface of the test material and the pure water dropped onto the surface of the membrane12. Each sample was measured 10 times, and the average value of the measurement was the value of the water contact angle of the membrane sample.\n\nTo study the effect of the blending of pluronic additives on the toughness of the PVC/ PF127 membrane, a membrane shrinkage test was performed. Three pieces of membrane at each PF127 concentration were slashed in wet conditions with a length and width of 5 and 1 cm, respectively. The membrane was dried in the oven for 24 hours at 80°C. The shrinkage of the membrane was calculated using Equation (3).\n\n\n\nWhere L0 = length of wet membrane (cm) and L1 = length of dried membrane (cm).\n\n\nResults and discussion\n\nThe results of the SEM analysis of the cross-sections of all the membranes are shown in Figure 3. The transverse part of the original PVC membrane has an asymmetric structure consisting of a thick, dense structure in the top layer and a finger-like structure in the center path of the cross-sectional area. The formation of a membrane structure fabricated using the wet inversion technique happens during coagulation, in which the DMAc solvent is leached out of the matrix of the dope solution and water as a non-solvent diffuses into the membrane. This phenomenon causes the formation of membrane pores and a finger-like macrovoid structure13. The structure of the PVC membrane changes after the addition of PF127 as the additive in forming the membrane pores9.\n\nAs shown in Figure 3a, the original PVC membrane has an upper layer that is thicker than the upper layer structure after the addition of PF127. The exchange of solvent from the polymer solution to the coagulation bath occur slowly in the case of the original PVC membrane, and contribute to the formation of a thick upper layer that is larger than in the other systems14. After the addition of PF127, the membrane surface becomes more hydrophilic and the affinity between the casting solution and water increases, so the polymer solution will attract more water and the diffusion process of water into the polymer matrix will be faster15. As a consequence of this mechanism, large macrovoids and a thin upper layer are formed. In other words, the increase in the PF127 concentration results in larger pores and a thinner upper layer of the membrane.\n\nThe measurement of the water contact angle is the simplest way to identify the degree of hydrophilicity and hydrophobicity of the membrane14. The hydrophilicity of the membrane, as measured by the water contact angle meter, is shown in Figure 4. The addition of PF127 is proven to improve the hydrophilicity of the membrane, as indicated by the decrease in the water contact angle. The existing PEO segment contained in the PF127 on the membrane surface contributed to the improvement in the membrane hydrophilicity. A number of studies on the mechanism of the decreased water contact angle of various membrane modifications with Pluronic have been reported by researchers16–19. The most hydrophilic membrane surface obtained in this study was found in the PVC membrane with the addition of 7% additive, with a contact angle of 67.2°.\n\nThe water permeability and rejection profile of the original PVC and PVC blend membrane are shown in Figure 5. The original PVC membrane has a water permeability of about 0.616 l/m2.h.atm. After the addition of 1 wt% of PF127 to the dope solution additive, the water permeability increases significantly. The PEO chain in PF127 increases the pore size of membrane. Therefore, the amount of water that passes through the membrane is higher than that of the membrane without the polymeric additive. The change in the bottom layer structure of the PVC blend membrane is also evidence of the increased water permeability (Figure 3). As reported by many authors, the addition of an appropriate amount of hydrophilic polymer to the dope solution might enhance the membrane pores7,14,20,21, and, consequently, high permeation would be obtained. In this work, the highest water permeability reached 45.618 l/m2.h.atm, which was obtained in the case of the blend membrane with the PF127 concentration of 7%.\n\nFigure 5 also shows the rejection efficiency of the dextran solution. The original PVC membrane is able to reject the dextran molecules by up to 100%. The addition of PF127 at a high concentration causes a decline in the rejection efficiency. The solution sample for the rejection experiment was prepared by dissolving a low molecular weight of dextran (i.e., 10.000 Da). This may be the reason for the reduction of rejection efficiency at high concentration of additive in this work. To achieve the best performance for permeation and selectivity, the optimization of the polymer solution could be designed by changing the relative composition of the PVC and the PF127.\n\nIn reference to the separation industry, membranes are expected to sustain in a wide range of temperature conditions. To determine the resistance of PVC/PF127 membranes in high-temperature conditions, a shrinkage test was performed by drying the membranes at 80°C; the results are shown in Table 1. The original PVC membrane did not suffer significant shrinkage after exposure to a temperature of 80°C, nor did the blending of PF127 into a polymer solution contribute seriously to the shrinkage of the PVC membrane. As shown in Table 1, an increase in the additive concentration of up to 7 wt% only has a small impact on the decrease in membrane size. PVC is one of the most widely used polymers in UF membranes, owing to its excellent physical and chemical properties. PVC polymer has a melting point of 212°C, and material fabricated from this polymer can only be degraded at high temperature22.\n\n\nConclusion\n\nThe fabrication of PVC membrane with PF127 as an additive has been performed in this work. The characteristics and performance of the membrane have been analyzed in terms of the morphology, hydrophilic or hydrophobic properties, water permeability, and solute rejection, as well as membrane shrinkage. Morphological analysis using SEM shows the increase in the membrane porosity after addition of PF127. A considerable change in the number and length of the macrovoid structure in the center section of the membrane increased the water permeability from 0.61 to 45.61 l/m2.hr.atm. Regarding the water contact angle observation, it is found that the hydrophilicity of the membrane improves as the proportion of PF127 is increased. On the basis of the results of the shrinkage test, it can be concluded that the PVC membrane obtained in this research is able to withstand extreme temperature conditions of up to 80°C. Regarding the experimental results, it can be concluded that PF127 succeeded in improving hydrophilic properties, filtration performance, and maintaining the stability of the membrane. Thus, the PVC-FP127 is useful to be applied in the water treatment industry.\n\n\nData availability\n\nDataset 1. Raw data for water permeability and solute rejection. DOI: 10.5256/f1000research.15077.d20640123.\n\nDataset 2. Raw data for water contact angle. DOI: 10.5256/f1000research.15077.d20640224.\n\nDataset 3. Raw data for membrane shrinkage test. DOI: 10.5256/f1000research.15077.d20640325.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declare that no grants were involved in supporting this work.\n\n\nReferences\n\nSaljoughi E, Mousavi SM: Preparation and characterization of novel polysulfone nanofiltration membranes for removal of cadmium from contaminated water. Sep Purif Technol. 2012; 90: 22–30. Publisher Full Text\n\nShockravi A, Vatanpour V, Najjar Z, et al.: A new high performance polyamide as an effective additive for modification of antifouling properties and morphology of asymmetric PES blend ultrafiltration membranes. Microporous Mesoporous Mater. 2017; 246: 24–36. Publisher Full Text\n\nMartín A, Arsuaga JM, Roldán N, et al.: Enhanced ultrafiltration PES membranes doped with mesostructured functionalized silica particles. Desalination. 2015; 357: 16–25. Publisher Full Text\n\nDiez B, Roldán N, Martín A, et al.: Fouling and biofouling resistance of metal-doped mesostructured silica/polyethersulfone ultrafiltration membranes. J Memb Sci. 2017; 526: 252–263. Publisher Full Text\n\nAryanti PTP, Yustiana R, Purnama RED, et al.: Performance and characterization of PEG400 modified PVC ultrafiltration membrane. Membr Water Treat. 2015; 6(5): 379–392. Publisher Full Text\n\nNasrul A, Bastian A, Sri M, et al.: Res J Chem Environ.\n\nArahman N, Mulyati S, Lubis MR, et al.: Modification of polyethersulfone hollow fiber membrane with different polymeric additives. Membr Water Treat. 2016; 7(4): 355–365. Publisher Full Text\n\nArahman N, Fahrina A, Amalia S, et al.: Effect of PVP on the characteristic of modified membranes made from waste PET bottles for humic acid removal [version 2; referees: 2 approved]. F1000Res. 2017; 6: 668. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRaslan R, Mohammad AW: Polysulfone/Pluronic F127 Blend Ultrafiltration Membranes: Preparation and Characterizations. J Appl Sci. 2010; 10(21): 2628–2632. Publisher Full Text\n\nLv C, Su Y, Wang Y, et al.: Enhanced permeation performance of cellulose acetate ultrafiltration membrane by incorporation of Pluronic F127. J Memb Sci. 2007; 294(1–2): 68–74. Publisher Full Text\n\nWang YQ, Su YL, Le Ma X, et al.: Pluronic polymers and polyethersulfone blend membranes with improved fouling-resistant ability and ultrafiltration performance. J Memb Sci. 2006; 283(1–2): 440–447. Publisher Full Text\n\nBahremand AH, Mousavi SM, Ahmadpour A, et al.: Biodegradable blend membranes of poly (butylene succinate)/cellulose acetate/dextran: Preparation, characterization and performance. Carbohydr Polym. 2017; 173: 497–507. PubMed Abstract | Publisher Full Text\n\nMavukkandy MO, Bilad MR, Kujawa J, et al.: On the effect of fumed silica particles on the structure, properties and application of PVDF membranes. Sep Purif Technol. 2017; 187: 365–373. Publisher Full Text\n\nMehrparvar A, Rahimpour A, Jahanshahi M: Modified ultrafiltration membranes for humic acid removal. J Taiwan Inst Chem Eng. 2013; 45(1): 275–282. Publisher Full Text\n\nLin J, Ye W, Zhong K, et al.: Enhancement of polyethersulfone (PES) membrane doped by monodisperse Stöber silica for water treatment. Chem Eng Process Process Intensif. 2016; 107: 194–205. Publisher Full Text\n\nLiu Y, Su Y, Zhao X, et al.: Improved antifouling properties of polyethersulfone membrane by blending the amphiphilic surface modifier with crosslinked hydrophobic segments. J Memb Sci. 2015; 486: 195–206. Publisher Full Text\n\nLoh CH, Wang R, Shi L, et al.: Fabrication of high performance polyethersulfone UF hollow fiber membranes using amphiphilic Pluronic block copolymers as pore-forming additives. J Memb Sci. 2011; 380(1–2): 114–123. Publisher Full Text\n\nBueno CZ, Dias AM, de Sousa HJ, et al.: Control of the properties of porous chitosan–alginate membranes through the addition of different proportions of Pluronic F68. \"Mater Sci Eng C Mater Biol Appl. 2014; 44: 117–125. PubMed Abstract | Publisher Full Text\n\nLoh CH, Wang R: Fabrication of PVDF hollow fiber membranes: Effects of low-concentration Pluronic and spinning conditions. J Memb Sci. 2014; 466: 130–141. Publisher Full Text\n\nJalali A, Shockravi A, Vatanpour V, et al.: Preparation and characterization of novel microporous ultrafiltration PES membranes using synthesized hydrophilic polysulfide-amide copolymer as an additive in the casting solution. Microporous Mesoporous Mater. 2016; 228: 1–13. Publisher Full Text\n\nEl-gendi A, Abdallah H, Amin A, et al.: Investigation of polyvinylchloride and cellulose acetate blend membranes for desalination. J Mol Struct. 2017; 1146: 14–22. Publisher Full Text\n\nXu J, Xu ZL: Poly(vinyl chloride) (PVC) hollow fiber ultrafiltration membranes prepared from PVC/additives/solvent. J Membr Sci. 2002; 208(1–2): 203–212. Publisher Full Text\n\nArahman N, Fahrina A, Wahab MY, et al.: Dataset 1 in: Morphology and performance of polyvinyl chloride membrane modified with Pluronic F127. F1000Research. 2018. Data Source\n\nArahman N, Fahrina A, Wahab MY, et al.: Dataset 2 in: Morphology and performance of polyvinyl chloride membrane modified with Pluronic F127. F1000Research. 2018. Data Source\n\nArahman N, Fahrina A, Wahab MY, et al.: Dataset 3 in: Morphology and performance of polyvinyl chloride membrane modified with Pluronic F127. F1000Research. 2018. Data Source"
}
|
[
{
"id": "34924",
"date": "21 Jun 2018",
"name": "Zulfan Adi Putra",
"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 is about adding PF127 as the additive of a PVC membrane. Membrane characteristics and performances are measured.\n\nIt is, however, not clear why the addition is only until 7%. Is there any reason for this? Looking at the results, better performance could be obtain for higher concentration of PF127. In this sense, we cannot conclude the way it is concluded now.\n\nThe same goes with the shrinkage test, where it is only tested at 80 degC. Hence, the respective conclusion is not correct. At 7%, the shrinkage is about 10%, is it not already significant shrinkage?\nFor the permeability test, it will be clearer if all various additives are put on the same graph, with the x-axis as the time and y-axis as the permeability. Then, we can see and compare the performance better.\n\nAll in all, the analysis of the data and the conclusion need to be revisited.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3789",
"date": "10 Jul 2018",
"name": "Nasrul Arahman",
"role": "Author Response",
"response": "Comment 1 It is, however, not clear why the addition is only until 7%. Is there any reason for this? Looking at the results, better performance could be obtain for higher concentration of PF127. In this sense, we cannot conclude the way it is concluded now. Response FIn this experiment, the best membrane is the membrane which has good both of permeability and selectivity. In other words, it has optimum filtration performance. As we know, permeability of membrane has opposite relation to its selectivity. Higher water permeability leads lower selectivity, therefore the optimum value of permeability and selectivity is the best performance of membrane. In this study, the addition of 7% PF127 produced highest water permeability, but it has the lowest rejection. In previous study, Lv (2007) investigated the addition of Pluronic F127 to CA membrane with the concentration 0 ; 4 ; 8 ; 12 ; 16 ; and 20 %. The result showed that the addition of PF127 higher than 8% enhance membrane porosity and decrease solute rejection. So that, in this study the use of PF127 is maximum 7%. The source of previous study: https://doi.org/10.1016/j.memsci.2007.02.011 Comment 2 The same goes with the shrinkage test, where it is only tested at 80 degC. Hence, the respective conclusion is not correct. At 7%, the shrinkage is about 10%, is it not already significant shrinkage? Response F Thank you for the comment. In this experiment, the shrinkage test is designed at 80 degC, and the shrinkage impact is not so significant. Comment 3 For the permeability test, it will be clearer if all various additives are put on the same graph, with the x-axis as the time and y-axis as the permeability. Then, we can see and compare the performance better. Response Fin this experimet, the all of permeability test were conducted in certain setting time. So, For that condition, the graph will be better if consist of the value of permeability and additives concentration. Comment 4 All in all, the analysis of the data and the conclusion need to be revisited. Response F the analysis and the conclussion in points has been re-written."
}
]
},
{
"id": "34922",
"date": "22 Jun 2018",
"name": "Heba Abdallah",
"expertise": [
"Reviewer Expertise Membrane technology",
"Production of polymeric membranes"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 studied the effect of addition of pluronic F127 on the PVC membrane morphology and performance.\n\nThe article needs Minor Revision.\n\nIntroduction\nThe authors should do a comparison between this work and other previous work in terms of different surfactant additions, rejection % and permeability.\n\nIn section membrane preparation\nI think this is wrong - thickness in micrometer (0.450 µm too thin) the wet thickness using most knifes or applicators can reach to 0.45 mm or 450 micrometer\n\nExplain where PF127 is added - is it added in polymeric solution or in coagulation bath?\n\nSEM section\nSEM section need more discussion, explain why the addition of 3% PF127 in fig 2b indicates the biggest finger like formation compared to 5 and 7%?\n\nMembrane hydrophilicity\nExplain how the PEO effects the membrane, with a reference?\n\nMembrane Shrinkage\n\nThe method of measuring shrinkage is not correct, it depends on the length of wet membrane, which means the blend membrane will absorb more water due to hydrophilicity, so I think that can affect the wet length of the membrane.\n\nSee this reference and repeat your shrinkage experiment:\n\nBilad M et al, 2015 1\n\nConclusion\nRe-write the conclusion in points to indicate your results clearly.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "3788",
"date": "10 Jul 2018",
"name": "Nasrul Arahman",
"role": "Author Response",
"response": "Heba Abdallah, Chemical Engineering and Pilot Plant Department, National Research Centre, Giza, Egypt Comment 1 The authors should do a comparison between this work and other previous work in terms of different surfactant additions, rejection % and permeability. Response F The comparison between previous works has been added to Introduction at third paragraph. The manuscript is written as follows: The hydrophilic polymers that are frequently used as an additive are polyvinylpyrrolidone, polyethylene glycol, Brij, Tetronic, and Pluronic 7, 8 . Of these, Pluronic is used as a surface modifying agent for many hydrophobic polymers. Raslan and Mohammad 9 added Pluronic F127 to a polysulfone membrane. The resulting membrane is resistant to fouling and possesses good pore distribution 9 . The highest flux reached up to around 800 L/m2.h at 4,5 bar with 4,8 g addition of PF127. Pluronic has also been used to improve the anti-fouling of cellulose acetate (CA) membranes 10 . The combination of CA polymer and Pluronic surfactant results in a membrane that is more resistant to fouling and has a more stable filtration profile. Hydrophilic surfactant Tween20 and Tween80 has been used to enhance the permeation and antifouling properties of PVC membrane. The surfactants were added as 0; 1; 3; 5; and 7 % to dope membrane. The result showed that higher concentration of the both surfactant resulting higher water flux. The highest water flux is 328,6 L/m2.h with 7% Tween20 addition. Contrary to the rejection performance of BSA, the result showed decreasing rejection value in higher surfactant addition. The BSA rejection by PVC original membrane is about 97%, after addition of surfactant additives the rejection value decreased up to around 86-87,5% at 7% Tween20 and Tween80. Comment 2 I think this is wrong - thickness in micrometer (0.450 µm too thin) the wet thickness using most knifes or applicators can reach to 0.45 mm or 450 micrometer F There was a typo and it has been revised in membrane preparation section as follows: The solution was then framed on a glass plate using an automatically adjustable applicator (YBA-3, Yoshimitsu, Japan) at a thickness of 450 µm. Comment 3 Explain where PF127 is added - is it added in polymeric solution or in coagulation bath? Response Fwe have rewrite the script in membrane preparation section To improve the performance of the membrane, PF127 was added at concentrations of 1, 3, 5 and 7 wt% to polymeric solutions and dissolve until homogene. Comment 4 SEM section need more discussion, explain why the addition of 3% PF127 in fig 2b indicates the biggest finger like formation compared to 5 and 7%? Response F The more discussion of finger like formation has been added in the second paragraph of membrane morphology discussion. The finger like structure forms at PVC membrane with 3wt% of PF127 is dominate a half of the crossection area of the membrane. While, in case of the addition of 5 and 7wt% of PF127, the finger-like structure formed in all cross-section area of the membrane Comment 5 Explain how the PEO effects the membrane, with a reference? Response F the more discussion of PEO effects has been written in membrane hydrophilicity section. the hydrophilic nature of PEO chain in Pluronic increased the diffusion rate of non-solvent during membrane formation. The increasing of nonsolvent in-diffusion rate promoted instantaneous demixing, which enhanced macrovoid formation. It has been reported that a rapid precipitation caused by the hydrophilicity of the additive results in higher surface porosity and more porous sublayer, leading to a higher water permeation. C.H. Loh et al. / Journal of Membrane Science 380 (2011) 114–123 Comment 6 The method of measuring shrinkage is not correct, it depends on the length of wet membrane, which means the blend membrane will absorb more water due to hydrophilicity, so I think that can affect the wet length of the membrane. See this reference and repeat your shrinkage experiment: Bilad M et al, 2015 1 Response F Thank you for your correction, we agree with your idea that the length of wet membrane has an effect on membrane shrinkage. We, therefore, calculated the membrane shrinkage by measuring the length of the wet cast and the dried membrane. Research group of Bilad (Bilad et al. 2015) designed the shrinkage calculation by measuring of the width of wet cast and the dried membrane. We revised the manuscript as follow: To study the effect of the blending of pluronic additives on the toughness of the PVC/ PF127 membrane, a membrane shrinkage test was performed. The wet cast of membrane at each PF127 concentration was dried in the oven for 24 hours at 80°C. The shrinkage of the membrane was calculated using Equation (3). Membrane shrink (%) = ((L0-L1)/(L0)) x 100 Where L 0 = length of wet membrane (cm) and L 1 = length of dried membrane (cm). Comment 7 Re-write the conclusion in points to indicate your results clearly. Response F the conclussion in points has been re-written. The fabrication of PVC membrane with PF127 as an additive has been performed in this work. The characteristics and performance of the membrane have been analyzed in terms of the morphology, hydrophilic or hydrophobic properties, water permeability, and solute rejection, as well as membrane shrinkage. From this recent study, it can be concluded as follows: Morphological analysis using SEM shows the increasing the membrane porosity after addition of PF127. The water permeability of PVC membrane increases from 0.61 to 45.61 l/m2.hr.atm after addition of 7% PF127. However, the optimum filtration result, water permeability and rejection are found on the membrane with 3% addition PF127. It reach 20,2 L/m2.h for permeability and 68,66% for solute rejection. The number and length of the macrovoid structure in the center section of the membrane increased and change after the presence of PF127 Regarding the water contact angle observation, it is found that the hydrophilicity of the membrane improves as the proportion of PF127 is increased. On the basis of the results of the shrinkage test, it can be concluded that the PVC membrane obtained in this research is able to withstand extreme temperature conditions of up to 80°C. Regarding the experimental results, it can be concluded that PF127 succeeded in improving hydrophilic properties, filtration performance, and maintaining the stability of the membrane. Thus, the PVC-FP127 is useful to be applied in the water treatment industry."
}
]
}
] | 1
|
https://f1000research.com/articles/7-726
|
https://f1000research.com/articles/7-1033/v1
|
09 Jul 18
|
{
"type": "Research Article",
"title": "A tiling array-based comparative genomic hybridization approach to predict copy number variations between Plasmodium falciparum field isolates from the Indian Sub-continent",
"authors": [
"Isha Pandey",
"Ramandeep Kaur",
"Amit Kumar Subudhi",
"P.A Boopathi",
"Raja C. Mugasimangalam",
"Sudha N. Rao",
"Mohammed Aiyaz",
"Sanjay Kochar",
"Dhanpat Kochar",
"Ashis Das",
"Isha Pandey",
"Ramandeep Kaur",
"Amit Kumar Subudhi",
"P.A Boopathi",
"Raja C. Mugasimangalam",
"Sudha N. Rao",
"Mohammed Aiyaz",
"Sanjay Kochar",
"Dhanpat Kochar"
],
"abstract": "Background: There are several techniques to analyse copy number variation in both research and clinical settings, such as whole genome amplification (sWGA), SNP arrays and one of the most commonly used techniques, array based comparative genomic hybridization (aCGH). In the latter, copy number comparison is obtained between differentially labelled target and reference DNAs by measuring ratio of fluorescence intensity of probes indicating loss or gain in the chromosomal region. Methods: Here we carry out a comparative analysis between two Plasmodium falciparum parasite isolates (Pf-isolate-2 and Pf-isolate-1) causing malaria using array CGH. The array contains approximately 418,577, 60mer custom-designed probes with an average probe spacing of 56 bp. The significant major variations (amplifications and deletions) copy number variations (CNV) in Pf-isolate-2 (Pf-2) in comparison with Pf-isolate-1 (Pf-1), are reported. Results: CNVs have been seen in all the chromosomes in Pf-2, most of the deletions have been seen mostly in sub-telomeric and telomeric regions of the chromosomes that comprises of variant surface antigen family genes. Apart from the subtelomeric regions other parts of the chromosomes have also shown CNVs. Novel variations , like continuous amplification of 28kb region (249817-278491) of chromosome-8, which covers for 3 genes two of which codes for conserved Plasmodium proteins with unknown function (MAL8P1.139, PF08_0122) and tRNA pseudouridine synthase, putative (PF08_0123). Amplifications in regions harboring genes like GTP cyclohydrolase I (GCH-1, PFL1155W) and ribosomal protein, L24, putative (PFL1150C) of chromosome 12 were seen. Conclusion: Other than known variations reported earlier, some novel variations have also been seen in the chromosomes of Pf-2. This is an experimental case study reporting major amplifications and deletions in Pf-isolate-2 in comparison with Pf-isolate-1 using a tiling array based comparative genomic hybridization approach.",
"keywords": [
"Plasmodium falciparum",
"Clinical isolates",
"Indian Subcontinent",
"tiling array",
"array CGH"
],
"content": "Introduction\n\nPlasmodium falciparum is the most virulent among the 6 Plasmodium species that causes malaria in humans. Current efforts to eliminate the disease have not been successful because of the rapid genetic adaptation and natural selection of the parasite. The currently known causes for genome sequence variations in the parasite are single nucleotide polymorphism (SNP), insertions, deletions of short sequences, major amplifications, translocations, inversions, and allelic variations (Cheeseman et al., 2009). Apart from the host factors these parasitic genetic variations may also lead to the disease severity., P.falciparum strains have long been established in continuous culture systems and most genomic analysis has naturally utilized the incredible benefit from such techniques. However, analysis of genome variations from culture adapted parasite lines remains intrinsically very different from analysis of in-vivo derived parasite nucleic acid material and the latter could have a major impact on understanding of the parasite biology (LeRoux et al., 2009). Environmental pressures have been known to introduce changes in genomes providing organisms an evolutionary advantage for adaptation and CNVs could be one such change that assists the parasite adaptation to the new environment. These CNVs could enable survival of the parasite under drug pressure. An amplification event in the GTP cyclohydrolase has been frequently reported and the high copy number of GCH-1 has been shown to confer resistance to anti-folates (Heinberg et al., 2013; Jiang et al., 2008; Kidgell et al., 2006; Nair et al., 2008; Ribacke et al., 2007). The major known parasite molecule that influences the severity of the disease are encoded by Variant Surface Antigen family genes that are located in the sub-telomeric part of the chromosome. P. falciparum 3D7 genome sequencing revealed the presence of ~60 var genes followed by other variant surface antigen families, rifin/stevor (~190), Pfmc-2TM (~12) and surfin (~10) (Frech & Chen, 2013; Scherf et al., 2008). Most studied var genes encodes for PfEMP-1 (Erythrocyte membrane protein-1) that play an important role in the pathogenesis by periodic switching and immune evasion. Along with the exported family members, these genes play an important role in cytoadhesion by binding to multiple cell receptors, including ICAM-1, CD36, E-selectin, NCAM and CD31 on micro vascular endothelia, which mediates sequestration that in-turn, leads to the severity of the disease. Recently another receptor, Endothelial protein C receptor (EPCR), a brain specific receptor has been identified as a binding partner of PfEMP-1 (Smith et al., 2013; Wassmer & Gray, 2017). Var genes are divided into 3 major groups A, B and C and 2 intermediate groups UpsB/A and B/C (Lavstsen et al., 2003). The higher transcript level of var group B and var group C family genes have been reported in severe malaria cases in comparison with the uncomplicated malaria cases from India (Subudhi et al., 2015).\n\nThere is limited information regarding complete nuclear genome variation for any of the human malaria parasites from field isolates from the Indian subcontinent, however, Tyagi et al. in 2014 have reported evolutionary history of Plasmodium falciparum by sequence analysis of mitochondrial genome of Indian isolates. Most studies have focused on analysis of sequence variations in single genes from different parts of India (Kumar et al., 2005; Kumar et al., 2012; Mishra et al., 2015; Rajesh et al., 2008; Zeeshan et al., 2012). These have focused on SNP's for \"drug resistance\" marker genes of Indian isolates. Garg et al., 2009; Pathak et al., 2014; Sharma et al., 2015).\n\nGenomics technologies like whole genome microarrays and more recently next-generation sequencing have enabled deeper analysis of various types of genomic variations. Using a reference clone/isolate, copy number variations of other isolates can be identified (Carret et al., 2005). An interesting study by Oyola et al. in 2016 details of SNP analysis from dried blood spots using Illumina based sequencing approaches.\n\nArray-based technologies have been used to distinguish CNV's in different diseases (Coughlin et al., 2012). Such techniques have been extensively used in cancer research to identify DNA copy number variability to represent structural variations at higher resolution and increased sensitivity (Liu, 2007; Mockler et al., 2005; Przybytkowski et al., 2011).\n\nThe objective of this study was to see whether we could at all distinguish differences between the two parasite isolates taken from individual patients showing differing disease severity using aCGH tilling array approach.\n\nThe microarray data discussed in this manuscript has been deposited in the NCBI Gene Expression Omnibus (GEO) under the GEO series accession number (GSE77165).\n\n\nMethods\n\nVenous blood samples were collected (5ml) from P. falciparum infected patients admitted to S.P. Medical College, Bikaner, India. Patient’s samples were collected on informed consent according to hospital guidelines. Sample collection was approved by the hospital ethical committee (No.F.(Acad) SPMC/2003/2395). Infection with P. falciparum was confirmed by microscopy and rapid diagnostic tests (OptiMal test; Diamed AG, cressier sur Morat, Switzerland, Falcivax test; Zephyr Biomedical System Goa, India) in the hospital. Blood was immediately (within 15 mins) subjected to density gradient-based separation (Histopaque 1077, Sigma Aldrich, USA) to separate the peripheral blood mononuclear cells (PBMCs) from the infected blood samples following manufactures instruction. Both the fractions were subsequently washed twice with phosphate buffered saline (PBS) and lysed using 4 volumes of TRI-reagent and stored at -80°C. Samples were transported in cold chain to BITS, processed and assessed by 18s r RNA gene based multiplex PCR to eliminate any possibility of P. vivax co-infection. Details of the PCR primers used for multiplex PCR for the detection of P. falciparum and P. vivax are presented in Table 1 as cited in the research articles (Pakalapati et al., 2013). PCR conditions used for the amplifications of 18S r RNA gene involved initial denaturaton at 93°C for 2 minutes followed by denaturation at 93°C for 1 minute 30 seconds, anneal-ing at 52°C for 2 minutes, extension at 72°C for 3 minutes, after initial denaturation all the reactions were run for 30 cycles. The TRI protocol was used to allow investigations of both RNA and DNA from the precious clinical samples. (Chomczynski, 1993) Criteria for determining complicated cases were based on World Health Organization guidelines (World malaria report, 2010). For CGH array hybridization, isolates from two patients were considered a) cerebral malaria (CM) whose Glasgow coma scale = 7/14 was considered as a complicated case (PFC) Pf-2 or test, b) uncomplicated case (PFU) Pf- 1.\n\nTotal DNA was isolated from samples using back extraction buffer following manufacture’s protocol (TRI- Reagent, Sigma Aldrich, and USA). Both the DNA samples (Pf-isolate-2 and Pf-isolate-1) were run on an agarose gel and found to be intact. Total DNA purity and concentration were assessed using Nanodrop ND-1000 UV-Visible spectrophotometer (Nanodrop technologies, Rockland, USA). The A260/A280 ratio suggests the absence of proteins and RNA.\n\n1.5 Micrograms of each DNA sample were restriction digested using Alu1 and Rsa1. This restricted DNA was random-primer labelled with Cyanine-3 (Cy3) and Cyanine-5-dUTP (Cy5) using Agilent Genomic DNA labeling kit (Part Number: 5190-0453). The labelled DNA was concentrated and quality assessed for yields and specific activity. Pf-2 which is a test sample was labelled with Cy3 and Pf-1 sample taken as a control was labelled with Cy5. 5 micrograms of the labelled samples were hybridized on a Genotypic designed P. falciparum 3D7 Custom 2x400K CGH Microarray, (Agilent Microarray Design ID AMADID: 025327). Hybridization of probes was carried out in Agilent Sure hyb chambers at 65°C with rotation for 40 hours. The hybridized slides were washed using Agilent aCGH wash buffer (Part No: 5188-5221/22) and scanned using the Agilent Microarray Scanner G2505C. The Scanned image was processed using Agilent Feature extraction software version 10.7 and probe intensities were normalized using the linear dye normalization method.\n\nA high resolution custom CGH array was designed with 418,577 probes, representing the entire P.falciparum 3D7 reference genome using the e-Array tool SureDesign v 5.0.1 by the Agilent certified microarray facility of Genotypic Technology, Bengaluru, India. Whole genome sequence for P. falciparum 3D7 strain was downloaded from NCBI genome database; the 60mer probes were designed based on optimal GC% and melting temperature using Probe parser V.1.0.exe (Perl program developed by Genotypic Technology, Bangalore, India) Genotypic Probe parser is the perl code developed by Genotypic Technology to design 60mer Probes from the downloaded genome sequence and calculate the GC% and melting temperature (Tm) for the probes. The 60 mer probes designed by the above mentioned program were then subjected to the Agilent eArray tool and the CGH array designed. The probes were designed by tiling over the DNA sequence. The tiling resolution was 56 bp. The probe length has been set as 60 bp which is the optimised length for Agilent comparative genomic hybridization workflow. The probes were designed by calculating GC% and optimum Tm values of (>65 genome sequence and calculate the GC% and melting temperature. The sequence information, including chromosomal location and direction of transcriptional regulation of P. falciparum 3D7 were retrieved from Plasmodium Genome Resource. [PlasmoDB.V.6.0] (Aurrecoechea et al., 2009; Gardner et al., 2002).\n\nImage analysis and data normalization was done by applying the linear dye normalization method using the Agilent Feature Extraction Software version 10.7. The Raw data were analysed using the Agilent Genomic Workbench Software version 6.5. A well-accepted ADM-2 algorithm was used (Aberration Detection Method II algorithm) developed by Agilent and included in the Agilent Genomic Workbench for CGH analysis. ADM-2 is a statistical algorithm which computes copy number differences between the sample and reference using an iterative procedure to identify all genomic regions from its expected value of 0 larger than a given threshold (In this case we have taken a minimum of 5 consecutive probes, with an average log ratio of 0.3). At each iteration, the region with the most significant score is reported. ADM- 2 statistical score is computed as the average normalized Cy3/Cy5 log2 ratios of all probes in the genomic interval multiplied by the square root of the number of these probes. Deletions are reported where the signal intensities of the probes representing the region in the test strain are near background and represents the absence of hybridization. Graphical visualization of the data for the significant aberrations was done using UCSC genome browser (Schneider et al., 2006).\n\n\nResults\n\nIn this study, a custom designed tiling microarray has been used to investigate the overall difference between the two clinical isolates i.e., Pf-isolate-2 (Pf-2) (complicated/cerebral malaria) and Pf-isolate-1 (Pf-1) (uncomplicated malaria). The reference genome 3D7 was used to design the probes. The tiling array used in this study enables large-scale detection of regions showing copy number variations across the whole genome. The CNV's were identified in Pf-2 (complicated) by comparison to Pf-1 (uncomplicated malaria). The well-accepted ADM-2 algorithm was used with filtering criteria of, minimum of 5 probes and the average log fold change of 0.3 for the analysis of CNV. Here we are reporting significant CNV regions seen in Pf-2 in comparison with the Pf-1. In this data, the term amplification and deletion are used to represent copy number of test strain Pf-2 compared to Pf-1. Deletions are reported where the signal intensities of the probes representing the region in the test strain are near background and represent the absence of hybridization.\n\nThe size of each chromosome with the number of genes reported here are as per PlasmoDB V.6.0 (Aurrecoechea et al., 2009; Gardner et al., 2002). Those regions that are re-annotated in the genome assembly of P. falciparum 3D7 and updated in recent version of PlasmoDB i.e PlasmoDB V.34 and are eliminated from the study to rule out any inadvertent errors due to the re-annotation processes. A circos plot of whole genome CNV data of Pf-Isolate-2 has been generated to show the regions in chromosomes exhibiting CNV (Figure 1).\n\nCircle of Genome variation of Plasmodium falciparum (Pf-2), representing probe based chromosome-wise Copy number variations. The outer red bars are showing amplifications i.e. regions with Cy3/ Cy5 ratio of ≥ 0.3 and green inner bars are showing deletions present in the region with Cy3/Cy5 ratio ≤ -0.3.\n\nA total of 687 unique genes was seen to exhibit CNVs in the genome of Pf-isolate-2 (Dataset 1, (Pandey et al., 2018a)). The loss/gains varied in size from 100 nucleotides to 47,000 bp and included intra and intergenic regions, as they are an essential part of the genome and might be involved in the regulation of the genes. From the total of 687 genes, amplification is reported in 511 and deletions in 176 genes. There are 27 genes in regions of chr-4,6,7,8,10,11,12,13 in Pf-2 in which both amplifications/gains and deletions/loss are present. This region is rich in Variant Surface Antigen family genes. The deletions or amplifications covered smaller regions and did not cover more than 2 contiguous genes. Most variations were seen in the sub-telomeric and telomeric regions.\n\nOur analysis revealed CNVs in 268 unique VSA gene families, (Dataset 2, (Pandey et al., 2018b)) out of which variation is present in 58 unique var genes in Pf-2 in comparison to Pf-1. Most of these genes showing CNVs belong to var group B and C. A previous report from our group has also reported differential expression in the transcriptome of these family genes in the isolates causing severe malaria. The longest CNV was reported in chr2: 86, 3297: 94, 7019, near the telomere covering about 20 genes. This region largely had deletions and covered genes belonging to the: var gene family (PfEMP-1) rifin, stevor, and exported family members Plasmodium exported protein (EPF-1, EPF-3, EPF-4) Pfmc-2TM Maurer's cleft two transmembrane protein (MC-2TM) and hypothetical proteins. Regions with multigene family genes showed deletions at both the telomeric and centromeric regions of the chromosomes. In chromosome-1 of Pf-isolate-2 a deletion of 16.3kb (563139-579488) region was seen, that includes 5 genes coding for Stevor, pseudogene (PFA0705C) rifin (RIF, PFA0710C) Plasmodium exported protein (hyp7), unknown function (PFA0715C) Plasmodium exported protein, unknown function, pseudogene (PFA0720W). There is also a 11.7 kb, deletion region in the same chromosome, which includes 4 genes coding for, Exported protein Family-3 (EPF-3, PFA0685C) Exported Protein Family-4 (EPF-4, PFA0690W) Erythrocyte Membrane Protein1 exon2, pseudogene (VAR) (PfEMP-1, VAR, Pseudogene, PFA0695C), and Plasmodium exported protein (hyp10) of unknown function (hyp10, PFA0700C). In previous reports SNPs have been detected in PFA0700C Plasmodium exported protein (hyp10) in malaria parasites from Guinean and Gambian populations (Mobegi et al., 2014). A study by Mok et al. in 2008 have shown a segmental duplication of PFA0675W, PFA0685C, PF0690W and PFA0700C in different strains, and another study from the same group have also reported the duplication of PFA0685C, PF0690W and PFA0695C located at the right end of the chromosome of 3D7 in a clinical isolate (UAM25) (Mok et al., 2008; Ribacke et al., 2007). In our clinical isolate showing complicated manifestation, these segments are showing deletion in the chromosome. A deletion of 11.9 kb region in chromosome-3 that includes conserved Plasmodium proteins with unknown functions (PFC0325C, PFC0330W, PFC0335C). It is interesting to note that although these genes are of unknown function their annotated GO functions suggest that these genes are involved in zinc ion binding activity, receptor activity and in ATP binding activity. A deletion of 14.8 kb, is present in chromosome number 11 and the region includes 6 genes rifin (RIF, PF11_0010), rifin (RIF, PF11_0011), RESA-like protein, pseudogene (RESA pseudogene, PF11_0012) Plasmodium exported protein (PHISTa), unknown function, pseudogene (PHISTa, pseudogene, PF11_0013), Pfmc-2TM Maurer's cleft two transmemberane protein (MC-2TM, PF11_0014) and RESA-like protein, pseudogene, (RESA, PF11_0015). This deletion region is present at the left end of the chromosome. A size variation of 13.2 kb has been seen in chromosome-13 from a starting position of 2828057-2841275 and covers 5 genes rifin (RIF, MAL13P1.495), rifin (RIF, MAL13P1.500), Stevor (Stevor, MAL13P1.505) erythrocyte membrane protein, exon-2, pseudogene (VAR) (PfEMP1, MAL13P1.510) rifin (RIF, MAL13P1.515). Other than larger deletions in chromosomes that cover for VSA, we have seen an amplification of size 13.2 kb present at the right end of chromosome-11 (2006705 - 201923) which covers 5 genes which encodes for rifin (RIF, PF11_0515), Stevor (PF11_0516), rifin (RIF, PF11_0517), rifin (RIF) pseudogene (RIF, PF11_0518) and rifin (RIF, PF11_0519). Details of the mentioned regions, and other regions covering VSA, are shown in Dataset 3 (Pandey et al., 2018c).\n\nOn analysis of the CNV data we have detected variations in all the 14 chromosomes; apart from the sub-telomeric regions of the chromosomes, we have also seen variations in other locations on the chromosomes, including intergenic regions.\n\nOther than the regions belonging to multigene family genes, which are mostly showing deletions in the chromosome, there are regions in which amplifications are present and represents genes that encodes for transcriptional co-activator ADA2 (ADA2), chromodomain-helicase-DNA-binding protein 1 homolog, putative (CHD1), polyubiquitin (PfpUB), non-SERCA-type Ca2+ -transporting P-ATPase (ATP4), etc. Major amplifications are present in chromosome-8, 10, and 12. An amplification of 28.6 kb is present in chromosome-8 (249817-278491) (Figure 2), near the centromeric location of the chromosome and covers 3 genes conserved Plasmodium protein, unknown function (MAL8P1139), conserved Plasmodium protein, unknown function (PF08_0122) tRNA pseudouridine synthase, putative (PF08_0123). In chromosome 10, A 7.4 kb and 5.6 kb amplification is present in genes that code for chromodomain helicase DNA binding protein 1 homolog, putative (CHD1, PF10_0232) and transcriptional co-activator ADA-2, (ADA-2, PF10_0143) Another region from 650889-653243 with an amplification of 2.3 kb cover for Glycophorin Binding Protein (GBP, PF10_0159). An amplification of 15.90 kb is present in close proximity to the centromeric region of chromosome 12 (Figure 3), which covers 6 genes that codes for conserved Plasmodium protein with unknown function (PFL1160C), Mitochondrial carrier protein, putative (PFL1145W) conserved plasmodium protein, unknown function (PFL1135C) GTP cyclohydrolase-I GCH1, (GCH-1, PFL1155W) ribosomal protein L24, putative (PFL1150C) and 50S integral memberane protein, putative (PFL1140W) An amplification of the region, that covers three genes, PFL1155W, PFL1150C, and PFL1145W have been seen previously in Dd2 strain of Plasmodium falciparum (Dharia et al., 2009). One of the genes showing amplification codes for the GTP cyclohydrolase enzyme, which is the 1st and the rate-limiting enzyme of the folate biosynthesis pathway. Amplification of this gene has been previously reported in parasite isolates from certain endemic countries (Kidgell et al., 2006; Nair et al., 2008). Another gene, mitochondrial carrier protein, putative (PFL1145W) is reported as a homologue of Yhm2 which was characterized from S. cerevisiae as an inner membrane DNA binding protein. It is speculated that this protein helps in mitochondrial maintenance by tethering mitochondrial DNA to inner membranes (van Dooren et al., 2006). Another region in chromosome-12 showing amplification covers 2 genes codes for Polyubiquitin (PfpUB, PFL0585W) and non-SERCA-type Ca transporting P-ATPase (ATP-4, PFL0590C).\n\nA continuous amplification of 28 kb region (249817-278491) of chromosome-8, the X-axis represents the baseline intensity of the probes, the Y-axis represents the hybridized probe intensity, red bars are showing amplifications present in the region and green bars are showing deletions present in the region. The genes present in the region, are near centromeric location on the chromosome-8 that covers for 3 genes (MAL8P1.139, PF08_0122) conserved Plasmodium membrane protein, unknown function, (PF08_0123) U2 snRNA/tRNA pseudouridine synthase, putative. The white double-headed arrows in single exons are representing the 5the 5exons are representing the 5epresents the the arrows in between the exons are representing introns and strand in which genes are present (Right-hand side arrow forward strand, left-hand side arrow representing reverse strand). The area without any bars are representing no variations.\n\nAmplification of 15.90kb is present near close proximity to the centromere region of chromosome 12 (Figure 3), which covers 6 genes, namely (PFL1160c), conserved Plasmodium protein, unknown function, (PFL1145w) mitochondrial carrier protein, putative, (PFL1135c) conserved Plasmodium protein, unknown function, (PFL1155w) GTP cyclohydrolase I (GCH1), (PFL1150c) ribosomal protein L24, putative, (PFL1140w). The X-axis represents the baseline intensity of the probes, the Y-axis represents the hybridized probe intensity, red bars are representing probe intensity as amplifications present in the region and green bars are showing deletions present in the region. Those without any bar are showing no variations.\n\nAn amplification of 3.3 kb is present in chromosome -13 (408857- 412219) in the single gene that codes for mRNA-decapping enzyme 2 putative (DCP2, PF13_0048), Amplification of 3.19 present on chromosome-14 covering a single gene coding for pyridine nucelotide transhydrogeanse, putative (PF14_0508) and 3.13 kb which is also present in chromosome-14 covering a single gene EMP-1 trafficking protein (PTP3, PF14_0758). It has been reported that disruption of the gene (PF14_0758) showed reduced levels of expression of PfEMP-1, which suggests that the protein encoded by PF14_0758 plays a role in trafficking and the display of the virulence protein PfEMP-1 on the host erythrocyte (Maier et al., 2008). All these genes showing amplifications in the data are important for the parasite to survive in the host environment.\n\n\nDiscussion\n\nSeveral studies at the genomic and transcriptomic level have been undertaken to understand the molecular basis of pathogenesis involved in severe disease manifestation. However, there is scarcity of information on genomic variation in parasites causing complicated manifestations, regions with seasonal malarial transmission. Besides, most of the CNV or SNP studies have been done on in-vitro parasite isolates or clinical isolates from other parts of the world, but very few studies are from parts of India (Cheeseman et al., 2009; Cheeseman et al., 2016). The lacunae in genome variation study still exists as there is no study from parts of India showing nuclear genomic variations between two parasites isolates with differing disease manifestations. CNVs were seen to be present throughout the chromosomes of Pf-2. Large amplifications and deletions of several kilo bases in regions covering Pf-EMP1 and exporter family member genes, have been seen, as it is a known fact that the variant surface antigen family genes are prone to variations and immune invasions. (Carret et al., 2005; Mok et al., 2008; Tan et al., 2009). Amplification in chromosome number 12 includes genes like GCH-1 which is a rate limiting enzyme of folate biosynthesis and is important for the parasite to survive, amplification of this gene is reported to cause antifolate resistance and drives the evolution of drug resistance (Kidgell et al., 2006; Nair et al., 2008). Amplification in Glycophorin binding protein (GBP, PF10_0159) was also seen in the PF-2, this gene is important for the parasite to invade the erythrocyte as it binds to the Glycophorin protein on the erythrocyte membrane. Amplification in sodium-dependent phosphate transporter (PiT, MAL13P1.206) has also seen, the uptake of Pi is essential for the parasite for the synthesis of DNA, RNA and numerous phosphorylated metabolic intermediates (Saliba et al., 2006).\n\n\nConclusion\n\nThe implications of deletions or amplifications in the coding regions of the genes remain unanswerable at this point. This study shows the presence of CNV based genome differences from two field isolates. Based on msp-1 and msp-2 genotyping it appears that both the isolates exhibit allelic families representing K1 (msp-1 block-2) and FC27 (msp-2 repeats) (Snounou et al., 1999). At this level of the study, it is not possible to speculate whether the differences that we are seeing in Pf-isolate-2 are instrumental in causing the severity of the disease. However, we are reporting here some putative and novel variations, which have not been reported before in single gene studies with field isolates. The Data clearly shows that there is a difference in the isolate which has been taken from the patient with severe malaria (Pf-2) than from that with uncomplicated malaria (Pf-1). Differences exist not only in coding regions, but also in the intergenic locations. Thus, there could be implications at the level of gene regulation. However, these differences could also be inherently natural differences between the two parasite isolates without any direct correlation with the disease states, which could be influenced by various other host factors. This case study is just a preliminary attempt to utilize aCGH tiling array using ADM-2 algorithm to report the variations if any or differences in the genomic segments of parasite isolate causing severe malaria with that of the isolate causing uncomplicated malaria. To get a real picture of the genomic variations related to the disease severity, a tiling array approach can be used to analyze a large set of clinical samples exhibiting different disease complications.\n\n\nData availability\n\nThe microarray data discussed in this manuscript has been deposited in the NCBI Gene Expression Omnibus (GEO) under the GEO series accession number (GSE77165).\n\nDataset 1: Total set of Genes showing Copy Number Variation in Pf-isolate- 2 in comparison with the Pf-isolate- 1. Includes data of total set of genes showing CNV, along with the Information about the Chromosome Number, start of the probe, end of the probe, size of the interval, number of probes in the interval, Amplification call, Deletion call, Gene symbols (PlasmoDB V.6.0), Updated gene symbol (PlasmoDB V.26), Description of gene symbol (PlasmoDB V. 34) log value of the regions, P-value of the region. 10.5256/f1000research.14599.d209655 (Pandey et al., 2018a)\n\nDataset 2: Copy Number Variations present in Variant Surface Antigen family genes of Pf-isolate-2. Includes data of the total set of VSA gene family genes that are showing variations. Along with the Information about Chromosome Number, start of the probe, end of the probe, the size of the interval, number of probes in the interval, Amplification call, Deletion call, Gene symbols (PlasmoDB V.6.0), Updated gene symbol (PlasmoDB v. 26), Description of gene symbol (PlasmoDB V. 28) log value of the regions, P-value of the region. 10.5256/f1000research.14599.d209656 (Pandey et al., 2018b)\n\nDataset 3: Genes from Variant Surface Antigen family genes that are showing more than 2 kb variations. Includes data from genes of regions covering for VSA family genes that are showing more than 2 kb variation along with the Information about Chromosome Number, start of the probe, end of the probe, size of the interval, number of probes in the interval, Amplification call, Deletion call, Gene symbols (PlasmoDB V.6.0), Updated gene symbol (PlasmoDB v. 26), Description of gene symbol (PlasmoDB V.34) log value of the regions, p-value of the region. 10.5256/f1000research.14599.d209657 (Pandey et al., 2018c)",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by a grant from the Department of Biotechnology (DBT), of the Ministry of Science and Technology, New Delhi, India [BT/PR7520/BRB/ 10/481/2006] to A.K.D., S.K.K. and D.K.K.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nWe thank all the patients and technical workers for their participation and support of this project. I.P. & R.K. acknowledge the Basic Scientific Research fellowship from University Grant Commission, New Delhi, India. A.K.S. acknowledges Senior Research Fellowship from the Council of Scientific and Industrial Research (CSIR), New Delhi, India and Project Assistantship from Department of Biotechnology (DBT), New Delhi, India. P.A.B. also acknowledges earlier Project Assistantship from Department of Biotechnology (DBT), New Delhi, India. A.K.D., S.K.K. and D.K.K. acknowledge and Birla Institute of Technology and Science, Pilani, India and S.P. Medical College, Bikaner, India for providing the required infrastructural facilities during this study.\n\n\nReferences\n\nAurrecoechea C, Brestelli J, Brunk BP, et al.: PlasmoDB: a functional genomic database for malaria parasites. Nucleic Acids Res. 2009; 37(Database issue): D539–D543. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarret CK, Horrocks P, Konfortov B, et al.: Microarray-based comparative genomic analyses of the human malaria parasite Plasmodium falciparum using Affymetrix arrays. Mol Biochem Parasitol. 2005; 144(2): 177–186. PubMed Abstract | Publisher Full Text\n\nCheeseman IH, Gomez-Escobar N, Carret CK, et al.: Gene copy number variation throughout the Plasmodium falciparum genome. BMC Genomics. 2009; 10: 353. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCheeseman IH, Miller B, Tan JC, et al.: Population Structure Shapes Copy Number Variation in Malaria Parasites. Mol Biol Evol. 2016; 33(3): 603–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChomczynski P: A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples. Biotechniques. 1993; 15(3): 532–4. PubMed Abstract\n\nCoughlin CR 2nd, Scharer GH, Shaikh TH: Clinical impact of copy number variation analysis using high-resolution microarray technologies: advantages, limitations and concerns. Genome Med. 2012; 4(10): 80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDharia NV, Sidhu AB, Cassera MB, et al.: Use of high-density tiling microarrays to identify mutations globally and elucidate mechanisms of drug resistance in Plasmodium falciparum. Genome Biol. 2009; 10(2): R21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrech C, Chen N: Variant surface antigens of malaria parasites: functional and evolutionary insights from comparative gene family classification and analysis. BMC Genomics. 2013; 14(1): 427. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGardner MJ, Hall N, Fung E, et al.: Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002; 419(6906): 498–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarg S, Saxena V, Kanchan S, et al.: Novel point mutations in sulfadoxine resistance genes of Plasmodium falciparum from India. Acta Trop. 2009; 110(1): 75–79. PubMed Abstract | Publisher Full Text\n\nHeinberg A, Siu E, Stern C, et al.: Direct evidence for the adaptive role of copy number variation on antifolate susceptibility in Plasmodium falciparum. Mol Microbiol. 2013; 88(4): 702–712. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJiang H, Yi M, Mu J, et al.: Detection of genome-wide polymorphisms in the AT-rich Plasmodium falciparum genome using a high-density microarray. BMC Genomics. 2008; 9: 398. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKidgell C, Volkman SK, Daily J, et al.: A systematic map of genetic variation in Plasmodium falciparum. PLoS Pathog. 2006; 2(6): e57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKochar DK, Saxena V, Singh N, et al.: Plasmodium vivax malaria. Emerg Infect Dis. 2005; 11(1): 132–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar N, Singh JP, Pande V, et al.: Genetic variation in histidine rich proteins among Indian Plasmodium falciparum population: possible cause of variable sensitivity of malaria rapid diagnostic tests. Malar J. 2012; 11(1): 298. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar SV, Ranjan S, Saxena V, et al.: Plasmodium falciparum: genetic diversity of C-terminal region of MSP-1 in isolates from Indian sub-continent. Exp Parasitol. 2005; 110(4): 384–388. PubMed Abstract | Publisher Full Text\n\nLavstsen T, Salanti A, Jensen AT, et al.: Sub-grouping of Plasmodium falciparum 3D7 var genes based on sequence analysis of coding and non-coding regions. Malar J. 2003; 2(1): 27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeRoux M, Lakshmanan V, Daily JP: Plasmodium falciparum biology: analysis of in vitro versus in vivo growth conditions. Trends Parasitol. 2009; 25(10): 474–481. PubMed Abstract | Publisher Full Text\n\nLiu XS: Getting started in tiling microarray analysis. PLoS Comput Biol. 2007; 3(10): 1842–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaier AG, Rug M, O’Neill MT, et al.: Exported proteins required for virulence and rigidity of Plasmodium falciparum-infected human erythrocytes. Cell. 2008; 134(1): 48–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMishra N, Prajapati SK, Kaitholia K, et al.: Surveillance of artemisinin resistance in Plasmodium falciparum in India using the kelch13 molecular marker. Antimicrob Agents Chemother. 2015; 59(5): 2548–2553. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMobegi VA, Duffy CW, Amambua-Ngwa A, et al.: Genome-wide analysis of selection on the malaria parasite Plasmodium falciparum in West African populations of differing infection endemicity. Mol Biol Evol. 2014; 31(6): 1490–1499. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMockler TC, Chan S, Sundaresan A, et al.: Applications of DNA tiling arrays for whole-genome analysis. Genomics. 2005; 85(1): 1–15. PubMed Abstract | Publisher Full Text\n\nMok BW, Ribacke U, Sherwood E, et al.: A highly conserved segmental duplication in the subtelomeres of Plasmodium falciparum chromosomes varies in copy number. Malar J. 2008; 7(1): 46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNair S, Miller B, Barends M, et al.: Adaptive copy number evolution in malaria parasites. PLoS Genet. 2008; 4(10): e1000243. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOyola SO, Ariani CV, Hamilton WL, et al.: Whole genome sequencing of Plasmodium falciparum from dried blood spots using selective whole genome amplification. Malar J. 2016; 15(1): 597. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPakalapati D, Garg S, Middha S, et al.: Comparative evaluation of microscopy, OptiMAL(®) and 18S rRNA gene based multiplex PCR for detection of Plasmodium falciparum & Plasmodium vivax from field isolates of Bikaner, India. Asian Pac J Trop Med. 2013; 6(5): 346–351. PubMed Abstract | Publisher Full Text\n\nPandey I, Kaur R, Subudhi AK, et al.: Dataset 1 in: A tiling array-based comparative genomic hybridization approach to predict copy number variations between Plasmodium falciparum field isolates from the Indian Sub-continent. F1000Research. 2018a. Data Source\n\nPandey I, Kaur R, Subudhi AK, et al.: Dataset 2 in: A tiling array-based comparative genomic hybridization approach to predict copy number variations between Plasmodium falciparum field isolates from the Indian Sub-continent. F1000Research. 2018b. Data Source\n\nPandey I, Kaur R, Subudhi AK, et al.: Dataset 3 in: A tiling array-based comparative genomic hybridization approach to predict copy number variations between Plasmodium falciparum field isolates from the Indian Sub-continent. F1000Research. 2018c. Data Source\n\nPathak A, Mårtensson A, Gawariker S, et al.: Characterization of drug resistance associated genetic polymorphisms among Plasmodium falciparum field isolates in Ujjain, Madhya Pradesh, India. Malar J. 2014; 13(1): 182. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrzybytkowski E, Ferrario C, Basik M: The use of ultra-dense array CGH analysis for the discovery of micro-copy number alterations and gene fusions in the cancer genome. BMC Med Genomics. 2011; 4(1): 16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRajesh V, Singamsetti VK, Vidya S, et al.: Plasmodium falciparum: genetic polymorphism in apical membrane antigen-1 gene from Indian isolates. Exp Parasitol. 2008; 119(1): 144–151. PubMed Abstract | Publisher Full Text\n\nRibacke U, Mok BW, Wirta V, et al.: Genome wide gene amplifications and deletions in Plasmodium falciparum Mol Biochem Parasitol. 2007; 155(1): 33–44. PubMed Abstract | Publisher Full Text\n\nSaliba KJ, Martin RE, Bröer A, et al.: Sodium-dependent uptake of inorganic phosphate by the intracellular malaria parasite. Nature. 2006; 443(7111): 582–585. PubMed Abstract | Publisher Full Text\n\nScherf A, Lopez-Rubio JJ, Riviere L: Antigenic variation in Plasmodium falciparum Annu Rev Microbiol. 2008; 62: 445–470. PubMed Abstract | Publisher Full Text\n\nSchneider KL: The UCSC Archaeal Genome Browser. Nucleic Acids Res. 2006; 34(Database issue): D407–D410. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSharma D, Lather M, Mallick PK, et al.: Polymorphism in drug resistance genes dihydrofolate reductase and dihydropteroate synthase in Plasmodium falciparum in some states of India. Parasit Vectors. 2015; 8(1): 471. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith JD, Rowe JA, Higgins MK, et al.: Malaria's deadly grip: cytoadhesion of Plasmodium falciparum-infected erythrocytes. Cell Microbiol. 2013; 15(12): 1976–1983. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSnounou G, Zhu X, Siripoon N, et al.: Biased distribution of msp1 and msp2 allelic variants in Plasmodium falciparum populations in Thailand. Trans R Soc Trop Med Hyg. 1999; 93(4): 369–374. PubMed Abstract | Publisher Full Text\n\nSubudhi AK, Boopathi PA, Garg S, et al.: Natural antisense transcripts in Plasmodium falciparum isolates from patients with complicated malaria. Exp Parasitol. 2014; 141: 39–54. PubMed Abstract | Publisher Full Text\n\nSubudhi AK, Boopathi PA, Pandey I, et al.: Disease specific modules and hub genes for intervention strategies: A co-expression network based approach for Plasmodium falciparum clinical isolates. Infect Genet Evol. 2015; 35: 96–108. PubMed Abstract | Publisher Full Text\n\nTan JC, Patel JJ, Tan A, et al.: Optimizing comparative genomic hybridization probes for genotyping and SNP detection in Plasmodium falciparum. Genomics. 2009; 93(6): 543–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTyagi S, Pande V, Das A: Mitochondrial genome sequence diversity of Indian Plasmodium falciparum isolates. Mem Inst Oswaldo Cruz. 2014; 109(4): 494–498. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Dooren GG, Stimmler LM, McFadden GI: Metabolic maps and functions of the Plasmodium mitochondrion. FEMS Microbiol Rev. 2006; 30(4): 596–630. PubMed Abstract | Publisher Full Text\n\nWassmer SC, Grau GE: Severe malaria: what’s new on the pathogenesis front? Int J Parasitol. 2017; 47(2–3): 145–152. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld malaria report 2010. Geneva: World Health Organization, 2010. Reference Source\n\nZeeshan M, Alam MT, Vinayak S, et al.: Genetic variation in the Plasmodium falciparum circumsporozoite protein in India and its relevance to RTS,S malaria vaccine. PLoS One. 2012; 7(8): e43430. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "37246",
"date": "14 Sep 2018",
"name": "Akash Ranjan",
"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\nAuthors report an aCGH based comparative genomic hybridization approach/case report to detect copy number variations CNVs) between genes/genomic locus of Plasmodium falciparum field isolates (Pf-isolate-1 and Pf-isolate-2) from the Indian Sub-continent. This is an experimental case study reporting major amplifications and deletions in Pf-isolate-2 in comparison with Pf-isolate-1 using a tiling array-based comparative genomic hybridization approach. Authors have used this approach to get a real picture of the genomic variations between the two isolates related to the disease severity.\n\nAuthors have identified CNVs in all the chromosomes in Pf-isolate-2 with most of the deletions have been seen in sub-telomeric and telomeric regions of the chromosomes that comprise off variant surface antigen family genes.\n\nApart from the subtelomeric regions, other parts of P. falciparum chromosomes have also shown CNVs. Authors have identified some novel variations, like an amplification of 28kb region (249817-278491) of chromosome-8, which covers for three genes two of which codes for conserved Plasmodium proteins with unknown function (MAL8P1.139, PF08_0122) and tRNA pseudouridine synthase, putative (PF08_0123). Also, amplifications in regions harboring genes like GTP cyclohydrolase I (GCH-1, PFL1155W) and ribosomal protein, L24, putative (PFL1150C) of chromosome 12 were seen.\n\nWhile authors have successfully used their approach to demonstrate the CNVs between the Pf-isolate-2 (an isolates from severe malaria case) and the Pf-isolate-1 (an isolate without severe malaria complication) very little efforts were made to understand this difference in terms of how these CNVs actually explain the disease severity case. This make the manuscript observational/case study type in nature. Nevertheless, this manuscript has provided a glimpse of P. falciparum chromosomal changes that may be associated or responsible for disease severity associated with P. falciparum infection in the field. It has potential to stimulate focused investigations to understand the role of specific amplifications and deletions in severity P. falciparum infection cases.\n\nMinor changes suggested\nPage 1: Please verify the affiliation mark for all the authors there could be a mistake in this, specifically the mark for Raja C. Mugasimangalam. English usage/writing eg articles, punctuations, hyphenation, spelling, capitalization, subject verb agreement etc:\nPage 3 paragraph 1 line 14: \"culture-adapted\" and not \"culture adapted\" Page 3 paragraph 1 line 2: \"that cause\" and not \"that causes \" Page 3 paragraph 1 line 17: \"the understanding\" and not \"understanding\" Page 3 paragraph 1 line 19: \"with an evolutionary advantage\" and not \"an evolutionary advantage\" Page 3 paragraph 1 line 40: \"microvascular\" and not \"micro vascular\" Page 3 paragraph 2 line 4: \"an evolutionary history\" and not \"evolutionary history\" Page 3 paragraph 2 line 5: \"of the mitochondrial genome\" and not \"of mitochondrial genome\" Page 3 paragraph 2 line 6: \"on the analysis of sequence variations\" and not \"on analysis of sequence variations\" Page 3 paragraph 7 line 17: \"gene-based\" and not \"gene based\" Page 3 paragraph 7 line 22: \"amplification\" and not \"amplifications\" Page 4 Table Title:\"amplification\" and not \"amplifications\" Page 4 paragraph 1 line 2: \"the World Health Organization\" and not \"World Health Organization” Page 4 paragraph 2 line 2: \"the manufacturer's\" and not \"manufacture’s\" Page 4 paragraph 2 line 6: \"Nanodrop Technologies\" and not \"Nanodrop technologies\" Page 4 paragraph 3 line 1: \"restriction-digested\" and not \"restriction digested\" Page 4 paragraph 3 line 4: \"Agilent Genomic DNA Labeling Kit\" and not \"Agilent Genomic DNA labeling kit \" Page 4 paragraph 3 line 12: \"Agilent SureHyb chambers\" and not \"Agilent Sure hyb chambers\" Page 4 paragraph 3 line 13: \"Agilent aCGH Wash Buffer\" and not \"Agilent aCGH wash buffer\" Page 4 paragraph 4 line 1: \"high-resolution\" and not \"high resolution\" Page 4 paragraph 4 line 10: \"the Perl code\" and not \"the perl code\" Page 4 paragraph 4 line 13: \"above-mentioned\" and not \"above mentioned\" Page 4 paragraph 4 line 23: \"the Plasmodium Genome Resource\" and not \"Plasmodium Genome Resource\" Page 4 paragraph 5 line 19: \"and represent\" and not \"and represents\" Page 4 paragraph 5 line 21: \"the UCSC genome browser\" and not \"UCSC genome browser\" Page 5 paragraph 2 line 4: \"in the recent version\" and not \"in recent version \" Page 5 paragraph 2 line 7: \"A Circos plot\" and not \"A circos plot\" Page 5 paragraph 3 heading: \"Genome-wide\" and not \"Genome wide\" Page 5 paragraph 3 line 4: \"intra- and intergenic regions\" and not \"intra and intergenic regions\" Page 6 paragraph 1 Line 9:“an 11.7 kb,“and not “a 11.7 kb,” Page 6 paragraph 1 Line 15:“In previous reports, “and not “In previous reports ” Page 6 paragraph 1 Line 31:“in the zinc-ion binding” and not “in zinc ion binding ” Page 6 paragraph 1 Line 38:“transmembrane” and not “transmemberane” Page 6 paragraph 1 last but one line:“other regions covering VSA,“and not “other regions covering VSA ” Page 6 Fig 2 legend: \"The genes present in the region are near centromere on the chromosome-8\" and not \"The genes present in the region, are near centromeric location on the chromosome-8\" Page 6 Fig 2 legend: \"the arrows\" and not \"the the arrows\" Page 6 Fig 2 legend: \"The white double-headed arrows in single exons are representing the 5the 5exons are representing the 5epresents the the arrows in between the exons are representing introns and strand in which genes are present (Right-hand side arrow forward strand, left-hand side arrow representing reverse strand).\" Not clear please rewrite. Page 7 paragraph 1 Line 1:“the 50S integral membrane protein” and not “50S integral memberane protein” Page 7 paragraph 1 Line 4:“the Dd2” and not “Dd2” Page 7 paragraph 2 Line 5:“nucleotide transhydrogenase” and not “nucelotide transhydrogeanse” Page 7 title:“The set of total genes” and not “Total set of Genes” Page 7 paragraph 3 Line 4:“a scarcity of information” and not “scarcity of information” Page 7 paragraph 3 Line 5:“in the regions with seasonal malaria transmission” and not “, regions with seasonal malarial transmission” Page 7 paragraph 3 Line 9:“A lacuna in genome variation study still exists” and not “The lacunae in genome variation study still exists” Page 7 paragraph 3 Line 14:“kilobases” and not “kilo bases” Page 7 paragraph 3 Line 15:“genes have been seen” and not “genes, have been seen”\n\nPage 7 paragraph 3 l ast Line :“rate-limiting” and not “rate limiting” Page 8 paragraph 2 Line 15:“the coding regions”and not “coding regions,” Page 8 paragraph 2 Line 21:“the ADM-2” and not “ADM-2”\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "38678",
"date": "12 Oct 2018",
"name": "Abhisheka Bansal",
"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 the present manuscript by Pandey et al, ‘A tiling array-based…. Indian Sub-continent’, the authors have used an array based comparative genome hybridization technique (aCGH) to identify genomic variations including copy number variations, amplifications, and deletions in field isolates from Indian sub-continent. The authors have used the technique to compare two Plasmodium falciparum field isolates with the objective of identifying genomic differences in Pf-isolate-2 (Plasmodium causing complicated malaria) and Pf-isolate-1 (uncomplicated malaria). The study is largely a case study reporting the genomic differences in Pf-isolate-2. The study has identified amplifications in all the chromosomes of Pf-isolate-2 that are largely limited to telomeric, and sub-telomeric regions and include multigene families such as vars, rifins, and stevors. Though, this is not the first report of aCGH technique being used for comparative genome analysis, it’s application in study of field isolates is demonstrated. The study holds merit for publication in the journal however, there are some criticisms that need to be addressed.\nMajor points:\nThe variations including amplifications and deletions are of common occurrence between two parasite strains isolated from different patients. These variations are more prominent in telomeric and sub-telomeric regions comprising of multigene families such as var genes, etc. Even two different isolates exhibiting identical phenotypic expression may show variations in multigene families. Since the present study is focused on identifying the genomic differences in Pf-isolate-2 (causing complicated malaria) with Pf-isolate-1 (causing uncomplicated malaria), there is no experimental evidence suggestive of direct association of the CNVs found in the study with the disease phenotype. Moreover, it would have been beneficial to incorporate more samples for comparison including parasites exhibiting identical phenotypes. This would have been useful in drawing meaningful conclusions on the CNVs and disease severity. What variations in results are expected if the study were to use Next Generation Sequencing (NSG) instead of aCGH, a summary in the discussion section will be appropriate? How the results might be affected if a newer version of Plasmodium genome database was used? Mention of each gene name in the text, to describe changes, is confusing and it can be presented in the table format making it easier to read and understand. There are genes such as msp2 (Chaorattanakawee et al, 2018; PMID: 29342212), etc that have been demonstrated to be associated with disease severity. Did the authors found any variation in msp2? The findings of such studies reporting genetic markers of disease severity should be discussed in detail in the discussion section.\n\nMinor points: Page 3, Introduction section: references for studies that have identified the receptors for EMP1s should be given such as ICAM-1, CD36, E-selectin, NCAM and CD31 Page 3 paragraph 3 last sentence is not necessary, just include the reference. Page 3 paragraph 5: Objective statement is too long. Reframe it to something like “The objective of this study was to identify genetic difference(s) between two parasite isolates, with varying disease severity, using aCGH tilling approach” Page 4 Table 1 legend: No need to mention the PCR condition here as it has already been mentioned in method section in Page 3 paragraph 7. Page 4 paragraph 5: Please mention what is the “GC%” and “melting temperature” considered for designing 60mer probe. Page 4 paragraph 5: Reframe the sentence “The probes were designed by calculating GC% and optimum Tm values of (>65 genome sequence and calculate the GC% and melting temperature.” The Plasmodium Genome Resource used was very old (V 6.0) need to be verified with the current version of Plasmo DB (V 39.0). Please be consistent with Plasmodium Gene nomenclature as the Gene ID used were from different version of Plasmo DB it makes it difficult to follow.\nGrammatical mistakes: whole manuscript needs a relook to correct the grammatical and typographical errors. Page 6 Figure 2 legend: Change the font of the sentence “conserved Plasmodium membrane protein, unknown function, (PF08_0123) U2 snRNA/tRNA pseudouridine synthase, putative” from italics to normal. Moreover, the legend has many typographical mistakes, for example: “5the”, “5exones”, “5epresents”, and “the the”. Page 7 Figure 3 legend: Similar to Figure 2 legend it has also few typographical errors that needs to be corrected Page 7 All the boxes: the “V” in PlasmoDB Version as well as “P” of “P-value” needs to be consistent. Page 7 last paragraph: Rephrase the sentence “However, there is scarcity of information on genomic variation in parasites causing complicated manifestations, regions with seasonal malarial transmission.” Page 7 last paragraph line 16: Replace “a known fact” with “established” Page 7 last paragraph line 18: replace “invasion” with “evasion” Page 8 paragraph 1 line 5: “Pf-2” and not “PF-2”\n\nReference for datasets presented in the manuscript is not required. It is an integral part of the manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1033
|
https://f1000research.com/articles/7-1032/v1
|
09 Jul 18
|
{
"type": "Research Article",
"title": "Absence of the CHEK2 c.1100delC mutation in familial breast and ovarian cancer in Colombia: a case-control study",
"authors": [
"Ana-Lucia Rivera-Herrera",
"Laura Cifuentes-C",
"JA Gil-Vera",
"Guillermo Barreto",
"Ana-Lucia Rivera-Herrera",
"JA Gil-Vera",
"Guillermo Barreto"
],
"abstract": "Background: BRCA1 and BRCA2 have been identified as high-penetrance breast cancer predisposition genes, but they only account for a small fraction of the inherited component of breast cancer. To explain the remaining cases, a polygenic model with a large number of low- to moderate-penetrance genes have been proposed; one of these, is the CHEK2 gene (Checkpoint Kinase 2). The objective of this study was to determine the role of the CHEK2 gene, specifically the c.1100delC mutation in familial breast cancer susceptibility in Colombian patients. Methods: We screened 131 high-risk breast and/or ovarian cancer patients (negative for mutations in BRCA1 and BRCA2) and 131 controls for the germline mutation CHEK2 c.1100delC by allele-specific PCR. Results: None of the cases or controls showed the CHEK2 c.1100delC mutation, neither as a homozygote nor as a heterozygote. Conclusions: Our results suggest that the CHEK2 c.1100delC mutation is not a risk factor for genetic susceptibility to familial breast or ovarian cancer in the Colombian population. The absence of the CHEK2 c.1100delC mutation in our population show the importance of considering ethnic background before offering a genetic test.",
"keywords": [
"CHEK2",
"familial breast and ovarian cancer",
"Colombia",
"CHEK2 c.1100delC",
"moderate-penetrance"
],
"content": "Introduction\n\nBreast cancer (BC) is the most common type of cancer among women1, and in Colombia, it is the main cause of death by cancer in women1. Of all cases of BC, approximately 5–10% have a strong inherited component, of which 25% is explained by germline mutations in the genes BRCA12 and BRCA23,4. To explain the BRCA1/2-negative cases, a polygenic model in which a large number of low- to moderate-penetrance genes as collectively responsible for the disease has been proposed5–7.\n\nCHEK2 has been proposed as a moderate penetrance BC susceptibility gene8. This gene controls cell cycle and apoptosis and is activated in response to DNA double-strand breakage8. Several mutations in the CHEK2 gene have been found, being the CHEK2 c.1100delC mutation the most studied; this is a truncating mutation in exon 10 that abolishes kinase activity of the protein9. The role of this mutation in breast cancer was confirmed by Meijers-Heijboer et al8. and in several other studies10–22. The mutation CHEK2 c.1100delC was identified in approximately 5% of families with BC that did not have mutations in either BRCA1 or BRCA2 and was estimated to confer moderate risk (20–25%) of developing breast cancer for female mutation carriers23.\n\nData on the contribution of moderate- or low-penetrance alleles to BC in South American populations are scarce. In this study, using a case-control design, we studied the CHEK2 c.1100delC mutation in order to investigate the potential influence of this variant on familial Breast and Ovarian Cancer (BOC) susceptibility in a Colombian cohort.\n\n\nMaterials and Methods\n\nThis was a case-control study conducted from 2009 to 2013 with 131 cases and 131 controls, which was carried out to determine the–association between the CHEK2 c.1100delC mutation and increased risk of developing breast cancer. This study was approved by the Ethical Board of the School of Medicine of the University of Valle. Informed consent was obtained from all the participants.\n\nCases. Our study cohort was composed of 131 Colombian familial breast/ovarian cancer (BOC) cases (BRCA1/BRCA2 negative). The patients were selected from the files of different health/Cancer centers after obtaining permission from each Center to participate in this study. These centers were located in the cities of Cali (Hospital Universitario del Valle, FUNCANCER, Clínica Rafael Uribe Uribe, Hematooncologos), Armenia (Oncólogos de Occidente), Cartagena (Fundación Hospital Infantil Napoleón Franco Pareja, Fundación mujeres por tus senos) and Bucaramanga (Insuasty Oncologia e Investigación). The selected cases met, at least, one of the following criteria: at least three family members with breast or ovarian cancer at any age; two first degree family members affected, at least one, with breast cancer before 41 years of age or with ovarian cancer at any age; one breast cancer case diagnosed before 35 years or less; one ovarian cancer case diagnosed before age 31. For all index cases, breast and ovarian cancers were verified by the original pathology report. After the selection of the patients based in the inclusion criteria (Dataset 1), they were asked to give a blood sample collected by a nurse or technician from each health/Cancer center and then sent to our laboratory.\n\nControls. The sample of healthy Colombian controls (n=131) consisted of unrelated individuals, with no personal or familial history of cancer, these individuals were interviewed and informed as to the aims of the study and who gave their informed written consent for anonymous testing. This control cohort was recruited in the same cities where the study cases were collected and also matched by age, sex and socioeconomic strata.\n\nGenomic DNA was extracted from the blood samples of 131 the BC cases and the 131 healthy controls, for both groups the blood samples were collected specifically for this study. Samples were analyzed using allele-specific PCR (ASPCR)24 to detect the presence of CHEK2 c.1100delC, using the primers reported by Rashid et al25. The products of the ASPCR were visualized by electrophoresis on an 8% polyacrylamide gel. The expected results were a band of 200 base pairs (bp) for the deletion of cytosine 1100 or a band 534 bp when the deletion is not present; if the sample showed both bands, it meant the presence of a heterozygote.\n\n\nResults\n\nA total of 262 samples (131 study cases and 131 matched-controls) were analyzed for the presence of the CHEK2 c.1100delC mutation. The clinical characteristics of the families included in this study are listed in Table 1 (Also see Dataset 1). After allele-specific PCR analysis, none of the cases or controls showed the CHEK2 c.1100delC mutation. All the samples showed only the heavier band (534bp) (Supplementary Figure 1). This indicates that the all the samples were homozygotes for the normal allele.\n\n\nDiscussion\n\nThis is the first study conducted in the Colombian population for CHEK2 c.1100delC mutation. Our results have shown that none of the 262 analyzed samples carried the CHEK2 c.1100delC mutation, suggesting that the frequency of this mutation is extremely low (or not present) in the Colombian population.\n\nSimilar results have been reported for populations in other South American countries. Although studies are scarce, in the few populations that have been evaluated for the presence of this variant, it has not been found in Chile26 and Mexico27, or has been found at very low frequencies in Brazilian population10,22,28,29.\n\nWorldwide, CHEK2 c.1100delC is absent in Spain30–31 and all Asian populations studied to date, including those in India32, Japan33, China34, Korea35, Singapore36, the Philippines22, Pakistan37 and Malaysia38. The mutation is present in populations of Galicia (northwest Spain)39–41, 1.4% of the Northern European countries including Finland, the UK and the Netherlands8,11,42 and the United States43–45.\n\nAn explanation for this variability, proposes CHEK2 c.1100delC as an allele whose frequency is distributed along a population-gradient, which would have originated in populations of Northern Europe (with high frequencies) and decreasing towards the regions of Southern Europe (Basque Country, Spain and Italy)39. This may explain the absence of the allele in the Colombian population., which is a mixture, in different proportions, of European-Spanish, African and Native American ancestry46. Hence, the probability that the allele would have reached our population is low. As reported in several studies, it is evident that the contribution of CHEK2 c.1100delC mutation to the burden of cancer varies according to the ethnic group, and from country to country47.\n\nWe found that the CHEK2 c.1100delC mutation is not present or is present at an extremely low frequency in familial BOC cases and controls in our Colombian cohort.\n\nBased on our findings, we suggest that genotyping of the CHEK2 c.1100delC mutation in genetic testing for breast cancer susceptibility in the Colombian population should not be recommended. However, further studies are required to confirm the contribution of this variant in the Colombian population.\n\nThe study of CHEK2 mutations in the Latin American population has been focused mainly in the c.1100delC mutation. However, databases like ExAC (Exome Aggregation Consortium) showed the presence of other germline mutations in the CHEK2 gene in Latin American samples that could generate cancer susceptibility48. Accordingly, it would be important to examine other mutations in the Colombian population and its association with the development of familial BOC.\n\n\nData availability\n\nDataset 1: Raw data for “Genotype and type of cancer present in the studied patients and their families” 10.5256/f1000research.13368.d20708449",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nBG received funding by COLCIENCIAS (Grant ID 1106-519-29134) and Universidad del Valle (Grant ID CI7830). 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 thank the families and institutions who participated in the research studies described in this article, Hospital Universitario del Valle, FUNCANCER, Clínica Rafael Uribe Uribe, Hematooncologos, Oncólogos de Occidente, Fundación Hospital Infantil Napoleón Franco Pareja, Fundación mujeres por tus senos and Insuasty Oncologia e Investigación.\n\n\nSupplementary Material\n\nSupplementary Figure 1: Allele specific PCR products for the CHEK2 c.1100delC mutation.\n\nPolyacrylamide gel electrophoresis at 8% (29:1 acrylamide:bisacrylamide) of allele specific PCR products for the CHEK2 c.1100delC variant for cases and controls. Lanes 1 - 5 show the amplified product of Colombian patients with breast and/or familial ovarian cancer; lanes 7-11 show the amplified product of healthy controls; Lane 6: molecular weight marker of 100 base pairs. all individuals show an amplified product of 537 bp which corresponds to the normal genotype. The letters “p.b.” in the picture are the spanish initials for base pairs (“pares de bases”).\n\nClick here to access the data.\n\n\nReferences\n\nFerlay J, Soerjomataram I, Dikshit R, et al.: Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015; 136(5): E359–E386. PubMed Abstract | Publisher Full Text\n\nMiki Y, Swensen J, Shattuck-Eidens D, et al.: A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science. 1994; 266(5182): 66–71. PubMed Abstract | Publisher Full Text\n\nWooster R, Bignell G, Lancaster J, et al.: Identification of the breast cancer susceptibility gene BRCA2. Nature. 1995; 378(6559): 789–792. PubMed Abstract | Publisher Full Text\n\nBarnes DR, Antoniou AC: Unravelling modifiers of breast and ovarian cancer risk for BRCA1 and BRCA2 mutation carriers: update on genetic modifiers. J Intern Med. 2012; 271(4): 331–343. PubMed Abstract | Publisher Full Text\n\nEaston DF: How many more breast cancer predisposition genes are there? Breast Cancer Res. 1999; 1(1): 14–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAntoniou AC, Pharoah PD, Mcmullan G, et al.: Evidence for further breast cancer susceptibility genes in addition to BRCA1 and BRCA2 in a population-based study. Genet Epidemiol. 2001; 21(1): 1–18. PubMed Abstract | Publisher Full Text\n\nAntoniou AC, Pharoah PD, Mcmullan G, et al.: A comprehensive model for familial breast cancer incorporating BRCA1, BRCA2 and other genes. Br J Cancer. 2002; 86(1): 76–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeijers-Heijboer H, van den Ouweland A, Klijn J, et al.: Low-penetrance susceptibility to breast cancer due to CHEK2(*)1100delC in noncarriers of BRCA1 or BRCA2 mutations. Nat Genet. 2002; 31(1): 55–59. PubMed Abstract | Publisher Full Text\n\nBell DW, Varley JM, Szydlo TE, et al.: Heterozygous germ line hCHK2 mutations in Li-Fraumeni syndrome. Science. 1999; 286(5449): 2528–2531. PubMed Abstract | Publisher Full Text\n\nBell DW, Kim SH, Godwin AK, et al.: Genetic and functional analysis of CHEK2 (CHK2) variants in multiethnic cohorts. Int J Cancer. 2007; 121(12): 2661–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCHEK2 Breast Cancer Case-Control Consortium: CHEK2*1100delC and susceptibility to breast cancer: a collaborative analysis involving 10,860 breast cancer cases and 9,065 controls from 10 studies. Am J Hum Genet. 2004; 74(6): 1175–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCybulski C, Wokolorczyk D, Huzarski T, et al.: A deletion in CHEK2 of 5,395 bp predisposes to breast cancer in Poland. Breast Cancer Res Treat. 2007; 102(1): 119–22. PubMed Abstract | Publisher Full Text\n\nDe Jong MM, van der Graaf W, Nolte IM: Increased CHEK2 1100delC genotype frequency (also) in unselected breast cancer patients. J Clin Oncol. 2004; 22(14 suppl): 9536. Publisher Full Text\n\nGhadirian P, Robidoux A, Zhang P, et al.: The contribution of founder mutations to early-onset breast cancer in French-Canadian women. Clin Genet. 2009; 76(5): 421–6. PubMed Abstract | Publisher Full Text\n\nKleibl Z, Novotny J, Bezdickova D, et al.: The CHEK2 c.1100delC germline mutation rarely contributes to breast cancer development in the Czech Republic. Breast Cancer Res Treat. 2005; 90(2): 165–7. PubMed Abstract | Publisher Full Text\n\nOffit K, Pierce H, Kirchhoff T, et al.: Frequency of CHEK2*1100delC in New York breast cancer cases and controls. BMC Med Genet. 2003; 4: 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRashid MU, Jakubowska A, Justenhoven C, et al.: German populations with infrequent CHEK2*1100delC and minor associations with early-onset and familial breast cancer. Eur J Cancer. 2005; 41(18): 2896–903. PubMed Abstract | Publisher Full Text\n\nThompson D, Seal S, Schutte M, et al.: A multicenter study of cancer incidence in CHEK2 1100delC mutation carriers. Cancer Epidemiol Biomarkers Prev. 2006; 15(12): 2542–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVahteristo P, Bartkova J, Eerola H, et al.: A CHEK2 genetic variant contributing to a substantial fraction of familial breast cancer. Am J Hum Genet. 2002; 71(2): 432–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeischer M, Bojesen SE, Ellervik C, et al.: CHEK2*1100delC genotyping for clinical assessment of breast cancer risk: meta-analyses of 26,000 patient cases and 27,000 controls. J Clin Oncol. 2008; 26(4): 542–8. PubMed Abstract | Publisher Full Text\n\nWeischer M, Bojesen SE, Tybjaerg-Hansen A, et al.: Increased risk of breast cancer associated with CHEK2*1100delC. J Clin Oncol. 2007; 25(1): 57–63. PubMed Abstract | Publisher Full Text\n\nZhang S, Phelan CM, Zhang P, et al.: Frequency of the CHEK2 1100delC mutation among women with breast cancer: an international study. Cancer Res. 2008; 68(7): 2154–7. PubMed Abstract | Publisher Full Text\n\nNagel JH, Peeters JK, Smid M, et al.: Gene expression profiling assigns CHEK2 1100delC breast cancers to the luminal intrinsic subtypes. Breast Cancer Res Treat. 2012; 132(2): 439–448. PubMed Abstract | Publisher Full Text\n\nWu DY, Ugozzolit L, Pal BK, Wallace RB: Allele-specific enzymatic amplification of beta-globin genomic DNA for diagnosis of sickle cell anemia. Proc Natl Acad Sci U S A. 1989; 86(8): 2757–2760. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRashid MU, Jakubowska A, Justenhoven C, et al.: German populations with infrequent CHEK2*1100delC and minor associations with early-onset and familial breast cancer. Eur J Cancer. 2005; 41(18): 2896–2903. PubMed Abstract | Publisher Full Text\n\nGonzález-Hormazábal P, Castro VG, Blanco R, et al.: Absence of CHEK2 1100delC mutation in familial breast cancer cases from a South American population. Breast Cancer Res Treat. 2008; 110(3): 543–5. PubMed Abstract | Publisher Full Text\n\nChaudhury A, Laukaitis C, Mauss C, et al.: Abstract P3-07-05: Frequent BRCA1 and BRCA2 mutations are found in Mexican and Mexican-American women with breast cancer. Cancer Res. 2013; 73(24 suppl). Publisher Full Text\n\nAbud J, Koehler-Santos P, Ashton-Prolla P, et al.: CHEK2 1100DELC germline mutation: a frequency study in hereditary breast and colon cancer Brazilian families. Arq Gastroenterol. 2012; 49(4): 273–8. PubMed Abstract | Publisher Full Text\n\nPalmero EI, Alemar B, Schüler-Faccini L, et al.: Screening for germline BRCA1, BRCA2, TP53 and CHEK2 mutations in families at-risk for hereditary breast cancer identified in a population-based study from Southern Brazil. Genet Mol Biol. 2016; 39(2): 210–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOsorio A, Rodríguez-López R, Diez O, et al.: The breast cancer low-penetrance allele 1100delC in the CHEK2 gene is not present in Spanish familial breast cancer population. Int J Cancer. 2004; 108(1): 54–56. PubMed Abstract | Publisher Full Text\n\nSánchez de Abajo A, de la Hoya M, Godino J, et al.: The CHEK2 1100delC allele is not relevant for risk assessment in HNPCC and HBCC Spanish families. Fam Cancer. 2005; 4(2): 183–186. PubMed Abstract | Publisher Full Text\n\nRajkumar T, Soumittra N, Nancy NK, et al.: BRCA1, BRCA2 and CHEK2 (1100 del C) germline mutations in hereditary breast and ovarian cancer families in South India. Asian Pac J Cancer Prev. 2003; 4(3): 203–208. PubMed Abstract\n\nSong CG, Hu Z, Yuan WT, et al.: [CHEK2 c.1100delC may not contribute to genetic background of hereditary breast cancer from Shanghai of China]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi. 2006; 23(4): 443–45. PubMed Abstract\n\nChen W, Yurong S, Liansheng N: Breast cancer low-penetrance allele 1100delC in the CHEK2 gene: not present in the Chinese familial breast cancer population. Adv Ther. 2008; 25(5): 496–01. PubMed Abstract | Publisher Full Text\n\nChoi DH, Cho DY, Lee MH, et al.: The CHEK2 1100delC mutation is not present in Korean patients with breast cancer cases tested for BRCA1 and BRCA2 mutation. Breast Cancer Res Treat. 2008; 112(3): 569–73. PubMed Abstract | Publisher Full Text\n\nLee AS, Ang P: CHEK2*1100delC screening of Asian women with a family history of breast cancer is unwarranted. J Clin Oncol. 2008; 26(14): 2419; author reply 2419-20. PubMed Abstract | Publisher Full Text\n\nQureshi Z, Mahjabeen I, Baig R, et al.: Correlation between selected XRCC2, XRCC3 and RAD51 gene polymorphisms and primary breast cancer in women in Pakistan. Asian Pac J Cancer Prev. 2014; 15(23): 10225–9. PubMed Abstract | Publisher Full Text\n\nThirthagiri E, Cheong LS, Yip CH, et al.: CHEK2*1100delC does not contribute to risk to breast cancer among Malay, Chinese and Indians in Malaysia. Fam Cancer. 2009; 8(4): 355–8. PubMed Abstract | Publisher Full Text\n\nMartínez-Bouzas C, Beristain E, Guerra I, et al.: CHEK2 1100delC is present in familial breast cancer cases of the Basque Country. Breast Cancer Res Treat. 2007; 103(1): 111–3. PubMed Abstract | Publisher Full Text\n\nGutiérrez-Enríquez S, de la Hoya M, Martínez-Bouzas C, et al.: Screening for large rearrangements of the BRCA2 gene in Spanish families with breast/ovarian cancer. Breast Cancer Res Treat. 2007; 103(1): 103–107. PubMed Abstract | Publisher Full Text\n\nFachal L, Santamariña M, Blanco A, et al.: CHEK2 c.1100delC mutation among non-BRCA1/2 Spanish hereditary breast cancer families. Clin Transl Oncol. 2013; 15(2): 164–5. PubMed Abstract | Publisher Full Text\n\nVahteristo P, Bartkova J, Eerola H, et al.: A CHEK2 genetic variant contributing to a substantial fraction of familial breast cancer. Am J Hum Genet. 2002; 71(2): 432–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOffit K, Pierce H, Kirchhoff T, et al.: Frequency of CHEK2*1100delC in New York breast cancer cases and controls. BMC Med Genet. 2003; 4: 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMateus Pereira LH, Sigurdson AJ, Doody MM, et al.: CHEK2:1100delC and female breast cancer in the United States. Int J Cancer. 2004; 112(3): 541–543. PubMed Abstract | Publisher Full Text\n\nFriedrichsen DM, Malone KE, Doody DR, et al.: Frequency of CHEK2 mutations in a population based, case-control study of breast cancer in young women. Breast Cancer Res. 2004; 6(6): R629-35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRojas W, Parra MV, Campo O, et al.: Genetic make up and structure of Colombian populations by means of uniparental and biparental DNA markers. Am J Phys Anthropol. 2010; 143(1): 13–20. PubMed Abstract | Publisher Full Text\n\nMarouf C, Hajji O, Diakité B, et al.: The CHEK2 1100delC allelic variant is not present in familial and sporadic breast cancer cases from Moroccan population. Springerplus. 2015; 4: 38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuauque-Olarte S, Rivera-Herrera AL, Cifuentes-C L: Mutations of the CHEK2 gene in patients with cancer and their presence in the Latin American population [version 1; referees: 3 approved with reservations]. F1000Research. 2016; 5: 2791. Publisher Full Text\n\nRivera-Herrera AL, Cifuentes-C L, Gil-Vera J, Barreto G, et al.: Dataset 1 in: Absence of the CHEK2 c.1100delC mutation in familial breast and ovarian cancer in Colombia: a case-control study. F1000Research. 2018. Data Source"
}
|
[
{
"id": "46435",
"date": "04 Apr 2019",
"name": "Taru A. Muranen",
"expertise": [
"Reviewer Expertise Cancer genetics",
"breast cancer"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nRivera-Herrera et al. report a carefully designed study on the occurrence of a moderate-penetrance breast cancer-susceptibility mutation CHEK2: c.1100delC. The paper is well written and the methods and results clearly articulated.\nc.1100delC is a founder mutation enriched in certain Northern-European populations. The two-to threefold increased risk associated with the mutation and the relatively high population carrier frequency (0.5-1.5%) e.g. in the United States, United Kingdom, The Netherlands, Germany, and Finland has prompted the inclusion of the mutation in the panels used for genetic counselling. However, in countries, where the c.1100delC frequency is very low (less than 0.1%), testing for the mutation may not be cost-effective. On the other hand, since c.1100delC is a causal mutation with verified risk effect, independent of the genetic background or ethnicity, even in countries with admixed population and very low carrier frequency the positive carrier status is informative for the rare mutation carriers and their families.\nThe report of Rivera-Herrera et al. serves well in the building of an overall view of breast cancer risk mutations present in Hispanic populations and would thus deserve visibility in the scientific community. However, they should address the limitations of their study and report the genotyping conditions in more detail. There are two issues:\nIn genotyping, they should use a positive control, i.e. a DNA sample from a c.1100delC carrier, to ensure proper performance of the assay. Referring to conditions and assays used in previous studies is not enough, because the assay ought to be optimized on site. If the authors had included a positive control, they should indicate it in the text. The authors should discuss the power of their analysis. Assuming underlying carrier frequency of 1.16% (as in Couch et al. 2017, JAMA Oncology1) in Colombian breast cancer families, the probability of by chance not detecting any carrier in 131 probands would be 21.7%, which is rather high for a conclusion that the mutation is not present in the country. The authors refer to other studies in Latin-American countries and claim that the c.1100delC mutation is absent or has very low frequency. These studies are also quite small. Furthermore, in the Brazilian studies the mutation frequency in families appeared comparable to the frequency in Couch et al.1, and thus should not be referred to as very low.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "49870",
"date": "01 Jul 2019",
"name": "Fernanda S. Jalil",
"expertise": [
"Reviewer Expertise Diagnostic laboratory",
"genes sequencing"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nRivera-Herrera et al. reported the analysis of CHEK2 c.1100delC in a population selected following the NCCN criteria, very well detailed in table 1. The MS is clearly written and thus, easy to follow.\nSpecific comments:\nThe actual nomenclature for this variant is of CHEK2 c.1100del.\n\nIt might be important to have the information of the ethnic origin of the probands analyzed since this pathogenic variant in common in Northern-European countries and to make a population conclusion it is critical.\n\nThe number of patients analyzed it is low based on the expected frequency in the northern countries that may have the highest, 0.5-1,16%, while it is expected based on the databases 100 times lower in Latin populations1. Although in Brazil it is reported 1.76%, based in 1 out 59 probands analyzed, a number still low to analyze a cohort of 131 patients and extrapolate to a population frequency.\n\nThe above comment does not deny the importance of the testing since the family diagnosis it is always very valuable for the counselling the relatives.\n\nIn methods it is very necessary to show the capacity to detect a positive mutation for CHEK2 c.1100del since a few reasons could account for a false negative result. The authors should include the DNA sequence profile for this positive control, in the case it was done.\n\nA brief explanation of the method used for BRCA1/2, as the patients were selected for not being a carrier of a mutation in those genes.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1032
|
https://f1000research.com/articles/7-1028/v1
|
09 Jul 18
|
{
"type": "Research Article",
"title": "A three minutes supine position test reveals higher risk of spinal anesthesia induced hypotension during cesarean delivery. An observational study.",
"authors": [
"Markos Erango",
"Arnoldo Frigessi",
"Leiv Arne Rosseland",
"Markos Erango",
"Arnoldo Frigessi"
],
"abstract": "Background: Cesarean delivery is performed under spinal anesthesia, and vasodilation is the main cause for a drop in blood pressure. The compression of the aorta and inferior vena cava by the gravid uterus is of additional clinical importance. Hypotension may occur during cesarean delivery even if prophylactic infusion of phenylephrine is practiced. We have tested if a 3 minute supine observation, can identify a subset of women with decreasing systolic arterial pressure (SAP) under spinal anesthesia. Methods: We performed a prospective observational study at Oslo University Hospital on healthy pregnant women for planned cesarean delivery. Continuous measurements of calibrated invasive SAP and estimated cardiac output were recorded for 76 women in a 3 minutes measurement with the woman in the left lateral position, followed by supine position for 3 minutes. Using functional data clustering, principal component analysis and curve smoothing, to filter way noise and reduce the dimensionality of the signal, we clustered the women into separate SAP groups.\n\nResults: We identified two significantly different groups of women during supine position; one characterized by initial drop in SAP, the other showed initial increase. After spinal anesthesia, the mean SAP curve of the women in the first group showed a drop in blood pressure, which was more rapid than for the other women. A minor difference in cardiac output was observed between the two groups of women with the mean cardiac output curve for the first group being higher. Conclusions: This work indicates that supine position affect clinically relevant cardiovascular measurements in pregnant women. A simple test may identify patients with increased risk of spinal anesthesia induced hypotension.",
"keywords": [
"Blood pressure",
"spinal anesthesia",
"cardiac output",
"cesarean delivery"
],
"content": "List of abbreviations\n\nSAP; systolic arterial pressure\n\nHb; hemoglobin concentration\n\nBMI; body mass index\n\nDAP; diastolic arterial pressure\n\nMAP; mean arterial pressure\n\nHR; heart rate\n\nSV; stroke volume\n\nCO; cardiac output\n\nSVR; systemic vascular resistance\n\n\nBackground\n\nSpinal anesthesia is a standard method in cesarean deliveries and is regarded as being safe for the mother and the baby, even though significant maternal hypotension often occur1. The vasodilatory effect of spinal anesthesia is the main cause for this drop in blood pressure2. The compression of the aorta and inferior vena cava by the gravid uterus is known to reduce venous return and this is estimated to be of clinical importance after approximately 20 weeks of pregnancy. Pregnant women may find the supine position uncomfortable and avoid this position close to delivery. During cesarean delivery left lateral tilt is recommended to oppose this effect, and theoretically this should reduce the negative effects on maternal circulation. However, in clinical practice the incidence of spinal induced hypotension is high if no prophylactic measures to oppose the vasodilatation are included1. We have previously compared two different prophylactic methods, lower leg compression and phenylephrine3. In that study continuous infusion of phenylephrine was superior to lower leg compression and placebo. Lower leg compression was found to increase heart stroke volume compared to placebo indicating a significant effect on venous return. Some women have hypotension during cesarean delivery even if prophylactic infusion of phenylephrine and/or lower leg compression is practiced. Phenylephrine continuous infusion is not included as standard prophylaxis globally4. Previous attempts to predict risk of spinal induced hypotension and to individualize the prophylactic treatment have not yet changed clinical practice.\n\nWe tested if a 3 minutes supine observation prior to spinal anesthesia with continuous invasive hemodynamic measurements\n\na) Allows identification of a subset of women with decreasing blood pressure and/or cardiac output;\n\nb) If this subset of women could be identified in left lateral positon;\n\nc) If this subset of women showed any special properties in terms of cardiac output and systolic arterial pressure during cesarean delivery with phenylephrine prophylaxis under spinal anesthesia.\n\n\nMethods\n\nThe patients participated in a randomized, double-blinded, placebo-controlled, parallel-group comparison of carbetocin and oxytocin given intravenously during elective cesarean delivery under spinal anesthesia. The study was conducted at the Department of Anesthesiology, Division of Emergency and Critical Care, Oslo University Hospital, Rikshospitalet between November 2009 and September 2011. The Birth Clinic at Rikshospitalet is a tertiary care center, but the majority of the laboring women are healthy and representative of the general population in central Norway. The protocol was approved by the Data Inspectorate’s local representative at Oslo University Hospital, the Regional Committee for Medical and Health Research Ethics of Southern Norway (Oslo, Norway)(2009/130), and the Norwegian Medicines Agency (Oslo, Norway)(09/07301–7). This approval covered both the RCT and sub-study. The randomized controlled trial was registered at clinicaltrials.gov (NCT00977769) and was conducted according to Good Clinical Trial practice, the CONSORT guidelines, and the principles of the Declaration of Helsinki. The results from the randomized controlled trial are published previously5.\n\nIn addition to the RCT we performed an observational study. Detailed information about the observational study was integrated in the patient information and consent form. The observational study was conducted and reported according to STROBE guidelines6.\n\nEligible participants were screened by the senior author (LAR) for inclusion at their last midwife consultation before their scheduled delivery. Oral and written information was given to each woman at least 24 hours before her delivery. Written informed consent was gained before inclusion. The women were informed about the intervention in the randomized controlled trial and the observational study being performed the last minutes before spinal anesthesia. They were asked to assess specifically the extra pain caused by to the arterial cannulation. The inclusion criteria included being in good health, a singleton pregnancy, being aged 18 years or older, and a scheduled planned cesarean section at a minimum gestational age of 36 weeks. The exclusion criteria included pre-eclampsia, placenta previa, placenta accreta, von Willebrand disease or other bleeding disorder, and systolic arterial pressure (SAP) < 90 mmHg. A total of 185 patients were screened, and 76 were included in the observational study. The indications for elective cesarean section were maternal request 43%, previous cesarean delivery 25%, breech presentation 8%, other obstetric reasons 13%, and other non-obstetric maternal or neonatal medical conditions 11%.\n\nIn addition to continuous beat-to-beat hemodynamic variables that were collected and stored electronically during left and supine position, we registered each patient’s age, height, weight, gestational age, preoperative hemoglobin concentration (Hb), hours fasting (liquids and solids), and the baby’s birth weight. High body mass index (BMI) may influence the hemodynamic response to both supine position and spinal anesthesia. Maternal weight upon inclusion was used to calculate the BMI (weight in kg / (height in meters)2). An arterial catheter was placed into the radial artery after infiltration of lidocaine (10–20 mg) immediately after arrival at the operating theatre. Peripheral intravenous (i.v.) catheters were placed on both forearms. Invasive blood pressure were calibrated and continuous beat-to-beat measurements of invasive SAP, diastolic arterial pressure (DAP), mean arterial pressure (MAP), and heart rate (HR) were recorded in 3 minutes with the woman in a left lateral position. Stroke volume (SV), cardiac output (CO), systemic vascular resistance (SVR), and other estimated variables based on continuous arterial waveform analysis system were recorded using PulseCO (PulseCOTM, Cambridge, United Kingdom), an integrated part of the LiDCOplus monitoring system (LiDCO Ltd., Cambridge, United Kingdom). We omitted the calibration of CO with the lithium dilution technique since the primary outcome was change in repeated measurements of hemodynamic variables. The hemodynamic observations were performed before spinal anesthesia was given. After the 3 minutes measurements in the left lateral position the women were turned into supine position for the 3 minutes test of aorto-caval compression. This experiment was not part of standard clinical procedure, and the women were encouraged to discontinue the test if they felt unwell.\n\nWith the woman in a right lateral position, spinal anesthesia was induced in the L2-L3 vertebral interspace, and bupivacaine 10 mg + fentanyl 20 µg was injected through a 25-G non-traumatic needle (Pencan®, B. Braun, Melsungen, Germany). Concomitantly, we started a rapid intravenous infusion of saline 0.9 mg/ml (37°C, 10 ml/kg) and a phenylephrine bolus (0.25 µg/kg). This was followed with a phenylephrine infusion (0.25 µg/kg/min). During surgery, the patient was supine with an operating wedge under her right hip (19° Tempur pillow). Hypotension (SAP< 90 mmHg) was treated with an extra i.v. bolus of phenylephrine 30 µg if the HR was above 60 beats per minute (bpm) or with i.v. ephedrine 5–10 mg if the HR was 60 bpm or below.\n\nThe hemodynamic data were stored prepared for statistical analysis according to methods described in detail previously5. The data set used in the analyses of the supine test is open access available for analyses (Dataset 1).\n\nPhenylephrine prophylaxis (25 µg/kg/min i.v.) was started when spinal anesthesia was given. Due to the phenylephrine prophylaxis we expected only a minor decrease in SAP after induction of spinal anesthesia. The analyses of spinal induced hypotension were based on the beat-to-beat measurements during the first 5 minutes, guided by inspection of the blood pressure curves. Based on a previously published study we expected a maximum decrease in SAP 3–4 minutes after spinal anesthesia3,7.\n\nIn order to capture the change of blood pressure, we considered the first order difference between two consecutive measurements di (t) = bi (t + 1) – bi (t), where bi (t) is the average blood pressure of woman i measured in the 5 seconds before time t. The time t is discrete, so that t = 1, …, T = 40 corresponds to the three minutes measurement. Because data were still noisy, we smoothed the first order difference blood pressure curve di (t), t = 1, …, T of every woman, using a roughness penalty method8, which led to a smoothed first order difference blood pressure curve denoted si (t), t = 1,…, T – 1. The method depends on a smoothing parameter λ. The results of the analysis below depend on the choice of λ. However, when clustering the women in separate groups, only a few women changed cluster when we changed the value of λ. To document such robustness, in the Supplementary material, we present results obtained for many values of λ (Supplementary Table 1, Supplementary File 1 and Supplementary Figure 1). We found that λ=10000 allows to filter noise and capture details in the best way and in the rest we show the results for λ=10000.\n\nWe computed the principal components of the smoothed first order difference blood pressure curves si (t) for all women. The first two principal components represented in total 76.9% of the total variation of the data and were used for clustering. We clustered the 76 women using these two first principal components, into two or three groups. We used the k-means method9, implemented in cclust package (Version 0.6–21)10. More details and an intuitive description of the statistical methods are given in the Supplementary material (Supplementary File 1).\n\nWe used the multivariate paired Hotelling’s T2 test11 to test the null hypothesis that the mean curves of the two clusters are equal for all time points against the alternative that the mean vectors of the two groups are not equal for all time points (H0: μ1 = μ2 vs. H1:μ1≠μ2) where μ1 is the mean curve vector for S1 and μ2 is the mean curve vector for S2. For comparison of the three clusters, we used the multivariate MANOVA to test the null hypothesis that the mean curves of are equal for all time points against the alternative that the mean vectors of the three groups are not equal for all time points (H0: μ1 = μ2 = μ3 vs. H1:μ1≠μ2 ≠ μ3), where μ1 is the mean curve vector for T1, μ2 is the mean curve vector for T2 and μ3 is the mean curve vector for T3. Statistical analyses were performed with R-software versions 3.2 to 3.3.2.\n\n\nResults\n\nBaseline measurements of HR, SAP, MAP, DAP, BMI, age, parity, gestational age, hours in fasting for solids, and estimated intake of liquids for the 76 included women are presented in Table 1. Mean supine SAP curves of the women with increasing values are available as Supplementary Figures 2. The principle component analysis identified two groups of women, by clustering the smoothed blood pressure difference curves during supine position into two groups as described in the methods. The first group, denoted S1, included 30 women, the second group, S2, the remaining 46 women. There were no statistically significant differences between the two clusters in baseline characteristics (see Supplementary Table 2).\n\nData presented as mean (SD, range, or total number).\n\nBMI; body mass index, GA; gestational age,\n\na; baseline values representing a 60 sec mean of intra-arterial blood pressure (AP), and heart rate (HR) with the patient in left lateral position.\n\nFigure 1 shows the mean SAP curves for the two groups, which are characterized by different trajectories. During supine position the mean SAP of the S2 patients dropped by about 10 mmHg in the first 100 seconds. After that the blood pressure was roughly constant. In the S1 group, the estimated mean SAP increased in the first 50 seconds of about 5 mmHg, and then started to decrease. The decrease was however less sharp than for group S2 (T2 = 216.89, p-value <0.001). The raw data curves of SAP for S1 and S2 groups are presented as Supplementary Figure 3.\n\nMean smoothed systolic arterial pressure (SAP) curves for S1 and S2 during supine position (mmHg) (panel a) and mean smoothed SAP curves for S1 and S2 during left lateral position (mmHg) (panel b). S1 in red, S2 in black.\n\nWe looked further into possible characterizations of the groups S1 and S2. We computed the standard deviation (SD) of the two groups for each time point. We found that S1 had minimum variation after 35 seconds, and then from the 100th seconds onwards had stable SD, whereas the SD in group S2 had a maximum at start and a minimum after 85 seconds (see Supplementary Figure 4).\n\nIn a second analysis, we clustered the 76 women into three groups. The reason for this was that the percentage (60%) of women in S2, characterized by a mean blood pressure drop, appeared to be larger than expected from clinical experience12. When repeating the same analysis with three groups, we found a first group T1 with 28 women (36.8%) and a second group T2 with 15 women (19.7%) for which the average blood pressure increased in the first 30–50 seconds before then starting to drop. The difference between these two groups was in the different rate of decay, stronger for the women in T2. The third group T3 (43.5%) was characterized by a sharp drop in the average blood pressure right away, of approximately 10 mmHg in the first 100 seconds. After that the mean blood pressure was roughly constant. The three groups are not equal (Wilks Lambda= 0.033481, p-value<0.001). Mean hemodynamic measurements in left lateral and supine position defined by the principle components are presented in Supplementary Table 3.\n\nWe investigated if the clusters defined on the basis of the measurements while in supine position could be associated with blood pressure patterns while on left lateral position. This was not the case, as the mean blood pressures while on left lateral position for groups S1 and S2 were not significantly different (Figure 1). The present data indicated that the drop in blood pressure when put in supine position for one group of women (S2), was not reflected in a blood pressure pattern when on the side which was different from the women in S1. Figure 2 shows how the first two principal components allow clustering the women into two groups (S1 and S2). The clustering algorithm puts each woman in one of two groups, so to divide them in the best way. The two clusters are nicely distinct, though there are borderline women.\n\nThere is one point per woman, with coordinates given with respect to the first and second principal component. PC= principal component.\n\nWhen comparing the two clusters (S1 and S2) in terms of their mean SAP curves, we found that for all women in both groups mean SAP increased roughly 8 mmHg in the first 80–100 seconds after spinal anesthesia. After that, mean SAP started to decrease in both groups (S1 and S2), but the blood pressure drop for the women in S2 was clearly sharper than for the women in S1 (Figure 3). We found that the group of women who experience a drop of blood pressure when put in supine position (S2) also have a drop which is faster and towards lower values compared to the women (in S1) who in supine position have a non-dropping blood pressure (S1) (Figure 3a). Mean cardiac output decreased in both groups but more in the group S1 than in S2 (Figure 3b). The mean SAP curves during anesthesia of the two groups were different (T2 = 370.5, p-value <0.001) as were CO (T2=294.3, p-value <0.001). Due to hypotension (SAP< 90 mmHg) three group S2 patients were given phenylephrine bolus 30 µg. None in group S1 received extra bolus phenylephrine.\n\nMean systolic arterial pressure (SAP)(mmHg) (panel a) and cardiac output (CO)(L/min) (panel b) after spinal anesthesia in the two clusters (S1 and S2).\n\nWhen we considered SAP difference under spinal anesthesia (SAP (time=250 sec) – SAP (time=10 sec)) compared with SAP difference during the supine period (SAP (time=180 sec) – SAP (time=10 sec)), the correlation in S1 was -0.158 (p-value =0.40) and the correlation in S2 -0.48 (p-value = 0.0008). When we consider SAP difference under spinal anesthesia in the period from 90 sec (peak SAP, see Figure 3a) and 250 sec (SAP (time=250 sec) – SAP (time=90 sec)) compared with SAP during supine (SAP (time=180 sec) – SAP (time=10 sec)), the correlation in S1 was 0.095 (p-value = 0.62) the correlation in S2 was 0.15 (p-value = 0.32).\n\nWhen repeating this analysis with the three groups division (T1, T2 and T3) we observed a comparable pattern (Figure 4). Mean SAP increased during the initial 50–100 seconds followed by a decrease which was different for the three groups. The group of women with the most prominent drop in blood pressure when put in supine position (T3), also had the largest drop in average blood pressure under spinal anesthesia, and was still falling after 250 seconds. The group of women with the smallest drop in blood pressure in supine position (T1) did not have any drop in blood pressure under spinal anesthesia, with mean SAP remained constant. The intermediate group of women (T2), with some drop of blood pressure when in supine position, also had some drop under spinal anesthesia, but it was a slower drop than what was observed for T3 (Figure 4a). Mean cardiac output decreased in all three groups (Figure 4b). The group differences in mean SAP and CO curves for all time points were statistically significant (p-value < 0.001)(MANOVA test, Wilks’ Lambda= 0.31862, F-value= 31.36, and Wilks’ Lambda = 0.17622, F-value= 48.307, respectively).\n\n\nDiscussion\n\nThe three minutes supine observation test prior to spinal anesthesia identified women with risk of spinal induced hypotension. This finding is in concordance with previous studies reporting an association between heart rate increase or hypotension in the supine position and subsequent development of spinal hypotension and/or vasopressor requirements13–15. Kinsella et al., studied non-invasive arterial pressure and heart rate during a 5 min supine position test13, and concluded that it was not possible to identify women with increased risk of spinal induced hypotension based on the observation of changes in arterial pressure. In contrast, the patients with a heart rate increase above 10 beats/min were more likely to develop hypotension. The novelty of our method is the potential benefit of in the future being able to predict hypotension and thereby individualize prophylactic therapy. We also demonstrated that this physiologic phenomenon present in the supine position was absent when observing the same women in left lateral position, supporting the concept of aorto-caval compression as the dominant mechanism of hypotension in the supine position13.\n\nThe cluster of supine women having a falling SAP trend developed lower mean SAP and higher mean CO after spinal anesthesia. The decrease in systemic vascular resistance following spinal anesthesia is the dominant mechanism of spinal hypotension2. A statistically significant difference in mean CO was observed between the two groups of women and the group with lowest SAP had a higher CO. The difference in CO was probably of limited clinical significance but indicates that the difference in the compensation of spinal anesthesia induced vasodilatation is the mechanism behind this observation. Women are more dependent on sympathetic tone during pregnancy than in non-pregnant state and the effect of spinal anesthesia is a rapid and prominent vasodilatation, leading to a decrease in blood pressure and increase in CO7,16. All women in this study received standard phenylephrine bolus and infusion, and this prophylaxis was effective as the mean SAP was stable (Supplementary Figure 2c). However, three women had a drop in SAP below 90 mmHg and needed extra doses of phenylephrine to keep SAP close to baseline. All three women belonged to the group characterized with drop in SAP during supine position.\n\nIt was evident that the principal component analysis could not be replaced by physician observation and visual assessment of the continuous SAP curves, as the variance is large and individual changes difficult to interpret (Supplementary Figure 2a). Functional data analysis with principal component analysis of continuous invasive measurements of SAP made a distinction between two separate clusters, one of the two characterized by decreasing mean SAP. It is important to notice that the two groups are not created because the women in S1 started with a lower average blood pressure than the women in S2. The two groups were created from the first order difference blood pressures, where the level does not have any influence. Principal component analyses have been used for making predictive models. In this population it would be interesting to identify patients at risk of severe spinal anesthesia induced hypotension and tailor the prophylactic treatment. The reliability of this method has to be proven in a new independent study.\n\nHeart rate variability analysis captures the autonomic nervous activity and has been shown to predict spinal anesthesia induced hypotension in pregnant women17,18, and the ratio of the low frequency and the high frequency bands (LF/HF) correlated with spinal induced hypotension. Spinal anesthesia induced hypotension is caused by the vasodilatory effects, and to a lesser extent by aorto-caval compression5, and the relationship between estimated autonomic activity (LF/HF) and spinal anesthesia induced hypotension is likely. The requirements of heart rate variability analyses limits the feasibility as stable ECG-signals has to be monitored at least 5 minutes to perform the frequency domain analyses19. Sakata et al., analyzed heart rate variability during postural changes and found a correlation between low-to-high frequency ratio shift and spinal induced hypotension during cesarean delivery18. Time domain analyses or continuous wavelet transform may be applicable20 and requires shorter heart rate recordings. However, the method is not yet validated in the obstetric population.\n\nPulse transit time measures the interval from the electrical activity to the physical pulse wave in the periphery. Pulse transit time provide rapidly available beat-to-beat information of changing hemodynamics and may estimate the real time vasodilatory effect of spinal anesthesia21, and is not suitable in a prediction model.\n\nThe external validity of a classification algorithm based on functional data analysis and clustering, as suggested in this paper, needs to be investigated in similar clinical assays. In this study all patients were treated according to a protocol including phenylephrine starting bolus and continuous infusion and we expected the blood pressure to be quite stable. Treatment protocols without intravenous phenylephrine prophylaxis will be different; Hypotension will be more profound and occur over a longer time sequence3.\n\nIn a randomized placebo-controlled trial we showed that lower leg wrapping was less effective than phenylephrine bolus + infusion regarding spinal induced hypotension3. Cardiac output increased in all groups. The study confirmed that spinal anesthesia is vasodilatory and that the changes in venous return, represented by heart stroke volume, were minor. Interestingly, heart stroke volume was higher in the lower leg wrapping group indicating some effect on venous return. Individual differences in the ability to compensate a fall in venous return may be the mechanism observed in the three minute supine test. Even though vasodilation is the main effect of spinal anesthesia it would be of interest to study how the prediction model works together with mechanical lower leg wrapping or when no hypotension prophylaxis is administered.\n\nThe algorithm should also be evaluated using a non-invasive continuous monitoring device. Non-invasive, continuous, beat-to-beat monitoring of blood pressure is available and included in clinical practice22 and invasive monitoring is not standard clinical practice in obstetric anesthesia. Two classes of devices have been developed; Tonometric devices and volume clamp devices. The limits-of-agreement analyses of the latter 2 device classes using invasive measures as a reference standard are promising22. Non-invasive measurement of blood pressure is not yet validated in obstetric anesthesia.\n\n\nConclusions\n\nThis work indicates the possibility to identify patients with increased risk of spinal anesthesia induced hypotension based on a simple pre-anesthesia test.\n\n\nData availability\n\nDataset 1: Complete continuous invasive hemodynamic measurements in both left lateral and supine positions 10.5256/f1000research.15142.d20936323",
"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\nThe authors thank MD PhD Guro Grindheim, Oslo University Hospital, Oslo, Norway, for contributing during planning of the project and for valuable advices in interpretation, and Professor PhD Magne Thoresen, Oslo Centre of Biostatistics and Epidemiology, and University of Oslo, Oslo, Norway for his advices in planning and interpretation of the statistical analyses.\n\n\nSupplementary material\n\nSupplementary File 1- File containing expansion of the statistical methods and the following supplementary tables:\n\nClick here to access the data.\n\nSupplementary Table 1: Baseline demographics in the cluster S1 and S2.\n\nSupplementary Table 2: Assignment of each woman to one of the two groups for various values of λ\n\nSupplementary File 2: File containing the following supplementary table:\n\nClick here to access the data.\n\nSupplementary Table 3: Hemodynamic variables in clusters defined by principle components\n\nSupplementary Figure 1: plot for each of the values of λ which were tested, the mean supine SAP of the women in each of the two groups. This is the mean of the smoothed supine SAP curves of each woman, and therefore depends on λ. With increasing values of λ, the two curves become more regular. There is little change, but noise is progressively reduced. By smoothing the data, noise is filtered out allowing for a better signal. The important observation is that the results are very stable for λ larger than 5000. This confirms that the choice of λ=10000 is appropriate, and that results would be qualitatively unchanged for other values of λ around 10000.\n\nClick here to access the data.\n\nSupplementary Figure 2: SAP of all women during lateral position (a), supine position (b) and during anesthesia (c), and (in bold red) average SAP at each time point. We did not notice any general clear trend, though the average SAP had a weak negative decay. Notice the large variance among the women.\n\nClick here to access the data.\n\nSupplementary Figure 3: Smoothed SAP curves for S1 and S2 during supine position and mean SAP (mmHg). Notice the large variability. The scale of the y-axis is much larger than in Figure 2 in the paper, to permit the plot of all women data.\n\nClick here to access the data.\n\nSupplementary Figure 4: Standard deviation (SD) of the smoothed SAP curves for S1 and S2 during three minutes supine position, time point per time point. There was a high variability in each group, clearly more in the S2 group. We therefore tried with three groups, so that variance in each groups will be less. Notice that the principle component analysis allows filtering out noise.\n\nClick here to access the data.\n\n\nReferences\n\nCyna AM, Andrew M, Emmett RS, et al.: Techniques for preventing hypotension during spinal anaesthesia for caesarean section. Cochrane Database Syst Rev. 2006; (4): Cd002251. PubMed Abstract | Publisher Full Text\n\nAssali NS, Prystowsky H: Studies on autonomic blockade. I. Comparison between the effects of tetraethylammonium chloride (TEAC) and high selective spinal anesthesia on blood pressure of normal and toxemic pregnancy. J Clin Invest. 1950; 29(10): 1354–1366. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuhn JC, Hauge TH, Rosseland LA, et al.: Hemodynamics of Phenylephrine Infusion Versus Lower Extremity Compression During Spinal Anesthesia for Cesarean Delivery: A Randomized, Double-Blind, Placebo-Controlled Study. Anesth Analg. 2016; 122(4): 1120–1129. PubMed Abstract | Publisher Full Text\n\nNgan Kee WD: The use of vasopressors during spinal anaesthesia for caesarean section. Curr Opin Anaesthesiol. 2017; 30(3): 319–325. PubMed Abstract | Publisher Full Text\n\nRosseland LA, Hauge TH, Grindheim G, et al.: Changes in blood pressure and cardiac output during cesarean delivery: the effects of oxytocin and carbetocin compared with placebo. Anesthesiology. 2013; 119(3): 541–551. PubMed Abstract | Publisher Full Text\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007; 370(9596): 1453–1457. PubMed Abstract | Publisher Full Text\n\nLangesaeter E, Rosseland LA, Stubhaug A: Continuous invasive blood pressure and cardiac output monitoring during cesarean delivery: a randomized, double-blind comparison of low-dose versus high-dose spinal anesthesia with intravenous phenylephrine or placebo infusion. Anesthesiology. 2008; 109(5): 856–863. PubMed Abstract | Publisher Full Text\n\nRamsay J, Silvermann B: Functional Data Analysis. Springer Series in Statistics. In: Wiley Online Library, 1998.\n\nKanungo T, Mount DM, Netanyahu NS, et al.: An efficient k-means clustering algorithm: Analysis and implementation. IEEE transactions on pattern analysis and machine intelligence. 2002; 24: 881–892. Publisher Full Text\n\nDimitriadou E, Hornik K, Hornik MK: Package ‘cclust’. 2015. Reference Source\n\nAnderson T: An Introduction to Multivariate Analysis. John Wiley & Sons (New York). 2003. Reference Source\n\nKinsella SM, Lohmann G: Supine hypotensive syndrome. Obstet Gynecol. 1994; 83(5 Pt 1): 774–788. PubMed Abstract\n\nKinsella SM, Norris MC: Advance prediction of hypotension at cesarean delivery under spinal anesthesia. Int J Obstet Anesth. 1996; 5(1): 3–7. PubMed Abstract | Publisher Full Text\n\nDahlgren G, Granath F, Wessel H, et al.: Prediction of hypotension during spinal anesthesia for Cesarean section and its relation to the effect of crystalloid or colloid preload. Int J Obstet Anesth. 2007; 16(2): 128–134. PubMed Abstract | Publisher Full Text\n\nJeon YT, Hwang JW, Kim MH, et al.: Positional blood pressure change and the risk of hypotension during spinal anesthesia for cesarean delivery: an observational study. Anesth Analg. 2010; 111(3): 712–715. PubMed Abstract | Publisher Full Text\n\nSharwood-Smith G, Drummond GB: Hypotension in obstetric spinal anaesthesia: a lesson from pre-eclampsia. Br J Anaesth. 2009; 102(3): 291–294. PubMed Abstract | Publisher Full Text\n\nHanss R, Bein B, Francksen H, et al.: Heart rate variability-guided prophylactic treatment of severe hypotension after subarachnoid block for elective cesarean delivery. Anesthesiology. 2006; 104(4): 635–643. PubMed Abstract\n\nSakata K, Yoshimura N, Tanabe K, et al.: Prediction of hypotension during spinal anesthesia for elective cesarean section by altered heart rate variability induced by postural change. Int J Obstet Anesth. 2017; 29: 34–38. PubMed Abstract | Publisher Full Text\n\nHeart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation. 1996; 93(5): 1043–1065. PubMed Abstract | Publisher Full Text\n\nWachowiak M, Hay D, Johnson M: Quantification of Wavelet Band Metrics for Assessing Heart Rate Variability. In: World Congress on Medical Physics and Biomedical Engineering, June 7-12, 2015, Toronto, Canada. Springer, 2015: 1026–1029. Publisher Full Text\n\nSharwood-Smith G, Bruce J, Drummond G: Assessment of pulse transit time to indicate cardiovascular changes during obstetric spinal anaesthesia. Br J Anaesth. 2006; 96(1): 100–105. PubMed Abstract | Publisher Full Text\n\nBartels K, Esper SA, Thiele RH: Blood Pressure Monitoring for the Anesthesiologist: A Practical Review. Anesth Analg. 2016; 122(6): 1866–1879. PubMed Abstract | Publisher Full Text\n\nErango M, Frigessi A, Rosseland LA: Dataset 1 in: A three minutes supine position test reveals higher risk of spinal anesthesia induced hypotension during cesarean delivery. An observational study. F1000Research. 2018. Data Source"
}
|
[
{
"id": "35875",
"date": "23 Jul 2018",
"name": "Kim Ekelund",
"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 present paper is a fine example of how simple questions can be answered by simple studies. The authors have observed that although a standardized approach is used to treat the spinal caused vasodilation during a cesarean section (CS) some women will still experience hypotension despite prophylactic treatment with phenylephrine. They try to identify subsets of women with different susceptibility to vasodilation by register the hemodynamic response after placing them in a supine position for a few minutes. The hypothesis is that some women may need a more aggressive anti-hypotensive treatment during CS and these women might react more when placed on the back, compressing the inferior caval vein and aorta.\nThe paper is very interesting because it presents new clinical relevant knowledge on a frequent challenge. The paper is well written, most of the time easy to understand, but is also at some points cut to tight, and I feel as a reader, I am missing information, and this stays with me, despite several attempts to understand.\n\nAbstract The abstract is maybe to condense, (maybe due to restriction from the publisher)… I have added a few comments to make the text easier to read… \"Cesarean delivery is performed under spinal anesthesia, and vasodilation is the main cause for a drop in blood pressure. [During supine position] The compression of the aorta and inferior vena cava by the gravid uterus is of additional clinical importance. [The spinal caused] Hypotension may occur during cesarean delivery even if prophylactic infusion of phenylephrine is practiced. We have tested if a 3 minute supine observation, can identify a subset of women with decreasing systolic arterial pressure (SAP) under spinal anesthesia”\n\nBackground “The compression of the aorta and inferior vena cava by the gravid uterus is known to reduce venous return and this is estimated to be of clinical importance after approximately 20 weeks of pregnancy. Pregnant women may find the supine position uncomfortable and avoid this position close to delivery. During cesarean delivery left lateral tilt is recommended to oppose this effect, and theoretically this should reduce the negative effects on maternal circulation” Q: Consider moving this paragraph. The paragraphs before and after is about vasodilation – this is not.\n\nMethods ”They were asked to assess specifically the extra pain caused by to the arterial cannulation” Q: What does this mean?\n\n“The indications for elective cesarean section were maternal request 43%, previous cesarean delivery 25%, breech presentation 8%, other obstetric reasons 13%, and other non-obstetric maternal or neonatal medical conditions 11%.” Q: Why is this important in the present study?\n\n“Concomitantly, we started a rapid intravenous infusion of saline 0.9 mg/ml (37°C, 10 ml/kg) and a phenylephrine bolus (0.25 µg/kg). This was followed with a phenylephrine infusion (0.25 µg/kg/min)”\nand later in the same column:\n”Phenylephrine prophylaxis (25 µg/kg/min i.v.) was started when spinal anesthesia was given. Due to the phenylephrine prophylaxis we expected only a minor decrease in SAP after induction of spinal anesthesia”. Q: I read this as identical information. And if so – one of the paragraphs should be deleted.\n\n“The analyses of spinal induced hypotension were based on the beat-to-beat measurements during the first 5 minutes, guided by inspection of the blood pressure curves. Based on a previously published study we expected a maximum decrease in SAP 3–4 minutes after spinal anesthesia3,7.” Q: I believe this should be moved to the Discussion. And be part of an explanation regarding why “three minutes” were chosen, over two, four, five or…. This is missing.\n\nStatistical analysis: Q: Although it seems to be a thorough explanation, I do not understand the majority. But I was intensively looking for a description of an algorithm, since this is mentioned later in the paper: P6: “The clustering algorithm puts each woman in one of two groups, so to divide them in the best way.W P8: ”The external validity of a classification algorithm based on functional data analysis and clustering, as suggested in this paper,…” P8: “The algorithm should also be evaluated using a non-invasive continuous monitoring device.” Q: What algorithm are the authors referring to?\n\nResults I have read and re-read the paper the sections regarding the three clusters. I have a hard time understanding the reasons why this subdivision is necessary and more important clinical relevant. Q: The authors compare the three clusters in supine position and after spinal. But in the present paper I am not presented with data from the supine position.\n\nFigure 4.\nQ: What are T1, T2 and T3, color-explanation is missing. Q: The. Red and black curves in Figure 3b and 4b are identical. Explanation? Q: Figure 4 text: “…T3.1”? Explanation of the colors should be added, please.\n\nDiscussion ”A statistically significant difference in mean CO was observed between the two groups of women and the group with lowest SAP had a higher CO. The difference in CO was probably of limited clinical significance but indicates that the difference in the compensation of spinal anesthesia induced vasodilatation is the mechanism behind this observation. Women are more dependent on sympathetic tone during pregnancy than in non-pregnant state and the effect of spinal anesthesia is a rapid and prominent vasodilatation, leading to a decrease in blood pressure and increase in CO7,16. All women in this study received standard phenylephrine bolus and infusion, and this prophylaxis was effective as the mean SAP was stable.” Q: Treat the patient not the picture! How were the women doing? Was this clinical relevant?\n\n“Principal component analyses have been used for making predictive models. In this. population it would be interesting to identify patients at risk of severe spinal anesthesia induced hypotension and tailor the prophylactic treatment.” Q: I would be interesting if the authors would share some ideas on how prevent hypotension – an anti-hypotensive adjusted prophylactic treatment?\n\nHeart rate variability, time domain analysis, and pulse transit time measures are all mentioned in the discussion, and maybe potentially relevant in this clinical setting. Q: However, as a reader, I cannot see the obvious link between these investigations and the three minutes supine position test. I would like to have more information, linking to the right population, and to the results of the present study in the discussion, if this information should be considered.\n\nQ: Finally, I would have preferred if the authors have been more strict on answering their Aims (P5) in the discussion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3850",
"date": "23 Jul 2018",
"name": "Leiv Arne Rosseland",
"role": "Author Response",
"response": "Thank you for the positive review of the manuscript. We will respond to the comments and questions and when we have received more referee assessments a revised manuscript will be prepared, including the editing of repeated phrases and the missing information in figure legends. Specific questions/answers: ”They were asked to assess specifically the extra pain caused by to the arterial cannulation” Q: What does this mean? A: The parturients were informed about that arterial cannulation is painful. They were asked to consider if this extra pain could be accepted before considering giving their consent. In the revised manuscript we will describe this better. “The indications for elective cesarean section were maternal request 43%, previous cesarean delivery 25%, breech presentation 8%, other obstetric reasons 13%, and other non-obstetric maternal or neonatal medical conditions 11%.” Q: Why is this important in the present study? A: The generalizability of the study relies on the sample being representative of the population and the information about the included women and the indication for cesarean is included for this reason. Statistical analysis: Q: Although it seems to be a thorough explanation, I do not understand the majority. But I was intensively looking for a description of an algorithm, since this is mentioned later in the paper: P6: “The clustering algorithm puts each woman in one of two groups, so to divide them in the best way.W P8: ”The external validity of a classification algorithm based on functional data analysis and clustering, as suggested in this paper,…” P8: “The algorithm should also be evaluated using a non-invasive continuous monitoring device.” Q: What algorithm are the authors referring to? In our paper we have used one after the other, three statistical procedures on our data, three algorithms, implemented as packages in the software language R. First we recognise that the measurements made for each women, though they come as a sequence of measures taken in discrete points, represent a function: measures of the pressure exist at any time point, and must be quite smooth. This means that measures made after each other, cannot change too much. The first algorithm estimates such a function for every women. The data for every women are then represented by such a function. The next procedure is finding the most important components of such function, by calculating the principal components and by using the first two components. At this point we have reduced the dimension of the very much, as every women is described by just two numbers (one per principal component). The third procedure is clustering: the women are now assigned to one of two clusters, so that within a cluster they are quite similar and across clusters they are quite different. The three procedures realise our algorithm. Results I have read and re-read the paper the sections regarding the three clusters. I have a hard time understanding the reasons why this subdivision is necessary and more important clinical relevant. Q: The authors compare the three clusters in supine position and after spinal. But in the present paper I am not presented with data from the supine position. A: The data from supine and left lateral position in the three clusters are presented in Supplementary Table 3.We agree that the clinical relevance of this information (the three clusters) is not obvious. Without this information one might question why we have chosen only the PCA analysis with two clusters. Was there more information hidden that would be revealed when applying three? Or four? Hence, this is kept for methodological reasons. Discussion ”A statistically significant difference in mean CO was observed between the two groups of women and the group with lowest SAP had a higher CO. The difference in CO was probably of limited clinical significance but indicates that the difference in the compensation of spinal anesthesia induced vasodilatation is the mechanism behind this observation. Women are more dependent on sympathetic tone during pregnancy than in non-pregnant state and the effect of spinal anesthesia is a rapid and prominent vasodilatation, leading to a decrease in blood pressure and increase in CO7,16. All women in this study received standard phenylephrine bolus and infusion, and this prophylaxis was effective as the mean SAP was stable.” Q: Treat the patient not the picture! How were the women doing? Was this clinical relevant? A: On average, the women were well treated with the combination of intra-venous fluid + phenylephrine prophylaxis/treatment. However, some patients will develop hypotension and the idea was to analyze if the simple supine position test could predict hypotension. The study should be regarded as an explorative assay, not a pragmatic clinical study. We believe we have demonstrated both the existence of a specific physiologic effect of supine position and that modern signal analyses may be useful in the development of prediction models. The clinical relevance of this has to be tested and proven in future trials. “Principal component analyses have been used for making predictive models. In this. population it would be interesting to identify patients at risk of severe spinal anesthesia induced hypotension and tailor the prophylactic treatment.” Q: I would be interesting if the authors would share some ideas on how prevent hypotension – an anti-hypotensive adjusted prophylactic treatment? A: Patients at risk according to this test (or further developments of the test, or other tests) should be treated state of the art to prevent hypotension. This includes lowering the local anesthetic dose injected spinally, increasing the dose of prophylactic phenylephrine (or at least included phenylephrine prophylaxis), left tilting the operating table adequately, and last but not least, consider using beat-to-beat monitoring, invasive or non-invasive monitor. We will include this information in the revised discussion. Heart rate variability, time domain analysis, and pulse transit time measures are all mentioned in the discussion, and maybe potentially relevant in this clinical setting. Q: However, as a reader, I cannot see the obvious link between these investigations and the three minutes supine position test. I would like to have more information, linking to the right population, and to the results of the present study in the discussion, if this information should be considered. A: We will like to keep this information in the discussion of different methods used and studied for prevention of spinal induced hypotension. If any of these methods prove sensitive and specific as prediction models, it is clinically relevant in the future. We will revise the discussion to make this clearer. Q: Finally, I would have preferred if the authors have been more strict on answering their Aims (P5). A: Thank you. We will try to indicate more specifically the answers to the aims in results and discussion."
}
]
},
{
"id": "36886",
"date": "08 Aug 2018",
"name": "Gianpaolo Scalia Tomba",
"expertise": [
"Reviewer Expertise Statistics in medicine",
"mathematical modelling of disease spread"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAn interesting study with potential clinical relevance for the identification of women possibly needing reinforced antihypotensive treatment during spinal anesthesia. The study identifies a group of women with stronger SAP reaction to spinal anesthesia based on pre-anesthesia SAP measurement dynamics. The study uses advanced statistical techniques such as smoothing, PCA and clustering to identify the relevant data patterns. Suppl Fig 2 shows how difficult this task would be without the statistical instruments. In the subsequent analysis, smoothing allows the signal to become clear, PCA to reduce the complexity of data and clustering to define groups. In conclusion, a study with a relevant aim, relevant data and relevant statistical analysis.\nDetailed comments\nIn the Discussion, the authors point out that a new independent study will be needed to show the reliability of the findings, a wise comment when interesting patterns are found in one data set... Maybe the authors could also comment on how this group finding could be used for an individual classification method... They hint at a individual classification algorithm, but maybe a possible realization of such an algorithm should be at least sketched, especially for a non-statistical readership.\nOtherwise, just minor details:\nAbstract,Background: \"...women with decreasing ...(SAP)...\" According to later data, all women eventually had decreasing SAP, the problem being approaching a hypotension limit... Abstract, Methods: way -> away Abstract, Conclusions: As commented above, here a \"simple test\" is mentioned, some more details would be useful later on... Background, para 1,l2: occur-> occurs Methods, para 1, l10: are published -> have been published Methods, para 3, l5: by to -> by Statistical analyses: Maybe the last sentence (about R) should be first... Statistical analyses, para 1, l10: in the rest -> in the rest of the paper Statistical analyses, para 2, l2: represented in total -> represented Statistical analyses, para 3, l6: the negation of H0 is that at least one is different, not that all are different...) Results, para 1, l3: why \"with increasing values\"? Results, para 1, l3: principle -> principal Results, para 1, l3: The whole sentence beginning on this line should be reformulated, since it is the clustering of the data (blood pressure difference curves) reduced in dimensionality by principal component analysis that allowed the identification of two or three groups of women. Results, Fig 1: it would be better, for comparability, if the two panes had the same y-axis Results, para 3: the relevance of the analysis of standard deviations is unclear, also whether these are computed on raw SAP or on first order differences. In the Suppl Fig 4, the group S2 seems clearly lower than S1. Does this mean anything? Apparently, these results lead to the clustering in 3 groups... Results, para 4: what are the relations between these 3 groups and S1/S2? Are the women in T1+T2 = S1 and T3=S2? Results, Comparison...supine/left, para 1, l7: maybe \"borderline points\" is better than \"borderline women\"... Results, Comparison...supine/left, para 1, l7: maybe Fig 2 (and the related last 3 sentences) should be in the preceding subsection... Discussion, para 1, l4: no comma (,) after Kinsella et al. Discussion, para 4, l9: requires -> require Discussion, para 5, l2: provide -> provides Discussion, para 6, l4: Hypotension -> hypotension\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36887",
"date": "13 Aug 2018",
"name": "Julian Stander",
"expertise": [
"Reviewer Expertise Applied Statistics. Computational Statistics. Data Science. The Use of R. Statistical Education"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 am pleased to have the opportunity to comment on this interesting article that has important implications for clinical practice. I am a statistician and will restrict my comments almost entirely to the statistical analysis.\n\nComments on the Statistical Analysis\nI find that the statistical analysis presented to be sound and appropriate.\n\nI very much appreciate the use of techniques from functional data analysis to extract understanding from these data. I was curious as to how the smoothing parameter lambda was selected. I understand that you have performed a thorough study to document the robustness of the results to the choice of lambda. I would, however, like to see a very brief discussion about whether lambda could be selected automatically, perhaps working in the Bayesian framework. Implementing and performing such an analysis is clearly beyond the scope of this paper, and I am not asking for it to be done.\nI did not understand from the paragraph beginning \"We used the multivariate paired Hotelling's T^2 test...\" on page 4 to what S1 and S2 refer. I think that S1 and S2 should be defined in the previous paragraph.\nFor the test H_0: mu_1 = mu_2 = mu_3, I think that H_1 is not precisely stated. I would state it simply as H_1: not H_0.\n\nI have no objection to you performing Hotelling's T^2 tests. I was, however, wondering whether there were specific tests for use with functional data. Perhaps your tests are specific to functional data. If they are not, a little discussion of this issue would be helpful.\nOn page 6 you state that \"The clustering algorithm puts each woman in one of two groups, so to divide them in the best way\". In what way is it best please? It is not quite clear to me from Figure 2, how you would choose the number of groups/clusters. I was wondering whether this could be done in some automatic way. Again, implementing and performing such automation is beyond the scope of this paper, but a brief discussion about it would be helpful and interesting.\nI was left a little perplexed by the paragraph beginning \"It was evident that the principal component analysis...\". I think that that paragraph needs to be revised with perhaps some more detail. There could be a relationship here with the comments above about automating certain choices. Further discussion would be valuable.\n\nGeneral Comment\nIn the Abstract, there should be no comma between \"observation\" and \"can\".\nI do not think that the Aims read like aims. I would like to see these re-written and then better referred to later in the article in a way that allows me to see that they have been satisfied.\nI think that it would be helpful if you were to remind us about the meaning of the colours on page 7.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1028
|
https://f1000research.com/articles/7-1024/v1
|
09 Jul 18
|
{
"type": "Research Article",
"title": "Chemical characterization of bamboo leaves (Gigantochloa albociliata and Dracaena surculosa) by sodium hydroxide treatment",
"authors": [
"Nadiah Ameram",
"Muhammad Afiq Che Agoh",
"Wan Farhana W. Idris",
"Arlina Ali",
"Muhammad Afiq Che Agoh",
"Wan Farhana W. Idris",
"Arlina Ali"
],
"abstract": "Background: Dracaena surculosa and Gigantochloa albociliata leaves are different in morphology and appearance. Sodium hydroxide (NaOH) is widely used in pulping of wood for making paper or regenerated fibers. NaOH is used to separate lignin from cellulose fibers, and this treatment is needed to identify the characteristics of leaves. This study was conducted in order to investigate the characteristics of D. surculosa and G.albociliata species under NaOH treatment.\n\nMethods: NaOH was applied to the leaves for 8 hours. Treated and untreated bamboo leaves were analysed using FTIR analysis, in order to identify the presence of functional groups in the leaves. Results: It was observed that these two species not only differ physically, but also chemically. The presence of OH, CH and alkynes functional groups in the leaf sample indicates that the species share similar properties but have a slight difference in the molecular bonds. From the morphological observation of D. surculosa and G. albociliata leaves, they are slightly different in terms of leaf appearance and characteristics. G. albociliata have thicker leaves compared to D. surculosa, and NaOH treatment shows that D. surculosa leaves are harder to dissolve into the solvent. Scanning electron microscope (SEM) analysis of these two species shows the initial structure of fibres in the leaves are intact but after NaOH treatment, the fibres are ruptured and appear in non-uniform shapes. Conclusions: The initial morphology of G. albociliata and D.surculosa is different in color and appearance. However after NaOH treatment, the color becomes almost the same. Regarding SEM analysis, after NaOH treatment the morphology of the bamboo leaves completely changes. Therefore, it can be concluded that the process of hemicellulose removal had occurred during treatment. The results show that lignin has been removed by NaOH treatment to enhance the characteristics of the bamboo leaves from different species.",
"keywords": [
"Bamboo leaves",
"Sodium hydroxide",
"Chemical treatment",
"FTIR",
"SEM"
],
"content": "Introduction\n\nBamboo is one of the most common plants available in Malaysia and is known for its multiple uses for construction, transportation and cultural purposes. About 70 species of bamboo (50 in Peninsular Malaysia, 30 in Sabah and 20 in Sarawak) can be found in Malaysia1. Being abundant throughout the whole country, bamboo can be treated as a source of income. Its leaf is claimed to have medicinal benefits among traditional practitioners2. Bamboo leaves, also known as Lophaterum, and bamboo shavings are commonly used as relievers for stomach aches, cooling effects and also help to counter the negative flow of qi3.\n\nAsian populations include bamboo in their dishes, especially Chinese individuals. In China and Southeast Asia, bamboo had been used as source of food and medicine for a long time4. Bamboo grows one third faster than the fastest growing tree5 and some species can grow up to 60 meters high. Bamboo can grow up to 1 meter per day and therefore it is easily accessible in a minimal amount of time. The optimum age for bamboo to be harvested is at 3–5 years old. For thousands of years, bamboo had been part of human’s diet.\n\nIn addition, the Indian elephant and the Giant Panda of China also utilize bamboo leaves as their main food source, which become their exclusive meal. The skeletal system of the panda is incredibly strong, and flexible and it can be related to their diet of bamboo leaves, which contains a great amount of silica6. Usage of silica includes silica as dessicants, as a cement, and in the production of tires. Silica gel desiccants are used as pharmaceutical-desiccants/silica gels to promote drying by absorbing water vapor and gases, like oxygen and hydrogen in humidity. It is very important in pharmaceutical industries as humidity and moisture can ruin pharmaceutical products7. Silica has been utilized as a cement substitution, as silica exists as quartz in the silica sand. It has also been used as a fractional substitution of cement during solid development8. In tires, mixing silica with carbon improves the tire performance on the road9, as silica helps to provide flexibility through low-heat build-up.\n\nDracaena surculosa is a fast growing species of bamboo and is therefore preferable as an indoor plant. This species also prefers partial shade for growth. Exposure to bright lights may cause the plant to lose or shed its leaves. A full grown D. surculosa leaf could be as long as 7 – 12 cm long. The surface of the leaf is leathery and smooth with obvious patterns of yellow or gold spots (Figure 1).\n\nPopular for its shoots, Gigantochloa albociliata is commonly planted and harvested as a food source and decoration. The shoots are edible and in Japan canned bamboo shoots are used. This bamboo species can easily grow in dry tropical mixed forest from low to medium elevations. The leaves are linear - lanceolate with the width of 15–20cm × 2–25cm (Figure 2). In the dry season, the leaves are shed10.\n\nThe abundance of bamboo in Malaysia may give a great advantage to the nation. However when harvested, not all parts of the bamboo are fully utilized, the leaves being especially discarded. This research aims to determine the organic compounds in bamboo leaves in two species, D. surculosa and G. albociliata after sodium hydroxide (NaOH) treatment. Along with sodium sulfide, NaOH is used to separate lignin from cellulose fibers in the craft process of making paper. Thus, this treatment is needed to separate lignin from fibers in order to identify the characteristics of leaves and was therefore used in the present study.\n\n\nMethods\n\nSodium hydroxide (Quality Reagent Chemical [QREC], 1.0 M), nitric acid (QREC, 1.0 M) and toluene (QREC, 0.1 M) were utilized in this study. Bamboo leaves were collected with special permissions from Lundang Committee of Kg Lundang (Dracaena surculosa: 6°06'00.0\"N 102°15'33.1\"E) and Director of UMK Jeli Campus for UMK Agropark Kelantan (Gigantochloa albociliata: 5°44'46.0\"N 101°51'58.3\"E). All other chemicals and reagents used were of the highest commercially available purity.\n\nThe bamboo leaves from G. albociliata and D. surculosa (5 g sample) was stirred in 125mL of 1.0M NaOH for 8 hours using a magnetic stirrer. The solution was filtrated (filer paper grade 1, 11 µm) to separate the solute (bamboo leaves) and the aqueous solution (sodium silicate). The solution was then titrated with 1.0M nitric acid to obtain less than 3 pH in order to obtain the gel of suspended silica. Subsequently, this solution was centrifuged at 4000 rpm for 10 minutes. Then, the solution was left for 4 days to allow the silica gel to form. The silica gel were then filtered (filter grade 1, 11 µm) and extracted using 50mL of toluene. Finally, the solution obtained was dried over anhydrous sodium sulfate (QREC 1.0 M). This was then left overnight at room temperature. If any liquid remains, this should be collected and discarded. The final product will be analyzed using Fourier-transform infrared spectroscopy (FTIR). FTIR analysis is used to identify functional groups in a molecule by producing an infrared absorption spectrum. The spectrums were identified using Perkin Elmer Spectrum 100 FT-IR Spectrometer with range spectrum (wavelength): 4000–400/cm. Scan rate: 16, 32. Universal ATR sampling accessory (not using KBr) was employed to prepare the sample.\n\nEach bamboo leaf sample was analysed using SEM to characterize the morphology of the specimen before and after treatment with NaOH. 30 g of each bamboo leaf was left untreated and ground to obtain a powder. In addition, 30g of bamboo leaves were treated with NaOH (30 mL) by stirring for 8 hours using a magnetic stirrer. The final solution was filtered using filter paper to obtain the solutes. The solutes were dried thoroughly before being ground into a powder to be analysed using SEM. The voltage used was between the ranges 1.5kV to 2.0kV. Secondary electron images were obtained.\n\n\nResults\n\nDuring NaOH treatment, D. surculosa leaves exhibit a rapid change in colour and texture. The initial colour for this sample is light green and yellowish. After being stirred with NaOH, the colour started to change to dark brown, and after 8 hours of treatment, the colour turned to very dark brown. A reason for the rapid colour change is due to the thickness of the leaves, as any pigment inside the leaf would be quickly released as the leaves are thin. In addition, after 8 hours of treatment, the leaves almost completely dissolved into the solution of NaOH. This makes it hard for the leaves to be used in the next treatment of solvent extraction.\n\nFor G. albociliata leaves, there was a slower reaction during treatment of NaOH. After a few hours of treatment, the colour of the leaves started to change to dark green (original color light brownish green) and after 8 hours, the colour turns to almost black. In contrast to D. surculosa, the leaves did not dissolve into the solution.\n\nFTIR spectra of changes in functional group of both untreated and treated (NaOH) bamboo leaves of G. albociliata is illustrated in Figure 3.\n\n(Graphs are representative of 2 repeated experiments.)\n\nThe absorption band at 3293.32cm-1 indicates that there is an O-H stretching vibrations between cellulose and hemicellulose11. Meanwhile the 2913.51cm-1 for treated and untreated bamboo leaves demonstrate the presence of CH2 groups. The band also contains functional group of alkynes from the peak at 2913.51 cm-1. In accordance to the presence of this peak, it also indicates that before the treatment with NaOH, pectin and waxes are present in the bamboo leaves sample of G. albociliata12. Different results are achieved with bamboo leaves treated with NaOH where the peaks have almost disappeared. This shows that hemicellulose is present and is gradually removed from the sample during treatment. The lists of functional groups are shown in Table 1.\n\nFTIR spectra of changes in functional group of untreated and treated D. surculosa leaf samples using 1.0M NaOH is illustrated in Figure 4.\n\n(Graphs are representative of 2 repeated experiments.)\n\nThe absorption band at 3294.48 cm-1 in D. surculosa leaves indicates that there is a O-H stretching vibrations between cellulose and hemicellulose13. Whereas the 2885.64 – 2895.44 cm-1 for treated and untreated bamboo leaves sample show the presence of CH2 groups. The band contains functional group of alkynes with triple bonds from the peak at 2117.32.51 cm-1. Different results are achieved with bamboo leaves treated with NaOH where the peaks had disappeared. The list of functional groups are shown in Table 2.\n\nThe study of morphology requires SEM analysis before and after treatment is done, as the treatment may affect the morphology of the species14. Figure 5 shows a 400x magnification of G. albociliata leaf surface before treatment with NaOH.\n\nWhite arrows shows the fiber bonding in the leaf.\n\nFrom this figure, it can be observed that the bonding between fibers in the leaf are still intact as shown by the arrows in Figure 5. Fibers play an important role in improving strength, providing higher initial modulus and reducing extensibility by improving the interfacial adhesion between them15.\n\nFrom Figure 6, it can be observed that the leaves had no particular shape and the morphology changes completely after being treated with NaOH. The degradation of hemicellulose may occur during treatment, which causes the leaf to be dissolved in the solvent itself. From this observation, it can be concluded that NaOH is one agent that can degrade the molecular structure of bamboo leaves. Furthermore, high concentration of NaOH used in this test (1.0M) may contribute to the rapid degradation of the hemicellulose. Another observation using SEM analysis shows the structure of the fibres are in non-uniform shape.\n\nThe leaves of untreated D. surculosa still have bonds between the hemicelluloses. After the grinding process, the leaves are crushed but remains in solid form, as seen in Figure 7, and are not powdery like G. albociliata.\n\nThe white ring shows that the fiber bond which is intact but not clearly shown.\n\nAfter treatment with 1.0M NaOH, the bamboo leaves become partially dissolved into the solvent, as can be seen in Figure 8. The final colour of the solution also changes to dark brown. This change of colour is caused by the removal of hemicellulose in the leaf itself16.\n\n\nDiscussion\n\nFrom the FTIR analysis, the functional group found in the specimens can be identified based on the presence of the bands in the graphs. It can be concluded that the process of treatment and extraction had contributed to the organic compounds present in the samples of different species6. Among the two samples, G. albociliata and D.surculosa, the initial morphology is different in two factors, colour and appearance. The initial colour of G. albociliata and D.surculosa is yellowish green and dark green, respectively. However after the treatment, the colour becomes almost the same. As for the SEM analysis, the treatment shows the morphology of the bamboo leaves completely changes due to the treatment with NaOH. Due to change in colour of NaOH, it can be concluded that the process of hemicellulose removal had occurred during the treatment. Bamboo is truly interesting to study as it has a unique anatomy and a superproductive behavior. Rhizomes are an essential part of bamboo anatomy. This study will be able to provide new data regarding the characteristics of G. albociliata and D. surculosa bamboo leaves after NaOH treatment\n\n\nData availability\n\nDataset 1: Replicates of FTIR analysis; bamboo leaves after NaOH treatment. DOI, http://dx.doi.org/10.5256/f1000research.15036.d20885917",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe thank UMK research short term grant SGJP for funding (number R/SGJP/A13.00/00692A/001/2018/000498).\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\nBrandis D: V. Remarks on the Structure of Bamboo Leaves. Transactions of the Linnean Society of London. 2nd Series: Botany. 1907; 7(5): 69–92. Publisher Full Text\n\nChen Q, Endo T, Wang Q: Characterization of Bamboo after Ionic Liquid-H2O Pretreatment for the Pyrolysis Process. BioResources. 2015; 10(2): 2797–2808. Publisher Full Text\n\nDeng ZH, Cheng CG, Wang XL, et al.: Preconcentration and Determination of Perfluoroalkyl Substances (PFASs) in Water Samples by Bamboo Charcoal-Based Solid-Phase Extraction Prior to Liquid Chromatography–Tandem Mass Spectrometry. Molecules. 2018; 23(4): pii: E902. PubMed Abstract | Publisher Full Text\n\nLv Z, Dong J, Zhang B: Rapid Identification And Detection Of Flavonoids Compounds From Bamboo Leaves By Lc-(Esi)-It-Tof/ms. BioResources. 2012; 7(2): 1405–1418. Publisher Full Text\n\nMarch R, Clark L: Sun-shade variation in bamboo (Poaceae: Bambusoideae) leaves. Telopea. 2011; 13(1–2): 93–104. Publisher Full Text\n\nNirmala C, Bisht MS, Laishram M: Bioactive compounds in bamboo shoots: Health benefits and prospects for developing functional foods. Int J Food Sci Technol. 2013; 49(6): 1425–1431. Publisher Full Text\n\nNugroho N, Ando N: Development of structural composite products made from bamboo I: Fundamental properties of bamboo zephyr board. J Wood Sci. 2000; 46(1): 68–74. Publisher Full Text\n\nGuo H, Fan KM, Qian JQ: Purification of Flavone C-Glycosides from Bamboo Leaves by Macroporous Adsorption Resin. Asian J Chem. 2014; 26(21): 7221–7225. Publisher Full Text\n\nRambo C, Martinelli J: Synthesis and Characterization of SiC from Bamboo. Key Eng Mat. 2001; 189–191: 9–15. Publisher Full Text\n\nSugiyanto K: Physical and chemical modification of the bamboo species Dendrocalamus asper: Submitted in total fulfilment of the requirement of the degree of Doctor of Philosophy. Creswick: University of Melbourne, 2011. Reference Source\n\nWang L, Gao B, Peng C, et al.: Bamboo leaf derived ultrafine Si nanoparticles and Si/C nanocomposites for high-performance Li-ion battery anodes. Nanoscale. 2015; 7(33): 13840–13847. PubMed Abstract | Publisher Full Text\n\nWang Q, Takahashi H, Takahashi Y, et al.: Characterization of liquefied waste bamboo and white-rotted wood. Energy and Sustainability VI. 2015; 195: 63–74. Publisher Full Text\n\nMarui A, Omoto S: Water consumption and retention of crushed bamboo as an agricultural organic material. 2014 ASABE Annual International Meeting. 2014; Publisher Full Text\n\nYang H: Study on the hemicellulose of Formosan bamboo (sinocalanus latiforus munre). Place of publication not identified: Agricultural Chemistry Department, National Taiwan University. (n.d.). Reference Source\n\nZhang Y, Ding X: The Bio-Antioxidative Activity of Functional Factors in Bamboo Leaves. Food for Health in the Pacific Rim. 2008; 266–273. Publisher Full Text\n\nZhu G, Zou XP, Cheng J: Synthesis and Characterization of Bamboo-Like Carbon Nanotubes. Adv Mat Res. 2008; 47–50: 355–358. Publisher Full Text\n\nAmeram NB, Afiq Che Agoh M, Idris WFW, et al.: Dataset 1 in: Chemical characterization of bamboo leaves (Gigantochloa albociliata and Dracaena surculosa) by sodium hydroxide treatment. F1000Research. 2018. Data Source"
}
|
[
{
"id": "35854",
"date": "18 Jul 2018",
"name": "Noorfatimah Yahaya",
"expertise": [
"Reviewer Expertise Analytical Chemistry"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nRef: Chemical characterization of bamboo leaves (Gigantochloa albociliata and Dracaena surculosa) by sodium hydroxide treatment\n\nThis article describes a chemical characterization of bamboo leaves (Gigantochloa albociliata and Dracaena surculosa) through sodium hydroxide treatment. The topic is interesting, but it requires a revision according to the following comments:\nIntroduction: The authors should state the novelty and significant of the study in their manuscript. Methods: ‘All other chemicals and reagents used were of the highest commercially available purity’. Instead, authors should state purity of each chemical and reagent used in the study. Methods: FTIR should be in separate section from solvent extraction method. Details description of FTIR and SEM instruments should be included in text. Results: FTIR spectrum should be improved. Only important peaks should be labelled in the spectrum. The spectrum should be improved in terms of resolution. Results: SEM images should be combined together, so a clear comparison could be made. Discussion: More discussion and explanation should be added in text to increase scientific values of the article. CONCLUSIONS should be included in the text.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
},
{
"id": "35857",
"date": "03 Aug 2018",
"name": "Nur Nadhirah Mohamad Zain",
"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\nPoor introduction – introduction is one of the important thing in the manuscript. In the introduction part, the authors should discuss the ‘fibers’/ cellulose/hemicellulose or pigment component inside the leaf. The major constituent inside the leaf. Besides that, the authors should discuss about the treatment of NaOH and previous work regarding various type of treatment. Insufficient data on the characterization part – the author should carried out more characterization for supportive data for morphological characterization. i.e atomic force microscopy (AFM) analysis. How did you determine/indicate the decolourisation occur? Please include the image that show colour changes for both species before and after treatment with NaOH.\n\nPage 1 – Abstract – Methods – Please include the SEM analysis of leaves to observe morphology after and before treatment of NaOH. Page 4 – The authors mention that “the solutes were dried thoroughly before being ground into a powder to be analysed using SEM”. Unfortunately, there is no condition or sample preparation included in the manuscript. Page 4 – FTIR analysis – Please change “….treated (NaOH) bamboo…” to “ ….treated (1.0 M NaOH) bamboo…”. Page 4 – FTIR analysis – What does it meant by “Different results are achieved…..the peaks have almost disappeared.” The peak is still presented after treated with NaOH but the peak intensity is decreased?. Besides that, the authors did not explain clearly regarding each peak is referred to which part structure of constituent. i.e. CH2 and alkenes (C double bond). Based on the Fig. 3 and Fig. 4, I could not find any differences of spectrum in the range of 4000 – 2000 cm-1. There no conclusion provided in the manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1024
|
https://f1000research.com/articles/7-1013/v1
|
06 Jul 18
|
{
"type": "Research Article",
"title": "In-hospital cardiac arrest resuscitation performed by the hospital emergency team: A 6-year retrospective register analysis at Danderyd University Hospital, Sweden",
"authors": [
"Hedwig Widestedt",
"Jasna Giesecke",
"Pernilla Karlsson",
"Jan G. Jakobsson",
"Hedwig Widestedt",
"Jasna Giesecke",
"Pernilla Karlsson"
],
"abstract": "Background: Cardiac arrest requires rapid and effective handling. Huge efforts have been implemented to improve resuscitation of sudden cardiac arrest patients. Guidelines around the various parts of effective management, the chain of survival, are available. The aim of the present retrospective study was to study sudden in-hospital cardiac arrest (IHCA) and the outcomes of emergence team resuscitation in a university hospital in Sweden. Methods: The Swedish Cardiopulmonary Resuscitation Registry was used to access all reported cases of IHCA at Danderyd Hospital from 2012 through 2017. Return of spontaneous circulation (ROSC), discharge alive, 30-day mortality and Cerebral Performance Scales score (CPC) were analysed. Results: 574 patients with cardiac arrests were included in the study: 307 patients (54%) had ROSC; 195 patients (34%) were alive to be discharged from hospital; and 191 patients (33%) were still alive at day-30 after cardiac arrest. Witnessed cardiac arrests, VT/VF as initial rhythm and experiencing cardiac arrest in high monitored wards were factors associated with success. However, 53% of patients’ alive at day-30 had a none-shockable rhythm, 16% showed initially a pulseless electrical activity and 37% asystole. CPC score was available for 188 out of the 195 patients that were alive to be discharged: 96.5% of patients where data was available had a favourable neurological outcome, a CPC-score of 1 or 2 at discharge, and only 6 of these patients had a CPC-score of 3 or higher (3%). Conclusions: One third of patients with sudden IHCA were discharged from hospital and alive at day-30, a clear majority without cognitive deficit related to the cardiac arrest. High monitored care, witnessed cardiac arrest and shockable rhythm were factors associated with high success; however, more than half of surviving patients had initially a none-shockable rhythm.",
"keywords": [
"Cardiac Arrest",
"in-hospital resuscitation",
"CPR",
"30-day mortality"
],
"content": "Introduction\n\nThe importance of prompt recognition of cardiac arrest and initiation of cardio-pulmonary resuscitation has been shown repeatedly1,2. The chain of survival, prompt recognition, early/bystander cardiopulmonary resuscitation (CPR) and early defibrillation is indeed of outmost importance3. Efforts to improve the results from out-of-hospital have been implemented and our hospital has likewise put training efforts into basic and advanced CPR.\n\nThe aim of the present retrospective register project was to study sudden in-hospital cardiac arrest (IHCA) and the outcomes of emergence team resuscitation at a hospital in Sweden.\n\n\nMethods\n\nThis is a retrospective single-centre register study in which the Swedish CPR Registry was used to access all reported cases of IHCA in Danderyd Hospital, Stockholm, Sweden. The study protocol was approved by Stockholm Ethical Review Board (EPN; 2017/4:10 approved 2017-11-08, Annika Sandström). Patient informed consent is not required for register studies in accordance with Swedish research regulations.\n\nAll reported cases of IHCA at Danderyd Hospital where CPR was initiated, from January 1st 2012 to December 31st 2017, were included in the study.\n\nPlace of cardiac arrest, witnessed cardiac arrest, bystander CPR, time to initiated CPR, initial rhythm, number of defibrillations, patients with return of spontaneous circulation (ROSC), patients discharged from hospital, Cerebral Performance Scales (CPC) score of discharged patients and 30-day mortality was studied.\n\nData is presented as mean and standard deviation and frequencies as applicable. Differences has been studied by Student’s t-test and ANOVA for continuous variables and Chi sqaured test for category data. A p<0.05 has been considered statistically significant. Statistics has been calculated with SPSS Statistics® for Macintosh version 24 (Armonk, New York, USA) and Microsoft Excel © 2017 version 16.9.\n\n\nResults\n\nA total of 574 patients with sudden IHCA were included in the study: 340 males and 234 females, with a mean age of 73 ± 14 years: 72 ± 13 for males and 75 ± 14 for the females (p<0.05).\n\nThe most common place for a sudden cardiac arrest was the Coronary Care Unit (CCU) followed by cardiology and medical wards. A majority (84%) of the cardiac arrests were witnessed, and bystander CPR was initiated within one minute in 96% of cardiac arrest cases. The most common initial rhythm was asystole (n=215) and the least common was VT/VF (n=147). The highest prevalence of VT/VF (57%) was seen in the percutaneous coronary intervention lab followed by the CCU and Intensive Care Unit (ICU) (33% and 27%).\n\nIn total, 333 (55.5%) of cardiac arrest patients were successfully resuscitated and had ROSC: 195 patients (34%) were discharged from hospital and 191 (33%) were still alive at day-30 after cardiac arrest (33%), see Figure 1.\n\nROSC, return of spontaneous circulation.\n\nThe highest 30-day survival rate was seen in patients with cardiac arrest in the PCI lab (61%), with the next to highest 30-day survival rate (46%) seen in the CCU.\n\nShockable rhythm was associated with success: CCU, VT/VF alive at day-30 had 21 out of 28 patients (75%); PCI, VT/VF alive at day-30 had 26 out of 35 patients (74%); and ICU, VT/VF alive at day-30 had 7 out of 12 patients (58%). Overall 89 out of the 167 patients (53%) alive at day-30 had an initial none-shockable rhythm. Age had an impact: patients alive at day-30 were significantly younger than those who were not alive at day-30 (69 vs 75 years; p=0.001) (Table 1).\n\nCA cardiac arrest, CPR cardio-pulmonary resuscitation, PEA pulse-less electric activity, VT ventricular tachycardia, VF ventricular fibrillation.\n\nP< 0.001 **, p< 0.05 *alive vs- dead day-30\n\nCPC-score was available for 188 out of the 195 patients that were alive to be discharged (96%). In total, 96.5% of patients where data was available had a favourable neurological outcome after cardiac arrest, i.e. a CPC score of 1 or 2 at discharge.\n\n\nDiscussion\n\nWe found that one third of patients suffering sudden IHCA were alive at day-30 and that patients alive to be discharged did not experience significant impairment of cognitive function. A majority of cardiac arrests were witnessed cardiac arrest and CPR had been initiated within 1–2 minutes. Having VT/VF as an initial rhythm and a lower age of the patient increased the chance of survival. However, it is worth noticing that more than half of the surviving patients had a non-shockable initial rhythm.\n\nA previous study at Danderyd Hospital in the late 1980s found only 9 out of 61 IHCA patients were alive to be discharged (15%)4. The survival rate seen in our study is higher than that presented from a study in Ireland on in-hospital resuscitation in 2011, one year before the start of our study, which found a 27% survival rate of discharge5. The average survival rate in our study is also higher than the survival rate reported from a US survey of in-hospital resuscitation including a total of 838,465 patients6. Data analysed from the Nationwide Inpatient Sample databases between 2003 and 2011 showed a 24.7% overall survival to hospital discharge6. A study conducted in Finland between 2009 and 2011, including 279 adult IHCA patients attended by the medical emergency team in a university hospital's general wards, found a 180-day survival rate of 19%7. They commented on the importance of shockable primary rhythm, monitored/witnessed event and low comorbidity score for survival. One should acknowledge that our study covered the period 2012 to 2017 and all hospital wards, including coronary and general intensive care departments. A study from China revealed a low survival rate where only 9.1% of patients were discharged alive8. Our results are however in line with a previous studies from Sweden. Herlitz et al. found a 43% survival rate for discharge among cardiac arrest patients suffering cardiac arrest in wards with monitoring facilities, and a 31% survival rate among cardiac arrest patients in general wards. They also found cerebral function to be favourable in most patients9,10.\n\nOur study does have limitations. We did not study the causes of cardiac arrest. It should be acknowledged that cardiac arrest cases throughout the hospital were included in the study, not only on cardiac arrest in high dependency wards and in patients with heart disease. There is missing data for initial rhythm in about 10% of cases, which means that conclusions concerning prevalence of different cardiac rhythms must be performed with caution. It should also be acknowledged that we do not have data on time to defibrillation.\n\nTo conclude, one third of IHCA patients resuscitated by the emergency team could be discharged alive and were still alive at day-30 in our study cohort, a majority without signs of cognitive impairment related to cardiac arrest. Most cardiac arrests were witnessed and CPR had been initiated within minutes. We found initial shockable rhythm VT/VF to be a factor related to successful CPR, which is similar to what has been shown for out-of-hospital CA; however, it should be noted that more than half of survivors had a none-shockable initial rhythm.\n\n\nData availability\n\nThe data has been retrieved from the Swedish CPR register (https://www.hlr.nu/svenska-hlr-registret/). This is a national database, supported by the Swedish and European Resuscitation Councils. The data can be retrieved by request from CPR register (https://shlrsjh.registercentrum.se/) following Ethical Review board approval on application (https://www.epn.se/en/start/).",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by the Department of Anaesthesia & Intensive Care, Danderyds Hospital. No external funding was provided.\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\nHasselqvist-Ax I, Riva G, Herlitz J, et al.: Early cardiopulmonary resuscitation in out-of-hospital cardiac arrest. N Engl J Med. 2015; 372(24): 2307–15. PubMed Abstract | Publisher Full Text\n\nMalta Hansen C, Kragholm K, Pearson DA, et al.: Association of Bystander and First-Responder Intervention With Survival After Out-of-Hospital Cardiac Arrest in North Carolina, 2010-2013. JAMA. 2015; 314(3): 255–64. PubMed Abstract | Publisher Full Text\n\nOng MEH, Perkins GD, Cariou A: Out-of-hospital cardiac arrest: prehospital management. Lancet. 2018; 391(10124): 980–988. PubMed Abstract | Publisher Full Text\n\nJakobsson J, Dahlqvist M, Rehnqvist N: Resuscitation of hospitalized patients. J Intern Med. 1990; 227(1): 15–8. PubMed Abstract | Publisher Full Text\n\nO'Sullivan E, Deasy C: In-hospital Cardiac Arrest at Cork University Hospital. Ir Med J. 2016; 109(1): 335–8. PubMed Abstract\n\nKolte D, Khera S, Aronow WS, et al.: Regional variation in the incidence and outcomes of in-hospital cardiac arrest in the United States. Circulation. 2015; 131(16): 1415–25. PubMed Abstract | Publisher Full Text\n\nTirkkonen J, Hellevuo H, Olkkola KT, et al.: Aetiology of in-hospital cardiac arrest on general wards. Resuscitation. 2016; 107: 19–24. PubMed Abstract | Publisher Full Text\n\nShao F, Li CS, Liang LR, et al.: Incidence and outcome of adult in-hospital cardiac arrest in Beijing, China. Resuscitation. 2016; 102: 51–6. PubMed Abstract | Publisher Full Text\n\nHerlitz J, Bång A, Aune S, et al.: Characteristics and outcome among patients suffering in-hospital cardiac arrest in monitored and non-monitored areas. Resuscitation. 2001; 48(2): 125–35. PubMed Abstract | Publisher Full Text\n\nHerlitz J, Aune S, Bång A, et al.: Very high survival among patients defibrillated at an early stage after in-hospital ventricular fibrillation on wards with and without monitoring facilities. Resuscitation. 2005; 66(2): 159–66. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "37733",
"date": "17 Sep 2018",
"name": "Søren Mikkelsen",
"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 aim of the study was to investigate sudden in-hospital cardiac arrests. The authors report that one third of the patients that suffer in-hospital cardiac arrest are discharged alive form the hospital. The authors describe that the chances of survival increases with shockable rhythm and lower age. Furthermore, and somewhat surprising, the authors report that more than half of the patients initially had a non-shockable rhythm.\nGeneral comments: The study is a single centre study in which the authors report a survival rate following in-hospital cardiac arrest somewhat higher than described in most other studies. Between the lines, this reviewer gets the notion that the medical emergency team is responsible for the favourable results. The study is, however, devoid of speculations on why this apparent increase in survival is found at Danderyd Hospital compared with other hospitals. This is somewhat disappointing. There is ample literature available to support discussions on the potential benefits of applying medical emergency teams and I feel that a discussion on caused for survival should have been discussed.\nThe authors report that more than half of the patients that survive to discharge are met with an initial non-shockable rhythm. This is an interesting finding and should have been discussed. As it stands, the paper is just a reporting on rather favourable outcomes following in-hospital cardiac arrest and is not delving deeper into causes and explanations.\nSpecific comments: One sentence is rather difficult to comprehend: “The aim of the present retrospective study was to study sudden in-hospital cardiac arrest (IHCA) and the outcomes of emergence team resuscitation in a university hospital in Sweden.” Should the sentence read: “The aim of the present retrospective study was to study sudden in-hospital cardiac arrest (IHCA) and the cerebral outcomes following resuscitation by an emergency team in a university hospital in Sweden.”\n\nStatistics: When applying means, standard deviations and t-test, the tested variables should follow a normal distribution. Have the authors assured that?\nReferences: The list of references is rather small and do not support a deeper discussion of results. Not one reference to the concept of medical emergency teams is made. Although apparently not the scope of this paper, a more comprehensive discussion of medical emergency teams would have been in order and the list of references should have reflected that.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "38715",
"date": "08 Oct 2018",
"name": "David A. Pearson",
"expertise": [
"Reviewer Expertise cardiac arrest"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 retrospective study demonstrates outcomes after in-hospital cardiac arrest as abstracted from a Swedish cardiac arrest registry. This study demonstrates a significant number of non-shockable initial arrest patients with a good neurological outcome, particularly those in PEA. To make this study more robust, it is essential to clarify the code response team, both personnel and process, as well as to ensure consistent definition of cardiac arrest, which in the study was defined as “CPR was initiated”, which needs more clarification (i.e., chest compressions initiated). Finally, there are multiple studies demonstrating similar positive results after in-hospital cardiac arrest. This study represents another study demonstrating the importance of resisting premature prognostication after in-hospital cardiac arrest, regardless of initial rhythm, as many survive with meaningful neurological recovery.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1013
|
https://f1000research.com/articles/7-722/v1
|
11 Jun 18
|
{
"type": "Research Article",
"title": "A cross-sectional study of the use and effectiveness of the Individual Development Plan among doctoral students",
"authors": [
"Nathan L. Vanderford",
"Teresa M. Evans",
"L. Todd Weiss",
"Lindsay Bira",
"Jazmin Beltran-Gastelum",
"L. Todd Weiss",
"Lindsay Bira",
"Jazmin Beltran-Gastelum"
],
"abstract": "Background: The Individual Development Plan (IDP) was introduced as a tool to aid in career planning for doctoral trainees. Despite the National Institutes of Health and academic institutions creating policies that mandate the use of IDPs, little information exists regarding the actual use and effectiveness of the career planning tool. Methods: We conducted a multi-institutional, online survey to measure IDP use and effectiveness. The survey was distributed to potential respondents via social media and direct email. IDP survey questions were formatted using a five-point Likert scale (strongly agree, agree, neutral, disagree and strongly disagree). For data analysis purposes, responses were grouped into two categories (agree versus does not agree/disagree). The data were summarized as one-way frequencies and the Pearson Chi-square test was used to determine statistical significance. Results: Usage of the IDP among doctoral students was low and the tool produces minimal effectiveness with regard to the perception of whether it is helpful to one’s career development. Further, our data suggests that the IDP is most effective when doctoral students complete the tool with faculty mentors with whom they have a positive relationship. Respondents who are confident about completing their doctoral training and their post-training career plans, and who take advantage of career development resources at their institution are also more likely to perceive that the IDP is useful for their career development. Conclusion: Given the nuanced use and effectiveness of the IDP, we call for more research to determine why IDP use and effectiveness is low, exactly how IDPs are being used, and whether there are unintended negative consequences created through the use of the tool. Furthermore, we recommend an enhancement of career development infrastructure that would include mentorship training for faculty in order to provide substantially more career planning support to doctoral trainees.",
"keywords": [
"biomedical research",
"career development",
"career planning",
"doctoral students",
"individual development plans",
"PhD training"
],
"content": "Introduction\n\nThe spotlight is bright today on the sustainability of the biomedical enterprise, especially regarding the support and general career outcomes of early career investigators and trainees1–3. There is a significant supply of PhDs and a weak market demand for faculty positions, and the majority of doctoral trainees are moving into non-faculty positions in academia, industry, government agencies, or entrepreneurship4,5. Greater career development support has been suggested by many as a key area of need to better support PhDs entering into this diverse workforce6.\n\nIn 2002, the U.S. Federation of American Societies for Experimental Biology created the Individual Development Plan (IDP) as a multi-component career planning worksheet that guides doctoral trainees through a self-assessment of skills, provides a platform for the exploration of scientific career paths, aids in the development of short and long term careers goals, and prompts the creation of action plans to achieve those goals7. In 2012, Science Careers launched a free online version of the IDP called myIDP8. In 2014, following the recommendation of the National Institutes of Health (NIH)’s Biomedical Research Workforce Working Group, the NIH implemented a policy requiring the reporting of IDP use by graduate students and postdoctoral researchers in grant progress reports9. Subsequently, many academic institutions have instituted policies dictating the use of the IDP for PhD trainees. Despite these policy implementations, studies investigating the use and effectiveness of the IDP have been limited to one report that was published in 2014, which studied 233 current postdoctoral researchers, 27 former postdoctoral researchers, and 337 mentors. This study demonstrated the low use of the IDP among postdoctoral researchers (19%) and their mentors (9%), but the perceived value of the instrument was high for those who had used the tool (71% for postdoctoral researchers and 90% for mentors)10. There have been recent calls to study the IDP more closely and for the NIH and other stakeholders to share the data collected on its use11.\n\nHerein, we describe the assessment of the use and effectiveness of the IDP among a sample of U.S. doctoral students. We surveyed doctoral students from at least 98 different U.S. universities in the spring and early summer of 2016 (March through June). We collected data from 663 respondents in PhD programs in the life/biological/medical (76.5%) or physical/applied sciences (23.5%), with the majority of respondents being female (70.9%) compared to their male (29.1%) counterparts (Supplementary File 1 and Supplementary File 2). We report evidence of the low usage and minimal effectiveness of the IDP, as measured by individuals’ perception of whether the tool is helpful to their career development. Further, our results suggest that the IDP is most effective when graduate students complete the tool with faculty mentors with whom they have a positive relationship. Confidence regarding the completion of doctoral training and post-training career plans and use of institutional career development resources are also associated with respondents being more likely to indicate that the IDP is helpful to their career development.\n\n\nMethods\n\nThis research was approved by the University of Kentucky (protocol 15-1080-P2H) and University of Texas Health San Antonio (protocol HSC20160025X) institutional review boards as a component of a health and wellbeing study. Respondents read a cover page and consented to the study by clicking the online survey web link. Subjects responded anonymously and were ensured of confidentiality.\n\nThe survey was conducted online using the secure web application REDCap. The survey was distributed to potential respondents through social media (primarily Twitter and LinkedIn) and direct email to subjects enrolled in life/biological/medical or physical/applied sciences doctoral programs across a number of different U.S. institutions (Supplementary File 1). Eligibility criteria included being currently enrolled in a life/biological/medical or physical/applied sciences doctoral program at a U.S. institution at the time the survey was conducted. Responses were collected over a three-month period, March 2016 to June 2016. The overall study sample size was dictated by the number of respondents fitting the eligibility criteria.\n\nSubjects were asked to respond to the IDP questions using the five-point Likert scale strongly agree, agree, neutral, disagree and strongly disagree. For data analysis, strongly agree and agree responses were grouped together as an agree category and neutral, disagree, and strongly disagree were grouped together in a does not agree/disagree category. The survey questions relevant to this study are included as Supplementary File 4.\n\nOne-way frequencies of the survey variables were calculated and the Pearson chi-square test was used to assess the univariate associations between the variables and the outcome “I Find the IDP Process Helpful to my Career Development.” All summaries and statistical analysis were performed in SAS 9.4.\n\n\nResults\n\nOnly 53.6% of respondents reported that they are required to complete a formal IDP, while only 37.4% do so with their faculty advisor. Strikingly, 26.1% complete the tool but do not discuss it with their advisor. Further, only one-third, 33.6%, of respondents feel that they can have an honest conversion with their advisor via the IDP process and only 33.7% feel that the IDP is helpful to their career development (Figure 1 and Supplementary File 2). In the 2014 study, only 8% of postdoctoral researchers were required to complete an IDP, although overall usage among respondents was approximately 19%10. There appears to be a modest increase in usage of the IDP among doctoral students versus postdoctoral researchers.\n\nWe found that respondents in the life/biological/biomedical sciences (36.7% versus 23.6% for physical/applied sciences) and females (36.9% versus 26.1% for males) are more likely to find the IDP process helpful to their career development (Figure 2A, B and Supplementary File 3). Additionally, respondents that are required to complete an IDP in general (49.4% versus 14.6% who are not) and those that complete the IDP with their advisor (56.2% versus 19.7% who do not) are also more likely to find the IDP helpful (Supplemental File 3).\n\nAcross several measures, positive mentorship relationships also associate with the perception that the IDP process is helpful. For example, of those respondents who found the IDP to be helpful to their career development, 66% indicated that they could have an honest conversation with their PI/advisor via the IDP process versus 17.7% who did not agree, and 36.1% said their PI/advisor is an asset to their academic and professional career versus 26% who did not agree (Figure 3 and Supplementary File 3). These data corroborate anecdotal testimonies suggesting that supportive mentors can positively influence one’s IDP experience whereas non-supportive mentors can have the opposite impact12.\n\nFurther, those respondents that are confident about completing their training (36.4% versus 25.9% who are not), their career prospects (39% versus 30.4% who are not), and their post-training career (37.8% versus 30.9% who are not) are also more likely to report the IDP as being helpful to their career development. Lastly, respondents who attend career development programs at their institution are more likely to report the IDP as helpful to their career development (Supplementary File 3).\n\n\nDiscussion\n\nMore than 15 years after the creation of the IDP and 4 years after the NIH required its use, do we know if the tool is working as it was intended? Unfortunately, the answer is no. The study focusing on postdoctoral researchers from 201410 and current study cannot fully answer this question, but rather these studies should serve to elicit further discussion on how to best use the IDP, especially in relation to the enforcement of the tool’s use and use with PIs/advisors. Further, this work should stimulate additional research on the general use and effectiveness of the tool.\n\nPolicymakers, leaders of academic institutions, individual faculty, and career development specialists should find it concerning that IDP use and effectiveness is not well understood, despite the tool’s general acceptance and use at countless U.S. universities and the NIH’s requirement for reporting on the use of the IDP. Should we not have known more about such an instrument prior to it being mandated as a policy? Is there potential harm being done by the use of IDPs? Anecdotally, some doctoral students and postdoctoral researchers report that faculty sometimes reject non-academic career trajectories within the context of the IDP and these faculty try to force trainees toward an academic career path12. Such negative mentorship relationships may partially explain the cause of the high rates of anxiety and depression in the doctoral student population14.\n\nWe have noticed that the structure of some IDPs has changed over time. For example, the University of Kentucky College of Medicine’s IDP has excluded the career exploration section of the tool15, which was prominently included in its original design. How widespread is such a change to the IDP? Could such a change have been made to appease stakeholders who are most interested in training PhDs to pursue faculty careers? Could such a change be driving a general increase in IDP usage among faculty mentors? These questions should be addressed in future research.\n\nWe recognize that there are limitations to our work. For example, this is a cross-sectional study that may not be representative of the entire U.S. research enterprise. Given the NIH’s adoption of the IDP, the agency should fund a more extensive longitudinal study with a larger sample size to understand the barriers that are preventing some trainees and mentors from using the IDP and to better understand the effectiveness of the IDP as doctoral students and postdoctoral researchers move through their PhD education and training experience. The IDP’s impact on specific outcomes, including career path decision making and long-term career outcomes, should be studied. Future work should also determine if there are any unintended negative consequences associated with IDP use.\n\nWe believe that our results call for the need for policymakers, funding agencies, and universities to focus attention on mentorship training for faculty and building career development infrastructure. If the NIH is to require the use of the IDP, they should require training of mentors on how to best support the career development of their mentees to obtain maximum impact, and institutional career development infrastructure is needed to achieve this. The NIH BEST program laid the foundation for building career development infrastructure at a limited number of institutions16. The National Institute of General Medical Sciences has recently incorporated career development components into their pre-doctoral T32 mechanism17, which is another good start to developing more widespread career development infrastructure. Other grant mechanisms should likewise be established so that a greater number of institutions can obtain NIH funds that will drive the creation of innovative career development programs across the U.S. Such programs should serve the needs of doctoral students and postdoctoral researchers and train faculty on the fine science and art of mentorship.\n\nThe NIH and several professional societies have been conducting “Train-the-Trainer” events to provide career and professional development training to faculty and staff. We recommend the extensive expansion of this program. The NIH could mandate such training for all faculty who pay doctoral students or postdoctoral researchers from NIH funds. Generally, it would likewise be prudent for universities to mandate that all faculty employing/supervising graduate students and postdoctoral researchers complete such training. The training could be developed and offered at each university through institutional career development offices.\n\nUltimately, the sustainability of the biomedical enterprise hinges upon the next generation of PhDs entering the diverse workforce. We should work to support this group of scientists with the same rigor and reproducibility that we strive for everyday as we conduct our experiments. The IDP is likely useful for supporting the career development of PhDs, but more work is needed to understand how best to use the tool.\n\n\nData availability\n\nDataset 1. Individual Development Plan survey data. Columns Q1–Q26 correspond to the questions listed in Supplementary File 4. DOI: 10.5256/f1000research.15154.d20639413.",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nN.L.V. is supported by the University of Kentucky’s Cancer Center Support Grant (NCI P30CA177558) and the Center for Cancer and Metabolism (NIGMS P20GM121327).\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 the Markey Cancer Center Research Communications Office for graphic design and formatting assistance; Dr. Paula Chambers, Versatile PhD, for her comments on and aid in distributing the study survey; and the Graduate School of Biomedical Sciences at the University of Texas Health San Antonio for providing partial funding for the study.\n\n\nSupplementary material\n\nSupplementary File 1. Self-reported institution of all respondents.\n\nClick here to access the data.\n\nSupplementary File 2. Response rates, separated by demographic characteristics.\n\nClick here to access the data.\n\nSupplementary File 3. Univariate analysis of demographic variables/responses to the questionnaire and perception of Individual Development Plan helpfulness.\n\nClick here to access the data.\n\nSupplementary File 4. Example copy of the survey questions relevant to this study.\n\nClick here to access the data.\n\n\nReferences\n\nAlberts B, Kirschner MW, Tilghman S, et al.: Rescuing US biomedical research from its systemic flaws. Proc Natl Acad Sci U S A. 2014; 111(16): 5773–5777. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlberts B, Kirschner MW, Tilghman S, et al.: Opinion: Addressing systemic problems in the biomedical research enterprise. Proc Natl Acad Sci U S A. 2015; 112(7): 1912–1913. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcDowell GS, Gunsalus KT, MacKellar DC, et al.: Shaping the Future of Research: a perspective from junior scientists [version 2; referees: 2 approved]. F1000Res. 2014; 3: 291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoach M, Sauermann H: The declining interest in an academic career. PLoS One. 2017; 12(9): e0184130. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSauermann H, Roach M: Science PhD career preferences: levels, changes, and advisor encouragement. PLoS One. 2012; 7(5): e36307. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFuhrmann CN: Enhancing Graduate and Postdoctoral Education To Create a Sustainable Biomedical Workforce. Hum Gene Ther. 2016; 27(11): 871–879. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClifford PS: Quality Time with Your Mentor. The Scientist. 2002; 16: 59. Reference Source\n\nHobin JA, Fuhrmann CN, Lindstaedt B, et al.: You Need a Game Plan. Science. 2012. Publisher Full Text\n\nNational Institutes of Health: Revised Policy: Descriptions on the Use of Individual Development Plans (IDPs) for Graduate Students and Postdoctoral Researchers Required in Annual Progress Reports beginning October 1, 2014. 2014. Reference Source\n\nHobin JA, Clifford PS, Dunn BM, et al.: Putting PhDs to work: career planning for today's scientist. CBE Life Sci Educ. 2014; 13(1): 49–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsai JW, Vanderford NL, Muindi F: Optimizing the utility of the individual development plan for trainees in the biosciences. Nat Biotechnol. 2018; 36(6): 552–553. PubMed Abstract | Publisher Full Text\n\nGould J: Career development: A plan for action. Nature. 2017; 548: 489–490. Publisher Full Text\n\nVanderford NL, Evans TM, Weiss LT, et al.: Dataset 1 in: A cross-sectional study of the use and effectiveness of the Individual Development Plan among doctoral students. F1000Research. 2018. Data Source\n\nEvans TM, Bira L, Gastelum JB, et al.: Evidence for a mental health crisis in graduate education. Nat Biotechnol. 2018; 36(3): 282–284. PubMed Abstract | Publisher Full Text\n\nUniversity of Kentucky College of Medicine: Individual Development Plan. 2018. http://graduate.med.uky.edu/news/uk-com-biomedical-education-releases-recommendations-individual-development-plans and https://research.med.uky.edu/news/revised-individual-development-plans-trainees. Reference Source\n\nNational Institutes of Health: Best Coordinating Center. 2018. Reference Source\n\nGammie AG, Gibbs K, Singh S: New NIGMS Institutional Predoctoral Training Grant Funding Opportunity Announcement. NIGMS Feedback Loop Blog. 2017. Reference Source"
}
|
[
{
"id": "34890",
"date": "21 Jun 2018",
"name": "Christopher L. Pickett",
"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 article “A cross-sectional study of the use and effectiveness of the Individual Development Plan among doctoral students” by Vanderford et al. examines the use of Individual Development Plans among biology and physics graduate students at U.S. universities. Through the use of an opt-in survey advertised on social media, the authors received over 650 responses about the usefulness of IDPs. The goal of the study was to provide information on how widespread the use of IDPs is and how effective they are in helping students explore careers.\n\nUnderstanding the efficacy of IDPs is important as this is an increasingly important piece of the career and skills development portfolio for grad students. However, I have serious reservations regarding the survey instrument and the interpretation of the resulting data.\nIt is not clear from the survey instrument or the methods section that all survey respondents completed an IDP. If respondents have never completed an IDP, their responses to questions about the efficacy of the IDP in Fig. 1 would likely skew the data in an unfavorable direction. The authors should describe whether they ensured that only those who have completed an IDP took the survey.\nIf they did not use this screen, the authors should describe how data from those who haven’t completed an IDP might affect their data on IDP efficacy and take this into account when drawing conclusions. If the authors did ensure that all respondents completed an IDP, this information should be clearly stated. In addition, the authors should better define the denominators of the percentages reported. For example, if 33.7% respondents find the IDP helpful, is this 33.7% of those completing the survey or 33.7% of those that completed an IDP? If a student completes multiple IDPs, isn't it possible the student could agree with questions 2 and 3 (complete IDP annually with advisor and complete IDP annually but don't talk to advisor)? Overlap in these populations, as is apparent in the supplemental data spreadsheet provided, could complicate the analysis of the data in Figs. 2 and 3 as it is not clear which experiences respondents may be referencing in their answers to the questions. How are we to think of the values presented here in light of the study referenced that analyzed the use of IDPs among postdocs? If the data here show that more grad students fill out IDPs than postdocs, this may provide insight into how and why different populations use IDPs.\n\nThe negative tone of the article is surprising. The authors rightly point out that the IDP is poorly studied and this survey is one of the first analyzing use by graduate students. Therefore, this study establishes the baseline for IDP use among grad students. If the authors wish to characterize the use of IDPs as low or ineffective, the authors should take time to discuss their expectations and what previous data/experiences were used to set those expectations.\nFor example, the overuse of “only” in the first paragraph of the results communicates these values fall below the authors’ expectations. Recognizing that these values will likely never reach 100%, what constitutes broad, acceptable use of an IDP? Furthermore, an IDP is supposed to help students set a path for developing skills relevant for their career. The authors should discuss what “minimal effectiveness” of IDPs means in the context of the respondents being students who do not have experience understanding how their IDP relates to securing subsequent jobs.\n\nThe interpretation of the data from the first question of the survey “My institution/college/department/PI/advisor requires me to compete a formal IDP” should be more guarded. This question asks respondents to comment on institutional or departmental policy. Respondents may disagree with the statement either because the institution does not require an IDP or because the student does not know institutional policies around IDPs. Additionally, some institutions had dozens of respondents. How consistent are responses to this and other questions when looking at respondents from the same institution?\n\nThe authors should provide figure legends beyond the figure titles to help the reader understand the data.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3805",
"date": "05 Jul 2018",
"name": "Nathan Vanderford",
"role": "Author Response",
"response": "Dear Dr. Pickett, Thank you for your review. Your critique has been helpful as we have revised the article. Below we address the major issues you raised. Within the text, we have clarified which respondents were analyzed in the article, thus addressing the confusion regarding the denominators used in the analysis. In the analysis of IDP use, one-way frequencies were calculated based on the responses from all the respondents. In the analysis of IDP effectiveness, we have reanalyzed our data and now present univariate associations that were calculated based on the responses from only the subset of respondents that completed an IDP. As such, our interpretation and reporting of the data has been adjusted accordingly. Of note, upon this reanalysis, the differences in IDP effectiveness between fields of study and gender were no longer significant. We apologize for the confusion regarding the reporting of the 33.7% of respondents that found the IDP helpful to their career development in the top section of the univariate analysis table (Supplementary File 3). We have now removed this data from the top section of Supplementary File 3, although this data can still be found in Supplementary File 2. We have neutralized the tone of the article in general and specifically we have removed words such as “only” and “minimal” within the context of our findings. We agree that our assessment of the expected level of IDP use and effectiveness was speculative. We also agree that as a baseline study, more work should be done to characterize an acceptable level of IDP use and effectiveness within the PhD trainee population. We have added an extensive limitations subsection (within the methods section) to the new version of the paper that speaks to several issues raised by the reviewers including your points about how respondents may or may not understand or be aware of their institution’s policies regarding IDP use. We also added cautions regarding variability in responses from subjects within the same institution. There are such variabilities within our dataset and it is difficult to assess the exact reasons for this, as it could be caused by, for example, general variability between respondents based on one’s individual perception of the IDP, different interpretations of the policies and procedures around IDP use, and/or the use of different IDP formats. Future work should help clarify this issue. We have added additional descriptions to our figure legends to help readers understand the data and our analysis. In closing, we hope that you will favorably review the revised version of the article in light of our changes based on your critique as well as that of the other reviewers. We thank you again for your comments and we strongly believe that your review has been critical in strengthening this work. Sincerely, Nathan L. Vanderford"
}
]
},
{
"id": "34887",
"date": "25 Jun 2018",
"name": "Jessica K Polka",
"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 “A cross-sectional study of the use and effectiveness of the Individual Development Plan among doctoral students,” the authors present a survey of graduate students conducted with the intention of determining experience with, and attitudes toward, Individual Development Plans. Given the widespread use of IDPs in the biomedical sciences, this study has the potential to provide insights that could improve the career planning process for many PhD students.\nLike reviewer 1, I am concerned that the overly-negative conclusions in this paper are influenced by the inclusion of irrelevant respondent populations. Therefore, in addition to the revisions suggested by reviewer 1, I recommend the following modifications:\nIt is perhaps unsurprising that overall usage of the IDP is low given that only 76.5% of respondents were in the life/biological/medical sciences programs. While it’s possible that the 23.5% of respondents in physical/applied sciences are in programs that require or encourage IDP completion, these students are likely unaffected by the NIH mandate mentioned in the introduction. As such, the author should not use this sample to imply that low IDP usage indicates a failure in NIH policy, as suggested by the first two sentences of the discussion.\n\nGiven the valid concerns that reviewer 1 has expressed, the authors should reanalyze their data to exclude respondents who have no experience with IDPs (assuming that the survey instrument did not do this already). While survey instrument did not address this question directly, the authors could at least take a subset of respondents who answered affirmatively to Q2 or Q3. If my calculations are correct, >50% of this subset finds the IDP process helpful for career development. Therefore, when the analysis is confined to students with confirmed IDP experience, the outlook for the IDP is less bleak than the tone of the paper makes it out to be.\n\nThe assessment of the value of the IDP should be placed into the context of overall cost of its implementation (which I suspect is extremely low). Assuming that 50% of students that complete it annual benefit from it, this is (in my opinion) a good payoff for a very small number of hours of work for mentors and students. I would be interested to read the author’s comments on cost/benefit ratio of this intervention.\n\nIn the discussion, the authors express concern that the IDP has been implemented without an attending study of the benefits. Later in the same section, they call for widespread career development infrastructure and “extensive” expansion of “Train-the-Trainer” events; to abide by their own logic, they should provide evidence that this change is supported by data. Again, the benefits of these interventions should be placed into the context of their cost.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3806",
"date": "05 Jul 2018",
"name": "Nathan Vanderford",
"role": "Author Response",
"response": "Dear Dr. Polka, Thank you for your review of our work. Your comments and critique have been critical in guiding our revisions. We respond to your major points below. We appreciate your thoughtful comments regarding the tone of the first version of this article. We have now revised the article in a way that neutrally presents and describes the data; our use of words such as “only” and “minimal” within the context of our findings have been removed. Additional work will need to be done to understand an appropriate and acceptable level of IDP use and effectiveness in the PhD trainee population. We have added a limitations subsection (within the methods section) that speaks to several points raised by all the reviewers. Within this section, we provide cautions regarding the generalizability of the data and differences between disciplines. We have reanalyzed the IDP effectiveness data to focus only on those respondents that completed an IDP. Of note, in relation to the previous comments about generalizability, upon reanalysis of the IDP effectiveness data, the differences between fields and gender are no longer significant and the text and figures have been revised accordingly. Regarding the cost/benefit ration of the IDP, this is a very interesting and important consideration, but ultimately we feel that our thoughts on this are too speculative to include in the article itself. That said, as stated at the end of the article, we do feel that the IDP can be an effective career planning tool when used “correctly.” However, we believe that more work needs to be done to assess the “correct” way to use the IDP, especially in a way that causes no harm to trainees. We believe that it is unacceptable for any trainee to be intimidated by and/or harmed through the use of the IDP, as has been suggested to occur. Ultimately, there is not enough information available on the use of the IDP to fully understand the costs, consequences, and/or benefits of its use. We have revised the discussion to clarify that our recommendations are based primarily on our own findings and we call for the evaluation of any new interventions that are put into place. In closing, we look forward to your second review in light of our revisions that are in response to your critiques and that of the other reviewers. We feel that your comments have been critical to the improvement of this work. Thank you again for your time and expertise. Sincerely, Nathan L. Vanderford"
}
]
},
{
"id": "35352",
"date": "27 Jun 2018",
"name": "Zeba Wunderlich",
"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 article, Vanderford, et al. examine the patterns of Individual Development Plans (IDP) usage and perceptions among biology and physics graduate students at U.S. universities. Given the increasing numbers of institutions that require IDPs for graduate students, the research questions posed in this article are timely and of potentially high impact. I recommend the following changes to the article:\nI strongly agree with the other reviewers that a key modification is to re-do the analysis, only considering those students who have experience completing an IDP, or to clarify if the study already followed this protocol.\n\nHow do the demographic data (gender, race, ethnicity) compare with the eligible survey population? If the data differ from the eligible population, do you expect this to affect the results? It might be appropriate to add a brief section to the beginning of the results section describing the demographics of the data and a comment to the discussion about the caveats of an opt-in survey. (I believe this is what you are hinting at in the paragraph starting \"We recognize that there are limitations...\" but it could be stated more explicitly.)\n\nAs the first reviewer, I did not always understand the denominators of the stated percentages. For example, do 37.4% of all respondents fill out an IDP with their advisor? Or 37.4% of those that are required to complete an IDP do so with their advisor? Please review the results section to clarify this and similar statements. Also, what fraction of students required to complete an IDP actually do it? And what fraction of students not required to complete an IDP do it?\n\n(Minor) The word \"only\" is used many times throughout the results section, which colors the interpretation of the results. I'd suggest keeping the results section more neutral, while saving the \"only\" statements for the discussion.\n\n(Minor) It might be worth noting that there is a two year gap in the 2014 postdoctoral study and the current study, so some of the increase in IDP usage may be due to an increase in usage over time.\n\n(Minor) What are the p-values of the reported differences in IDP effectiveness between males/females, physics/biology students? I found these in the Supplement, but they are worth mentioning in the main text or depicting on the figures/figure legends.\n\n(Minor) I appreciate the paragraph in the discussion about the changes to IDP structure. It may be worth commenting on the differences of IDP structure between institutions or individuals. My own experience with IDPs (Vincent, et al. 2015 Molecular Cell) didn't use the myIDP platform and may have influenced my perception of IDPs. I suspect I'm not alone.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3807",
"date": "05 Jul 2018",
"name": "Nathan Vanderford",
"role": "Author Response",
"response": "Dear Dr. Wunderlich, We greatly appreciate your review, which has aided in shaping our revised article. We have responded to your major critiques/comments below. In response to your critique and that of the other reviewers, we have reanalyzed the IDP effectiveness data to include only those respondents that completed an IDP. We have revised the text and figures accordingly. Of note, upon this reanalysis, the differences in IDP effectiveness between fields of study and gender were no longer significant. Given that our study was based on an online survey that was conducted, in part, through an open call on social media, we have no way of knowing the demographic makeup of the all the potential respondents. That said, we have characterized the demographic makeup of all the actual respondents (see Supplementary File 1 and 2). We agree that there are caveats to our methods and thus we have added an extensive description of the limitations to this work within the methods section of the article. We apologize that there was confusion regarding the denominator used in our analysis. We have now clarified which respondents were used in the analysis. Namely, in the analysis of IDP use, one-way frequencies were calculated based on the responses from all the respondents. In the analysis of IDP effectiveness, again, we have reanalyzed our data and now present univariate associations that were calculated based on the responses from only the subset of respondents that completed an IDP. This has been more clearly stated in the revised version of the article. We have revised the article such that it is now written in a neural tone; we have removed such words as “only” and “minimal” in relation to describing our own findings. Future work should address what is a reasonable/acceptable level of IDP use and effectiveness. We appreciate your comments regarding the comparison of our data to the 2014 postdoctoral study. We have revised the text such that we now point out the study, but we do not specifically comment on how the prior data may relate to our findings. Future work should address the current use and effectiveness of the IDP in postdoctoral researchers. We have now included p-values in the text and in the figure legends were applicable. We have added comments in our limitations section regarding how differences in IDP structures could influence our data. In summary, we have revised our article according to your critique/comments and that of the other reviewers, and we feel that the collective reviews have significantly strengthened our work. Thank you for your time and we look forward to reading your next review. Sincerely, Nathan L. Vanderford"
}
]
}
] | 1
|
https://f1000research.com/articles/7-722
|
https://f1000research.com/articles/7-1012/v1
|
05 Jul 18
|
{
"type": "Research Article",
"title": "Neuronal pentraxin receptor-1 is a new cerebrospinal fluid biomarker of Alzheimer’s disease progression",
"authors": [
"Ilijana Begcevic",
"Magda Tsolaki",
"Davor Brinc",
"Marshall Brown",
"Eduardo Martinez-Morillo",
"Ioulietta Lazarou",
"Mahi Kozori",
"Fani Tagaraki",
"Stella Nenopoulou",
"Mara Gkioka",
"Eutichia Lazarou",
"Bryant Lim",
"Ihor Batruch",
"Eleftherios P. Diamandis",
"Ilijana Begcevic",
"Magda Tsolaki",
"Davor Brinc",
"Marshall Brown",
"Eduardo Martinez-Morillo",
"Ioulietta Lazarou",
"Mahi Kozori",
"Fani Tagaraki",
"Stella Nenopoulou",
"Mara Gkioka",
"Eutichia Lazarou",
"Bryant Lim",
"Ihor Batruch"
],
"abstract": "Background: Alzheimer’s disease (AD) is the most common type of dementia, with progressive onset of clinical symptoms. The main pathological hallmarks are brain deposits of extracellular amyloid beta plaques and intracellular neurofibrillary tangles (NFT). Cerebrospinal fluid reflects pathological changes in the brain; amyloid beta 1-42 is a marker of amyloid plaques, while total and phosphorylated tau are markers of NFT formation. Additional biomarkers associated with disease pathogenesis are needed, for better prognosis, more specific diagnosis, prediction of disease severity and progression and for improved patient classification in clinical trials. The aim of the present study was to evaluate brain-specific proteins as potential biomarkers of progression of AD. Methods: Overall, 30 candidate proteins were quantified in cerebrospinal fluid (CSF) samples from patients with mild cognitive impairment (MCI) and mild, moderate and severe AD dementia (n=101) using mass spectrometry-based selected reaction monitoring assays. ELISA was used for neuronal pentraxin receptor-1 (NPTXR) confirmation. Results: The best discrimination between MCI and more advanced AD stages (moderate and severe dementia) was observed for protein NPTXR (area under the curve, AUC=0.799). A statistically different abundance of this protein was observed between the two groups, with severe AD patients having progressively lower levels (p<0.05). ELISA confirmed lower levels in AD, in a separate cohort that included controls, MCI and AD patients. Conclusions: We conclude that NPTXR protein in CSF is a novel potential biomarker of AD progression and could have important utility in assessing treatment success in clinical trials.",
"keywords": [
"Alzheimer’s disease",
"biomarkers",
"cerebrospinal fluid",
"mass spectrometry",
"selected reaction monitoring",
"neuronal pentraxin receptor-1",
"Alzheimer’s disease progression",
"dementia"
],
"content": "Introduction\n\nAlzheimer’s disease (AD) is a neurodegenerative disease characterized by progressive cognitive decline, behavioural problems and impairment of daily living activities. The main pathological hallmarks of AD are brain extracellular deposits known as amyloid β (Aβ) plaques, composed of aggregated Aβ fragments and intracellular neurofibrillary tangles (NFT), containing hyperphosphorylated protein tau (p-tau) fibrils1. The diagnosis of AD is currently made based on core clinical criteria, including medical history, mental status testing, and neurological and physical assessment2. With such criteria, only probable dementia due to AD can be diagnosed, while definitive diagnosis of AD can be made only post mortem, by neuropathological examination of different brain regions. More recently, a blood test for AD diagnosis has been suggested but it has not as yet been clinically validated3. Three stages of AD have been recognized by the National Institute on Aging (NIA) and the Alzheimer’s Association newly revised diagnostic and research criteria: preclinical stage, mild cognitive impairment (MCI) due to AD, and dementia due to AD4. Preclinical stage describes asymptomatic individuals with existing early brain pathology, while MCI due to AD includes patients with prodromal, mild symptoms as a result of disease pathology. Patients with dementia due to AD have impaired memory, thinking and behavioural functions, accompanied by severe pathological brain changes. Clinical symptoms typically appear gradually, indicating different levels of dementia severity: mild dementia (or early stage), moderate dementia (or middle stage) and severe dementia (or late stage)5.\n\nCerebrospinal fluid (CSF) is a proximal fluid of the central nervous system, residing in direct contact with the brain parenchyma and thus can reflect physical and pathological changes in the brain6. As such, CSF may be the most promising source of AD biomarkers; especially highly specific, brain-related protein biomarkers. The best evaluated AD biomarkers to date are CSF Aβ1-42, total tau (t-tau) and p-tau levels7. These core AD biomarkers reflect main pathological hallmarks: Aβ1-42 peptide is a marker of Aβ plaque formation, while t-tau and p-tau are biomarkers of neuronal injury. Decreased CSF levels of Aβ1-42 and increased levels of t-tau and p-tau have been observed in AD patients, compared with healthy controls8. Still, these extensively studied biomarkers are not widely used in the clinic, largely due to the lack of method standardization, and are mostly utilized in research settings, as also suggested by the new AD diagnostic guidelines2. In addition, current CSF biomarkers have been tested in clinical trials of different AD therapeutic approaches. The results were contradictory, questioning their usefulness as indicators of efficacy of new therapies9. It has also been previously reported that current AD biomarkers do not correlate well with cognitive decline in AD patients10,11.\n\nThere is a clinical need for novel biomarkers of AD progression. Such biomarkers could accurately and proactively identify evolving cases of AD and could be invaluable in clinical trials for patient enrichment and/or as surrogate endpoints. Moreover, such biomarkers could contribute to the better understanding of the underlying pathological mechanisms of AD.\n\nDifferential expression of proteins specific to a particular tissue can have strong disease specificity, pinpointing to pathology unique to that tissue. Some of these tissue-specific proteins have already shown promise as potential biomarkers, such as in male infertility (testis-specific protein TEX101) and in cerebral hemorrhagic stroke (brain-specific proteins NFM, α-Inx and β-Syn)12,13. In our recent study, we identified a set of brain-specific proteins that are consistently detected in the normal CSF proteome14. The brain-specific proteins were retrieved from the Human Protein Atlas (HPA) tissue-specific database15 and encompassed tissue-enriched (mRNA expression at least five times higher in the particular tissue (i.e. brain) relative to other tissues) and group-enriched proteins (mRNA expression at least five times higher in the group of 2–7 tissues (including brain), relative to all other tissues). These proteins were also secreted and/or were membrane-bound (as defined by HPA). We have further developed targeted mass spectrometry-based assays for quantification of 30 of these highly specific brain proteins in CSF16. The main objective of the present study is to evaluate these 30 brain-related proteins for their ability to differentiate various stages of AD severity, i.e. MCI, mild, moderate and severe AD dementia, by utilizing state-of-the-art mass spectrometry-based selected reaction monitoring (SRM) assays. Considering that the apolipoprotein E (APOE) ε4 allele is the strongest genetic risk factor for developing AD, associated with disease pathology17, we have further evaluated if the abundance of our candidates was related to the APOE phenotypes.\n\n\nMethods\n\nA multiplexed, scheduled, SRM assay was developed for 30 brain-related proteins and is described in detail elsewhere16. A protein previously found not to change in AD CSF (extracellular matrix protein 1, ECM1) was included as a negative control. Also included was a protein primarily related to demyelinating diseases (myelin basic protein, MBP). The SRM method for MBP has been described elsewhere13. A peptide corresponding to apolipoprotein B (APOB) protein (a plasma protein) was also monitored, to check for blood contamination18. For peptides containing methionine, both oxidized and non-oxidized forms of the peptide were monitored. Four peptides that represent different APOE phenotypes where additionally added to the assay, including an APOE peptide for total APOE, as a control. The APOE method was previously published18.\n\nCSF samples were thawed and volumes equivalent to 15 µg of total protein were denatured with 0.05% RapiGest detergent (Waters, Milford, USA) and reduced with 5 mM dithiothreitol (Sigma-Aldrich, Oakville, Canada) at 60°C for 40 min. Alkylation was achieved with 15 mM iodoacetamide (Sigma-Aldrich) for 60 min in the dark at 22°C. A mixture of APOE heavy peptides was spiked into samples prior to addition of trypsin, while a mixture of 32 heavy peptides (30 candidates, ECM1, MBP) plus a heavy peptide for total APOE were spiked into the mixture after digestion, followed by addition of 1% trifluoroacetic acid. Digestion was carried out for 24 hours at 37°C with 1:30 trypsin-to-total protein ratio. Samples were then centrifuged at 1,000 g for 30 min and the supernatants retained. Peptides were purified using OMIX C18 tips, eluted in 4.5 µL of acetonitrile solution (65% acetonitrile, 0.1% formic acid) and finally diluted with 54 µL of water-formic acid mix (0.1% formic acid).\n\nSamples were analysed with a triple quadrupole mass spectrometer, TSQ Quantiva: (Thermo Scientific, San Jose, USA). Each sample (18 μL) was injected into an in-house-packed 3.3 cm pre-column (5 μm C18 particle, column inner diameter 150 μm), followed by a 15 cm analytical column (3 μm C18 particle, inner diameter 75 μm, tip diameter 8 μm). The liquid chromatography, EASY-nLC 1000 system (Thermo Fisher, Odense, Denmark) was coupled online to the TSQ Quantiva mass spectrometer with a nano-electrospray ionization source. A 37-min LC gradient was applied, with an increasing percentage of buffer B (0.1% formic acid in acetonitrile) for peptide elution at a flow rate of 300 nL/min. The SRM assay parameters were set up as follows: positive-ion mode, optimized collision energy values, adjusted dwell time, 0.7 Th Q1 resolution of full width at half-maximum and 0.7 Th in Q3 resolution. LC peaks for all peptides were manually inspected to ensure acquisition of minimum 10 points per LC peak. Raw data were uploaded and analyzed with Skyline software (University of Washington, Seattle, USA).\n\nWe used the RayBio Human NPTXR ELISA kit, as recommended by the manufacturer (catalog # ELH-NPTXR, Ray Biotech, Norcross, GA, USA). All CSF samples were analyzed after a 25-fold dilution. For this independent validation, we used CSF samples from 12 AD patients, 21 patients with MCI and 23 control subjects. This cohort was used previously for mass spectrometric analyses, as outlined elsewhere19. The samples were obtained by lumbar puncture and stored at -80°C until use. The Institutional Review Board of the Technical University of Munich approved the study and all patients signed an informed consent form.\n\nAge and sex data were collected from all participants. In total, 101 CSF samples were retrospectively collected at the memory and dementia clinic of the 3rd Department of Neurology, “G. Papanikolaou”, School of Medicine, Aristotle University of Thessaloniki, Greece and from the Day Centers of the Greek Association of Alzheimer’s Disease and Related Disorders (GAARD), Thessaloniki, Greece. A summary of patient characteristics is shown in Table 1.\n\nAD, Alzheimer’s disease.\n\na Expressed as median (25th, 75th percentile)\n\nb Expressed as mean (standard deviation)\n\nc Mini-Mental State Examination\n\nPatients suspected of having AD were examined by a specialist neuropsychiatrist and diagnosis was made based on the NINCDS/ADRDA criteria for probable AD20. Disease severity was determined based on the Mini-Mental State Examination (MMSE) and clinical dementia rating (CDR) scores and patients were categorized as having mild (MMSE=20–26, CDR=1), moderate (MMSE=10–19, CDR=2) and severe (MMSE=0–9, CDR=3) dementia. Diagnosis of MCI was based on the description by Petersen, which is almost equivalent to the NIH-AA criteria for MCI due to AD21. This study was approved by the GAARD scientific and ethics committees and by the Institutional Review Boards of Aristotle University and the University of Toronto. All participants signed an informed consent form.\n\nA fraction of CSF samples were analyzed for core AD biomarkers (Aβ1-42, t-tau, p-tau) using an Innotest ELISA kit (Fujirebio Europe)22. Overall, 54 participants were tested for Aβ1-42 (distributed by groups, MCI: n=10, mild: n=7, moderate: n=23, severe: n=14), 42 for t-tau (distributed by groups, MCI: n=9, mild: n=6, moderate: n=16, severe: n=11) and 43 for p-tau (distributed by groups, MCI: n=9, mild: n=5, moderate: n=21, severe: n=8).\n\nAll CSF samples were collected by lumbar puncture, inspected macroscopically for blood contamination, centrifuged and stored at -80°C in polypropylene tubes. Samples were shipped to the Lunenfeld−Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, Canada and stored at -80°C until processing. Ethics approval was obtained from the Mount Sinai Hospital Research Ethics Board for use of these samples.\n\nClinical samples were randomized and ran in duplicate. The raw files were uploaded to Skyline software (version 3.5.0.9319), which was used for peak integration and quantification of the area under the curve (AUC). Relative quantification was performed as previously described16. For peptides with amino acid methionine in the sequence, AUClight/AUCheavy was calculated as: AUC (oxidized + non-oxidized)light/AUC (oxidized + non-oxidized)heavy. SRM data were manually evaluated and samples with poor integration were excluded. Identification of APOE phenotype was determined as described in our previous report18.\n\nStatistical analysis was performed with R statistical and graphics software, version 3.5.0. Means, medians, standard deviations, interquartile ranges and coefficients of variations were calculated. Linear regression was used to test for differences in ages. For tests involving a dichotomous variable, such as sex, the Fisher’s exact test was used. Tests for differences in candidate protein abundance, MMSE score, CSF Aβ1-42, t-tau and p-tau across disease stages were adjusted by age and sex using multivariate linear regression. Correlation analyses for MMSE score and protein abundance were performed using Spearman’s rank correlation test. ROC curves were prepared for the most significant proteins and AUC values with 95% confidence intervals were calculated using the bootstrap method. AUC values were covariate-adjusted by age or sex when there was a significant association (p<0.05) between a marker and the covariates in controls23. P-values for comparison between groups were reported as non-adjusted and adjusted for multiple comparison by the Holm method and p<0.05 was considered statistically significant.\n\n\nResults\n\nCSF samples from MCI and AD patients with different dementia severity (n=101) were randomized into two sets. The rationale for the randomization was to confirm the validity of our findings in separate assays, performed on different days. In the first set, 8 patients were diagnosed with MCI, 11 with mild, 24 with moderate and 15 with severe dementia, while in the second set, 6 patients had MCI, 8 mild, 16 moderate and 13 severe dementia (Table 1).\n\nThe MMSE cognitive test was significantly different (p<0.001) in both sets between the four groups (Figure 1, Table 1). As expected, MCI patients had the highest MMSE score, followed by mild, moderate, and severe AD. In the first set, the mean age (years) was 74.5 for MCI, 71.4 for mild, 75.7 for moderate and 74.4 for severe dementia. In the second set, the mean age (years) was 67.6 for MCI, 76.2 for mild, 78.2 for moderate and 71.1 for severe dementia. In set 1 there were 3 females each in the MCI and mild dementia groups, 13 in moderate and 6 in severe dementia groups, whereas in set 2, 5 females were in the MCI group, 3 in mild, 6 in moderate and 2 in severe dementia groups. Between groups, there was no difference in age (p=0.514) or sex (p=0.504) in set 1, while a small difference was found for age (p=0.041) and sex (p=0.047) between the four groups in set 2.\n\nThe cognitive test Mini-Mental State Examination (MMSE) was compared between MCI, mild, moderate and severe AD dementia patients. A statistically significant difference in cognitive performance was observed among the four groups, in both sets (p<0.001). Horizontal lines represent medians. The number of patients per group is mentioned in Table 1.\n\nCurrent CSF biomarkers were tested in a fraction of MCI, mild, moderate and severe AD patients. A statistical difference was observed between disease groups for Aβ1-42 (decreasing with severity) and t-tau (increasing with severity) (p<0.05); Aβ1-42 levels were differentially expressed between MCI vs. moderate AD dementia and MCI vs. severe AD dementia (Supplementary Table 1). The distributions of Aβ1-42, t-tau and p-tau in the four groups are shown in Figure 2.\n\nThe concentrations of these proteins were compared between mild cognitive impairment (MCI) (n=10) mild (n=7), moderate (n=23) and severe Alzheimer’s disease (n=14). Aβ1-42 and t-tau were significantly different between the tested groups (p<0.05). Horizontal lines represent medians.\n\nFor evaluation of the 30 biomarker candidates, 101 CSF samples from MCI and AD patients were randomized into two separate sets.\n\nOverall, the majority of the proteins showed similar distribution patterns across AD stages, with a trend towards a decline in CSF concentration with disease progression. Among all proteins, only NPTXR showed a statistically significant difference between MCI vs. combined moderate and severe AD groups in both sets of patients (set 1: p=0.004, set 2: p=0.039). The concentration of NPTXR decreased in advanced stages. However, this significance did not remain after multiple comparison correction by the Holm’s method.\n\nSeveral other proteins also showed decreases in advanced stages of AD but did not consistently achieve statistical significance. In the first data set, proteins NPTXR, NPY and VGF were significantly different between the four groups (p=0.014, 0.033, 0.038, respectively), before correction for multiple comparison testing. After correction, the significance disappeared. Likewise, the findings observed in the first data set were not always-replicated in the second data set, but some proteins showed differential levels when comparing MCI vs. moderate and severe AD. These included BAI2, ECM1, FRRS1L, NPTXR, NPY, SLITRK1 and VGF (p=0.044, 0.033, 0.042, 0.004, 0.004, 0.048, 0.005 respectively). Proteins NPTXR, NPY and VGF were the most consistent, showing reductions in concentration with increasing AD severity.\n\nControl protein ECM1 did not differ among MCI, mild, moderate and severe AD patients (set 1 p=0.200, set 2 p=0.926) but differed between MCI vs. moderate and severe AD groups, only in set 1. However when multiple correction was applied, the difference disappeared.\n\nStatistical analysis of all candidates between the four groups and between MCI vs. moderate and severe AD dementia is shown in Supplementary Table 2.\n\nThe reproducibility of the assays for control samples (pools of non-pathological CSFs) and clinical samples was <20% (data not shown). The distributions of all candidate proteins between the four disease groups in sets 1 and 2 are shown in Figure 3 and Figure 4.\n\nCandidate proteins were measured with SRM assay and compared between mild cognitive impairment (MCI) (n=8), mild (n=11), moderate (n=24) and severe AD (n=15). Full gene names can be found in the website of the human gene nomenclature committee (https://www.genenames.org/).\n\nCandidate proteins were measured with SRM assay and compared between mild cognitive impairment (MCI) (n=6), mild (n=8), moderate (n=16) and severe AD (n=13). Full gene names can be found in the website of the human gene nomenclature committee (https://www.genenames.org/).\n\nDiagnostic performance was evaluated by calculating the AUC for discriminating MCI vs. moderate and severe AD dementia. Based on the performance of candidates in both sets, only NPTXR protein showed a significant and reproducible separation between the two groups. In the first set, the AUC for NPTXR was 0.799 (95% CI: 0.628, 0.928) and in the second set was 0.799 (95% CI: 0.586, 0.960). Figure 5 shows ROC curves for this protein in both sets.\n\nROC curve of NPTXR protein in set 1 and set 2; area under the curve value for set 1 was 0.799 (95% CI: 0.628, 0.928) and in set 2 was 0.799 (95% CI: 0.586, 0.960).\n\nPairwise Spearman’s rank correlation was used to assess if there is a correlation between protein candidates and the cognitive test MMSE score. A few proteins showed a significant positive correlation with MMSE score (Supplementary Table 3), which means that a lower score was associated with a lower protein concentration in CSF. Spearman’s rank correlation coefficients between level of these candidates and the cognitive test (for pairs significant at 0.05 level) was: 0.21 for BAI2, 0.23 for NCAN, 0.29 for NPY, 0.22 for OPCML, 0.29 for RTN4RL2, 0.26 for SCG2, 0.23 for SEZ6L, 0.25 for SST and 0.32 for VGF. The Spearman’s coefficient for NPTXR was 0.20 (not significant).\n\nOverall, there was 31% APOE ε4 carriers among disease patients (14% among MCI, 21% among mild, 30% among moderate and 46% among severe AD dementia). APOE ε4 homozygous patients were present only in mild (n=2) and severe AD (n=2) groups. There was no significant difference in distribution of ε4 carriers between disease patients with different severity (p=0.138). Overall, five APOE phenotypes were identified in all subjects, ε2/ε3, ε2/ε4, ε3/ε3, ε3/ε4 and ε4/ε4, with no difference in the APOE phenotype frequencies among tested groups (p=0.160). In set 1 all five APOE phenotypes were present, ε2/ε3 (n=2), ε2/ε4 (n=1), ε3/ε3 (n=38), ε3/ε4 (n=13) and ε4/ε4 (n=4), while in set 2 only three: ε2/ε3 (n=3), ε3/ε3 (n=27), ε3/ε4 (n=13). The frequencies of APOE phenotypes are shown in Table 2.\n\nAPOE, apolipoprotein E; AD, Alzheimer’s disease.\n\nIn both set of samples, none of the proteins showed a reproducible difference in abundance between APOE phenotypes (data not shown). Only FRRS1L protein showed a modest significance and only in the first set (p=0.040, when not adjusted for multiple comparison). There was no difference in proteins in set 2 between different phenotypes (p>0.05).\n\nIn order to validate our findings of decreased NPTXR in CSF of MCI and AD patients, we analyze CSF NPTXR by sandwich ELISA assay. For this independent validation, we used CSF samples from 12 AD patients, 21 patients with MCI and 23 control subjects. The results of CSF NPTXR concentration in the three groups of patients are shown in Figure 6. Controls had the highest level, followed by MCI and AD. The differences between controls and MCI were not statistically significant by the Mann-Whitney non-parametric test (p=0.52). Also, the differences were not significant between MCI and AD patients (p=0.10). However, the differences between controls and AD were highly significant (p=0.004). These results further support our hypothesis that NPTXR is a new CSF biomarker of AD, decreasing progressively with disease severity.\n\nThe differences were statistically significant only between controls and AD patients by Mann-Whitney test (p=0.004). Horizontal lines represent means and 25–75 percentiles. This cohort has been described elsewhere19. For more details see text.\n\n\nDiscussion\n\nThe main objective of the present study was to evaluate 30 brain-related proteins as CSF biomarkers of AD severity and progression. These highly specific brain proteins were assessed in AD patients with different stages, including MCI, mild, moderate and severe AD dementia. Protein NPTXR showed potential as a biomarker of disease progression. Significant and consistent differences in CSF NPTXR levels were observed between MCI vs. combined moderate and severe AD dementia groups.\n\nNPTXR protein is a member of the neuronal pentraxin family, expressed predominately in the brain, with the highest expression observed in the hippocampus and cerebellum24. This transmembrane presynaptic protein was suggested to be involved in the activation of both excitatory and inhibitory neurons25. NPTXR has been suggested previously as a potential prognostic biomarker26, more specific for AD, compared to Parkinson’s disease27. Differential abundance of NPTXR has been observed in asymptomatic carriers of AD familial mutations, compared to non-carriers, with elevated levels observed in asymptomatic carriers28. Wildsmith et al. observed a similar trend of NPTXR abundance between MCI and AD groups, as found here; i.e. increased levels in the MCI group, with AD patients having lower levels26. A decreasing abundance of NPTXR was also observed in longitudinally followed AD patients. This data are in accord with our study, since NPTXR levels declined proportionally with advanced AD dementia stages. Additional accumulating evidence by Hendrickson et al. further suggests that NPTXR represents a new biomarker of AD disease progression, decreasing with the severity of AD29. The latter study identified VGF as an additional progression biomarker, as we also found here.\n\nThe observed pattern of decline in abundance with disease severity for a few proteins, in addition to NPTXR, could be partially explained by their brain-specificity. Since our candidates were selected to have high specificity for brain tissue expression, these proteins could be expressed throughout various brain regions and their decrease may represent the overall decline in cortical volume, thus serving as markers of brain atrophy. This suggestion could be confirmed in future studies, by comparing brain imaging data with abundance of specific proteins in CSF. The superiority of NPTXR over other proteins may be related to its high expression in the hippocampus.\n\nAPOE ε4 allele represents the main genetic risk factor for developing AD. Carriers for APOE ε4 have earlier age of disease onset and more pronounced amyloid pathology compared to non-carriers. For example, amyloid plaques are more abundant in ε4 carriers30, with lower CSF concentration of Aβ1-42 in AD patients31. In addition to enhanced Aβ plaques in the brain, ε4 carriers exhibit vascular Aβ deposition31. Therefore, we aimed to evaluate if levels of our candidate biomarkers change with APOE phenotype. In this study, none of the candidates showed any reproducible difference in abundance between APOE phenotypes. However, this needs to be further investigated due to our limited sample size and considering that not all APOE phenotypes were identified in both set of samples.\n\nSome limitations of our study are associated with the selection of the 30 brain-specific proteins. Our focus was on predicted secreted and membrane-bound proteins, since the majority of HPA brain-enriched proteins are membrane and/or secreted15 and most of the CSF proteins are membrane-bound or secreted32. Intracellular proteins, which were excluded from this study, may have lower abundance in normal CSF but under pathological conditions they could be released into the CSF. Other limitations are related to the patients included in the study. Our cohorts did not include preclinical AD, which would allow assessment of the proposed candidates from the very early stage of developing AD. Cognitively healthy, age-matched controls were not included in this study since we aimed to test biomarkers in different stages of disease progression. Moreover, only a subset of patients had information on current AD biomarkers (Aβ1-42, t-tau, p-tau). Therefore, the diagnostic accuracy (ROC curve analysis) of NPTXR was not compared to the existing AD CSF biomarkers.\n\nOur candidate biomarkers should be further tested in individuals encompassing the whole AD continuum, from preclinical to more advanced clinical stages. Preferentially, longitudinally followed patients should be monitored to assess the prognostic potential of the candidates, over sufficient period of time, allowing disease progression to the next stage. The approximate annual rate of progression of MCI to AD dementia is 10 to 15%33.\n\nIn summary, in this study, we evaluated 30 brain-specific proteins as candidate CSF biomarkers of AD severity, utilizing multiplex mass spectrometry-based quantification. The protein NPTXR showed the most promise as a potential biomarker of disease progression. Interestingly, at least two other previous studies have also identified NPTXR as a highly promising biomarker of progression of AD. CSF NPTXR levels decline proportionally, as AD becomes more severe. This finding needs to be validated in a larger, longitudinally followed cohort. We suggest that NPTXR may have value as a CSF biomarker for assessing the efficacy of new therapies for AD.\n\n\nData availability\n\nDataset 1: Raw data for the results included in this study. DOI, http://dx.doi.org/10.5256/f1000research.15095.d20830434.",
"appendix": "Competing interests\n\n\n\nDr. Eleftherios P. Diamandis holds a consultant/advisory role with Abbott Diagnostics.\n\n\nGrant information\n\nThis work was supported by Mount Sinai Hospital.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary Table 1: Statistical analysis between four patient groups for Aβ1-42, t-tau and p-tau.\n\nClick here to access the data.\n\nSupplementary Table 2: Statistical analysis of candidates between four patient groups and MCI vs. moderate and severe AD dementia.\n\nClick here to access the data.\n\nSupplementary Table 3: Spearman’s rank correlation between candidate levels and MMSE score.\n\nClick here to access the data.\n\n\nReferences\n\nBallard C, Gauthier S, Corbett A, et al.: Alzheimer's disease. Lancet. 2011; 377(9770): 1019–1031. PubMed Abstract | Publisher Full Text\n\nMcKhann GM, Knopman DS, Chertkow H, et al.: The diagnosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement. 2011; 7(3): 263–269. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakamura A, Kaneko N, Villemagne VL, et al.: High performance plasma amyloid-β biomarkers for Alzheimer's disease. Nature. 2018; 554(4691): 249–54. PubMed Abstract | Publisher Full Text\n\nJack CR Jr, Albert MS, Knopman DS, et al.: Introduction to the recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement. 2011; 7(3): 257–262. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWHO: Dementia: a public health priority. WHO. 2012. Reference Source\n\nKroksveen AC, Opsahl JA, Aye TT, et al.: Proteomics of human cerebrospinal fluid: discovery and verification of biomarker candidates in neurodegenerative diseases using quantitative proteomics. J Proteomics. 2011; 74(4): 371–388. PubMed Abstract | Publisher Full Text\n\nBlennow K, Hampel H, Weiner M, et al.: Cerebrospinal fluid and plasma biomarkers in Alzheimer disease. Nat Rev Neurol. 2010; 6(3): 131–144. PubMed Abstract | Publisher Full Text\n\nBlennow K, de Leon MJ, Zetterberg H: Alzheimer's disease. Lancet. 2006; 368(9533): 387–403. PubMed Abstract | Publisher Full Text\n\nKhan TK, Alkon DL: Alzheimer's Disease Cerebrospinal Fluid and Neuroimaging Biomarkers: Diagnostic Accuracy and Relationship to Drug Efficacy. J Alzheimers Dis. 2015; 46(4): 817–836. PubMed Abstract | Publisher Full Text\n\nSunderland T, Wolozin B, Galasko D, et al.: Longitudinal stability of CSF tau levels in Alzheimer patients. Biol Psychiatry. 1999; 46(6): 750–755. PubMed Abstract | Publisher Full Text\n\nWilliams JH, Wilcock GK, Seeburger J, et al.: Non-linear relationships of cerebrospinal fluid biomarker levels with cognitive function: an observational study. Alzheimers Res Ther. 2011; 3(1): 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDrabovich AP, Dimitromanolakis A, Saraon P, et al.: Differential diagnosis of azoospermia with proteomic biomarkers ECM1 and TEX101 quantified in seminal plasma. Sci Transl Med. 2013; 5(212): 212ra160. PubMed Abstract | Publisher Full Text\n\nMartinez-Morillo E, Garcia Hernandez P, Begcevic I, et al.: Identification of novel biomarkers of brain damage in patients with hemorrhagic stroke by integrating bioinformatics and mass spectrometry-based proteomics. J Proteome Res. 2014; 13(2): 969–981. PubMed Abstract | Publisher Full Text\n\nBegcevic I, Brinc D, Drabovich AP, et al.: Identification of brain-enriched proteins in the cerebrospinal fluid proteome by LC-MS/MS profiling and mining of the Human Protein Atlas. Clin Proteomics. 2016; 13: 11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUhlén M, Fagerberg L, Hallström BM, et al.: Proteomics. Tissue-based map of the human proteome. Science. 2015; 347(6220): 1260419. PubMed Abstract | Publisher Full Text\n\nBegcevic I, Brinc D, Dukic L, et al.: Targeted mass spectrometry-based assays for relative quantification of 30 brain-related proteins and their clinical applications. J Proteome Res. 2018. PubMed Abstract | Publisher Full Text\n\nLiu CC, Liu CC, Kanekiyo T, et al.: Apolipoprotein E and Alzheimer disease: risk, mechanisms and therapy. Nat Rev Neurol. 2013; 9(2): 106–118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartínez-Morillo E, Nielsen HM, Batruch I, et al.: Assessment of peptide chemical modifications on the development of an accurate and precise multiplex selected reaction monitoring assay for apolipoprotein e isoforms. J Proteome Res. 2014; 13(2): 1077–1087. PubMed Abstract | Publisher Full Text\n\nBegcevic I, Brinc D, Brown M, et al.: Brain-related proteins as potential CSF biomarkers of Alzheimer’s disease: A targeted mass spectrometry approach. J Proteomics. 2018; 182: 12–20. PubMed Abstract | Publisher Full Text\n\nMcKhann G, Drachman D, Folstein M, et al.: Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 1984; 34(7): 939–944. PubMed Abstract | Publisher Full Text\n\nPetersen RC, Smith GE, Waring SC, et al.: Mild cognitive impairment: clinical characterization and outcome. Arch Neurol. 1999; 56(3): 303–308. PubMed Abstract | Publisher Full Text\n\nToledo JB, Zetterberg H, van Harten AC, et al.: Alzheimer's disease cerebrospinal fluid biomarker in cognitively normal subjects. Brain. 2015; 138(Pt 9): 2701–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJanes H, Longton G, Pepe M: Accommodating Covariates in ROC Analysis. Stata J. 2009; 9(1): 17–39. PubMed Abstract | Free Full Text\n\nDodds DC, Omeis IA, Cushman SJ, et al.: Neuronal pentraxin receptor, a novel putative integral membrane pentraxin that interacts with neuronal pentraxin 1 and 2 and taipoxin-associated calcium-binding protein 49. J Biol Chem. 1997; 272(34): 21488–21494. PubMed Abstract | Publisher Full Text\n\nLee SJ, Wei M, Zhang C, et al.: Presynaptic Neuronal Pentraxin Receptor Organizes Excitatory and Inhibitory Synapses. J Neurosci. 2017; 37(5): 1062–1080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWildsmith KR, Schauer SP, Smith AM, et al.: Identification of longitudinally dynamic biomarkers in Alzheimer's disease cerebrospinal fluid by targeted proteomics. Mol Neurodegener. 2014; 9: 22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYin GN, Lee HW, Cho JY, et al.: Neuronal pentraxin receptor in cerebrospinal fluid as a potential biomarker for neurodegenerative diseases. Brain Res. 2009; 1265: 158–170. PubMed Abstract | Publisher Full Text\n\nRingman JM, Schulman H, Becker C, et al.: Proteomic changes in cerebrospinal fluid of presymptomatic and affected persons carrying familial Alzheimer disease mutations. Arch Neurol. 2012; 69(1): 96–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHendrickson RC, Lee AY, Song Q, et al.: High Resolution Discovery Proteomics Reveals Candidate Disease Progression Markers of Alzheimer's Disease in Human Cerebrospinal Fluid. PLoS One. 2015; 10(8): e0135365. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchmechel DE, Saunders AM, Strittmatter WJ, et al.: Increased amyloid beta-peptide deposition in cerebral cortex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc Natl Acad Sci U S A. 1993; 90(20): 9649–9653. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrince JA, Zetterberg H, Andreasen N, et al.: APOE epsilon4 allele is associated with reduced cerebrospinal fluid levels of Abeta42. Neurology. 2004; 62(11): 2116–2118. PubMed Abstract | Publisher Full Text\n\nZhang Y, Guo Z, Zou L, et al.: A comprehensive map and functional annotation of the normal human cerebrospinal fluid proteome. J Proteomics. 2015; 119: 90–99. PubMed Abstract | Publisher Full Text\n\nHansson O, Zetterberg H, Buchhave P, et al.: Association between CSF biomarkers and incipient Alzheimer's disease in patients with mild cognitive impairment: a follow-up study. Lancet Neurol. 2006; 5(3): 228–234. PubMed Abstract | Publisher Full Text\n\nBegcevic I, Tsolaki M, Brinc D, et al.: Dataset 1 in: Neuronal pentraxin receptor-1 is a new cerebrospinal fluid biomarker of Alzheimer’s disease progression. F1000Research. 2018. Data Source"
}
|
[
{
"id": "36054",
"date": "26 Jul 2018",
"name": "Georgios Pampalakis",
"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\nAD is the most common neurodegenerative diseases. Unfortunately, there no molecular biomarkers for the disease yet. In the present study the authors used a well described cohort of AD patients to identify new protein-based biomarkers in CSF that can determine the progression of AD. The procedure was based on identifying potential protein biomarkers from Human Protein Atlas (30 brain-specific proteins were selected) and design analytical method based on MS for their determination in CSF. After careful examination of clinical specimens, the authors identified a new biomarker that was validated with ELISA, the NPTXR. The authors did not find any correlation of the potential biomarkers tested (including NPTXR) with the status of ApoE polymorphism. Finally, the limitations of their study are well-described. A major limitation as noted is the absence of normal samples for the analysis of the potential biomarkers. Although not entirely required for this study, analysis of normal CSF samples will help to identify biomarkers for AD diagnosis. However, the authors validated NPTXR in control samples with ELISA assay and found not statistically different levels between controls and patients with mild cognitive impairment. In conclusion, the present study is well-designed and adds new information on the development of new AD biomarker. It will worth in future studies to examine the levels of NPTXR in other neurodegerative diseases such as Parksinson to determine the specificity.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36373",
"date": "06 Aug 2018",
"name": "Edward W. Randell",
"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 are currently no blood based or CSF biomarkers commonly used on a routine basis for screening, diagnosis, or monitoring progression of Alzheimer’s disease. This work is an effort to fill the gap as it examines 30 candidate proteins from CSF and their relationship with disease progression. This was done by examining levels of the CSF proteins in patients at various stages of disease ranging from mild cognitive impairment to severe dementia. Initial assay of the 30 proteins was carried out using tryptic digestion and analysis of signature peptides by LC-MS/MS selective reaction monitoring technique. An attempt to validate results was undertaken by examination in a second set of patients representing the four stages of disease, from mild cognitive impairment to mild, moderate, and severe dementia. The study shows that the biomarker, neuronal pentraxin receptor-1 best discriminates mild cognitive impairment from advanced stages of the disease. This finding was confirmed by ELISA assay of the same protein. The study concludes that this CSF protein in a potential biomarker of Alzheimer’s disease progression with potential utility in monitoring treatment. The patient population was well described, and the experimental design sound. Findings of the study are well discussed with previous work and with limitations identified.\nAlthough, not explicitly stated this study apparently measures the 30 candidate biomarkers in CSF from 101 different individuals at different disease stages. Future use of neuronal pentraxin receptor-1, or any biomarker, for disease monitoring purposes at least partially involves establishing a baseline for individual patients and possible serial measurement. Based on information in the box and whisker plots, results for neuronal pentraxin receptor-1 in patients with mild cognitive impairment varied over a range that was about 3-fold and showed significant overlap with results of patients from the moderate and severe dementia groups. There were also noted outliers in the latter two groups (moderate and severe) showing results well within and even higher than in patients with mild cognitive impairment. Moreover, most of the biomarkers examined showed similar outliers in both sets of analyses. This presents many questions. It is not clear is these outliers for the different biomarkers are from the same patients (or samples) or if different patients and samples produce different biomarker outliers. Also, it is not clear if the neuronal pentraxin receptor-1 outliers, which represent 5 to 10% of the samples from patients with moderate and severe disease, are causes by pre-analytical issues concerning samples or confounding physiological/pathological processes in these patients. It is also not clear how variable levels of this protein is in CSF from an individual patient; and the imprecision of the two different assays (LC-MS/MS method and ELISA) for neuronal pentraxin receptor-1 used was not stated. Between-run assay imprecision and intra-individual biological variability will have bearing on the usefulness of a CSF neuronal pentraxin receptor-1 measurement in practice. It would also be interesting to see how levels change in individual patients as Alzheimer’s disease progresses. But, the requirement of repeated lumbar puncture to assess disease course using this or any CSF biomarkers will be a difficult sell given the absence of effective disease modifying therapy. Nevertheless, this does not take away from the value of this early work in highlighting potential value of neuronal pentraxin receptor-1 for disease progression with potential future clinical value if effective disease modifying treatments become available. But as implied by the authors, what would be most useful is a biomarker, like neuronal pentraxin receptor-1, that not only correlates with disease severity but is also predictive of Alzheimer’s disease progression and related outcomes.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36375",
"date": "08 Aug 2018",
"name": "Julie Shaw",
"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\nCurrently there are no biomarkers for AD used in clinical practice. Current methods for evaluating patients only allow for diagnoses of probable AD while definitive diagnosis can only take place at the time of post-mortem examination. This study aims to identify much needed biomarkers for AD in CSF using mass spectrometry. The authors used tandem mass spectrometry to measure the concentrations of 30 CSF-specific proteins, which they previously identified, in patient specimens. The patient cohort included those with mild cognitive impairment, mild, moderate and severe AD, as determined by the MMSE and CDR ratings. The authors identified one candidate protein, neuronal pentraxin receptor -1 (NPTXR), as promising in its ability to differentiate MCI from moderate and severe AD. Independent assessment of NPTXR concentrations in patient specimens using an ELISA confirmed that concentrations of this protein are lower in CSF from patients with moderate and severe AD compared to control patients.\n\nThis manuscript is well written and I only have minor comments, which I will outline below.\n\nIn the results section, the authors state that the reproducibility of the assays for control samples and clinical samples was <20%. It is unclear to me what this refers to and it would be helpful for the authors to clarify, especially since the data are not shown.\n\nHow was cognitive level assessed for the patient specimens used for the ELISA measurements? It appears that these specimens came from a different institution than those used for the mass spectrometry analysis. If a different assessment scheme was employed for these specimens, this could be confounding. I would suggest that the authors comment on this.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "36772",
"date": "15 Aug 2018",
"name": "Mohd M. Khan",
"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 the US alone, Alzheimer’s disease (AD) is the 6th leading cause of death. Unfortunately there are no clinical biomarkers of AD that could be used to efficiently and successfully diagnose the disease onset, monitor the AD progression, and evaluate the efficiency of therapeutic interventions against AD. According to Alzheimer's association, early and accurate AD diagnosis could save $7.9 trillion in hospital-associated costs. Hence to fill this diagnostic gap, present work aims to identify and validate AD biomarkers in cerebrospinal fluid (CSF) patient samples using mass spectrometry-based targeted proteomics. The manuscript is well written and methods are either sufficiently described and/or appropriate earlier works are cited (particularly for targeted method development and protein quantitation). However I suggest authors should consider submitting the raw MS data to ProteomeXchange if not done already.\n\nTargeted proteomics was used to quantify 30 candidate proteins in CSF samples obtained from 101 patients that had mild to severe AD as well as patients with mild cognitive impairment (MCI). In both the sets, 28 severe AD, 40 moderate AD, 19 mild AD, and 14 MCI patient samples were quantified for candidate biomarkers using SRM method. One biomarker candidate of interest neuronal pentraxin receptor-1 (NPTXR), quantified by mass spectrometry and validated in a separate set by ELISA, was found useful in differentiating AD from control samples. Severe AD patients had lower levels of NPTXR in CSF than those in control CSF samples; in contrast, considerable variation in the levels of NPTXR levels in MCI patients was observed. However the rationale behind not segregating samples in 5 groups (control, MCI, mild-, moderate-, and severe-dementia) for ELISA validation is missing and could be added. For the NPTXR validation, authors could have done secondary validation of NPTXR levels in present cohort (large group; n=101) followed by an independent validation in a separate cohort as done in Figure 6 (n=56). Nonetheless this early work on AD biomarker discovery is interesting and a longitudinal study will be helpful in establishing NPTXR as a true biomarker of AD progression and disease severity.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1012
|
https://f1000research.com/articles/7-500/v1
|
26 Apr 18
|
{
"type": "Research Article",
"title": "Scepticaemia: The impact on the health system and patients of delaying new treatments with uncertain evidence; a case study of the sepsis bundle",
"authors": [
"Robin Blythe",
"David Cook",
"Nicholas Graves",
"David Cook",
"Nicholas Graves"
],
"abstract": "Background: A sepsis care bundle of intravenous vitamin C, thiamine, and hydrocortisone was reported to improve treatment outcomes. The data to support it are uncertain and decision makers are likely to be cautious about adopting it. The objective of this study was to model the opportunity costs in dollars and lives of waiting for better information before adopting the bundle. Methods: A decision tree was built using information from the literature. We modelled the impact of bundle adoption under three scenarios using a simulation in which the bundle was effective as reported in the primary trial, less effective based on other information, and ineffective. The measurements were health services costs, quality-adjusted life years, and transition probabilities. Results: If the bundle proves to be effective under either scenario, it will save billions of dollars and millions of life-years in the United States. This is the opportunity cost of delaying an adoption decision and waiting for better quality evidence. We suggest that hospital decision-makers consider implementing the bundle on a trial basis while monitoring costs and outcomes data even while the evidence base is uncertain. Conclusions: If the decision maker is unwilling to use the best available evidence now, but rather wishes to wait for definitive evidence they are risking incurring large costs for health care systems and for the patients they serve. An explicit analysis of uncertain clinical outcomes is a useful adjunct to guide decision making where there is clinical ambiguity. This approach offers a valid alternative to the default of clinical inactivity when faced with uncertainty.",
"keywords": [
"Sepsis",
"Costs and Cost Analysis",
"Evidence-based practice"
],
"content": "Introduction\n\nSepsis arises frequently among patients admitted to hospitals in the US and elsewhere. It is often fatal, accounting for 30% to 50% of inpatient deaths1. Those who survive incur large costs from increased risk of organ damage2. A recent paper by Marik et al. (2017) described the effectiveness of a treatment bundle made up of intravenous vitamin C, thiamine and hydrocortisone3. Because of small sample size, non-randomised design, large observed treatment effects and the simple and low cost characteristics of the bundled intervention, there was scepticism among clinicians, administrators and payers regarding adoption4–6. Rapid adoption without good evidence has risks and costs, yet delaying an effective treatment imposes larger costs to patients and the health system. Waiting for conclusive evidence might be a poor and costly strategy and should be balanced against the likely costs of rapid adoption with uncertain and low quality evidence.\n\nThe aim of this paper is to estimate the economic consequences of a decision to adopt the Marik sepsis bundle early, under conditions of large uncertainty. This is compared to the alternative, which is to wait one year; the time period between publication of the first paper in June 20173 and the completion of the National Institutes of Health funded trial due in mid-2018. The trial will ideally reduce the uncertainty in an adoption decision. This paper predicts the cost savings and health benefits of adopting the Marik bundle immediately with uncertain evidence for the entire United States health system, compared to waiting for better quality evidence. Our estimates examine likely changes to costs and health outcomes if the treatment is found to be effective, less effective and ineffective.\n\n\nMethods\n\nAn incremental cost-effectiveness analysis was conducted on the change to costs and Quality-Adjusted Life Years (QALYs) of standard sepsis care compared to the early adoption of the Marik bundle3. Values for expected costs and health utilities were taken from the literature, with probabilities of treatment outcomes taken from Marik’s paper3. The time horizon was 5 years in order to sufficiently measure the long-term outcomes of acute kidney injury (AKI), an important and common result of severe septic shock2. A decision tree was programmed in TreeAge Pro 2017 R2.1 (Williamstown, MA2017) and prior statistical distributions of costs, outcomes and probabilities used to include uncertainties in the data, see Figure 1. Patients receiving either modality progress through a chance node to death or survival. Surviving patients progress through a second chance node, where they would either recover or recover with organ damage. AKI is a common outcome in cases of sepsis, and typically require renal replacement therapy, which can take the form of either dialysis or organ transplantation2.\n\nThe costs of care include the hospital cost of a sepsis episode, the additional cost of the Marik treatment bundle and the annual ongoing cost of renal replacement therapy (RRT). Hospital costs were taken from a systematic review by Arefian et al. (2017) for a US perspective with a mean of $32,421 and standard deviation of $15,051 per stay7. For the Marik bundle, we used an estimate of $528 and a standard deviation of 5% at $26.50 per patient per course. The annual cost of RRT for end stage renal disease varies depending on treatment modality, such as dialysis or transplant, and a weighted average of $76,936, with a standard deviation of $15,387 was taken from the 2017 USRDS Annual Data Report8.\n\nThe health utility score of patients reflects the value of their heath state and is a measure of health related quality of life. It is expressed on a range between zero, the worst health state, and one, the best possible health state9. The scores for patients who recover from septic shock depend on whether or not they suffered AKI. Patients who did not suffer AKI still underwent the stress of intensive care unit recovery, and their mean health utility scores and standard deviations were taken from Cuthbertson et al. (2010) at 1, 2.5, and 5 years as 0.666 (0.280), 0.701 (0.281), and 0.677 (0.301), respectively10. Survivors of AKI were found to have a utility of 0.40 with a standard deviation of 0.37 at 60 days. We did not find evidence of improved health utilities over the 5 year period11. The health utility of death was valued at 09.\n\nThe transition probabilities of patients moving through the decision tree were informed by data from Marik’s study. This includes the probabilities of survival and AKI, see Table 13. In Marik’s study, the bundle’s probability of death and AKI were 0.085 and 0.097, with standard deviations of 0.041 and 0.043 respectively. These compared to the standard care mortality and AKI probabilities of 0.404 and 0.367, with standard deviations of 0.071 and 0.070 respectively.\n\nScenario 1 uses probabilities from the literature for mortality and AKI; costs and utilities were unchanged. The attributable mortality risk and rate of AKI in the general population vary in the literature and are different to those reported in Marik’s control cohort. We used the mortality risk identified in Rhee et al. (2017) of 21.2% and the rate of sepsis-associated AKI in Angus et al. (2001) of 22.0%12,13. These are lower than the rates seen in Marik’s observation group. We estimated large standard deviations for mortality and AKI at 30% and 28% of the means respectively to include Marik’s treatment group results as possible outcomes within 95% confidence intervals.\n\nScenario 2 represents a worst-case scenario, in which Marik’s bundle is ineffective and there is no difference in probabilities between the bundle and standard care. We modelled scenario 2 using the same transition probabilities found in Rhee et al. and Angus et al. for both arms of the decision tree. As in scenario 1, costs and utilities remained unchanged.\n\nProbabilistic sensitivity analysis was conducted on the decision tree by taking 1000 random resamples of values from the prior statistical distributions of model parameters. This propagates all uncertainty in the data forward to the results and provides useful information for decision making.\n\n\nResults\n\nThe total economic cost of patients under standard care, including 5 year estimations of RRT, was $116,427 compared to the total economic cost of patients receiving the Marik bundle of $66,744, an expected saving of $49,683 per patient. The annual incidence of sepsis in the United States is 1.5m cases suggesting a total expected cost saving of $74.5bn USD. Over the same 5-year period, patients gain an additional 1.26 QALY per case, or 1.9m QALYs per year for the population. Given the reduced costs and improved outcomes, the Marik bundle dominates standard care, as it both saves costs and increase health outcomes at the same time. The probabilistic sensitivity analysis shows this conclusion arises 98.0% of the time.\n\nScenario 1 reduces the mean expected economic cost of standard care by $17,166 to $100,244 per patient, compared to the Marik bundle at $67,581. The expected cost saving between the bundle and standard care $32,663 per patient, a total saving of $49.0bn USD, down from $74.7bn. Altering the transition probabilities to values from the literature also reduces the QALYs gained, with patients gaining only 0.55 QALY per case, or 825,000 QALY per year over 1.5m cases per year. In this scenario, the Marik bundle still dominates standard care by saving costs and improving health outcomes. The probabilistic sensitivity analysis shows this conclusion arises 90.8% of the time.\n\nScenario 2 is that there is no change to mortality outcomes and AKI between Marik’s bundle and standard care. The mean cost per patient over 5 years is $98,932USD for standard care and $99,459USD for the Marik bundle, an increase of $527 per patient over a 5 year period. There is no difference in QALYs between treatment alternatives. We were unable to find any evidence of harm to the patient as a result of the bundle. Hydrocortisone and thiamine are already present in routine sepsis care, and the dosage of 6g of vitamin C per day has been shown to be safe unless contraindicated14–16. In this scenario, assuming the intervention is universally adopted, no patients will have been harmed, but healthcare payers and providers would have spent an additional $791m USD per year. Outcomes from the scenario analyses are listed in Table 2 below.\n\n\nDiscussion\n\nWe found that adopting the Marik bundle has a high likelihood to save billions of dollars and generate millions of extra QALYs under the conditions outlined in Marik’s paper and in an alternate scenario that uses other data. Under the ineffective treatment of scenario 2 costs are increased to health services by $0.791bn. These results reveal substantial opportunity costs in dollars and lives if we fail to implement and the bundle is ultimately found to be effective, even if the treatment effect is lower than purported by Dr Marik. If the bundle does not work then some costs have been incurred by hospitals for no heath gain. Not adopting the bundle because the evidence for effectiveness is currently uncertain could well be a poor strategy.\n\nThe scenario analysis uses values at which the Marik bundle is less effective, specifically by aligning observation group figures with the literature. The mortality rate of sepsis under standard care was suggested by Marik to be 40.4%. Sepsis mortality has been declining in the US, from 46.9% in the early 1990s to 21.2% in 2014, declining by about 3% per annum and driven by improved organ support systems and protocoled early recognition and treatment13,17,18. It might be that Marik’s study featured unusually sick patients in the control group. Study participants are often unrepresentative of the general population and the intervention group may have been less likely to suffer an adverse outcome19. This progress in sepsis care management is complicated by the fact that claims data has shown concurrently increasing sepsis incidence and decreasing mortality, so the literature is conflicted20,21. It is possible that the increased incidence from claims data is due to increased reimbursement received by US hospitals for sepsis compared to other diseases, and patients are being misdiagnosed. An increasing rate of misdiagnosis of patients that do not have sepsis increases the denominator of septic cases while mortality stays the same in the numerator, creating the illusion of declining mortality13,21.\n\nRegarding the mortality of Marik’s treatment group, we note the significant improvement in outcomes recommending the use of steroids in adults with septic shock by Annane et al. (2018)22. Patients randomised to the steroid group (n = 614) showed a 6% absolute reduction in 90-day all-cause mortality when compared to placebo (n = 627). Therefore, the figures in Marik’s patient cohort may have been unusual, but they are plausible.\n\nThere is no current body of evidence that suggests this bundle is dangerous. Indeed, the combination of vitamin C and thiamine with steroids would have to cause an attributable death once in every 10 sepsis patients - by a hitherto unimagined and novel mechanism - to negate the modelled benefits of its early adoption.\n\nOur study excluded the costs of bundle implementation, including training, labour and the potential for high costs of de-implementing an unsuccessful treatment. These would have increased the cost of bundle implementation, so the incremental cost difference of the Marik bundle may be understated. We noted that hydrocortisone, as part of the bundle, was found by Venkatesh et al. (2018) to speed up resolution of shock and reduce the need for blood transfusions, so the in-hospital cost-savings for the treatment group may also be understated16. There is also considerable uncertainty around the parameter estimates that are available. We did not have access to primary data, including clinical data on costs and transition probabilities for each patient, and were reliant on the literature.\n\nAn average hospital in the US may treat around 230 sepsis patients per year12. By implementing the bundle, it will spend an additional $528 per patient, or $121,440 per year. Conservative estimates from the scenario analysis shows cost savings of $32,644 and a gain of 0.55 QALYs per patient. If the treatment was effective for just four patients out of 230, or 1.7%, it will have paid for itself in terms of total economic costs. Comparing implementing the sepsis bundle to other hospital-based treatment studies shows that for 230 patients, the bundle costs less than a tenth of a standard phase I clinical trial, which run from $1.4m-$6.5m23.\n\nThe case for not adopting the Marik bundle has several components. Scientific and empirical evidence is thin, and a single-site Vitamin C trial showing remarkable results only to be proven ineffective after a multi-site RCT is a clinical trope. If the bundle was ineffective, health systems will have added an unnecessary load to clinicians and implementers, which would then have to be de-implemented. Administrators and clinicians may be less likely to adopt novel treatments in the future. Introducing the bundle on the current evidence may also set a bad precedent for novel treatments, eroding the authority of the presiding physician and giving more credence to largely unproven interventions. The proposal in this paper does not replace the need for the clarity provided by good science and empirical research, but these are not always immediately available. We provide an approach to explicitly guide the interim decisions that must be made under these circumstances.\n\nThe Marik bundle is a somewhat unusual case relative to most ‘miracle’ interventions later found to be ineffective. As the analysis shows, it is an extremely cheap treatment with the potential to reduce rates of mortality and kidney damage at no risk to the patient. The delay between publication of the pilot study and results of the large RCT due in mid-2018 could create substantial opportunity costs in dollars and lives, and while there is not perfect evidence, under the circumstances it might be sufficient for hospitals and health systems to choose whether to conduct their own trials and not only independently verify their results, but also publish their findings and improve the availability of evidence around the treatment bundle.\n\nThe implementation decision of the Marik bundle relies upon the willingness of health administrators to use the available evidence to influence policy. We attempted to make the bundle choice as intuitive as possible, using straightforward trade-offs, simple modelling techniques, and a realistic decision process. Merlo et al. (2014) showed that for research to be accessible to decision-makers, it must be contextually relevant, contain little jargon, and put the terms of the implementation decision into terms that specify a trade-off that will be familiar to decision-makers24. If the decision maker is unwilling to use the best available evidence now, but rather wishes to wait for definitive evidence they are risking incurring large costs for health care systems and for the patients it serves.\n\n\nConclusion\n\nAn explicit analysis of uncertain clinical outcomes is a useful adjunct to guide decision making where there is clinical ambiguity. This approach offers a valid alternative to the default of clinical inactivity when faced with uncertainty.\n\n\nData availability\n\nDataset 1. Sepsis Distributions Table: Data on distributions used in the analysis. 10.5256/f1000research.14619.d20141125",
"appendix": "Competing interests\n\n\n\nThe authors have no financial interests, affiliations, gifts, grants, equipment or drugs to disclose.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nLiu V, Escobar GJ, Greene JD, et al.: Hospital deaths in patients with sepsis from 2 independent cohorts. JAMA. 2014; 312(1): 90–2. PubMed Abstract | Publisher Full Text\n\nAlobaidi R, Basu RK, Goldstein SL, et al.: Sepsis-associated acute kidney injury. Semin Nephrol. 2015; 35(1): 2–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarik PE, Khangoora V, Rivera R, et al.: Hydrocortisone, Vitamin C and Thiamine for the Treatment of Severe Sepsis and Septic Shock: A Retrospective Before-After Study. Chest. 2017; 151(6): 1229–38. PubMed Abstract | Publisher Full Text\n\nRezane S: The Marik Protocol: Have We Found a \"Cure\" for Severe Sepsis and Septic Shock? R.E.B.E.L. EM; 2017. Reference Source\n\nFarkas J: PulmCrit - Metabolic sepsis resuscitation: the evidence behind Vitamin C. emcrit.org; 2017. Reference Source\n\nSGEM: SGEM#174: Don't Believe the Hype - Vitamin C Cocktail for Sepsis. thesgem.com; 2017. Reference Source\n\nArefian H, Heublein S, Scherag A, et al.: Hospital-related cost of sepsis: A systematic review. J Infect. 2017; 74(2): 107–17. PubMed Abstract | Publisher Full Text\n\nUSRDS: 2017 ADR Reference Table K: Economic Costs of ESRD. 2017. Reference Source\n\nTorrance GW: Measurement of health state utilities for economic appraisal. J Health Econ. 1986; 5(1): 1–30. PubMed Abstract | Publisher Full Text\n\nCuthbertson BH, Roughton S, Jenkinson D, et al.: Quality of life in the five years after intensive care: a cohort study. Crit Care. 2010; 14(1): R6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohansen KL, Smith MW, Unruh ML, et al.: Predictors of health utility among 60-day survivors of acute kidney injury in the Veterans Affairs/National Institutes of Health Acute Renal Failure Trial Network Study. Clin J Am Soc Nephrol. 2010; 5(8): 1366–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAngus DC, Linde-Zwirble WT, Lidicker J, et al.: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med. 2001; 29(7): 1303–10. PubMed Abstract | Publisher Full Text\n\nRhee C, Dantes R, Epstein L, et al.: Incidence and Trends of Sepsis in US Hospitals Using Clinical vs Claims Data, 2009-2014. JAMA. 2017; 318(13): 1241–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMallat J, Lemyze M, Thevenin D: Do not forget to give thiamine to your septic shock patient! J Thorac Dis. 2016; 8(6): 1062–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOhno S, Ohno Y, Suzuki N, et al.: High-dose vitamin C (ascorbic acid) therapy in the treatment of patients with advanced cancer. Anticancer Res. 2009; 29(3): 809–15. PubMed Abstract |\n\nVenkatesh B, Finfer S, Cohen J, et al.: Adjunctive Glucocorticoid Therapy in Patients with Septic Shock. N Engl J Med. 2018; 378(9): 797–808. PubMed Abstract | Publisher Full Text\n\nStevenson EK, Rubenstein AR, Radin GT, et al.: Two decades of mortality trends among patients with severe sepsis: a comparative meta-analysis*. Crit Care Med. 2014; 42(3): 625–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevy MM, Rhodes A, Phillips GS, et al.: Surviving Sepsis Campaign: association between performance metrics and outcomes in a 7.5-year study. Intensive Care Med. 2014; 40(11): 1623–33. PubMed Abstract | Publisher Full Text\n\nKalata P, Martus P, Zettl H, et al.: Differences between clinical trial participants and patients in a population-based registry: the German Rectal Cancer Study vs. the Rostock Cancer Registry. Dis Colon Rectum. 2009; 52(3): 425–37. PubMed Abstract | Publisher Full Text\n\nGlück T: Mortality from Severe Sepsis Decreasing in the U.S. www.jwatch.org: NEJM. 2013. Reference Source\n\nDressler DD: Is Sepsis Incidence and Mortality Changing in U.S. Hospitals?www.jwatch.org: NEJM. 2017. Reference Source\n\nAnnane D, Renault A, Brun-Buisson C, et al.: Hydrocortisone plus Fludrocortisone for Adults with Septic Shock. N Engl J Med. 2018; 378(9): 809–18. PubMed Abstract | Publisher Full Text\n\nSertkaya A, Wong HH, Jessup A, et al.: Key cost drivers of pharmaceutical clinical trials in the United States. Clin Trials. 2016; 13(2): 117–26. PubMed Abstract | Publisher Full Text\n\nMerlo G, Page K, Ratcliffe J, et al.: Bridging the gap: exploring the barriers to using economic evidence in healthcare decision making and strategies for improving uptake. Appl Health Econ Health Policy. 2015; 13(3): 303–9. PubMed Abstract | Publisher Full Text\n\nBlythe R, Cook D, Graves N: Dataset 1 in: Scepticaemia: The impact on the health system and patients of delaying new treatments with uncertain evidence; a case study of the sepsis bundle. F1000Research. 2018. Data Source"
}
|
[
{
"id": "33910",
"date": "11 May 2018",
"name": "Paul E. Marik",
"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 outstanding paper that objectively evaluates the potential benefit of the vitamin C protocol for sepsis.This evaluation was impeccably conducted and reported. The impact of this paper is significant.\nA few comments that will improve the paper:\n1.The authors state \"The delay between publication of the pilot study and results of the large RCT due in mid-2018 could create substantial opportunity costs\". This study which the authors cite is the NIH funded study due for completion in mid-2018 which is investigating vitamin c (ALONE) and not the combination AND in patients with ARDS (not specifically sepsis). Secondly, this study is not powered for a mortality difference. Currently according to Clintrials.gov there are 8 RCTS testing the combination and 3 testing Vitamin C alone. This suggests that it may take at least until the end of 2019 before the results of any of these studies are available and we can make firm conclusions.\n2. The authors refer to the \"sepsis bundle\"; I would avoid using this term as it will create confusion with the Surviving Sepsis Bundle.. widely known as the Sepsis Bundle. I would call it the Vitamin C protocol, the Marik Protocol, or something similar.\n3. Approximately 50% of sepsis survivors develop the post-sepsis syndrome, defined as severe physical and cognitive dysfunction not unlike PTSD. This syndrome is associated with enormous suffering and high costs. It is likely that the Marik Protocol will reduce the incidence and severity of this syndrome. While the cost savings are difficult to quantitate, I think that this proposition should be mentioned in the discussion section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "34658",
"date": "04 Jun 2018",
"name": "Adam Visser",
"expertise": [
"Reviewer Expertise Critical care medicine",
"sepsis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 fascinating paper uses a simple health economics model to argue that delaying the widespread implementation of the Vitamin C protocol as described by Marik and colleagues would lead to the loss of enormous potential savings to the health industry. The argument is well presented, easily followed, and sound, with some exceptions.\n\nThere is likely to be some inaccuracy in the figures presented. The authors state that there are 1.5 million cases of sepsis in the United States yearly (no reference supplied, but data easily available via the CDC). Whilst there may indeed be this number who meet the recently updated definition of sepsis, this large cohort of patients are different to those who have a mortality rate of 40% with standard care. This group, as described in Marik’s paper, are a much sicker population. Their number is likely to be significantly fewer than 1.5 million. Whilst it is not known whether Vitamin C may have benefits in a less sick population, it is unlikely that ‘early adopters’ of this therapy will consider it in those that are not towards the severe end of the sepsis spectrum. Reducing the exposed population in the model from 1.5 million will reduce the proposed savings by a proportional amount.\n\nSimilarly, the quoted incidence of requirement for renal replacement therapy is overstated. It is true that the Marik Vitamin C-treated group had an incidence of RRT-requiring AKI of 9.7%. However, the Marik paper does not state that these patients all went on to require ongoing RRT for ESRF (at the authors’ quoted cost of $77,000 per year). The majority of patients who require RRT in the setting of sepsis-induced AKI will recover their renal function either completely, or enough that they do not require ongoing dialysis. Again, reducing the stated financial benefits of this aspect of the Vitamin C treatment will reduce the long-term economic attractiveness of the therapy.\n\nOverall however, the authors are correct in their statement that the economic implications of adopting the ‘bundle’ will be positive. Furthermore, the economic implications should be secondary to the beneficial effects the treatment may have on mortality. Again, this therapy is unlikely to be offered to 1.5 million patients per year, but whatever the number, those who receive it are likely to benefit.\n\nA major strength of the paper is their use of a more pessimistic scenario (scenario 1) than published in the Marik study to demonstrate the robustness of their figures. The Marik-stated improvement in mortality, from 40% to 8%, is likely to not be repeated in larger studies, as it is simply too great to be plausible.\n\nReducing the relative risk reduction such that mortality is reduced from 21% to 8% may be more realistic, and yet the economic figures still add up to considerable savings.\n\nIt would perhaps also be useful to see a model where baseline mortality of 40% (ie the very sick cohort from Marik) have their mortality reduced to, say, 30% (ie a more plausible magnitude of effect from Vitamin C).\n\nThe authors have done well to pick this particular therapy to demonstrate their model. The Vitamin C bundle is unique in recent years, in that it is a cheap and seemingly safe therapy. Their model is unlikely to be able to be used for supporting other therapies that emerge in coming years, given the likelihood that these therapies will come from the pharmaceutical industry and have large price tags attached.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-500
|
https://f1000research.com/articles/7-519/v1
|
30 Apr 18
|
{
"type": "Software Tool Article",
"title": "eXamine: Visualizing annotated networks in Cytoscape",
"authors": [
"Philipp Spohr",
"Kasper Dinkla",
"Gunnar W. Klau",
"Mohammed El-Kebir",
"Philipp Spohr",
"Kasper Dinkla",
"Gunnar W. Klau"
],
"abstract": "eXamine is a Cytoscape app that displays set membership as contours on top of a node-link layout of a small graph. In addition to facilitating interpretation of enriched gene sets of small biological networks, eXamine can be used in other domains such as the visualization of communities in small social networks. eXamine was made available on the Cytoscape App Store in March 2014, has since registered more than 7,200 downloads, and has been highly rated by more than 25 users. In this paper, we present eXamine's new automation features that enable researchers to compose reproducible analysis workflows to generate visualizations of small, set-annotated graphs.",
"keywords": [
"functional subnetwork modules",
"visualization",
"enrichment",
"graph drawing",
"network community",
"Euler diagram"
],
"content": "Introduction\n\nThe Cytoscape app eXamine visualizes a small graph and a collection of node sets. The main purpose of eXamine is to aid in the interpretation of a small subnetwork module extracted from a large biological network1. Cytoscape apps like jActiveModules2 or external tools like Heinz3 extract subnetwork modules from a protein-protein interaction network given gene expression data. To interpret the identified subnetwork module, a frequent follow-up analysis is to compute enrichment of the nodes of the identified module in terms of known annotations such as from the Gene Ontology (GO)4,5 or from the Kyoto Encyclopedia of Genes and Genomes (KEGG)6. These annotations are a collection of node sets. eXamine provides a visual analysis approach that facilitates interpretation of the subnetwork module and the identified node sets by biologists. Given a small collection of node sets, eXamine generates a visualization of a small graph as a node-link layout together with contours for the selected sets (Figure 1). More specifically, the layout is computed using a variation of the algorithm described in 7. This algorithm preserves topological distances along inter-module links as much as possible, while making sure that none of the nodes overlap. In addition, spanning graphs are derived for those sets that have been selected by the user. These graphs are included in the computation of topological distances between nodes, pulling the nodes closer together. The spanning graphs are also used to draw the set contours, by dilating and eroding the associated links to form the rounded shapes that visually encompass and connect nodes.\n\nThe graphs generated in the two use cases using eXamine’s automation features.\n\nAs an alternative to eXamine, the existing group layout of Cytoscape can be used to show node partitions by visualizing disjoint sets in separate circles. The Venn and Euler diagram app8 for Cytoscape visualizes overlapping sets. However, in both cases network and group analysis are visually separate. The RBVI Cytoscape plugins9 provide a group viewer that aggregates groups into meta-nodes, without making group overlaps explicit.\n\nHere, we present a new version of eXamine that uses Cytoscape’s recently introduced automation features. We demonstrate eXamine’s new automation features using two use cases. The first use case replicates the case study provided in the original eXamine publication1. The second use case, the analysis of a social network, demonstrates that eXamine is applicable to other domains beyond computational biology. eXamine’s new automation features enable the composition of reproducible analysis workflows that generate appealing visualizations of small, set-annotated graphs.\n\n\nMethods\n\neXamine is implemented in Java and is available as an app for Cytoscape 3. We used WebCola algorithms http://ialab.it.monash.edu/webcola/ to simultaneously lay out nodes, links, and set contours. We refer to the original publication1 for additional implementation details regarding the used visualization techniques. Since the typical analysis workflows of eXamine consist of relatively simple commands that do not require streaming of complex data, we implemented the automation features through the ‘Command’ interface rather than the ‘Function’ interface. As a result, eXamine’s commands can be either executed from the command line or through Cytoscape’s REST (CyREST) interface from a Jupyter notebook.\n\nAs eXamine requires Cytoscape 3.6 to run, eXamine has the same system requirements as Cytoscape 3.6. eXamine can be operated through the Cytoscape graphical user interface (GUI) or through the new ‘Command’ interface of Cytoscape. We refer to the original publication1 for GUI instructions. In the following, we will describe and use the new ‘Command’ interface of eXamine.\n\nTable 1 provides a summary of the API of the eXamine commands and their parameters. To enable workflow authors to use eXamine’s automation most effectively, we also generated Swagger-based documentation that describes the commands and arguments. This documentation can be accessed in the Cytoscape menu: Help → Automation → CyREST Command API. Figure 2 shows a typical workflow of eXamine analysis.\n\nFigure 2 shows an example workflow using the below commands.\n\nFirst, we import a network, followed by importing node annotations that associate each node with of set of groups. Next, we optionally select a smaller subnetwork to visualize. We generate an internal representation of the groups, and import additional group annotations. After selecting the groups to visualize, we export an image of the visualization. Alternatively, we can launch a window that allows the user to select different groups.\n\n\nUse cases\n\nTo illustrate the new ‘Command’ interface of eXamine, we present two use cases. In the first use case, we describe a workflow to study a small subnetwork module extracted from the KEGG mouse network6. In the second use case, we describe a workflow to study Zachary’s karate club, a well-known social network. Both workflows are available as Jupyter notebooks. The workflows require Cytoscape v3.6 and a recent Python version (Python v2.7 or Python v3.6).\n\nThe Human Cytomegalovirus (HCMV) is a highly-contagious herpes virus. Previously1, we used eXamine to interpret a small subnetwork module (17 nodes and 18 edges) extracted from the KEGG mouse network using Heinz3 given gene expression data of an HCMV-infected mouse cell line. Node sets of this subnetwork module were annotated using enriched pathways from KEGG and enriched terms from GO. Below, we provide Python code that uses Cytoscape’s and eXamine’s automation feautures to generate Figure 1.\n\n1. To begin, we define a helper function that creates HTTP GET requests.\n\n\n\n2. We import the full KEGG mouse network using the ‘network import url‘ command provided by Cytoscape.\n\n\n\n3. We import node annotation, containing set membership information, using the ‘table import url‘ command provided by Cytoscape.\n\n\n\n4. Using the ‘network select’ command provided by Cytoscape, we select the nodes that comprise the subnetwork module identified by Heinz3.\n\n\n\n5. Using the ‘examine generate groups’ command, we generate group nodes for each member of the specified sets that occur in the module (the nodes we previously selected).\n\n\n\n6. We import additional annotations that describe the generated groups using the ‘import table url’ command of Cytoscape.\n\n\n\n7. As described in 1, we consider groups from the sets Pathway and Function using the ‘examine update settings’ command. In addition, we specify the columns that contain node labels and URLs as well as group scores and annotations.\n\n\n\n8. Using the ‘examine select groups’ command, we select the same groups as described in 1.\n\n\n\n9. Finally, we export the visualization to a scalable vector graphics (SVG) file using the ‘examine export’ command. Figure 1a shows the resulting graph.\n\n\n\nAlternatively, using the ‘examine interact’ command, we can launch an interactive visualization window that allows us to select different groups.\n\n\n\nWe consider the graph ‘Zachary’s karate club’, which is an undirected social network of friendships between 34 members of a karate club at a US university in the 1970s10. Here, we use eXamine to visualize the six overlapping communities of this network identified in 11.\n\n1. We use the same helper function as defined in the previous use case.\n\n2. We import the network using the ‘network import url‘ command provided by Cytoscape.\n\n\n\n3. We import node annotation, containing set membership information, using the ‘table import url‘ command provided by Cytoscape.\n\n\n\n4. Using the ‘network select’ command provided by Cytoscape, we select all nodes of the network.\n\n\n\n5. Using the ‘examine generate groups’ command, we generate group nodes for each member of the specified sets that annotate the nodes of the network.\n\n\n\n6. We consider groups from the set Community using the ‘examine update settings’ command.\n\n\n\n7. We select all six groups using the ‘examine select groups’ command.\n\n\n\n8. Finally, we export the visualization to an SVG file using the ‘examine export’ command. Figure 1b shows the resulting graph.\n\n\n\n\nDiscussion\n\neXamine is limited to visualizing small, relatively sparse networks. It is not possible to use eXamine to construct a comprehensive layout if the network consists of hundreds of nodes or if there are dozens of annotation sets to visualize at the same time. This is a natural limitation of any visualization approach based on node-link diagrams and set contours.\n\neXamine currently uses Cytoscape’s CyREST API to import networks and their annotations. If apps that provide gene enrichment analysis functionality, such as BiNGO12, would expose this functionality through the CyREST API, we envision updating the workflow to include this type of upstream analysis.\n\n\nSummary\n\neXamine is a Cytoscape app for a set-oriented visual analysis approach for small annotated graphs that displays set membership as contours on top of a node-link layout (Figure 1). In this paper, we presented new automation features for eXamine that are accessible through Cytoscape’s REST API (Table 1). As such, researchers can embed eXamine in reproducible and well-documented workflows that generate appealing visualizations of small, set-annotated graphs (Figure 2). We demonstrated two such workflows in the context of computational network biology and social network analysis.\n\n\nData and software availability\n\n1. Cytoscape app store: http://apps.cytoscape.org/apps/examine\n\n2. Source code: https://github.com/ls-cwi/eXamine\n\n3. Latest commit as of release: https://github.com/ls-cwi/eXamine/commit/adeea322cdd6a787baca98adfe70c4165c4d892c\n\n4. Archived source code as at time of publication: https://doi.org/10.5281/zenodo.121856913\n\n5. License: GPL v2.0\n\n6. Jupyter notebooks: https://github.com/ls-cwi/eXamine/blob/master/doc/tutorial/eXamineNotebook/eXamineTutorial.ipynb and https://github.com/ls-cwi/eXamine/blob/master/doc/tutorial/eXamineNotebook/eXamineTutorial2.ipynb.",
"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\nDinkla K, El-Kebir M, Bucur CI, et al.: eXamine: exploring annotated modules in networks. BMC Bioinformatics. 2014; 15: 201. PubMed Abstract | Publisher Full Text | Free 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–S240. PubMed Abstract | Publisher Full Text\n\nDittrich MT, Klau GW, Rosenwald A, et al.: Identifying functional modules in protein-protein interaction networks: an integrated exact approach. Bioinformatics. 2008; 24(13): i223–31. 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\nThe Gene Ontology Consortium: Expansion of the Gene Ontology knowledgebase and resources. Nucleic Acids Res. 2017; 45(D1): D331–D338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanehisa M, Furumichi M, Tanabe M, et al.: KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res. 2017; 45(D1): D353–D36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDwyer T, Koren Y, Marriott K: IPSEP-COLA: an incremental procedure for separation constraint layout of graphs. IEEE Trans Vis Comput Graph. 2006; 12(5): 821–828. PubMed Abstract | Publisher Full Text\n\nVenn and Euler Diagrams: Venn and Euler Diagrams. Reference Source\n\nRBVI Cytoscape Plugins: RBVI Cytoscape Plugins - Cytoscape Group Support. Reference Source\n\nZachary WW: An information flow model for conflict and fission in small groups. J Anthropol Res. 1977; 33(4): 452–473. Publisher Full Text\n\nChen W, Liu Z, Sun X, et al.: A game-theoretic framework to identify overlapping communities in social networks. Data Min Knowl Discov. 2010; 21(2): 224–240. Publisher Full Text\n\nMaere S, Heymans K, Kuiper M: BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics. 2005; 21(16): 3448–3449. PubMed Abstract | Publisher Full Text\n\nDinkla K, Spohr P, El-Kebir M, et al.: ls-cwi/eXamine: Automation Submission Version (Version v2.1.1). Zenodo. 2018. Data Source"
}
|
[
{
"id": "33593",
"date": "18 May 2018",
"name": "Barry Demchak",
"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 describes the use of new Commands exposed by the eXamine app.\n\nThe article is well written and adequately describes the Commands. My comments are advisory, and I hope they can improve the paper.\nThe procedure for using Commands should be well known to a proficient Cytoscape user. However, there's an important class of user that wants to use Automation but isn't a Cytoscape expert. To help him/her along good reference Cytoscape manual (Commands Tool), and good to point out that the commands can be used from the Commands dialog, a separate Commands script (via the Cytoscape command line or the Tools | Execute Command File menu item), or via Python. (See last paragraph of Methods section.) Good, too, to remind the user that eXamine commands are in the examine namespace, and that different namespaces resolve to Cytoscape and other apps.\nThis paper is about Python usage. Good to remind user that similar calls can be made from R.\nIn the Introduction, it would be helpful to justify eXamine automation by giving an example of workflows that become possible. This motivates the user. This could be as simple as summarizing Use cases.\nFor Discussion, it may be worth speculating on the value of providing Python and R libraries that act as cover functions for the eXamine REST calls. Python and R programmers really want to think in terms of Python and R, not REST.\nThe color coding in Figure 2 is very effective. Can you also color code the corresponding REST calls in the Use Case section? This may require changing the color scheme so the examples can be easily read. (Maybe Red/Green or Red/Blue??)\nIn Use Cases, good to say where the Jupyter notebooks are ... even if you identify them at the end of the paper.\nAs a side note, I notice that all eXamine endpoints are GET. The Commands convention we're using now is POST, with GET being deprecated. In this case, it doesn't matter much, as the parameter list isn't long and there isn't any return result that would benefit from JSON encoding. For a future release, good to consider POST versions, too. That way, if an error occurs, you'll be able to return it in a CIResponse structure.\nFor the BASE_URL, the datasets are necessary to run the examples. Can you list them at the end of the paper? Of course, they must be persistently available for the life of the paper, correct??\nIn Use Case 1, step 2, it would be good to give a sentence or two explaining how/why a user would have created this dataset.\nThat's it ... nice job!\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3785",
"date": "05 Jul 2018",
"name": "Mohammed El-Kebir",
"role": "Author Response",
"response": "We thank the referee for the positive feedback. Below our point-by-point response to the referee's comments. We included the following sentences along with a reference to the Cytoscape documentation: ‘Figure 1 shows a typical workflow of eXamine analysis, where the commands provided are in the `examine' namespace (red) and the commands provided by Cytoscape are colored blue.’ ‘We note that the commands described in the two use cases can also be directly executed from Cytoscape via the `Commands dialog' or a separate `Commands script'.3’ We now provide R code for the two use cases. We reworded the last paragraph of the introduction to more clearly describe the benefits of eXamine's new automation features. We wrote the following sentences in Discussion: ‘Finally, it would be good to enhance the API to return richer R and Python data types rather than a flag indicating whether the command succeeded. For instance, upon a `table import' it would be good to actually return a dataframe. This, however, will require switching from the `Command' interface to the `Function' interface.’ Thank you for the excellent suggestion about the coloring. We colored eXamine commands red and Cytoscape commands blue. We included URLs to the Jupyter and R markdown documents. We now use POST commands. We now include links to the datasets. They are indeed persistently archived using zenodo. We included the following sentences in step 2 of use case 1: ‘This a protein-protein interaction network that is typically used for the analysis of high-throughput biological data in the context of a biological network. The Cytoscape app KEGGscape provides functionality for importing pathways from KEGG [9].’"
}
]
},
{
"id": "34810",
"date": "20 Jun 2018",
"name": "Richa Batra",
"expertise": [
"Reviewer Expertise patient stratification",
"disease endotyping",
"clustering",
"network analysis",
"pathway enrichment",
"classification"
],
"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 presents new features to an existing cytoscape app namely eXamine. eXamine was initially developed to visualize small graphs and node sets. This new version of the app has automation features for reproducible research. Given the importance of reproducible research, this new version of the app is highly relevant.\n\nFigure 2 should really be Figure 1. For someone new to eXamine, it aptly summarizes the app. Also the summary at the end is very clear.\n\nIt would have been great to have a small dataset and step by step guide through the cytoscape app. Then small R script that I can run to generate the same results. Could you extend it for R users? Do you plan it in future work?\nIf the paper is about automation of eXamine via python, then it should reflect in title. The current name of the paper is misleading.\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3784",
"date": "05 Jul 2018",
"name": "Mohammed El-Kebir",
"role": "Author Response",
"response": "We thank the referee for the positive feedback. Below our point-by-point response to the referee's comments. We have merged Figures 1 and 2 in the revised manuscript. We included R example code in the revised manuscript."
}
]
},
{
"id": "34808",
"date": "21 Jun 2018",
"name": "Markus List",
"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 “eXamine: Visualizing Annotated Networks in Cytoscape”, Spohr et al. describe a new version of their Cytoscape tool eXamine. eXamine allows user to enrich the visualization of small networks via colored contours which represent specific annotations of subnetworks, e.g. functional annotation. eXamine is very useful because it not only highlights group properties in a visually appealing way but also allows the representation to be changed interactively. The software always attempts to achieve an optimal layout in which overlaps are kept at a minimum. Here, the authors have adapted eXamine to make use of the new Cytoscape automation features which allow for programmatic control of the plotting process, thus allows for generating reproducible results that can further be embedded in scripts.\nThe manuscript is well written and describes the new features and their usage clearly.\nThe code of eXamine is hosted on github and contains the two use cases from the manuscript in the form of popular Jupyter notebooks. Each step of the examples is well described and illustrates to the user how eXamine can be used in a scripting environment such as Python. In particular less obvious steps like extracting network ids from a JSON result help new users to achieve results fairly quickly.\nI have only minor comments:\nUnfortunately the second use case (Jupyter notebook) in the git repository did not work due to the use of a local file path instead of a public URL.\n\nAdd the end of the first paragraph the manuscript says “by dilating and eroding the associated links”. I’m not sure what is meant here with erosion.\n\nUse case 1, first paragraph, there is a typo: feautures -> features\n\nThe github readme should briefly describe the jupyter notebooks and link to them.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3783",
"date": "05 Jul 2018",
"name": "Mohammed El-Kebir",
"role": "Author Response",
"response": "We thank the referee for the positive feedback. Below our point-by-point response to the referee's comments. We fixed the link to the git repository for the second use case. We changed the wording of the sentence where we previously used 'erosion'. We fixed the typo. Thank you for pointing this out. We updated the Github README.md file."
}
]
},
{
"id": "33592",
"date": "25 Jun 2018",
"name": "Alexander Pico",
"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\nMany researchers arrive at point in enrichment analysis where they have tables of enriched ontology terms and pathways, but don't know how to go beyond presenting those data as tables or perhaps bar graphs. The use of networks is an intuitive and powerful visualization option. The eXamine app for Cytoscape provides a nice approach for displaying a simplified version of enrichment results. A couple steps are missing, however, and a few minor edits are suggested.\nMissing steps:\nAll sources of enrichment that I'm aware of provide rows of terms and a list of genes in a single field per row. It would be nice if you provided a bit of Python to transform that common input data type to that required by eXamine.\n\nI was wondering about genes that belong to more than one term (often the case). I looks like your tool handles these via a piped list. Is that right? Please describe the correct syntax in the text so readers don't have to guess.\n\nMinor suggestions:\nGroups in Cytoscape are no longer provided as a separate plugin, but rather are now integrated into the core of Cytoscape. Your reference to \"RBVI Cytoscape plugins\" should be updated to simply mention the Cytoscape manual and point to the section on groups: http://manual.cytoscape.org/en/stable/Creating_Networks.html#grouping-nodes\n\nThe LaTeX formatting for your Python code is unfortunately rendering single quotes in a way that throws SyntaxError in Python. Can you update the rendering so one can simply copy/paste snippets?\n\nIn the next version of this paper, I'd highly recommend using py2cytoscape (https://github.com/cytoscape/py2cytoscape). This will make the code you have write much more concise and easier to maintain or adapt. Likewise, I'd really like to see R examples and you can leverage RCy3 to make that code easy to write as well (http://bioconductor.org/packages/release/bioc/html/RCy3.html).\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3782",
"date": "05 Jul 2018",
"name": "Mohammed El-Kebir",
"role": "Author Response",
"response": "We thank the referee for the positive feedback. Below our point-by-point response to the referee's comments. Regarding the missing details on generating the enrichment input, the DAVID webtool (https://david.ncifcrf.gov/) yields a file where the rows are genes containing a list of terms. Implementing a generic Python function that transforms enrichment output that comes in various to our required format is a challenging task. Ideally, the enrichment step should be part of the workflow, as described in Discussion. We handle genes with multiple terms using lists, and the default separator used by Cytoscape’s import table functionality is indeed a pipe character. We updated the tutorials and text to describe this. We updated the reference to \"RBVI Cytoscape plugins\". We replaced all single quotes by double quotes, which are retained by the used LaTeX package for displaying the code (package is ‘listing’). Simple copy-paste of the commands into the Python interpreter should work now. We thank the referee for the suggestions regarding R and Python. We now use the RCy3 library for the two use cases in R. As for py2cytoscape, we found that the functionality that we need is not supported directly and requires separate script files for each command. We found that using the REST interface directly was easier."
}
]
}
] | 1
|
https://f1000research.com/articles/7-519
|
https://f1000research.com/articles/7-57/v1
|
16 Jan 18
|
{
"type": "Research Article",
"title": "Crystal structures of a llama VHH antibody BCD090-M2 targeting human ErbB3 receptor",
"authors": [
"Igor E. Eliseev",
"Anna N. Yudenko",
"Vera V. Vysochinskaya",
"Anna A. Svirina",
"Anna V. Evstratyeva",
"Maria S. Drozhzhachih",
"Elena A. Krendeleva",
"Anna K. Vladimirova",
"Timofey A. Nemankin",
"Viktoria M. Ekimova",
"Andrey B. Ulitin",
"Maria I. Lomovskaya",
"Pavel A. Yakovlev",
"Anton S. Bukatin",
"Nickolay A. Knyazev",
"Fedor V. Moiseenko",
"Oleg B. Chakchir",
"Anna N. Yudenko",
"Vera V. Vysochinskaya",
"Anna A. Svirina",
"Anna V. Evstratyeva",
"Maria S. Drozhzhachih",
"Elena A. Krendeleva",
"Anna K. Vladimirova",
"Timofey A. Nemankin",
"Viktoria M. Ekimova",
"Andrey B. Ulitin",
"Maria I. Lomovskaya",
"Pavel A. Yakovlev",
"Anton S. Bukatin",
"Nickolay A. Knyazev",
"Fedor V. Moiseenko",
"Oleg B. Chakchir"
],
"abstract": "Background: The ability of ErbB3 receptor to functionally complement ErbB1-2 and induce tumor resistance to their inhibitors makes it a unique target in cancer therapy by monoclonal antibodies. Here we report the expression, purification and structural analysis of a new anti-ErbB3 single-chain antibody. Methods: The VHH fragment of the antibody was expressed in E. coli SHuffle cells as a SUMO fusion, cleaved by TEV protease and purified to homogeneity. Binding to the extracellular domain of ErbB3 was studied by surface plasmon resonance. For structural studies, the antibody was crystallized by hanging-drop vapor diffusion in two different forms. Results: We developed a robust and efficient system for recombinant expression of single-domain antibodies. The purified antibody was functional and bound ErbB3 with KD = 1 μM. The crystal structures of the VHH antibody in space groups C2 and P1 were solved by molecular replacement at 1.6 and 1.9 Å resolution. The high-quality electron density maps allowed us to build precise atomic models of the antibody and the putative paratope. Surprisingly, the CDR H2 existed in multiple distant conformations in different crystal forms, while the more complex CDR H3 had a low structural variability. The structures were deposited under PDB entry codes 6EZW and 6F0D. Conclusions: Our results may facilitate further mechanistic studies of ErbB3 inhibition by single-chain antibodies. Besides, the solved structures will contribute to datasets required to develop new computational methods for antibody modeling and design.",
"keywords": [
"cancer",
"therapeutic antibodies",
"receptor tyrosine kinase",
"HER3",
"single-domain antibody",
"nanobody",
"crystal structure"
],
"content": "Introduction\n\nReceptor tyrosine kinases ErbB1-4 (HER1-4) receive inputs from growth factors and transmit signals to the cell nucleus, thus regulating key cellular processes such as growth, differentiation, migration, and apoptosis1. Aberrations of ErbB signaling, caused by mutations or receptor overexpression, are associated with the development of a wide variety of cancers. The essential role of ErbB receptors in tumor development makes them a unique target in cancer therapy by monoclonal antibodies2. Therapeutic antibodies often act on the first stage of signal transduction by inhibiting ligand binding or receptor dimerization.\n\nThe first two members of the family, ErbB1 (EGFR, HER1) and ErbB2 (HER2/neu), were early recognized as promising drug targets because of their frequent overexpression in tumors. The examples of successful application of anti-ErbB antibodies in cancer treatment include cetuximab (anti-EGFR) in head and neck cancer therapy and trastuzumab (anti-HER2) in breast cancer treatment. The role of the third member, ErbB3, has long been underestimated because it lacks intrinsic tyrosine kinase activity. However, its ability to form functional dimers with ErbB1-2 and to confer resistance to their inhibitors makes ErbB3 an important drug target3. Particularly, it was shown that inhibition of ErbB2 with lapatinib caused transcriptional up-regulation of ErbB3, which was then phosphorylated by residual ErbB2 kinase activity thus limiting antitumor effect4. A comprehensive clinical study revealed that ErbB3 overexpression was a significant marker of reduced survival in patients with breast cancer5.\n\nThis new data stimulated the development of anti-ErbB3 antibodies, which are at various stages of clinical trials6. The rational antibody design requires knowledge of molecular mechanism of ErbB3 inhibition. Recently, several structures of ErbB3-antibody complexes were solved7–9. Surprisingly, these structures showed that antibodies target entirely different epitopes on the receptor: extracellular domain I7, domains II and IV8, or domain III9.\n\nIn Russia, anti-ErbB3 antibodies are developed by BIOCAD biotechnology company. The phage display selection of antibodies from immunized llamas and subsequent sequencing allowed the identification of several anti-ErbB3 single-chain antibodies. As a part of our ongoing effort to elucidate the molecular mechanism of ErbB3 inhibition and ultimately open up a possibility of therapeutic application of these antibodies, we study their thermodynamic stability10, functional properties, and structure. In this work, we describe the expression, purification, crystallization and structural analysis of the variable fragment of an antibody BCD090-M2 (RRID:AB_2721074), which demonstrated an affinity to the extracellular domain of ErbB3 in preliminary experiments.\n\n\nMethods\n\nGene fragment encoding the VHH fragment of the antibody was cloned into pSolSUMO expression vector (Lucigen), following the manufacturer’s recommendations. Briefly, the fragment was PCR amplified using the primers: forward 5’-aatctgtacttccagggtcaggtgcagctggtgcag-3’, reverse 5’-gtggcggccgctctattatgaggagacggtgaccgt-3’, with the first 18 nucleotides in both primers matching the ends of linearized pSolSUMO vector. Following the amplification, the fragment was mixed with linearized pSolSUMO vector and used to transform chemically competent E. cloni 10G cells (Lucigen). The resultant plasmid pSolSUMO-BCD090-M2, encoding SUMO with N-terminal hexahistidine tag fused to BCD090-M2 through a TEV recognition site, was sequenced and used for further protein expression in E. coli SHuffle T7 Express cells (NEB).\n\nChemically competent E. coli SHuffle T7 Express cells (NEB) were transformed by pSolSUMO-BCD090-M2, and single colonies were used to start small-scale overnight cultures. Then 2–4 l bacterial cultures were inoculated by 1:100 volume of overnight culture and grown in 2xYT supplemented with 50 µg/ml kanamycin at 37°C. At OD 0.6–0.8, protein expression was induced by the addition of L-Rhamnose to a final concentration 5 mM, temperature was lowered to 27–29°C and cells were grown overnight for additional 14–15 h. Cells were then harvested by centrifugation at 10000g (5 min), resuspended in IMAC buffer (50 mM Na2HPO4 pH 8.0, 0.3 M NaCl, 5 mM Imidazole) with 1 mM PMSF, 0.5 mM EDTA as protease inhibitors and lysed by ultrasonication. Cell debris were pelleted by centrifugation at 40000g (20 min), and the cell extract supernatant was filtered through 0.22µm membrane. The solution was loaded on 1 ml IMAC column cOmplete (Roche) at 0.5–1 ml/min, the column was washed with 20–40 column volumes of IMAC buffer, and then the protein was eluted by IMAC buffer with 0.3 M Imidazole.\n\nAfter elution from IMAC column, sample purity was usually higher than 90% as judged by SDS-PAGE, and the protein was cleaved by TEV protease. Sample was first dialyzed against TEV buffer (30 mM Tris pH 8.0, 0.5 mM EDTA, 1 mM DTT) overnight at 4°C, then mixed with TEV protease at 1:40 to 1:80 enzyme to substrate ratio and cleaved for 4 h at 25°C with mild agitation. Histidine-tagged SUMO was then removed by three repeats of negative IMAC chromatography in batch mode. Sample in TEV buffer with 50 mM NaCl and 5 mM Imidazole was mixed with 200µl Ni-NTA agarose resin (Qiagen) and incubated for 30 mins with mild agitation, the agarose beads were pelleted by a short centrifugation, and the supernatant was taken and used for the next round of SUMO depletion.\n\nFinally, the cleaved VHH antibody BCD090-M2 was purified by an additional polishing step of high-resolution cation exchange chromatography on a MonoS 5/50 GL column (GE Healthcare). The protein was dialyzed against IEX buffer (20 mM Na Acetate pH 6.0) overnight at 4°C, loaded on a pre-equilibrated column, and eluted by IEX buffer with 0–0.5 M NaCl gradient over 20 column volumes. The peak fractions analyzed by SDS-PAGE were pooled, dialyzed overnight against Sample buffer (20 mM HEPES pH 7.5, 50 mM NaCl), and concentrated on a 10 kDa Amicon centrifuge concentrators (Millipore). Protein concentration was measured spectrophotometrically with the parameters ε=27055 M−1cm−1, MW=13955 Da calculated from the amino acid sequence with the ProtParam tool11.\n\nFor affinity measurements, we produced an extracellular domain (residues 21–643) of the human ErbB3 receptor using a pEE vector with a CMV promoter carrying the ErbB3 gene fragment followed by a hexahistidine tag and a FLAG-tag. CHO-T-HC cells were transfected with PEI and grown one day in HyCell TransFx-C media (GE Healthcare) at 37°C. On the day 2, the temperature was lowered to 32°C, and cells were grown for an additional 8 days. Then the cells were harvested by sterile filtration through Opticap XL capsule filters (Millipore), the clarified culture fluid was supplemented with 1 mM NiCl2 and 10 mM Imidazole, and loaded on a HisTrap HP (GE Healthcare) column equilibrated with IMAC buffer. The column was washed with 10 volumes of IMAC buffer, and the protein was eluted with 0.3 M Imidazole. After elution from IMAC column, the protein was further purified by size-exclusion chromatography on HiLoad 16/600 Superdex 200pg (GE Healthcare), dialyzed against PBS and concentrated on a 10 kDa Amicon centrifuge concentrators (Millipore).\n\nInteraction of the recombinantly expressed VHH antibody BCD090-M2 with the extracellular domain of the ErbB3 receptor was studied by surface plasmon resonance technique using a Biacore T200 instrument (GE Healthcare). The purified extracellular domain was diluted in 10 mM Na Acetate buffer pH 4.5 to a final concentration of 260µg/ml and immobilized on CM5 chip via amine coupling with EDC/NHS, following the manufacturer’s recommendations. The BCD090-M2 stock solution was serially diluted in HBS-P buffer (10 mM HEPES pH 7.4, 0.15 M NaCl, 0.005% v/v surfactant P20) to prepare concentration series from 4 µM down to 60 nM. Each sample was injected to a cell with immobilized receptor and a reference cell at 10 µl/min flow rate and association/dissociation time of 120/600 s. Between the samples the chip surface was regenerated by a 30 s pulse of glycine-HCl pH 2.0 and equilibrated with HBS-P. The sensograms were reference-subtracted and analyzed in Biacore T200 evaluation software. The equilibrium dissociation constant was obtained by fitting the response measured at 5 s before the end of association phase as a function of analyte concentration.\n\nThe BCD090-M2 crystallization conditions were screened using the commercial sparse-matrix screens Classics I and II, AmSO4 (Qiagen), Clear Strategy I and II, Morpheus (Molecular Dimensions). Crystallization experiments were set up by sitting-drop vapor diffusion method in 96-well plates at 19°C. Each crystallization drop consisted of 100 nl protein solution at a concentration of 17 mg/ml in Sample buffer (20 mM HEPES pH 7.5, 50 mM NaCl) and 100 nl reservoir solution. The screening revealed the two classes of promising conditions, one with salts of carboxylic acids and PEG (Morpheus #73, #76), and the other with divalent cations and PEG (Classics II #64). The crystallization experiments with the identified conditions were reproduced in 24-well Linbro plates by hanging-drop vapor diffusion method with 2 µl drop volume, 1:1 ratio, and 0.5 ml reservoir volume at 20°C. In the case of first crystallization condition (Morpheus #73), well-formed crystals of 0.2–0.3 mm size appeared after 3–4 days in the hanging-drop experiments. Preliminary X-ray experiments showed diffraction up to 1.6 Å, therefore no further optimization was attempted; the crystals for data collection were obtained using the original precipitant solution #73 from Morpheus screen. In the case of the second crystallization condition (Classics II #64), optimization experiments were made to increase crystal size and improve morphology. It appeared that among divalent cations only Cd2+ was essential for crystallization, the optimized reservoir solution had the following composition: 0.1 M MES pH 6.5, 12% PEG 3350, 5 mM CdSO4. Large crystals up to 0.7 mm usually appeared after 4–5 days and diffracted below 2.0 Å.\n\nThe crystals grown in the first crystallization condition (Morpheus #73) were mounted in loops, cryoprotected in the mother liquor with 25% glycerol, and flash cooled in cold nitrogen gas stream. The crystals grown with Cd2+ deteriorated upon soaking in different cryoprotectant solutions, and so were mounted in thin-walled quartz capillaries (Hampton Research) for room-temperature data collection. All diffraction data were collected on a Kappa Apex II diffractometer (Bruker AXS) using CuKα radiation. The datasets were integrated with SAINT V8.18C and scaled with SADABS v. 2008/1 software12. The crystal grown in the first condition diffracted to 1.6 Å, the unit cell parameters were a=65.76 Å, b=38.93 Å, c=47.48 Å, α=γ=90°, β =102.24°, the space group C2 was determined with XPREP v. 2008/212. Notably, the crystal grown in the second condition with Cd2+ appeared triclinic with unit cell parameters a=35.77 Å, b=41.53 Å, c=46.49 Å, α=89.99°, β =67.92°, γ=76.06° and two copies of the VHH antibody in the asymmetric unit, and diffracted to a slightly lower resolution of 1.9 Å. The details of data collection and processing are summarized in Table 1.\n\nStatistics for the highest-resolution shell are shown in parentheses.\n\n\nResults and discussion\n\nThe llama VHH antibody BCD090-M2 was expressed in soluble form in the cytoplasm of E. coli SHuffle T7 Express cells as a SUMO fusion. The SHuffle strain13 has deletions of the genes trxB and gor, and constitutively expresses a chromosomal copy of the disulfide bond isomerase DsbC to promote formation of correct disulfide bonds in recombinant proteins. The system proved very efficient for single-chain antibody expression, the typical protein yield in our experiments was 50 mg of a fusion protein per liter of a bacterial culture after the first IMAC step. For the IMAC we used new chromatography media cOmplete (Roche), which withstands high EDTA and DTT concentrations and demonstrates a different binding strength and specificity compared to traditional Ni-NTA resins. Particularly, our histidine-tagged fusion protein eluted at 30 mM Imidazole concentration in gradient elution experiments, and addition of only 5 mM Imidazole to cell extract and wash buffer efficiently suppressed non-specific binding, resulting in greater than 90% purity after the first chromatography step. The protein was then cleaved by TEV protease to obtain untagged antibody BCD090-M2 with nearly native N-terminus, differing from the original sequence by a single additional glycine residue left from the TEV recognition site. The extent of cleavage was monitored by SDS-PAGE and typically was more than 70% at 1:80 enzyme ratio and more than 85% at 1:40 ratio, as shown in Figure 1 (left panel).\n\nThe antibody was expressed in the cytoplasm of E. coli SHuffle cells as His6-SUMO fusion, purified by IMAC and cleaved with TEV protease at 1:80 or 1:40 enzyme:substrate ratio. The bottom band marked with a red arrow corresponds to BCD090-M2, the middle band marked with a blue arrow is a histidine-tagged SUMO, and a top band is an uncleaved protein. After the cleavage, BCD090-M2 was further purified by negative IMAC and cation-exchange chromatography on MonoS, the SDS-PAGE analysis of the purified antibody is presented in the right panel.\n\nThe bottom band with apparent molecular mass 14 kDa corresponds to processed BCD090-M2, the middle band corresponds to His6-SUMO (13 kDa) and migrates anomalously slow probably due to positively charged histidine tag, and the top band is the uncleaved protein. The processed BCD090-M2 was separated from His6-SUMO and intact fusion protein by negative IMAC chromatography in batch regime, and then polished by an additional step of cation exchange chromatography on a MonoS column. After elution from MonoS, the VHH antibody was almost pure as judged by SDS-PAGE shown in Figure 1 (right panel).\n\nThe surface plasmon resonance experiment confirmed that the recombinant VHH antibody was functional and efficiently bound to the immobilized receptor. The experimental binding sensograms are shown in Figure 2A, and the processed data fitted with an equilibrium binding model is shown in Figure 2B. The obtained equilibrium dissociation constant for monovalent binding was 1 µM. Although the monovalent affinity is a fundamental characteristic of antibody-antigen interaction, the avidity of a full-length bivalent antibody can be much higher. Typically, the enhancement of avidity due to a bivalent interaction is 3–4 orders of magnitude and depends strongly on the surface concentration of antigen14.\n\nThe purified extracellular domain (residues 21–643) was immobilized on CM5 chip surface. A) binding sensograms measured for different concentrations of BCD090-M2; B) fitting of the measured response as a function of concentration of free BCD090-M2 in solution with a simple bimolecular equilibrium binding model gives KD=1 μM.\n\nThe protein was successfully crystallized in two different forms: in space group C2 with a single copy of BCD090-M2 in the asymmetric unit and in P1 with two molecules and two cadmium ions in the unit cell. After data collection and processing, all further data analysis procedures, including phasing, model building, and refinement, were conducted in Phenix software suite v. 1.11.1_257515. To solve the structures by molecular replacement, we selected a set of single-domain antibodies from PDB with the highest homology with BCD090-M2. We then processed the search models with Sculptor16 to delete residues that were not aligned with the target and to prune side chains by the Schwarzenbacher algorithm17, and performed molecular replacement using Phaser v. 2.7.1618. The best molecular replacement solutions for both datasets were obtained with a nanobody targeting complement receptor Vsig4, PDB:5IMK19. Then we used phenix.autobuild20 to automatically build the framework regions of the BCD090-M2 and Coot v. 0.8.6.121 to manually fit the missing CDRs into the experimental electron density. The structures were refined using phenix.refine22 (see Table 1 for refinement statistics) and deposited to the Protein Data Bank under entry codes 6EZW and 6F0D. All figures were generated using PyMOL.\n\nThe overall structure of BCD090-M2 crystallized in space group C2 is shown in Figure 3. The framework regions are drawn as grey ribbons, and the CDRs are colored in orange (H1), blue (H2), and red (H3). The CDRs were defined according to Kabat23 as shown in Figure 4.\n\nFramework regions are grey and CDRs are colored.\n\nThe numbering is the same as in the PDB files 6EZW and 6F0D. The CDR regions according to Kabat definition are indicated with color and frames. In the case of CDR H1, the AbM definition (dashed frame, residues 27–36) was used.\n\nFor the CDR H1, we also used AbM definition, a combination of Kabat and Chotia24 definitions, which better matches the loop in protein structure. The high resolution of the dataset and good quality of the electron density maps allowed us to build a precise atomic model of the antibody. We then used a PyIgClassify25 to analyze the structures of BCD090-M2 CDRs using a comprehensive database of antibody CDR loop conformations. The CDR H1 and H2 belonged to two large common clusters, H1-13-1 and H2-10-2 respectively. The CDR H3 was not assigned to any known structure cluster, probably due to size (18 residues) and complex structure of the loop, which has a cis-proline and four aromatic residues. It is consistent with an observation that CDR H3 loops are very diverse in structure and a few clusters of significant size are present in the database.\n\nThe structure of the BCD090-M2 dimer crystallized in space group P1 with cadmium ions is shown in Figure 5. Two cadmium ions in the unit cell are pictured as green spheres, and two symmetry-related ions from the neighbor unit cell are shown in dots. Each cadmium ion is bound by residues Asp100, Glu114, and Asp116 belonging to CDR H3, and by N-terminal glycine residue of an antibody molecule from neighbor unit cell. Thus, intermolecular interaction through cadmium ions effectively define the lattice of a crystal. This finding is consistent with a previous general observation that cadmium can induce the formation of protein crystals or improve their quality26.\n\nAsymmetric unit contains two copies of the VHH antibody and two cadmium ions which are pictured as green spheres, two symmetry-related ions from the neighbor unit cell are shown in dots. Framework regions are grey and CDRs are colored.\n\nFinally, we analyzed the structural variability of the overall antibody fold and the putative paratope by comparing the three solved structures 6EZW, 6F0D (A), and 6F0D (B). We first performed the structural alignment based on Cα atoms of the framework regions, which is shown in Figure 6A.\n\nA) the three structures, 6EZW, 6F0D chain A, and 6F0D chain B, were superimposed, the structural alignment was based on Cα atoms of framework regions. CDRs H1 and H3 showed little structural variation, while CDR H2 adopted different conformations in 6EZW and 6F0D, which are shown in blue and green color. B) Electron density map (F-obs, Phi-model, 1.0 σ) for CDR H2 (residues 51–67) of 6EZW. The quality of electron density maps allowed us to unambiguously trace the loop and place amino-acid side chains.\n\nThe overall fold of the antibody had a very low structural variability, the RMSD between framework Cα atoms in 6EZW and 6F0D was 0.19 Å. The conformational mobility of the short CDR H1 was also low, with Cα RMSD between 6EZW and 6F0D 0.28 and 0.31 Å for chains A and B, respectively. Surprisingly, the most complex CDR H3 had a rigid conformation, which was not significantly altered by binding of a cadmium ion and intermolecular interaction in the asymmetric unit of 6F0D. The Cα RMSD for CDR H3 between two structures was 0.61 (chain A) and 0.57 Å (chain B). The largest structural variation was observed in the loop H2, as seen in Figure 6A. The Cα RMSD was 1.81 and 1.77 Å for chains A and B, respectively, and the largest difference was observed in the region Arg53-Leu57. Despite the probable flexibility of this loop, in both crystals it was well-resolved in the electron density maps, as shown in Figure 6B, allowing us to unambiguously place all residues. As a result of the observed structural change, the CDRs H2 in 6F0D were attributed to a different from 6EZW distant structural clusters H2-10-6 (chain A) or H2-10-7 (chain B) by PyIgClassify25.\n\n\nConclusions\n\nIn conclusion, here were present expression and purification, functional and structural analysis of a new single-domain llama antibody against human ErbB3. We crystallized the antibody in two different forms and solved high-resolution structures, giving an insight to the organization of the putative ErbB3 binding paratope. We believe this data may facilitate further studies of mechanisms of ErbB3 inhibition by single-chain antibodies. Besides, the solved structures will contribute to datasets that are required to develop new robust computational methods for antibody modeling and design.\n\n\nData availability\n\nThe atomic coordinates and structure factors can be accessed under PDB codes 6EZW and 6F0D.\n\nDataset 1: Uncropped gel from Figure 1 and raw output data from Biacore software. DOI, 10.5256/f1000research.13612.d19041927\n\nThe plasmids and recombinant proteins used in this study are available from Igor Eliseev (corresponding author) upon request.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work is funded by the Ministry of Education and Science of the Russian Federation (contract 14.577.21.0217, unique identifier RFMEFI57716X0217) and co-funded by CJSC Biocad.\n\n\nAcknowledgements\n\nWe acknowledge technical support by the SPC facility at EMBL Hamburg. We also thank Ms. Maria V. Mitkevich for remarkable administrative support.\n\n\nReferences\n\nCitri A, Yarden Y: EGF-ERBB signalling: towards the systems level. Nat Rev Mol Cell Biol. 2006; 7(7): 505–516. PubMed Abstract | Publisher Full Text\n\nPolanovski OL, Lebedenko EN, Deyev SM: ERBB oncogene proteins as targets for monoclonal antibodies. Biochemistry (Mosc). 2012; 77(3): 227–245. PubMed Abstract | Publisher Full Text\n\nKol A, Terwisscha van Scheltinga AG, Timmer-Bosscha H, et al.: HER3, serious partner in crime: therapeutic approaches and potential biomarkers for effect of HER3-targeting. Pharmacol Ther. 2014; 143(1): 1–11. PubMed Abstract | Publisher Full Text\n\nGarrett JT, Olivares MG, Rinehart C, et al.: Transcriptional and posttranslational up-regulation of HER3 (ErbB3) compensates for inhibition of the HER2 tyrosine kinase. Proc Natl Acad Sci U S A. 2011; 108(12): 5021–5026. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChiu CG, Masoudi H, Leung S, et al.: HER-3 overexpression is prognostic of reduced breast cancer survival: a study of 4046 patients. Ann Surg. 2010; 251(6): 1107–1116. PubMed Abstract | Publisher Full Text\n\nZhang N, Chang Y, Rios A, et al.: HER3/ErbB3, an emerging cancer therapeutic target. Acta Biochim Biophys Sin (Shanghai). 2016; 48(1): 39–48. PubMed Abstract | Publisher Full Text\n\nMirschberger C, Schiller CB, Schräml M, et al.: RG7116, a therapeutic antibody that binds the inactive HER3 receptor and is optimized for immune effector activation. Cancer Res. 2013; 73(16): 5183–5194. PubMed Abstract | Publisher Full Text\n\nGarner AP, Bialucha CU, Sprague ER, et al.: An antibody that locks HER3 in the inactive conformation inhibits tumor growth driven by HER2 or neuregulin. Cancer Res. 2013; 73(19): 6024–6035. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee S, Greenlee EB, Amick JR, et al.: Inhibition of ErbB3 by a monoclonal antibody that locks the extracellular domain in an inactive configuration. Proc Natl Acad Sci U S A. 2015; 112(43): 13225–13230. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEliseev IE, Yudenko AN, Besedina NA, et al.: Thermodynamic analysis of the conformational stability of a single-domain therapeutic antibody. Tech Phys Lett. 2017; 43(12): 1088–1091, (in press).\n\nGasteiger E, Hoogland C, Gattiker A, et al.: Protein identification and analysis tools on the ExPASy server. In JM Walker, editor, The Proteomics Protocols Handbook. Humana Press, The address of the publisher. 2005; 571–607. Publisher Full Text\n\nBruker: SAINT, SADABS, XPREP. Bruker AXS Inc., Madison, Wisconsin, USA, 2008–2012.\n\nLobstein J, Emrich CA, Jeans C, et al.: SHuffle, a novel Escherichia coli protein expression strain capable of correctly folding disulfide bonded proteins in its cytoplasm. Microb Cell Fact. 2012; 11(1): 56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaufman EN, Jain RK: Effect of bivalent interaction upon apparent antibody affinity: experimental confirmation of theory using fluorescence photobleaching and implications for antibody binding assays. Cancer Res. 1992; 52(15): 4157–4167. PubMed Abstract\n\nAdams PD, Afonine PV, Bunkóczi G, et al.: PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 2): 213–221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBunkóczi G, Read RJ: Improvement of molecular-replacement models with Sculptor. Acta Crystallogr D Biol Crystallogr. 2011; 67(Pt 4): 303–312. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwarzenbacher R, Godzik A, Grzechnik SK, et al.: The importance of alignment accuracy for molecular replacement. Acta Crystallogr D Biol Crystallogr. 2004; 60(Pt 7): 1229–1236. PubMed Abstract | Publisher Full Text\n\nMcCoy AJ, Grosse-Kunstleve RW, Adams PD, et al.: Phaser crystallographic software. J Appl Crystallogr. 2007; 40(Pt 4): 658–674. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWen Y, Ouyang Z, Schoonooghe S, et al.: Structural evaluation of a nanobody targeting complement receptor Vsig4 and its cross reactivity. Immunobiology. 2017; 222(6): 807–813. PubMed Abstract | Publisher Full Text\n\nTerwilliger TC, Grosse-Kunstleve RW, Afonine PV, et al.: Iterative model building, structure refinement and density modification with the PHENIX AutoBuild wizard. Acta Crystallogr D Biol Crystallogr. 2008; 64(Pt 1): 61–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEmsley P, Lohkamp B, Scott WG, et al.: Features and development of Coot. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 4): 486–501. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAfonine PV, Grosse-Kunstleve RW, Echols N, et al.: Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallogr D Biol Crystallogr. 2012; 68(Pt 4): 352–367. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKabat EA, Te Wu T, Foeller C, et al.: Sequences of proteins of immunological interest. NIH Publication No. 91-3242, 5th edition, 1991. Reference Source\n\nAl-Lazikani B, Lesk AM, Chothia C: Standard conformations for the canonical structures of immunoglobulins. J Mol Biol. 1997; 273(4): 927–948. PubMed Abstract | Publisher Full Text\n\nAdolf-Bryfogle J, Xu Q, North B, et al.: PyIgClassify: a database of antibody CDR structural classifications. Nucleic Acids Res. 2015; 43(Database issue): D432–D438. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrakhanov S, Kreimer DI, Parkin S, et al.: Cadmium-induced crystallization of proteins: II. Crystallization of the Salmonella typhimurium histidine-binding protein in complex with L-histidine, L-arginine, or L-lysine. Protein Sci. 1998; 7(3): 600–604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEliseev IE, Yudenko AN, Vysochinskaya VV, et al.: Dataset 1 in: Crystal structures of a llama VHH antibody BCD090-M2 targeting human ErbB3 receptor. F1000Research. 2018. Data Source"
}
|
[
{
"id": "30112",
"date": "05 Mar 2018",
"name": "Shi Hu",
"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\nFor the immunoglobulin-γ (IgG) antibodies, the structure usually is assembled from two identical heavy (H)-chain and two identical light (L)-chain polypeptides is well established and highly conserved in mammals. The L chain of these immunoglobulins comprises two domains, whereas the H chain folds into four domains. Currently, most of therapeutic antibodies used in clinical development are belong to such type. One exception to this conventional mammalian IgG structure is found in sera of Camelidae. Unlike the conventional heterotetrameric antibodies these sera possess special IgG antibodies. IgG antibodies, known as heavy-chain antibodies (HCAbs), are devoid of the L chain polypeptide and are unique because they lack the first constant domain (CH1). At its N-terminal region, the H chain of the homodimeric protein contains a dedicated variable domain, referred to as VHH, which serves to associate with its cognate antigen. The VHH in an HCAb is the structural and functional equivalent of the Fab fragment (antigen-binding fragment) of conventional antibodies.\nIn this report, Igor et al expressed a recombinant single-domain antibody (VHH form llamas) targeting ErbB3 Protein crystallography study was used for determine the structure of VHH in different crystal forms. It is a well done crystallography report providing readers more data for the structure of VHH, and I only have a few minor suggestions:\n\nDo the authors have any information or data about the epitope of BCD090-M2?\n\nSingle-domain antibodies from llamas are new generation of therapeutical antibody drugs, can authors make another 3-D Alignment figure to compare BCD090-M2 with human antibodies?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3796",
"date": "04 Jul 2018",
"name": "Igor Eliseev",
"role": "Author Response",
"response": "Dear Dr. Shi,We thank you for reading and commenting on our manuscript. Let us respond to your comments: Initially, we hypothesized that the antibody BCD090-M2 binds to the extracellular domain III of ErbB3 (ECDIII). To test this hypothesis, we have conducted pairwise SPR binding experiments with intact ErbB3 ECD and another anti-ErbB3 antibody which targets ECDIII but saw simultaneous binding of two antibodies. Then we expressed and purified ECDIII (amino acids 329-532 of ErbB3), and also observed no interaction with BCD090-M2 in direct binding experiments. We are now performing additional experiments to determine the epitope of BCD090-M2 on the ErbB3 extracellular domain and elucidate the binding mechanism. The alignment of the framework regions of our llama antibody and heavy chains of human antibodies is indeed very good. However, the CDRs of our antibody does not align well with human antibodies, e.g., those anti-ErbB3 antibodies we cited in the manuscript (target ECD I, III, II and IV). It is interesting whether llama and human antibodies recognize the same molecular features, and we certainly plan to compare our llama antibody with human anti-ErbB3 antibodies after identification of its epitope."
}
]
},
{
"id": "32141",
"date": "10 Apr 2018",
"name": "Kathryn M. Ferguson",
"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 purify a VHH fragment corresponding to an camelid single-chain antibody previously selected from a Llama derived phage antibody library. They measure binding of this VHH (BCD090-M2) to the extracellular region of ErbB3 that was immobilized to a CM5 biosenor chip, using Surface plasmon resonance (SPR). An apparent KD of 1μM is reported. The authors solve the X-ray crystal structure of the VHH in two different crystal forms and observe differences in the conformation of CDR H2.\n\nThe SPR sensograms do not come to a plateau during each injection, which is most obvious at the higher concentrations, so it is not a saturation binding response for each concentration. The Response data that are plotted in Figure 2B for each concentration of VHH do not correspond to the response values 5 seconds before the end of the injection in the sensograms plotted in Figure 2A. It looks more like the values 50 seconds before end of injection. There is something wrong here. Perhaps the data in A and B are from two different experiments? No error is given on the KD value so it is not clear if these data were repeated. I also worry about use of regeneration using glycine at pH 2. It is highly likely that this will denature the immobilized ErbB3. These data are thus questionable. No information on binding for the intact antibody is referenced or provided. The VHH is clearly folded. The question is whether it really binds to ErbB3.\n\nThe structure determination is technically sound.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3797",
"date": "09 Jul 2018",
"name": "Igor Eliseev",
"role": "Author Response",
"response": "Dear Prof. Ferguson, We appreciate your comments on the SPR experiment, which appeared both insightful and very helpful for us. We first addressed the regeneration issue. The initial choice of regeneration buffer (glycine-HCl pH 2.0) was made because it efficiently eluted several anti-ErbB3 VHH fragments we were working with. We probed the receptor with a different antibody and saw no significant loss of binding after treatment with glycine buffer. However, this alternative antibody targeted a different epitope on the receptor, while the epitope of BCD090-M2 appeared very sensitive to low pH, as you suggested. We tested various regeneration conditions, and although their effect on the receptor was not as detrimental as that of glycine, we still saw some memory effects. Therefore, we finally adopted a protocol with a long dissociation phase (30 min) followed by a 10 min equilibration with running buffer, which was sufficient for complete elution of the antibody. Then we increased the length of the injection to 30 min to ensure that all binding curves reach equilibrium at the end of association phase. The measured binding response as a function of antibody concentration was fitted with simple 1:1 binding model by a nonlinear least-squares method which gave equilibrium constant KD=14 nM. We repeated the experiment twice and saw a remarkable reproducibility; the mean KD was 15±1 nM. Finally, we updated the Figures 2A and 2B to present new experimental data."
}
]
}
] | 1
|
https://f1000research.com/articles/7-57
|
https://f1000research.com/articles/6-1502/v1
|
17 Aug 17
|
{
"type": "Case Report",
"title": "Case Report: Synchronous primary malignancy including the breast and endometrium",
"authors": [
"Elham Sadat Banimostafavi",
"Sepideh Tayebi",
"Maryam Tayebi",
"Sepideh Tayebi",
"Maryam Tayebi"
],
"abstract": "Breast and endometrial cancer are the most common types of female cancers, but the incidence of both of these malignancies in a single patient is a rare event. Multiple primary malignancy has been increasingly reported over the past decade, and double primary cancer is considered as the most common type. In this study, we present a 53-year-old woman with synchronous primary malignancy of breast and endometrium. This patient had a history of breast and endometrial cancer in her family. Mammography and chest CT of the patient revealed a mass in the right breast and left supraclavicular region. However, the patient did not want to initiate treatment. Subsequently, the patient returned with a chief complaint of persistent abnormal vaginal bleeding. Abdominopelvic CT scan of the patient revealed a huge soft tissue mass in the pelvic cavity. She underwent hysterectomy, and pathology revealed endometrioid carcinoma, which had invaded the full thickness of uterine wall. Since this type of malignancy is rare and several risk factors are associated with it, it is worth being considered by clinicians when making decisions about screening or strategy for prevention.",
"keywords": [
"Breast",
"Endometrium",
"Cancer",
"Multiple Primary Malignancy"
],
"content": "Introduction\n\nBreast cancer (BC) is the most frequently diagnosed malignancy worldwide and is the first cause of cancer death in women1. The common metastatic sites of breast cancer are the lungs, bones, liver and brain. Metastases to the gynecologic and gastrointestinal tract are rare2,3; endometrial cancer (EC) is considered as the most frequent of this type.\n\nMultiple primary malignancy (MPM) has increased over the past decade. It is a term defined as spreading of the primary malignancy to two or more parts of the body distinct from each other. In addition to being distinct, these tumours must have definite featured of malignancy, and the possibility that one is the metastasis of the other must be ruled out4,5. Double primary cancers are the most common types of MPM.\n\nMultiple mechanisms such as hereditary, immune and environmental factors, e.g. chemical, viruses and chemotherapeutic regimens, are considered as the pathogenesis of MPM6. Tumours that are diagnosed simultaneously or within six months are known as synchronous; a longer interval time and the tumours are metachronous.\n\nWe present a patient with two primary malignant tumours, including BC (invasive ductal carcinoma) and EC (endometroid cell type), which can be considered as synchronous MPM.\n\n\nCase report\n\nThe following case is a 53-year-old woman who was referred to hospital from a local doctor in December 2016 with a palpable mass in her left supraclavicular region. Mammography and chest CT scan revealed the presence of a speculated mass 3.8×3.7 cm in the right breast (Figure 1), and additionally a soft tissue mass 5.8×5.1 cm in the left supraclavicular region (Figure 2). Core needle biopsy for the right breast mass was preformed, and invasive ductal carcinoma with metastatic supraclavicular lymph nodes was confirmed. Although the mass was diagnosed as BC, the patient personally refused to get any treatment. She has a positive family history of breast cancer and uterine cancer in her sister.\n\nAn irregular speculated hyperdensity mass in the right breast upper outer quadrant.\n\nA soft tissue mass in the left supraclavicular region consistent with metastatic lymph node (yellow arrow).\n\nOne month later, the patient returned with a chief complaint of persistent abnormal vaginal bleeding. She had the history of bleeding 4 years ago and it had worsened over the previous 7 months. Abdominopelvic CT scan of the patient revealed a huge soft tissue mass 14×11 cm in the pelvic cavity with right external iliac and para-aortic lymphadenopathy and dilatation of renal calyces and ureters on both sides (Figure 3).\n\nA soft tissue mass in the pelvic cavity with right external iliac and para-aortic lymphadenopathy.\n\nIn January 2017, a total abdominal hysterectomy was performed with no complication, and the pathology revealed EC (with extensive coagulative necrosis and few retained tissue), which had invaded the full thickness of the uterine wall. Pelvic wall mass resection and cervix excision revealed the invasion of the tumour, but peritoneal fluid cytology was negative for malignancy. After two days she discharged from hospital with relative improvement.\n\nWe could not follow up the patient because she moved to another city for further treatment; this is one limitation of our study. At the final follow-up, the patient was referred to the oncology department in a different hospital to initiate chemotherapy.\n\n\nDiscussion\n\nThe diagnosis of synchronous primary cancers in an individual is rare and difficult, especially in the case of finding the same type of cancer. In the present case, clinicopathological criteria was used to distinguish the two similar cancers7.\n\nThe risk of a new primary cancer in cancer survivors is 20% higher than in the general population8. In addition, it has been was shown that the risk of developing a new malignancy is 1.29 times more than those who have never been diagnosed9. The possibility of synchronous BC and EC in one person is extremely low, as reported in one study the diagnosis of EC within one year after the diagnosis of primary BC is less than 0.05%8.\n\nThe coexistence of breast and endometrial cancer reflects the fact that there are many environmental and hormonal risk factors that may predispose the patient to both BC and EC, such as genetics, hormonal, environmental or treatment-related factors, and obesity (i.e. high BMI)10,11. Some of these factors are controversial. For instance, high BMI increases the risk of BC in postmenopausal women; however, it has opposite effect on premenopausal women12,13. By contrast, high BMI increases the risk of EC in both pre and postmenopausal women14,15.\n\nThere are many other situations that are correlated with an increasing risk of EC, such as age (i.e. more common in older patients), postmenstrual period16,17, nulliparous, and a positive history of irregular menstrual cycle14. Our case had some of these risk factors, such as being postmenopausal and having a high BMI.\n\nBesides these factors, hormonal status has an important role in endometrial carcinogenesis. Lower exposure to estrogen and higher exposure to progesterone reduce the risk of EC18. The conversion of adrenal hormones into estrogen may be done by fat cells in obese women, so obesity may increase the risk of EC in this way19. Obesity, nullipara and irregular menstrual cycle may represent less progesterone exposure, so they may contribute to EC development. In addition, EC may develop in association with tamoxifen treatment for BC, particularly in the case of long-term administration and high cumulative doses of tamoxifen20–22. The patient in our study did not have any risk factors related to treatment because she did not start BC radio or chemotherapy before presentation of EC symptoms; therefore, we cannot consider the effects of tamoxifen usage in BC as a risk factor of EC in this patient.\n\nGenetic and/or epigenetic changes and other plausible molecular mechanisms might be important in patients with synchronous double cancers23. The present case had a family history of breast and uterine cancer, so heredity could be counted as one of the strongest risk factors for this patient.\n\nIn addition to many similar environmental and hormonal risk factors, the same embryological origin of the endometrium and breast can constitute as an additional factor6,24. MPMs can generally be categorized into three major groups depending on the main etiologic factor. The first group are treatment-related neoplasms, the second group are syndromic cases (like Cowden syndrome), and the third group are neoplasms that may share common etiologic factors, such as genetic predisposition or the same environmental factors25. According to this classification, our patient can be categorized in the third group.\n\nTo conclude, finding a patient with simultaneous presentation of endometrial and breast cancer is rare; however both of these primary malignancies are considered as the most common cancers in females. Several associated risk factors to this event have been described above. In our case, a high BMI, postmenopausal status and hereditary are probably the most relevant risk factors. Hence, all these factors should be taken into account by clinicians when making a decision concerning screening or strategy for prevention.\n\n\nConsent\n\nWritten informed consent for the publication of the patient’s clinical details and images was obtained from the patient.",
"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\nFast Stats: Most frequent cancers: both sexes [Internet]. WHO. 2012; [cited 2017]. Reference Source\n\nArslan D, Tural D, Tatlı AM, et al.: Isolated uterine metastasis of invasive ductal carcinoma. Case Rep Oncol Med. 2013; 2013: 3, 793418. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbid A, Moffa C, Monga DK: Breast cancer metastasis to the GI tract may mimic primary gastric cancer. J Clin Oncol. 2013; 31(7): e106–e7. PubMed Abstract | Publisher Full Text\n\nAydiner A, Karadeniz A, Uygun K, et al.: Multiple primary neoplasms at a single institution: differences between synchronous and metachronous neoplasms. Am J Clin Oncol. 2000; 23(4): 364–70. PubMed Abstract\n\nDerwinger K, Gustavsson B: A study of aspects on gender and prognosis in synchronous colorectal cancer. Clin Med Insights Oncol. 2011; 5: 259–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakellakis M, Peroukides S, Iconomou G, et al.: Multiple primary malignancies: a report of two cases. Chin J Cancer Res. 2014; 26(2): 215–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYeh CC, Wang PH, Lai CR, et al.: Synchronous breast invasive ductal carcinoma and endometrial endometrioid adenocarcinoma: case report. J Obstet Gynaecol Res. 2011; 37(9): 1246–9. PubMed Abstract | Publisher Full Text\n\nTravis LB, Hill D, Dores GM, et al.: Cumulative absolute breast cancer risk for young women treated for Hodgkin lymphoma. J Natl Cancer Inst. 2005; 97(19): 1428–37. PubMed Abstract | Publisher Full Text\n\nSchoenberg BS: Multiple primary malignant neoplasms. The Connecticut experience, 1935–1964. Recent Results Cancer Res. 1977; 58: 1–173. PubMed Abstract | Publisher Full Text\n\nMellemkjaer L, Friis S, Olsen JH, et al.: Risk of second cancer among women with breast cancer. Int J Cancer. 2006; 118(9): 2285–92. PubMed Abstract | Publisher Full Text\n\nAmir E, Freedman OC, Seruga B, et al.: Assessing women at high risk of breast cancer: a review of risk assessment models. J Natl Cancer Inst. 2010; 102(10): 680–91. PubMed Abstract | Publisher Full Text\n\nIatrakis G, Zervoudis S, Saviolakis A, et al.: Women younger than 50 years with endometrial cancer. Eur J Gynaecol Oncol. 2006; 27(4): 399–400. PubMed Abstract\n\nBallard-Barbash R, Swanson CA: Body weight: estimation of risk for breast and endometrial cancers. Am J Clin Nutr. 1996; 63(3 Suppl): 437S–41S. PubMed Abstract\n\nSoliman PT, Oh JC, Schmeler KM, et al.: Risk factors for young premenopausal women with endometrial cancer. Obstet Gynecol. 2005; 105(3): 575–80. PubMed Abstract | Publisher Full Text\n\nBianchini F, Kaaks R, Vainio H: Overweight, obesity, and cancer risk. Lancet Oncol. 2002; 3(9): 565–74. PubMed Abstract | Publisher Full Text\n\nSaadat M, Truong PT, Kader HA, et al.: Outcomes in patients with primary breast cancer and a subsequent diagnosis of endometrial cancer : comparison of cohorts treated with and without tamoxifen. Cancer. 2007; 110(1): 31–7. PubMed Abstract | Publisher Full Text\n\nBurke TW, Fowler WC Jr, Morrow CP: Clinical aspects of risk in women with endometrial carcinoma. J Cell Biochem Suppl. 1995; 59(S23): 131–6. PubMed Abstract | Publisher Full Text\n\nDossus L, Allen N, Kaaks R, et al.: Reproductive risk factors and endometrial cancer: the European Prospective Investigation into Cancer and Nutrition. Int J Cancer. 2010; 127(2): 442–51. PubMed Abstract | Publisher Full Text\n\nSiiteri PK: Adipose tissue as a source of hormones. Am J Clin Nutr. 1987; 45(1 Suppl): 277–82. PubMed Abstract\n\nvan Leeuwen FE, Benraadt J, Coebergh JW, et al.: Risk of endometrial cancer after tamoxifen treatment of breast cancer. Lancet. 1994; 343(8895): 448–52. PubMed Abstract | Publisher Full Text\n\nBergman L, Beelen ML, Gallee MP, et al.: Risk and prognosis of endometrial cancer after tamoxifen for breast cancer. Comprehensive Cancer Centres' ALERT Group. Assessment of Liver and Endometrial cancer Risk following Tamoxifen. Lancet. 2000; 356(9233): 881–7. PubMed Abstract | Publisher Full Text\n\nWang PH, Chao HT: A reconsideration of tamoxifen use for breast cancer. Taiwan J Obstet Gynecol. 2007; 46(2): 93–5. PubMed Abstract | Publisher Full Text\n\nFletcher O, Houlston RS: Architecture of inherited susceptibility to common cancer. Nat Rev Cancer. 2010; 10(5): 353–61. PubMed Abstract | Publisher Full Text\n\nHemminki K, Aaltonen L, Li X: Subsequent primary malignancies after endometrial carcinoma and ovarian carcinoma. Cancer. 2003; 97(10): 2432–9. PubMed Abstract | Publisher Full Text\n\nTakalkar U, Asegaonkar BN, Kodlikeri P, et al.: An elderly woman with triple primary metachronous malignancy: A case report and review of literature. Int J Surg Case Rep. 2013; 4(7): 593–6. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "26730",
"date": "12 Oct 2017",
"name": "Weibo Yu",
"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\nBreast cancer and endometrial cancer are two most frequent hormone-related cancers among women. The authors presented a rare case with synchronous primary malignancy of breast and endometrium. This is a well-written case report. It clearly described a particular individual’s history with a disease presentation and progress.\n\nOne limitation is the patient lost to follow up.\n\nIt is advised to add pathology images demonstrating their findings.\n\nPatients with germline mutations in BRCA1 and BRCA2 genes are highly susceptible to breast cancer. These mutations are also related to an increased risk of developing endometrial cancer. Do author consider to perform an investigation in genetic level? It would be adequate to add more discussion in this point.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "25974",
"date": "16 Nov 2017",
"name": "Minas Sakellakis",
"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 needs to be significantly improved before it gets accepted for publication. Here are my suggestions to improve this article:\n\nCOMMENTS ON INTRODUCTION:\n“Metastases to the gynecologic and gastrointestinal tract are rare2,3; endometrial cancer (EC) is considered as the most frequent of this type.” Please correct: Endometrial cancer is not a metastasis.\n\n“Multiple primary malignancy (MPM) has increased over the past decade. It is a term defined as spreading of the primary malignancy to two or more parts of the body distinct from each other. In addition to being distinct, these tumours must have definite featured of malignancy, and the possibility that one is the metastasis of the other must be ruled out” Please correct: Multiple primary malignancy is not spreading of the primary malignancy to two or more parts. The term refers to separate cancers.\n\n“Multiple mechanisms such as hereditary, immune and environmental factors, e.g. chemical, viruses and chemotherapeutic regimens, are considered as the pathogenesis of MPM” Many times (if not most times) MPM are the result of pure chance. For instance if a disease A has a prevalence of e.g. 3% and disease B has a prevalence of e.g. 2%, then a small number of patients will have both diseases only as a matter of chance. The same thing implies here for BC and EC.\n\nCOMMENTS ON CASE PRESENTATION:\nAdditional information should be given regarding the tumor stage. There is no information regarding the lymph node status in the right axilla. Does the text imply that the disease was spread directly to the left supra-clavicular lymph nodes. Did the authors rule out a second tumor in the left breast? Did they detect metastatic disease elsewhere?\n\nThe authors don't mention the status of estrogen and progesterone receptors, nor do they report the HER2/NEU status.\n\nPathology images should be added.\n\nCOMMENTS ON DISCUSSION:\nLanguage editing is strongly adviced before the authors submit a revision of this manuscript. For instance the phrase:\"The diagnosis of synchronous primary cancers in an individual is rare and difficult, especially in the case of finding the same type of cancer. In the present case, clinicopathological criteria was used to distinguish the two similar cancer\" needs to rephrased completely.\n\nDiscuss that the coexistence of BC and EC in this patient might be only a coincidence.\n\nInformation regarding menopausal status and BMI should be given in the case presentation section.\n\nWhat are the treatment implications?\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/6-1502
|
https://f1000research.com/articles/5-1384/v1
|
15 Jun 16
|
{
"type": "Research Article",
"title": "An end to end workflow for differential gene expression using Affymetrix microarrays",
"authors": [
"Bernd Klaus"
],
"abstract": "In this article, we walk through an end–to–end Affymetrix microarray differential expression workflow using Bioconductor packages. This workflow is directly applicable to current “Gene” type arrays, e.g. the HuGene or MoGene arrays but can easily adapted to similar platforms. The data re–analyzed is a typical clinical microarray data set that compares inflammed and non–inflammed colon tissue in two disease subtypes. We will start from the raw data CEL files, show how to import them into a Bioconductor ExpressionSet, perform quality control and normalization and finally differential gene expression (DE) analysis, followed by some enrichment analysis. As experimental designs can be complex, a self contained introduction to linear models is also part of the workflow.",
"keywords": [
"gene expression",
"Affimetrix",
"microarrays",
"workflow"
],
"content": "Introduction\n\nIn this article we introduce a complete workflow for a typical (Affymetrix) microarray analysis. Data import, preprocessing, differential expression and enrichment analysis are discussed. We also introduce some necessary mathematical background on linear models along the way.\n\nThe data set used1 is from a paper studying the differences between patients suffering from Ulcerative colitis (UC) or Crohn’s disease (CD). This is a typical clinical data set consisting of 14 UC and 15 CD patients from which inflamed and non–inflamed colonic mucosa tissue was obtained via a biopsy. Our aim is to analyze differential expression (DE) between the tissues in the two diseases.\n\n\nRequired packages and other preparations\n\n\n\n\nDownload the raw data from from ArrayExpress\n\nThe first step of the analysis is to download the raw data CEL files. These files are produced by the array scanner software and contain the probe intensities measured. The data have been deposited at ArrayExpress and have the accession code E-MTAB-2967.\n\nEach ArrayExpress data set has a landing page summarizing the data set, and we use the ArrayExpress Bioconductor package to obtain the ftp links to the raw data files (Data from Palmieri et. al. on ArrayEpress).\n\n\nInformation stored in ArrayExpress\n\nEach dataset at ArrayExpress is stored according to the MAGE–TAB (MicroArray Gene Expression Tabular) specifications as a collection of tables bundled with the raw data. The MAGE–TAB format specifies up to five different types of files, namely the Investigation Description Format (IDF), the Array Design Format (ADF), the Sample and Data Relationship Format (SDRF), the raw data files and the processed data files.\n\nFor use, the IDF and the SDRF file are important. The IDF file contains top level information about the experiment including title, description, submitter contact details and protocols. The SDRF file contains essential information on the experimental samples, e.g. the experimental group(s) they belong to.\n\n\nDownload the raw data and the annotation data\n\nWith the code below, we download the raw data from ArrayExpress2. It is saved in the directory raw_data_dir which defaults to the subdirectory rawDataMAWorkflow of the current working directory. The names of the downloaded files are returned as a list.\n\n\n\nWe now import the SDRF file directly from ArrayExpress in order to obtain the sample annotation.\n\nThe raw data consists of one CEL file per sample (see below) and we use the CEL file names as row names for the imported data. These names are given in a column named Array.Data.File in the SDRF table. We turn the SDRF table into an AnnotatedDataFrame from the Biobase package that we will need later to create an ExpressionSet for our data3.\n\n\n\nBefore we move on to the actual raw data import, we will briefly introduce the ExpressionSet class contained in the Biobase package. It is commonly used to store microarray data in Bioconductor.\n\nGenomic data can be very complex, usually consisting of a number of different bits and pieces, e.g. information on the experimental samples, annotation of genomic features measured as well as the experimental data itself In Bioconductor the approach is taken that these pieces should be stored in a single structure to easily manage the data.\n\nThe package Biobase contains standardized data structures to represent genomic data. The ExpressionSet class is designed to combine several different sources of information (i.e. as contained in the various MAGE–TAB files) into a single convenient structure. An ExpressionSet can be manipulated (e.g., subsetted, copied), and is the input to or output of many Bioconductor functions.\n\nThe data in an ExpressionSet consist of\n\n• assayData: Expression data from microarray experiments.\n\n• metaData: A description of the samples in the experiment (phenoData), metadata about the features on the chip or technology used for the experiment (featureData), and further annotations for the features, for example gene annotations from biomedical databases (annotation).\n\n• experimentData: A flexible structure to describe the experiment.\n\nThe ExpressionSet class coordinates all of these data, so that one does not have to worry about the details. However, some constrains have to be met. In particular, the rownames of the phenoData (which holds the content of the SDRF file) have to match the column names of the assay data (as they represent the sample identifiers), while the row names of the expression data have to match the row names of the featureData (as they represent the feature identifiers). This is illustrated in the figure.\n\nYou can use the functions pData and fData to extract the sample and feature annotation respectively from an ExpressionSet. The function exprs will return the expression data itself as a matrix.\n\nThe analysis of Affymetrix arrays starts with CEL files. These are the result of the processing of the raw image files using the Affymetrix software and contain estimated probe intensity values. Each CEL file additionally contains some metadata, such as a chip identifier.\n\nThe function read.celfiles from the oligo4 can be used to import the files. The package automatically uses pd.hugene.1.0.st.v1 as the chip annotation package as the chip–type is also stored in the .CEL files.\n\nWe specify our AnnotatedDataFrame created earlier as phenoData. Thus, We have to be sure that we import the CEL files in the order that corresponds to the SDRF table — to enforce this, we use the column Array.Data.File of the SDRF table as the filenames argument.\n\nFinally, we check whether the object created is valid. (e.g. sample names match between the different tables).\n\nWe collect the information about the CEL files and import and them into the variable raw_data:\n\nWe now inspect the raw data a bit and retain only those columns that are related to the experimental factors of interest (identifiers of the individuals, disease of the individual and the mucosa type).\n\n\n\n\n\n\n\n\n\n\n\nThe first step after the intial data import is the quality control of the data. Here we check for outliers and try to see whether the data clusters as expected, e.g. by the experimental conditions. We use the identifiers of the individuals as plotting symbols.\n\n\n\n\n\nThe PCA (performed on the log–intensity scale) plot of the raw data shows that the first principal component differentiates between the diseases. However, the intensity boxplots show that the intensity distributions of the individual arrays are quite different, indicating the need of an appropriate normalization, which we will discuss next.\n\nA wide range of quality control plots can be created using the package arrayQualityMetrics5. The package produces an html report, containing the quality control plots together with a description of their aims and an identification of possible outliers. We don’t discuss this tool in detail here, but the code below can be used to create a report for our raw data.\n\n\n\n\nBackground adjustment, calibration, summarization and annotation\n\nAfter the initial import and quality assessment, the next step in processing of microarray data is background adjustment. This is essential because a part of the measured probe intensities are due to non-specific hybridization and the noise in the optical detection system. Therefore, observed intensities need to be adjusted to give accurate measurements of specific hybridization.\n\nWithout proper normalization across arrays, it is impossible to compare measurements from different array hybridizations due to many obscuring sources of variation. These include different efficiencies of reverse transcription, labeling or hybridization reactions, physical problems with the arrays, reagent batch effects, and laboratory conditions.\n\nAfter normalization, summarization is needed because on the Affymetrix platform transcripts are represented by multiple probes. For each gene, the background adjusted and normalized intensities need to be summarized into one quantity that estimates an amount proportional to the amount of RNA transcript.\n\nAfter the summarization step, the summarized data can be annotated with various information, e.g. gene symbols and EMSEMBL gene identifiers. There is an annotation database available from Bioconductor for our platform, namely the package hugene10sttranscriptcluster.db.\n\nYou can view its content like this\n\n\n\nAdditional information is available from the reference manual of the package. Essentially, the package provides a mapping from the transcript cluster identifiers to the various annotation data.\n\nTraditionally, Affymetrix arrays (the so–called 3’ IVT arrays) were probeset based: a certain fixed group of probes were part of a probeset which represented a certain gene or transcript (note however, that a gene can be represented by multiple probesets).\n\nThe more recent “Gene” and “Exon” Affymetrix arrays are exon based and hence there are two levels of summarization. The exon level summarization leads to “probeset” summary. However, these probesets are not the same as the probesets of the previous chips, which usually represented a gene/transcript. Furthermore, there are also no longer designated match/mismatch probes present on “Gene” type chips.\n\nFor the newer Affymetrix chips a gene/transcript level summary is given by “transcriptct clusters”. Hence the appropriate annotation package is called hugene10sttranscriptcluster.db.\n\nTo complicate things even a bit more, note that the “Gene” arrays were created as affordable versions of the “Exon” arrays by taking the “good” probes from the Exon array. So the notion of a probeset is based on the original construction of the probesets on the Exon array, which contains usually at least four probes.\n\nBut since Affymetrix selected only a the subset of “good” probes for the Gene arrays, a lot of the probesets on the “Gene” arrays are made up of three or fewer probes. Thus, a summarization on the probeset/exon level is not recommended for “Gene” arrays but nonetheless possible by using the hugene10stprobeset.db annotation package.\n\nThe package oligo allows us to perform background correction, normalization and summarization in one single step using a deconvolution method for background correction, quantile normalization and the RMA (robust multichip average) algorithm for summarization.\n\nThis series of steps as a whole is commonly referred to as RMA algorithm, although strictly speaking RMA is merely a summarization method6–8.\n\n\n\n\n\nThe parameter target defines the degree of summarization, the default option of which is “core”, using transcript clusters containing “safely” annotated genes. Other options for target include “extended” and “full”. For summarization on the exon level (not recommended for Gene arrays), one can use “probeset” as the target option.\n\nAlthough other methods for background correction and normalization exist, RMA is usually a good default choice. RMA shares information across arrays and uses the versatile quantile normalization method that will make the array intensity distributions match. However, it is preferable to apply it only after outliers have been removed. The quantile normalization algorithm used by RMA works by replacing values by the average of identically ranked (with a single chip) values across arrays. A more detailed description can be found on the Wikipedia page about it.\n\nAn alternative to quantile normalization is the vsn algorithm, that performs background correction and normalization by robustly shifting and scaling log–scale intensity values within arrays9. This is less “severe” than quantile normalization.\n\nA generic model for the value of the intensity Y of a single probe on a microarray is given by\n\nWe now produce a clustering and another PCA plot using the calibrated data. In order to display a heatmap of the sample–to–sample distances, we first compute the distances using the dist function. We need to transpose the expression values since the function computes the distances between the rows (i.e. genes in our case) by default. The default distance is the Euclidean one. However this can be changed and we choose the manhatten distance here (it uses absolute instead of squared distances). We set the diagonal of the distance matrix to NA in order to increase the contrast of the color coding. Those diagonal entries do not contain information since the distance of a sample to itself is always equal to zero.\n\n\n\n\n\nThe second PC roughly separates Crohn’s disease from ulcerative colitis, while the first separates the tissues. This is what we expect: the samples cluster by their experimental conditions. On the heatmap plot we also see that the samples do not cluster strongly by tissue, confirming the impression from the PCA plot that the separation between the tissues is not perfect. The stripe in the heatmap might correspond to outlier that could potentially remove. The arrayQualityMetrics package produces reports that compute several metrics that can be used for outlier removal.\n\nWe now filter out lowly expressed genes. Microarray data commonly show a large number of probes in the background intensity range. They also do not change much across arrays. Hence they combine a low variance with a low intensity. Thus, they could end up being detected as differentially expressed although they are barely above the “detection” limit and are not very informative in general. We will perform a “soft” intensity based filtering here, since this is recommended by limma’s10,11 user guide (a package we will use below for the differential expression analysis). However, note that a variance based filter might exclude a similar set of probes in practice. In the histogram of the gene–wise medians, we can clearly see an enrichment of low medians on the left hand side. These represent the genes we want to filter.\n\nIn order to infer a cutoff from the data, we inspect the histogram of the median–intensities. We visually fit a central normal distribution given by 0.5 · N(5.1, 1.18) to the probe–wise medians, which represents their typical behavior in the data set at hand.\n\nThen we use the 5% quantile of this distribution as a threshold, We keep only those genes that show an expression higher than the threshold in at least as many arrays as in the smallest experimental group.\n\n\n\n\n\nIn our case this would be 14.\n\n\n\n\n\n\n\n\n\nBefore we continue with the linear models for microarrays and differential expression we describe how to add “feature Data”, i.e. annotation information to the transcript cluster identifiers stored in the featureData of our ExpressionSet. We use the function select from AnnotationDbi to query the gene symbols and associated short descriptions for the transcript clusters. For each cluster, we add the gene symbol and a short description of the gene the cluster represents.\n\n\n\nMany transcript–cluster identifiers will map to multiple gene symbols. We compute a summary table in the code below to see how many there are.\n\n\n\n\n\n\n\n\n\nWe have over 2000 transcript–clusters that map to multiple gene symbols. It is difficult to decide which mapping is “correct”. Therefore, we exclude these transcript–clusters. Additionally, we also exclude transcript–clusters that do not map to gene symbols.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nAlternatively, one can re–map the probes of the array to a current annotation, a workflow to do this for Illumina arrays is given in 12. Essentially, the individual probe sequences are re–aligned to an in–silico “exome” that consists of all annotated transcript exons.\n\nIn any case, the package pdInfoBuilder can be used to build custom annotation packages for use with oligo. In order to do this, PGF/CLF files (called “library files” on the Affymetrix website) as well as the probeset annotations are required. The probesets typically represent a small stretches of the genome (such as a single exon) and multiple probesets are then used to form a transcript cluster.\n\nThe CLF file contains information about the location of individual probes on the array. The PGF file then contains the individual probe sequences and shows the probeset they belong to. Finally, The probeset annotation .csv then contains information about which probesets are used in which transcript cluster. Commonly, multiple probesets are used in one transcript cluster and some probesets are contained in multiple transcript clusters.\n\n\nA short overview of linear models\n\nI am afraid this section is rather technical. However general experience shows that most questions on the Bioconductor support site about packages using using linear models like limma10, DESeq213 and edgeR14 are actually not so much about the packages themselves but rather about the underlying linear models. It might also be helpful to learn a bit of linear algebra to understand the concepts better. The Khan Academy offers nice (and free) online courses. Mike Love’s and Michael Irizzary’s genomics class is also a very good resource, especially its section on interactions and contrasts.\n\nIn regression models we use one variable to explain or predict the other. It is customary to plot the predictor variable on the x–axis and the predicted variable on the y–axis. The predictor is also called the independent variable, the explanatory variable, the covariate, or simply x. The predicted variable is called the dependent variable, or simply y.\n\nIn a regression problem the data are pairs (xi, yi) for i = 1, . . . , n. For each i, yi is a random variable whose distribution depends on xi. We write\n\nThe above expresses yi as a systematic or explainable part g(xi) and an unexplained part εi. Or more informally: response = signal + noise. g is called the regression function. Once we have an estimate g^ of g, we can compute ri := yi − g(xi). The ri’s are called residuals. The εi’s themselves are called errors.\n\nResiduals are used to evaluate and assess the fit of models for g. Usually one makes distributional assumption about them, e.g. that they are independent and identically normally distributed with identical variance σ2 and mean zero:\n\nLinear regression is a special case of the general regression model. Here, we combine the predictors linearly to produce a prediction. If we have only single predictor x, the simple linear regression model is:\n\n\n\nWith X being the so called design matrix:\n\nTo get an idea of what design matrices look like, we consider several examples. It is important to know some fundamentals about design matrices in order to be able to correctly transfer a design of a particular study to an appropriate linear model.\n\nWe will use the base R functions:\n\n• formula\n\n• model.matrix\n\n\n\n\n\nThis design is called treatment contrast parameterization for an obvious reason: the first column of the design matrix represents a “base level”, i.e the mean β0 for group one and the second column, corresponding to β1, represents the difference between the group means since all group two samples have means represented by β0 + β1. As β0 is the mean of group 1, β1 corresponds to the difference of the means of group two and group one and thus shows the effect of a “treatment”.\n\nHowever, this design is not orthogonal, i.e. the columns of the design matrix are not independent. We can construct an equivalent orthogonal design as follows:\n\nHere, we loose the nice direct interpretability of the coefficients. Now β1 is simply the mean of the second group. We will discuss the extraction of interesting contrasts (i.e. linear combinations of coefficients) from a model like this below.\n\nWe explicitly excluded the intercept by specifying it as zero. Commonly it makes sense to include an intercept in the model, especially in more complex models. We can specify a more complex design pretty easily: if we have two independent factors, the base mean now corresponds to the first levels of the two factors.\n\n\n\n\n\nThe “drop one” strategy is the default method for creating regression coefficients from factors in R. If a factor has d levels, adding it to the model will give you d − 1 regression coefficients corresponding to d − 1 columns in a design matrix. Apart from excluding the intercept, you can also use the I function to treat a covariate as it is without using the formula syntax. The code below includes z2 as a covariate.\n\n\n\n\n\n\n\n\n\n\n\n\n\nWhatever your model matrix looks like, you should make sure that it is non–singular. Singularity means that the measured variables are linearly dependent and leads to regression coefficients that are not uniquely defined. In linear algebra terms, we say that the matrix does not have full rank, which for design matrices means that the actual dimension of the space spanned by the column vectors is in fact lower than the apparent one. This leads to a redundancy in the model matrix, since some columns can be represent by linear combinations of other columns.\n\nFor design matrices, which contain factors, this happens if two conditions are confounded, e.g. in one experimental group there are only females and in the other group there are only males. Then the effect of sex and experimental group cannot be disentangled.\n\nLet’s look at an example. We set three factors, of which the third one is nested with the first two. We can check the singularity of the model matrix by computing its so called singular value decomposition and check it’s minimal singular value. If this is zero, the matrix is singular. As we can see, this is indeed the case here.\n\n\n\n\n\n\n\n\n\nWe have one column in the design matrix that can be represented by a linear combination of the other columns, thus the column space has actually a lower dimension than the apparent one. For example, we can represent column 5 (“zm”) by a linear combination of the first two columns “intercept” and “x2”:\n\n\n\n\n\n\n\n\n\nI.e., in mathematical notation this means\n\nThus, the corresponding regression coefficients are not uniquely determined and the model does not make much sense. Therefore, the non–singularity of the model matrix should always be checked beforehand.\n\nIn differential expression analysis, our most important covariates will be factors that differentiate between two or more experimental groups, e.g. the covariate Xp = (xp1, . . . , xpn) is either zero or one depending on which group the sample belongs to.\n\nWe will illustrate this concept using a small data set called toycars from the DAAG package. The data set toycars gives the distance traveled by one of three different toy cars on a smooth surface, starting from rest at the top of a 16 inch long ramp tilted at varying angles. We have the variables:\n\n• angle: Angle\n\n• distance: Traveled distance\n\n• car: Car number (1, 2 or 3)\n\nWe transform car into a factor so that R performs the necessary parameterization of the contrasts automatically.\n\n\n\nBy looking at the box plots of distance by car, we can clearly see differences between the three types of cars. We can now fit a linear model with distance as the dependent variable and car and angle as the predictors. As we can see from the linear model output, the treatment contrast parameterization was used, with car 1 being the base level.\n\n\n\n\n\nThe estimated coefficients now give us the difference between the distance traveled between car 1 and car 2 (0.11) and car 1 and car 3 (-0.08), and the associated t–tests of these coefficients. However we cannot see a test of car 2 vs. car 3. This contrast test would correspond to testing the difference between the car 2 and car 3 regression coefficients.\n\nThus, contrasts of interest to us may not readily correspond to coefficients in a fitted linear model. However, one can easily test general linear hypotheses of the coefficients of the form:\n\nWhere C is the contrast matrix containing the between group differences of interest, q is the total number of comparisons to be performed and α contains the difference to be tested, this is usually a vector of zeros. If one tests multiple coefficients at once (e.g. β1 = 0 and β2 = 0 ) the corresponding test statistic is F–distributed. If one just tests linear combinations of coefficients, e.g. β1 − β2 = 0, β1 − β2 − 2β3 = 0 or something similar the test statistic has t–distribution. The function summary for lm reports the results of β1 = β2 = . . . = βp = 0 (F–test) and the t–tests of βj = 0 each of the coefficients.\n\nNote that the model does not actually have to be refitted in order to test the contrasts. This makes contrast matrix based testing more efficient and convenient than reformulating the model using a new paramerization of the factors to obtain the desired tests. We can use the function glht from the multcomp package to test these general linear hypotheses15. Let us assume we want all pairwise comparisons between the cars, this can be achieved by defining a specific contrast matrix for our current model as given below.\n\nThere are three such comparisons and we can print the results by using the summary function. As we can see the estimates for the contrasts already contained in the original model agree with the obtained results from contrast fit.\n\n\n\n\n\n\n\n\n\nNote that the term “contrast” is used in the context of (re)parameterization of the original model (as in “treatment contrasts”) and in the testing of linear hypotheses about the model coefficients. This can lead to some confusion, however usually it should be clear from the context whether a reparameterization or test of linear hypotheses is intended.\n\nWe can also fit a linear model without an intercept to the toycars data set. Now, the coefficients derived from the “car” factor represent car–wise means. Thus, the contrasts we have to form change, however, the results for the group comparisons do not.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nLinear models for microarrays\n\nWe now apply linear models to microarrays. Specifically, we discuss how to use the limma for differential expression analysis. The package is designed to analyze complex experiments involving comparisons between many experimental groups simultaneously while remaining reasonably easy to use for simple experiments. The main idea is to fit a linear model to the expression data for each gene. Empirical Bayes and other shrinkage methods are used to borrow information across genes for the residual variance estimation leading to a “moderated” t–statistics, and stabilizing the analysis for experiments with just a small number of arrays11.\n\nIn the following, we use appropriate design and contrast matrices for our linear models and fit a linear model to each gene separately.\n\nThe original paper is interested in changes in transcription that occur between inflamed and adjacent non–inflamed mucosal areas of the colon. This is studied in both inflammatory bowel disease types.\n\nSince we have two arrays per individual, the first factor we need is a blocking factor for the individuals that will absorb differences between them. Then we create a factors that give us the grouping for the diseases and the tissue types. We furthermore simplify the names of the diseases to UC and DC, respectively. Then, we create two design matrices, one for each of the two diseases as we will analyze them separately in order to follow the analysis strategy of the original paper closely (one could also fit a joint model to the complete data set, however, the two diseases might behave very differently so that a joint fit might not be appropriate).\n\n\n\nWe can inspect the design matrices and test their rank.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nWe now fit the linear models and define appropriate contrasts to test hypotheses of interest. We want to compare the inflamed to the the non–inflamed tissue. Thus, we create a contrast matrix consisting of one row. limma ’s function makeContrasts creates this matrix from a synbolic description of the contrast of interest. We can fit the linear model, compute the moderated t–statistics by calling the eBayes function and finally extract the number of differentially expressed genes while controlling the FDR by requiring BH–corrected p–value below a certain threshold.\n\n\n\nResults can be extracted by use of the topTable function. We extract the comparisons for both Crohn’s disease as well as ulcerative colitis and sort the results by their absolute t–statistics. As a diagnostic check, we also plot the p–value histogram: We expect a uniform distribution for the p–values that correspond to true null hypotheses, while the a peak near zero shows a enrichment for low p–values corresponding to differentially expressed (DE) genes. A p–value less than 0.001 was used in the original paper as a significance cutoff leading to 298 (CD) and 520 (UC) DE–genes for the two diseases.\n\nWe call around 500/1000 genes in the two conditions at the same cutoff, this higher number of DE genes identified is probably due to the increased power from the blocking according to the individuals and the moderated variance estimation that limma performs.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nWe now compare our list of differentially expressed genes to the results obtained in the paper. The paper results can be downloaded as excel files from http://links.lww.com/IBD/A795. We save it in an .xlsx file named palmieri_DE_res.xlsx. The paper results are given as identified differentially expressed genes with a p–value less than 0.001, which corresponds to an FDR of 0.05 in Crohn’s disease and 0.02 in ulcerative colitis. There are four tables in total giving the list of up and downregulated genes in CD and UC respectively. In the code below, we extract the gene symbols from the excel table and then compare them to the differentially expressed genes we identify at a p–value of 0.001.\n\n\n\n\n\n\n\n\n\nWe see that we get a moderate overlap of 0.678 for CD and 0.692 for UC. Note that is recommended to always to choose an FDR cutoff instead of a p–value cutoff, since this way you control an explicitly defined error rate and the results are easier to interpret and to compare. In what follows, we choose an FDR cutoff of 10%.\n\n\nGene ontology (GO) based enrichment analysis\n\nWe can now try characterize the identified differentially expressed genes a bit better by performing an GO enrichment analysis. Essentially the gene ontology (http://www.geneontology.org/) is a hierarchically organized collection of functional gene sets16–18.\n\nThe function genefinder from the genefilter19 will be used to find a background set of genes that are similar in expression to the differentially expressed genes. We then check whether the background has roughly the same distribution of average expression strength as the foreground.\n\nWe do this in order not to select a biased background since the gene set testing is performed by a simple Fisher test on a 2×2 table. Note that this approach is very similar to commonly used web tools like GOrilla20. Here we focus on the CD subset of the data.\n\nFor every differentially expressed gene, we try to find genes with similar expression.\n\n\n\n\n\n\n\n\n\n\n\nWe can see that the matching returned a sensible result and can now perform the actual testing. For this purpose we use the topGO which implements a nice interface to Fisher testing and also has additional algorithms taking the GO structure into account, by e.g. only reporting the most specific gene set in the hierarchy21.\n\nThe GO has three top ontologies, cellular component (CC), biological processes (BP), and molecular function (MF). For illustrative purposes we limit ourselves to the BP category here.\n\nWe first create a factor all_genes which indicates for every gene in our background/universe, whether it is differentially expressed or not.\n\n\n\nWe now initialize the topGO data set, using the GO annotations contained in the annotation data base for the chip we are using. The nodeSize parameter specifies a minimum size of a GO category we want to use: i.e. here categories with less than 10 genes are not included in the testing.\n\n\n\nNow the tests can be run. topGO offers a wide range of options, for details see the paper or the package vignette.\n\nWe run two common tests: an ordinary Fisher test for every GO category, and the “elim” algorithm, which tries to incorporate the hierarchical structure of the GO and tries “decorrelate” it in order to report the most specific significant term in the hierarchy.\n\nThe algorithm starts processing the nodes/GO categories from the highest (bottommost) level and then iteratively moves to nodes from a lower level. If a node is scored as significant, all of its genes are marked as removed in all ancestor nodes. This way, the “elim” algorithm aims at finding the most specific node for every gene.\n\nThe tests uses a 0.01 p–value cutoff by default.\n\n\n\nWe can now inspect the results. We look at the top 100 GO categories according to the “Fisher elim” algorithm. The function GenTable produces a table of significant GO categories, the function printGenes gives significant genes annotated to them.\n\n\n\n\n\nA graph of the results can also be produced. Here we visualize the three most significant nodes according to the Fisher elim algorithm in the context of the GO hierarchy.\n\n\n\nWe can see that indeed GO categories related to inflammation, signalling and immune response show up as significant. Gene set enrichment analysis has been a field of very extensive research in bioinformatics. For additional approaches see the topGO vignette and the references therein and also in the GeneSetEnrichment view.\n\n\nA pathway enrichment analysis using reactome\n\nThe package ReactomePA offers the possibility to test enrichment of specific pathways using the free, open-source, curated and peer reviewed pathway Reactome pathway database22,23. The package requires entrez identifiers, so we convert our PROBEIDs (trancript cluster identifiers) to entrez identifiers using the function mapIDs from the package AnnotationDbi. This will create a named vector that maps the PROBEIDs to the entrez ones.\n\n\n\nWe can now run the enrichment analysis that performs a statistical test based on the hypergeoemtric distribution that is the same as a one sided Fisher–test, which topGO calls “Fisher–classic”. Details can be found in the vignette of the DOSE package24.\n\n\n\n\n\n\n\nNote that we trimmed pathway names to 20 characters.\n\nThe reactomePA package offers nice visualization capabilities. The top pathways can be displayed as a bar char that displays all categories with a p–value below the specified cutoff.\n\n\n\nThe “enrichment map” displays the results of the enrichment analysis as a graph, where the color represents the p–value of the pathway and the edge–thickness is proportional to the number of overlapping genes between two pathways.\n\n\n\nAgain, we see pathways related to signalling and immune response.\n\nThe package clusterProfiler25 can also perform these analyses using downloaded KEGG data. Furthermore, the package EnrichmentBrowser26 additionally offers network–based enrichment analysis of individual pathways. This allows the mapping of the expression data at hand to known regulatory interactions.\n\n\nSession information\n\nAs the last part of this document, we call the function sessionInfo, which reports the version numbers of R and all the packages used in this session. It is good practice to always keep such a record of this as it will help to track down what has happened in case an R script ceases to work or gives different results because the functions have been changed in a newer version of one of your packages. By including it at the bottom of a script, your reports will become more reproducible.\n\nThe session information should also always be included in any emails to the Bioconductor support site along with all code used in the analysis.\n\n\n\n\n\n\nData and software availability\n\nThis article is based on an R markdown file (MA-Workflow.Rmd) which is available as Dataset 1 (Dataset 1. R markdown document to reproduce the results obtained in the article, 10.5256/f1000research.8967.d124759)31 and will also become available as a Bioconductor workflow. This file allows the reader to reproduce the analysis results obtained in this article. All data analyzed are downloaded from ArrayExpress.",
"appendix": "Competing interests\n\n\n\nThe author declares that there are 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\nThe author would like to thank Vladislava Milchevskaya. Julian Gehring and Mike Smith for helpful comments on and small contributions to the workflow. This workflow draws a lot of inspiration from the Bioconductor books27,28 as well as Love et al.’s workflow for gene level analysis of RNA–Seq data29. James W. MacDonald provided valuable information on the evolution of Affymetrix arrays in some of his posts of on the Biocondctor mailing list/support site. The author would also like to thank him for some friendly personal correspondence about the annotation resources available for microarrays in Bioconductor.\n\n\nReferences\n\nPalmieri O, Creanza TM, Bossa F, et al.: Genome-wide Pathway Analysis Using Gene Expression Data of Colonic Mucosa in Patients with Inflammatory Bowel Disease. Inflamm Bowel Dis. 2015; 21(6): 1260–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKolesnikov N, Hastings E, Keays M, et al.: ArrayExpress update--simplifying data submissions. Nucleic Acids Res. 2015; 43(Database issue): D1113–D1116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarvalho BS, Irizarry RA: A framework for oligonucleotide microarray preprocessing. Bioinformatics. 2010; 26(19): 2363–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKauffmann A, Gentleman R, Huber W: arrayQualityMetrics--a bioconductor package for quality assessment of microarray data. Bioinformatics. 2009; 25(3): 415–6. PubMed Abstract | Publisher Full Text | Free 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\nBolstad BM, Irizarry RA, Astrand M, et al.: A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003; 19(2): 185–193. PubMed Abstract | Publisher Full Text\n\nIrizarry RA, Hobbs B, Collin F, et al.: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. 2003; 4(2): 249–264. PubMed Abstract | Publisher Full Text\n\nHuber W, von Heydebreck A, Sültmann H, et al.: Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics. 2002; 18(Suppl. 1): S96–S104. PubMed Abstract | Publisher Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015; 43(7): e47. 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(1): 1–25, Article3. PubMed Abstract | Publisher Full Text\n\nArloth J, Bader DM, Röh S, et al.: Re-Annotator: Annotation Pipeline for Microarray Probe Sequences. PLoS One. 2015; 10(10): e0139516. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHothorn T, Bretz F, Westfall P: Simultaneous inference in general parametric models. Biom J. 2008; 50(3): 346–363. PubMed Abstract | Publisher Full Text\n\nGene Ontology Consortium: Gene Ontology Consortium: going forward. Nucleic Acids Res. 2015; 43(Database issue): D1049–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\ndu Plessis L, Skunca N, Dessimoz C: The what, where, how and why of gene ontology--a primer for bioinformaticians. Brief Bioinform. 2011; 12(6): 723–735. 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\nBourgon R, Gentleman R, Huber W: Independent filtering increases detection power for high-throughput experiments. Proc Natl Acad Sci U S A. 2010; 107(21): 9546–9551. 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\nAlexa A, Rahnenfuhrer J, Lengauer T: Improved scoring of functional groups from gene expression data by decorrelating GO graph structure. Bioinformatics. 2006; 22(13): 1600–1607. 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–D477. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFabregat A, Sidiropoulos K, Garapati P, et al.: The Reactome pathway knowledgebase. Nucleic Acids Res. 2016; 44(D1): D481–D487. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu G, Wang LG, Yan GR, et al.: DOSE: an R/Bioconductor package for disease ontology semantic and enrichment analysis. Bioinformatics. 2015; 31(4): 608–609. PubMed Abstract | Publisher Full Text\n\nYu G, Wang LG, Han Y, et al.: clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012; 16(5): 284–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeistlinger L, Csaba G, Zimmer R: Bioconductor’s EnrichmentBrowser: seamless navigation through combined results of set- & network-based enrichment analysis. BMC Bioinformatics. 2016; 17(1): 45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHahne F, Huber W, Gentleman R, et al.: Bioconductor Case Studies. Use R!. Springer, 2008. Publisher Full Text\n\nGentleman R, Carey VJ, Huber W, et al.: Bioinformatics and Computational Biology Solutions Using R and Bioconductor. Statistics for Biology and Health. Springer, 2005. Publisher Full Text\n\nLove MI, Anders S, Kim V, et al.: RNA-Seq workflow: gene-level exploratory analysis and differential expression [version 1; referees: 2 approved]. F1000Res. 2015; 4: 1070. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu G, He QY: ReactomePA: an R/Bioconductor package for reactome pathway analysis and visualization. Mol Biosyst. 2016; 12(2): 477–479. PubMed Abstract | Publisher Full Text\n\nKlaus B: Dataset 1 in: An end to tend workflow for differential gene expression using Affymetrix microarrays. F1000Research. 2016. Data Source"
}
|
[
{
"id": "14392",
"date": "05 Jul 2016",
"name": "James W. MacDonald",
"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 manuscript is intended to take the reader through a complete analysis of Affymetrix Gene ST arrays, based on a set of arrays downloaded from ArrayExpress. The author covers each step from quality control of the raw data all the way to making comparisons using linear models and testing and visualizing pathways or gene sets.\nMajor comments\nWhile this manuscript is technically correct (e.g., the code does what the author claims, the explanatory text is valid), I am not sure it is as useful as it could be for its intended audience. In other words, this manuscript is intended to provide an inexperienced reader (inexperienced in either R/Bioconductor or statistics or both) with a road map they can follow to learn how to analyze microarray data. However, both the code and the statistical explanations are far too complex to be useful for such a person.\nAs an example, the author explains in mathematical terms what the background correction and summarization steps are intended to accomplish. While this is an important step in the analysis, this could instead be explained in heuristic terms that would be far more approachable for a less statistically savvy audience, while still conveying the general idea.\nThe section on linear modeling and design matrices are similarly impenetrable for non-statisticians. The limma User's Guide has dozens of examples of model matrices, with clear interpretations of the model coefficients. Yet as the author notes, this is probably the number one question on the Bioconductor support site. Rather than re-explaining something that most people clearly don't understand, it would be much more helpful to focus on a single model matrix, and provide a clear, heuristic explanation. The easiest to understand is the most basic orthogonal model that computes the mean of each group, followed by a contrast matrix to make comparisons of interest.\n\nThe overview on linear models is well beyond the scope of this manuscript, and should be excised. The same is true for the section on testing general linear hypotheses. Ideally, an analyst using these tools would understand what they are doing from a statistical perspective, but it's difficult enough for a novice user to comprehend what the code is doing without trying to also understand the statistics.\nSimilarly, the code is more complex than necessary. If the goal is to teach novice users about Bioconductor packages, then the code should be restricted as much as possible to those packages and base R. While ggplot2 style graphics, magrittr style function piping and dplyr two-table verbs may be useful for more advanced R users, in this context they are an added distraction.\n\nAn example of overly-complex code is the filtering step. There are any number of ways to filter out genes that are arguably not expressed. The example is a very sophisticated way to perform this task, but a novice who is just learning doesn't require sophistication, they require something they can understand. Choosing a cutoff based on the distribution of probesets across each array, and then filtering all genes where fewer than 14 samples exceed this cutoff is not very sophisticated, but it is easy to understand, and would only require a few lines of code.\nThe learning curve for R and Bioconductor is steep, and making this manuscript both simpler in terms of the code, and more focused by excluding most of the statistics and any example analyses that do not involve the microarray data would make it more approachable for inexperienced users.\n\nMinor comments\nIn the download step, getAE() is used to download the data, but then the SDRF file is downloaded directly. This is confusing, as getAE() has already downloaded that file. Is there a particular reason for the extra step?\nTo test for model matrices that are not full rank, it's easier to use either nonEstimable() or is.fullrank() from the limma package.\nConclusion\nAs noted above, this manuscript is technically accurate, and the code does exactly what the author claims. But it is too ambitious for the intended audience. Most people who could benefit from this manuscript are not statistically savvy, and the statistical sections will simply confuse them. In addition, it is not likely that many potential readers will be familiar with packages from the 'Hadleyverse', and incorporating those packages in the manuscript makes the code more difficult to understand.\nParing the manuscript down to very simple heuristic statistical explanations, and limiting the code to functions from base R and the various Bioconductor packages being illustrated would make the manuscript more useful for its intended audience.",
"responses": [
{
"c_id": "3735",
"date": "03 Jul 2018",
"name": "Bernd Klaus",
"role": "Author Response",
"response": "# Major comments ## comment 1 While this manuscript is technically correct (e.g., the code does what the author claims, the explanatory text is valid), I am not sure it is as useful as it could be for its intended audience. In other words, this manuscript is intended to provide an inexperienced reader (inexperienced in either R/Bioconductor or statistics or both) with a road map they can follow to learn how to analyze microarray data. However, both the code and the statistical explanations are far too complex to be useful for such a person. As an example, the author explains in mathematical terms what the background correction and summarization steps are intended to accomplish. While this is an important step in the analysis, this could instead be explained in heuristic terms that would be far more approachable for a less statistically savvy audience, while still conveying the general idea. Answer: As suggested, the general style of the workflow was changed a lot in order to guide inexperienced readers through the analysis of microarray data. Code is explained in greater detail, and all technical parts not directly related to the data at hand have either been removed or improved. However, we have decided to keep the the mathematical explanation of background correction and summarization, as we feel that the (rather simple) formulas are easier to understand than text. ## comment 2 The section on linear modeling and design matrices are similarly impenetrable for non-statisticians. The limma User's Guide has dozens of examples of model matrices, with clear interpretations of the model coefficients. Yet as the author notes, this is probably the number one question on the Bioconductor support site. Rather than re-explaining something that most people clearly don't understand, it would be much more helpful to focus on a single model matrix, and provide a clear, heuristic explanation. The easiest to understand is the most basic orthogonal model that computes the mean of each group, followed by a contrast matrix to make comparisons of interest. The overview on linear models is well beyond the scope of this manuscript, and should be excised. The same is true for the section on testing general linear hypotheses. Ideally, an analyst using these tools would understand what they are doing from a statistical perspective, but it's difficult enough for a novice user to comprehend what the code is doing without trying to also understand the statistics. Answer: We gladly adopted this point of criticism and eliminated the rather technical part in which the theory behind design matrices and linear modelling is explained. Instead, we focused on explaining the design matrix we haved used in the dataset at hand in easy-to-follow terms and present its use in the context of single linear model for a single gene. ## comment 3 Similarly, the code is more complex than necessary. If the goal is to teach novice users about Bioconductor packages, then the code should be restricted as much as possible to those packages and base R. While ggplot2 style graphics, magrittr style function piping and dplyr two-table verbs may be useful for more advanced R users, in this context they are an added distraction. An example of overly-complex code is the filtering step. There are any number of ways to filter out genes that are arguably not expressed. The example is a very sophisticated way to perform this task, but a novice who is just learning doesn't require sophistication, they require something they can understand. Choosing a cutoff based on the distribution of probesets across each array, and then filtering all genes where fewer than 14 samples exceed this cutoff is not very sophisticated, but it is easy to understand, and would only require a few lines of code. Answer: While not all of the code from the packages like ggplot2, magrittr and dplyr was removed, an effort was made to explain the code in more detail in order to give the inexperienced user the possibility to understand what it does. In particular, we think that ggplot2 is by now very commonly used and probably becoming a defacto standard plotting package. In the low-intensity filtering step, the fitting of a normal distribution to the histogram was eliminated for the sake of clarity. Instead, we now visually set a vertical cutoff line to the histogram and filter genes with a lower intensity than the cutoff in at least as many samples as the smallest experimental group has. ## comment 4 The learning curve for R and Bioconductor is steep, and making this manuscript both simpler in terms of the code, and more focused by excluding most of the statistics and any example analyses that do not involve the microarray data would make it more approachable for inexperienced users. Answer: Thank you for this useful comment. We have made some effort to tailor the revised version to the beginner in R / Bioconductor and / or statistics. The revised article contains a more focused workflow with more in depth explanations of the code. # Minor comments In the download step, getAE() is used to download the data, but then the SDRF file is downloaded directly. This is confusing, as getAE() has already downloaded that file. Is there a particular reason for the extra step? Answer: There is no particular reason for this step, the code has been changed in order to import the SDRF file from the already downloaded data. To test for model matrices that are not full rank, it's easier to use either nonEstimable() or is.fullrank() from the limma package. Answer: This section was removed, as it is quite technical and is really more of a \"sanity check\" than a necessity."
}
]
},
{
"id": "14390",
"date": "08 Jul 2016",
"name": "Andrea Rau",
"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\nKlaus illustrates an analysis workflow of Affymetrix microarrays for an experimental design with paired samples from two tissues in two separate diseases. The workflow covers steps including downloading and loading raw CEL files into R/Bioconductor, pre-processing and normalization, differential analysis via limma, and functional/pathway enrichment analysis, with detailed R scripts throughout. The workflow is a nice complement to resources that are already available on similar topics (notably the Limma user's guide) as it unites into a single document topics that have been discussed elsewhere in diverse forums.\nMajor remarks:\nAlthough the normalization/linear model/contrast sections are technically correct, I find them to be overly disruptive to the presentation of the analysis. In particular, I would suggest moving the normalization section to an appendix. For the linear models section, presentation of ordinary least squares estimators, matrix notation, distributional assumptions of residuals, etc. seems outside of the scope of this work (or could potentially be presented in an appendix as well). I think that it would be more helpful if, rather than using a separate illustrative example based on the toycars data, the author described in detail: 1) the design matrix for the Palmieri example; 2) a linear model for a single gene from the Palmieri data; 3) a linear model fit independently for each gene, sharing information across all genes (limma), with a brief discussion about the advantage of such an approach; 4) a simplified discussion of various contrasts relevant for this particular study.\n\nI was surprised to see that differential analysis conclusions are based on the raw p-values rather than those adjusted for multiple testing.\n\nI have two additional suggestions that could be of great practical interest to many readers:\n\nI think it would be nice to have a discussion (or at least some references) discussing what can be done if hypotheses are NOT met (e.g., the raw p-values are not uniformly distributed between 0 and 1 for genes under H0). I have in mind the author's recent discussion about using fdrtool to estimate the variance of the null model in the context of differential analyses via DESeq2 (http://www-huber.embl.de/users/klaus/Teaching/DESeq2Predoc2014.html).\n\nIn addition to the fixed patient effect model presented here, I think it would be interesting to present an alternative strategy possible with limma for such a design: estimating correlation within patient blocks using duplicateCorrelation(), see for example section 17.3.6 in the limma User's Guide. Note that I am not suggesting a detailed comparison or presentation of mixed models (that is of course well beyond the scope of this work!), but it may be useful to discuss the code/results for such an approach.\n\nMinor remarks:\nI would suggest specifying in the abstract that the study \"...compares paired inflamed and non-inflamed colon tissue...\", as this was not immediately clear otherwise. In addition, on page 2/paragraph 1, the author writes that the original paper studied \"differences between patients suffering UC or CD\", which makes it sound like the disease comparison was of interest, when in fact it was the intra-patient tissue differences that were studied. It may in fact be helpful to present the experimental design in greater detail at the start of the paper; it was not clear to me until page 23 that the design actually involved paired samples from individuals.\n\nPage 12, the author says \"In order to infer a cutoff from the data, we inspect the histogram of the median–intensities. We visually fit a central normal distribution given by 0.5 · N(5.1, 1.18) to the probe–wise medians\". From the R code, it appears that the distribution fit was N(emp_mu = 5.3, emp_sd = 1.19), where emp_mu and emp_sd were estimated from the median expression values -- this is unclear from the expression \"visually fit\" in the text description. Also unclear to me is why prop_cental was set to be 0.5?\n\nThe phrase \"from the XXXX\" is often used when referencing R packages, rather than \"from XXXX\" or \"from the XXX package\".\n\nThe comparison to the original paper results (page 26) seems unnecessary.\n\nI found it a bit distracting that none of the figures are explicitly referred to by number in the text (e.g., \"In Figure 1, ...\").\n\nThere are occasionally very verbose outputs included in the text (e.g. head(pData(raw_data))) that produce multiple pages of output that are not particularly useful.\n\nThere are a few English typos throughout, e.g. \"inflammed\" should be \"inflamed\", \"constrains\" should be \"constraints\", \"transcriptct clusters\" should be \"transcript clusters\", etc.\n\nSince a very large number of R packages are loaded for the analysis on page 2, it would likely be helpful to either reduce the number of packages needed to be the strict minimum necessary, or to group packages by theme/functionality via comments, for example:\n\nGeneral Bioconductor packages library(Biobase) library(oligoClasses)\n\nAnnotation and data packages library(pd.hugene.1.0.st.v1)\n\nQuality control and pre-processing packages library(oligo)\n\nAnalysis and statistics packages library(limma) library(mvtnorm)\n\nPlotting and color options packageslibrary(gplots) library(ggplot2)\n\nFormatting/documentation packages library(knitr) library(BiocStyle) library(dplyr)",
"responses": [
{
"c_id": "3734",
"date": "03 Jul 2018",
"name": "Bernd Klaus",
"role": "Author Response",
"response": "Klaus illustrates an analysis workflow of Affymetrix microarrays for an experimental design with paired samples from two tissues in two separate diseases. The workflow covers steps including downloading and loading raw CEL files into R/Bioconductor, pre-processing and normalization, differential analysis via limma, and functional/pathway enrichment analysis, with detailed R scripts throughout. The workflow is a nice complement to resources that are already available on similar topics (notably the Limma user's guide) as it unites into a single document topics that have been discussed elsewhere in diverse forums. #Major remarks: # remark 1 Although the normalization/linear model/contrast sections are technically correct, I find them to be overly disruptive to the presentation of the analysis. In particular, I would suggest moving the normalization section to an appendix. For the linear models section, presentation of ordinary least squares estimators, matrix notation, distributional assumptions of residuals, etc. seems outside of the scope of this work (or could potentially be presented in an appendix as well). I think that it would be more helpful if, rather than using a separate illustrative example based on the toycars data, the author described in detail: 1) the design matrix for the Palmieri example; 2) a linear model for a single gene from the Palmieri data; 3) a linear model fit independently for each gene, sharing information across all genes (limma), with a brief discussion about the advantage of such an approach; 4) a simplified discussion of various contrasts relevant for this particular study. Answer: The linear model sections not directly related to the dataset at hand were removed from the workflow. In general, we have tried to tailor the revised workflow to beginners in R and Bioconductor. Therefore, rather than discussing elaborate tools / techniques in detail, we try to point to the relevant literature where appropriate. This is helpful for the reader who needs to deviate from our workflow for his or her own data and maintains a straightforward workflow design. Concerning the single points mentioned in the comment above: 1) The design matrix for the Palmieri example is now explained in detail. 2) In the subsequent part of the workflow, the advise of the reviewer to implement a linear model for a single gene from the Palmieri data was adopted by fitting a linear model to the \"CRAT\" gene, and testing it for differential expression. 3) The advantage of shared information across all genes is discussed briefly when introducing the \"eBayes\"-method in the \"Contrasts and Hypotheses tests\" part. 4) In this workflow, we restrict the contrast analysis to the contrast between non-inflamed and inflamed tissue, as it is the one that is also analyzed in the original paper, and do not include additional contrasts in order to keep the workflow concise. # remark 2 I was surprised to see that differential analysis conclusions are based on the raw p-values rather than those adjusted for multiple testing. Answer: This is indeed a possible cause of confusion. In the workflow, parts of the analysis are based on raw p-values in order to make the results comparable to the original paper. However, we now mention the caveats of this approach more, introduce FDRs in a \"hands-on\" manner explicitly and caution against the use of raw p--values in practice. For the subsequent enrichment analyses, DE genes are identified using an FDR cutoff of 10%. I have two additional suggestions that could be of great practical interest to many readers: # remark 3 I think it would be nice to have a discussion (or at least some references) discussing what can be done if hypotheses are NOT met (e.g., the raw p-values are not uniformly distributed between 0 and 1 for genes under H0). I have in mind the author's recent discussion about using fdrtool to estimate the variance of the null model in the context of differential analyses via DESeq2 (http://www-huber.embl.de/users/klaus/Teaching/DESeq2Predoc2014.html). Answer: As mentioned above, and in concordance with the requests of Jim, we tried to avoid introducing any additional (advanced) methods in order to not confuse the novice R/Bioconductor user. Nevertheless, for workflow users encountering the problem of unusual p-value distributions, we have included references to this phenomenon and its implications by referring to the article on false discovery rate estimation by Korbininan Strimmer and chapter 1-6 of Efron's book on Large-Scale Inference. # remark 4 In addition to the fixed patient effect model presented here, I think it would be interesting to present an alternative strategy possible with limma for such a design: estimating correlation within patient blocks using duplicateCorrelation(), see for example section 17.3.6 in the limma User's Guide. Note that I am not suggesting a detailed comparison or presentation of mixed models (that is of course well beyond the scope of this work!), but it may be useful to discuss the code/results for such an approach. Answer: We acknowledge the advantages in certain applications of a mixed model using duplicateCorrelation() compared to a fixed patient effect model used in this workflow. Given that the workflow is tailored towards unexperienced users, we consider the fixed patient effect model more intuitively understandable. However, we now note explicitly that such an analysis could be performed. Minor remarks: # remark 5 I would suggest specifying in the abstract that the study \"...compares paired inflamed and non-inflamed colon tissue...\", as this was not immediately clear otherwise. In addition, on page 2/paragraph 1, the author writes that the original paper studied \"differences between patients suffering UC or CD\", which makes it sound like the disease comparison was of interest, when in fact it was the intra-patient tissue differences that were studied. It may in fact be helpful to present the experimental design in greater detail at the start of the paper; it was not clear to me until page 23 that the design actually involved paired samples from individuals. Answer: The experimental design is now described in greater detail in abstract and introduction. Also, the sentence that the original paper studied \"differences between patients suffering from UC or CD\" was removed. # remark 6 Page 12, the author says \"In order to infer a cutoff from the data, we inspect the histogram of the median–intensities. We visually fit a central normal distribution given by 0.5 · N(5.1, 1.18) to the probe–wise medians\". From the R code, it appears that the distribution fit was N(emp_mu = 5.3, emp_sd = 1.19), where emp_mu and emp_sd were estimated from the median expression values -- this is unclear from the expression \"visually fit\" in the text description. Also unclear to me is why prop_cental was set to be 0.5? Answer: In an attempt to make low-intensity-filtering more intuitive, we removed the fitting a normal distribution the the histogram of the probe-wise medians. Instead, we now visually set a vertical cutoff line to the histogram and filter genes with a lower intensity than the cutoff in at least as many samples as the smallest experimental group has. # additional minor remarks The phrase \"from the XXXX\" is often used when referencing R packages, rather than \"from XXXX\" or \"from the XXX package\". Answer: Thank you, the respective sentences were corrected accordingly. The comparison to the original paper results (page 26) seems unnecessary. Answer: The comparison to the original paper serves as a \"proof of principle\" for the implemented workflow (to show that what we are doing makes sense). Additionally, we wanted to show that it is relatively straightforward to re--analyze publicly available microarray data using R and Bioconductor. I found it a bit distracting that none of the figures are explicitly referred to by number in the text (e.g., \"In Figure 1, ...\"). Answer: Thank you, the respective parts were changed accordingly. There are occasionally very verbose outputs included in the text (e.g. head(pData(raw_data))) that produce multiple pages of output that are not particularly useful. Answer: The output was shortened where appropriate. The pdata call in particular was retained, as we feel that it is interesting to the readers to see what kind of information comes with a dataset from Array Express. There are a few English typos throughout, e.g. \"inflammed\" should be \"inflamed\", \"constrains\" should be \"constraints\", \"transcriptct clusters\" should be \"transcript clusters\", etc. Answer: Thanks, we tried to perform thorough spell-checking in the revised version. Since a very large number of R packages are loaded for the analysis on page 2, it would likely be helpful to either reduce the number of packages needed to be the strict minimum necessary, or to group packages by theme/functionality via comments, for example: General Bioconductor packages library(Biobase) library(oligoClasses) Annotation and data packages library(pd.hugene.1.0.st.v1) Quality control and pre-processing packages library(oligo) Analysis and statistics packages library(limma) library(mvtnorm) Plotting and color options packageslibrary(gplots) library(ggplot2) Formatting/documentation packages library(knitr) library(BiocStyle) library(dplyr) Answer: Thank you for this very useful suggestion, we implemented it in the revised version; the package-import statements at the beginning of the workflow is now grouped by topic."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1384
|
https://f1000research.com/articles/5-2926/v1
|
28 Dec 16
|
{
"type": "Software Tool Article",
"title": "A Bioconductor workflow for processing and analysing spatial proteomics data",
"authors": [
"Lisa M. Breckels",
"Claire M. Mulvey",
"Kathryn S. Lilley",
"Laurent Gatto",
"Lisa M. Breckels",
"Claire M. Mulvey",
"Kathryn S. Lilley"
],
"abstract": "Spatial proteomics is the systematic study of protein sub-cellular localisation. In this workflow, we describe the analysis of a typical quantitative mass spectrometry-based spatial proteomics experiment using the MSnbase and pRoloc Bioconductor package suite. To walk the user through the computational pipeline, we use a recently published experiment predicting protein sub-cellular localisation in pluripotent embryonic mouse stem cells. We describe the software infrastructure at hand, importing and processing data, quality control, sub-cellular marker definition, visualisation and interactive exploration. We then demonstrate the application and interpretation of statistical learning methods, including novelty detection using semi-supervised learning, classification, clustering and transfer learning and conclude the pipeline with data export. The workflow is aimed at beginners who are familiar with proteomics in general and spatial proteomics in particular.",
"keywords": [
"Bioconductor",
"R Package",
"proteomics",
"spatial proteomics",
"protein sub-cellular localisation",
"mass spectromery",
"machine learning",
"transfer learning"
],
"content": "Introduction\n\nQuantitative mass spectrometry based spatial proteomics involves elaborate, expensive and time consuming experimental protocols and considerable effort is invested in the generation of such data. Multiple research groups have described a variety of approaches to establish high quality proteome-wide datasets (see for example 2 for a review, and 3–5 for recent examples). However, data analysis is as critical as data production for reliable and insightful biological interpretation. Here, we walk the reader through a typical pipeline for the analysis of such data using several Bioconductor packages for the R statistical programming environment.\n\nThe main package to analyse protein localisation data is pRoloc, which offers a set of dedicated functions for the analysis of such data. pRoloc itself relies on MSnbase to manipulate and process quantitative proteomics data. Many other packages are used by pRoloc for clustering, classification and visualisation. Support for interactive visualisation is offered by the pRolocGUI package.\n\nIn this workflow, we will describe how to prepare the spatial proteomics data starting from a spreadsheet containing quantitative mass spectrometry data, through to some essential data processing steps, and finish with different applications of machine learning (Figure 1). We focus on a recent pluripotent mouse embryonic stem cells experiment3. These data, as well as additional annotated and pre-formatted datasets from various species are readily available in the pRolocdata package.\n\nInstallation of Bioconductor packages is documented in detail on the Bioconductor installation help page. Below, we show how to install the four main packages used in this workflow:\n\n\n\nThis procedure is also applicable to any packages from CRAN as well as GitHub. Once a package has been installed, it needs to be loaded for its functionality to become available in the R session; this is done with the library function e.g. to load the pRoloc one would type library(\"pRoloc\") after installation.\n\nIf you have questions about this workflow in particular, or about other Bioconductor packages in general, they are best asked on the Bioconductor support site following the posting guidelines. Questions can be tagged with specific package names or keywords. For more general information about mass spectrometry and proteomics, the readers are invited to read the RforProteomics package vignettes and associated papers6,7.\n\n\nReading and processing spatial proteomics data\n\nAs a use-case, we analyse a recent high-throughput spatial proteomics dataset from pluripotent mouse embryonic stem cells (E14TG2a)3. The data was generated using hyperplexed LOPIT (hyperLOPIT), a state-of-the-art method relying on improved sub-cellular fractionation and more accurate quantitation, leading to more reliable classification of protein localisation across the whole sub-cellular space. The method uses an elaborate sub-cellular fractionation scheme, enabled by the use of Tandem Mass Tag (TMT)8 10-plex and application of the MS data acquisition technique named synchronous precursor selection MS3 (SPS-MS3)9, for TMT quantification with high accuracy and precision. Three biological replicates were generated from the E14TG2a experiment, the first was to target low density fractions and the second and third were to emphasis separation of the denser organelles. The intersect of replicates 1 and 2 was treated as a 20-plex dataset for the analysis. As discussed in the publication3, it has been shown that combining replicates from different gradients can increase spatial resolution10. The combination of replicates resulted in 5032 proteins common to both experiments.\n\nThese, as well as many other data are directly available as properly structured and annotated datasets from the pRolocdata experiment package. In this workflow, we will start with a description of how to generate these ad hoc data objects starting from an arbitrary spreadsheet, as produced by many popular third-party applications.\n\nWhile we focus here on a LOPIT-type dataset, these analyses are relevant for any quantitative spatial proteomics data, irrespective of the fractionation or quantitation (i.e. labelled or label-free) methods.\n\nTo make use of the full functionality of the pRoloc software one needs to import their data into R and prepare them as an MSnSet. The MSnSet is a dedicated data structure for the efficient manipulation and processing of mass spectrometry and proteomics data in R. Figure 2 illustrates a simplified view of the MSnSet structure; there exists 3 key sub-parts (termed slots) to such a data object: (1) the exprs (short for expression data) slot for storing the quantitation data, (2) the fData slot (short for feature-metadata) for storing the feature metadata, and finally (3) the pData slot (short for pheno-metadata, i.e. sample phenotypic data) for storing the sample metadata.\n\nFeature metadata typically contains general annotation about the proteins (accession numbers, description, …), information related to the identification search (confidence scores, number of peptides, …) as well as annotation about know sub-cellular location (see in particular the Markers section) and results from data analysis. The sample metadata would, for example, record what stable isotope labels were used for the respective fraction (when labelled quantitation is used), replicate number, fraction number along the gradient and pooling information.\n\nAnother slot of interest is processingData, that logs the processing MSnSet objects undergo. The processing log can be accessed with the processingData function and is displayed under Processing information in the textual object summary when an MSnSet’s name is typed in the R console.\n\nThere are a number of ways to import quantitation data and create an MSnSet instance. All methods are described in the MSnbase input/output capabilities vignette. One suggested simple method is to use the function readMSnSet2. The function takes a single spreadsheet file name as input and extracts the columns containing the quantitation data, as identified by the argument ecol, to create the expression data, while the other columns in the spreadsheet are appended to the feature metadata slot. By example, in the code chunk below we read in the csv spreadsheet containing the quantitation data from the intersect of replicates 1 and 2 of the mouse map3, using the readMSnSet2 function. The data is as available online with the manuscript (see tab 2 of the xlsx supplementary Dataset 1 in 3, which should be exported as a text-based spreadsheet). It is also available as a csv file in the Bioconductor pRolocdata data package, which we make use of below.\n\nTo use the readMSnSet2 function, as a minimum one must specify the file path to the data and which columns of the spreadsheet contain quantitation data. In the code chunk below, we start by identifying the file that we want to use. The system.file function is used to return the path to the extdata directory from the pRolocdata package, which is where our file of interest resides. We then use the dir function to list the content of that directory and store the path that matches the file name of interest in the csvfile. Note that these two lines are only needed here to locate a file in a package; in a more general use case, the user would define the csvfile variable containing the file name of interest directly.\n\nA common pitfall here is to provide only the file name, rather than full path to the file (which is what is shown below with basename; we don’t print the full path, as it will vary from computer to computer). Note that only specifying the name of the file is sufficient when it exists in the working directory (i.e. the directory in which R is running, which can be queried and changed with the getwd and setwd functions respectively).\n\n\n\nNote that the file is compressed (as indicated by the gz, for gzip, extension), and will be decompressed on-the-fly when read into R.\n\nNext, we need to identify which columns in the spreadsheet contain the quantitation data. This can be done using the getEcols function inside R. The spreadsheet deposited by the authors contains two headers, with the second header containing information about where the quantitation data is stored.\n\nWe can display the names of the second header by calling the getEcols function with the argument n = 2 (the default value is n = 1), to specify that we wish to display the column names of the second line. We also specify the name of the spreadsheet file (defined as csvfile above) and the separator that splits cells.\n\n\n\nIt is now easy for one to identify that the quantitation data, corresponding to the 10 TMT isobaric tags, is located in columns 8 to 27. We now have the two mandatory arguments to readMSnSet2, namely the file name (stored in the csvfile variable) and the quantitation column indices. In addition to these, it is also possible to pass the optional argument fnames to indicate which column to use as the labels by which to identify each protein in the sample. Here, we use fnames = 1 to use the UniProt identifiers contained in the first (unnamed) column of the spreadsheet. We also need to specify to skip the first line of the file (for the same reason that we used n = 2 in getEcols above) to read the csv data and convert it to an MSnSet object, named hl (for hyperLOPIT).\n\n\n\nBelow, we display a short summary of the data. The data contains 5032 proteins/features common across the 2 biological replicates for the respective 2 × 10-plex reporter tags (20 columns or samples), along with associated feature metadata such as protein markers, protein description, number of quantified peptides etc (see below).\n\n\n\nBelow, we examine the quantitative information along the whole gradient for the first 5 proteins. It is also possible to access specific rows and columns by naming the proteins and TMT tag channels of interest.\n\n\n\nThe feature metadata is stored in the fData slot and can be accessed by fData(hl). When using readMSnSet2, automatically, everything that is not defined as quantitation data by ecol or the feature names by fnames is deposited to the fData slot.\n\nWe see the fData contains 25 columns describing information such as the number of peptides, associated markers, machine learning results etc. To identify the feature variable names we can use the function fvarLabels. We see that the first 6 feature variable names contain non-discriminatory label names, so we relabel them to help us identify what feature data information is stored in the associated columns.\n\n\n\nNote that when using the simple readMSnSet2 procedure, the pData slot which is used to store information about the samples/channels is kept empty. As illustrated below, one can use the $ operator to access (or create) individual columns in the metadata slot. It is advised to annotate the channels as well. Below, we annotate the replicate from which the profiles originate and the TMT tag (extracted from the sample/channel names). To do so, we use the sample names that were assigned automatically using the quantiation column names and remove leading X and trailing .1 using the sub function.\n\n\n\nThroughout this workflow we refer to the different columns that are found in the exprs (expression data) slot as channels (short for TMT channels). In the frame of LOPIT and hyperLOPIT these channels constitute the relative abundance of each protein (along the rows) in the channel of interest. Each TMT channel originates from fractions collected from the density gradient, or a set of pooled fractions or may be a sample originating from an alternative preparation e.g. such as from the chromatin enrichment performed in Christoforou et al.3. Information about which gradient fractions were used for which tag should also be stored in the sample metadata pData slot.\n\nThe sample metadata that is distributed with the pRolocdata package for Christoforou’s hyperLOPIT experiment and (as above) the quantitation data file, are located in the extdata in the pRolocdata package on the hard drive.\n\nIn the code chunk below we again use the dir function to locate the filepath to the metadata csv file and then read it into R using read.csv. We then append the metadata to the pData slot. Information about the gradient fractions used and the associated subcellular fraction densities in each replicate are stored here.\n\n\n\nNormalisation. There are two aspects related to data normalisation that are relevant to spatial proteomics data processing. The first one focuses on reducing purely technical variation between channels without affecting biological variability (i.e. the shape of the quantitative profiles). This normalisation will depend on the underlying quantitative technology and the experimental design, and will not be addressed in this workflow. The second aspect, and more specific to spatial proteomics data, is scaling all the organelle-specific profiles into a same intensity interval (typically 0 and 1) by, for example, dividing each intensity by the sum of the intensities for that quantitative feature. This is not necessary in this example as the intensities for each replicate have already been re-scaled to 1 in Proteome Discoverer v1.4 Thermo Fisher. However, if the data require normalisation, the user can execute the normalise function as demonstrated in the below code chunk.\n\n\n\nThis transformation of the data assures cancellation of the effect of the absolute intensities of the quantitative features along the rows, and focus subsequent analyses on the relative profiles along the sub-cellular channels.\n\nThe same normalise function (or normalize, both spellings are supported) can also be applied in the first case described above. Different normalisation methods, such as mean or median scaling, variance stabilisation or quantile normalisation, to cite a few, can be applied to accomodate different needs (see ?normalise for available options).\n\nAs previously mentioned, before combination, the two replicates in the hl data that we read into R were separately normalised by sum (i.e. to 1) across the 10 channels for each replicate respectively. We can verify this by summing each rows for each replicate:\n\n\n\nWe see that some features do not add up exactly to 1 due to rounding errors after exporting to intermediate files. These small deviations do not bear any consequences here.\n\nThe spreadsheet that was used to create the hl MSnSet included the two replicates within one csv file. We also provide individual replicates in the pRolocdata package. Below, we show how to combine MSnSet objects and, subsequently, how to filter and handle missing values. We start by loading the pRolocdata package and the equivalent replicates using the data function.\n\n\n\nAt the R prompt, typing\n\n\n\nwill list the 54 datasets that are available in pRolocdata.\n\nCombining data is performed with the combine function. This function will inspect the feature and sample names to identify how to combine the data. As we want our replicates to be combined along the columns (same proteins, different sets of channels), we need to assure that the respective sample names differ so they can be identified from one another. The function updateSampleNames can be used do this.\n\n\n\nIn addition to matching names, the content of the feature metadata for identical feature annotations must match exactly across the data to be combined. In particular for these data, we expect the same proteins in each replicate to be annotated with the same UniProt entry names and descriptions, but not with the same coverage of number of peptides or peptide-spectrum matches (PSMs).\n\n\n\nBelow, we update the replicate specific feature variable names and remove the shared annotation. In the first line, we update only the feature variable names 3 to 5 (by appending a 1) and in the second line, we apply the updateFvarLabels function to update all feature variable names (by appending a 2).\n\n\n\nWe can now combine the two experiments into a single MSnSet:\n\n\n\nMore details about combining data are given in the dedicated Combining MSnSet instances section of the MSnbase tutorial vignette.\n\nMissing data are a recurrent issue in mass spectrometry applications, and should be addressed independently of this workflow11,12. In 13, we have described how a high content in missing values in spatial proteomics data and their inappropriate handling leads to a reduction of sub-cellular resolution. Missing data can be imputated using MSnbase’s impute function. The method underlying the imputation method is then determined by a methods parameter (see ?impute for available options). To impute missing values using nearest neighbour imputation, one would:\n\n\n\nIn our particular case, missing values are indicative of protein groups that were not acquired in both replicates (Figure 4).\n\nNote that the features are re-ordered to highlight cluster of proteins with similar numbers of missing values.\n\nWe prefer to remove proteins that were not assayed in both replicated experiments. This is done with the filterNA function that removes features (proteins) that contain more than a certain proportion (default is 0) missing values. The Processing information section summarises how combining and filtering missing values (subsetting) changed the dimensions of the data.\n\n\n\n\n\n\n\nWhen more than 2 datasets are to be combined and too many proteins have not been consistently assayed, leading to too many proteins being filtered out, we suggest to implement an ensemble of classifiers voting on protein-sub-cellular niche membership over the output of several experiments (see section Supervised machine learning for the description of sub-cellular assignments).\n\n\nQuality control\n\nData quality is routinely examined through visualisation to verify that sub-cellular niches have been separated along the gradient. Based on De Duve's principle14 proteins that co-localise in a cell, exhibit similar quantitation profiles across the gradient fractions employed. One approach that has been widely used to visualise and inspect high throughput mass spectrometry-based proteomics data is principal components analysis (PCA). PCA is one of many dimensionality reduction methods, that allows one to effectively summarise multi-dimensional data in to 2 or 3 dimensions to enable visualisation. Very generally, the original continuous multi-dimensional data is transformed into a set of orthogonal components ordered according to the amount of variability that they describe. The plot2D and plot3D methods in pRoloc allows one to plot the principal components (PCs) of a dataset against one another, by default the first two components are plotted on the x- and y-axis for the plot2D function, and first three components are loaded for the plot3D function, respectively (the dims argument can be used to plot other PCs). If distinct clusters are observed, we assume that there is organellar separation present in the data. Although, representing the multi-dimensional data along a limited set of PCs does not give us a hard quantitative measure of separation, it is extremely useful summarising complex experimental information in one figure, to get a simplified overview of the data.\n\nIn the code chunk below we produce a 2-dimensional PCA plot of the mouse stem cell dataset (Figure 5). Each point on the plot represents one protein. We can indeed see several distinct protein clusters. We specify fcol = NULL to ignore feature metadata columns and not annotate any feature (protein) with a colour. We will see later how to use this argument to annotate the PCA plot with prior information about sub-cellular localisation.\n\n\n\nEach dot represents a single protein, and cluster of proteins represent proteins residing in the same sub-cellular niche. The figure on the right bins proteins and represent the bins density to highlight the presence of protein clusters.\n\nIn the first instance we advise one to visualise their data without any annotation (i.e. with fcol = NULL), before proceeding with data annotation. The identification of well resolved clusters in the data, constitutes an unbiased assessment of the data structure, demonstrating the successful separation of sub-cellular clusters.\n\nIt is also useful to visualise the relative intensities along the gradient to identify channels displaying particularly low yield. This can be done using the plotDist and boxplot functions, that plot the protein profiles occupancy along the gradient (we also display the mean channel intensities) and a boxplot of the column intensities. In the two plots displayed on Figure 6, we re-order the TMT channels to pair corresponding channels in the two replicates (rather than ordering the channels by replicate).\n\nThe red dots represent the mean relative intensity for each channel.\n\n\nMarkers\n\nIn the context of spatial proteomics, a marker protein is defined as a well-known resident of a specific sub-cellular niche in a species and condition of interest. Applying this to machine learning (ML), and specifically supervised learning, for the task of protein localisation prediction, these markers constitute the labelled training data to use as input to a classification analyses. Defining well-known residents, and obtaining labelled training data for ML analyses can be time consuming, but it is important to define markers that are representative of the multivariate data space and on which a classifier will be trained and generated. pRoloc provides a convenience function, addMarkers, to directly add markers to an MSnSet object, as demonstrated in the code chunk below. These marker sets can be accessed using the pRolocmarkers() function. Marker sets are stored as a simple named vector in R, and originate from in-house user-defined sets of markers or from previous published studies13, which are continuosly updated and integrated.\n\n\n\n\n\nThese markers can then be mapped to an MSnSet’s featureNames. The mouse dataset used here has Uniprot IDs stored as the featureNames (see head(featureNames(hl))) and the names of the vector of the mouse markers stored in pRoloc (mmus markers) are also Uniprot IDs (see head(mrk) in the code chunk below, that displays the 6 first markers), so it is straightforward to match names between the markers and the MSnSet instance using the addMarkers function.\n\n\n\nWe recommend at least 13 markers per sub-cellular class (see the Optimisation section for details about the algorithmic motivation of this number). Markers should be chosen to confidently represent the distribution of genuine residents of a sub-cellular niche. We generally recommend a conservative approach in defining markers to avoid false assignments when assigning sub-cellular localisation of proteins of unknown localisation. A more relaxed definition of markers, i.e. one that broadly or over-confidently defines markers, risks the erroneous assignment of proteins to a single location, when, in reality, they reside in multiple locations (including the assumed unique location). One can not expect to identify exact boundaries between sub-cellular classes through marker annotation alone; the definition of these boundaries is better handled algorithmically, i.e. after application of the supervised learning algorithm, using the prediction scores (as described in the Classification section, in particular Figure 16).\n\nIf the protein naming between the marker sets and the MSnSet dataset are different e.g. the markers are labelled by Uniprot accession numbers and the dataset entries are labelled by Uniprot entry names, one will have to convert and match the proteins according to the appropriate identifier. Sometimes, we find the equivalent entry name, Uniprot ID or accession number is stored in the feature metadata, which makes conversion between identifers relatively straightforward. If this is not the case however, conversion can be performed using biomaRt, the Bioconductor annotation resouces or any conversion softwares available online.\n\nIt is also possible for users to use their own marker list with the addMarkers function. The user needs to create a named vector of marker localisation, or a create a csv file with two columns (one for the protein names, one for the corresponding sub-cellular marker annotation) and pass the vector or file name respectively to the function. As previously mentioned, the protein names of these markers must match some (but not necessarily all) of the MSnSet’s feature names. See ?addMarkers for more details.\n\nIn general, the Gene Ontology (GO)15, and in particular the cellular compartment (CC) namespace are a good starting point for protein annotation and marker definition. It is important to note however that automatic retrieval of sub-cellular localisation information, from pRoloc or elsewhere, is only the beginning in defining a marker set for downstream analyses. Expert curation is vital to check that any annotation added is in the correct context for the biological question under investigation.\n\nHaving added a the mouse markers to our fData from the pRolocmarkers, we can now visualise these annotations along the PCA plot using the plot2D function and then use the addLegend function to map the marker classes to the pre-defined colours. We also display the data along the first and seventh PCs using the dims argument. Note that in these calls to the plot2D function, we have omitted the fcol argument and use of the default \"markers\" feature variable to annotate the plot. We choose to display PCs 1 and 7 to illustrate that while upper principal components explain much less variability in the data (2.23% for PC7, as opposed to 48.41% for PC1), we see that the mitochondrial (purple) and peroxisome (dark blue) clusters can be differentiated, despite the apparent overlap in the visualisation of the two first PCs (Figure 7).\n\nThe data can also be visualised along three PCs using the plot3D function (Figure 8). When produced interactively, the plot can be rotated and zoomined using the mouse.\n\n\n\nThe default colours for plotting have been defined so as to enable the differentiation of up to 30 classes. If more are provided, different character symbols (circles, squares, and empty and solid symbols) are used. The colours and the default plotting characters (solid dots for the markers and empty circles for the features of unknown localisation) can of course be changed, as described in the setStockcol manual page.\n\nAs demonstrated in 3 and illustrated in the PCA plot (Figure 7), the Golgi apparatus proteins (dark brown) display a dynamic pattern, noting sets of Golgi marker proteins that are distributed amongst other subcellular structures, an observation supported by microscopy. As such, we are going to reset the annotation of Golgi markers to unknown using the fDataToUnknown function. It is often used to replace empty strings (\" \") or missing values in the markers definition to a common definition of unknown localisation.\n\n\n\nAnother useful visualisation that relies on marker annotation is the representation of the protein profiles occupancy along the gradient using the plotDist function. While the PCA plot enables efficient visualisation of the complete dataset and assessment the relative separation of different sub-cellular niches, comparing profiles of a few marker clusters is useful to assess how exactly they differ (in terms of peak channels, for example). On Figure 9, we plot the profile of the mitochondrial and peroxisome markers to highlight the differences in profiles between these two sets of markers in the channels labelled with tag 129C, as represented above along the 7th PC on the PCA plot on Figure 7.\n\n\n\n\n\nFinally, in addition to plot2D, the plot3D function allows to interactively explore a 3-dimensional plot of the data.\n\nIn addition to adding annotation using the addMarkers function, one can store specific sets of proteins by using the Features of interest infrastructure from the MSnbase package. If users have specific subsets of proteins they wish to highlight in their data (possibly across multiple experiments) they would first create a FeaturesOfInterest object. For example, if we wanted to highlight a proteins with the accession numbers Q8CG48, Q8CG47, Q8K2Z4, and Q8C156, which are some of the proteins known to form part of the 13S condensin complex, we would create a first create a FeaturesOfInterest object, and subsequently highlight their location on a PCA plot with the highlightOnPlot function.\n\n\n\nUsers can also create several sets of FeaturesOfInterest object and store them in a FoICollection.\n\nIt is also worthy of note that it is possible to search for a specific protein of interest by featureNames or using any identifying information found in the fData columns by using the search box on the pRolocVis application part of the pRolocGUI package (see section on interactive visualisation). This can be handy for quickly searching and highlighting proteins on the fly, the disavanatge here is that proteins can only be searched for a one-by-one basis.\n\n\nReplication\n\nWith the aim of maximising the sub-cellular resolution and, consequently, the reliability in protein sub-cellular assignments, we follow the advice in 10 and combine replicated spatial proteomics experiments as described above. Indeed, Trotter et al. have shown a significant improvement in protein–organelle association upon direct combination of single experiments, in particular when these resolve different subcellular niches.\n\nDirect comparisons of individual channels in replicated experiments does not provide an adequate, goal-driven assessment of different experiments. Indeed, due to the nature of the experiment and gradient fraction collection, the quantitative channels do not correspond to identical selected fractions along the gradient. For example, in the table below (taken from hl’s pData) TMT channels 127C (among others) in both replicates originate from different sets of gradient fractions (gradient fractions 7–9 and 8–9 for each replicate, respectively). Different sets of gradient fractions are often pooled to obtain enough material and optimise acurate quantitation.\n\nThe more relevant comparison unit is not a single channel, but rather the complete protein occupancy profiles, which are best visualised experiment-wide on a PCA plot. As such, we prefer to focus on the direct, qualitative comparison of individual replicate PCA plots, assuring that each displays acceptable sub-cellular resolution. Note that in the code chunk below, we mirror the x-axis to represent the two figures with the same orientation. The interactive \"compare\" application part of the pRolocGUI package is also useful for examining replicate experiments (see the next section interactive visualisation for details).\n\n\n\nIn addition, the reproducibility can be assessed by performing independent classification analyses on each replicate (see the section on Supervised machine learning below) and comparing the the results. Even when the gradient conditions different (for unexpected technical or voluntary reasons, to maximise resolution when combining experiments10), one expects agreement in the most confident organelle assignments.\n\n\nInteractive visualisation\n\nVisualisation and data exploration is an important aspect of data analyses allowing one to shed light on data structure and patterns of interest. Using the pRolocGUI package we can interactively visualise, explore and interrogate quantitative spatial proteomics data. The pRolocGUI package is currently under active development and it relies on the shiny framework for reactivity and interactivity. The package currently distributes 3 different GUI’s (main (default), compare or classify) which are wrapped and launched by the pRolocVis function.\n\nIn the code chunk below we launch the main app (note, we do not need to specify the argument, app = \"main\" as it is the default).\n\n\n\nAs diplayed in the screenshot in Figure 12, the main application is designed for exploratory data analysis and is divided into 3 tabs: (1) PCA, (2) Profiles and (3) Table selection. The default view upon loading is the PCA tab, which features a clickable interface and zoomable PCA plot with an interactive data table for displaying the quantitation information. Particular proteins of interest can be highlighted using the text search box. There is also a Profiles tab for visualisation of the protein profiles, which can be used to examine the patterns of proteins of interest. The Table selection tab provides an interface to control data table column selection. A short animation https://github.com/lmsimp/bioc-pRoloc-hyperLOPIT-workflow/blob/master/Figures/pRolocVis_pca.gif illustrating the interface is available in the manuscript repository16.\n\nThe compare application is useful for examining two replicate experiments, or two experiments from different conditions, treatments etc. The compare application is called by default if the input object to pRolocVis is an MSnSetList of 2 MSnSets, but it can also be specified by calling the argument app = \"compare\". For example, in the code chunk below we first create an MSnSetList of replicates 1 and 2 of the hyperLOPIT data, this is then passed to pRolocVis.\n\n\n\nThe comparison app loads the two PCA plots side-by-side. Only common proteins between the two datasets are displayed. As per the main application, proteins can be searched, identified and highlighted on both PCA plots and in the dedicated profiles tab. One key feature of the compare application is the ability to re-map the second dataset onto the PCA data space of the first (reference) dataset (see ?pRolocVis and the argument remap = TRUE). Using the first dataset as the reference set, PCA is carried out on the first dataset and the standard deviations of the PCs (i.e. the square roots of the eigenvalues of the covariance/correlation matrix) and the matrix of variable loadings (i.e. a matrix whose columns contain the eigenvectors) are stored and then used to calculate the principal components of the second dataset. Both datasets are scaled and centered in the usual way. The first dataset appears on the left, and the second re-mapped data appears on the right. The order of the first (the reference data for remapping) and second dataset can be changed through regeneration/re-ordering of the MSnSetList object.\n\nThe final application classify, has been designed to view machine learning classification results according to user-specified thresholds for the assignment of proteins to its sub-cellular location, as discussed later in the subsection thresholding in the supervised machine learning section.\n\n\nNovelty detection\n\nThe extraction of sub-cellular protein clusters can be difficult owing to the limited number of marker proteins that exist in databases and elsewhere. Furthermore, given the vast complexity of the cell, automatic annotation retrieval does not always give a full representation of the true sub-cellular diversity in the data. For downstream analyses, such as supervised ML, it is desirable to obtain reliable markers that cover as many sub-cellular niches as possible, as these markers are directly used in the training phase of the ML classification. We find that a lack of sub-cellular diversity in the labelled training data leads to prediction errors, as unlabelled instances can only be assigned to a class that exists in the training data17. In such scenarios, novelty detection can be useful to identify data-specific sub-cellular groupings such as organelles and protein complexes. The phenotype discovery (phenoDisco) algorithm17 is one such method and is available in pRoloc. It is an iterative semi-supervised learning method that combines the classification of proteins on existing labelled data with the detection of new clusters.\n\nIn addition to extracting new phenotypes, novelty detection methods are also useful for confirming the presence of known or postulated clusters in an unbiased fashion. For example, in 3 the phenoDisco algorithm was used to confirm the data-specific presence of the nucleus and nucleus sub-compartments. In the code chunk below, we demonstrate how to do this analysis, highlighting some of the optional arguments and parameters available for phenotype extraction and give some advice on how to interpret the output.\n\nAs the phenoDisco algorithm is semi-supervised it uses both labelled (markers) and unlabelled data to explore the data structure and find new sub-cellular data clusters. Thus the first step is to define some input labelled data i.e. markers, that the algorithm will use as input for the supervised learning aspect of the algorithm. As described in3 we define a set of markers to use as input for the analyses that cover well-known residents from three distinct organelle structures; the mitochondria, plasma membrane and ER, and from three well-known and abundant protein complexes; the proteasome and two ribosomal subunits, 40S and 60S. These input markers are stored in the featureData column of hl where fcol = \"phenoDisco.Input\". We can use the convenience accessor function getMarkers to print out a table of the markers contained in this marker set. These initial markers were manually curated using information from the UniProt database, the Gene Ontology and the literature.\n\n\n\nIn the code chunk below we show how to run the phenoDisco function and return a novelty detection result, according to the specified parameters. The algorithm parameters times (number of iterations) and GS (minimum number of proteins required to form a new phenotype) are passed to the function, along with the fcol to tell the algorithm where the input training data is contained.\n\n\n\nNote: We do not evaluate this code chunk in this document as the algorithm is computationally intensive and best parallelised over multiple workers. This phenoDisco analysis took 24 hours to complete when parallelised over 40 workers.\n\nThe argument times indicates the number of times we run unsupervied Gaussian Mixture Modelling before defining a new phenotype cluster. The recommended minimum and default value is 100. In the above code chunk we increase the value to times = 200 as we have found for larger datasets (e.g. 5000+ proteins) a higher times is requried for convergence. GS defines the minimum number of proteins allowed per new data cluster and thus heavily influences what type of new clusters are extracted. For example, if a user is interested in the detection of small complexes they may wish to use a small GS = 10, or GS = 20 etc. If they wish to detect larger, more abundant sub-cellular niches a much higher GS would be preferable. Specifying a small GS can be more time consuming than using a larger GS, and there is a trade off between finding interesting small complexes and those that may not be of interest as we find there is a tendancy to find more noise when using a small GS compared to using a higher one.\n\nOne may also consider increasing the search space for new data clusters by increasing the value of the parameter G. This defines the number of GMM components to test and fit; the default is G = 1:9 (the default value in the mclust package18). One should note that the decreasing the GS, and increasing the values of the arguments times, G (among other function arguments, see ?phenoDisco) will heavily influence (increase) the total time taken to run the algorithm. phenoDisco supports parallelisation and we strongly suggest you make use of a parallel processing to run these analyses.\n\nThe ouput of running the phenoDisco algorithm is an MSnSet containing the new data clusters, appended to the featureData under the name pd. The results can be displayed by using the getMarkers function. We see that 5 new phenotype data clusters were found.\n\n\n\nWe can plot the results using the plot2D function (Figure 14). The five new phenotype data clusters can be extracted and examined. In the code chunk below we write the results to a csv file using the write.exprs function. We use the argument fDataCols to specify which columns of the featureData to write.\n\n\n\n\n\nWe can also examine each phenotype interactively and visualise their protein profiles by using the pRolocVis function in the pRolocGUI package. We found that phenotype 1 was enriched in nucleus associated proteins, phenotype 2 in chromatin associated proteins, phenotype 3 in cytosolic and phenotypes 4 and 5 in both lysosomal and endosomal proteins.\n\n\nSupervised machine learning\n\nSupervised machine learning, also known as classification, is an essential tool for the assignment of proteins to distinct sub-cellular niches. Using a set of labelled training examples i.e. markers, we can train a machine learning classifier to learn a mapping between the data i.e. the quantitative protein profiles, and a known localisation. The trained classifier can then be used to predict the localisation of a protein of unknown localisation, based on its observed protein profiles. To date, this method has been extensively used in spatial quantitative proteomics to assign thousands of proteins to distinct sub-cellular niches3,10,19,20–22.\n\nThere are several classification algorithms readily available in pRoloc, which are documented in the dedicated pRoloc machine learning techniques vignette. We find the general tendancy to be that it is not the choice of classifier, but the improper optimisation of the algorithmic parameters, that limits classification accuracy. Before employing any classification algorithm and generating a model on the training data, one must first find the optimal parameters for the algorithm of choice.\n\nIn the code chunk below we use a Support Vector Machine (SVM) to learn a classifier on the labelled training data. As previously mentioned, one first needs to train the classifier’s parameters before an algorithm can be used to predict the class labels of the proteins with unknown location. One of the most common ways to optimise the parameters of a classifier is to partition the labelled data into training and testing subsets. In this framework parameters are tested via a grid search using cross-validation on the training partition. The best parameters chosen from the cross-validation stage are then used to build a classifier to predict the class labels of the protein profiles on the test partition. Observed and expected classication results can be compared, and then used to assess how well a given model works by getting an estimate of the classifiers ability to achieve a good generalisation i.e. that is given an unknown example predict its class label with high accuracy. In pRoloc, algorithmic performance is estimated using stratified 80/20 partitioning for the training/testing subsets respectively, in conjuction with five-fold cross-validation in order to optimise the free parameters via a grid search. This procedure is usually repeated 100 times and then the best parameter(s) are selected upon investigation of classifier accuracy. We recommend a minimum of 13 markers per sub-cellular class for stratified 80/20 partitioning and 5-fold cross-validation; this allows a minimum of 10 examples for parameter optimisation on the training partition i.e. 2 per fold for 5-fold cross-validation, and then 3 for testing the best parameters on the validation set.\n\nClassifier accuracy is estimated using the macro F1 score, i.e. the harmonic mean of precision and recall. In the code chunk below we demonstrate how to optimise the free parameters, sigma (the inverse kernel width for the radial basis kernel) and cost (the cost of constraints violation), of a classical SVM classifier with a Gaussian kernel using the function svmOptimisation. As the number of labelled instances per class varies from organelle to organelle, we can account for class imbalance by setting specific class weights when generating the SVM model. Below the weights, w are set to be inversely proportional to the class frequencies.\n\n\n\n\n\nAs mentioned previously, we rely on the default feature variable \"markers\" to define the class labels and hence do not need to specify it in the above code chunk. To use another feature variables, one needs to explicitly specify its name using the fcol argument (for example fcol = \"markers2\").\n\nThe output params is an object of class GenRegRes; a dedicated container for the storage of the design and results from a machine learning optimisation. To assess classifier performance we can examine the macro F1 scores and the most frequently chosen parameters. A high macro F1 score indicates that the marker proteins in the test dataset are consistently and correctly assigned by the the algorithm. Often more than one parameter or set of parameters gives rise to the best generalisation accuracy. As such it is always important to investigate the model parameters and critically assess the best choice. The f1Count function counts the number of parameter occurences above a certain F1 value. The best choice may not be as simple as the parameter set that gives rise to the highest macro F1 score. One must be careful to avoid overfitting, and choose parameters that frequently provide high classification accuracy. Below, we see that only a sigma of 0.1 produces macro F1 scores above 0.6, but that a cost of 16 is much more frequently chosen than lower values.\n\n\n\nThe parameter optimistion results can also be visualised as a boxplot or heatmap, as shown in Figure 15. The plot method for GenRegRes object shows the respective distributions of the 100 macro F1 scores for the best cost/sigma parameter pairs, and levelPlot shows the averaged macro F1 scores, for the full range of parameter values. These figures also indicate that values of 0.1 and 16 for sigma and cost consistently deliver better classification scores.\n\n\n\nColours indicate class membership and point size are representative of the classification confidence.\n\nBy using the function getParams we can extract the best set of parameters. Currently, getParams retrieves the first best set automatically, but users are encouraged to critically assess whether this is the most wise choice (which it is, as demonstrated above).\n\n\n\nOnce we have selected the best parameters we can then use them to build a classifier from the labelled marker proteins.\n\nWe can use the function svmClassification to return a classification result for all unlabelled instances in the dataset corresponding to their most likely sub-cellular location. The algorithm parameters are passed to the function, along with the class weights. As above, the fcol argument does not need to be specified as we use the labels defined in the default \"markers\" feature variable.\n\n\n\nIn the code chunk above, we pass the whole params parameter results and, internally, the first pair that return the highest F1 score are returned (using the getParams function above). It is advised to always check that these are actually good parameters and, if necessary, set them explicitly, as shown below.\n\n\n\nAutomatically, the output of the above classification, the organelle predictions and assignment scores, are stored in the featureData slot of the MSnSet. In this case, they are given the labels svm and svm.scores for the predictions and scores respectively. The resultant predictions can be visualised using plot2D. In the code chunk below plot2D is called to generate a PCA plot of the data and fcol is used to specify where the new assignments are located e.g. fcol = \"svm\".\n\nAdditionally, when calling plot2D we can use the cex argument to change the size of each point on the plot to be inversely proportional to the SVM score. This gives an initial overview of the high scoring localisations from the SVM predictions.\n\nThe adjustment of the point size intuitively confers important information that is more difficult to define formally (we will address in the next section). The classifier (SVM in our case, but this is also valid of other classifiers) defines boundaries based on the labelled marker proteins. These class/organelle boundaries define how non-assigned proteins are classified and with what confidence.\n\nIt is common when applying a supervised classification algorithm to set a specific score cutoff on which to define new assignments, below which classifications are kept unknown/unassigned. This is important as in a supervised learning setup, proteins can only be predicted to be localised to one of the sub-cellular niches that appear in the labelled training data. We can not guarantee (and do not expect) that the whole sub-cellular diversity to be represented in the labelled training data as (1) finding markers that represent the whole diversity of the cell is challenging (especially obtaining dualand multiply-localised protein markers) and (2) many sub-cellular niches contain too few proteins to train on (see above for a motivation of a minimum of 13 markers).\n\nDeciding on a threshold is not trivial as classifier scores are heavily dependent upon the classifier used and different sub-cellular niches can exhibit different score distributions, as highlighted in the boxplot below. We recommend users to set class-specific thresholds. In the code chunk below we display a boxplot of the score distributions per organelle (Figure 17).\n\nThere are many ways to set thresholds and the choice of method will depend on the biological question and experimental design at hand. One viable approach in the frame of the above experimetal design would be to manually set a FDR, say 5%, per organelle. To do this the user would examine the top scoring predictions for each organelle, and then set a threshold at the score at which they achieve 5% of false assignments per organelle. The definintion of a false assignment would depend on the information available, for example, validity or lack of validity for the localisation from another experiment as reported in the literature or a reliable database. If such information is not available, one crude method is to set a threshold per organelle by extracting the median or 3rd quantile score per organelle. For example, in the code chunk below, we use the orgQuants function to extract the median organelle scores and then pass these scores to the getPredictions function to extract the new localisations that meet this scoring criteria. Any sub-cellular predictions that fall below the specified thresholds are labelled as unknown.\n\n\n\n\n\n\n\n\n\n\n\nThe organelle threshold (ts above) can also be set manually using an interactive app (see below) or by using a named vector of thresholds, as shown in the putative example below for 4 organelles.\n\n\n\nThe output of getPredictons is the original MSnSet dataset with a new feature variable appended to the feature data called fcol.pred (i.e. in our case svm.pred) containing the prediction results. The results can also be visualised using plot2D function (Figure 18) and extracted by retrieving that specific column from the feature metadata using, for example, fData(hl)$svm.pred.\n\n\n\nThere is also a dedicated interactive application to help users examine these distributions in the pRolocGUI package (Figure 19). This app can be launched via the pRolocVis function and specifying the argument app = \"classify\" along with the relevent fcol, scol and mcol which refer to the columns in the feature data that contain the new assignments, assignment scores and markers respectively (see also fvarLabels(svmres)).\n\n\n\nThe data is loaded and displayed on a PCA plot and a boxplot is used to display the classifier scores by data class. On the left, there is a sidebar panel with sliders to control the thresholds upon which classifications are made. There are two types of cut-off that the user can choose from: (1) Quantile and (2) User-defined. By default, when the application is launched quantile scoring is selected and set to 0.5, the median. The class-specific score thresholds that correspond to selecting the desired quantile are shown as red dots on the boxplot. The assignments on the PCA plot are also updated according to the selected threshold. The quantile threshold can be set by moving the corresponding quantile slider. If the users wishes to set their own cut-offs, the User-defined radio button must be selected and then the sliders for defining user-specified scores become active and the scores are highlighted on the boxplot by blue dots. For more information we refer users to the pRolocGUI tutorial vignette.\n\n\nTransfer learning\n\nIn addition to high quality MS-based quantitative proteomics data, there exist a number of other sources of information that are freely available in the public domain that may be useful to assign a protein to its sub-cellular niche. For example, imaging from immunofluorescence microscopy, protein annotations and sequences, and protein-protein interactions among others, represent a rich and vast source of complementary information. We can integrate this auxiliary information with our primary MS-based quantitative data using a paradigm known as transfer learning (TL). The integration of data between different technologies is one of the biggest challenges in computational biology to date and the pRoloc package provides functionality to do such analyses. We recently developed two TL algorithms using a k-NN and SVM framework and applied them to the task of protein localisation prediction23. In this section we will begin with explaining the concept of TL and then show how to apply this in the frame of spatial proteomics and protein localisation prediction.\n\nIn TL one typically has a primary task that one wishes to solve, and some complementary (often heterogeneous) auxiliary information that is related to the primary learning objective, that can be used to help solve the primary goal. For example, here our primary task is to assign proteins to their sub-cellular niche with high generalisation accuracy from data collected from quantitative MS-based experiments. We have already seen that straightforward supervised ML works well for these types of experiments, however, TL is particularly useful for classes that are not as well separated.\n\nIn the example below we extract GO information to use as an auxiliary data source to help solve our task of protein localisation prediction.\n\nUsing the functions setAnnotationParams and makeGoSet we can contruct an auxiliary MSnSet of GO terms, from the primary data’s features i.e. the protein accession numbers. All the GO terms associated to each accession number are retrieved and used to create a binary matrix where a 1 (0) at position (i, j) indicates that term j has (not) been used to annotate protein i. The GO terms are retrieved from an appropriate repository using the biomaRt package. The specific Biomart repository and query will depend on the species under study and the type of identifiers. The first step is to construct the annotation parameters that will enable to perform the query, which is done using setAnnotationParams. Typing into the R console par <- setAnnotationParams() will present two menus, firstly asking you to identify the species of study, and then what type of identifier you have used to annotate the proteins in your MSnSet. It is also possible to pass arbitrary text to match the species e.g. in the code chunk below we pass \"Mus musculus\", and the identifier type for our data (see featureNames(hl)) which is \"Uniprot/Swissprot\", for the Biomart query.\n\n\n\nNow we have contructed the query parameters we can use the makeGoSet function to retrieve and build an auxiliary GO MSnSet as described above. By default, the cellular component terms are downloaded, without any filtering on evidence codes. It is also possible to download terms from the molecular function and biological process GO namespaces, and also apply filtering based on evidence codes as desired, see ?makeGoSet for more details.\n\n\n\nThe function makeGoSet uses the biomaRt package to query the relevent database (e.g. Ensembl, Uniprot) for GO terms. All GO terms that have been observed for the 5032 proteins in the hyperLOPIT dataset are retrieved. Users should note that the number of GO terms used is also dependent on the database version queried and thus is always subject to change. We find it is common to see GO terms with only one protein assigned to that term. Such terms do not bring any information for building the classifier and are thus removed using the filterBinMSnSet function.\n\n\n\nNow that we have generated our auxiliary data, we can use the k-NN implementation of transfer learning available in pRoloc to integrate this with our primary MS-based quantitative proteomics data using the functions knntlOptimisation to estimate the free-parameters for the integration, and knntlClassification to do the predictions. We have shown that using transfer learning results in the assignment of proteins to sub-cellular niches with a higher generalisation accuracy than using standard supervised machine learning with a single source of information23.\n\nThe first step, as with any machine learning algorithm, is to optimise any free paramaters of the classifier. For the k-NN TL classifier there are two sets of parameters that need optimising: the first set are the k’s for the primary and auxiliary data sources required for the nearest neighbour calculations for each data source. The second set of parameters (noted by a vector of θ weights) that require optimising are the class weights, one per subcellular niche, that control the proportion of primary and auxiliary data to use for learning. A weight can take any real value number between 0 and 1. A weight of θ = 1 indicates that all weight is given to the primary data (and this implicitly implies that a weight of 1 − θ = 0 is given to the auxiliary data), and similarly a weight of θ = 0 implies that all weight is given to the auxiliary data (so 0 is given to the primary source). If we conduct a parameter seach and test weights θ = 0, 1/3, 2/3, 1 for each class, and if we have, for example 10 subcellular niches, this will result in 410 different combinations of parameters to test. The parameter optimisation is therefore time consuming and as such we recommend making use of a computing cluster (code and submissing scripts are also available in the supporting information). The markers in the hl dataset contain 14 subcellular classes. If we examine these markers and classes on the PCA plot above we can see that in particular the two ribosomes and two nuclear compartments are highly separated along the first two components, this is also evident from the profiles plot which gives us a good indication that these subcellular niches are well-resolved in the hyperLOPIT dataset. Transfer learning is particularly useful for classes that are not as well separated. We find that subcellular niches that are well-separated under hyperLOPIT and LOPIT obtain a class score of 1 (i.e. use only primary data from transfer learning23). Therefore, for the optimisation stage of the analyses we can already infer a subcellular class weight of 1 for these niches and only optimise over the remaining organelles. This can significantly cut down optimisation time as by removing these 4 classes from the optimisation (and not the classification) we only have 410 class weight combinations to consider instead of 414 combinations.\n\nIn the example below we remove these 4 classes from the marker set, re-run the knnOptimisation for each data source and then run the knntlOptimisation with the 10 remaining classes. (Note: this is not run live as the hl dataset with 10 classes, 707 markers and 410 combinations of parameters takes around 76 hours to run on the University of Cambridge HPC using 256 workers).\n\nTo remove the 4 classes and create a new column of markers in the feature data called tlmarkers to use for the analysis:\n\n\n\nTL optimisation stage 1 Run knnOptimisation to get the best k’s for each data source.\n\n\n\nFrom examining the parameter seach plots as described in section Optimisation, we find the best k’s for both the primary and auxiliary are 3.\n\nTL optimisation stage 2 Run knntlOptimisation to get the best transfer learning weights for each sub-cellular class.\n\n\n\n\n\nThe results of the optimisation can be visualised using the plot method for tlopt optimisation result:\n\n\n\nLooking at the bubble plot displaying the distribution of best weights over the 50 runs we find that for many of the subcellular niches a weight of 1 is most popular (i.e. use only primary hyperLOPIT data in classification), this is unsuprising as we already know the dataset is well resolved for these classes. We see that the most popular weights for the proteasome and lysosome tend to be towards 0, indicating that these niches are well-resolved in the Gene Ontology. This tells us that we would benefit from including auxiliary GO information in our classifier for these subcellular compartments. The plasma membrane weights are relatively equally spread between using hyperLOPIT and GO data. Using the getParams function we can return the best weights and then use this as input for the classification.\n\nOne of the benefits of the algorithm is the ability to manually select weights for each class. In the optimisation above, for time constraints, we removed the two ribosomal subunits and the two nuclear compartments, and therefore in the code chunk below when we extract the best parameters, these subcellular niches are not included. To include these 4 subcellular niches in the next classification step we must include them in the parameters. We define a weight of 1 for each of these niches, as we know they are well resolved in hyperLOPIT. We then re-order the weights according to getMarkerClasses and perform the classification using the function knntlClassification.\n\n\n\nThe results from the classification results and associated scores are appended to the fData slot and named knntl and knntl.scores respectively. Results can be visualised using plot2D, scores assessed and cutoffs calculated using the classify app in pRolocVis, predictions obtained using getPredictions in the same way as demonstrated above for the SVM classifier.\n\nIn pRoloc’s transfer learning vignette, we demonstrate how to use imaging data from the Human Protein Atlas24 via the hpar package25 as an auxiliary data source.\n\n\nUnsupervised machine learning\n\nIn pRoloc there is functionality for unsupervsied machine learning methods. In unsupervised learning, the training data consists of a set of input vectors e.g. protein profiles, ignoring the information about the class label e.g. localisation, other than for annotation purposes. The main goal in unsupervised learning is to uncover groups of similar features within the data, termed clustering. Ordination methods such as PCA also fall into the category of unsupervised learning methods, where the data can be projected from a high-dimensional space down to two or three dimensions for the purpose of visualisation.\n\nAs described and demonstrated already above, PCA is a valuable and powerful method for data visualisation and quality control. Another application uses hierarchical clustering to summarise the relation between marker proteins using the mrkHClust function, where the euclidean distance between average class-specific profiles is used to produce a dendrogram describing a simple relationship between the sub-cellular classes (Figure 21). The mrkHClust uses the same defaults as all other function, using the markers feature variable to define marker proteins. In the code chunk, we adapt the figure margins to fully display the class names.\n\n\n\nWe generally find supervised learning more suited to the task of protein localisation prediction in which we use high-quality curated marker proteins to build a classifier, instead of using an entirely unsupervised approach to look for clusters and then look for enrichment of organelles and complexes. In the latter we do not make good use of valuable prior knowledge, and in our experience unsupervised clustering can be extremely difficult due to poor estimates of the number of clusters that may appear in the data.\n\nEach row displays the frequency of observed weights (along the columns) for a specific sub-cellular class, with large dots representing higher observation frequencies.\n\n\nWriting and exporting data\n\nAn MSnSet can be exported from R using the write.exprs function. This function writes the expression values to a text-based spreadsheet. The fcol argument can be used to specify which featureData columns (as column names, column number or logical) to append to the right of the expression matrix.\n\nIn the code chunk below we write the hl object to a csv file. The file argument is used to specify the file path, the sep argument specifies the field separator string, here we use a comma. Finally, as we want to write all the information in the featureData to the file, as well as the expression data, we specify fvarLabels(hl) that returns all feature variable names, and write the resulting data to the file \"hl.csv\".\n\n\n\nExporting to a spreadsheet however loses a lot of important information, such as the processing data, and the sample metadata in the phenoData slot. Other objects, such as parameters from the machine learning optimisation, cannot be represented as tabular data. To directly serialise R objects to disk, on can use the standard save function, and later reload the object using save. For example, to save and then re-load the parameters from the SVM optimisation,\n\n\n\n\nSession information\n\nThe function sessionInfo provides a summary of all packages and versions used to generate this document. This enables us to record the exact state of our session that lead to these results. Conversely, if the script stops working or if it returns different results, we are in a position to re-generate the original results using the adequate software versions and retrace changes in the software that lead to failure and/or different results.\n\n\n\nWe also recommend that users regularly update the packages as well as the R itself. This can be done with the biocLite function.\n\n\n\nIt is always important to include session information details along with a short reproducible example highlighting the problem or question at hand.\n\n\nConclusions\n\nThis workflow describes a step-by-step guide for analysing and interpreting contemporary spatial proteomics data. We described the software architecture, how to import and process data, quality control and visualisation as well as a range of machine learning methods available through the pRoloc suite of Bioconductor packages.\n\n\nData and software availability\n\nThe software and data presented in this workflow are part of the Bioconductor1 project. Version numbers for all packages used are shown in the Session information section.\n\nThe source of this document, including the code necessary to reproduce the analyses and figures is available at: https://github.com/lmsimp/bioc-pRoloc-hyperLOPIT-workflow/16. An archived version as at the time of publication is available at: DOI 10.5281/zenodo.19706826.",
"appendix": "Author contributions\n\n\n\nLMB and LG developed the software presented in this workflow. All authors wrote and approved the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were dislcosed.\n\n\nGrant information\n\nLMB and CMM are supported by a Wellcome Trust Technology Development Grant (grant number 108441/Z/15/Z). KSL is a Wellcome Trust Joint Investigator (110170/Z/15/Z). LG is supported by the BBSRC Strategic Longer and Larger grant (Award BB/L002817/1).\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 thank Dr. Stuart Rankin from the High Performance Computing Service for his support. Part of this work was performed using the Darwin Supercomputer of the University of Cambridge High Performance Computing Service, provided by Dell Inc. using Strategic Research Infrastructure Funding from the Higher Education Funding Council for England and funding from the Science and Technology Facilities Council.\n\n\nReferences\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with bioconductor. Nat Methods. 2015; 12(2): 115–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGatto L, Vizcaíno JA, Hermjakob H, et al.: Organelle proteomics experimental designs and analysis. Proteomics. 2010; 10(22): 3957–69. PubMed Abstract | Publisher Full Text\n\nChristoforou A, Mulvey CM, Breckels LM, et al.: A draft map of the mouse pluripotent stem cell spatial proteome. Nat Commun. 2016; 7: 8992. PubMed Abstract | Publisher Full Text | Free Full Text\n\nItzhak DN, Tyanova S, Cox J, et al.: Global, quantitative and dynamic mapping of protein subcellular localization. eLife. 2016; 5: pii: e16950. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJean Beltran PM, Mathias RA, Cristea IM: A Portrait of the Human Organelle Proteome In Space and Time During Cytomegalovirus Infection. Cell Syst. 2016; 3(4): 361–373.e6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGatto L, Christoforou A: Using R and Bioconductor for proteomics data analysis. Biochim Biophys Acta. 2014; 1844(1 Pt A): 42–51. PubMed Abstract | Publisher Full Text\n\nGatto L, Breckels LM, Naake T, et al.: Visualization of proteomics data using R and Bioconductor. Proteomics. 2015; 15(8): 1375–1389. ISSN 1615-9861. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThompson A, Schäfer J, Kuhn K, et al.: Tandem mass tags: a novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS. Anal Chem. 2003; 75(8): 1895–904. PubMed Abstract | Publisher Full Text\n\nMcAlister GC, Nusinow DP, Jedrychowski MP, et al.: Multinotch MS3 enables accurate, sensitive, and multiplexed detection of differential expression across cancer cell line proteomes. Anal Chem. 2014; 86(14): 7150–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrotter MW, Sadowski PG, Dunkley TP, et al.: Improved sub-cellular resolution via simultaneous analysis of organelle proteomics data across varied experimental conditions. Proteomics. 2010; 10(23): 4213–4219 ISSN 1615-9861. PubMed Abstract | Publisher Full Text\n\nWebb-Robertson BJ, Wiberg HK, Matzke MM, et al.: Review, evaluation, and discussion of the challenges of missing value imputation for mass spectrometry-based label-free global proteomics. J Proteome Res. 2015; 14(5): 1993–2001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLazar C, Gatto L, Ferro M, et al.: Accounting for the Multiple Natures of Missing Values in Label-Free Quantitative Proteomics Data Sets to Compare Imputation Strategies. J Proteome Res. 2016; 15(4): 1116–25. PubMed Abstract | Publisher Full Text\n\nGatto L, Breckels LM, Burger T, et al.: A foundation for reliable spatial proteomics data analysis. Mol Cell Proteomics. 2014; 13(8): 1937–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Duve C, Beaufay H: A short history of tissue fractionation. J Cell Biol. 1981; 91(3 Pt 2): 293s–299s. PubMed Abstract | 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): 25–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBreckels LM, Gatto L: A bioconductor workflow for processing and analysing spatial proteomics data. 2016. Reference Source\n\nBreckels LM, Gatto L, Christoforou A, et al.: The effect of organelle discovery upon sub-cellular protein localisation. J Proteomics. 2013; 88: 129–40. PubMed Abstract | Publisher Full Text\n\nFraley C, Raftery AE, Murphy TB, et al.: mclust Version 4 for R: Normal Mixture Modeling for Model-Based Clustering, Classification, and Density Estimation. 2012. Reference Source\n\nGroen AJ, Sancho-Andrés G, Breckels LM, et al.: Identification of trans-golgi network proteins in Arabidopsis thaliana root tissue. J Proteome Res. 2014; 13(2): 763–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall SL, Hester S, Griffin JL, et al.: The organelle proteome of the DT40 lymphocyte cell line. Mol Cell Proteomics. 2009; 8(6): 1295–1305. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDunkley TP, Hester S, Shadforth IP, et al.: Mapping the Arabidopsis organelle proteome. Proc Natl Acad Sci U S A. 2006; 103(17): 6518–6523. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTan DJ, Dvinge H, Christoforou A, et al.: Mapping organelle proteins and protein complexes in Drosophila melanogaster. J Proteome Res. 2009; 8(6): 2667–78. PubMed Abstract | Publisher Full Text\n\nBreckels LM, Holden SB, Wojnar D, et al.: Learning from Heterogeneous Data Sources: An Application in Spatial Proteomics. PLoS Comput Biol. 2016; 12(5): e1004920. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUhlen M, Oksvold P, Fagerberg L, et al.: Towards a knowledge-based Human Protein Atlas. Nat Biotechnol. 2010; 28(12): 1248–1250. ISSN 1546-1696. PubMed Abstract | Publisher Full Text\n\nGatto L: hpar: Human Protein Atlas in R. R package version 1.16.0. 2016. Reference Source\n\nGatto L, Breckels L: lmsimp/bioc-pRoloc-hyperLOPIT-workflow: Submission to F1000research [Data set]. Zenodo. 2016. Data Source"
}
|
[
{
"id": "18813",
"date": "11 Jan 2017",
"name": "Leonard J. Foster",
"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 a Bioconductor workflow for analyzing subcellular proteomics data. It is very detailed and comprehensive and will be useful for others in the field. A few comments:\nSome clearer statement early on would help to clarify for readers what types of data this works with. I know that the authors indicate that the example they use is 10-plex TMT and that it can be used with label-free or other labels, but that is not what I am referring to. Rather, structure of the experiment. That is, that one needs systematic quantitative data on all the different relevant fractions from a cell, as opposed to someone who perhaps did a differential centrifugation experiment to isolate a couple fractions and then wants to apply this (my understanding is that this latter example would not be usable).\n\nHow do the authors recommend collapsing replicates? This could be covered in the section dedicated to the Compare function. Two replicates will (almost) never agree 100% so how are discrepancies handled?",
"responses": [
{
"c_id": "3671",
"date": "03 Jul 2018",
"name": "Laurent Gatto",
"role": "Author Response",
"response": "This manuscript describes a Bioconductor workflow for analyzing subcellular proteomics data. It is very detailed and comprehensive and will be useful for others in the field. Many thanks for your comments, we have responded to them inset below. Some clearer statement early on would help to clarify for readers what types of data this works with. I know that the authors indicate that the example they use is 10-plex TMT and that it can be used with label-free or other labels, but that is not what I am referring to. Rather, structure of the experiment. That is, that one needs systematic quantitative data on all the different relevant fractions from a cell, as opposed to someone who perhaps did a differential centrifugation experiment to isolate a couple fractions and then wants to apply this (my understanding is that this latter example would not be usable). From a purely technical point of view, we need a data matrix with features (typically proteins or protein groups) along the rows and sub-cellular fractions along the columns. Given the requirement for complete (or near-complete) quantitative vectors along all fractions to assure best results, a data set that would only contain one quantitation value per fraction would not work. However, separation using differential centrifugation, or any separation that generates organelle-specific separation profiles is a good fit for pRoloc. For instance, the type of data generated by methods such as described in Itzhak et al. (2016) are well fitted for our software: > library(pRoloc) > library(pRolocdata) > data(itzhak2016stcSILAC) > ## 6 combined replicates of 5 fractions each > dim(itzhak2016stcSILAC) [1] 5265 30 > plot2D(itzhak2016stcSILAC) (see outout here) How do the authors recommend collapsing replicates? This could be covered in the section dedicated to the Compare function. Two replicates will (almost) never agree 100% so how are discrepancies handled? Currently, we recommend to visualise different replicates on their own, to confirm that they are of sufficient quality, and then combine them, retaining the proteins that have been quantified over all replicatedd experiments. This allows to obtain localisation information over all replicated data. We however do not explicitly assess the variability using this approach. This could be done by analysing replicated independently and then compare the coherence of the classification results. Proteins that are only observed in some replicates could be rescued by repeating the analysis using only the relevant (possibly unique) replicate(s)."
}
]
},
{
"id": "18815",
"date": "18 Jan 2017",
"name": "Daniel J. Stekhoven",
"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\nBreckels et al. have written a very nice piece on analysing appropriate proteomics data for subcellular localisation. I particularly like the \"workshop characteristics\" of the text. Which allows a novice, but interested reader to work through the analysis stepwise and reproduce the results described therein. The authors took great care in keeping this ideal up during their text and this is also where I have put my greatest reservation to the manuscript in its present form - since a reader cannot work through the code presented in the manuscript, since there at at least two situations where a readily available HPC and quite some time is required. This kind of leaves a dent in my impression - however, given this can be resolved as well as some typos - the workflow report is superb.\nMajor comments:\nNext to reducing the dimensions of data for visualisation, PCA also offers a way to understand how the variability is distributed across the multidimensional data by providing linear combinations of the variables which then constitute the actual PCs. On that note it would be nice to mention this in Visualising markers section on page 16, where PC7 explains not much variability but due to the correct weighing of the variables we do get a separation between mitochondrial and peroxisome. This then can be further motivated with Figure 9 - where we probably can see that the weights for the fractions where the two localisations differ are larger than otherwise.\n\nI was unable to reproduce Figure 13 comparing the two MSnSets. While I was able to look at each set separately using pRolocVis(hllst@x[[I]]), where i is 1 or 2, I only got an error using the code from the manuscript:\n> pRolocVis(hllst, app=“compare”) Subsetting MSnSetList to their common feature names 5032 features in common Remapping data to the same PC space Error in (function (od, vd) :\n\nobject and replacement value dimnames differ Error in pRolocVis_compare(object, ...) : object 'idDT' not found\nWhen using ‘remap=FALSE’ it actually works, but since this makes barely sense it is of no use - but just as a hint at debugging it.\n\nYou really need to make the results from the phenoDisco classification available too. It is super disappointing that one cannot continue reproducing the code from page 23 on, because it takes 24 hours to compute it using 40 cores…\n\nThe above comment is of course also true for the KNN TL Optimisation on page 33 - this needs to be downloadable, since not everyone has access to Cambridge’s HPC and probably even less have 76 hours to spare.\n\nYour comment on the increase suitability of classification instead of clustering (when additional information on classes is available) at the bottom of page 35 could be more pronounced - for educational reasons.\n\nMinor comments:\nI was not able to naïvely reproduce the workflow from the R commands in the article due to an error installing pRolocdata on a Windows machine. On OS X it was smooth.\n\nOn page 10 line 2 there is a ‘to’ missing.\n\nI never came across the verb imputate in the context of missing values, I guess the proper term is impute.\n\nOn page 11 the image2 function is called after the filterNA function a couple of lines above. This however would result in an only black heat map (since there are no more missing). The image2 function should be called before the filterNA function. Since the reader does not see the chunk options, it could be puzzling.\n\nFor completeness sake there should also be an install.packages(c(“hexbin”, “rgl”)) somewhere to generate the second PCA-plot and the 3D plot. Moreover, Mac users will need to install xquartz to use rgl properly.\n\nOn page 14 the plotting code chunk is off track - in the middle of the marker sets output.\n\nOn page 18: …wanted to highlight a proteins with the… -> lose the a and later in the sentence there is a ‘create a’ too many.\n\nDirect comparisons of individual channels in replicated experiments do not provide…\n\nYou may want to consider adding a layout(1) or similar, after changing the mfrow argument of the parameters to accommodate 2 panels, such that the uncanny reader does not get confused.\n\nI would prefer links to referred sections of the text, but that may be personal taste…\n\nPage 23: One should note that the decreasing the GS, and increasing the … at least one the too many, probably two.\n\nOn page 25: We find the general tendancy to be that it is not the choice … tendency?\n\nOn page 28 you refer to ‘…the code chunk below…’ for Figure 17, however, the following code chunk is generating Figure 16 (which is above and btw not referenced in the text). Maybe force your figures a little to float where you want them/refer to them.\n\nOn page 28: …by extracting the median or 3rd quantile score per organelle… do you mean quartile? Otherwise I do not follow.\n\nOn page 32: …package to query the relevent database … relevant?\n\nOn page 32 - there is something wrong with this sentence: To remove the 4 classes and create a new column of markers in the feature data called tlmarkers to use for the analysis:\n\nOn page 34: From examining the parameter seach plots as described in section Optimisation… search!\n\nOn page 36: …and later reload the object using save. -> that would be ‘load’ then!\n\nOn page 38 - I fully agree with the following sentence, but right after the updating comment it kind of seems ‘misplaced’? Maybe add a title like ‘Getting help’?\nIt is always important to include session information details along with a short reproducible example highlighting the problem or question at hand.",
"responses": [
{
"c_id": "3672",
"date": "03 Jul 2018",
"name": "Laurent Gatto",
"role": "Author Response",
"response": "Thank you for your comments. Please find our responses to these inset below. Next to reducing the dimensions of data for visualisation, PCA also offers a way to understand how the variability is distributed across the multidimensional data by providing linear combinations of the variables which then constitute the actual PCs. On that note it would be nice to mention this in Visualising markers section on page 16, where PC7 explains not much variability but due to the correct weighing of the variables we do get a separation between mitochondrial and peroxisome. This then can be further motivated with Figure 9 - where we probably can see that the weights for the fractions where the two localisations differ are larger than otherwise. We have added a paragraph to the 'Visualising markers' section of the manuscript reiterating the purpose of PCA and motivating the choice of looking at PC's 1 and 7. Figure 9 now follows on from this (now Figure 8), along with the corresponding code and an explanation of the `plotDist` function. I was unable to reproduce Figure 13 comparing the two MSnSets. While I was able to look at each set separately using `pRolocVis(hllst@x[[I]])`, where i is 1 or 2, I only got an error using the code from the manuscript. When using `remap=FALSE` it actually works, but since this makes barely sense it is of no use - but just as a hint at debugging it. We can not reproduce this error. Have you updated to the latest version of R and the latest version of `pRolocGUI`? If you still get this error message could you please post this as an issue (https://github.com/ComputationalProteomicsUnit/pRolocGUI/issues) on the `pRolocGUI` Github page along with your `sessionInfo()` and we will certainly attempt to solve this. You really need to make the results from the phenoDisco classification available too. It is super disappointing that one cannot continue reproducing the code from page 23 on, because it takes 24 hours to compute it using 40 cores? The results are already available as a `RDS` file and stored in `pRolocdata` for users. This is what is called in the manuscript under the hood: ``` f0 <- dir(extdatadir, full.names = TRUE, pattern = \"bpw-pdres.rds\") pdres <- readRDS(f0) hl <- addMarkers(hl, pdres, mcol = \"pd\", verbose = FALSE) ``` We have made this code available in the manuscript in an appendix so users can continue to produce the exact plots as they see in this workflow. The above comment is of course also true for the KNN TL Optimisation on page 33 - this needs to be downloadable, since not everyone has access to Cambridge's HPC and probably even less have 76 hours to spare. The same as for the `phenoDisco` analysis and `svm`, the TL results are stored as a `RDS` in `pRolocdata` and are loaded in the background. We have added the code required to the appendix so that users can load the results directly. Your comment on the increase suitability of classification instead of clustering (when additional information on classes is available) at the bottom of page 35 could be more pronounced - for educational reasons. To address the above comment on suitability we have added a few additional points on the challenges of using clustering for this type of data. We generally find supervised learning more suited to the task of protein localisation prediction in which we use high-quality curated marker proteins to build a classifier, instead of using an entirely unsupervised approach to look for clusters and then look for enrichment of organelles and complexes. In the latter we do not make good use of valuable prior knowledge, and in our experience unsupervised clustering can be extremely difficult due to (i) the loose definition of what constitutes a cluster (for example whether it is defined by the quantitative data or the localisation information), (ii) the influence of the algorithm assumption on the cluster identification (for example parametric or non-parametric) and (iii) poor estimates of the number of clusters that may appear in the data. I was not able to naively reproduce the workflow from the R commands in the article due to an error installing pRolocdata on a Windows machine. On OS X it was smooth. We didn't experience any Windows-specific problems. If you re-try the installation and please let us know if you still have any issues by opening an issue (https://github.com/lgatto/pRolocdata/issues) or by posting this issue on the Bioconductor Support site (https://support.bioconductor.org). On page 10 line 2 there is a *to* missing. In the version have, we currently can't find the missing 'to'. I never came across the verb imputate in the context of missing values, I guess the proper term is impute. This has been changed to read \"We can impute missing data...\" On page 11 the image2 function is called after the filterNA function a couple of lines above. This however would result in an only black heat map (since there are no more missing). The image2 function should be called before the filterNA function. Since the reader does not see the chunk options, it could be puzzling. This was an editing mistake and has now been rectified. For completeness sake there should also be an install.packages(c(\"hexbin\", \"rgl\")) somewhere to generate the second PCA-plot and the 3D plot. Moreover, Mac users will need to install xquartz to use rgl properly. A footnote has been added here to tell users that the package `rgl` may need to be installed with `install.packages(\"rgl\")` and mac users may need to install xquartz if it's not already installed. LG: this one is still missing. On page 14 the plotting code chunk is off track - in the middle of the marker sets output. This has now been rectified. On page 18: ...wanted to highlight a proteins with the ... -> lose the a and later in the sentence there is a 'create a' too many. These typos have been rectified. Direct comparisons of individual channels in replicated experiments do not provide? You may want to consider adding a layout(1) or similar, after changing the mfrow argument of the parameters to accommodate 2 panels, such that the uncanny reader does not get confused. We would prefer to keep the code as it is and not introduce more noise with calls to other functions such as `layout`. The workflow is not aimed at teaching R. Users should have some basic knowledge of R before tackling this tutorial. I would prefer links to referred sections of the text, but that may be personal taste? This is a comment for F1000. We cannot control the linking of sections in the final version. Page 23: One should note that the decreasing the GS, and increasing the ...at least one the too many, probably two. We have reworded this sentence as requested. On page 25: We find the general tendancy to be that it is not the choice ...tendency? This typo has been rectified. On page 28 you refer to '...the code chunk below...' for Figure 17, however, the following code chunk is generating Figure 16 (which is above and btw not referenced in the text). Maybe force your figures a little to float where you want them/refer to them. We have now referenced Figure 16 in the text and made sure that the code chunks and figures follow inline where they are referenced in the text. On page 28: ...by extracting the median or 3rd quantile score per organelle? do you mean quartile? Otherwise I do not follow. Thank you, yes this is a typo and has now been changed to 'quartile'. On page 32: package to query the relevent database *relevant* This typo has been rectified. On page 32 - there is something wrong with this sentence: To remove the 4 classes and create a new column of markers in the feature data called tlmarkers to use for the analysis: This sentence is not needed here and so it has now been removed as it essentially reiterates what is said in the above paragraph. On page 34: From examining the parameter seach plots as described in section Optimisation... search! This typo has been rectified. On page 36: and later reload the object using save. -> that would be `load` then! This typo has been rectified. On page 38 - I fully agree with the following sentence, but right after the updating comment it kind of seems misplaced? Maybe add a title like Getting help? We have changed the title of this section to 'Session information and getting help' to clarify this section of the tutorial. A pdf version of our replies is also available here."
}
]
}
] | 1
|
https://f1000research.com/articles/5-2926
|
https://f1000research.com/articles/7-77/v1
|
17 Jan 18
|
{
"type": "Systematic Review",
"title": "A systematic review and critical analysis of cost-effectiveness studies for coronary artery disease treatment",
"authors": [
"Victoria McCreanor",
"Nicholas Graves",
"Adrian G Barnett",
"Will Parsonage",
"Gregory Merlo",
"Nicholas Graves",
"Adrian G Barnett",
"Will Parsonage",
"Gregory Merlo"
],
"abstract": "Background: Cardiovascular disease remains the primary cause of death among Australians, despite dramatic improvements in overall cardiovascular health since the 1980s. Treating cardiovascular disease continues to place a significant economic strain on the Australian health care system, with direct healthcare costs exceeding those of any other disease. Coronary artery disease accounts for nearly one third of these costs and spending continues to rise. A range of treatments is available for coronary artery disease yet evidence of cost-effectiveness is missing, particularly for the Australian context. Cost-effectiveness evidence can signal waste and inefficiency and so is essential for an efficient allocation of healthcare resources. Methods: We used systematic review methods to search the literature across several electronic databases for economic evaluations of treatments for coronary artery disease. We critically appraised the literature found in searches, both against the CHEERS statement for quality reporting of economic evaluations and in terms of its usefulness for policy and decision-makers. Results: We retrieved a total of 308 references, 229 once duplicates were removed. Of these, 26 were excluded as they were not full papers (letters, editorials etc.), 55 were review papers, 50 were not cost-effectiveness analyses and 93 related to a highly specific patient sub-group or did not consider all treatment options. This left five papers to be reviewed in full. Conclusions: The current cost-effectiveness evidence does not support the increased use of PCI that has been seen in Australia and internationally. Due to problems with accessibility, clarity and relevance to policy and decision-makers, some otherwise very scientifically rigorous analyses have failed to generate any policy changes.",
"keywords": [
"Coronary artery disease",
"cost-effectiveness analysis",
"economic analysis",
"review",
"health policy",
"health services research"
],
"content": "Introduction\n\nCardiovascular disease is the primary cause of death for Australians and places enormous strain on the health care system. Treatments for cardiovascular disease consume 12% of Australian health care spending, AUD $7 billion annually, with coronary artery disease responsible for 27% of the cost1. Australian health services continue to increase spending in this area, with cardiovascular disease treatment costs doubling between 2000–01 and 2008–091. This increase in spending is occurring despite improvements in the cardiovascular health of Australians, resulting from improved lifestyle factors, most importantly reduced rates of tobacco smoking2.\n\nThere have been changes in the preferences for different treatments. Since 1998, percutaneous coronary intervention (PCI) has overtaken coronary artery bypass graft (CABG) as the most common revascularisation procedure in Australia3. Between 2000–01 and 2007–08 the number of PCIs performed increased by 57%2. In 2012–13, 93% of PCIs involved the insertion of one or more stents4. Accompanying the increase in PCIs, there was a 19% reduction in the number of CABGs performed5. Since then, rates of PCI have remained high4. The increase in PCI also suggests that more patients who in the past would have been treated conservatively, with medical therapy only, are now also undergoing PCI.\n\nInvasive treatments for coronary artery disease are costly, involving surgery, expensive equipment and consumables, yet there is no adequate assessment of the cost-effectiveness of the treatments provided. An important question is whether the extra costs incurred are adequately compensated by gains to health. In an era of non-increasing health budgets, changes to practice should be accompanied by improvements in health outcomes, particularly if increased costs are involved. Cost-effectiveness evidence should be used to assess whether changes in costs are justified by changes to health outcomes associated with new services or changes in practice.\n\nIn the case of treatment for coronary artery disease, it is not clear if there is sufficient evidence to support the recent changes in treatment preferences from CABG to PCI, or a move from medical therapy alone to more invasive treatment such as PCI. While many economic evaluations have been undertaken, much of the literature assessing the cost-effectiveness of coronary artery disease treatments compares only two options at a time, with a large focus on the differences between drug-eluting stents (DES) and bare metal stents (BMS). However, comparing only two treatments at a time is limited. It assumes the chosen baseline comparator is a good quality service, and omits other available treatment options. It is sub-optimal for high-level budgetary decisions to be made without more comprehensive information about all competing treatment choices. An analysis comparing BMS only with DES fails to consider treatments other than stents and may therefore overestimate the cost-effectiveness of one type of stent, over other treatments.\n\nClinical trial evidence has failed to show a mortality benefit of PCI over medical therapy in the treatment of stable disease, but there is some evidence of greater symptom relief6–9. Coronary artery bypass graft surgery on the other hand has been shown to provide mortality benefit in some circumstances, and more prolonged symptom relief compared with PCI8,10. However, PCI is an expensive procedure, and CABG even more so. The question, therefore, is whether the additional costs are sufficiently offset by the greater symptom relief afforded by PCI and mortality benefit and symptom relief of CABG, when compared with medical therapy alone.\n\nThe aims of this review are to evaluate the literature that describes the cost-effectiveness of all treatment options for coronary artery disease: PCI including stent insertion, CABG and medical therapy, and then to critique the literature based on the quality of the cost-effectiveness evaluations and usefulness of the findings of the research for real-world applications. Usefulness for real-world applications includes the applicability of the outcomes for informing decisions about the allocation of resources, particularly in the Australian context, and the ability to translate the findings into practice.\n\nPrior to undertaking any new research, it is important to undertake a review of the literature, to reduce the chance of duplication of effort, and to avoid tackling questions that have already been answered11. This review is designed to identify gaps in the knowledge about the cost-effectiveness of treatments for coronary artery disease, and therefore to inform future research in this area. Our goal is to provide insights useful for clinicians, healthcare service budget holders, and policy-makers about the best use of scarce resources for the treatment of coronary artery disease.\n\n\nMethods\n\nThe literature published between January 1995 (after the use of stents was approved in the United States in 1994) and May 2017 was searched in PubMed, Embase, Scopus, CINAHL (via Ebscohost) and EconLit (via ProQuest). The searches focussed on extracting papers that examined the cost-effectiveness of PCI (including stent insertion), CABG and medical therapy together. Due to the large volume of research on coronary artery disease, searches were limited to subject headings where possible. A slightly broader approach was taken to capture medical therapy as this is described less consistently in the literature. Only research published in English was included. The search terms used are in Box 1. There is no specific MeSH for cost-effectiveness analysis and the suggested heading is Cost-benefit Analysis. Search terms were modified slightly to fit the subject heading structure of each database (See Supplementary File 1).\n\n(percutaneous coronary intervention[MeSH Terms] OR stents[MeSH Terms]) AND coronary artery bypass[MeSH Terms] AND (Cost-benefit Analysis[MeSH Terms] OR Models, economic[MeSH Terms]) AND (((medical OR conservative) AND (therapy OR treatment)) OR primary prevention OR secondary prevention)\n\nSearch results were imported into EndNote X7 software, duplicates were removed and articles then reviewed according to the inclusion and exclusion criteria outlined in Table 1.\n\nWe used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement12 and checklist in retrieving and reviewing articles for inclusion in this review. Due to the nature of this review, not all items were relevant. Our completed checklist is available in Supplementary File 2. Titles of all papers were reviewed and the abstract or full text examined in detail where required to assess against inclusion and exclusion criteria.\n\nWe used the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) Statement13 to assess the quality of reporting of the economic analyses (Supplementary File 3). The CHEERS Statement and checklist was developed to improve and promote quality of reporting health economic evaluations14. Economic evaluations are designed to assist in health service decision-making and resource allocation. Therefore, due to the opportunity costs of acting on poor-quality evidence, it is particularly important to ensure high-quality reporting in economic evaluations14. The CHEERS Statement checklist consists of 24 items which should be reported and guidance regarding the specific details required. We extracted information from each of the included studies for each item on the checklist, to assess the quality of reporting.\n\nIn addition to high-quality reporting, economic evaluations need to be useful to decision-makers, as their purpose is to provide evidence to improve the efficiency of use of healthcare resources. Decision-makers need to understand the potential impact of acting on cost-effectiveness evidence and making changes to healthcare services; most notably what will be the effect on health outcomes and costs, and how certain are these projections? To assess the usefulness of the evaluations for decision-makers, we also extracted data on the interventions compared, the effectiveness measures, whether the analysis applied to a specific patient group, the structure or type of analysis used and the time period of the analysis. We then assessed usefulness of the reporting by rating the reporting of outcomes, costs, uncertainty as: useful, not useful, or partly useful, based on whether the results could be used by a decision-maker. We also looked for a clear statement about the policy implications or direction that should follow based on the outcomes, and gave each paper an overall usefulness rating of low, medium or high, depending on the other elements assessed. We acknowledge that these ratings are subjective and have not been validated, nevertheless we think they are practically useful.\n\n\nResults\n\nSearches in all databases, except EconLit, revealed potentially relevant articles. A total of 308 results were retrieved, and 229 remained after duplicates were removed. The numbers of papers retrieved from each database are in Figure 1.\n\nMany articles tagged under the cost-benefit or cost-effectiveness subject headings were not cost-effectiveness analyses but simply mentioned cost-effectiveness as a factor for consideration. In addition, most of the cost-effectiveness papers did not examine medical therapy, percutaneous coronary intervention and bypass graft surgery, but focussed on only two treatments. Papers were excluded for other reasons including a focus on rehabilitation following cardiac procedures, screening of cardiac patients, being commentary only or reviews. The results of the review process are shown in Figure 1.\n\nThe results of the review process left only five papers for consideration. Table 2 provides a summary of the papers included in the full review. The results varied across the five studies, but across the scenarios analysed most concluded that medical therapy was the cost-effective treatment, and three concluded that CABG was cost-effective. In no scenario was PCI reported as being cost-effective compared with the alternatives (Table 2). Three studies included quality of life measures in at least part of their analyses and only two studies undertook projections over the lifetime of patients (Table 2).\n\nMost papers reported more than one timeframe. Shading indicates treatment reported as cost-effective.\n\n* CABG = coronary artery bypass grafting, LY = life years, MI = myocardial infarction, OMT = optimal medical therapy, PCI = percutaneous coronary intervention, QALY = quality-adjusted life-year.\n\n* *N.B. Fidan et al.16 paper did not make direct comparisons of the relevant treatments\n\nA summary of the assessment against the CHEERS Statement are shown in Table 4. Overall, the quality of reporting was high, with studies adequately reporting against 50 to 100% of relevant items on the checklist (Table 3). A more detailed table is available in Supplementary File 3.\n\nDue to their nature, not all CHEERS items were relevant to all studies.\n\n* ACRE = Appropriateness Coronary REvascularisation, BMS = bare metal stent, CABG = coronary artery bypass grafting, DES = drug-eluting stent, LY = life years, MI = myocardial infarction, OMT = optimal medical therapy, PCI = percutaneous coronary intervention, PTCA = percutaneous transluminal coronary angioplasty, QALY = quality-adjusted life-year, RCT = randomised controlled trial.\n\nAll studies adequately reported on the CHEERS Statement items relating: to model/analysis description, background and reasons for undertaking economic evaluations, relevant patient groups and sub-groups, comparators, time horizons and choice of health outcomes included in their analyses. The most poorly reported element related to reasons for choice of model (part of item 15). Only three studies did this. The others described the analysis undertaken, but did not give reasons for the chosen strategy.\n\nWhile for the most part the studies reported their analyses and findings to a good standard when assessed against the CHEERS Statement, their usefulness to decision-makers is arguably of greater importance. The summary data extracted in relation to usefulness of each paper is shown in Table 4. Our assessment of the usefulness of the reporting for decision-makers is in Table 5.\n\nOf the five papers reviewed in full, three were trial-based analyses, one used a cohort modelling approach and there was one meta-analysis (Table 4). The timeframes analysed ranged from 1 year post-intervention to a lifetime horizon. The studies came from a wide range of countries. Three of the five studies used quality of life measures in their analyses, one of which only considered QALYs as part of sensitivity analyses. The others used length of life measures to assess cost-effectiveness and one also used the clinical endpoint of myocardial infarction.\n\nIn assessing the usefulness of reporting, we found that while most studies reported on various items, reporting was not always easy to interpret in the context of decision-making. In judging the reporting we were looking for a clear direction or suggestion about how the results of the analysis could be used to improve the efficiency of healthcare resource use. Only two studies made a clear statement about changing the allocation of resources or how the outcomes are relevant to policy16,17. We rated two studies as low usefulness, two medium and only one highly useful for decision-makers.\n\nCaruba et al.15 carried out a meta-analysis of cost-effectiveness studies. After concluding that there was no statistically significant difference between treatment strategies on clinical endpoints of myocardial infarction or death, the analysis was conducted on costs only, over 1 and 3 years15. As a result, the analysis focusses primarily on cost differences across the treatments. They estimated that substantial cost savings could be made through the management of patients with stable angina, using medical therapy15.\n\nA more detailed examination of the outcomes reported by Caruba et al.15 revealed that while no statistically significant differences were found, there appears to be some clinically significant difference in treatment effectiveness. The confidence intervals of hazard ratios reported for both death and myocardial infarction are very wide. For example, at three years follow-up, confidence intervals related to estimates of risk of death range from a halving to a doubling of risk for all comparator treatments15. Similarly, the probabilities of being the best treatment vary widely; from 0.49 for drug-eluting stents to 0.05 for percutaneous transluminal coronary angioplasty (for risk of death at three years follow-up)15. These results suggest both that there is a high degree of uncertainty regarding estimates of effectiveness, and therefore that a clinically significant difference between the treatments is possible. This could greatly affect estimates of cost-effectiveness. In addition to uncertainty regarding estimates of effectiveness, the authors highlight important limitations and high levels of uncertainty in the overall findings.\n\nWe rated this meta-analysis as of low use to decision-makers due to the level of uncertainty described, making it difficult to interpret how the findings might be used to direct policy or practice. While a high level of uncertainty in results should not disqualify an analysis from being useful, the authors did not make any statements which assist in determining how the findings might be used. In addition, the difference in effectiveness of treatments was not fully explored, adding even more uncertainty to the findings.\n\nAn analysis by Fidan et al.16 modelled the life years gained for 36 condition–treatment scenarios for coronary artery disease. These included everything from acute myocardial infarction to primary prevention using statins16. They used the IMPACT model; a large cell-based mortality model of coronary heart disease risk and treatment16,20. Cost-effectiveness ratios were reported, however these are not presented as incrementally, which makes it difficult to do head-to-head treatment comparisons. All treatments were examined against the baseline mortality rates, and they found that medical and surgical treatments prevented or postponed over 25,000 deaths in patients with coronary artery disease16. The approach ranked different interventions, showing a 100-fold difference in cost-effectiveness across all treatments, but it does not provide insight into incremental costs associated with new technology or interventions16. Again, the analysis only considers length of life, not quality of life. While this provides useful information about length of life it does not consider the full effect of different treatments. Included in this assessment are treatments for chronic angina. It is a somewhat counterintuitive approach to only examine length of life gains from a treatment that targets symptom relief.\n\nWhile the reporting in this study is clear, we rated it as of medium usefulness for decision-makers because it does not include an assessment of the full effect of treatments on health outcomes (i.e. it only assessed length of life). The authors did, however, present a general policy suggestion, stating that investment in secondary prevention was likely to produce gains in length of life for lower costs.\n\nPrior to the analysis of the MASS-II trial, there had been no cost-effectiveness analysis based on a trial comparing percutaneous intervention, surgery and medical therapy together19. To address this gap, Griffin et al.17 conducted an economic analysis using the Appropriateness of Coronary REvascularisation (ACRE) study cohort. The ACRE study rated patients as appropriate for percutaneous coronary intervention and/or coronary artery bypass grafting but followed them according to the treatment they actually received17. Economic analysis of the ACRE study data concluded that coronary artery bypass grafting was cost-effective compared with percutaneous coronary intervention in patients classified as appropriate for bypass grafting only or for both bypass grafting and percutaneous intervention17. The analysis also found that percutaneous coronary intervention was not cost-effective when compared with medical therapy for patients classified as appropriate for percutaneous coronary intervention only17. The results of this analysis are useful because they include quality of life outcomes. However, the approach used averaged quality of life over the 6-year period using a regression model18. This gives some good information about the average quality of life of patients receiving different treatments over the time period, but does not account for events during which patients might experience reduced quality of life, such as a period of hospitalisation for a subsequent procedure.\n\nWe rated this analysis as of high usefulness to decision-makers. While the estimates of quality of life could be improved, the authors make clear statements about the changes in costs and health outcomes achieved through different treatments. They also make a clear statement of how to make changes to resource allocation based on their results, which could benefit the health service. One limitation to the usefulness of the outcomes presented is that they only cover the six-year trial period. This may not be long enough to see the full cost-effectiveness of treatment for a chronic disease and the authors foresee extending the model over a lifetime horizon in future work17.\n\nThe analysis by Hlatky et al. in 2009 examined the cost-effectiveness of revascularisation procedures in patients with type-2 diabetes, using data from the Bypass Angioplasty Revascularization Investigation 2 Diabetes trial (BARI 2D)18. The BARI 2D study randomised patients with type 2 diabetes to medical therapy alone or medical therapy with immediate revascularisation (either PCI or CABG)21. While the effectiveness of treatment for coronary artery disease has been shown to be affected by the presence of diabetes22–25, due to the high prevalence of type 2 diabetes in this patient population and more generally, this analysis was not excluded on grounds of being relevant only to a specific group. The rates of diabetes in other included studies are 36% in MASS II19, 15% in the ACRE study17, and 9 to 33% in the studies included in the meta-analysis by Caruba et al.15. The economic evaluation of the BARI 2D study outcomes concluded that medical therapy was cost-effective compared with revascularisation (PCI or CABG), in the short-term (4 years)18. When using lifetime projections of cost-effectiveness, however, medical therapy was cost-effective compared with PCI, and CABG was cost-effective compared with medical therapy18.\n\nThe BARI 2D trial used a pragmatic approach which reflects the realities of clinical practice; patients undergoing revascularisation were not randomised to a particular revascularisation strategy (i.e. PCI or CABG); this was directed by clinicians ahead of randomisation to either prompt revascularisation or medical therapy21. The effect of this is that patients were stratified into groups based on clinical markers of disease severity. The results of the study are therefore useful for choosing between medical therapy and PCI in patients with less severe disease, or between medical therapy and CABG in patients with severe disease. They are also only relevant to diabetic patients, however, as prevalence of type 2 diabetes is increasing globally, this is relevant to an increasing number of patients.\n\nThe overall results in the BARI 2D trial are based on length of life measures; quality of life measures were only used in sensitivity analyses. It was concluded that the quality of life measures did not affect the estimates of cost-effectiveness, based on life-years only. It is unclear why this choice was made, when quality of life measures provide a more comprehensive assessment of treatment effect. We rated this study as of medium use for decision-makers as it presents an analysis of real-world practice, but does not account for the full effect of treatment on patients’ health, by all but ignoring quality of life measures. Decision-makers wishing to know the full effect of different treatments on patient health outcomes need information beyond length of life.\n\nVieira et al. also conducted a trial-based analysis using data from the MASS II Trial (Medical Angioplasty or Surgery Study)19. This was the only trial revealed in searches which randomised patients to each of the three treatment options. Its major conclusions were that medical therapy was cost-effective compared to CABG, and CABG was cost-effective compared to PCI19. While this analysis did use QALYs they were not calculated using conventional health related quality of life surveys, but estimated based on the average time to event and angina free proportion of the population in each group19. This is unlikely to provide good estimates of quality of life in these patients as the measurement assumes that in the period between events the patient has full quality of life and that those with angina have no quality of life. These estimates produced average QALYs of 2.07 to 2.81, over 5 years which if averaged over that time give utility weights of 0.41 to 0.56 (e.g. 2.07/5). These values are far below the estimates of 0.69 to 0.86 used in other analyses of patients with coronary artery disease26–30. Values of less than 0.5 are generally seen only in very debilitating conditions. The quality of life estimates in the MASS II study analysis therefore substantially undervalue the quality of life of patients with coronary artery disease. While the outcomes of that analysis do examine both costs and effectiveness of the three different treatments for coronary artery disease, the outcomes reported are not useful for those making decisions about resource allocation because they do not allow comparison with other areas of healthcare or report the incremental cost per QALY gained.\n\nWe rated the analysis by Vieira et al.19 as of low usefulness to decision-makers because although quality of life was included in the analysis, it was not done in a way that makes it comparable to other studies. In addition to these novel methods of QALY estimation, the authors did not conduct incremental analyses, nor did they discuss any uncertainty in their findings.\n\nIt is worth noting that further research has been undertaken using the MASS II trial data and a validated quality of life instrument.31 Unfortunately, only a conference abstract was available and it was therefore not included in the analysis. The results available in that abstract show much higher average health utility weights of 0.77 to 0.8131, aligning them with the values seen in other analyses of coronary artery disease26–30. When published, the full analysis will add greatly to the current knowledge.\n\n\nDiscussion\n\nWhen operating under conditions of scarce resources there is responsibility to promote cost-effective care, achieving larger health gains from available resources. In an ideal world, decision-makers would have information about the long-term costs and health outcomes achievable through different configurations of health services and be able to invest accordingly. However, without good evidence of cost-effectiveness, it is impossible for decision-makers to fulfil this responsibility with any confidence.\n\nIn the case of coronary artery disease, we have some information about the comparative cost-effectiveness of optimal medical therapy, PCI and CABG, but it is difficult to interpret in the context of healthcare resource allocation. Overall, the results of cost-effectiveness analyses suggest that in most scenarios optimal medical therapy is cost-effective compared with alternatives and CABG is cost-effective for certain patient groups.\n\nHowever, only three of the five studies included in this review used quality of life as an effectiveness measure. This is a key outcome measure for cost-effectiveness and good decision making. For chronic diseases, improvements in quality of life are equally as relevant as improvements in length of life. In the case of coronary artery disease, relief from chest-pain is a key objective of treatment. If quality of life is not measured, two treatments affording a patient equal length of life are valued equally even where one restored the patient to better health than the other. However, if given the choice, patients and health service providers would choose the option most likely to provide the best improvements to quality of life. Therefore, analyses based solely on length of life measures do not provide a full picture of the effectiveness of each treatment.\n\nAnother omission from the literature is to neglect the lifetime costs and health outcomes. Important information about the longevity of treatment effect may be overlooked. This is particularly important for chronic diseases such as coronary artery disease, where important costs and health consequences are missed when they occur beyond the timeframe of a clinical trial.\n\nThe current information about cost-effectiveness of treatments for coronary artery disease suggests that either optimal medical therapy or CABG could be cost-effective, over a 1-year to lifetime timeframe. There is no evidence from the papers included in this review that PCI is cost-effective when compared with other competing treatment options. Therefore the current cost-effectiveness evidence does not support the increased use of PCI that has been seen in Australia and internationally, and there is increasingly reduced evidence of clinical effectiveness32.\n\nHowever, it is unlikely that healthcare decision-makers would be confident making changes to the allocation of resources based on the economic evidence outlined in this review. Our evaluation of the relevant cost-effectiveness evidence showed that overall, information is not presented in a way useful for decision-making. We found only one study to be of high usefulness in this context (Table 5).\n\nOur assessment of the usefulness of the cost-effectiveness studies examined suggests that poor reporting may contribute to the problem. We rated two out of the five studies as ‘low usefulness’ for decision-makers. Reporting was either too complex, making interpretation challenging, or uncertainty was not reported in way that made clear the effect of acting on the evidence. It is also apparent from our analysis, that the CHEERS Statement, while encouraging comprehensive reporting, is not sufficient alone, to assess the usefulness of economic evaluations.\n\nOthers have explored barriers to the use of economic evaluation by decision-makers33. A review by Merlo et al.33 used an accessibility and acceptability framework developed by Williams and Bryan34 to categorise barriers to use of economic evaluations. Accessibility refers to the ability of decision-makers to interpret and use economic evidence, and includes issues of complexity and timeliness of economic evaluations33,34. Acceptability includes factors associated with scientific rigor, applicability to the institution in which decisions are to be made, and ethical considerations such as equity33,34.\n\nWhile some studies we examined included comprehensive reporting, they did not always present results in a way conducive to decision-making. For example reporting large results tables covering many different clinical scenarios, as seen in the paper by Fidan et al.16, demonstrates the complexity but does nothing to assist those wanting to make higher-level resource allocation decisions. In fact, presenting all clinical complexities can mean the overall message is lost in the details, making research less accessible to decision-makers and decreasing the chance that any improvement to health services will follow. This is unfortunate, considering the purpose of economic evaluation is to provide evidence for resource allocation decisions to improve service delivery.\n\nMost of the economic evaluations we have assessed suffer from a number of accessibility problems which prevent them from being useful for decision-making; the interpretation of results and their applicability to policy are generally lacking. The information in Table 5 reveals that only one study, by Griffin et al.17, included a clear expression of the confidence in estimates of cost-effectiveness, useful in the context of decision-making. Hlatky et al.18 made a less-clear statement. Only two studies, Griffin et al.17 and Fidan et al.16, made clear statements of what direction policy should take based on results of their research (Htlaty et al.18 and Vieira et al.19 expressed a direction but less-clearly). The result of this highly complex reporting and lack of clear policy direction to follow, is that some otherwise very scientifically rigorous analyses have failed to generate any policy changes or perhaps even reach their intended audience.\n\nIn addition to these complexities, while the patient populations may be similar to those in Australia, none of these studies have been carried out in the Australian context. This means that from an Australian perspective, it is even more unlikely that resource allocation decisions would be able to be made on the basis of the evidence in this review.\n\nThe evaluation of the studies in this review highlights a lack of information useful for making decisions about the allocation of resources for coronary artery disease. It is concerning that over $2 billion of Australia’s annual healthcare budget is being spent on cardiovascular disease, with inadequate economic evidence. Since the mid to late 1990s, increased spending has been directed towards PCI with stenting, over coronary artery bypass grafting, but there is insufficient economic evidence to support this transition. Compounding this, the current cost-effectiveness evidence which would suggest a move away from PCI and stents remains too unclear and uncertain for policy-makers to be confident in making changes. In a time of increased pressure on health budgets, economic evidence should be fundamental to resource allocation decisions.\n\nFor those wishing to make resource allocation decisions to improve the efficiency in treatment of coronary artery disease, the current evidence is insufficient. A transparent, structured, lifetime analysis of all competing treatments, incorporating quality of life measures, would be valuable for decision makers. The analysis should account for fluctuations in the quality of life of patients over their lifetimes, related to symptom relief, repeat procedures and acute events.\n\nTo be of use to decision-makers and have a better chance of generating policy change, analyses must be accessible; economic evaluations should include a clear indication for the direction of policy or changes to practice that should follow and a statement of the probability that such changes will the generate the predicted improvements. In their systematic review of barriers and facilitators to use of evidence by policymakers, Oliver et al. named clarity, relevance and reliability as some of the top barriers to use of evidence35. For decision-makers to be able to act on the economic evidence, the expected effect of making changes based on the results needs to be clear. In cases where there is too much uncertainty, a strategy to improve the analysis should be outlined.\n\nWe suggest that to improve reporting of economic evaluations for decision-making, an additional item could be included in the CHEERS Statement, relating to implications for policy and practice. Ideally, this would be a statement describing the implications of acting on the evidence presented; encompassing both expected improvements to health outcomes and confidence in the effect.\n\n\nData availability\n\nDataset 1: Endnote library of retrieved references - Data related to this review are available in an EndNote Library, containing all references retrieved using the search terms described. This library also contains subfolders used to categorise papers during the review process. 10.5256/f1000research.13616.d19056236",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work forms part of the PhD Candidature of Victoria McCreanor, funded by the Capital Markets Cooperative Research Centre and supported by the Queensland University of Technology.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary file 1: Search terms used in each database searched.\n\nClick here to access the data.\n\nSupplementary file 2: Completed PRISMA Checklist.\n\nClick here to access the data.\n\nSupplementary file 3: CHEERS Statement Checklist – completed for reviewed studies.\n\nClick here to access the data.\n\n\nReferences\n\nAustralian Institute of Health and Welfare: Health-care expenditure on cardiovascular diseases 2008–09. Canberra: AIHW. 2014. Reference Source\n\nWaters AM, Trinh L, Chau T, et al.: Latest statistics on cardiovascular disease in Australia. Clin Exp Pharmacol Physiol. 2013; 40(6): 347–356. PubMed Abstract | Publisher Full Text\n\nAustralian Institute of Health and Welfare: Coronary revascularisation in Australia, 2000. Canberra: AIHW. 2003. Reference Source\n\nAustralian institute of Health and Welfare: Cardiovascular disease, diabetes and chronic kidney disease—Australian facts: Morbidity–Hospital care. Canberra: AIHW. 2014. Reference Source\n\nAustralian Institute of Health and Welfare: Cardiovascular disease: Australian facts 2011. Canberra: AIHW. 2011. Reference Source\n\nBoden WE, O'Rourke RA, Teo KK, et al.: Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med. 2007; 356(15): 1503–1516. PubMed Abstract | Publisher Full Text\n\nWeintraub WS, Spertus JA, Kolm P, et al.: Effect of PCI on quality of life in patients with stable coronary disease. N Engl J Med. 2008; 359(7): 677–687. PubMed Abstract | Publisher Full Text\n\nHueb W, Lopes N, Gersh BJ, et al.: Ten-Year Follow-Up Survival of the Medicine, Angioplasty, or Surgery Study (MASS II): A Randomized Controlled Clinical Trial of 3 Therapeutic Strategies for Multivessel Coronary Artery Disease. Circulation. 2010; 122(10): 949–957. PubMed Abstract | Publisher Full Text\n\nBoden WE, O'Rourke RA, Teo KK, et al.: Impact of optimal medical therapy with or without percutaneous coronary intervention on long-term cardiovascular end points in patients with stable coronary artery disease (from the COURAGE Trial). Am J Cardiol. 2009; 104(1): 1–4. PubMed Abstract | Publisher Full Text\n\nWeintraub WS, Grau-Sepulveda MV, Weiss JM, et al.: Comparative effectiveness of revascularization strategies. N Engl J Med. 2012; 366(16): 1467–1476. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLund H, Brunnhuber K, Juhl C, et al.: Towards evidence based research. BMJ. 2016; 355: i5440. 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. PLoS Med. 2009; 6(7): e1000097. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHusereau D, Drummond M, Petrou S, et al.: Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. Value Health. 2013; 16(2): e1–e5. PubMed Abstract | Publisher Full Text\n\nHusereau D, Drummond M, Petrou S, et al.: Consolidated Health Economic Evaluation Reporting Standards (CHEERS)--explanation and elaboration: a report of the ISPOR Health Economic Evaluation Publication Guidelines Good Reporting Practices Task Force. Value Health. 2013; 16(2): 231–250. PubMed Abstract | Publisher Full Text\n\nCaruba T, Katsahian S, Schramm C, et al.: Treatment for stable coronary artery disease: a network meta-analysis of cost-effectiveness studies. PLoS One. 2014; 9(6): e98371. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFidan D, Unal B, Critchley J, et al.: Economic analysis of treatments reducing coronary heart disease mortality in England and Wales, 2000–2010. QJM. 2007; 100(5): 277–289. PubMed Abstract | Publisher Full Text\n\nGriffin SC, Barber JA, Manca A, et al.: Cost effectiveness of clinically appropriate decisions on alternative treatments for angina pectoris: Prospective observational study. BMJ. 2007; 334(7594): 624. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHlatky MA, Boothroyd DB, Melsop KA, et al.: Economic outcomes of treatment strategies for type 2 diabetes mellitus and coronary artery disease in the Bypass Angioplasty Revascularization Investigation 2 Diabetes trial. Circulation. 2009; 120(25): 2550–2558. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVieira RD, Hueb W, Hlatky M, et al.: Cost-effectiveness analysis for surgical, angioplasty, or medical therapeutics for coronary artery disease: 5-year follow-up of medicine, angioplasty, or surgery study (MASS) II trial. Circulation. 2012; 126(11 Suppl 1): S145–150. PubMed Abstract | Publisher Full Text\n\nCapewell S, Beaglehole R, Seddon M, et al.: Explanation for the Decline in Coronary Heart Disease Mortality Rates in Auckland, New Zealand, Between 1982 and 1993. Circulation. 2000; 102(13): 1511–1516. PubMed Abstract | Publisher Full Text\n\nThe BARI 2D Study Group, Frye RL, August P, et al.: A Randomized Trial of Therapies for Type 2 Diabetes and Coronary Artery Disease. N Engl J Med. 2009; 360(24): 2503–2515. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbizaid A, Costa MA, Centemero M, et al.: Clinical and economic impact of diabetes mellitus on percutaneous and surgical treatment of multivessel coronary disease patients: insights from the Arterial Revascularization Therapy Study (ARTS) trial. Circulation. 2001; 104(5): 533–538. PubMed Abstract | Publisher Full Text\n\nBakhai A, Stables RH, Prasad S, et al.: Trials comparing coronary artery bypass grafting with percutaneous transluminal coronary angioplasty and primary stent implantation in patients with multivessel coronary artery disease. Curr Opin Cardiol. 2000; 15(6): 388–94. PubMed Abstract | Publisher Full Text\n\nFarkouh ME, Dangas G, Leon MB, et al.: Design of the Future REvascularization Evaluation in patients with Diabetes mellitus: Optimal management of Multivessel disease (FREEDOM) trial. Am Heart J. 2008; 155(2): 215–223. PubMed Abstract | Publisher Full Text\n\nThe BARI Investigators: Influence of Diabetes on 5-Year Mortality and Morbidity in a Randomized Trial Comparing CABG and PTCA in Patients With Multivessel Disease: the Bypass Angioplasty Revascularization Investigation (BARI). Circulation. 1997; 96(6): 1761–9. PubMed Abstract | Publisher Full Text\n\nBagust A, Grayson AD, Palmer ND, et al.: Cost effectiveness of drug eluting coronary artery stenting in a UK setting: cost-utility study. Heart. 2006; 92(1): 68–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBowen J, Hopkins R, He Y, et al.: Systematic review and cost-effectiveness analysis of drug eluting stents compared to bare metal stents for percutaneous coronary interventions in Ontario. Ontario: Ontario Ministry of Health & Long-term Care; 2005. Reference Source\n\nEkman M, Sjögren I, James S: Cost-effectiveness of the Taxus paclitaxel-eluting stent in the Swedish healthcare system. Scand Cardiovasc J. 2006; 40(1): 17–24. PubMed Abstract | Publisher Full Text\n\nHill R, Bagust A, Bakhai A, et al.: Coronary artery stents: a rapid systematic review and economic evaluation. Health Technol Assess. 2004; 8(35): iii–iv, 1–242. PubMed Abstract | Publisher Full Text\n\nNeyt M, Van Brabandt H, Devriese S, et al.: Cost-effectiveness analyses of drug eluting stents versus bare metal stents: A systematic review of the literature. Health Policy. 2009; 91(2): 107–120. PubMed Abstract | Publisher Full Text\n\nBrandao SM, Hueb W, Polanczyk CA, et al.: Utility measures and quality-adjusted life years in patients with symptomatic multivessel coronary artery disease assigned to surgery, angioplasty or medical treatment-mass ii trial. Value Health. 2016; 19(7): A658. Publisher Full Text\n\nAl-Lamee R, Thompson D, Dehbi HM: Percutaneous coronary intervention in stable angina (ORBITA): a double-blind, randomised controlled trial. Lancet. 2018; 391(10115): 31–40. PubMed Abstract | Publisher Full Text\n\nMerlo G, Page K, Ratcliffe J, et al.: Bridging the gap: exploring the barriers to using economic evidence in healthcare decision making and strategies for improving uptake. Appl Health Econ Health Policy. 2015; 13(3): 303–309. PubMed Abstract | Publisher Full Text\n\nWilliams I, Bryan S: Understanding the limited impact of economic evaluation in health care resource allocation: a conceptual framework. Health Policy. 2007; 80(1): 135–143. PubMed Abstract | Publisher Full Text\n\nOliver K, Innvar S, Lorenc T, et al.: A systematic review of barriers to and facilitators of the use of evidence by policymakers. BMC Health Serv Res. 2014; 14(1): 2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCreanor V, Graves N, Barnett A, et al.: Dataset 1 in: A review and critical analysis of cost-effectiveness studies for coronary artery disease treatment. F1000Research. 2018. Data Source"
}
|
[
{
"id": "30777",
"date": "02 Mar 2018",
"name": "Elizabeth A. Geelhoed",
"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 presents a valuable critical analysis of cost-effectiveness studies and highlights serious issues in relation to potential inefficiencies in funding for Coronary Artery Disease (CAD) treatment. As identified by the authors, heart disease is a significant area of expenditure for health in Australia and this paper contributes important information for policy consideration. The paper is well-written and raises a number of recommendations useful for general presentation of cost-effectiveness analyses in the literature. In particular, these recommendations are useful in terms of the quality of life measurement and the application of the CHEERS checklist. One interesting issue raised here – which could potentially be further explored - is in terms of high-level policy changes. All new procedures presented for listing on Medicare or the PBS necessarily undergo a rigorous process of evaluation, including assessment of cost-effectiveness evidence, which often includes commissioned modelling studies to apply international clinical findings for Australian scenarios. The economic evaluation requires comparison of the new technology or drug with the status quo (or best current alternative) to appraise the incremental cost-effectiveness for the specific population group defined in the submission. However the point raised here is an important one – that incremental benefits and costs in cost-effectiveness may not reflect the bigger picture if the underlying comparator is flawed. Nonetheless it is a much greater undertaking to address wider policy ramifications, particularly where evidence continues to accrue. (Notably the current MBS taskforce is endeavouring to address such challenges.) One major criticism however is that it is not made clear to the reader that the analysis specifically targets stable angina, which is a subgroup of CAD patients. In particular, the abstract suggests that the results apply to treatment of all CAD patients and hence the results are potentially misleading when considered across all patients with CAD. It is important to know what proportion of CAD patients present with stable disease (and hence are eligible for all three treatments (CABG, PCI, MT). The information is implicit since it is clear from the population group of each of the included trials that patients with acute disease have been excluded. The fairly recent RCT which provides the most enlightening data, has quite strict exclusion criteria, which lists a number of presentations. The conclusions of the review are valid, but the subgroup requires transparency. This leads to a second point, which questions the importance of the clinical diagnosis. NICE clinical guidelines recommend drug treatment for stable angina unless symptoms are not satisfactorily controlled, in which case interventional procedures should be considered. This suggests a value judgement in selecting patients for interventional care. Does this potentially present a continuum that has shifted over time? It is important to assess what proportion of the cited 57% increase in PCI has been for the treatment of stable CAD. A related query is whether the non-RCT studies were subject to possible selection bias. A fundamental question then arises as to whether the cost-effectiveness studies should stand apart from the clinical data. That is, if the clinical evidence is changing (and clearly the controversy around interventional procedures for stable disease reflects the growing evidence) – it may be most useful for the cost-effectiveness to be presented for the specific sub groups that are most controversial – for example based on the differential burden of ischaemia.\n\nQueries and comments: ·\n\nReviews were excluded from consideration – were they considered for comparison of findings and were they scrutinised for potentially missed references? ·\n\n‘..most of the .. papers did not examine medical therapy.’ Was this because they pertained to acute presentations or was medical therapy overlooked for stable presentation?\n\nI disagree with some of the statements made around usefulness for policy. For example the statement that presenting clinical complexities ‘does nothing to assist those wanting to make higher level resource allocation decisions’. Complexity may limit relevance in some cases, but if cost-effectiveness varies according to clinical diagnosis (and evidence supports this), then the heterogeneity may be an important consideration. Also, the statement referring to publications that consider only 2 treatments as not being useful to policy – if the policy decision relates to critical cases such as myocardial infarction or if it relates to a minor change to policy already in existence, then comparison of more than 2 options may not be relevant. Thirdly, non-Australian evidence may be appropriate if the clinical findings are relevant to an Australian population and Australian cost data can be applied. It may be in the wording of some of these statements, given the separation of higher level allocation decisions and those relating to small changes in resource allocation. However most policy decisions assess a well-defined intervention, compared to a current alternative, for a clearly defined patient group (which is necessary to prevent leakage), so commonly the detailed breakdown of cost-effectiveness provides critical information for those decisions. The paper provides an interesting supplement to the current clinical literature on the controversy around treatment of stable coronary artery disease. The discussion of the importance of policy-relevant evidence will inform future cost-effectiveness analyses of trials currently in progress.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Partly\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Not applicable\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly",
"responses": [
{
"c_id": "3765",
"date": "03 Jul 2018",
"name": "Victoria McCreanor",
"role": "Author Response",
"response": "We agree with these comments regarding the Australian regulatory system, however, we don’t feel this paper is the right place for a discussion of Australian regulatory processes. We also agree that economic evidence may change as new evidence accrues, however, cost-effectiveness analysis based on imperfect information should still be used as long the caveats are noted and the decision makers are fully informed about the problems with any analysis. Using imperfect information to make decisions is better than making them with no information. As with clinical evidence and practice, economic evaluations can be updated as new evidence becomes available. We have updated the text to make it clear that the review relates to patients with stable disease. We agree with the comments regarding the importance of clinical diagnosis, however there is mounting evidence, for example from the first results from the ORBITA trial published in the Lancet, that in stable patients, interventional treatment may not improve patient outcomes to the extent originally expected. Regarding the change in preference towards PCI, we have updated the text to include the proportion increase in PCI related to non-AMI patients. The point raised about non-RCT studies being subject to selection bias is valid, however we would suggest that perhaps these better reflect real-world practice and therefore may be more useful for cost-effectiveness analysis and policy than results from highly-controlled clinical trials. RCTs aim for high internal validity, which comes at the expense of generalisability, and generalisability is more important for economic evaluation and policy decisions. We agree that where clinical evidence changes, economic analyses may also need to be updated. However, we think it is important to use the evidence currently available, rather than waiting for perfect information. We agree that where the effectiveness is affected by particular clinical sub-groups, it will likely be important to present the cost-effectiveness evidence accordingly. In this paper we only attempted to review cost-effectiveness analyses which compared PCI, CABG and OMT in groups of patients where all three were considered appropriate treatment options. The other reviewer also raised queries about relevance across different patient sub-groups and we noted the following: This would introduce three additional reviews (PCI v CABG, OMT v CABG and OMT v PCI) - All are relevant but of course, by necessity, only pertaining to more and more selected groups. Thus, comparisons would be hard to interpret, if not, impossible. In response to the queries and comments: We have updated the text to note that we examined review papers for potentially missed references. The sentence referred to, ‘..most of the .. papers did not examine medical therapy.’, meant that most studies did not examine all three treatment groups together. However, most compared interventional treatments with each other and did not consider medical therapy as an alternative treatment on its own. Some pertained to acute presentations and others to stable, however, as noted this paper relates only to stable disease. In response to comments about statements regarding usefulness for policy: We agree that clinical complexity is important in many instances. In the comment highlighted, we refer specifically to the Fidan et al paper, which presents 34 scenarios together, making it difficult to interpret in the context of resource allocation decision-making. We have updated the text to clarify this. Regarding comments about two-treatment comparisons not being useful for policy, we agree that if only two treatments are relevant to a certain cohort, then an analysis of only those two would be useful. We have also addressed in this in an earlier comment. Perhaps it is our wording, but we do not mean to imply that two-treatment comparisons are irrelevant in all contexts. We do state that if a two-treatment analysis omits another relevant treatment, then the analysis is limited. We agree with the comments regarding the Australian context and have updated the text to note that Australian costs could be applied to an existing analysis. We agree that detailed breakdown of cost-effectiveness analysis and caveats is important for policy decisions and have discussed this in some of our earlier comments. Our main point, however, is that presenting too much detail can detract from the overall message."
}
]
},
{
"id": "34101",
"date": "31 May 2018",
"name": "James P Howard",
"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 excellent review of cost-effectiveness studies comparing percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG) and optimal medical (OMT) therapy for stable coronary artery disease (CAD). It is well-written, easy to follow, and the interpretations are sound.\nMy only query about the design of the study is I'm not entirely sure why studies comparing only PCI vs OMT were excluded, given the authors go on to rightly raise concerns about a lack of high-quality data. We must remember that the majority of of patients with stable CAD are not candidates under current guidelines for CABG (e.g. if they lack 3 vessel or proximal LAD disease) and so to argue a cost-effectiveness study should use data where all three are compared because CABG could be a superior options to PCI seems strange to me. Even if CABG was deemed more effective than PCI in a study, the finding would not be applicable to most of our patients, as only a small subset of our patients would be eligible for randomisation in whatever study that data was based on.\nI also wonder whether the authors think this data will change in the next few years? The two most high profile studies of stable CAD in the last few years are probably FAME-2 and ORBITA. The former has shown improvements when measured as a reduction in a composite of death, myocardial infarction and urgent revascularisation (although this latter largely powers this and there have been significant concerns about this endpoint in unblinded settings). ORBITA initially showed no benefits in improvements in exercise time from PCI versus optimal medical therapy, but the recent secondary analysis in Circulation showed a significant reduction in the number of patients reporting angina following PCI versus placebo (when all patients were on OMT). I appreciate that the authors are reviewing cost-effectiveness studies, and it may be several years before these new data are included in such studies, but it appears that under this current design (including only cost-effectiveness studies involve CABG) would mean the only blinded trial of PCI versus OMT that I am aware of would be ineligible for analysis?\nSome minor points are there isn’t discussion about the role of direct versus indirect cost measurements in the different studies. This is crucial when considering the applicability of results to different healthcare settings, where sometimes indirect costs are borne by providers but sometimes not.\nAlso, the authors do query about the applicability of data to Australia. Whilst this may be an issue, it seems likely that evidence from Europe and USA are of relatively-good applicability to Australia as similarly developed nations with, in some cases, similar healthcare systems.\nFinally, a tiny point is I think it would be useful to order tables 4 and 5 the same way; I foolishly initially assumed that the top study in both (the most recent study using meta-analysis in table 4 and the high quality study in table 5) were the same study, when actually the most recent study was deemed of 'low' usefulness). Indeed, table 5 may be better as a traffic light system with colour coding to make things more clear, but I will leave that up to the authors and editorial team.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "3764",
"date": "03 Jul 2018",
"name": "Victoria McCreanor",
"role": "Author Response",
"response": "We agree with the comments regarding appropriateness of treatment in different patient groups. However, the problem with reviewing papers comparing two treatments is complexity. This would introduce three additional reviews (PCI v CABG, OMT v CABG and OMT v PCI) - All are relevant but of course, by necessity, only pertaining to more and more selected groups. Thus, comparisons would be hard to interpret, if not, impossible.Regarding the comments about data changing in the coming years, in short, yes, we think data are likely to change. Invasive procedures for stable disease are currently being subject to increasing scrutiny, particularly in the wake of ORBITA, as you suggest. The secondary analysis of ORBITA provides a well-needed and important contribution to the field and it will be interesting to see how those results affect cost-effectiveness estimates. We have included a comment in the discussions to that effect. Other important data will come from the ISCHEMIA trial, which randomises patients with stable disease to conservative or invasive strategies, particularly in relation to long-term outcomes.In response to the other points, all studies used direct costs and we have updated the text accordingly.We agree that evidence from Europe and USA is of relatively-good applicability to Australia and have updated the text to include a comment that existing analyses could be applied to the Australian context if Australian costs were used.We thank you for the useful suggestion to make the tables clearer. We have updated tables 4 and 5 so the studies are in the same order and included a traffic light system for overall usefulness."
}
]
}
] | 1
|
https://f1000research.com/articles/7-77
|
https://f1000research.com/articles/7-978/v1
|
02 Jul 18
|
{
"type": "Research Article",
"title": "Factors associated with induced demand for services in Iran’s healthcare system",
"authors": [
"Ghahraman Mahmoudi",
"Ghanbar Roohi",
"Mohammad Asadi",
"Fatemeh Rasooly Kalamaki",
"Samira Abam",
"Mansoor Khojamli",
"Masoomeh Abdi Talarposhti",
"Ghanbar Roohi",
"Mohammad Asadi",
"Fatemeh Rasooly Kalamaki",
"Samira Abam",
"Mansoor Khojamli",
"Masoomeh Abdi Talarposhti"
],
"abstract": "Background: One of the most important subjects in health economics and healthcare management is the theory of induced demand; that is, caring for or providing and selling unnecessary services to users of healthcare systems, which is accompanied by the exercising of power by the service providers. Methods: This study was performed on physicians, nurses, and laboratory and radiology technicians working in Medical Science universities. Random sampling was conducted from five areas: the center, north, west, east and south of Iran. Data were gathered by a questionnaire, with a Cronbach's alpha of >0.7, consisting of nine dimensions on existence of induced demand and its associated factors. Results: The results showed that overall, 65.2% of the participants agreed with the existence of induced demand. Chi-squared test showed there was no difference in the level of induced demand between the regions of the country, education level and occupation. However, there was a significant difference in terms of gender (P<0.005). The Kruskal-Wallis test indicated a significant relationship between the associated factors and induced demand (P<0.005). Conclusions: Results showed that induced demand was influenced by factors including service recipients’ awareness, personal benefits of service providers, the extent they cared about health, supervision of insurance companies, industrialization of the health sector, diversity and increased number of trained experts and the quality of methods of training the service providers. Therefore, policymakers and planners should consider raising awareness of health service recipients, supervising insurance companies, reforming teaching methods, social culture making and changing the beliefs of society.",
"keywords": [
"Induced demand",
"Healthcare services",
"Hospital",
"Healthcare system"
],
"content": "Introduction\n\nThe provision of healthcare services is faced with serious challenges due to economic factors such as the adequacy or lack of financial resources, access to technology and medical equipment, and social determinants of health1. Adverse selection, moral hazard and induced demand are three important phenomena that affect patients' behavior2. Among these, induced demand is an issue that economists consider to have side effects and to affect interactions between patient and physician. Economic transactions generally involve two parties, a seller and a buyer. However, demand from patients (as buyers) in the health sector is mediated by other factors (i.e. physicians and sometimes healthcare agencies3. In fact, induced demand is defined as inducing caring for or providing and selling unnecessary services to clients of healthcare system that is accompanied with exercise of power by the service providers. This may result in providing services that do not have a positive impact on the health of the client4. The phenomenon of induced demand supplier; in other words, the information asymmetry between the service provider and the customer or service applicant in the healthcare market due to the poor knowledge of patients, the importance of health and concern over the consequences of diseases, clients tend to comply with the orders of the service providers5. For example, some patients assume that the services provided in hospitals are of higher quality than those in clinics, and prefer to be treated in hospitals even if treatment is available in clinics6. The assumption of asymmetric information between the physician and patient is the background for the induced demand theory7. The induced demand hypothesis emphasizes the motivations of doctors as healthcare providers to create induced demand8. The patients also affect the induction of demand as they have unlimited desire for, and tendency toward, services9. Moreover, economic incentives focused on insurance can affect patients’ behavior and the profit of healthcare organizations10. Many studies have confirmed physicians' induced demand, identified and investigated various supply and demand factors11–15. A study by Bogg et al in India showed that the government's financial support has resulted in induced demand7. Recent studies also demonstrated the impact of physicians’ payment mechanisms on motivating health-seeking behavior in patients16. A study by Karimi et al showed that induced demand for medical and pharmaceutical services is more common among general physicians17. Cutler et al revealed a correlation between causes of induced demand from patients and physicians in providing services in different regions18. A study by Yuda showed that a 1% reduction in medical costs by the service providers could increase induced demand for services by up to 7.5%19. In their study in France, Delattre and Dormont showed that increasing the number of physicians relative to the population reduced the number of patients seeking medical advice/counseling; however, this reduction was minimal and a decline in the number of consulting physicians did not increase the volume of medical services sought20. A study in Italy by Magazzino and Mele showed that the gross national product, unemployment rate, number of hospital beds, degree of urbanization and percentage of the population with high school education had a significant direct impact on health expenditure21.\n\nThe issue of induced demand is highly complicated due to the information asymmetry in medical services22. The possibility of induced demand and the severity of its effects are associated with interactions between a wide range of factors that affect patients’ behavior, including market, behavior, rules and regulations. These factors could act as incentives and disincentives for patients’ participation in or resistance against the phenomenon of induced demand4. Considering the extent of factors affecting induced demand and lack of a comprehensive study on this issue in Iran, the purpose of this paper is to examine the factors of knowledge and awareness of service providers, the personal interests of service providers, their sensitivity to health, the supervision of insurance companies, the industrialization of the health sector, the increase and diversity and the number of specialist forces and the quality of practices training to providers in hospitals.\n\n\nMethods\n\nThis applied, cross-sectional, descriptive-analytical study was conducted from January to December 2015 in a number of health centers and hospitals in Iran. Overall, 511 subjects consisting of 78 physicians, 350 nurses, 26 laboratory experts and 28 radiology technicians were included in this study. The subjects were selected by a mixed sampling method, which included stratified random sampling (the country was classified into five regions: center, north, west, east and south) and cluster sampling of levels was conducted within hospitals and health centers (each center or hospital was considered as a single cluster).\n\nThe centers and hospitals studied were as follows:\n\nNorth: Imam Khomeini Hospital (Sari), Imam Reza Hospital affiliated to Sari University of Medical Sciences (Amol), 5th Azar Hospital (Gorgan), Shahid Sayad Shirazi Hospital affiliated to Golestan University of Medical Sciences (Gorgan).\n\nSouth: Golestan Hospital affiliated to Ahvaz University of Medical Sciences (Ahvaz).\n\nEast: Imam Reza Hospital (Mashhad), Rasool-e-Akram Hospital (Zabol).\n\nWest: Imam Khomeini Hospital affiliated to Ardebil University of Medical Sciences (Ardebil).\n\nCenter: Rasool-e-Akram Hospital affiliated to Iran University of Medical Sciences (Tehran).\n\nFinally, sampling within clusters was done by classification per ratio of physicians, nurses, laboratory experts and radiology technicians working in the medical centers. The main goal of the study was to determine the demand induction rate of services in Iran's health system and its related factors. The following formula was used to calculate the sample size, with a correction factor of 1.25 used for more precision.\n\n\n\nAt least 480 samples were selected for this study, consisting of 350 females (72.9%) and 130 (27.1%) males aged 22–57 years. Inclusion criteria consisted of at least 6 months of work experience and willingness to participate in the study.\n\nData were collected by a questionnaire (Supplementary File 1), which consisted of two parts: (a) background characteristics and (b) factors affecting induced demand for health services. The individuals completed the questionnaires in writing in the presence of the researcher. This questionnaire contains 33 items and 9 dimensions and uses the Likert scale. The items have 5 options that are as follows: I totally agree with the highest score (score of 1), I agree score of 2), I somewhat agree (score of 3), I disagree (score 4), and I totally disagree with the lowest score (score of 5). The dimension of service recipients’ awareness had 3 related item with the minimum score of 3 and maximum score of 15, the dimension of the extent the service providers cared about their health had 8 related item with a minimum score of 8 and maximum score of 40, the dimension of supervision and planning of insurance companies had 3 related item with the minimum score of 3 and maximum score of 15, the dimension of personal benefits of service providers had 3 related item with the minimum score of 3 and maximum score of 15, the dimension of industrialization of the health sector had 3 related item with the minimum score of 12 and maximum score of 20, the dimension of diversity and increased number of trained experts had 4 related item with the minimum score of 4 and maximum score of 20, the dimension of service providers’ awareness had 3 related item with the minimum score of 3 and maximum score of 15, the dimension of quality of models and methods used for training service providers had 3 related item with the minimum score of 3 and maximum score of 15, and the total score of questionnaire ranged between 33 and 165. The validity and reliability of the questionnaire was confirmed by a Master's thesis (S. Montazeri, Master’s thesis). The scientific validity of its contents was verified by 10 faculty members. Cronbach's alpha of 0.80 was achieved in a preliminary study on 30 samples. Cronbach's alpha of more than 0.7 was achieved for the nine dimensions. Table 1 shows the dimensions of the questionnaire of induced demand for health services in Iran.\n\nData were analyzed using SPSS (version 23) and descriptive statistics (the mean and standard deviation) were produced. Considering the difference in the number of options and scores of each factor, scores were first converted to a 0–100 scale. Non-parametric tests were used to determine the relationship between factors and induced demand. The Kruskal–Wallis test was used due to absence of normality assumption for factors associated with induced demand. The chi-squared test was also used for data analysis at a significance level of P<0.05.\n\nEthical approval was obtained from the relevant organization (Islamic Azad University of Sari Branch with code of ethics IR.IAU.SARI.REC.1397.7) and written consent of the participants (samples), and they were assured of the confidentiality of all data collected.\n\n\nResults\n\nOverall, 511 subjects, who were working in emergency departments (17.9%), wards (71.5%), laboratories (6.8%) and clinics (3.8%), completed the questionnaire. The highest and lowest level of agreement with induced demand was recorded in the west (71%) and south (57.5%) of Iran, respectively. Moreover, 65.2% of the subjects in the whole country agreed with the existence of induced demand, while only 4.4% did not agree with the existence of induced demand. The chi-squared test showed no significant difference between different regions of the country (Table 2).\n\nMost individuals with bachelors’ degree (67.7%) agreed with the presence of induced demand, while only 38.1% of specialists agreed with the presence of induced demand. However, there was no significant relationship between education level and induced demand.\n\nLaboratory staff (73.1%) and physicians (55.1%) had the highest and lowest level of agreement with the existence of induced demand, respectively. Moreover, 68.1% and 3.8% of the nurses agreed and disagreed with the existence of induced demand, respectively. Furthermore, 57.1% and 4.4% of the radiology technicians agreed and disagreed with the existence of induced demand, respectively. Chi-square test showed no significant relationship between induced demand and job categories (P=0.194).\n\nThe results also showed that women (69.4%) agreed significantly more with the existence of induced demand than men (53.8%) (Table 3). The results of this study indicate that 64.9% of the subjects believed that sensitivity of clients toward their own health affected the incidence of induced demand. In addition, 62.1% of the subjects believed that the monitoring and planning of insurance companies affected the incidence of induced demand. Kruskal-Wallis test showed a significant relationship between induced demand and its related factors (Table 4).\n\n\nDiscussion\n\nThe results of the present study showed that the majority of participants (65.2%) agreed with the existence of induced demand, which is consistent with previous studies11–15. Bogg et al showed that the government’s financial support has led to induced demand7 In Japan, Sekimoto and Ii studied the supply and induced demand for chronic diseases, and reported an increase in induced demand for hypertension and diabetes24. The findings of Crivellei et al. in Switzerland25 Delattre and Dormont in France20 were in line with the results of this study. However, a study in America showed no correlation between demand for services and services provided by physicians through Medicare26. A study by Amporfu indicated that salaried physicians have no incentive to induce demand27. Although few studies strongly reject the existence of induced demand, some experts believe that identification and determination of the incidence of induced demand are health economics issues due to the complexity of treatment, medical decisions, clinical uncertainty and modernization of health needs. It is not simple to distinguish between real demand and induced demand. In this regard, Richardson and Peacock claimed that medical decision-making is often complex and uncertain even if all medical standards were met28. Lien et al also stated that the development of needs, epidemiological changes and diversity of tastes have further complicated patient diagnosis and treatment, and the distinction between real and induced demand29. It seems that the implementation of healthcare reform in Iran and significant reductions in patients’ out-of-pocket payments have resulted in increased demand for services. This issue is not observed in countries with more stable healthcare system. On the other hand, cultural differences and health beliefs affect the amount of demand in different regions.\n\nInduced demand occurs because of multiple economic and structural factors22. The present study found a significant relationship between induced demand and the related factors of knowledge and awareness of service recipients, sensitivity of clients toward their own health, monitoring and planning of insurance companies, personal interests of service providers, industrialization of the healthcare sector, increases in the variety and number of specialists, knowledge and awareness of service providers, quality of education models and training methods for service providers. Leone conducted a study in three provinces of India in 2014, which indicated the role of several factors in induced supply and demand30. The present study is also in line with a study by Keyvanara et al, which demonstrated factors such as the role of supplemental health insurance, lack of strict supervision on insurance companies, increased profit of equipment companies, diagnostic centers, pharmaceutical companies, the over-trusting of physicians, incorrect demands of patient from the physicians, lack of awareness, patients’ free access to physicians, patients’ willingness to make greater use of free services and franchising affect the incidence of induced demand4. Bazyar et al indicated inadequate supervision on the insurance system as the main cause of induced demand31. A study conducted in China reported that the high percentage of cesarean sections is due to the decision of pregnant women. However, the physicians’ impact on the increased demand for cesarean section cannot be ruled out32. This imbalance between clients and physicians' knowledge may result in the physicians’ tendency to provide questionable services3. A physician has the expertise and knowledge to have a dual role of a consultant and service provider for the patient33. Borhanzade claims that poor monitoring and control by insurance companies on the payment system leads to easier demand for excess services34. The findings of other studies indicate that consumers compensate for the increased costs of healthcare by using supplemental insurance35. Madden et al believe that different methods of payment affect the professional behavior of service providers36. According to a study by Dosoretz, physicians and other providers maximize their benefit when clinical values are not defined37. The difference between the results of this study and those of other studies is due to the focus of other studies on methods of payment to physicians and other components of health economics, whereas the present study emphasized more on social factors, knowledge, attitude, commitment to medical ethics and insurance company monitoring. Authorities and decision-makers in the healthcare sector should aim to increase public awareness about diseases and treatment processes, and offer appropriate solutions to increase physicians’ occupational commitment by changing the medical education system. In the meantime, other interfering factors in increased demand such as payment methods, performance of public and supplementary insurances should be monitored closely. In this regard, the mass media could be helpful in raising public awareness and modifying pubic beliefs.\n\nThe present study showed that the industrialization of the healthcare sector and an increase in the variety and numbers of specialists are effective on induced demand. Nowadays, market competition has been created between the providers of healthcare services. Ferguson demonstrated that induced demand results from the increasing role of the market in medical care38. Economics is the most important factor affecting the medical institutions, which use induced demand as a means of increasing profits. An increase in unnecessary services leads to increased profit, and even distribution of money between various providers39. Bhatia emphasizes that the pharmaceutical market is highly competitive, and each company is constantly marketing to maximize their revenue40. Palesh et al claimed that market forces such as advertising are significantly effective on the unnecessary use of technology. Moreover, advertisements of importers and manufacturers increase demand41. However, Bickerdyke et al suggest that induced demand is influenced by the patient's goals (including good health, vitality, longevity, information, autonomy, usefulness of process, regular source of care, confidence, access to benefits of social security or insurance payments and pain relief), medical history and environmental factors (including financial situation, family/work commitments, cultural background and attitude of the society)42. Nevertheless, it seems that this previous study was patient-centered, and mostly focused on the quality of services rather than social and economic aspects.\n\nAlthough healthcare services are the main indicator of social development, the provision of such services is tremendously costly. The ever-increasing cost, along with induced demand for services, causes certain problems in society. The results of this study showed that the phenomenon of induced demand is influenced by various factors, including the knowledge and awareness of service recipients, sensitivity of clients toward their own health, monitoring and planning of insurance companies, personal interests of service providers, industrialization of the healthcare sector, increases in the variety and number of specialists, knowledge and awareness of service providers, quality of education models and training methods for service providers. Therefore, it is suggested that decision makers and policymakers in the community arena can improve health services by increasing the quality of services and satisfaction of patients reducing direct payments from people, reducing household expenses, observing tariffs, lowering under-payment and unnecessary inductive demand. To be efficient and optimal.\n\n\nData availability\n\nDataset 1. Responses of all participants to all questions. DOI: 10.5256/f1000research.14377.d20159823",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe would like to thank the Research of Islamic Azad University, Sari Branch, Iran, and those who participated in the project.\n\n\nSupplementary material\n\nSupplementary File 1. Demographic and induced demand questionnaire.\n\nClick here to access the data.\n\n\nReferences\n\nWolsko PM, Eisenberg DM, Davis RB, et al.: Insurance coverage, medical conditions, and visits to alternative medicine providers: results of a national survey. Arch Intern Med. 2002; 162(3): 281–7. PubMed Abstract | Publisher Full Text\n\nAtella V, Holly A, Mistretta A: Disentangling adverse selection, moral hazard and supply induced demand: An empirical analysis of the demand for healthcare services. CEIS Working Paper No 389. 2016; 60. Publisher Full Text\n\nTunis SR: Economic analysis in healthcare decisions. Am J Manag Care. 2004; 10(5): 301–4. PubMed Abstract\n\nKeyvanara M, Karimi S, Khorasani E, et al.: Experts’ perceptions of the concept of induced demand in healthcare: A qualitative study in Isfahan, Iran. J Educ Health Promot. 2014; 3: 27. PubMed Abstract | Free Full Text\n\nIzumida N, Urushi H, Nakanishi S: An empirical study of the physician-induced demand hypothesis: The cost function approach to medical expenditure of the elderly in Japan. Rev Popul Soc Policy. 1999; 8: 11–25. Reference Source\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\nBogg L, Diwan V, Vora KS, et al.: Impact of Alternative Maternal Demand-Side Financial Support Programs in India on the Caesarean Section Rates: Indications of Supplier-Induced Demand. Matern Child Health J. 2016; 20(1): 11–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHansen BB, Sørensen TH, Bech M: Variation in utilization of health care services in general practice in Denmark. University of Southern Denmark, Institute of Public Health–Health Economics. 2008. Reference Source\n\nWonderling D: Introduction to health economics. McGraw-Hill Education (UK); 2011. Reference Source\n\nMcHugh MD, Aiken LH, Eckenhoff ME, et al.: Achieving Kaiser Permanente quality. Health Care Manage Rev. 2016; 41(3): 178–88. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaeda T, Babazono A, Nishi T, et al.: Investigation of the Existence of Supplier-Induced Demand in use of Gastrostomy Among Older Adults: A Retrospective Cohort Study. Medicine (Baltimore). 2016; 95(5): e2519. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeeks WB, Jardin M, Dufour JC, et al.: Geographic variation in admissions for knee replacement, hip replacement, and hip fracture in France: evidence of supplier-induced demand in for-profit and not-for-profit hospitals. Med care. 2014; 52(10): 909–17. PubMed Abstract | Publisher Full Text\n\nShigeoka H, Fushimi K: Supplier-induced demand for newborn treatment: evidence from Japan. J Health Econ. 2014; 35: 162–78. PubMed Abstract | Publisher Full Text\n\nWatts BV, Shiner B, Klauss G, et al.: Supplier-induced demand for psychiatric admissions in Northern New England. BMC psychiatry. 2011; 11: 146. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith R: Why medicine is overweight. Don’t forget inconvenient truth of supplier induced demand. BMJ. 2010; 340: c3334. PubMed Abstract | Publisher Full Text\n\nIssel LM: Forty years: a brief history of health care considered. Health Care Manage Rev. 2015; 40(1): 1. PubMed Abstract | Publisher Full Text\n\nKarimi S, Khorasani E, Keyvanara M, et al.: Factors affecting physicians' behaviors in induced demand for health services. Int J Educ Psychol Res. 2015; 1(1): 43–51. Publisher Full Text\n\nCutler D, Skinner J, Stern AD, et al.: Physician beliefs and patient preferences: a new look at regional variation in health care spending. Nat Bur Econ Res. 2013. Publisher Full Text\n\nYuda M: Medical fee reforms, changes in medical supply densities, and supplier-induced demand: Empirical evidence from Japan. Hitotsub J Econ. 2013; 54(1): 79–93. Publisher Full Text\n\nDelattre E, Dormont B: Fixed fees and physician-induced demand: a panel data study on French physicians. Health Econ. 2003; 12(9): 741–54. PubMed Abstract | Publisher Full Text\n\nMagazzino C, Mele M: The determinants of health expenditure in Italian regions. Int J Econ Finance. 2012; 4(3): 61–72. Publisher Full Text\n\nMahbubi M, Ojaghi S, Ghiyasi M, et al.: Supplemental insurance and induce demand in veterans. Med Veterans J. 2010; 2(8): 18–22.\n\nMahmoudi G, Ghanbar R, Asadi M, et al.: Dataset 1 in: Factors associated with induced demand for services in Iran’s healthcare system. F1000Research. 2018. Data Source\n\nSekimoto M, Ii M: Supplier-induced demand for chronic disease care in Japan: multilevel analysis of the association between physician density and physician-patient encounter frequency. Value Health Reg Issues. 2015; 6: 103–10. Publisher Full Text\n\nCrivelli L, Filippini M, Mosca I: Federalism and regional health care expenditures: an empirical analysis for the Swiss cantons. Health Econ. 2006; 15(5): 535–41. PubMed Abstract | Publisher Full Text\n\nChang CH, Stukel TA, Flood AB, et al.: Primary care physician workforce and Medicare beneficiaries' health outcomes. JAMA. 2011; 305(20): 2096–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmporfu E: Private hospital accreditation and inducement of care under the ghanaian national insurance scheme. Health Econ Rev. 2011; 1(1): 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichardson JR, Peacock SJ: Supplier-induced demand: reconsidering the theories and new Australian evidence. Appl Health Econ Health Policy. 2006; 5(2): 87–98. PubMed Abstract | Publisher Full Text\n\nLien HM, Albert Ma CT, McGuire TG: Provider-client interactions and quantity of health care use. J Health Econ. 2004; 23(6): 1261–83. PubMed Abstract | Publisher Full Text\n\nLeone T: Demand and supply factors affecting the rising overmedicalization of birth in India. Int J Gynaecol Obstet. 2014; 127(2): 157–62. PubMed Abstract | Publisher Full Text\n\nBazyar M, Soofi M, Rashidian A: Ways to Control Moral Hazard in Health System: Demand-Side and Supply-Side Interventions (Review Article). Tolooe Behdasht. 2012; 11(1): 110–22. Reference Source\n\nHuang K, Tao F, Bogg L, et al.: Impact of alternative reimbursement strategies in the new cooperative medical scheme on caesarean delivery rates: a mixed-method study in rural China. BMC Health Serv Res. 2012; 12(1): 217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndersen LB, Serritzlew S: Type of services and supplier-induced demand for primary physicians in Denmark. 2007. Reference Source\n\nBorhanzade A: Induced demand and the cost of tests and its impact on cost and family health. Iranian Association of Clinical Laboratory Doctors. 2011. Reference Source\n\nKim J, Ko S, Yang B: The effects of patient cost sharing on ambulatory utilization in South Korea. Health Policy. 2005; 72(3): 293–300. PubMed Abstract | Publisher Full Text\n\nMadden D, Nolan A, Nolan B: GP reimbursement and visiting behaviour in Ireland. Health Econ. 2005; 14(10): 1047–60. PubMed Abstract | Publisher Full Text\n\nDosoretz AM: Reforming Medicare IMRT (intensity modulated radiation therapy) reimbursement rates: A study investigating increasing IMRT utilization rates and doctors' incentives. Tufts University; 2011. Reference Source\n\nFerguson BS: Issues in the demand for medical care: Can consumers and Doctors be trusted to make the right choices? Atlantic Institute for Market Studies; 2002.\n\nVan De Voorde C, Van Doorslaer E, Schokkaert E: Effects of cost sharing on physician utilization under favourable conditions for supplier‐induced demand. Health Econ. 2001; 10(5): 457–71. PubMed Abstract | Publisher Full Text\n\nBhatia T: An empirical analysis of physician prescription behavior. Northwestern University; 2006.\n\nPalesh M, Tishelman C, Fredrikson S, et al.: \"We noticed that suddenly the country has become full of MRI\". Policy makers' views on diffusion and use of health technologies in Iran. Health Res Policy Syst. 2010; 8(1): 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBickerdyke I, Dolamore R, Monday I, et al.: Supplier-induced demand for medical services. Canberra: Productivity Commission Staff Working Paper. 2002. Reference Source"
}
|
[
{
"id": "53621",
"date": "27 Sep 2019",
"name": "Hanna Tiirinki",
"expertise": [
"Reviewer Expertise Health management",
"health service research",
"health quality",
"Lean",
"primary health care"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 allowing me to have the opportunity to review interesting manuscript \"Factors associated with induced demand for service in Iran's healthcare system\". Overall, the study is well organized. Study is a cross-sectional descriptive-analytical and the data has collected by a questionnaire. Analyze were done using SPSS.\n\nAs a whole, the paper has well written. In the background section of the manuscript, it would be good to open a bit Iran's healthcare system, especially for wide international readers to understand better the context of research. Please clarify the value of this research for international and national level. I agree there is a significant value of this kind of health economic and management research for the Iran's healthcare. But please clarify how.\n\nPlease, be critical with the cited literature. Literature cited in article are partly quite old.\nMethods: Overall, 511 subjects consisting of 78 physicians, 350 nurses, 26 laboratory experts and 28 radiology technicians. Still 29 is missing? Clarify please.\nResults: The tables of the paper are clear. It would be good to have some deeper understanding for the interesting results of this paper, comparing chosen method \"descriptive-analytical\".\n\nDiscussion: Authors do a comparisons and reflection between results of the present study and those obtained in other similar cited studies.\nAuthors explicites what this paper adds to knowledge of induced demand in healthcare. Authors should also explicit what it adds to Iran's healthcare and from what perspective the research of this topic need to be continue.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-978
|
https://f1000research.com/articles/7-521/v1
|
01 May 18
|
{
"type": "Research Article",
"title": "Enterobacter hormaechei subsp. hoffmannii subsp. nov., Enterobacter hormaechei subsp. xiangfangensis comb. nov., Enterobacter roggenkampii sp. nov., and Enterobacter muelleri is a later heterotypic synonym of Enterobacter asburiae based on computational analysis of sequenced Enterobacter genomes.",
"authors": [
"Granger G. Sutton",
"Lauren M. Brinkac",
"Thomas H. Clarke",
"Derrick E. Fouts",
"Lauren M. Brinkac",
"Thomas H. Clarke",
"Derrick E. Fouts"
],
"abstract": "Background: The predominant species in clinical Enterobacter isolates is E. hormaechei. Many articles, clinicians, and GenBank submissions misname these strains as E. cloacae. The lack of sequenced type strains or named species/subspecies for some clades in the E. cloacae complex complicate the issue. Methods: The genomes of the type strains for Enterobacter hormaechei subsp. oharae, E. hormaechei subsp. steigerwaltii, and E. xiangfangensis, and two strains from Hoffmann clusters III and IV of the E. cloacae complex were sequenced. These genomes, the E. hormaechei subsp. hormaechei type strain, and other available Enterobacter type strains were analysed in conjunction with all extant Enterobacter genomes in NCBI’s RefSeq using Average Nucleotide Identity (ANI). Results: There were five recognizable subspecies of E. hormaechei: E. hormaechei subsp. hoffmannii subsp. nov., E. hormaechei subsp. xiangfangensis comb. nov., and the three previously known subspecies. One of the strains sequenced from the E. cloacae complex was not a novel E. hormaechei subspecies but rather a member of a clade of a novel species: E. roggenkampii sp. nov.. E. muelleri was determined to be a later heterotypic synonym of E. asburiae which should take precedence. Conclusion: The phylogeny of the Enterobacter genus, particularly the cloacae complex, was re-evaluated based on the type strain genome sequences and all other available Enterobacter genomes in RefSeq.",
"keywords": [
"Enterobacter",
"hormaechei",
"steigerwaltii",
"oharae",
"xiangfangensis",
"hoffmannii",
"roggenkampii",
"Prokaryote Code"
],
"content": "Introduction\n\nThe name Enterobacter hormaechei was created for a taxon at the rank of species that had previously been called Enteric Group 75. O’Hara et al1. defined the type strain to be ATCC 49162T from the 23 strains they studied. Twelve of the strains were shown to be closely related via DNA-DNA hybridization (DDH) and less closely related to other Enterobacter species. Numerous biochemical assays were performed on the 23 strains to characterize and differentiate the new species.\n\nHoffmann and Roggenkamp2 investigated the genetic structure of the E. cloacae complex (the set of species included in this complex has varied over time) by a combination of sequencing of the three housekeeping genes hsp60, rpoB, and hemB; and PCR-restriction fragment length polymorphism (PCR-RFLP) analysis of ampC. They defined 12 genetic clusters (I-XII) based most exhaustively on the hsp60 sequencing. Three of the clusters (cluster III, 58 strains; cluster VI, 28 strains; and cluster VIII, 59 strains) accounted for 70% of the 206 strains studied. The authors noted that “Only 3% of our study strains clustered with the type strain of E. cloacae.” (cluster XI), “We found that 3% of our study strains clustered around the E. hormaechei type strain.” (cluster VII), and “Our clusters VI and VIII were closely related to E. hormaechei cluster VII. DDH studies are needed to verify whether these clusters form a common DNA relatedness group allowing emending and broadening of the species description of E. hormaechei.”.\n\nHoffmann et al3. followed up with a characterization of clusters VI, VII, and VIII asserting based on DDH that these clusters were subspecies of the same species. Since cluster VII contained the type strain for E. hormaechei Hoffmann et al. named cluster VII E. hormaechei subsp. hormaechei, cluster VI E. hormaechei subsp. oharae, and cluster VIII E. hormaechei subsp. steigerwaltii. Forty-eight strains were characterized using 129 biochemical tests showing that there were phenotypic differences between the subspecies. Unfortunately the authors did not decide to include the other predominant cluster (III) in their analysis, nor did they validly publish these subspecies names. This was rectified recently in Validation List no. 1724.\n\nGu et al5. defined E. xiangfangensis using a phylogenetic tree based upon concatenated partial rpoB, atpD, gyrB and infB gene sequences from a novel isolate and existing type strains where E. xiangfangensis grouped closest to E. hormaechei. Biochemical assays were performed and E. xiangfangensis strains were differentiable from the E. hormaechei type strain.\n\nDuring analysis of the E. cloacae complex and E.(now Klebsiella6) aerogenes strains looking at antimicrobial resistance patterns7, many of the Hoffmann et al. clusters were rediscovered using whole genome comparisons such as SNP analysis and average nucleotide identity (ANI). The clusters were identifiable by the hsp60 sequences deposited by the Hoffmann group. The three subspecies of E. hormaechei defined by Hoffmann et al. fell within the expected ANI range for bacterial species, being greater than 95% ANI between subspecies and greater than 98% ANI within a subspecies. Unexpectedly Hoffmann cluster III also met the ANI criteria to be an E. hormaechei subspecies. Further, genomes named E. xiangfangensis in GenBank fell within the E. hormaechei subsp. steigerwaltii cluster rather than a separate cluster. Moreover, most of the genomes in these clusters were mistakenly identified as E. cloacae when they were submitted to GenBank. To resolve the naming inconsistencies of these genomes the type strains for E. hormaechei subsp. steigerwaltii, E. hormaechei subsp. oharae, E. xiangfangensis, Hoffmann cluster III, and Hoffmann cluster IV were sequenced.\n\n\nBackground\n\nTools for bacterial species assignment have changed over time8,9. Initially, morphology as viewed through a microscope and later aided by staining such as Gram staining10 to distinguish cell wall differences was used. Biochemical assays and other methods to determine phenotype followed. Use of the genome started with DNA-DNA hybridization (DDH) where a 70% threshold for species followed later by a 79% threshold for subspecies were proposed. Widespread use of marker genes in particular the 16S rRNA gene made assays easier. A threshold of less than 97% identity for the 16S rRNA gene was used to determine a new species but values above 97% could not guarantee that isolates were the same species. The sequence of other less conserved marker genes such as hsp60 has also been used to differentiate species. More recently multiple marker genes are sequenced and a combined alignment is used. With the advent of inexpensive genome sequencing, computing average nucleotide identity (ANI), which correlates very closely with DDH, has largely supplanted other methods. Studies have shown that an ANI threshold between 94-96.5% correlates well with existing species definitions and 97-98% for subspecies11–19. DDH has been shown to not only correlate with ANI but also with how many of the genes or what fraction of the genomes are shared in common so some ANI based tools take this measurement into account as well17–19. Most definitions of new species involve sequencing the genome and taking ANI and shared gene content into account in some fashion but many species definitions predate genome sequencing and some type strains have not been sequenced. There is no generally accepted method for reconciling older species definitions with genome comparisons but usually ANI and shared gene content form a basis for the analysis.\n\nAs Hoffmann2,3 and others20–26 discovered the predominant species in clinical Enterobacter isolates is E. hormaechei. Unfortunately many articles, clinicians, and GenBank submissions misname these strains as E. cloacae perhaps as a short hand for the E. cloacae complex and possibly due to the E. hormaechei subspecies not being validly published until recently. Another issue was the lack of sequenced type strains or named species/subspecies for some clades. The definition of what species/subspecies make up the E. cloacae complex has been in flux2,27,28 and even what species are in the genus Enterobacter29–31.\n\nThe E. cloacae complex was shown to have 18 clades (A-R)7, 12 of which corresponded to 11 of the 12 clusters defined previously by Hoffmann2. Hoffmann cluster X is E. nimipressuralis which has been reclassified as Lelliottia nimipressuralis29. Table 1 incorporates more recently sequenced genomes and published papers adding four clades (S-V) and incorporating the latest literature. For example, clade R (Hoffmann cluster IX) was recently defined to be E. bugandensis31.\n\nE. lignolyticus and E. timonensis have not been validly published and are deemed to be outside of the E. cloacae complex. E. siamensis and E. tabaci do not have sequenced genomes but based on their 16S rRNA genes may be in the E. cloacae complex. Proxy indicates whether a type or proxy strain was available. The last two columns are for the clade (A-V) and Hoffmann cluster (I-XII).\n\n\nResults\n\nAll RefSeq genomes labelled as being in the genus Enterobacter were downloaded from NCBI RefSeq resulting in 1,249 genomes. A fast approximate ANI tool, called MASH32, was used to generate a pairwise ANI based distance matrix and average linkage hierarchical clustering was used to generate the tree shown in Figure 1. 1,216 genomes were assigned to 22 clades (A-V Table 1) in the E. cloacae complex (Supplemental Table 1) while 30 genomes were deemed to be outliers and not in the Enterobacter genus (best MASH matches in Supplemental Table 2) as well as 2 E. lignolyticus genomes and 1 E. timonensis genome deemed to be outside of the E. cloacae complex. Two species of Enterobacter: E. siamensis and E. tabaci do not have sequenced genomes and their type strains’ 16S rRNA sequences while having full length matches at 98% and 99% respectively to some E. cloacae complex genomes did not have definitive matches to any particular clade. The type strains for E. asburiae and E. muelleri fall within the same clade (J – Hoffmann cluster I). All 78 genomes in this clade are above the 95% ANI species cut-off (Table 2) but using a 98% ANI subspecies cut-off produces 8 subclades of sizes 1, 1, 2, 2, 2 (E. muelleri), 3 (E. asburiae), 24, and 43. Thus E. muelleri33 is a later heterotypic synonym of E. asburiae34 which should take precedence. Whether the 8 subclades of E. asburiae should be treated as subspecies is beyond the scope of this paper.\n\nMean and standard deviation are shown above and the minimum and maximum pairwise values below. The last two rows show E. lignolyticus (Li) and E. timonensis (Ti) which have consistently lower ANI values.\n\nFive clades (A-E) are above the 95% ANI cut-off to be considered the same species (Table 2). Almost all within-clade pairwise ANIs are greater than between-clade ANIs (Table 2) and all genomes within a clade had the highest pairwise ANI to the type strain for that clade, supporting that these are distinct subspecies. Based on hsp60 sequences, clade A containing the E. xiangfangensis type strain is Hoffmann cluster VI; clade B containing the E. hormaechei subsp. steigerwaltii type strain is Hoffmann cluster VIII; clade C containing the E. hormaechei subsp. oharae type strain is also Hoffman cluster VI; clade D containing the Hoffmann cluster III type strain (proposed name E. hormaechei subsp. hoffmannii subsp. nov.) is Hoffmann cluster III; and clade E containing the E. hormaechei subsp. hormaechei type strain is Hoffmann cluster VII.\n\nTo explore the gene content differences of the E. cloacae complex and the E. hormaechei subspecies in particular, the pan-genome of the 1216 E. cloacae complex genomes was determined using PanOCT35. The pan-genome generates orthologous gene clusters that delineate which genes are in common between the clades and which genes differentiate the clades (Supplemental Table 3 and Supplemental Table 4). There were 2966 genes in “common to all” of the clades (present in 90% of the genomes of each clade). The number of genes “specific to” a clade (present in 90% of the genomes of that clade and in less than 10% of genomes from any other clade) varied from 0 (L) to 465 (V). The number of genes “missing from” a clade (present in less than 10% of the genomes of that clade and present in at least 90% of the genomes of all other clades) varied from 0 (A,C,H,K,O) to 40 (U). While ANI is the primary determinant of drawing distinctions between species and subspecies, gene content plays a role in generating phenotypic differences which might rationalize segregating a clade from species into subspecies. The clades which represent named species and subspecies show no qualitative difference in gene content from clades with no named species (Supplemental Table 4). In particular, clade D which is the proposed E. hormaechei subsp. hoffmannii has more genes specific to it than 3 of the 4 recognized subspecies. The gene content numbers need to be looked at carefully since they depend on the number of genomes in a clade (T has 187 clade specific genes but this is based on a single genome which means it is really strain specific genes rather than species specific), the distance from other clades (V the most distant clade has 465 specific genes and also has only 3 genomes), and sampling bias such as if most genomes in a clade are from a clonal outbreak. ANI appears to have less of these subjective issues to deal with.\n\nBiochemical and other properties of the E. hormaechei subspp. clades have been previously published3,5 except for clade D. With the availability of whole genome sequences and pan-genome analysis tools some of the observed phenotypic traits can be assigned to genetic features, such as the presence or absence of protein coding genes for known metabolic pathways. E. hormaechei subsp. hormaechei was previously distinguished from E. hormaechei subsp. oharae and E. hormaechei subsp. steigerwaltii by growth on dulcitol (a.k.a. galactitol) as the sole carbon source3. This phenotype can be explained by the presence of a gat operon7,36 exclusively within the hormaechei subsp.. Also, in the same genomic location, between the D-galactarate dehydratase gene and the 16S rRNA methyltransferase gene, the steigerwaltii, oharae, xiangfangensis, and hoffmannii subspp. have a related, but different operon, encoding for N-acetyl galactosamine metabolism (a.k.a., the aga locus)7,37. Similarly, steigerwaltii isolates can be distinguished from hormaechei, oharae, xiangfangensis, and hoffmannii by their ability to grow on adonitol (a.k.a. ribitol) and D(+)-arabitol; both 5 carbon sugar alcohols known as penitols. The rbt and dal operons known from Klebsiella aerogenes, which metabolize ribitol and D(+)-arabitol respectively7,38, account for this difference and are found almost exclusively in steigerwaltii. E. hormaechei subsp. hoffmannii has 25 clade specific genes 10 of which (clusters 28856-28865 Supplemental Table 3) occur as a unit between core clusters (16694-5) and another 6 (15153-15156, 27141-2) occur between core clusters (17653-4). These clusters have no or vague annotation but are intriguing targets to provide functional phenotypic differences.\n\n\nMethods\n\nMASH32 is a very fast tool for determining approximate pairwise ANI values given sequenced genomes. A PERL script was used to invoke the following command to generate a set of MASH (version 2.0) sketches of k-mer size 16 for the 1249 downloaded Enterobacter genomes:\n\nmash sketch -k 16 -o Enter.Sketch.file [List of the Genomes]\n\nThe resulting sketches file was then used to compare all the genomes against each other with an additional PERL script which calls MASH (version 2.0) with the command:\n\nMash dist Enter.Sketch.file [List of the Genomes]\n\nwhich generated data that could be extracted into an all versus all ANI comparison (Supplemental Table 5). We used the GGRaSP R package (version 1.0) which generated an ultramateric tree by using the R hclust function with average linkage from the distance matrix calculated by subtracting 100 from the MASH ANI results. The result was translated into Newick format with the APE39 R package (Supplemental File 1) rendered with metadata annotated using the Interactive Tree of Life40 into Figure 1.\n\nBased on the tree 30 genomes were deemed to be outliers and probably not in the Enterobacter genus as well as 2 E. lignolyticus genomes and 1 E. timonensis genome deemed to be outside of the E. cloacae complex. These 30 genomes were compared to all genome sequenced bacterial type strains from NCBI RefSeq (Supplemental Table 2) using MASH which confirmed that these genomes were likely misnamed as Enterobacter. The decision to leave E. lignolyticus and E. timonensis outside of the E. cloacae complex was based on two reasons: historically neither has been included in the complex, and there is a quantitative difference in the mean ANI values between genomes of these two species and genomes included in the 22 clades within the complex (last two rows of Table 2). The highest mean ANI for E. lignolyticus and E. timonensis to genomes included in the 22 clades within the complex is 86.2% for E. timonensis to clade S; whereas, the lowest mean ANI within the complex is 86.5% between clades P and U.\n\nFrom the all versus all MASH ANI comparison GGRaSP was used to generate average linkage clusters and the medoids of those clusters at both the 95% (species) and 98% (subspecies) levels. If type strains existed at the subspecies level those clusters were used (E. hormaechei and E. cloacae) otherwise species level clusters were used resulting in 22 clades (A-V). If a type strain genome sequence existed for a clade it was selected otherwise the medoid was selected as a proxy. The one exception for this was clade J where two different type strains existed: E. asburiae and E. muelleri where both were retained for the typing. These 23 representative genomes were used to “type” all 1216 Enterobacter cloacae complex genomes (Supplemental Table 1). For typing the best MASH ANI match was used and resolved to either the species or subspecies level. As expected the typing was in complete agreement with the clades in the MASH ANI tree (Figure 1). The MASH sketches for these 22 clade representatives (after removing the redundant E. muelleri) can be used as a fast categorization tool for novel Enterobacter cloacae complex genomes.\n\nGGRaSP was similarly used to select the 250 most diverse genomes including the outliers from the 1249 downloaded genomes while eliminating very closely related genomes. PanOCT35,41 run at the nucleotide level was used to generate the orthologous clusters for a pan-genome. The primary use of this was to validate the approximate MASH ANI values. PanOCT determines pairwise ANI values by looking at every orthologous cluster shared by a pair of genomes. The percent identity of each match is weighted by the length of the match, summed over all relevant clusters, and divided by the sum of match lengths which is consistent with previous calculations of ANI. Supplemental Figure 1 shows that the MASH ANI estimate is very strongly correlated (98.9) with the PanOCT ANI measurement. For PanOCT ANI values greater than 94% the estimate is very tight (mean error 0.34±0.22) versus less than 94% (1.15±0.70). The clades and tree at the clade level remained the same using PanOCT ANI values.\n\nFor the PanOCT run with 1,216 genomes to determine gene content similarities, PanOCT was run as part of the JCVI pan-genome pipeline in hierarchical fashion with the following batches of genomes run by PanOCT at level 1: (combined 3 E. mori, 3 E. soli, 8 E. cancerogenus, 8 E. cloacae complex clade K, 13 E. cloacae complex clade L, 11 E. cloacae complex clade N, 4 E. cloacae complex clade O, 4 E. cloacae complex clade P, 5 E. cloacae complex clade S, 1 E. cloacae complex clade T); (combined 45 E. cloacae subsp. cloacae, 9 E. cloacae subsp. dissolvens); (randomly split into 4 groups 169 E. hormaechei subsp. hoffmannii); (7 E. hormaechei subsp. hormaechei); (68 E. hormaechei subsp. oharae); (randomly split into 8 groups 325 E. hormaechei subsp. steigerwaltii); (randomly split into 6 groups 255 E. hormaechei subsp. xiangfangensis); (78 E. asburiae); (30 E. bugandensis); (71 E. kobei); (29 E. ludwigii); and (70 E. roggenkampii). The level 1 clusters were then combined using PanOCT at level 2 and the final output generated using the PanOCT (version 3.27) command line:\n\npanoct.pl -R matchtable.txt -f genomes.list -g combined.att_file -P combined.fasta -b final_panoct_run -c 0,95\n\nThe diverse 250 genome PanOCT run and the level 1 PanOCT batch runs used the PanOCT (version 3.27) command line:\n\npanoct.pl -b results -t combined.blast -f genomes.list -g combined.att -P combined.fasta -S yes -L 1 -M Y -H Y -V Y -N Y -F 1.33 -G y -c 0,50,95,100 -T\n\nThe hierarchical PanOCT run of 1,216 genomes produced a matrix of orthologous gene clusters (Supplemental Table 3) where the rows are clusters and the columns are genomes with the cells containing the RefSeq IDs for the gene from the corresponding genome. This matrix was used to determine genes common to all, specific to, and missing from clades A-V. Individual PanOCT runs were also done for clade J, D, and M. Clade J to insure that PanOCT ANI values confirmed MASH ANI values that E. asburiae and E. muelleri are the same species which they did and these ANI values were used to determine the 8 subclades at 98% ANI using hierarchical clustering (hclust in R) average linkage. Clade D to confirm the MASH ANI values for E. hormaechei subsp. hoffmannii which they did. Clade M was done likewise to confirm E. roggenkampii which they did.\n\n\nDiscussion\n\nThe Background section reviews how the tools for defining a species have evolved. In a recent review of the genus Mycobacterium, the authors proposed that any newly defined bacterial species must have a genome sequence and an ANI comparison carried out against existing sequenced type strains to justify a novel species assignment42. ANI analysis should not be relied on in isolation for defining a species since historical or clinical phenotypic distinctions may be important for example in distinguishing between E. coli and Shigela which by ANI are the same species. However, genome sequencing appears to be outstripping the taxonomic definition of species within some genera. For the 22 clades of the E. cloacae complex identified here 9 do not have named type strains (7 if the two proposed here are adopted). For important pathogens where clinical practice may rely on proper classification the ability to name these clades/species and provide resources for identifying them could be pivotal. Unfortunately, the current established journal for validly publishing bacterial species’ names insists on phenotypic characterization and deposition of the type strain before naming is valid. This prevents computational based methods from moving quickly. Paradoxically almost all species identifying diagnostic tests are genotype not phenotype based so genotype is good enough for diagnosis but not species definition. Further, delineating what is acceptable to define as a new species is also genotype not phenotype based whether via DDH, marker genes, or more recently ANI. Worse there are no published standards for what defines the minimal set of phenotypic biochemical assays that must be performed. As the Mycobacterium review authors state: “The easy and affordable availability of reliable whole-genome sequences raises doubts about the real added value of investigating phenotypic traits when a new species is described. Actually, different taxonomists use their own panels of tests, often not standardized, to produce results of no use for colleagues and absolutely incomprehensible to the community of mycobacteriologists who have dismissed such approach since the ‘90s. For the genus Mycobacterium the major phenotypic traits that cannot be disregarded should include growth rate and pigmentation of colonies, while the classical investigation of biochemical activities is clearly obsolete.”. If there were accepted standards for minimal phenotypic characterization then culture collection repositories could choose to provide the characterization as fee for service or even for free for type strains as an incentive for deposition. With the rapid growth in synthetic genomics capabilities one could argue that the deposition of a high quality complete genome might suffice rather than a culture. We propose allowing “placeholder” species or subspecies names such as “E. cloacae complex clade S” in order to enable the most robust use of computational and genomic resources for clinical diagnosis while awaiting the designation and deposition of a type strain with a valid name possibly with some minimal phenotypic characterization.\n\n\nConclusion\n\nComputational analysis supports the reassignment of E. xiangfangensis to E. hormaechei subsp. xiangfangensis. We propose to name clade D / Hoffman cluster III as E. hormaechei subsp. hoffmannii in honor of Harald Hoffmann’s work elucidating the phylogenetic structure of the E. cloacae complex2 in particular the subspecies of E. hormaechei3. We propose to name clade M / Hoffmann cluster IV Enterobacter roggenkampii after Andreas Roggenkamp for his work on elucidating the phylogenetic structure of the E. cloacae complex2. The analysis also shows that E. muelleri33 is a later heterotypic synonym of E. asburiae34 which should take precedence.\n\nE. hormaechei subsp. xiangfangensis (xi.ang.fang.en′sis. N.L. gen. m. adj. xiangfangensis pertaining to Xiangfang, a district located in Harbin, Heilongjiang Province, where the bacterium was first isolated).\n\nBasonym: Enterobacter xiangfangensis5.\n\nThe species description is unchanged from its description as Enterobacter xiangfangensis5.\n\nThe type strain is strain 10–17T ( = LMG 27195T = NCIMB 14836T = CCUG 62994T), isolated from traditional sourdough in Heilongjiang Province, China.\n\nThe GenBank accessions for the complete genome sequence of E. hormaechei subsp. xiangfangensis are PRJNA259658, SAMN05581746, ASM172978v1, and CP017183.1.\n\nE. hormaechei subsp. hoffmannii (hoff.mannʹi.i. N.L. gen. m. Hoffmann, in honor of Harald Hoffmann, a German microbiologist who helped elucidate the phylogenetic structure of the E. cloacae complex in particular the subspecies of E. hormaechei).\n\nHoffmann and Roggenkamp2 determined clusters within the E. cloacae complex using marker genes, primarily hsp60. Hoffman et al3. followed up on three closely grouping clusters to define the three current subspecies of E. hormaechei based on DDH and phenotypic tests. Chavda et al7. determined groups for the E. cloacae complex using SNPs from whole genome alignments. ANI analysis showed that the Chavda groups were highly similar at levels associated with species or subspecies groupings. This paper performs a more detailed analysis of gene content and ANI across a larger set of genomes supporting the Chavda groups A-E as E. hormaechei subspecies. E. hormaechei subsp. hoffmannii subsp. nov. has similar gene content and ANI characteristics as the previously defined four subspecies.\n\nHoffmann deposited the type strain, EN-114, for Enterobacter hormaechei subsp. hoffmannii in Leibniz-Institut DSMZ-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, accession DSM-14563, and recently the strain was also deposited in BCCM/LMG Bacteria Collection, accession LMG-30171. The GenBank accessions for the complete genome sequence are PRJNA259658, SAMN05581748, ASM172974v1, CP017186.1, and CP017187.1.\n\nAccording to2, the strain was isolated from the respiratory tract of a clinical patient. The DSMZ database indicates that the sample was isolated prior to 2002 in Bavaria, Germany.\n\nE. roggenkampii (rog.gen.kampʹi.i. N.L. gen. m. Roggenkamp, in honor of Andreas Roggenkamp, a German microbiologist who helped elucidate the phylogenetic structure of the E. cloacae complex).\n\nHoffmann and Roggenkamp2 determined clusters within the E. cloacae complex using marker genes, primarily hsp60. Chavda et al7. determined groups for the E. cloacae complex using SNPs from whole genome alignments. ANI analysis showed that the Chavda groups were highly similar at levels associated with species or subspecies groupings. Enterobacter roggenkampii sp. nov. is the type strain for Hoffmann cluster IV and Chavda group M. This paper performs a more detailed analysis of gene content and ANI across a larger set of genomes supporting the Chavda groups A-R and adding S-V. E. roggenkampii sp. nov. has similar gene content and ANI characteristics as previously defined species in the E. cloacae complex.\n\nHoffmann deposited the type strain, EN-117, for Enterobacter roggenkampii in Leibniz-Institut DSMZ-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, accession DSM-16690, and recently the strain was also deposited in BCCM/LMG Bacteria Collection, accession LMG-30172. The GenBank accessions for the complete genome sequence are PRJNA259658, SAMN05581750, ASM172980v1, CP017184.1, and CP017185.1.\n\nAccording to2, the strain was isolated from the stool of a clinical patient. The DSMZ database indicates that the sample was isolated in 2000 in Germany.\n\nThe GenBank accessions for the complete genome sequence of E. hormaechei subsp. steigerwaltii are PRJNA259658, SAMN05581751, ASM172972v1, and CP017179.1.\n\nThe GenBank accessions for the complete genome sequence of E. hormaechei subsp. oharae are PRJNA259658, SAMN05581749, ASM172970v1, and CP017180.1.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work has been funded in whole or in part with federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services under award number U19AI110819.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank: Jason Inman from JCVI for help with pan-genome runs; Karen Beeri, Karrie Goglin, and Kelly Colt from the JCVI sequencing core for growth and sequencing of the type strains; and Elke Lang and Claudine Vereecke for help getting the type strains into the BCCM/LMG Bacteria Collection.\n\n\nSupplementary material\n\nSupplemental Table 1. ANI clades compared to MASH best match assignment for 1,216 Enterobacter cloacae complex genomes.\n\nClick here to access the data.\n\nSupplemental Table 2. MASH typing of 30 outlier genomes falling outside of the Enterobacter cloacae complex but labelled as Enterobacter in RefSeq.\n\nClick here to access the data.\n\nSupplemental Table 3. PanOCT generated orthologous clusters for 1,216 Enterobacter cloacae complex genomes. Rows are clusters, columns are genomes, cells contain RefSeq gene identifiers.\n\nClick here to access the data.\n\nSupplemental Table 4. Gene counts for genes common to all genomes, specifc to a clade, or missing from a clade.\n\nClick here to access the data.\n\nSupplemental Table 5. Pairwise MASH Average Nucleotide Identity (ANI) values for 1,249 genomes labelled Enterobacter in RefSeq.\n\nClick here to access the data.\n\nSupplemental Figure 1. Graph of MASH estimated versus PanOCT calculated Average Nucleotide Identity (ANI) for 250 representative genomes.\n\nClick here to access the data.\n\nSupplemental File 1. Newick formatted tree generated from Supplemental Table 5 and used to generate Figure 1.\n\nClick here to access the data.\n\n\nReferences\n\nO'Hara CM, Steigerwalt AG, Hill BC, et al.: Enterobacter hormaechei, a new species of the family Enterobacteriaceae formerly known as enteric group 75. J Clin Microbiol. 1989; 27(9): 2046–9. PubMed Abstract | Free Full Text\n\nHoffmann H, Roggenkamp A: Population genetics of the nomenspecies Enterobacter cloacae. Appl Environ Microbiol. 2003; 69(9): 5306–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoffmann H, Stindl S, Ludwig W, et al.: Enterobacter hormaechei subsp. oharae subsp. nov., E. hormaechei subsp. hormaechei comb. nov., and E. hormaechei subsp. steigerwaltii subsp. nov., three new subspecies of clinical importance. J Clin Microbiol. 2005; 43(7): 3297–303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOren A, Garrity GM: List of new names and new combinations previously effectively, but not validly, published. Int J Syst Evol Microbiol. 2016; 66(11): 4299–305. PubMed Abstract | Publisher Full Text\n\nGu CT, Li CY, Yang LJ, et al.: Enterobacter xiangfangensis sp. nov., isolated from Chinese traditional sourdough, and reclassification of Enterobacter sacchari Zhu et al. 2013 as Kosakonia sacchari comb. nov. Int J Syst Evol Microbiol. 2014; 64(Pt 8): 2650–6. PubMed Abstract | Publisher Full Text\n\nTindall BJ, Sutton G, Garrity GM: Enterobacter aerogenes Hormaeche and Edwards 1960 (Approved Lists 1980) and Klebsiella mobilis Bascomb et al. 1971 (Approved Lists 1980) share the same nomenclatural type (ATCC 13048) on the Approved Lists and are homotypic synonyms, with consequences for the name Klebsiella mobilis Bascomb et al. 1971 (Approved Lists 1980). Int J Syst Evol Microbiol. 2017; 67(2): 502–504. PubMed Abstract | Publisher Full Text\n\nChavda KD, Chen L, Fouts DE, et al.: Comprehensive Genome Analysis of Carbapenemase-Producing Enterobacter spp.: New Insights into Phylogeny, Population Structure, and Resistance Mechanisms. mBio. 2016; 7(6): pii: e02093-16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStaley JT: The bacterial species dilemma and the genomic-phylogenetic species concept. Philos Trans R Soc Lond B Biol Sci. 2006; 361(1475): 1899–909. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeorgiades K, Raoult D: Defining pathogenic bacterial species in the genomic era. Front Microbiol. 2011; 1: 151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBARTHOLOMEW JW, MITTWER T: The Gram stain. Bacteriol Rev. 1952; 16(1): 1–29. PubMed Abstract | Free Full Text\n\nKim M, Oh HS, Park SC, et al.: Towards a taxonomic coherence between average nucleotide identity and 16S rRNA gene sequence similarity for species demarcation of prokaryotes. Int J Syst Evol Microbiol. 2014; 64(Pt 2): 346–51. PubMed Abstract | Publisher Full Text\n\nKonstantinidis KT, Ramette A, Tiedje JM: The bacterial species definition in the genomic era. Philos Trans R Soc Lond B Biol Sci. 2006; 361(1475): 1929–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChan JZ, Halachev MR, Loman NJ, et al.: Defining bacterial species in the genomic era: insights from the genus Acinetobacter. BMC Microbiol. 2012; 12: 302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoris J, Konstantinidis KT, Klappenbach JA, et al.: DNA-DNA hybridization values and their relationship to whole-genome sequence similarities. Int J Syst Evol Microbiol. 2007; 57(Pt 1): 81–91. PubMed Abstract | Publisher Full Text\n\nRichter M, Rosselló-Móra R: Shifting the genomic gold standard for the prokaryotic species definition. Proc Natl Acad Sci U S A. 2009; 106(45): 19126–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang W, Du P, Zheng H, et al.: Whole-genome sequence comparison as a method for improving bacterial species definition. J Gen Appl Microbiol. 2014; 60(2): 75–8. PubMed Abstract | Publisher Full Text\n\nVarghese NJ, Mukherjee S, Ivanova N, et al.: Microbial species delineation using whole genome sequences. Nucleic Acids Res. 2015; 43(14): 6761–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeier-Kolthoff JP, Auch AF, Klenk HP, et al.: Genome sequence-based species delimitation with confidence intervals and improved distance functions. BMC Bioinformatics. 2013; 14: 60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nColston SM, Fullmer MS, Beka L, et al.: Bioinformatic genome comparisons for taxonomic and phylogenetic assignments using Aeromonas as a test case. mBio. 2014; 5(6): e02136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavin-Regli A, Bosi C, Charrel R, et al.: A nosocomial outbreak due to Enterobacter cloacae strains with the E. hormaechei genotype in patients treated with fluoroquinolones. J Clin Microbiol. 1997; 35(4): 1008–10. PubMed Abstract | Free Full Text\n\nPaauw A, Caspers MP, Leverstein-van Hall MA, et al.: Identification of resistance and virulence factors in an epidemic Enterobacter hormaechei outbreak strain. Microbiology. 2009; 155(Pt 5): 1478–88. PubMed Abstract | Publisher Full Text\n\nCampos LC, Lobianco LF, Seki LM, et al.: Outbreak of Enterobacter hormaechei septicaemia in newborns caused by contaminated parenteral nutrition in Brazil. J Hosp Infect. 2007; 66(1): 95–7. PubMed Abstract | Publisher Full Text\n\nWenger PN, Tokars JI, Brennan P, et al.: An outbreak of Enterobacter hormaechei infection and colonization in an intensive care nursery. Clin Infect Dis. 1997; 24(6): 1243–4. PubMed Abstract | Publisher Full Text\n\nOhad S, Block C, Kravitz V, et al.: Rapid identification of Enterobacter hormaechei and Enterobacter cloacae genetic cluster III. J Appl Microbiol. 2014; 116(5): 1315–21. PubMed Abstract | Publisher Full Text\n\nGuérin F, Isnard C, Sinel C, et al.: Cluster-dependent colistin hetero-resistance in Enterobacter cloacae complex. J Antimicrob Chemother. 2016; 71(11): 3058–61. PubMed Abstract | Publisher Full Text\n\nMorand PC, Billoet A, Rottman M, et al.: Specific distribution within the Enterobacter cloacae complex of strains isolated from infected orthopedic implants. J Clin Microbiol. 2009; 47(8): 2489–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMezzatesta ML, Gona F, Stefani S: Enterobacter cloacae complex: clinical impact and emerging antibiotic resistance. Future Microbiol. 2012; 7(7): 887–902. PubMed Abstract | Publisher Full Text\n\nPaauw A, Caspers MP, Schuren FH, et al.: Genomic diversity within the Enterobacter cloacae complex. PLoS One. 2008; 3(8): e3018. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrady C, Cleenwerck I, Venter S, et al.: Taxonomic evaluation of the genus Enterobacter based on multilocus sequence analysis (MLSA): proposal to reclassify E. nimipressuralis and E. amnigenus into Lelliottia gen. nov. as Lelliottia nimipressuralis comb. nov. and Lelliottia amnigena comb. nov., respectively, E. gergoviae and E. pyrinus into Pluralibacter gen. nov. as Pluralibacter gergoviae comb. nov. and Pluralibacter pyrinus comb. nov., respectively, E. cowanii, E. radicincitans, E. oryzae and E. arachidis into Kosakonia gen. nov. as Kosakonia cowanii comb. nov., Kosakonia radicincitans comb. nov., Kosakonia oryzae comb. nov. and Kosakonia arachidis comb. nov., respectively, and E. turicensis, E. helveticus and E. pulveris into Cronobacter as Cronobacter zurichensis nom. nov., Cronobacter helveticus comb. nov. and Cronobacter pulveris comb. nov., respectively, and emended description of the genera Enterobacter and Cronobacter. Syst Appl Microbiol. 2013; 36(5): 309–19. PubMed Abstract | Publisher Full Text\n\nStephan R, Grim CJ, Gopinath GR, et al.: Re-examination of the taxonomic status of Enterobacter helveticus, Enterobacter pulveris and Enterobacter turicensis as members of the genus Cronobacter and their reclassification in the genera Franconibacter gen. nov. and Siccibacter gen. nov. as Franconibacter helveticus comb. nov., Franconibacter pulveris comb. nov. and Siccibacter turicensis comb. nov., respectively. Int J Syst Evol Microbiol. 2014; 64(Pt 10): 3402–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoijad S, Imirzalioglu C, Yao Y, et al.: Enterobacter bugandensis sp. nov., isolated from neonatal blood. Int J Syst Evol Microbiol. 2016; 66(2): 968–74. PubMed Abstract | Publisher Full Text\n\nOndov BD, Treangen TJ, Melsted P, et al.: Mash: fast genome and metagenome distance estimation using MinHash. Genome Biol. 2016; 17(1): 132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKämpfer P, McInroy JA, Glaeser SP: Enterobacter muelleri sp. nov., isolated from the rhizosphere of Zea mays. Int J Syst Evol Microbiol. 2015; 65(11): 4093–9. PubMed Abstract | Publisher Full Text\n\nBrenner DJ, McWhorter AC, Kai A, et al.: Enterobacter asburiae sp. nov., a new species found in clinical specimens, and reassignment of Erwinia dissolvens and Erwinia nimipressuralis to the genus Enterobacter as Enterobacter dissolvens comb. nov. and Enterobacter nimipressuralis comb. nov. J Clin Microbiol. 1986; 23(6): 1114–20. PubMed Abstract | Free Full Text\n\nFouts DE, Brinkac L, Beck E, et al.: PanOCT: automated clustering of orthologs using conserved gene neighborhood for pan-genomic analysis of bacterial strains and closely related species. Nucleic Acids Res. 2012; 40(22): e172. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNobelmann B, Lengeler JW: Molecular analysis of the gat genes from Escherichia coli and of their roles in galactitol transport and metabolism. J Bacteriol. 1996; 178(23): 6790–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReizer J, Ramseier TM, Reizer A, et al.: Novel phosphotransferase genes revealed by bacterial genome sequencing: a gene cluster encoding a putative N-acetylgalactosamine metabolic pathway in Escherichia coli. Microbiology. 1996; 142( Pt 2): 231–50. PubMed Abstract | Publisher Full Text\n\nWu J, Anderton-Loviny T, Smith CA, et al.: Structure of wild-type and mutant repressors and of the control region of the rbt operon of Klebsiella aerogenes. EMBO J. 1985; 4(5): 1339–44. PubMed Abstract | Free Full Text\n\nParadis E, Claude J, Strimmer K: APE: Analyses of Phylogenetics and Evolution in R language. Bioinformatics. 2004; 20(2): 289–90. PubMed Abstract | Publisher Full Text\n\nLetunic I, Bork P: Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Res. 2016; 44(W1): W242–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChan AP, Sutton G, DePew J, et al.: A novel method of consensus pan-chromosome assembly and large-scale comparative analysis reveal the highly flexible pan-genome of Acinetobacter baumannii. Genome Biol. 2015; 16: 143. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTortoli E, Fedrizzi T, Meehan CJ, et al.: The new phylogeny of the genus Mycobacterium: The old and the news. Infect Genet Evol. 2017; 56: 19–25. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "33699",
"date": "17 May 2018",
"name": "Mark J. Pallen",
"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 in general a well written and well argued paper that represents a valuable addition to attempts to bring bacterial taxonomy into the genomic age. I can find no fault with the methodologies used nor with the general interpretation of results. I agree with the authors that all future bacterial taxonomy and nomenclature should be based on genomic data and they have fallen in line with an emerging consensus of how to make that work using ANI.\nIt is clear that bacterial taxonomy is broken and needs fixing and the only suitable response to the tyranny of The International Committee on Systematic Bacteriology is subversion by publishing papers like this that ignore its ridiculous and outdated requirements.\nTo quote Darwin: \"Our classifications will come to be, as far as they can be so made, genealogies\"\nI have just a handful of minor criticisms/suggestions for improvement:\nI don't see the need for the separate Introduction and Background sections. According to the guidelines for authors, papers in this journal should follow the usual IMRAD format, so I think that the two sections should simply become sub-sections of the Introduction, perhaps with brief explanatory headers.\n\nI am not sure why the authors abdicate responsibility for determining whether \"8 subclades of E. asburiae should be treated as subspecies\". Why not roll their approach out to cover these lineages too?\n\nThe authors discuss the concept of \"placeholder\" species and subspecies in the Discussion, but fail to mention the \"Candidatus\" designation, which is recognised by the current bacterial taxonomy apparatchiks:\nhttp://ijs.microbiologyresearch.org/content/journal/ijsem/10.1099/00207713-45-1-186\nhttps://en.wikipedia.org/wiki/Candidatus\nThey should include some discussion of this designation that includes a recognition of its major shortcoming in requiring phenotypic data in addition to genome sequence.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3654",
"date": "18 May 2018",
"name": "Granger Sutton",
"role": "Author Response",
"response": "The three reviewer suggestions all have merit and we will try to address them once more reviews are received. 1) We can certainly conform to IMRaD by using subheadings. 2) The issue of species versus subspecies should be addressed in the discussion. Our feeling was that when it is already problematic to validly publish names for species it is even more burdensome to do so for subspecies. What is the appropriate criteria to go to the trouble to differentiate subspecies: clinical significance, number of exemplars of each subspecies, and/or amount of core gene content difference between subspecies (this can only be determined once there are enough exemplars of each subspecies)? 3) We were unaware of the Candidatus designation and appreciate this being pointed out. While it does not appear to be a good fit for the case where genome sequences exist and species/subspecies are determined computationally since it was designed for environmental or unculturable samples with limited sequence data but at least some phenotypic or morphological data, it does suggest that some similar designation be used for \"placeholder\" names. We do not want to assign potentially permanent names with a notation indicating they are provisional but would like the name itself to indicate it is provisional and to be replaced when someone does the hard work of depositing a type strain and any required minimal phenotypic information. Again we should address this in the discussion."
}
]
},
{
"id": "34159",
"date": "07 Jun 2018",
"name": "Trinad Chakraborty",
"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\nConsiderable genome data is now available for isolates of many members of the family Enterobacteriaceae. As we move away from well-defined species such as E. coli and Salmonella, taxonomic assignments become blurred and there is now a great need to develop standardized tools for proper classification. A particular case is that of the species Enterobacter, where only 12 of the 35 historically classified species in this genus are valid.\nThe present manuscript reevaluates taxonomic allocation of members of the Enterobacter cloacae complex using whole genome sequences (WGS). It is important to remember that the dataset comprises primarily of draft genome sequences of varying quality and with only a very small number representing truly closed genomes.\nIsolates of the E. hormaechei complex are often associated with clinical disease. Based on the data from this study there are now two novel subspecies of E. hormaechei designated as E. hormaechei subsp. hoffmannii and E. hormaechei subsp. xiangfangensis respectively. In addition, a new species E. roggenkampii is proposed. Overall the study predicts the existence of 7 additional species within the genus Enterobacter.\nThe bulk of the analysis is based on a single tool viz. MASH-based ANI and is supplemented by the panOCT tool developed by the authors. The authors should consider the use of additional software tools to determine the overall genome-related index (OGRI).\nSpecific comments:\nClade A-E represent the five subspecies of E. hormaechei. The average nucleotide identity (ANI) for the clades A-D and E are at the borderline ANI-species definition.\n\nIn view of the fact that data is based mainly on draft genomes, the utility of supportive assignments based on the total numbers of unique genes must be considered carefully.\n\nFor such closely related clades, multi-tool-based analysis of taxonomy are helpful to reassure the claims. To support the species/subspecies distinction, particularly for those closely related clades, the use of widely used taxonomic tools such as the digital DNA-DNA hybridization tool, GGDC should be employed to strengthen the claims.\n\nANI values can vary when using different calculation tools as for e.g. with JSpecies and ANI calculator. The use of MASH algorithm leads to minor variation in ANI values and makes the borderline species definitions presented here difficult to interpret.\n\nTo confirm separation of E. timonensis and E. lignolyticus from the genus Enterobacter, comparison with members of the closest genera (for e.g., Klebsiella, Citrobacter etc.) should be added.\nFinally, biochemical and fermentation characteristics are key indicators for phenotypic characterization of isolates in diagnostic laboratories.\nThe final paragraph on biochemical properties is inadequate and could lead to confusion of phenotypes and undo the very purpose of the proposed classification scheme. Thus the gat operon is not exclusive to E. hormaechei subspecies hormaechei as stated, but is also present for e.g. in type strain E. bugandensis EB-247T.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3768",
"date": "29 Jun 2018",
"name": "Granger Sutton",
"role": "Author Response",
"response": "We thank Dr. Pallen for the thoughtful review and respond to issues below. “I don't see the need for the separate Introduction and Background sections. According to the guidelines for authors, papers in this journal should follow the usual IMRAD format, so I think that the two sections should simply become sub-sections of the Introduction, perhaps with brief explanatory headers.” We removed the Background and Conclusion section headings to conform to the IMRAD format. “I am not sure why the authors abdicate responsibility for determining whether \"8 subclades of E. asburiae should be treated as subspecies\". Why not roll their approach out to cover these lineages too?” We now address this in the Discussion section. “The authors discuss the concept of \"placeholder\" species and subspecies in the Discussion, but fail to mention the \"Candidatus\" designation, which is recognised by the current bacterial taxonomy apparatchiks: http://ijs.microbiologyresearch.org/content/journal/ijsem/10.1099/00207713-45-1-186 https://en.wikipedia.org/wiki/Candidatus They should include some discussion of this designation that includes a recognition of its major shortcoming in requiring phenotypic data in addition to genome sequence.” We thank Dr. Pallen for pointing this out to us and have included this in the Discussion section. We thank Dr. Chakraborty for the thoughtful review and respond to issues below. “The bulk of the analysis is based on a single tool viz. MASH-based ANI and is supplemented by the panOCT tool developed by the authors. The authors should consider the use of additional software tools to determine the overall genome-related index (OGRI).” and “For such closely related clades, multi-tool-based analysis of taxonomy are helpful to reassure the claims. To support the species/subspecies distinction, particularly for those closely related clades, the use of widely used taxonomic tools such as the digital DNA-DNA hybridization tool, GGDC should be employed to strengthen the claims.” We have included the comparison of GGDC to MASH and PanOCT ANI in the Methods section. “Clade A-E represent the five subspecies of E. hormaechei. The average nucleotide identity (ANI) for the clades A-D and E are at the borderline ANI-species definition.” This is certainly true but is also true of the already existing E. hormaechei subspecies: clade B E. hormaechei ssp. steigerwaltii, clade C E. hormaechei ssp. oharae, and clade E E. hormaechei ssp. hormaechei. While in the absence of previous taxonomic assignments one might choose to be reluctant to combine clades B, C, and E into a single species based on ANI because they have already been grouped as a species the borderline ANI values are not strong enough to argue for changing this. Given this adding clades A and D to E. hormaechei is strongly confirmed by the ANI values between clades A, B, C, and D. “In view of the fact that data is based mainly on draft genomes, the utility of supportive assignments based on the total numbers of unique genes must be considered carefully.” We have noted this concern in the results section. Gene content is not a primary consideration in our proposed new species designation but rather a possible reason to delineate at the subspecies level. In our experience most recent draft genome sequences are of high quality and the RefSeq genomes we used are screened by NCBI to meet certain quality requirements. Draft genome breaks tend to be at and due to repetitive elements such as transposons which would not affect the representation of most genes. We also try to take this into account by using a 90% rather than a 100% threshold. “ANI values can vary when using different calculation tools as for e.g. with JSpecies and ANI calculator. The use of MASH algorithm leads to minor variation in ANI values and makes the borderline species definitions presented here difficult to interpret.” ANI values for the newly proposed type strains were backed up by PanOCT ANI and now by GGDC and are not borderline except as consistent with previous taxonomy. “To confirm separation of E. timonensis and E. lignolyticus from the genus Enterobacter, comparison with members of the closest genera (for e.g., Klebsiella, Citrobacter etc.) should be added.” We have added this analysis to the Methods section. “Finally, biochemical and fermentation characteristics are key indicators for phenotypic characterization of isolates in diagnostic laboratories.” As the paper mentions we are not opposed to the biochemical characterization of type strains but need a standard that can be implemented by culture collections so that computationalists can acquire this data. The DSMZ for instance supports doing some of this characterization but does not claim it to be standard. In addition, DSMZ supports storing this characterization data in “The Bacterial Diversity Metadatabase” (BacDive) such as for the E. bugandensis type strain (https://bacdive.dsmz.de/strain/132404). What is interesting is that most biochemical characterization is not used to define a species in current practice. Researchers no longer collect phenotypic features and cluster based on a feature vector. Rather, genotypic characteristics are captured such as 16S or hsp60 or rpoB or WGS which are used to define a cluster of strains and then phenotypic characterization of those strains is performed and used as part of the species definition no matter how divergent those features may be. Computational taxonomy provides a structure by which strains can be clustered, named, referenced, discussed and compared to related clades. Biologists should follow up on clinically or otherwise interesting clades. We are not sure whether Dr. Chakraborty is arguing for historical consistency in what characterization is minimally required for a type strain or is arguing that there is little or no value in computational taxonomy without phenotypic characterization because it is required for clinical diagnosis. We would disagree with both since with the advent of whole genome sequences (or even DDH) phenotype is not needed to define species and clinical diagnosis can be done with molecular markers. “The final paragraph on biochemical properties is inadequate and could lead to confusion of phenotypes and undo the very purpose of the proposed classification scheme. Thus the gat operon is not exclusive to E. hormaechei subspecies hormaechei as stated, but is also present for e.g. in type strain E. bugandensis EB-247T.” We apologize for being unclear. We were summarizing what is already in the literature for distinguishing E. hormaechei subspecies from each other. We have been more precise and clarified this issue in the Results section."
}
]
}
] | 1
|
https://f1000research.com/articles/7-521
|
https://f1000research.com/articles/7-969/v1
|
29 Jun 18
|
{
"type": "Opinion Article",
"title": "Ten considerations for open peer review",
"authors": [
"Birgit Schmidt",
"Tony Ross-Hellauer",
"Xenia van Edig",
"Elizabeth C Moylan",
"Tony Ross-Hellauer",
"Xenia van Edig",
"Elizabeth C Moylan"
],
"abstract": "Open peer review (OPR), as with other elements of open science and open research, is on the rise. It aims to bring greater transparency and participation to formal and informal peer review processes. But what is meant by `open peer review', and what advantages and disadvantages does it have over standard forms of review? How do authors or reviewers approach OPR? And what pitfalls and opportunities should you look out for? Here, we propose ten considerations for OPR, drawing on discussions with authors, reviewers, editors, publishers and librarians, and provide a pragmatic, hands-on introduction to these issues. We cover basic principles and summarise best practices, indicating how to use OPR to achieve best value and mutual benefits for all stakeholders and the wider research community.",
"keywords": [
"open peer review",
"open science",
"good practice",
"research integrity"
],
"content": "Introduction\n\nPeer review is heralded as the bedrock of quality assurance in scholarly communication, used to scrutinise, select, and improve manuscripts for publication (and further applied in many other contexts, including the review of grant proposals, conference papers, etc). However there are differences in the way various models of peer review are implemented. What is often termed ‘traditional’ or ‘conventional’ peer review is generally (1) partially or completely anonymous, with either the reviewer unknown to the author (single-blind review) or both author and reviewer unknown to each other (double-blind review); (2) selective, with reviewers invited by editors; and (3) opaque, with neither the review, editorial process nor the review reports themselves ever made public. Large-scale surveys continuously show that researchers hold peer review to be beneficial, but that processes are potentially often sub-optimal and open to criticism for being, for example, biased or slow (e.g. 1,2,3). In response to these criticisms, and as the wider agenda towards open research is taking hold, variations of open peer review (OPR) are increasingly being offered by publishers and third-party vendors as a regular or additional feature of the publication process.\n\nSo what is OPR? OPR means different things to different people and communities and has been defined as “an umbrella term for a number of overlapping ways that peer review models can be adapted in line with the aims of open science”4. Its two main traits are “open identities”, where both authors and reviewers are aware of each other’s identities (i.e., non-blinded), and “open reports”, where review reports are published alongside the relevant article. These traits can be combined, but need not be, and may be complemented by other innovations, such as “open participation”, where the wider community are able to contribute to the review process, “open interaction”, where direct reciprocal discussion between author(s) and reviewers, and/or between reviewers, is allowed and encouraged, and “open pre-review manuscripts”, where manuscripts are made immediately available in advance of any formal peer review procedures (either internally as part of journal workflows or externally via preprint servers).\n\nAll these features aim towards either increased transparency, rigour, or inclusivity in research processes, as well as recognizing reviewers’ contribution to published research literature, driven by a wide range of considerations. In this sense, this article sets out the following 10 items that outline how to apply OPR in a way such that it becomes a valuable exercise for you as an author, reviewer or editor.\n\nThe results of this papers have been derived from a review of a broad range of research studies (in some the authors have been involved), as well as practices and experiences at publishing houses that have implemented OPR (including those to which the authors are affiliated, for the literature review we build on4).\n\n\nItem 1: Understand what kind of peer review you’re dealing with\n\nThe term “open peer review” is often used to refer to a number of distinct innovations in peer review, which are combined in many different ways4. As an author, reviewer or editor it is essential that you take the time to understand the choices and obligations you have under each system. Must reviewer identities be revealed or is this optional? Will reviewer reports be published upon acceptance, or even if manuscripts are rejected? Will authors and reviewers be brought into discussion with each other? Familiarising yourself with the particular aspects of an OPR process will help avoid any surprises later on. If in doubt, do not hesitate to contact the journal editor to clarify any questions. Editors play an important role in moderating the review process and are glad to provide additional guidance.\n\n\nItem 2: Open peer review relies on, and encourages mutual trust, respect, and openness to criticism\n\nWhatever the degree of openness in a peer review process, as a standard form of academic best practice, it is essential to act with an attitude of charity and in accordance with the highest moral principles5. First of all, as a reviewer you should start with acknowledging the authors’ effort in presenting their results — i.e. review on the assumption that the text makes sense, that it is important and interesting, even if it does not seem so at first glance (cf. Davidson’s “principle of charity”6). In the review process, authors and reviewers typically collaborate on the improvement of a manuscript, which is in principle submitted as ready for publication. A notable exception here is the registered report article format in which the rationale for a study and the proposed methodology — the “study protocol” — are pre-registered with the journal and submitted for peer review before the actual gathering of data and research takes place7. During the review process, authors and reviewers may subjectively agree or disagree on how to present the results and/or what needs improvement, amendment or correction. Now imagine all this happens with readers able to see the process ‘live’ or after the article has been published! It is therefore essential that reviewers ensure that they communicate their points in a clear and civil way, to maximise the chances that it will be received as valuable feedback by the author(s). Authors should also be able to respond in kind (i.e., treating peer review as a dialogue, not a monologue), accepting comments and critique as a process of constructive collaboration in ensuring their work is of the highest quality for publication, and refrain from anything which could be interpreted as a vengeful action.\n\n\nItem 3: Open peer review enables constructive and efficient quality assurance\n\nIn situations where manuscripts are made available as a preprint and review reports are disclosed, all steps of the scientific quality assurance process can be traced and examined. Experiences gained since the early 2000s have shown that submissions that are posted publicly for interactive public peer review start off with a higher quality compared to those submitted in closed peer review processes8. Although some early studies found no overall change in quality between single-blind versus identities revealed to the authors (i.e. “open identities”)9, other studies have indicated that there may be an improvement of the overall quality of review reports under OPR, particularly that comments are more constructive10,11.\n\nSome OPR approaches rely on a consensus-building process between the reviewers (e.g., eLife takes this approach), which is an efficient way of providing feedback to the authors. Consistent feedback saves the authors’ time and is far easier to take into account than contradictory reports. Signed reviews enable direct communication between authors and reviewers and thus may also enhance constructive quality assurance, although there is a risk that non-anonymous reviews will be less critical.\n\nOPR can also address aspects of the replication crisis making it easier for experts and non-experts to evaluate the reliability of findings.\n\n\nItem 4: Open peer review increases transparency and accountability\n\nTransparency in peer review — i.e. with review reports and/or reviewer identities disclosed — can be beneficial for all parties –authors, reviewers, editors, and readers. Transparency of the peer review process enables readers to see how any (dis)agreements (and potential biases) were addressed and how the final version emerged (who argued for what, what arguments were adopted, how controversial points were addressed, etc.). This is especially the case when the authors’ responses to the reviewers’ comments are also published. An OPR process also makes the reviewers accountable for their comments, and indeed the editor accountable for their choice of reviewers and the final acceptance decision. On this basis, the research community and wider public can assess all comments made by authors, reviewers and editors, and may even participate in the discussion. By fostering such transparency, OPR reports can help dispel persistent concerns about the rigour of peer review processes12, or even highlight places where these concerns might be perfectly legitimate.\n\nIn fact, were OPR the standard, it could help distinguish journals, authors, editors, and reviewers who follow good practices from those that do not. Furthermore, the growing body of openly available publications together with review reports also enables the mining of content and perceptions as well as directions of research, including the assessment of “quality and quantity of contributions to the peer-review process alongside publication record as an additional measure of a researcher’s impact in his or her field”13. However, how to best archive and preserve published review reports and related comments has yet to be addressed comprehensively.\n\n\nItem 5: Open peer review facilitates wider, and more inclusive discussion\n\nResearch findings emerge in a complex network of scholarly communication, but only a small part of this process is currently recorded and made publicly available to authors, reviewers, and readers. Publicly available peer review reports, comments and discussions broaden the perspective on the research presented. They document how reviewers and authors, as representatives of the research community, evaluated the work’s achievements, merits and shortcomings in its early stages of being ‘made ready’ for publication. Depending on the topic, papers may receive from just a few to up to a hundred comments8 (for a paper which received many comments see e.g. 14). Again, journal editors can play an important role in encouraging and moderating comments from the research community and the wider public.\n\nIf the process is opened up to the wider community, additional constructive input (in addition to the reviewer reports) can help to further enhance the quality of a manuscript. Examples for journals with such discussions are the Journal Economics, the interactive journals of Copernicus Publications, and SciPost. This type of open commenting facilitates discussion and (in some approaches) consensus building between reviewers and authors. Finally, as an author, you may cite reviews and comments in your revised version, and thus acknowledge these contributions.\n\n\nItem 6: Open peer review gives reviewers recognition and makes reviews citable\n\nPeer review, done well, is hard work — usually taking between 5 hours (median) and 8.5 hours (mean) per person per review15. Yet, traditionally there have been few obvious incentives for reviewers, beyond a quid pro quo status of mutualism. A recent survey of almost 3,000 reviewers found that 4 in 5 agreed peer review is currently insufficiently recognised, and that reviewers would invest extra effort if review activities were formally acknowledged in research assessments, promotion processes and funding applications16. OPR can facilitate this by making review activities visible, open to inspection, and formally citable (e.g. by assigned DOIs to review reports). Traditionally, as a reviewer, it was only possible to indicate the journals that one had reviewed for — now with open reports linked to reviewer identities, reviews become creditable research outputs in their own right. Crossref has recently adopted a metadata schema that allows DOIs to be assigned to reviews, enabling peer review reports to be persistent research outputs, which can be listed on CVs and ORCID profiles17,18.\n\n\nItem 7: Open peer review is gaining popularity\n\nWhile OPR is not without its challenges (e.g. 19,20,21), support is growing across various fields. It has long been recognized to be feasible in practice13 and in recent optional trials, the majority of authors are willing to publish the full peer review history if given the opportunity (e.g. 22,23). However, there are still strong concerns against disclosing reviewer identities, with more than half believing it would make peer review worse24. Anecdotal evidence from editors suggests it can be harder to find reviewers who are willing to agree to OPR, however this is not insurmountable in practice25 and can be outweighed by the advantages discussed above. OPR is certainly happening across many research fields and looks set to rise in the future26.\n\n\nThing 8: Open peer review offers learning opportunities and facilitates training\n\nWith every review process, both early career and established reviewers can learn something, typically about a new finding in their field of research but also through providing feedback and advice to other researchers. Where reports of fellow reviewers are made available, it is worthwhile to read them carefully and learn from them in writing your next review5. OPR opens up this opportunity to a wider audience, and adds further learning points. Consider using open review reports as instructive material in training sessions with early career researchers on how to write a good review. Open review comments offer further opportunities: they can serve as a testing ground for brief feedback or more comprehensive additional reviews.\n\nHowever, as an early career researcher you may not feel comfortable to step into a fully open review process, in particular if you have strong opinions about the paper under review. Perhaps, a potential compromise is to start with a journal that allows you to become gradually open. For example, Royal Society Open Science and PeerJ encourage but do not require reviewers to sign their reports, and authors are allowed to decide whether reviewer reports and author responses are published alongside articles. In addition, turning to a trusted mentor for advice whilst undertaking a first-time OPR may be helpful.\n\n\nItem 9: There is room to practice open peer review even if it has not been formally introduced yet\n\nEven where journals do not formally operate an OPR process, as an author or reviewer you will still have room to practice openness. For example, you can often choose to sign your review, although this can depend on the individual editor or peer review workflow whether this information will be passed to the author. When signing reviews you may consider doing so with reference to the Open Science Peer Review Oath27. A more radical step is for a reviewer to publish the review report, however, this would require express agreement in advance with the authors and journal.\n\nThe most radical approach from a reviewer perspective was recently featured in WIRED magazine: neuroscientist Niko Kriegeskorte advocates only reviewing manuscripts that have already been made available online on a preprint server. When he receives a review request, he first emails the author to identify himself and advise of his review policy, requesting the author upload a preprint online. He then reviews the paper, posts the review on his blog and in conclusion sends the review to the editor28. In a similar spirit, the PRO initiative encourages reviewers to make openness of underlying research data and related materials a pre-condition of comprehensively reviewing a paper.\n\nAs an author there is less room for direct interaction in the traditional journal setting but with the emergence of preprint servers in many fields you have new opportunities to collect feedback and comments from the scientific community and wider public, both in public and private. If you want to crowd-source additional comments and reviews for your manuscript via open participation, this is still possible even if the venue you submit to does not practice such peer review — simply post your pre-review manuscript to a preprint-server simultaneously with submission and invite feedback through commenting features, e.g. services for collaborative review of preprints29, or just ask people to email or contact you via social media (again, do acknowledge these contributions in your revised version). However, as some journals consider posting on a preprint server as pre-publication you need to check beforehand if your preferred journal permits posting on preprint servers (cf. the SHERPA RoMEO database and Wikipedia’s Academic journals by posting policy).\n\n\nItem 10: We need more analysis of and research on open peer review\n\nIt is well recognized that the diverse practice of peer review is not without its flaws (for a summary see e.g. 30). Although OPR may help address some of them, it will not solve them all or suit every community. OPR can provide incentives for robust open research practices but will not be able to completely prevent undesirable behavior or even misconduct. However, it provides a means to make such cases much more transparent, even in cases of retractions (for an example discussed on Retraction Watch see 31).\n\nThere is still need for further research into OPR, especially in terms of both desirable and unintended outcomes as well as efficacy compared to conventional processes. We need additional evidence from authors, reviewers and editors on how effective OPR actually is in various fields. We also need further research into which issues and biases in peer review still need to be addressed4,24,32, and how the publication environment can be further improved in order to better support diversity and inclusivity in peer review. OPR is not a panacea and other models, experiments and initiatives with peer review may help to address some of the concerns with traditional peer review.\n\n\nDiscussion\n\nIn this section we would like to briefly outline where further investigation is needed especially with regard to unintended effects, possible biases, how to mitigate those effects and what role “opt-out” may play in the above good-practice recommendations.\n\nAre there any rules missing? Possible “Opt out if there is a sound reason to do so” could be considered. For example, in pilots with OPR experiments10, about 1 in 10 reviewers declined to review due to a potential conflict of interest and 1 in 4 for personal reasons. However, time was the most important factor (over 2/3). In the context of this paper, we have refrained from introducing an “opt out” related to any of the traits of OPR as the option to decline to review seems sufficient.\n\nOne further open question is who benefits most from OPR, e.g. is it the case that there are more positive reviews for well-established researchers or for papers which tackle more trendy topics? Which biases play out in OPR? Are language skills a factor? Are there country, subject or gender biases? How can such effects be controlled/mitigated? For the latter there certainly is a role for e.g. editors, who can amend policies and provide additional instruction and education. Is “open reports” the best we can have given the strong competition in academia?\n\nSome of these questions are currently being explored, e.g. in the context of the EU-funded OpenUp project. In addition to exploring workflows and the behavior of all actors involved, OpenUp proposes to collect and aggregate data across publishers in order to evaluate the efficacy of OPR processes33.\n\nHowever, the good news is that in OPR settings biases can be monitored and inform interventions. E.g. a recent study revealed that all-female economics papers remain six months longer in peer review than all-male papers, and the author expressed that hope that open peer review may prevent such behavior34. An investigation based on Wellcome Open Research from the first year of operation (based on 142 papers, gender of first author) showed that reviewers took only a few days longer to review papers of female first authors (19.5 vs. 14 days for female vs. male first authors)35. Further investigation would be desirable, in order to monitor undesirable behavior and identify opportunities for editor intervention.\n\n\nConclusion\n\nOPR is an innovation in scholarly communication that deserves further attention. As we have outlined in the 10 items above, it places a research work in the context of a discussion, and gives authors, readers and others a chance to better understand the process from the initial manuscript submission to final published version. As such it provides excellent learning opportunities and the potential to improve scholarly communication and research towards a more transparent, collaborative and participative undertaking.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Competing interests\n\n\n\nBS is head of Open Science projects at Göttingen State and University Library. She contributes to several committees, e.g. the EC's Horizon2020 expert group on Future of Scholarly Publishing and scholarly Communication and Knowledge Exchange's Open Access Expert Group.\r \r BS and TRH are affiliated with OpenUP, an EC funded project which addresses key aspects and challenges of the currently transforming science landscape and aspires to come up with a cohesive framework for the review-disseminate-assess phases of the research life cycle that is fit to support and promote Open Science. \r \r BS is, and TRH was, affiliated with the OpenAIRE2020 project, an EC-funded initiative to implement and monitor Open Access and Open Science policies in Europe and beyond. \r \r TRH is Editor-in-Chief of Publications (ISSN 2304-6775), an open access journal on scholarly publishing published quarterly by MDPI and Senior Researcher at Know-Center GmbH, Graz, Austria, a research centre for data-driven business innovative information and communication technologies.\r \r XvE is business development manager at Copernicus GmbH, and open-access publisher (Copernicus Publications) and professional congress organizer (Copernicus Meetings). Since September 2015 she has been a member of the board of directors of the Open Access Scholarly Publishers Association (OASPA). \r \r ECM supported and initiated the development and implementation of new approaches to peer review and related policy at BMC (part of Springer Nature) until June 2018. She is also an Editorial Board Member for Research Integrity and Peer Review and a member of the Advisory Board for EnTIRE (an EU proposal for Mapping the research ethics and research integrity framework).\n\n\nGrant information\n\nParts of this work were funded by the European Commission H2020 projects OpenUP (Grant agreement 710722, Call: H2020-GARRI-2015-1) and OpenAIRE2020 (Grant agreement: 643410, Call: H2020-EINFRA-2014-1).\n\nTRH is Senior Researcher at Know-Center GmbH, Graz, Austria. The Know-Center is funded within the Austrian COMET program—Competence Centers for Excellent Technologies – under the auspices of the Austrian Federal Ministry of Transport, Innovation and Technology, the Austrian Federal Ministry of Economy, Family and Youth, and the State of Styria. COMET is managed by the Austrian Research Promotion Agency FFG.\n\nThe funders provided support in the form of salaries for authors [BS, TRH], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors have been articulated using CrediT.\n\n\nAcknowledgements\n\nThe authors would like to thank Margo Bargheer, Andrea Bertino, Jean-Sébastien Caux, Sally Chambers, Keti Glonti, Lynn Kamerlin, Louisa Kulke, Bahar Mehmani, Michael Markie, Jon Tennant, and Korinna Werner-Schwarz for providing valuable comments to a draft version of this paper. In addition, we are grateful for the helpful comments and suggestions provided by the reviewers and audience of the PEERE conference, held in May 2018 in Rome.\n\n\nReferences\n\nSense about Science: Peer review survey 2009: Full report. Technical report, 2009. Reference Source\n\nWare M: Publishing research consortium peer review survey 2015. Technical report, 2016. Reference Source\n\nPeer review – a global view. Technical report, Taylor & Francis, 2016. Reference Source\n\nRoss-Hellauer T: What is open peer review? A systematic review [version 2; referees: 4 approved]. F1000Res. 2017; 6: 588. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBourne PE, Korngreen A: Ten simple rules for reviewers. PLoS Comput Biol. 2006; 2(9): e110. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavidson D: Inquiries into Truth and Interpretation. Clarendon Press, Oxford, 1984. Reference Source\n\nChambers C, Munafo M: Trust in science would be improved by study pre-registration. The Guardian. 2013. Reference Source\n\nPöschl U: 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. 2012; 6: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Rooyen S, Godlee F, Evans S, et al.: Effect of open peer review on quality of reviews and on reviewers’ recommendations: a randomised trial. BMJ. 1999; 318(7175): 23–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMehmani B: Is open peer review the way forward? 2016. Reference Source\n\nKowalczuk MK, Dudbridge F, Nanda S, et al.: Retrospective analysis of the quality of reports by author-suggested and non-author-suggested reviewers in journals operating on open or single-blind peer review models. BMJ Open. 2015; 5(9): e008707. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWicherts JM: Peer Review Quality and Transparency of the Peer-Review Process in Open Access and Subscription Journals. PLoS One. 2016; 11(1): e0147913. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGodlee F: Making reviewers visible: openness, accountability, and credit. JAMA. 2002; 287(21): 2762–2765. PubMed Abstract | Publisher Full Text\n\nHansen J, Sato M, Hearty P, et al.: Ice melt, sea level rise and superstorms: evidence from paleoclimate data, climate modeling, and modern observations that 2 °C global warming could be dangerous. Atmos Chem Phys. 2016; 16(6): 3761–3812. Publisher Full Text\n\nWare M, Monkman M: Peer Review in Scholarly Journals: An international study of the perspective of the scholarly community. Technical report, Marc Ware Consulting, 2008. Reference Source\n\nWarne V: Rewarding reviewers - sense or sensibility? A Wiley study explained. Learned Publishing. 2016; 29(1): 41–50. Publisher Full Text\n\nHendricks G, Lin J: Making peer reviews citable, discoverable, and creditable. 2017. Reference Source\n\nWrigley A: #RecognizeReview with ORCID. 2016. Reference Source\n\nPros and cons of open peer review. Nat Neurosci. 1999; 2(3): 197–198. PubMed Abstract | Publisher Full Text\n\nBoughton S: What are the challenges of open peer review? 2016. Reference Source\n\nYoder J: The Fourth Reviewer: What problem is open peer review trying to solve? 2016. Reference Source\n\nWho’s Afraid of Open Peer Review? PeerJ Blog. 2015. Reference Source\n\nTransparent peer review one year on. Nat Commun. 2016; 7: 13626. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoss-Hellauer T, Deppe A, Schmidt B: Survey on open peer review: Attitudes and experience amongst editors, authors and reviewers. PLoS One. 2017; 12(12): e0189311. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKowalczuk MK, Samarasinghe M: Comparison of Acceptance of Peer Reviewer Invitations by Peer Review Model: Open, Single-blind, and Double-blind Peer Review. Peer Review Congress, 2017. Reference Source\n\nAmsen E: What is open peer review? 2014. Reference Source\n\nAleksic J, Alexa A, Attwood TK, et al.: An Open Science Peer Review Oath [version 2; referees: 4 approved, 1 approved with reservations]. F1000Res. 2015; 3: 271. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThe Rogue Neuroscientist on a Mission to Hack Peer Review. 2017. Reference Source\n\nHindle S, Saderi D: PREreview – a new resource for the collaborative review of preprints. 2017. Reference Source\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 | Free Full Text\n\nAuthors retract much-debated blockchain paper from F1000. 2017. Reference Source\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\nRoss-Hellauer T, Görögh E: Application framework and transformation scenarios for open peer review. OpenUP Deliverable 3.3. Technical report, 2018. Reference Source\n\nHengel E: Publishing while female: Are women held to higher standards? Evidence from peer review. 2017. Publisher Full Text\n\nSchmidt B: WOR: Wellcome Open Research - Exploration of year one data. 2018; original-date: 2018-01-30T14:33:31Z. Reference Source"
}
|
[
{
"id": "35996",
"date": "23 Jul 2018",
"name": "Richard Smith",
"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'm glad that this article is already available to be read by those who want to know more about open peer review, particularly young researchers. The article is a useful guide.\nIronically, I think that little value will be added by me peer reviewing it openly ot otherwise. If the article had been submitted to a more traditional journal, including many that review openly, then it would not be available for people to read. It might have languished for months while it progressed slowly through the bureaucracy of science journals. Indeed, it would probably be rejected because there is little or nothing that is new in the article. The article would then work its way down the food chain of journals before appearing in an obscure journal where few people would read it, especially if it appeared in a journal behind a paywall.\nWhy have the authors submitted the paper to a journal? Why haven't they simply posted it as a blog and used social media to promote the article? They might easily achieve a higher readership, and presumably it is readers that they want. Or is it? I think that these authors are primarily interested in readers, but sadly the main motivation for many authors is simply to add to their CVs and increase their chances of raising more grants and getting promoted.\nI am letting my cynicism about the whole process of science publishing shine through, and I would have liked this article more if it had been less accepting of the status quo and more questioning. But that is probably not the role of a simple guide.\nI don't have any major criticisms of the article, and it will be fine if it published and indexed as it is. But here are some comments in descending order of importance.\n1. I think that the authors could have done a better job of discerning the quality of the evidence they quote. As I've argued elsewhere peer review is more faith than evidence based, https://breast-cancer-research.biomedcentral.com/articles/10.1186/bcr2742 and scientists who would not advance string views on their own subject without evidence are quite happy to make statements on peer review without evidence. For example, in Item 3 the authors write: “Although some early studies found no overall change in quality between single-blind versus identities revealed to the authors (i.e. “open identities”), other studies have indicated that there may be an improvement of the overall quality of review reports under OPR, particularly that comments are more constructive.” They do not make clear that the first study was a randomised trial, a much more reliable method than the later studies, one of which was simply a survey and the other a retrospective study.\n2. This is not a scientific study, but the authors do outline what might be described as their methods in the fourth paragraph. They might usefully expand this, providing more information and explaining how this paper builds on their systematic review.\n3. This is my bias, but I'd have liked to see a short section pointing out that there is little empirical as opposed to anecdotal evidence to support peer review but lots demonstrating its problems. At the end the cite my article in the JRSM, but they would do better to cite my article in Breast Cancer Research1, which is more up to date and contains more data. They might also have discussed more the objectives of peer review.\n4. The authors suggest that tyro reviewers might start with journals, like Royal Society Open Science that allows reviewers to choose whether or not to be named. This is a tangential point, but I think that offering this option is probably a bad idea. I've experienced a case where a journal rejected a paper on the basis of an unsigned review while also sending the authors a glowing signed review. You can imagine that this seemed unfair to authors. The other strategy the authors suggest--of consulting a mentor--is a much better idea.\n5. Perhaps the authors might say more about the ethics of open peer review. We introduced open peer review at the BMJ when I was the editor largely on ethical grounds as our studies did not show that open peer review was superior to closed peer review2. All publishing should be about credit and accountability, and closed review is weak on both.\n6. I, like most readers, find it annoying when authors introduce things I've never heard of without saying more. I'd like them not just to give a reference to Open Science Peer Review Oath but tell me a bit more in this paper. Similarly they refer to \"the PRO initiative.\" Is PRO a reference to the Peer Review Oath? I guess that it probably is, but I don't want to have to guess.\n7. I wish the authors had not succumbed to the academy disease of acronyms. I'd lime them to spell out OPR, and there is something particularly silly about referring, as they do, to \"the process of OPR.\" 8. I hadn't heard of the process the authors mention adopted by Niko Kreigeskorte, but I think it a great idea. Like him I will post these comments on my blog as well as on F1000Research.\nConclusion As I said at the beginning, I’m sure that some researchers and others will find this guide useful, and I’m glad that it’s already published. The authors might want to revise the paper in the light of my comments, but I don’t think that it’s essential. I support the paper being indexed whether or not they revise the paper.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "35998",
"date": "01 Aug 2018",
"name": "Darko Hren",
"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 nice introductory article to a reader who wishes to learn about open peer review. Authors introduce the main concepts and lay down ten items to be considered when dealing with open peer review. It will be useful to early stage researchers and general readership new to the issue. This is not a research article, but a non-systematic review of existing literature with recommendations from experienced researchers and professionals in the field. In my opinion, the article can be indexed without revisions.\nMinor point:\n\npp 4 bottom right says \"Thing 8\" - should be \"Item 8\" (?)\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-969
|
https://f1000research.com/articles/7-968/v1
|
29 Jun 18
|
{
"type": "Software Tool Article",
"title": "Norwegian e-Infrastructure for Life Sciences (NeLS)",
"authors": [
"Kidane M. Tekle",
"Sveinung Gundersen",
"Kjetil Klepper",
"Lars Ailo Bongo",
"Inge Alexander Raknes",
"Xiaxi Li",
"Wei Zhang",
"Christian Andreetta",
"Teshome Dagne Mulugeta",
"Matúš Kalaš",
"Morten B. Rye",
"Erik Hjerde",
"Jeevan Karloss Antony Samy",
"Ghislain Fornous",
"Abdulrahman Azab",
"Dag Inge Våge",
"Eivind Hovig",
"Nils Peder Willassen",
"Finn Drabløs",
"Ståle Nygård",
"Kjell Petersen",
"Inge Jonassen",
"Kidane M. Tekle",
"Sveinung Gundersen",
"Kjetil Klepper",
"Lars Ailo Bongo",
"Inge Alexander Raknes",
"Xiaxi Li",
"Wei Zhang",
"Christian Andreetta",
"Teshome Dagne Mulugeta",
"Matúš Kalaš",
"Morten B. Rye",
"Erik Hjerde",
"Jeevan Karloss Antony Samy",
"Ghislain Fornous",
"Abdulrahman Azab",
"Dag Inge Våge",
"Eivind Hovig",
"Nils Peder Willassen",
"Finn Drabløs",
"Ståle Nygård"
],
"abstract": "The Norwegian e-Infrastructure for Life Sciences (NeLS) has been developed by ELIXIR Norway to provide its users with a system enabling data storage, sharing, and analysis in a project-oriented fashion. The system is available through easy-to-use web interfaces, including the Galaxy workbench for data analysis and workflow execution. Users confident with a command-line interface and programming may also access it through Secure Shell (SSH) and application programming interfaces (APIs). NeLS has been in production since 2015, with training and support provided by the help desk of ELIXIR Norway. Through collaboration with NorSeq, the national consortium for high-throughput sequencing, an integrated service is offered so that sequencing data generated in a research project is provided to the involved researchers through NeLS. Sensitive data, such as individual genomic sequencing data, are handled using the TSD (Services for Sensitive Data) platform provided by Sigma2 and the University of Oslo. NeLS integrates national e-infrastructure storage and computing resources, and is also integrated with the SEEK platform in order to store large data files produced by experiments described in SEEK.\n\nIn this article, we outline the architecture of NeLS and discuss possible directions for further development.",
"keywords": [
"Data management and sharing",
"compute and storage infrastructure",
"microservices",
"federated authentication",
"integration API",
"Galaxy",
"ELIXIR Norway"
],
"content": "1. Introduction\n\nThe Norwegian ELIXIR node is coordinated by the University of Bergen (UiB) and comprises the University of Oslo (UiO), The Arctic University of Norway (UiT), the Norwegian University of Science and Technology (NTNU), and the Norwegian University of Life Sciences (NMBU). The node provides services, training, and support to a broad range of national users, largely life-science researchers and students1. These scientists usually work in collaborative projects and need to store, analyze, and share data sets, often large in size, throughout all stages of the project, and between various platforms and computational resources. However, many of these users do not feel comfortable using a command-line interface, and have limited programming, system administration, or data management skills.\n\nCommercial workbenches such as the BaseSpace Sequence Huba and Geneiousb aim at user accessibility, but offer computation and data sharing only within their closed and expensive platform setups. On the other hand, the open-source SEEK serves as a platform for sharing systems biology project data, transparently and for free2,3. Notably, the integrative open-source GenomeSpace enables organizing and sharing data not only between users, but also between various workbenches and computational resources4–6. Although powerful, its setup could not be adapted to integrate with the available and required e-infrastructure resources in Norway.\n\nThe Norwegian e-Infrastructure for Life Sciences (NeLS) was built upon the previous experiences with developing and using bioinformatics workbenches in Norway, for example: the Genomic HyperBrowser7–9, an extension of Galaxy10,11; the easy-to-use UiO Bioportal12, later replaced by a Galaxy-based Lifeportal13; or eSysbio, a workbench prototype for data sharing and systems biology workflowsc. NeLS provides users in Norway and their collaborators abroad an integrated system for data storage and sharing, as well as data processing and analysis. NeLS allows users to efficiently and safely store, analyze, and share their genomics-scale data and analyses, all through the use of web interfaces. Most Norwegian users can log in using the credentials – user name and password – they use at their home institution, other users need to register.\n\nNeLS has a three-layered architecture (Figure 1). The intermediate layer (Storage Level I) provides data storage intended to be used by projects in an active analysis phase (with data being kept in this storage layer for months). Data can be accessed (and up- and downloaded) through a web portal, as well as through Secure Shell (SSH) and application programming interfaces (APIs). The latter two provide command-line-confident users with a more efficient way to work with data. The top level constitutes the data analysis workbench of NeLS. For this, we have chosen the popular Galaxy, an open-source, web-based workbench for accessible, reproducible, and transparent computational omics. Galaxy allows computational workflows to be set up and used without the need of programming skills. Our Galaxy instances have limited storage capacity and it is therefore intended that data resides on this level only for short periods of time, in the range of weeks. The bottom layer (Storage Level II) offers long-term storage, provided by the National Infrastructure for Research Data (NIRD), a generic e-infrastructure operated by Sigma2d. Here, data can be stored for years, requiring a more strictly organized structure with defined types and metadata.\n\nNeLS includes Galaxy instances hosted at the five universities constituting the Norwegian ELIXIR node. These instances have a basic catalogue of tools and workflows that are relevant for researchers in life sciences, as well as more specialized ones depending on the focus of the hosting institution. NeLS provides tools integrated into Galaxy to easily push and pull data from the persistent Storage Level I. Some of the Galaxy servers are integrated with high-performance computing (HPC) resources – provided by Sigma2 – for transparent execution of computationally intensive tools.\n\nIn this article, we describe the architecture of this integrated e-infrastructure and examples of its usage, and outline the possible directions of future developments.\n\n\n2. The architecture of NeLS\n\nThe Norwegian e-Infrastructure for Life Sciences was not built as a top-down, grand design and implementation exercise. Rather, it was implemented through time by focusing on different parts of the problem at a time and always striving to make a functional whole. It was decided early on to avoid re-inventing the wheel and rather base the system on proven solutions and practices whenever possible. In addition, addressing different concerns in isolation while keeping the big picture in mind has proven to be an effective way for constructing the NeLS system. In the end, there were many components of different flavors: off-the-shelf systems, adaptations of available open-source packages, and also custom in-house developed systems. Figure 2 shows a componentized architecture of the NeLS system. In the following subsections, we describe the components of the NeLS system.\n\nAuthentication is a process by which the user’s credentials are verified against a user-information catalogue in order to determine whether the user is who they claim to be, before granting access to resources. NeLS supports multiple identity sources based on the Security Assertion Markup Language 2.0 (SAML 2.0) standarde. Currently, it supports the Norwegian Federated Electronic Identity service (FEiDEf) and NeLS’s own identity provider. NeLS’s identity provider was constructed by configuring Simple-SAMLphpg, an open-source security software system. Integration with the ELIXIR Authentication and Authorisation Infrastructure (AAI)15 as an identity provider has been tested technically, and can in the future be used to differentiate resource allocations further in the NeLS network of services.\n\nThe first time a user logs in using any of the supported identity providers, NeLS creates a user profile and subsequent derived identities. Secure Shell (SSH) access credentials are generated for the user and can be fetched from the NeLS portal (central hub). This coordination layer holds metadata about projects and associated membership of users. The user-profile management and coordination block provides a Representational State Transfer (REST, \"RESTful\") web application programming interface (API)16–18 for other units, and enables asynchronous job management by leveraging an off-the-shelf message-queuing system, RabbitMQh. Structured logging, e-mail communication, and related management tasks are supported in this block.\n\nThe NeLS portal is the central hub of the whole system. It is a Java-based web application with multiple responsibilities and uses the Spring Security packagei to interface with the different SAML 2.0 identity providers. Upon first login, NeLS creates a profile for the user and initializes all necessary components. Following are the four distinct responsibilities of the NeLS portal:\n\n(a) Web-based file-system browser to the NeLS Storage Level I (see subsection 2.6).\n\n(b) Initiate and monitor asynchronous jobs for copying, moving, and sharing files within the Level I storage layer, as well as transfer across storage Levels I and II.\n\n(c) Facilitate OAuthj token provision. The NeLS portal acts as a bridge towards the identity providers and avails NeLS metadata for the OAuth service (see subsection 2.4).\n\n(d) Interface with external systems. The NeLS portal has been successfully integrated with a national solution for sensitive data, the TSDk 19. NeLS makes two-factor authentication of the TSD easier for users, and provides an easy web-based way for initiating data-transfer jobs between the NeLS and TSD infrastructures.\n\nThe Galaxy block is the workhorse of the NeLS system. It gives the user a curated set of tools and workflows supported by the ELIXIR Norway help desk1. Technically, the Galaxy block comprises Galaxy in a remote-user configuration, and an authentication layer in front to interface with the same set of identity providers as the NeLS portal. SimpleSAMLphp in service provider (SP) configuration with its AuthMemCookiel solution on top of Memcachedm is used to interface with the Apache web server, transforming the SAML 2.0 authorization information into a Galaxy-compatible format. NeLS also provides Galaxy tools for data import and export, which work in tandem with the NeLS portal to give the user the possibility of pulling data from the Storage Level I in NeLS into a Galaxy history, and also be able to push results of Galaxy jobs into the NeLS Storage Level I.\n\nNeLS provides different Galaxy instances hosted by different institutions.\n\nThe NeLS compute block executes NeLS Galaxy jobs. The jobs are either executed on the same high-spec (fat) servers as the Galaxy server, or they can be submitted to a high-performance computing (HPC) cluster for parallel execution. The job execution details, including HPC job management, are hidden for the user. The HPC jobs are run using a pre-allocated compute quota.\n\nWe use the Light-weight Runner (LWR, now renamed to Pulsarn) Galaxy services to submit computationally intensive Galaxy jobs to an HPC system (such as the Stallo supercomputer in the UiT Galaxyo). LWR communicates with Galaxy via the Galaxy API and a RabbitMQ Advanced Message Queuing Protocol (AMQP) message queue. It specifies the required parameters for the tool and executes a wrapper script for the tool. The wrapper creates temporary directories, submits tool jobs to the HPC scheduler (PBS/Torquep) with selected parameters, saves results, and deletes temporary files. Once the jobs are completed, LWR transfers the data back to Galaxy for the user to inspect.\n\nThis layer of NeLS provides flexible storage with advanced access control list to allow appropriate sharing and data protection in scientific projects.\n\nThe NeLS Storage Level I layer features a dedicated private directory for each user’s personal data as well as a project-based shared storage area for collaboration and sharing. A user can be added to a NeLS project with three possible roles: member, power-user, or principal investigator (PI). Each role has a predefined set of permissions allowed in the project area. Technically, the NeLS Storage Level I is built using FreeBSDq with its support for the ZFSr file system. It employs advanced access control lists and also provides SSH access to more tech-savvy users. It provides a RESTful web API (Java) and command-line management tools (Python).\n\nLevel II of the NeLS storage, also known as the StoreBioinfo layer, enforces more strict organization of datasets, and is facilitated through integration with national storage resources provided by the Norwegian Infrastructure for Research Data (NIRD). Its purpose is to act as a long-term storage and data warehouse, with capacity to hold all of a project’s generated data, from raw to final results, including any data replicated to the Storage Level I. It has a metadata database and interfaces with the data-warehousing system, iRodss, via specialized server-side scripts. The NeLS Storage Level II provides a RESTful web API and is orchestrated via the NeLS central coordination block (3. Central Hub in Figure 2).\n\nThe NeLS public API is a RESTful web service targeted towards external systems. It supports implicit and authorization-code OAutht grant profiles. It exposes a well-defined navigation and linking mechanism into the structured data of NeLS Storage Level II. The OAuth service is an in-house built Python-based system that uses Tornadou and python-oauth2v libraries by interfacing with the NeLS portal. In collaboration with Digital Life Norwayw, the public API is developed to support integration with the SEEK data management system2,3, to allow a resolvable URL to a dataset in NeLS be referenced in SEEK.\n\n\n3. Operation\n\nNeLS is inherently a distributed infrastructure of multiple microservices which naturally would be deployed on different servers. The scale and availability of the different resources to be integrated – such as compute, storage, identity providers, databases, etc. – heavily influences its deployment. In Norway, the NeLS production instance is deployed on 7 different servers, including 2 storage master systems (additional slave storage nodes – sub-systems – are not counted).\n\nFor testing or a proof-of-concept setup, all microservices and web components are possible to run on a single host with 2–4 cores and 64GB of memory, while the storage levels would naturally require their own setups.\n\n\n4 Workflows\n\nTo cover the most prominent NeLS user needs, Galaxy workflows for analyzing RNA sequence data (prokaryotic and eukaryotic), and workflows for both the taxonomic and functional profiling of metagenomic data have been developed. In addition, workflows for the analysis of miRNAs and ChiP-seq analysis are available, see Table 1. All workflows are maintained in order to provide the state-of-the-art tools for the analysis to the users. Upon demand, each work-flow can be modified to accommodate specific user needs, e.g. that a tool is replaced by another tool, or version. A complete overview of the NeLS workflows and links to the Galaxy instance in which a workflow is available can be found in the NeLS portal.\n\nTo ensure data reproducibility and to reduce the compute time for the user, a common data repository with pre-indexed reference genomes has been built. The repository is available across all five Galaxy instances.\n\nFor first-time users of the NeLS Galaxy, a quick start guide that contains information on the Galaxy basics is available on each NeLS Galaxy start page, and more detailed documentation and tutorials on the NeLS workflows are also available there. Finally, the user can contact the national help desk or access a Q&A forum directly from the NeLS Galaxy.\n\n\n5 Use cases\n\nIn a project with non-sensitive data, a user will perform the following steps (See Figure 3e (a)-(e))\n\n(a) Log in to the NeLS portal. If the user does not have an account, and neither has a FEiDE account, they can easily apply for a NeLS account.\n\n(b) Upload data to NeLS repository using SSH or Filezillax. If the data are generated by a Norwegian high-throughput sequencing core facility, e.g. one linked with the Norwegian Consortium for Sequencing and Personalized Medicine (NorSeqy), the user is offered to have the data uploaded directly from the core facility.\n\n(c) (Recommended) Synchronize data to Storage Level II (StoreBioinfo) for annotation and long-term storage.\n\n(d) Log in to one of the Galaxy instances, e.g. ?https://galaxy-ntnu.bioinfo.no.\n\n(e) Get data from NeLS Storage Level I to Galaxy, and run Galaxy tools and/or workflows. Share or publish Galaxy history.\n\n(f) Copy new results to Storage Level I and synchronize to Storage Level II (recommended).\n\nMETA-pipe23–25, developed within the ELIXIR Marine metagenomics project26,27, efficiently produces full-length annotated genes from metagenomic assemblies, and offers the extensive annotation options, flexibility, and visualization needed to pick interesting targets for further investigation. The NeLS version of META-pipe provides taxonomic and functional analysis from whole-genome shotgun sequence data. It supports high-throughput sequencing data and provides assembly, focusing the analysis on full-length genes. The pipeline consists of three major modules: preprocessing, taxonomic classification, and functional analysis. All modules are available as individual workflows, except for assembly in pre-processing, which is run manually on either a high-memory computer or our supercomputer. Workflows can be tailored to the specific needs for the analysis of a sample and it is also possible to add additional steps or to omit some of the steps.\n\nTo use META-pipe, the user follows the generic steps for login and data upload above, using the NeLS Galaxy instance of UiTaa and the NeLS portal to administrate the data of the metagenomics project. Input files are transferred from the NeLS project in Storage Level I, to the Galaxy history of the user using the provided NeLS data transfer tools in Galaxy tool menu.\n\nTo analyze the data, the user selects the META-pipe tool in Galaxy and then configures the pipeline parameters such as which tools to run, input files, reference database versions, and output formats. Once the workflow is configured, the user presses the execute button in Galaxy to execute the pipeline in the background. This will create a history element in Galaxy where the user can view the current status of the job. Currently queued or running jobs are colored yellow, and completed jobs are colored green. When the job is done, the user can examine the output data in the Galaxy view panel, transfer results to the NeLS project in Storage Level II, or download the files to their own computer.\n\nNeLS was not designed for hosting sensitive data such as human genome data from Norwegian patients. ELIXIR Norway is collaborating with the TSD project and infrastructureab 19, created by USIT (The University Centre of Information Technology) at the University of Oslo, to offer a service to researchers in Norway for storing and processing sensitive data, including health data.\n\nNeLS allows for seamless data transfer from the NeLS storage services (Layer I) to the TSD File Lock servers for import of supplementary non-sensitive data that user projects would need available inside TSD to interpret their sensitive data.\n\nWorkflow development for sequencing data analysis etc. performed in ELIXIR Norway is implemented either as tools and workflow definitions for Galaxy or as software containers, such that workflows can be deployed in the appropriate compute environment to facilitate analysis of both sensitive and non-sensitive data. ELIXIR Norway and TSD are working towards a Galaxy service in TSD, and the aspect of workflow mobility is also a key aspect in the Nordic project Tryggve2ac, with ELIXIR partners from Norway, Finland, Sweden, and Denmark, and respective national infrastructures for sensitive data.\n\nTo illustrate how NeLS is used in daily operation of the help desk of ELIXIR Norway, we include an example. The help desk was contacted by a researcher wanting to analyze 15 RNA-Seq samples from the Atlantic salmon. We decided that the already prepared RNA-Seq workflow in our Galaxy instance at NTNU Trondheim consisting of HISAT2 alignment22, followed by read assignment by featureCounts (subRead package)28 and differential expression analysis in voom29, would be suitable for the initial analysis of the data. Atlantic salmon is not an organism with pre-processed genome and transcriptome readily available, so our help desk first had to create a HISAT2-indexed reference genome from the original Atlantic salmon genome FASTA file and transcriptome GFF file downloaded from SalmoBase30. The initial indexing (using HISAT2 indexing with 1.5TB of memory) was done by the ELIXIR Norway staff at NMBU Ås responsible for SalmoBase. This step only needs to be performed once for any reference genome, and can be reused for other users targeting the same organism. The indexed reference was made available for selection in the workflow by ELIXIR Norway staff at NTNU. To run the workflow, the NTNU help desk created a shared project in NeLS Storage Level I layer, and shared the project with the researcher. They in turn uploaded the raw sequencing data to the shared project in NeLS (then becoming available to the responsible person in the help desk), who ran the RNA-Seq workflow in the Galaxy environment. The workflow made use of the dataset collection feature in Galaxy to run alignment and read-assignment on all 15 samples in a single step. Sample group assignments for comparison in differential expression analysis can be defined at the beginning of the workflow, or by adding assignment and comparisons during a re-run of the last step in the workflow (voom analysis). In this way, the user only needs to run the computationally demanding alignment and assignment steps once, but still have the flexibility to change samples assignments and group comparisons in subsequent analysis. In total, four group comparisons were made, reporting differentially expressed genes in each comparison. The total processing and analysis were done with a minimal effort for the user who basically only had to 1) upload the data, 2) define a dataset collection, 3) select the correct organism reference genome (for alignment) and transcriptome (for counting), and 4) define the sample groups assignments and the groups to be compared (differential analysis).\n\n\n6 Unified service toward data-generating platforms\n\nNational or other large data-generating platforms, such as the Norwegian Consortium for Sequencing and Personalized Medicine (NorSeqad) produce user-requested sequencing for multiple purposes. The data are produced on receipt of DNA samples, and these may be of both human and non-human nature, requiring different data handling procedures. The goal is to provide a unified and seamless user experience, in which the user is provided with a resulting dataset in an environment that is suitably equipped with compute resources, relevant analytical tools, reference data, and initial analysis results. All of this should be provided and documented with no action required from the user after the initial agreement. This requires a tight collaboration between the data-generating platform, the national hardware (storage, compute) resources, as well as the ELIXIR help desk facilitating for the user experience in providing relevant tools, workflows, support, and documentation.\n\nNon-sensitive data handling is coordinated through the use of NeLS and its layered architecture. NorSeq staff uploads the generated data to a NeLS project area created by ELIXIR Norway help desk on behalf of the research group ordering the sequencing. After verification of the uploaded data, ELIXIR Norway help desk assists the research group in synchronizing the data also to the Store-Bioinfo services at the Norwegian Infrastructure for Research Data (NIRD), and provides access in different roles to the different members of the research project. Users may then analyze the data utilizing the Galaxy front ends in the NeLS ecosystem, and receive support and training from the ELIXIR Norway help desk.\n\nSensitive data handling is achieved by utilization of Services for Sensitive Data (TSD)19. NorSeq staff uploads the generated data to a special TSD project that allows for initial data analysis using ELIXIR Norway provided workflows jointly by NorSeq and ELIXIR Norway help desk staff, before the raw and processed data are made fully available to the user’s TSD project.\n\n\n7 Discussion\n\nWe have described the NeLS system developed to serve a broad spectrum of bioinformatics users, with focus on Norwegian users and on genomics data. The system supports data storage, sharing, and data analysis in a project-oriented fashion. A strength of the system is that it utilizes a federated identity provider allowing most users to use their institutional login. Furthermore, it integrates storage and compute resources offered by the generic einfrastructure Sigma2, set-up and funded to support users across all research fields in Norway. This avoids duplication of effort and caters for a more harmonized policy with respect to allocations of compute and storage between life science and other fields. An additional strength of the system is that it has interfaces adapted to both advanced users through a programmatic interface (API) and SSH, and to less computer-savvy users through a web portal. This allows different categories of users to work efficiently with the system, and to collaborate through joint projects. The system has been in production since 2015, and has been adapted according to user feedback accumulated over a series of workshops.\n\nThe system has been designed to use existing open-source solutions whenever possible. We believe this strategy produces a system that is easier to maintain and therefore more sustainable. NeLS has been developed in an iterative fashion with short agile development cycles facilitating adaptation to changing needs.\n\nUntil now, we have had one instance of NeLS running at the University of Bergen, linked with five instances of Galaxy, one at each of the partner institutions in ELIXIR Norway. For the future, we are investigating a more dynamic approach launching Galaxy instances on demand.\n\nNeLS itself does not provide the level of security required for handling sensitive data. To support such projects, NeLS is linked with the TSD (Services for Sensitive Data) platform in Oslo19. NeLS and TSD are integrated, allowing transfer of data and workflows between the systems, making for more resource-efficient support of both types of projects benefiting both their operation and their users.\n\nNeLS uses the national Federated Electronic Identity provider (FEiDE) linking all Norwegian universities. The technology used is the same as that used for the ELIXIR Authentication and Authorisation Infrastructure (AAI)15. It is therefore possible to extend NeLS to also support ELIXIR AAI identity provision. NeLS has so far been designed and resourced for supporting Norwegian projects, and new policies – and ideally also new funding mechanisms – would be needed to extend the scope beyond Norwegian projects.\n\nThe NeLS system can be used as an example of how to set up a flexible and relatively light-weight system providing bioinformatics projects with data storage, sharing, and analysis. The NeLS source code is available on GitHub and can helpin the building of similar projects elsewhere, although adaptations must be expected, for example to integrate with storage and compute resources.\n\nThe modularity of NeLS allows its parts to be reused in other contexts. An example is the integration of NeLS with the SEEK platform2,3, where users can link data sets in NeLS with their metadata in SEEK. Future work may include functionality for allowing users to export annotated data from NeLS (optionally integrating linked metadata from SEEK) into public data repositories such as ArrayExpress31 and PRIDE32.\n\n\n8 Conclusions\n\nThe NeLS system is in production and serves as an important platform for the operation of ELIXIR Norway and its help desk for users in molecular life sciences. The system will therefore be maintained and supported in the foreseeable future. We benefit from sharing experiences with other similar projects within and beyond ELIXIR, through the wide adoption of Galaxy across many ELIXIR nodes.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nBio.Tools33 ID: NeLS (https://bio.tools/nels)\n\nRRID: SCR_016301\n\nNeLS is available at https://nels.bioinfo.no, without extra registration for all Norwegian academic users (via FEiDEae), and with registration upon request for all other users.\n\nThe source code of the core NeLS modules is available at https://github.com/elixir-no-nels/nels-core, under the Apache License 2.0af.\n\nArchived source code at the time of publication is available here: http://doi.org/10.5281/zenodo.1251639 under an Apache License 2.034.\n\n\nUse cases\n\nThe source code of the module integrating NeLS with the national Service for Sensitive Data (TSD), can be found within the above repository and archive in elixir-no-nels/nels-core/tsd-proxy.\n\nThe source code of the integration module of META-pipe with NeLS, including Galaxy front end and HPC back end (Stallo supercomputer at UiT), is available under the MIT licenseag at https://gitlab.com/uit-sfb/meta-pipe-galaxy-wrapper, archived in 35.\n\n\nNotes\n\nahttps://basespace.illumina.com\n\nbhttps://www.geneious.com\n\nc14, pp. 53–56, 61–64, https://bora.uib.no/bitstream/ handle/1956/10658/thesis.pdf#page=61\n\ndhttps://www.sigma2.no\n\nehttps://wiki.oasis-open.org/security/FrontPage\n\nfhttps://www.feide.no/introducing-feide\n\nghttps://simplesamlphp.org\n\nhhttps://www.rabbitmq.com\n\nihttps://projects.spring.io/spring-security/\n\njhttps://tools.ietf.org/html/rfc5849\n\nkServices for Sensitive Data, in Norwegian Tjenester for Sensitive Data\n\nlhttps://zenprojects.github.io/Apache-Authmemcookie-Module/\n\nmhttps://memcached.org\n\nnhttps://galaxyproject.org/admin/config/pulsar/\n\nohttps://galaxy-uit.bioinfo.no\n\nphttps://www.adaptivecomputing.com/products/open-source/torque\n\nqhttps://www.freebsd.org/\n\nrhttps://en.wikipedia.org/wiki/ZFS\n\nshttps://irods.org/\n\nthttps://tools.ietf.org/html/rfc5849\n\nuhttps://pypi.org/project/tornado/\n\nvhttps://github.com/joestump/python-oauth2\n\nwhttps://digitallifenorway.org/\n\nxhttps://filezilla-project.org/\n\nyhttps://www.norseq.org\n\nzlicebase.org\n\naahttps://galaxy-uit.bioinfo.no\n\nabServices for Sensitive Data, in Norwegian Tjenester for Sensitive Data\n\nachttps://neic.no/tryggve2/\n\nadhttps://www.norseq.org\n\naehttps://www.feide.no/introducing-feide\n\nafhttps://www.apache.org/licenses/LICENSE-2.0.html\n\naghttps://opensource.org/licenses/MIT",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe work was funded by the ELIXIR.NO (208481/F50) and ELIXIR2 (270068) infrastructure grants from the Research Council of Norway, as well as the Tryggve and Tryggve2 projects from the Nordic e-Infrastructure Collaboration (NeIC).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe acknowledge valuable user feedback from NeLS users and our collaborators in the national sequencing platform NorSeq. We acknowledge Prof. Atle M. Bones and Mahsa Jalili for allowing us to include their project as an example of a bioinformatics analysis in NeLS.\n\n\nReferences\n\nNygård S, Jonassen I: Norwegian Bioinformatics Platform. NBS Nytt. 2014; 38(2): 32–35. Reference Source\n\nWolstencroft K, Owen S, Krebs O, et al.: Semantic Data and Models Sharing in Systems Biology: The Just Enough Results Model and the SEEK Platform. In Proceedings of the 12th International Semantic Web Conference - Part II. ISWC ’13, New York, NY, USA, Springer-Verlag New York, Inc. 2013; 212–227. Publisher Full Text\n\nWolstencroft K, Owen S, Krebs O, et al.: SEEK: a systems biology data and model management platform. BMC Syst Biol. 2015; 9(1): 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReich M, Liefeld T, Ocana M, et al.: GenomeSpace: an environment for frictionless bioinformatics. F1000Posters. Poster. 2013. 4: 804. Reference Source\n\nGaramszegi S, Mesirov JP, The GenomeSpace Team: GenomeSpace: An environment for frictionless bioinformatics [v1; not peer reviewed]. F1000Res. Poster. 2015; 4(ISCB Comm J.): 349. Publisher Full Text\n\nQu K, Garamszegi S, Wu F, et al.: Integrative genomic analysis by interoperation of bioinformatics tools in GenomeSpace. Nat Methods. 2016; 13(3): 245–247. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSandve GK, Gundersen S, Rydbeck H, et al.: The Genomic HyperBrowser: inferential genomics at the sequence level. Genome Biol. 2010; 11(12): R121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSandve GK, Gundersen S, Johansen M, et al.: The Genomic HyperBrowser: an analysis web server for genome-scale data. Nucleic Acids Res. 2013; 41(Web Server Issue): W133–W141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimovski B, Vodák D, Gundersen S, et al.: GSuite HyperBrowser: integrative analysis of dataset collections across the genome and epigenome. GigaScience. 2017; 6(7): 1–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiardine B, Riemer C, Hardison RC, et al.: Galaxy: a platform for interactive large-scale genome analysis. Genome Res. 2005; 15(10): 1451–1455. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAfgan E, Baker D, van den Beek M, et al.: The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res. 2016; 44(W1): W3–W10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar S, Skjaeveland A, Orr RJ, et al.: AIR: A batch-oriented web program package for construction of supermatrices ready for phylogenomic analyses. BMC Bioinformatics. 2009; 10(1): 357. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar S, Krabberød AK, Neumann RS, et al.: BIR Pipeline for Preparation of Phylogenomic Data. Evol Bioinform Online. 2015; 11: 79–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalaš M: Efforts towards accessible and reliable bioinformatics. PhD thesis, University of Bergen, Norway, 2015. Publisher Full Text\n\nLinden M, Procházka M: ELIXIR Authentication and Authorisation Infrastructure (AAI) [version 1; not peer reviewed]. F1000Res. Poster. 2016; 5(ELIXIR): 332. Publisher Full Text\n\nFielding RT: Architectural Styles and the Design of Network-based Software Architectures. PhD thesis, University of California, Irvine, 2000. Reference Source\n\nRichardson L, Ruby S, Hansson DH: RESTful Web Services. O’Reilly, 2007. Reference Source\n\nRichardson L, Amundsen M, Ruby S: RESTful Web APIs. O’Reilly, 2013. Reference Source\n\nAzab A, Domanska D: Software Provisioning Inside a Secure Environment as Docker Containers Using STROLL File-System. In 2016 16th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid). IEEE, 2016; 674–683. Publisher Full Text\n\nEichner C, Dondrup M, Nilsen F: RNA sequencing reveals distinct gene expression patterns during the development of parasitic larval stages of the salmon louse (Lepeophtheirus salmonis). J Fish Dis. 2018; 41(6): 1005–1029. PubMed Abstract | Publisher Full Text\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim D, Langmead B, Salzberg SL: HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015; 12(4): 357–360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobertsen EM, Kahlke T, Raknes IA, et al.: META-pipe-Pipeline Annotation, Analysis and Visualization of Marine Metagenomic Sequence Data. CoRR. 2016; abs/1604.04103. Reference Source\n\nAgafonov A, Mattila K, Tuan CD, et al.: META-pipe cloud setup and execution [version 2; referees: 1 approved, 1 approved with reservations]. F1000Res. 2018; 6(ELIXIR): 2060. Publisher Full Text\n\nRaknes IA, Bongo LA: META-pipe Authorization service [version 1; referees: 2 approved with reservations]. F1000Res. 2018; 7(ELIXIR): 32. Publisher Full Text\n\nRobertsen EM, Raknes IA, Tartari G, et al.: ELIXIR Pilot: Marine metagenomics [version 1; not peer reviewed]. F1000Res. Poster. 2016; 5(ELIXIR): 864. Publisher Full Text\n\nRobertsen EM, Denise H, Mitchell A, et al.: ELIXIR pilot action: Marine metagenomics – towards a domain specific set of sustainable services [version 1; referees: 1 approved, 2 approved with reservations]. F1000Res. 2017; 6(ELIXIR): 70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014; 30(7): 923–930. PubMed Abstract | Publisher Full Text\n\nLaw CW, Chen Y, Shi W, et al.: voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014; 15(22): R29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSamy JKA, Mulugeta TD, Nome T, et al.: SalmoBase: an integrated molecular data resource for Salmonid species. BMC Genomics. 2017; 18(1): 482. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParkinson H, Sarkans U, Shojatalab M, et al.: ArrayExpress--a public repository for microarray gene expression data at the EBI. Nucleic Acids Res. 2005; 33(Database issue): D553–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVizcaíno JA, Reisinger F, Côté R, et al.: PRIDE: Data Submission and Analysis. Curr Protoc Protein Sci. 2010; Chapter 25: Unit 25.4. PubMed Abstract | Publisher Full Text\n\nIson J, Rapacki K, Ménager H, et al.: Tools and data services registry: a community effort to document bioinformatics resources. Nucleic Acids Res. 2016; 44(D1): D38–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTekle KM, Li X, Zhang W, et al.: elixir-no-nels/nels-core: NeLS core 2018-05-16. Zenodo. 2018. Data Source\n\nBongo AL, Raknes IA: META-pipe Galaxy Wrapper. Software version. Zenodo. 2018. Data Source"
}
|
[
{
"id": "35600",
"date": "16 Jul 2018",
"name": "Olivier Collin",
"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 the architecture of NeLS, the Norwegian e-Infrastructure for Life Sciences. Use cases provide an overview on how the NeLS infrastructure operates. The infrastructure can deal with sensitive and non-sensitive data.\n\nThe microservices oriented architecture allowing several universities to propose a national service is well explicited. The integration with data-generating platforms will help users to manage efficiently their research processes.\n\nThe article is well written and provides a good overview of the infrastructure. The description of the use case improves the understanding of the operations. Additional informations and source code are available on a GitHub repository.\n\nComments\n\nMajor remark :\n\nThe data management could be more explicitely described in the article.\n\nThe data analysis workflow is well described from the data analysis point of view but the usage of SEEK in the NeLS infrastructure is only briefly mentioned. Some additional description on SEEK usage and interest for NeLS users could improve the article.\n\nIn the Figure 1 : there is a mention of \"data curation into structured storage\". This data curation is not explained in the text and this raises some questions. Is this data curation using SEEK ?\n\nMinor remarks :\n\nIn the abstract, it is stated that \"NeLS is also integrated with the SEEK platform in order to store large data files produced by experiments described in SEEK\". This should be rephrased since it gives the impression that the SEEK platform is used to store large data files.\n\nIn the second chapter of the introduction, the flow of ideas is not clear. A comparison is made between BaseSpace, Geneious and SEEK before mentioning GenomeSpace. Why compare BaseSpace, Geneious and SEEK that are tools that do not really compare since the first two are focused on data analysis and the last on data management? Maybe there could be some rephrasing needed (this could also help to improve the data management focus mentioned in my first comment).\nConcerning BaseSpace and Geneious, it could be less judgemental to say that their service is chargeable instead of expensive. And to say that their platform is private. It could be interesting to describe briefly why the GenomeSpace setup could not be adapted.\n\nIn the data reproducibility chapter some reference genomes are made available on the five Galaxy instances. Is it possible to describe how this is achieved and how everything is synced?\n\nSome quantitative data about the number of users are needed in order to better estimate the community covered.\n\nIn the abstract there is a mention of a \"project-oriented fashion\". This expression is strange to me. Maybe replacing fashion by way or method or mode?\n\nTypo :\n\nThe headlines are numbered and sometimes there is a dot after the number and sometimes not. And they are not numbered in the templates if I am not mistaken.\n\nIn 2.1 Authentication whether the user is who he claims to be instead of whether the user is who they\n\nPage 9 col 1 line 29 : e-infrastructure\n\nPage 9 col 2 line 15 : some strange font glitches in the pdf file (but not online)\n\nPage 9 col 2 line 17 : can help in instead of helpin\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "35602",
"date": "13 Sep 2018",
"name": "Wolfgang Mueller",
"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 am a member of ELIXIR, and as member of the FAIRDOM project, I am heavily involved with the SEEK system whose link to NeLS is hinted inside the paper. However, I have been only losely involved with that, so I was deemed to be an appropriate reviewer. In order to make things crystal clear, I hereby make my status visible.\nIn my view, this paper has diverse functions:\nIt provides an architectural view on how a comparatively lightweight combination of a variety of pre-existing tools can yield a powerful national research infrastructure.\n\nIt provides a short justification of many architectural decisions.\n\nIt describes main steps in an ordinary NeLS project.\n\nIt provides a description of typical use cases and a reference of workflows that can be run by users.\n\nI find the paper clear and readable.\nI second the major remark by Olivier Collin.\nI was also surprised to find a whole paper about an infra structure that does not reference the FAIR principles explicitly. I think they could provide some point of reference regarding the handling of metadata, identifiers, licenses.\nHowever, it is only a matter of making existing links to FAIR more explicit.\n\nOne suggestion: Please adjust the numbering in the drawings with the numbering of subsections in the paper. This would make things much easier.\nBullet point 2.3 (c) references subsection 2.4, shouldn't this be 2.8, the public API? Please clarify.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-968
|
https://f1000research.com/articles/7-960/v1
|
28 Jun 18
|
{
"type": "Case Report",
"title": "Case Report: Severe back pain, epigastric distress and refractory nausea; an unusual presentation of mediastinal bronchogenic cyst",
"authors": [
"Saeed Ali",
"Abdul Rauf",
"Ling Bing Meng",
"Zeeshan Sattar",
"Sana Hussain",
"Umair Majeed",
"Saeed Ali",
"Abdul Rauf",
"Ling Bing Meng",
"Sana Hussain",
"Umair Majeed"
],
"abstract": "Background: Bronchogenic cysts are congenital malformations from abnormal budding of embryonic foregut and tracheobronchial tree. We present a case of bronchogenic cyst with severe back pain, epigastric distress and refractory nausea and vomiting.\n\nCase Presentation: A 44-year-old Hispanic female presented with a 3-week history of recurrent sharp interscapular pain radiating to epigastrium with refractory nausea and vomiting. She underwent cholecystectomy 2-years ago. Computed tomography (CT) abdomen at that time showed a subcarinal mass measuring 5.4 X 5.0 cm. Subsequent endoscopic ultrasound diagnosed it as a bronchogenic cyst. Endobronchial ultrasound (EBUS) guided aspiration resulted in incomplete drainage and she was discharged after partial improvement. Current physical examination showed tachycardia and tachypnea with labs showing leukocytosis, elevated inflammatory markers, and hypokalemic metabolic alkalosis. CT chest showed an increased size of the bronchogenic cyst (9.64 X 7.7 cm) suggestive of possible partial cyst rupture or infected cyst. X-ray esophagram ruled out esophageal compression or contrast extravasation. Patient’s symptoms were refractory to conservative management. The patient ultimately underwent right thoracotomy with cyst excision that resulted in complete resolution of symptoms. Conclusion: Bronchogenic cysts are the most common primary cysts of mediastinum with the prevalence of 6%. The most common symptoms are chest pain, dyspnea, cough, and stridor. Diagnosis is made by chest X-Ray and CT chest. Magnetic resonance imaging chest and EBUS are more sensitive and specific. Symptomatic cysts should be resected unless surgical risks are high. Asymptomatic cysts in younger patients should be removed due to low surgical risk and potential late complications. Watchful waiting has been recommended for asymptomatic adults or high-risk patients. This case presents mediastinal bronchogenic cyst as a cause of back, nausea and refractory vomiting. Immediate surgical excision in such cases should be attempted, which will lead to resolution of symptoms and avoidance of complications.",
"keywords": [
"bronchogenic",
"cyst",
"back",
"pain"
],
"content": "Introduction\n\nBronchogenic cysts are congenital malformations of the bronchial tree. They result from the anomalous development of ventral foregut and tracheobronchial tree1. They can present as a mediastinal mass that may enlarge and cause local compression. They are the most common primary cysts of the mediastinum with a prevalence of 6%2. Approximately 79% of cysts are located in the middle mediastinum, 17% in posterior mediastinum and 3% in the anterior mediastinum3.\n\nAlmost 75% of bronchogenic cysts are asymptomatic. Symptoms vary with age at presentation and with size and location of the cyst. The common symptoms include chest pain (22%), dyspnea (12%), cough (7%), stridor (7%) and respiratory compromise due to tracheal/bronchial compression (10%). Unusual manifestations are dysphagia (1%), pneumothorax (1%), and superior vena cava syndrome (1%)3.\n\nDiagnosis is made by chest X-Ray and computed tomography (CT) chest although magnetic resonance imaging (MRI) chest and endobronchial ultrasound are highly sensitive and specific4. MRI chest can provide additional information about the consistency and nature of the cyst depending upon the presence of proteinous contents in the fluid. In general, bronchogenic cyst appears hypo-intense on T1-weighed images and hyper-intense on T2-weighed images5. Endoscopic ultrasound (EUS) is a relatively invasive procedure for the diagnosis of the bronchogenic cyst.\n\nTreatment options depend on patient’s age and symptoms. Symptomatic bronchogenic cyst are managed surgically with resection, Endobronchial ultrasound (EBUS) guided aspiration, and video-assisted thoracoscopic surgery being a minimally invasive procedure6. Thoracotomy is performed for difficult cases. Asymptomatic cysts in younger patients should be removed due to low surgical risk and potential late complications such as infection, hemorrhage or neoplasia. Watchful waiting has been recommended for asymptomatic adults or high-risk patients. Percutaneous drainage or alcohol ablation has been performed in selected cases4. We present a case of a mediastinal bronchogenic cyst in a 44-year-old female presenting in the form of severe back pain, epigastric distress and nausea.\n\n\nCase report\n\nA 44-year-old Hispanic female presented with a three-week history of recurrent sharp interscapular pain radiating to the mid-sternal and epigastric region associated with refractory nausea and vomiting. She underwent cholecystectomy for intermittent epigastric pain two years ago. CT abdomen at that time showed a subcarinal mass measuring 5.4 X 5.0 cm (Figure 1). Subsequent EUS diagnosed it as a bronchogenic cyst. EBUS guided aspiration resulted in an incomplete drainage and she was discharged after partial improvement.\n\nCurrent physical examination showed a heart rate of 126/min (normal range: 60–100/min) and respiratory rate of 20/min (normal range: 12–20/min). Initial labs showed white cell count of 10.58X103/uL (normal range: 4000–11X103uL), elevated inflammatory markers [ESR of 63mm/hr (normal range: 0–20 mm/hr); CRP of 116 mg/L (normal range: <3.0 mg/L)], and hypokalemic metabolic alkalosis. Electrocardiogram showed non-specific T wave changes. Chest X-ray showed right posterior mediastinal mass (Figure 2).\n\nCT chest showed an increase in the size of the bronchogenic cyst (9.64 X 7.7 cm) with small right pleural effusion (Figure 3).\n\nThe X-ray and CT findings were consistent with partial cyst rupture or an infected cyst. X-ray esophagogram ruled out esophageal compression or contrast extravasation. The patient’s symptoms were refractory to conservative analgesic and antiemetic measure like Dilaudid (hydromorphone) 1 mg IV every 3 hourly and Zofran (Ondansetron) 4 mg IV every 4 hourly for pain and nausea/vomiting respectively. Cardiothoracic surgery was consulted and the patient underwent right thoracotomy and surgical cyst excision. Cyst pathology was consistent with severe inflammatory changes. Within 24–48 hours after the surgery, the resolution in the patient’s symptoms were noted in terms of decrease in need of pain and nausea medications. Repeated labs showed resolution of leukocytosis.\n\n\nDiscussion\n\nBronchogenic cysts are the rare benign congenital malformation resulting from the anomalous budding of ventral foregut and tracheobronchial tree1. They are part of the bronchopulmonary foregut malformations. They are more commonly found in the mediastinum in the paratracheal and subcarinal regions. Less commonly they are found in the lung parenchyma.\n\nBronchogenic cyst may present with unusual symptoms posing a diagnostic challenge. Signs and symptoms of bronchogenic cyst mainly depend upon its location, size, and compression of surrounding structures like esophagus, trachea, and bronchus7. Most common presentation in adult patients includes chest pain, cough, dyspnea and dysphagia8.\n\nOur patient presented with unusual symptoms of severe backache, epigastric discomfort, refractory nausea, and vomiting. It is believed that the back pain is caused by stretching of the nerves supplying the parietal pleura while the epigastric distress is caused by the stimulation of the vagal nerve4,9.\n\nIn our case, repeat CT chest confirmed an increase in the size of a bronchogenic cyst with small right pleural effusion. Considering that approximately 10% of the patients develop respiratory problems due to tracheal or bronchial compression, we performed X-ray esophagogram and ruled out esophageal compression or contrast extravasation.\n\nAt present management of symptomatic bronchogenic cyst is surgical as discussed. Management of asymptomatic cyst is controversial. It has been suggested that as most of the cysts eventually cause some symptoms or serious complications like respiratory distress from airway compression, infection and airway fistulae, surgical resection in asymptomatic patients is recommended. Also, postoperative surgical complications are more common in patients with symptomatic cysts as compared to asymptomatic cysts further implying the benefits of surgical resection of asymptomatic cysts10.\n\n\nConclusion\n\nThis case highlights the importance of recognizing bronchogenic cyst as a cause of severe back pain, refractory nausea, and vomiting. Back pain is caused by stretching of nerves supplying the parietal pleura; while nausea is caused by stimulation of vagus nerve. Prompt surgical excision can lead to complete symptom resolution and avoidance of future complications.\n\n\nConsent\n\nWritten informed consent was obtained from the patient for the publication of this case report and any accompanying images.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "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 case was previously presented at World Congress of Gastroenterology (WCG) at American College of Gastroenterology (ACG) 2017 as a poster presentation: P1135 - Severe Back Pain, Epigastric Distress and Refractory Nausea: An Unusual Presentation of Mediastinal Bronchogenic Cyst.\n\n\nReferences\n\nLimaïem F, Ayadi-Kaddour A, Djilani H, et al.: Pulmonary and mediastinal bronchogenic cysts: a clinicopathologic study of 33 cases. Lung. 2008; 186(1): 55–61. PubMed Abstract | Publisher Full Text\n\nLyerly H, Sabiston D: Primary neoplasms and cysts of the mediastinum. Pulmonary diseases and disorders. 1988; 3: 2090–1.\n\nMcAdams HP, Kirejczyk WM, Rosado-de-Christenson ML, et al.: Bronchogenic cyst: imaging features with clinical and histopathologic correlation. Radiology. 2000; 217(2): 441–6. PubMed Abstract | Publisher Full Text\n\nTiwari MK, Yadav R, Mathur RM, et al.: Mediastinal bronchogenic cyst presenting with dysphagia and back pain. Lung India. 2010; 27(2): 86–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuramatsu T, Shimamura M, Furuichi M, et al.: Thoracoscopic resection of mediastinal bronchogenic cysts in adults. Asian J Surg. 2011; 34(1): 11–4. PubMed Abstract | Publisher Full Text\n\nWeber T, Roth TC, Beshay M, et al.: Video-assisted thoracoscopic surgery of mediastinal bronchogenic cysts in adults: a single-center experience. Ann Thorac Surg. 2004; 78(3): 987–91. PubMed Abstract | Publisher Full Text\n\nChe WC, Zang Q, Zhu Q, et al.: Lipoma-Like Bronchogenic Cyst in the Right Chest Sidewall: A Case Report and Literature Review. Ann Thorac Cardiovasc Surg. 2016; 22(6): 370–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPages ON, Rubin S, Baehrel B: Intra-esophageal rupture of a bronchogenic cyst. Interact Cardiovasc Thorac Surg. 2005; 4(4): 287–8. PubMed Abstract | Publisher Full Text\n\nWCOGACG2017: Severe Back Pain, Epigastric Distress and Refractory Nausea: An Unusual Presentation of Mediastinal Bronchogenic Cyst. Orlando: ACG; 2017. Reference Source\n\nPatel SR, Meeker DP, Biscotti CV, et al.: Presentation and management of bronchogenic cysts in the adult. Chest. 1994; 106(1): 79–85. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "37054",
"date": "19 Sep 2018",
"name": "Dipak K. 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\nBronchogenic cyst diagnosis on EBUS and complication of Infection.\nNeeds to alert that Needling with EBUS needle is a recognised complication and some practitioners caution against using EBUS needle if diagnosis can be secured by other means but sometime if there is uncertainty about the diagnosis one has to do the EBUS needling.\nOne has to understand the potential chance of recurrence in absence of complete removal of cyst wall after surgery and possibly need mentioning.\nOn the whole the article is well written.\nsimilar article we presented: Mogal et al1\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "37985",
"date": "20 Sep 2018",
"name": "Ali Dahhan",
"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 case report of bronchogenic cyst. It highlights the importance of imaging, including computed tomography. The article is well-written. One minor suggestion is to consider adjusting the position of the arrows in Figures 1 and 3 as it points at the right lung rather than the cyst - instead an asterix over the cyst can also be practical without hiding parts of the figures.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-960
|
https://f1000research.com/articles/7-592/v1
|
15 May 18
|
{
"type": "Method Article",
"title": "Preparation of organotypic brain slice cultures for the study of Alzheimer’s disease",
"authors": [
"Cara L. Croft",
"Wendy Noble"
],
"abstract": "Alzheimer's disease, the most common cause of dementia, is a progressive neurodegenerative disorder characterised by amyloid-beta deposits in extracellular plaques, intracellular neurofibrillary tangles of aggregated tau, synaptic dysfunction and neuronal death. There are no cures for AD and current medications only alleviate some disease symptoms. Transgenic rodent models to study Alzheimer’s mimic features of human disease such as age-dependent accumulation of abnormal beta-amyloid and tau, synaptic dysfunction, cognitive deficits and neurodegeneration. These models have proven vital for improving our understanding of the molecular mechanisms underlying AD and for identifying promising therapeutic approaches. However, modelling neurodegenerative disease in animals commonly involves aging animals until they develop harmful phenotypes, often coupled with invasive procedures. In vivo studies are also resource, labour, time and cost intensive. We have developed a novel organotypic brain slice culture model to study Alzheimer’ disease which brings the potential of substantially reducing the number of rodents used in dementia research from an estimated 20,000 per year. We obtain 36 brain slices from each mouse pup, considerably reducing the numbers of animals required to investigate multiple stages of disease. This tractable model also allows the opportunity to modulate multiple pathways in tissues from a single animal. We believe that this model will most benefit dementia researchers in the academic and drug discovery sectors. We validated the slice culture model against aged mice, showing that the molecular phenotype closely mimics that displayed in vivo, albeit in an accelerated timescale. We showed beneficial outcomes following treatment of slices with agents previously shown to have therapeutic effects in vivo, and we also identified new mechanisms of action of other compounds. Thus, organotypic brain slice cultures from transgenic mouse models expressing Alzheimer’s disease-related genes may provide a valid and sensitive replacement for in vivo studies that do not involve behavioural analysis.",
"keywords": [
"Organotypic brain slice culture",
"neurodegeneration",
"amyloid-β",
"tau",
"Alzheimer’s disease",
"transgenic mice",
"reduction"
],
"content": "\n\n\n\nNeurodegeneration in AD is progressive and age-dependent and therefore many experiments rely upon animals developing a moderate to severe phenotype prior to assessment of disease parameters\n\nAging slices in a culture dish obviates this requirement, as well as providing an in vitro system in which all neural cell types are present, and functional and anatomical connectivity is retained\n\nBrain slice cultures can be used as an alternative to some in vivo experiments thereby reducing the numbers of animals required for such studies\n\nDisease phenotypes are accelerated in brain slice culture models so equivalent data can be obtained from neonatal mice in 1 month of in vitro experiments rather than 12 months of aging in vivo from fewer total mice\n\n36 brain slices can be cultured from a single mouse brain, so studies can be refined as it is possible to assess several parameters or disease-modifying agents in a single brain thereby minimising biological variation\n\nThis model is considerably more time and cost effective when compared to in vivo studies\n\nThe 3xTg-AD slice culture model is suitable for biochemical and immunohistochemical research into Alzheimer’s disease and related tauopathies\n\nLong-term organotypic brain slice cultures have the potential for use in all aspects of neuroscience\n\nCultures from other tissues will have utility for many other fields of biological and biomedical research\n\n\nIntroduction\n\nAlzheimer’s disease (AD), the most common cause of dementia, currently affects around 35 million people worldwide and carries huge societal and economical costs. AD is a multi-factorial disease with two major pathological hallmarks; extracellular plaques composed of β-amyloid (Aβ) and intracellular neurofibrillary tangles containing aggregated post-translationally modified tau. The only available treatments for AD target the symptoms of disease, but not disease course. Intensive research efforts are ongoing to better understand the biological causes of disease so that effective disease-modifying therapies can be developed.\n\nPerhaps the most accepted models for AD research are transgenic rodents that express wild-type or mutant human AD-associated genes and recapitulate key molecular phenotypes of AD. Mice are generally one of the best accepted animal models in neuroscience research since there is significant homology between the human and mouse genome, mice have a relatively short life span, well-defined genetic backgrounds, are amenable to further genetic manipulations, enable assessment of changes in behaviour, cognition, brain biochemistry and physiology during disease progression, and a battery of well-characterised tasks are available to study behaviour and cognition1.\n\nOur calculations, based on a PubMed search using the terms “Alzheimer’s + transgenic + mouse” and assuming an average of 30 mice for each of the 700 papers published, suggest that over 20,000 transgenic mice were used in AD research in 2017. Due to the age-related neurodegenerative nature of the disease, this research often involves aging several cohorts of mice to observe disease progression. Allowing mice expressing AD-related genes to reach the terminal stages can result in severe phenotypes, and some studies are coupled with invasive procedures such as advanced live imaging or collection of interstitial or cerebrospinal fluids.\n\nAlternatives to in vivo AD research in mammalian systems include rodent and human cell lines manipulated to express genes of interest, however these can be criticised for lacking key features of differentiated post-mitotic neurons and can be prone to artefacts resulting from protein over-expression. Dissociated neural cell cultures are commonly used as a readily tractable model in which pathways of interest can be manipulated, however even co-culture systems do not completely replicate the complex connections between different neural cell types and the brain vasculature, and they cannot model the synaptic and anatomical connectivity of mammalian brain. The latter is also true for neural cells derived from human induced pluripotent stem cells (iPSCs). Recent reports using iPSC-derived neurons also highlights the extensive time in culture required before even subtle disease relevant changes are observed in these human neural cells2.\n\nOrganotypic brain slice cultures are a well-established technique. Slice cultures maintain a three-dimensional organisation with the preservation of cytoarchitecture and cell populations, and are an accessible system lending their application to electrophysiology, morphology and biochemical analyses3,4. The interface-slice culture method established by Stoppini and colleagues in 19915 is the most common method to culture brain sections ex vivo. This relatively simple method cultures brain tissue explants from neonatal mice/rats on a porous membrane insert that acts as an interface between the humidified incubator atmosphere and the culture medium that provides nutrition5. Cultures can be maintained for several weeks in culture after explant and continue to develop and mature once plated6,7. Importantly, slice cultures are prepared from neonatal mice precluding aging transgenic mice to a stage when they develop a detrimental phenotype. In addition, 36 slices containing the cortex and hippocampus can be prepared from one postnatal mouse brain allowing multiple variables to be tested or manipulated in tissue from the same mice, thereby considerably reducing the number of mice required as well as minimising experimental variation.\n\nAD researchers are beginning to embrace organotypic brain slice culture models, with recent papers describing AD-relevant changes in slice cultures prepared from mice that overexpress amyloid precursor protein or are seeded with Aβ, which show some accumulation of Aβ and synaptic alterations8–10. Others have shown that slice cultures prepared from mice that overexpress human tau can accumulate phosphorylated and some sarkosyl-insoluble tau10,11. We recently demonstrated that slice cultures prepared from 3xTg-AD mice12 overproduce Aβ, accumulate somatodendritic and synaptic phosphorylated tau at an accelerated rate compared to 3xTg-AD mice13, allowing study of Aβ-tau interactions and AD disease pathways ex vivo.\n\nThe utility of slice cultures for drug discovery efforts has previously been reviewed14, and we have validated 3xTg-AD slice cultures for this purpose by showing that the effects of disease-modifying compounds observed in vivo can be recapitulated in slice culture15. We also identified novel targets for compounds, further demonstrating the usefulness of slice cultures for therapeutic development. In this paper, we provide detailed methods for the preparation of organotypic brain slice cultures for the study of AD and we discuss the advantages of this model system in terms of the 3Rs in AD research, most specifically in reducing mouse numbers. We believe that this model system will be of most benefit to researchers in the neurodegeneration field, who are either focussed on understanding the biological mechanisms underpinning disease or who aim to screen and test the efficacy of novel disease-modifying therapeutics.\n\n\nMethods overview\n\n3xTg-AD mice were obtained under a material transfer agreement from Professor Frank LaFerla (University of California Irvine, USA) and maintained as a breeding colony at King’s College London. 3xTg-AD mice express mutant human PS1 (M146V), APP (Swe, K670N, M671L) and tau (P301L) transgenes12. Wild-type (WT) mice of an identical background strain (F2 hybrid: C57BL/6J and 129S1/SvImJ) were maintained as background controls. All housing and experimental procedures were carried out in compliance with the local ethical review panel of King’s College London under a UK Home Office project licence held in accordance with the Animals (Scientific Procedures) Act 1986 and the European Directive 2010/63/EU. Male and female mice were used in this study. After weaning, mice were housed in single sex groups in standard 40 x 25 x × 12 cm plastic cages. Bedding consisted of kiln dried aspen shavings and paper sizzle nest material (Datesand Ltd, Manchester, UK). Water and food were available (Picolab rodent diet 20; #5053; Lab Diet, St Louis, MO) ad libitum. Animals were housed at 19–22°C, humidity 55%, 12 h:12 h light: dark cycle with lights on at 07:30. Cages were cleaned once every two weeks, with mice handled by the tail by experienced animal care staff to transfer them between cages.\n\nOrganotypic brain slice cultures were prepared from postnatal day 8-9 3xTg-AD16,17 and background control wild-type mice as previously described11. Briefly, pups were culled by decapitation in accordance with the UK Animals in Scientific Procedures Act (1986). Brains from pups were bisected into hemi-brains by a single cut along the midline. The cerebellum, thalamus and brainstem were removed and discarded to leave the cortex, hippocampus and connecting areas. These were kept in ice-cold dissection buffer (1.25 mM KH2PO4 pH 7.4, 124 mM NaCl, 3 mM KCl, 8.19 mM MgSO4, 2.65 mM CaCl2, 3.5 mM NaHCO3, 10 mM glucose, 2 mM ascorbic acid, 39.4 µM ATP in ultrapure H2O, sterile filtered (0.2 µm)) with constant oxygenation throughout the preparation procedure. 350 µm coronal slices were cut using a McIlwain Tissue Chopper (Stoelting Europe, Ireland). Eighteen slices from each hemi-brain were collected and 3 consecutive slices per well were positioned on interface-style Millicell culture inserts (Millipore (UK) Ltd.) in 6 well culture plates (ThermoFisher Scientific, UK) containing 1 mL of sterile slice culture medium (Basal medium eagle (BME), 26.6 mM HEPES, pH 7.1, 19.3 mM NaCl, 5 mM NaHCO3, 511 µM ascorbic acid, 40 mM glucose, 2.7 mM CaCl2, 2.5 mM MgSO4, 1 % (v/v) GlutaMAX (Life Technologies, Paisley, UK), 0.033% (v/v) insulin, 0.5% (v/v) penicillin/streptomycin (Life Technologies), in ultrapure H2O, sterile filtered (0.2 µm), plus 25% (v/v) heat inactivated horse serum (ThermoFisher, UK). Three hours after plating, the culture medium was removed by aspiration and replaced with 1 mL of pre-warmed fresh sterile culture medium. Brain slices were incubated at 37°C and the culture medium was changed from the bottom of each well every 2 to 3 days. Slices are maintained for a minimum of 14 days in vitro prior to treatment/harvesting.\n\nSlice cultures can be pharmacologically or genetically modified using a number of methodologies. These methods are out of the scope of this publication but have previously been published by ourselves and others (for example, 8–11,13,15).\n\nOrganotypic brain slice cultures can be fixed on their membrane inserts in 4% PFA for 4 h and stained according to Croft et al.13. In brief, slice cultures are cut whilst still on their membranes and then treated as free-floating sections. Slice cultures are permeabilised for 18h in 0.5% Triton X-100 at 4°C and then blocked in 20% bovine serum albumin (BSA) for 4h at RT. Slice cultures are incubated with primary antibodies overnight at 4°C in 5% BSA, washed and then incubated with fluorophore-coupled secondary antibodies for 4h at ambient temperature. Slice cultures are washed a final time before mounting on slides with fluorescent mounting medium (Dako Ltd., Ely, UK) prior to imaging.\n\nAlternatively, tissue can be lysed for subcellular fractionation and/or biochemical analysis as described by us and others9–11,13,15. To prepare lysates for immunoblotting, slice culture medium is aspirated and slices washed once with ice-cold PBS. Slices are collected via scraping into ice-cold PBS. Cellular matter is pelleted by centrifugation at 7,000 g (av) for 30 seconds at ambient temperature. The supernatant is discarded and tissue pellets lysed in 100 μL ice-cold extra strong lysis buffer (10 mM Tris-HCl (pH 7.5), 0.5% (w/v) sodium dodecyl sulphate (SDS), 20 mM sodium deoxycholate, 1% (v/v) Triton-X-100, 75 mM sodium chloride, 10 mM ethylenediaminetetraacetic acid (EDTA), 2 mM sodium orthovanadate, 1.25 mM sodium fluoride) and protease inhibitor cocktail for mammalian tissues (Roche Diagnostics, UK). The suspension is then sonicated briefly (10 seconds) using a Vibra-Cell™ probe sonicator to improve sample handling. Slice lysates are centrifuged at 23,000 g(av) for 20 minutes at 4°C and the supernatant collected. The protein content of the slice lysates can be determined using a BCA protein assay (Pierce Endogen, Rockford, USA) and protein concentration normalised prior to immunoblotting or ELISA. Slice lysates are mixed with an equal volume of 2x sample buffer before immunoblotting.\n\nCulture medium can also be collected for analysis of its components, as we recently described for tau and Aβ13. Slice culture medium is replaced with Hank’s Balanced Salt Solution (HBSS; Life Technologies Ltd). HBSS is collected from the slice cultures and centrifuged at 12,000g for 10 min at 4°C to remove cell debris. Protein content in HBSS can be determined by ELISA by standard or sandwich ELISA.\n\n\nFull protocol for the model development\n\nMcIllwain Tissue Chopper (RRID:SCR_015798; Mickle Laboratory Engineering Co. Ltd., Surrey, UK)\n\nStereomicroscope for tissue dissection\n\nChopping discs (Product code: 752TC-CT; Campden Instruments Ltd., Loughborough, UK)\n\nBlades (Product code: TC752-1; Campden Instruments Ltd., Loughborough, UK)\n\n55mm diameter ashless filter paper (Product code: WHA1442055; Merck, UK)\n\nFlat (cover glass) forceps (Product code: 11074-02; Fine Science Tools, Heidelberg, Germany)\n\nCohan-Vannas Spring Scissors (Product code: 15000-02; Fine Science Tools, Heidelberg, Germany)\n\nMayo scissors (Product code: 14010-17; Fine Science Tools, Heidelberg, Germany)\n\nFine scissors, sharp (Product code: 14060-11; Fine Science Tools, Heidelberg, Germany)\n\nMoria MC17C perforated spoon-mini (Product code: 10370-19; Fine Science Tools, Heidelberg, Germany)\n\nMini hippocampal dissection tool (Product code: 10099-12; Fine Science Tools, Heidelberg, Germany)\n\nCell culture treated 6-well plates (Product code: 140675; ThermoFisher Scientific, UK)\n\nCell culture treated 10cm dish (Product code: 172931; ThermoFisher Scientific, UK)\n\nPTFE 30mm tissue culture inserts 0.4μm (Product code: PICM03050; Millipore, Fisher Scientific, UK)\n\nPasteur pipettes – sterile – individually wrapped (Product code: Z350621-400EA; Merck, UK)\n\n30ml Pyrex beaker (Produce code: CLS100030; Merck, UK)\n\nFine paintbrush\n\nCarbogen (95% Oxygen / 5% Carbon Dioxide)\n\nStandard tissue culture incubator (37°C / 5% Carbon Dioxide)\n\nSlice Culture Dissection buffer: 1.25 mM KH2PO4 pH 7.4, 124 mM NaCl, 3 mM KCl, 8.19 mM MgSO4, 2.65 mM CaCl2, 3.5 mM NaHCO3, 10 mM glucose, 2 mM ascorbic acid, 39.4 µM ATP in ultrapure H2O, sterile filtered (0.2 µm).\n\nSlice culture medium: Basal medium eagle (BME), 26.6 mM HEPES, pH 7.1, 19.3 mM NaCl, 5 mM NaHCO3, 511 µM ascorbic acid, 40 mM glucose, 2.7 mM CaCl2, 2.5 mM MgSO4, 1% (v/v) GlutaMAX (Life Technologies, Paisley, UK), 0.033% (v/v) insulin, 0.5% (v/v) penicillin/streptomycin (Life Technologies), in ultrapure H2O, sterile filtered (0.2 µm), plus 25% (v/v) heat inactivated horse serum (ThermoFisher, UK).\n\nBrain extraction:\n\nExperiments are performed under sterile conditions with tools sterilised by autoclaving prior to use. 70% EtOH is used to sterilize equipment and surfaces throughout the experiment. Postnatal day 8 or 9 WT or 3xTg-AD mice are used (Figure 1A–C).\n\n(A) After removal from the skull, brains are bisected along the midline using a razor blade. (B) The thalamus, cerebellum and brain stem are removed using the hippocampal dissection tool leaving the cortex, hippocampus and connected brain regions. (C) Two hemi-brains are kept in oxygenated dissection buffer throughout the procedure; one hemi-brain is stored whilst the other is processed. (D) A hemi-brain is placed on dampened filter paper on the cutting surface of a McIlwain tissue chopper. (E–F) 350 µm coronal slices are cut. (G–J) Slice cultures are sequentially separated under a dissection microscope using a hippocampal dissection tool. (K) Three slices are plated per well on Millipore membrane inserts in 6 well plates. Three consecutive slices can be placed in each well or slices plated randomly depending on experimental needs. Cultures are maintained by replacing the culture medium every 2–3 days.\n\n1. Pups are decapitated using Mayo scissors and death confirmed.\n\n2. Heads are transferred to a 10cm tissue culture dish containing oxygenated ice-cold dissection buffer.\n\n3. Fine scissors are used to remove hair and skin, cutting anteriorly from the base of the skull along the midline, revealing the brain and skull.\n\n4. Small spring scissors are used to carefully cut through the midline of the skull.\n\n5. The brain is bisected through the midline while remaining in the skull using a razor blade.\n\n6. The hippocampal dissection tool is used to cleanly remove the remove brainstem, cerebellum and thalamus which are discarded. The cortex, hippocampus and associated regions remain intact.\n\n7. The remaining tissue is removed from the skull and transferred to a glass beaker containing oxygenated dissection buffer.\n\n8. Repeat for the other hemi-brain. One hemi-brain will remain in dissection buffer, regularly re-oxygenated, while the other is processed.\n\nSlice culture preparation (Figure 1D–K):\n\n9. 1 mL of slice culture medium is added to each well of a 6-well culture plate. Flat cover glass forceps are used to add a Millicell culture insert into each well. The plate is returned to a 37°C incubator to ensure that culture medium is pre-warmed before slices are plated.\n\n10. A plastic chopping disc and fresh filter paper is placed on the cutting stage of a McIlwain tissue chopper. Three to four drops of dissection solution are used to dampen the filter paper and allow the hemi-brain to remain in place.\n\n11. A hemi brain is placed onto the filter paper and oriented for coronal sectioning (the front of the brain on the right-hand side).\n\n12. A drop of dissection solution is added to the brain to prevent the cutting blade from sticking.\n\n13. The section size on the McIllwain tissue chopper is adjusted to 350 µm. The blade should be manually positioned adjacent to the frontal region of the brain.\n\n14. The tissue chopper is started - the automated razor blade will cut 350 µm sections until manually switched off. The speed of cutting can be adjusted if necessary.\n\n15. The hemi-brain, remaining on the filter paper, is transferred to a fresh 10cm dish containing oxygenated dissection solution, and the dish is placed under a dissection microscope.\n\n16. Slices are manually separated using the hippocampal dissection tool, taking care to avoid ripping tissue. Using a plastic Pasteur pipette, individual slices are transferred to culture inserts in 6-well plate.\n\n17. From the leading edge (frontal cortex), three consecutive slices are positioned in each culture insert. Each hemi-brain is sectioned into 18 slices, equivalent to one 6-well plate. Alternatively, slices can be plated randomly and distributed throughout wells allowing the study of frontal, middle and rostral sections within each well.\n\n18. Care should be taken to ensure that slices do not overlap or make contact with the sides of the insert. A fine paintbrush can be used to move the slices and to ensure that no areas of the slices are folded or wrinkled.\n\n19. Excess dissection solution is removed from the slice culture inserts, and the 6-well plate is returned to an incubator and maintained in humid conditions at 37°C with 5% CO2.\n\nSlice culture maintenance:\n\n20. In sterile conditions, approximately three hours after plating a glass Pasteur pipette is used to aspirate culture medium. 1 mL pre-warmed fresh sterile culture medium is then added.\n\n21. Brain slices are incubated in humid conditions at 37°C with 5% CO2. Culture medium is replaced every 2 to 3 days, taking care not to move the inserts within each well. Any excess medium which collects in the insert is also removed during media changes, taking care not to disturb the slices.\n\n22. Slice cultures can be analysed from 14 days after plating, at which time lactate dehydrogenase release should have returned to basal levels10.\n\n• P8-P9 pups are used in this method, however other ages of pups are described in publications from other groups. It is likely that some optimisation may be required depending on the strain or transgenic line of mice being used.\n\n• Slice cultures are initially white in colour, but become translucent after 7 to 10 days in culture. White tissue remaining at this point is likely to signify unhealthy or dead tissue.\n\n• It is important to cleanly remove all of the thalamus, cerebellum and brainstem during dissection since these are detrimental to slice survival under these conditions. These tissues can be cultured but using alternative protocols.\n\n• It is important that excess dissection buffer is removed from the slice cultures once they have been plated since prolonged exposure to this buffer in culture can affect slice health.\n\n\nResults\n\nWe have previously characterised organotypic brain slice cultures prepared from 3xTg-AD mice in comparison to brain from aged in vivo 3xTg-AD mice13. We examined abnormalities in β-amyloid and tau that accumulate in AD brain. We found that 3xTg-AD slice cultures show an accelerated development of highly phosphorylated and oligomeric/64kDa tau species, some of which redistributed to synaptic compartments by 28 days in vitro (DIV). Similar changes in vivo are typically observed from 12 months of age. An accelerated accumulation of potentially pathogenic Aβ species were also observed in brain slice cultures from 3xTg-AD mice, with significantly increased Aβ1-42 levels detected at 28 DIV in slices. In comparison, we could only detect significant changes in Aβ1-42 amounts in 3xTg-AD brain in 12-month old mice. Thus, disease-associated protein species show an accelerated accumulation in long-term brain slice cultures in comparison to in vivo. Using differential centrifugation approaches we were also able to show the differential accumulation of phosphorylated and dephosphorylated tau species in synaptic compartments and at membranes, in agreement with previous reports using human tissue and primary cell cultures18,19. Table 1 provides a summary of molecular changes in the slice culture model in comparison to findings made using tissue from aged 3xTg-AD mice. Primary data is available here (Dataset 120).\n\nPrimary references are shown. AD: Alzheimer's disease; DIV: days in vitro.\n\nThere is also a great deal of versatility in the methods that can be used to assess disease changes in this model. We have confirmed that methods including, but not limited to the following, can be used with slice cultures; biochemical changes can be assessed by ELISA or immunoblotting, slice cultures can be examined by immunohistochemistry and confocal microscopy, sufficient material is present to allow differential centrifugation to enrich cell compartments such as synaptosomes, membrane and cytosol. Additionally, cell death can be measured using lactate dehydrogenase assays and the release of disease-associated proteins into culture medium can be quantified13,21. Others have also shown that brain slice cultures are amenable to ultrastructural analysis22, and live calcium imaging11.\n\nIn addition to comparing molecular features of AD in slice cultures in comparison to in vivo, we also validated the use of slice cultures for studying the effects on tau phosphorylation of acute application of compounds in comparison to their reported effects in previously published in vivo studies.\n\nLithium chloride (LiCl) can inhibit activity of the prominent tau kinase, glycogen synthase kinase-3β (GSK-3)23, which targets many of the tau residues known to be aberrantly phosphorylated in AD16. Treatment of 12-month-old 3xTg-AD mice for 3 months with the GSK-3 inhibitor, LiCl, was shown to reduce phosphorylation of tau at Thr 181, Ser 202/Thr 205, Thr 231, and Thr 212/Ser 214, but not Ser 396/40424. We found that application of LiCl for 4 h to 3xTg-AD brain slice cultures at 28 days days in vitro (DIV) resulted in significantly reduced tau phosphorylation at the Ser396/404 and Ser202/Thr205 epitopes, in addition to causing a subtle reduction in total tau amounts when compared to slices treated with control (NaCl). There was also a notable shift in the apparent molecular weight of tau in lysates from LiCl treated slice cultures, which is characteristic of reduced tau phosphorylation15.\n\nAnother potential therapeutic approach for AD is to use microtubule-stabilising agents to recover the loss of function, which occurs following the detachment of phosphorylated tau from the microtubule cytoskeleton23. Neuroprotective effects of the peptide NAPVSIPQ have previously been reported in 12-month-old 3xTg-AD mice. Mice administered NAPVSIPQ for 3 months showed reduced phosphorylation of tau at Ser 202/Thr 205 and Thr 231, but not at Ser 202 alone. Treatment of 28 DIV 3xTg-AD slice cultures with 100 nM NAPVSIPQ for 24 h significantly reduced tau phosphorylation at the Thr231 epitope, but did not alter the total amount of tau when compared to control cultures15, thus we find that equivalent treatment of 3xTg-AD organotypic brain slice cultures recapitulates previous in vivo findings conducted in 3xTg-AD mice24,33,34.\n\nPrimary data is available at https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5547074/15.\n\nWe also used the slice culture model to identify novel tau-directed effects of BTA-EG415, a compound that had previously shown Aβ-binding effects and synaptic protection35–37. A growing body of publications have further demonstrated the tractability of the slice culture system, including pharmacological manipulation of Aβ production9 and tau aggregation11, in addition to modulation of microglial composition to examine the phagocytic action of microglia on Aβ deposits22.\n\nThese data suggest that potential therapeutic agents can be sensitively examined in organotypic brain slice culture models. Since a number of methods can be applied to study slice culture tissues, this system should be considered as a replacement for in vivo studies with molecular and cellular study parameters and when end-points do not include life-span or behavioural assessment. Certainly, in academic and industrial laboratories, slice cultures should provide an excellent system for medium throughput drug screening or range-finding studies.\n\n\nDiscussion\n\nDepending on the nature of the experiment, one postnatal day 8 or 9 pup can provide an n=36 for immunohistochemical analysis, and a single well containing three slices can be combined to give n=12 for biochemical analysis or compound screening. For example, the preparation of slice cultures from only six postnatal pups would allow the opportunity to study 12 time-points in six different animals, a reduction in numbers of 91% in comparison to the 72 mice that would be required for an in vivo aging study. In addition, since multiple time points will be assessed in tissues from the same animals, experimental within-group variation is substantially reduced.\n\nTake-up of this method within academic laboratories in the UK appears to be growing, however it is very difficult to accurately quantify the number of animals that have not been used as a result of researchers preparing slice cultures in preference to in vivo experimentation. Within our own laboratories, we estimate that our in vivo experimentation has reduced by approximately 20% as we train more researchers in the method of brain slice culture preparation.\n\nWhile we have focussed on AD research in this article, organotypic brain slice cultures are equally suitable for research into a range of other neurodegenerative and neurological conditions, as well as for basic neuroscientific research (reviewed by 8). Slice cultures can also be prepared specifically from the hippocampus9 or from other tissues such as spinal cord38; the latter being used to investigate prion-like properties of mutant SOD1 proteins in amyotrophic lateral sclerosis. The technique is not limited to mice, rats are commonly used8 and methods are emerging to allow long-term culture of human organotypic brain slice cultures39. There are no major restrictions on uptake of this model since it requires only modest investment in terms of equipment provision. Training in tissue dissection and slicing may be beneficial, but the technique can readily be learned with practice.\n\n\nConclusions\n\nHere, we describe a detailed method for the preparation of long-term organotypic brain slice cultures from postnatal mice. We describe our work previously showing that slice cultures prepared from 3xTg-AD mice recapitulate important molecular and cellular features of in vivo disease development and the human disease phenotype. We also summarise the versatility of the model for drug discovery and the acute screening of compounds. Slice cultures show a significant acceleration in the timescale in which disease features develop, with relevant pathological changes observed at 28 days in vitro as opposed to 12 months in vivo in 3x-TgAD mice. We suggest that organotypic brain slice cultures can be used to replace several in vivo studies and that their widespread uptake could reduce the number of animals used in neurodegenerative disease research by 20–50%. This could be achieved if slice cultures were used in place of purely biochemical and immunohistological studies, and for experiments not reliant on behavioural outcomes.\n\n\nData availability\n\nDataset 1: Primary data for molecular changes in the slice culture model in comparison to findings made using tissue from aged 3xTg-AD mice. DOI: 10.5256/f1000research.14500.d20083220",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by funding from the National Centre for Replacement, Refinement and Reduction of Animals in Research (NC/K500343/1 to WN) and the Biotechnology and Biological Sciences Research Committee (BB/L502601/1 to WN).\n\n\nSupplementary material\n\nSupplementary File 1: ARRIVE checklist.\n\nClick here to access the data.\n\n\nReferences\n\nEllenbroek B, Youn J: Rodent models in neuroscience research: is it a rat race? Dis Model Mech. 2016; 9(10): 1079–1087. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArber C, Lovejoy C, Wray S: Stem cell models of Alzheimer's disease: progress and challenges. Alzheimers Res Ther. 2017; 9(1): 42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGähwiler BH, Capogna M, Debanne D, et al.: Organotypic slice cultures: a technique has come of age. Trends Neurosci. 1997; 20(10): 471–477. PubMed Abstract | Publisher Full Text\n\nHailer NP, Jarhult JD, Nitsch R: Resting microglial cells in vitro: analysis of morphology and adhesion molecule expression in organotypic hippocampal slice cultures. Glia. 1996; 18(4): 319–331. PubMed Abstract | Publisher Full Text\n\nStoppini L, Buchs PA, Muller D: A simple method for organotypic cultures of nervous tissue. J Neurosci Methods. 1991; 37(2): 173–182. PubMed Abstract | Publisher Full Text\n\nDe Simoni A, Griesinger CB, Edwards FA: Development of rat CA1 neurones in acute versus organotypic slices: role of experience in synaptic morphology and activity. J Physiol. 2003; 550(Pt 1): 135–147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Simoni A, Yu LM: Preparation of organotypic hippocampal slice cultures: interface method. Nat Protoc. 2006; 1(3): 1439–1445. PubMed Abstract | Publisher Full Text\n\nHumpel C: Organotypic vibrosections from whole brain adult Alzheimer mice (overexpressing amyloid-precursor-protein with the Swedish-Dutch-Iowa mutations) as a model to study clearance of beta-amyloid plaques. Front Aging Neurosci. 2015; 7: 47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarwell CS, Coleman MP: Synaptophysin depletion and intraneuronal Aβ in organotypic hippocampal slice cultures from huAPP transgenic mice. Mol Neurodegener. 2016; 11(1): 44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuff K, Noble W, Gaynor K, et al.: Organotypic slice cultures from transgenic mice as disease model systems. J Mol Neurosci. 2002; 19(3): 317–320. PubMed Abstract | Publisher Full Text\n\nMessing L, Decker JM, Joseph M, et al.: Cascade of tau toxicity in inducible hippocampal brain slices and prevention by aggregation inhibitors. Neurobiol Aging. 2013; 34(5): 1343–1354. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOddo S, Caccamo A, Shepherd JD, et al.: Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron. 2003; 39(3): 409–421. PubMed Abstract | Publisher Full Text\n\nCroft CL, Wade MA, Kurbatskaya K, et al.: Membrane association and release of wild-type and pathological tau from organotypic brain slice cultures. Cell Death Dis. 2017; 8(3): e2671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSundstrom L, Morrison B 3rd, Bradley M, et al.: Organotypic cultures as tools for functional screening in the CNS. Drug Discov Today. 2005; 10(14): 993–1000. PubMed Abstract | Publisher Full Text\n\nCroft CL, Kurbatskaya K, Hanger DP, et al.: Inhibition of glycogen synthase kinase-3 by BTA-EG4 reduces tau abnormalities in an organotypic brain slice culture model of Alzheimer's disease. Sci Rep. 2017; 7(1): 7434. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo T, Noble W, Hanger DP: Roles of tau protein in health and disease. Acta Neuropathol. 2017; 133(5): 665–704. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrundke-Iqbal I, Iqbal K, Quinlan M, et al.: Microtubule-associated protein tau. A component of Alzheimer paired helical filaments. J Biol Chem. 1986; 261(13): 6084–6089. PubMed Abstract\n\nPerez-Nievas BG, Stein TD, Tai HC, et al.: Dissecting phenotypic traits linked to human resilience to Alzheimer's pathology. Brain. 2013; 136(Pt 8): 2510–2526. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPooler AM, Usardi A, Evans CJ, et al.: Dynamic association of tau with neuronal membranes is regulated by phosphorylation. Neurobiol Aging. 2012; 33(2): 431.e427–438. PubMed Abstract | Publisher Full Text\n\nCroft CL, Noble W: Dataset 1 in: Preparation of organotypic brain slice cultures for the study of Alzheimer’s disease. F1000Research. 2018. Data Source\n\nFontaine SN, Zheng D, Sabbagh JJ, et al.: DnaJ/Hsc70 chaperone complexes control the extracellular release of neurodegenerative-associated proteins. EMBO J. 2016; 35(14): 1537–1549. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHellwig S, Masuch A, Nestel S, et al.: Forebrain microglia from wild-type but not adult 5xFAD mice prevent amyloid-β plaque formation in organotypic hippocampal slice cultures. Sci Rep. 2015; 5: 14624. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNoble W, Pooler AM, Hanger DP: Advances in tau-based drug discovery. Expert Opin Drug Discov. 2011; 6(8): 797–810. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaccamo A, Oddo S, Tran LX, et al.: Lithium reduces tau phosphorylation but not A beta or working memory deficits in a transgenic model with both plaques and tangles. Am J Pathol. 2007; 170(5): 1669–1675. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlenner GG, Wong CW: Alzheimer's disease and Down's syndrome: sharing of a unique cerebrovascular amyloid fibril protein. Biochem Biophys Res Commun. 1984; 122(3): 1131–1135. PubMed Abstract | Publisher Full Text\n\nO'Brien RJ, Wong PC: Amyloid precursor protein processing and Alzheimer's disease. Annu Rev Neurosci. 2011; 34: 185–204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMasliah E, Mallory M, Alford M, et al.: Altered expression of synaptic proteins occurs early during progression of Alzheimer's disease. Neurology. 2001; 56(1): 127–129. PubMed Abstract | Publisher Full Text\n\nKurbatskaya K, Phillips EC, Croft CL, et al.: Upregulation of calpain activity precedes tau phosphorylation and loss of synaptic proteins in Alzheimer's disease brain. Acta Neuropathol Commun. 2016; 4: 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTai HC, Serrano-Pozo A, Hashimoto T, et al.: The synaptic accumulation of hyperphosphorylated tau oligomers in Alzheimer disease is associated with dysfunction of the ubiquitin-proteasome system. Am J Pathol. 2012; 181(4): 1426–1435. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGylys KH, Fein JA, Yang F, et al.: Synaptic changes in Alzheimer's disease: increased amyloid-beta and gliosis in surviving terminals is accompanied by decreased PSD-95 fluorescence. Am J Pathol. 2004; 165(5): 1809–1817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOddo S, Caccamo A, Kitazawa M, et al.: Amyloid deposition precedes tangle formation in a triple transgenic model of Alzheimer's disease. Neurobiol Aging. 2003; 24(8): 1063–1070. PubMed Abstract | Publisher Full Text\n\nRevilla S, Ursulet S, Álvarez-López MJ, et al.: Lenti-GDNF gene therapy protects against Alzheimer's disease-like neuropathology in 3xTg-AD mice and MC65 cells. CNS Neurosci Ther. 2014; 20(11): 961–972. PubMed Abstract | Publisher Full Text\n\nMatsuoka Y, Gray AJ, Hirata-Fukae C, et al.: Intranasal NAP administration reduces accumulation of amyloid peptide and tau hyperphosphorylation in a transgenic mouse model of Alzheimer's disease at early pathological stage. J Mol Neurosci. 2007; 31(2): 165–170. PubMed Abstract | Publisher Full Text\n\nMatsuoka Y, Jouroukhin Y, Gray AJ, et al.: A neuronal microtubule-interacting agent, NAPVSIPQ, reduces tau pathology and enhances cognitive function in a mouse model of Alzheimer's disease. J Pharmacol Exp Ther. 2008; 325(1): 146–153. PubMed Abstract | Publisher Full Text\n\nInbar P, Li CQ, Takayama SA, et al.: Oligo(ethylene glycol) derivatives of thioflavin T as inhibitors of protein-amyloid interactions. Chembiochem. 2006; 7(10): 1563–1566. PubMed Abstract | Publisher Full Text\n\nMegill A, Lee T, DiBattista AM, et al.: A tetra(ethylene glycol) derivative of benzothiazole aniline enhances Ras-mediated spinogenesis. J Neurosci. 2013; 33(22): 9306–9318. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHabib LK, Lee MT, Yang J: Inhibitors of catalase-amyloid interactions protect cells from beta-amyloid-induced oxidative stress and toxicity. J Biol Chem. 2010; 285(50): 38933–38943. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAyers JI, Diamond J, Sari A, et al.: Distinct conformers of transmissible misfolded SOD1 distinguish human SOD1-FALS from other forms of familial and sporadic ALS. Acta Neuropathol. 2016; 132(6): 827–840. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwarz N, Hedrich UBS, Schwarz H, et al.: Human Cerebrospinal fluid promotes long-term neuronal viability and network function in human neocortical organotypic brain slice cultures. Sci Rep. 2017; 7(1): 12249. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "34021",
"date": "18 May 2018",
"name": "Claire S. Durrant",
"expertise": [
"Reviewer Expertise Organotypic hippocampal slice cultures",
"Alzheimer's Disease"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis well written method article describes the process of making whole brain organotypic slices from 3xTg-AD mice to use for studies of Alzheimer’s related pathology. The main focus of the paper is on how this method can be used to reduce and refine the use of animals in Alzheimer’s disease (AD) research. The methods of creating and maintaining the slice cultures are clearly described in detail allowing for future replication by other groups. Specific protocols for preparing slice material for immunostaining, immunoblotting and ELISA are also described effectively. The experimental results given from previously published work, as well as the cited work of others, serve as excellent examples of how this system can be utilised in AD studies with particular emphasis on the 3Rs. In particular, the authors highlight the ability to test multiple timepoints or multiple compounds on tissue from the same animal resulting in not only a reduction in the number of animals required, but also better controls for biological variability. I agree with the authors’ conclusions that brain slice models of Alzheimer’s disease can partly replace biochemical/cellular/molecular studies in older animals that have developed pathology.\nThe authors provide a useful comparison between changes observed in human brain, aged 3xTgAD brain and brain slice cultures, with the raw data provided. Their observations that slice cultures show an accelerated phenotype when compared to adult mice is interesting and in agreement with what we have seen in hippocampal slices cultures from TgCRND8 mice (reference 9 in this paper). Future work seeking to elucidate why this is the case could be very informative.\nWhilst the methods in this paper have been previously described elsewhere, the detailed description of culture preparation, experimental design and methods for analysis in the context of AD are a useful and timely addition to the literature. The tips relating to assessment of culture health (a white slice becoming translucent) and issues to look out for that may harm the health of the slices (excess medium on the membrane insert) are completely aligned with our own experience and very useful for someone seeking to perform slice culture experiments for the first time. I have no reservations in approving this manuscript.\nMinor comments:\nWhilst actin is a valid control protein in western blots, it could be argued that beta-III tubulin (or alternative neuronal markers) may be a more informative control when normalising synaptic protein levels. This controls for any difference in neuronal number between samples which may alter synaptic protein levels in the absence of a synaptic specific change. This would be more important in cases where synaptic protein levels are down, but there is also evidence for neuronal loss.\n\nIt would be interesting to see whether the 3xTg-AD brain slices lose synaptic proteins if maintained for longer than 28 days in vitro. It may be that this phenotype appears after the observed changes in tau and Aβ in this culture system.\n\nIt should be stated whether the 3xTg-AD mice used in this study are homozygous or heterozygous (I am presuming homozygous but this is not explicitly stated). It appears that the wild-type mice used as controls for the 3xTg-AD are background-matched, but not littermates. This is standard practice for many homozygous mouse lines but care should be taken to ensure backcrossing of the two lines to prevent genetic drift between colonies that have remained separated for a long time. Ideally, littermates would be used, but I understand that the homozygosity of this mouse model would complicate this and result in a greater number of animals being required to generate all genotypes. If separate colonies for transgenic and wild type pups are necessary, litters should ideally be synchronised to control for minor differences in culture preparation/ harvesting on different days.\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "34019",
"date": "22 May 2018",
"name": "Christian Humpel",
"expertise": [
"Reviewer Expertise Alzheimer",
"brain slices",
"transgenic",
"diagnosis",
"platelets",
"monocytes",
"plaques",
"tau"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 done contribution by a group having good experience in organotypic brain slice cultures. The MS is well written, the methods clear and the pictures of good quality. I have a few suggestions:\nPlease provide more details in the Abstract, especially, on the organotypic brain slice model (3xTg mice, chopper slices, 350 µm, postnatal d8-9, cultured for > 2weeks on 0.4 µm pore semipermeable membranes). Please also add this information in the Box research highlights. This is important to see on one view which kind of slices you mean.\n\nI suggest that you are more careful about the used number of animals. In Abstract you write an estimated 20,000 mice per year for dementia research and on page 3, right col. sec. paragr. you write ... were used in AD research, and you write 20,000 transgenic; assuming from 30 mice per paper. What do you mean: (1) per year in UK, Europe or the world?, (2) total mice per year or only transgenic per year? (3) based on which fact you assume 30 mice per paper (value 30 is just your thoughts or is this based on an official reference?). I suggest that you check an EU report on the total number of animals/mice/transgenic, or you check your national institution (and also give the official reference). Such rough values could be very problematic.\n\nPage 3 third paragr: .... importantly, slice cultures are prepared from neonatal mice ... this (neonatal) should be explained as it is unclear.\n\nIt would be helpful to give the average weight of the postnatal d8-9 mice. We usually weigh them, to be sure about the stage. Or do you have other rules to be sure on the correct stage. Also mention up to which age does your model work?\n\nPage 5, point 6: typo: 2x remove\n\nPage 7 right col, last paragr: ....when reporting on the phosphorylation of tau via LiCl and GSK-3, the authors may cite the recent paper Foidl and Humpel (2018) in Frontiers Aging Neurosci. on hyperphosphorylation of tau in organotypic brain slices induced by okadaic acid. In this respect also the review on brain culture slices by Humpel (2015) in Neuroscience should be cited (page 4, left col, line, 14).\n\nPlease explain better your definition of the n-number (page 8, right col.) .... one postnatal d8 or 9 pup can provide an n=36. I think that one mouse only gives an n=1; for an experiment you need at least 6-8 different mice (n=6-8); but one mouse can give up to 36 different treatments (or in duplicate 2x18). What is your suggestion on power calculations (how many slices per group?). Please explain better any statistical issues.\n\nFinally, please state in your conclusion the limits of the brain slices; how to study genetic (familiar) versus sporadic AD, how comparable is the developmental stage with the adult stage (maturation of neurons after 2 weeks in culture?), how long do you need to culture to get at least an early comparable adult stage, what about axonal networks and the problem of axotomy.\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "34402",
"date": "29 May 2018",
"name": "Susan C. Barnett",
"expertise": [
"Reviewer Expertise Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact. Additional expertise: glia cell biologist",
"CNS co-cultures to study myelination",
"spinal cord injury"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 clearly explained method for making slice cultures from 3xTg-AD mice for the study of Alzheimer’s disease. It is useful to have a step by step account with hints to describe this method as in general slice cultures can be difficult to reproduce and be consistent. There is a good description of how this method fulfils NC3Rs criteria.\nMinor points:\nIt would be very helpful to have a little more justification of why this is a novel method when many publications can be seen using slice cultures for Alzheimers research, even from transgenic mice.\n\nIt was a little bit confusing on the section stating: In brief, slice cultures are cut whilst still on their membranes and then treated as free-floating sections. Is this illustrated in Figure 1? If so could be made clearer?\n\nNot convinced the n=12 from one pup statement is correct, as this would be replicates. N usually refers to different biological repeats.\n\nIt would be really useful to see example of the slices when validated, with examples of staining. Can microglia be seen and how good is the anatomy of the slices.\n\nThe LDH was not explained enough and perhaps the time course could have been shown.\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-592
|
https://f1000research.com/articles/7-920/v1
|
27 Jun 18
|
{
"type": "Research Article",
"title": "Who and why do researchers opt to publish in post-publication peer review platforms? - findings from a review and survey of F1000 Research",
"authors": [
"Jamie Kirkham",
"David Moher",
"David Moher"
],
"abstract": "Background: Preprint servers and alternative publication platforms enable authors to accelerate the dissemination of their research. In recent years there has been an exponential increase in the use of such servers and platforms in the biomedical sciences, although little is known about who, why and what experiences researchers have with publishing on such platforms. In this article we explore one of these alternative publication platforms, F1000 Research, which offers immediate publication followed by post-publication peer review. Methods: From an unselected cohort of articles published between 13th July 2012 and 30th November 2017 in F1000 Research, we provided a summary of who and what was published on this platform and calculated the percentage of published articles that had been indexed on a bibliographic database (PubMed) following successful post-publication peer review. We also surveyed corresponding authors to further understand the rationale and experiences of those that have published using this platform.\n\nResults: A total of 1865 articles had been published in the study cohort period, of which 80% (n=1488) had successfully undergone peer review and were indexed on PubMed within a minimum period of six months since first publication. Nearly three-quarters of articles passed the peer review process with their initial submission. Survey responses were received from 296 corresponding authors. Open access, open peer review and the speed of publication were the three main reasons why authors opted to publish with F1000 Research.\n\nConclusions: Many who published with F1000 Research had a positive experience and indicated that they would publish again with this same platform in the future. Nevertheless, there remained some concerns about the peer review process and the quality of the articles that were published.",
"keywords": [
"F1000 Research",
"Journalology",
"Peer Review",
"Rapid Publication"
],
"content": "Introduction\n\nThe conventional method of journal publication involves manuscript submission, peer review and editorial oversight, revision and publication. While this process is presumed to ensure the scientific integrity of the research undertaken, the availability of the research findings entering the public domain may take several months or even years depending on factors such as a journal editors decision to publish or reject an article, peer reviewer availability or a journal’s publication frequency. The efficiency of peer review was underlined in a recent peer review survey conducted in 2018 by ASAPbio (Accelerating Science and Publication in biology). The survey revealed that for their most recent published article, about 50% of the authors surveyed (132/259) submitted their article to two or more journals, with 7% (18/259) submitting to five or more [http://asapbio.org/peer-review/survey]. This process can result in a substantial delay in research findings entering the public domain. Traditionally, authors have not been able to add such research findings to their curriculum vitae and/or grant applications. In some scientific fields such as pandemics or humanitarian emergencies, the time to deliver research findings may be as equally as important as research quality, and may be critical to health care provision.\n\nWhile some journals may offer a ‘fast track’ service to publication, preprint servers offer rapid publication on all articles but without systematic refereeing, which brings significant benefits to the authors, the presentation of the article and the readers. The profile of preprint servers is increasing, with many major funders (e.g. Wellcome Trust and National Institute of Health (NIH)), now endorsing the use of preprint servers, particularly for grant applications [https://wellcome.ac.uk/news/we-now-accept-preprints-grant-applications, http://www.sciencemag.org/news/2017/03/nih-enables-investigators-include-draft-preprints-grant-proposals].\n\nHowever, due to lack of peer review, preprint servers in a life sciences setting have been criticised as they may lack quality and subsequently have the potential to report flawed research which may harm patients1. To improve scientific integrity, new emerging options towards publication are being considered under the Open Science Initiative. An example includes peer review before the results are known ‘registered reports’, which aims to eliminate questionable research practices and poor research design [https://cos.io/rr/]. The registered reported publishing format is currently being used by over 100 journals.\n\nF1000 Research is an example of an alternative ‘platform’ which offers the advantages of a preprint server in terms of immediate publishing on a variety of research article types linked to biomedical research, with the added advantage of post-publication open peer review. Once peer review is complete (at least two approved referee reviews, or one approved plus two approved with reservations reviews) the article is subsequently indexed in a bibliographic database such as PubMed.\n\nAs of May 2018, F1000 Research has published over 2000 articles since its inception in 2012. Little is known about who, and the reasons why authors publish in F1000 Research. The aim of this study is to provide a descriptive summary of the research that has been published in F1000 Research, and to determine how much of this published research has been accepted for bibliographic database indexing. We also survey authors who have published in F1000 Research to further understand the rationale and experiences of those that have published using this publication platform.\n\n\nMethods\n\nWe studied a cohort of all article types that were first published on F1000 Research between 13th July 2012 (earliest publication) and 30th November 2017. A data extraction form was developed and piloted on the first page of 20 listed publications. For each article, the following information was extracted; article type, the year of publication, funding sources, the country of the first listed corresponding author and the peer review status. The peer review status for all articles was last verified on May 30th 2018, i.e. six months after the last published article in the study cohort. At the same time, we also checked whether articles were indexed on the bibliographic database, PubMed.\n\nWith the exception of Editorials and F1000 Faculty Critiques which are published by invitation only and not subject to external peer review, the first listed corresponding author of all published studies in the study cohort were contacted via a personalised email. We removed any duplicate email addresses such that a corresponding author who had published multiple articles were contacted only once. Participants were asked to participate in a short online survey with regards to their main reasons and experiences of publishing with F1000 Research. The survey was constructed using the online survey software, REDcap [https://www.project-redcap.org/], and was open for responses between 6th April 2018 and 10th May 2018. The survey questions are available in Supplementary File 1 and reflect the importance of a series of factors that may influence the decision to submit to F1000 Research as rated on a five-point Likert scale ranging from ‘very important’ to ‘not important’. Similarly, on a 5-point Likert scale we asked about the importance of articles being indexed on a bibliographic database, the importance of a transparent peer review process and the likelihood that they would submit future manuscripts to F1000 Research or recommend the platform to others. Participants could also provide free text comments on positive or negative experiences associated with submitting or publishing with the platform. Non-responders were contacted periodically if a response to the survey was not received. The data were presented as the frequency distribution for each level of response. All positive and negative experiences were independently reviewed by both authors and categorized into common topics. Any discrepancies were solved via discussion.\n\nThe University of Liverpool Ethics Committee was consulted and granted ethical approval for this study (Reference 3233). Informed consent was assumed if a participant responded to the survey.\n\n\nResults\n\nA total of 1865 articles were published in F1000 Research between the period 13th July 2012 and 30th November 2017. Just over a third of articles published were research articles (677/1865; 36%) with no more than 10% of the remaining articles published representing a different article type (Table 1). The majority of articles published received non-commercial funding (1054/1865; 57%), while a large proportion also declared no funding source (745/1865; 40%). The first corresponding author in nearly 80% of articles published were from high income countries (1480/1865; 79%) and less than 2% were from low income countries (Table 1, Figure 1). The six countries with more than 50 articles published were USA (618 articles), UK (232 articles), Germany (91 articles), Australia (84 articles), India (82 articles) and Canada (78 articles). There appeared to be a gradual increase in the number of articles published over time with over 400 articles published in each of the last two years in the study cohort (Table 1).\n\naStudies that were partially funded by industry (e.g. pharmaceutical) were classified as ‘commercial funding’\n\nbEconomic status was classified according to the World Bank list of economies (June 2017)\n\nAllowing for a minimum of six months from the first publication of all articles, 80% (1488/1865) had successfully undergone peer review (with the exception of 51 editorials that were not subject to peer review) and were indexed in the bibliographic database, PubMed (Table 2). For the remaining 20% of articles, the lack of indexing was because peer review was incomplete (n=317), peer review had discontinued (n=36) or the article had been removed by authors (n=3). When peer review was incomplete, the peer review process had been ongoing for over 12 months for 80% of articles (253/317). In a small number of articles (n=14) the peer review process was complete but the article had not yet appeared in PubMed (Table 2). Of the articles that were published in F1000 Research, 74% (1065/1448) passed the peer review process with the initial submission, 29% (n=336) after one revision, 3% (n=42) after two revisions, while five articles required four or five revisions.\n\naTwo of these articles were not published by F1000 Research (South Asian Journal of Cancer (case report), La Tunisie Medicale (research article)).\n\nbOne article was indexed on a bibliographic database but the peer review process was incomplete.\n\nc253: peer review ongoing for over 12 months since the article was first submitted.\n\ndFour: peer review completed but article not indexed on a bibliographic database within 12 months of last publication date.\n\nAfter excluding 58 articles that did not undergo peer review (editorials and or F1000 Faculty Critiques), there were 1476 unique first listed corresponding author email addresses (out of 1807 articles) that were targeted in the survey. Notably two authors were listed as the same corresponding author on 18 (Germany) and 16 articles, respectively (USA/India).\n\nResponses to the survey were received from 296 corresponding authors. An exact response rate was difficult to estimate but we approximate this to be between 25–30% given the number of returned survey emails that had invalid and/or expired email addresses or ‘out of offices’ during the period the survey was live. The majority of responders were academic affiliated (74%; 219/296), while a minor proportion represented non-profit organisations (9%; 22/296), industry (5%; 14/296) and government (7%; 22/296). The remaining 14 represented other entities such as independent self-employed private researchers, consulting agencies, schools and hospitals. There was an increasing trend in terms of the respondents research experience with 7% (21), 18% (52), 33% (98) and 42% (125) out of the 296 respondents representing trainee, early-, mid- and senior researchers, respectively.\n\nThe importance of factors that influenced an author’s decision to submit an article to F1000 Research are presented in Figure 2. The open access policy, open peer review policy and the speed of publication were the three top reasons for publishing with F1000 Research, with more than 70% of participants reporting these factors as either important or very important. Linked to peer review policy, the transparency of the F1000 Research peer review system that includes reviewer’s names was rated as important or very important by nearly 70% of respondents (202/295). Of less importance was the recommendation to publish in F1000 Research by colleagues, had previously peer reviewed for F1000 Research and for promotion and tenure. Nevertheless, 80% (237/295) of respondents said they would either likely or very likely recommend F1000 Research to a colleague.\n\nThe respondents listed 58 additional new items relating to three themes that influenced their decision to publish with F1000 Research. One theme was linked to promotional activities connected directly with F1000 Research which included fee waivers (n=14), personal invitations to submit an article (n=2) and commissioned calls for specific articles (n=3). In another theme, the reasons were unconnected to F1000 Research but were a result of failures to publish in alternative journals (n=16), reasons cited included; other journals not interested in the type of article/analysis, bad reviews, biased peer review and editorial biases. In a third theme, authors chose to submit to F1000 Research due to specific characteristics of the publishing platform. Characteristics included, accessibility of previous versions of the article and the ability to access and respond to reviewer comments (n=5), ability to be able to publish negative findings and material on controversial topics (n=3), no size limit on articles (n=2), ability to share public datasets (n=1), quality of publication of images (n=1) and good altmetrics (n=1). The remaining ten items related to the fact that authors were intrigued to test out a new publication platform.\n\nNearly 90% of respondents (261/295) stated that it was either important or very important that following immediate publication with F1000 Research, their article was later indexed on a bibliographic database after the article was approved following peer review. The 261 corresponding authors who thought this was important listed 198 reasons for this. The most common reason was related to making the article visible and easily accessible (n=81), and ensuring that the article was sufficiently exposed (n=27) to its intended readership. Others stated that indexing enhanced the credibility and quality (n=32) of articles as it was the benchmark that the article had undergone peer review. Article indexing for assessment purposes (n=37) was also seen to be important in terms of the assessment of research impact, assessing scientists (e.g. for promotion), assessing institutes (e.g. the Research Excellence Framework in the UK) and when applying for grant income. Finally, some respondents thought that the indexing of articles was important for personal distinctions (n=21), examples included personal recognition amongst peers, citation counts and the enhancement of personal portfolios.\n\nThe main criticism of those submitting to F1000 Research was related to the peer review process following publication. Authors found the process of nominating and reviewers agreeing to review challenging (n=9) given the strict criteria for selection [https://f1000research.com/for-authors/tips-for-finding-referees], this it was felt led to typically a longer period of peer review (n=12) than most other journals. Some authors also questioned the quality of the reviews (n=9), with the suggestion that author selected reviewers may be biased or reserved in terms of lack of criticism given the public record and naming of reviewers when a platform operates an open peer review policy. A few authors felt ‘trapped’ in the peer review process and felt that once the article was published there was ‘no way out’ if reviewers could not be found or reviewers had stopped providing reviews. Publication fees was seen as major barrier for this publication platform, particularly in some areas of research where funds available for publishing is limited (n=10). The impact of articles published with F1000 Research was also seen as a limitation (n=16), and while this was not necessarily the authors personal concern, the perceived reputation that this would be considered a low-quality publication and poorly cited on a platform with no reputable impact factor were foreseen as issues with peers and within scientific organisations. Three authors provided condemning reviews of the platform and suggested that it ‘provided an easy opportunity to publicly criticise the work of others in an act that constitutes unwarranted bullying’ and were subsequently forced into using the platform to correct and refute the criticisms to protect personal reputation. A number of authors (n=5) also commented that the platform was difficult for editing and writing purposes and was particularly tedious when making a data deposition (n=2).\n\nDespite some negative feedback regarding F1000 Research, there were also many positive responses with a number stating that this was ‘their best experience in publishing’ with the hope that this publication style becomes ‘dominant in the future’. Based on their experience, 74% (218/296) of the respondents said that they would likely or very likely submit to F1000 Research in the future. The speed and efficiency of publication (n=11) was the main reason that authors felt the experience was positive, while others (n=8) thought the extent and the transparency of reviews was both helpful and important. Some authors also found the editorial staff to be cooperative and professional (n=7) while other benefits included the ease of use of the platform and the standard of the publication.\n\n\nDiscussion\n\nThe explosion in the number or preprint platforms and the number of researchers submitting to preprint servers and alternative platforms in the biomedical sciences is rising exponentially [http://www.prepubmed.org/monthly_stats/], in yet relatively little is known about them. F1000 Research offers a unique publishing platform, which like preprint servers offer immediate publishing but with the added advantage of post-publication peer review and eventual article indexing on a bibliometric database. The speed of the publication, alongside the open access and open peer review policy were particularly attractive traits to authors who submitted their research to this platform. Having an article indexed on a bibliometric database was seen to be important by the majority of the respondents and this study revealed that 80% of the articles achieved indexing within six months of submission to F1000 Research. Visibility and accessibility of research articles were deemed to be the most common reason for submission. The visibility of publishing research without peer review plays another important role. The peer review for about 15% of the articles had either been ongoing for more than a year or discontinued completely meaning that many of these articles may have contributed to the vast quantity of inaccessible unpublished literature (and the potential for publication bias) had these articles been subjected to the standard peer review before publication model.\n\nThe F1000 Research publication model was not without criticisms. Some found that the peer review process took longer than standard journals because there was more emphasis on the authors rather than editors to find peer reviewers. There was also a sense that there was the potential for an article to become caught up in the process, immediate publication meant that there was limited scope to remove or submit elsewhere if peer reviewers could not be found or existing reviewers failed to provide subsequent reviews. While these criticisms may have reflected the experiences of some survey respondents, this process is not dissimilar from the many journals/publishers of standard journals which request names of peer reviewers, and in some instances release articles if peer reviewers cannot be found in a reasonable timeframe.\n\nThe majority of respondents generally saw open peer review as a good thing, but some respondents felt this process could lead to inferior and reserved poorer quality reviews that lacked criticism, this was perhaps evidenced by the fact that 75% of articles passed the peer review stage based on the first submitted version. Despite this finding, a randomised trial has found that asking a reviewer to consent to be identified to the author had no important effect on the quality of the review but it may significantly increase the likelihood of reviewers declining to review2.\n\nThe strength of this study is that we evaluated a large, unselected cohort of articles that were published with F1000 Research. The response to the survey was quite poor with only 296 corresponding authors engaging with the survey from the potential 1476 unique email addresses identified. Nevertheless, calculating an exact response rate was particularly challenging given several hundred of those contacts were found to be invalid or expired, a consequence of targeting some authors that published their articles several years ago. Even with the potential for such response bias, the open text comments received appeared to be relatively balanced in terms positive and negative experiences of publishing with F1000 Research with key themes identified. There was also a general sense that the F1000 Research platform appeared to be ‘modern’ and yet was potentially ‘less attractive to the early career researcher’ because the need to publish in recognised journals with high impact factors is still considered the standard by the vast majority of researchers and institutes to gain promotion and tenure.\n\nIt was clear that researchers from all around the world have published on the F1000 Research platform. The importance of alternative publications platforms is beginning to extend beyond an authors choice to submit to them. For example, recently Public Library of Science (PLOS) have partnered with the preprint server bioRxiv, and from May 1st 2018, authors will have the option to post their submitted manuscript on to the preprint server in order to disseminate their work prior to peer review [http://blogs.plos.org/plos/2018/04/one-small-step-for-preprints-one-giant-step-forward-for-open-scientific-communications/]. To a large extent this mimics the idea of the already existent F1000 Research publication model.\n\nIn conclusion, there is undoubtedly and increase in the use of researchers publishing their research on alternative platforms for biomedical sciences, but there still remains a level of dogma surrounding their use by many, and there remains concerns about the quality of the articles published on these platforms.\n\n\nData availability\n\nDataset 1: Articles published in F1000 Research between the period 13th July 2012 and 30th November 2017 10.5256/f1000research.15436.d2083083",
"appendix": "Competing interests\n\n\n\nDavid Moher is an advisory board member for F1000 Research. Jamie Kirkham declares no conflicts of interest.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary Appendix A: Survey questions sent to corresponding authors of articles published in F1000 Research between 13th July 2012 and 30th November 2017\n\nClick here to access the data.\n\n\nReferences\n\nChalmers I, Glasziou P: Should there be greater use of preprint servers for publishing reports of biomedical science? [version 1; referees: not peer reviewed]. F1000Res. 2016; 5: 272. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Rooyen S, Godlee F, Evans S, et al.: Effect of open peer review on quality of reviews and on reviewers’ recommendations: a randomised trial. BMJ. 1999; 318(7175): 23–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKirkham J, Moher D: Dataset 1 in: Who and why do researchers opt to publish in post-publication peer review platforms? - findings from a review and survey of F1000 Research. F1000Research. 2018. Data Source"
}
|
[
{
"id": "35531",
"date": "04 Jul 2018",
"name": "Larry Peiperl",
"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\nF1000 research provides a platform for immediate publishing of research followed by open peer review and, if approved by reviewers, indexing in PubMed or other databases. In the current article, the authors, who are leaders in the fields of metaresearch and journalology, provide a descriptive summary of the 1865 articles that were published in F1000 Research from earliest publication in July 2012 through November 2017, verifying peer review status through May 2018. They also conduct an online survey in April-May 2018 through email requests to corresponding authors of articles that had been subject to external peer review.\nThe results are descriptive and presented in summary format, which seems appropriate to the study design. No statistical inference is undertaken.\n\nA spreadsheet of study-specified characteristics for all 1865 F1000 articles, and survey questions (not responses, which may possibly reflect an effort to preserve confidentiality of respondents) are provided as supplementary files.\n\nThe bibliometric analysis appears straightforward. The authors may wish to clarify the following points, most of which pertain to the survey and interpretation of its (partially qualitative) results.\n\nMethods\n\n1. “All positive and negative experiences were independently reviewed by both authors and categorized into common topics. Any discrepancies were solved via discussion.” Can you provide more detail on how this qualitative analysis was performed? For example, were the “common topics” pre-defined? If not, how did you arrive at the 3 thematic categories that you present in the Results? Especially in light of the appropriately declared author competing interest, and the apparent relevance of the conclusions to the platform on which the authors are publishing the article, further information about how the study methodology may have supported objective interpretation of the qualitative data could strengthen the presentation.\n\nResults\n2. Thirty-six articles had peer review “discontinued”. Under what circumstances does discontinuation of peer review occur?\n\n3. Of 1476 unique corresponding author email addresses targeted in the survey, you received 296 responses, representing some 20% of the corresponding authors. You identify the low response rate as a limitation of the study, but can you comment further on the representativeness of this sample? Are you able to draw any useful conclusions based on the data you obtained for the full set of articles (year of publication, country income level, etc) regarding the extent to which responders may have differed from non-responders? If such an analysis is not feasible, more caution would seem appropriate in drawing conclusions on the basis of the survey results.\n\nDiscussion\n\nThe authors might clarify their reasoning at several points:\n4. Article: “There was also a sense that there was the potential for an article to become caught up in the process, immediate publication meant that there was limited scope to remove or submit elsewhere if peer reviewers could not be found or existing reviewers failed to provide subsequent reviews. While these criticisms may have reflected the experiences of some survey respondents, this process is not dissimilar from the many journals/publishers of standard journals which request names of peer reviewers, and in some instances release articles if peer reviewers cannot be found in a reasonable timeframe.”\n\nComment: If an author believes that an article published prior to peer review in F1000 cannot then be submitted to another journal, that author’s situation would seem different from when a journal rejects an article after peer reviewers cannot be found. In the latter situation, the author has the option to pursue indexing by submitting the article to another journal. It’s not clear from this paragraph whether or not authors have this option for articles that haven’t passed peer review in F1000.\n\n5. While there are some similarities, I’m not sure it’s fair to say that the PLOS partnership with bioRxiv “largely mimics the idea of the existing F1000 Research publication model.”\n\nPLOS is partnering with bioRxiv in a way that other journals may choose to adapt, to complement the traditional journal publishing model by speeding dissemination while maintaining author choice regarding preprint posting and ultimate publication venue. Within this collaboration, bioRxiv does not function as an end-to-end publishing platform like F1000Research. Preprints posted to bioRxiv are not identified as having been submitted to a PLOS journal prior to publication in that journal, and authors are free to submit their posted work to another journal if the PLOS journal does not accept it.\n\n6. Article: “There still remains a level of dogma surrounding [preprint platform] use by many, and there remain concerns about the quality of the articles published on these platforms.”\nComment: It's not clear what you refer to as “dogma.\" The term seems dismissive, yet your research article identifies a number of reasonable concerns pertaining to preprint platform use.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "35535",
"date": "10 Jul 2018",
"name": "David Mellor",
"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 provide an appropriate and needed review of a leading platform in post publication peer review and preprints, F1000. They collect information on the country of origin, the \"success\" at being formally indexed, and survey authors on their experiences with the process (which is rarely shared about any other platform, so this information is particularly needed). My critiques below are all modest points for improvement.\n\n\"Traditionally, authors have not been able to add such research findings to their curriculum vitae and/or grant applications.\" The only harm implied by this lag is to the author. This roundabout system is also inefficient for reviewers, journals, and for dissemination to potentially interested readers.\n\n\"In some scientific fields such as pandemics or humanitarian emergencies, the time to deliver research findings may be as equally as important as research quality, and may be critical to health care provision.\" The authors should couch this as a balance between two conflicting needs: speed for potentially urgent information and ensuring that health-related information is of the highest possible quality.\n\n\"Open Science Initiative\" I am not aware of a formal \"initiative\". Recommend revise to: \"...new emerging options towards publication are being considered as part of the the \"Open Science\" movement.\"\nAn additional benefit of Registered Reports is to address publication bias towards significant or novel results.\n\nIt might be worthwhile to mention that F1000 itself offers the RR workflow. https://blog.f1000.com/2017/10/12/transparency-meets-transparency/\n\nIn the intro, the following citations would provide a bit more context for the reader\n\nSurvey on open peer review: Attitudes and experience amongst editors, authors and reviewers1 Altmetric Scores, Citations, and Publication of Studies Posted as Prep2 And for the section mentioning the rationale for Registered Reports, \"Instead of “playing the game” it is time to change the rules: Registered Reports at AIMS Neuroscience and beyond\"3\n\n\"Non-responders were contacted periodically if a response to the survey was not received.\" Please provide your rule for re-contact (e.g. once per week until response was received or four contacts were made).\n\n\"All positive and negative experiences were independently reviewed by both authors and categorized into common topics. Any discrepancies were solved via discussion.\" What does this mean? That each free response, positive/negative opinion was scored by this study's authors (since you're describing a survey or authors, please clarify who the \"authors\" are in this sentence).\n\n\"with no more than 10% of the remaining articles published representing a different article type\" Case report was slightly over 10% (this is obviously a nitpick, but it did lead me to wonder if I was looking at the correct column in the correct table).\n\n\"The majority of articles published received non-commercial funding\" Recommend revise to \"The majority of articles published reported receiving non-commercial funding...\"\n\"Economic status was classified according to the World Bank list of economies (June 2017)\" Please provide a specific link to this list.\n\"For the remaining 20% of articles, the lack of indexing was because peer review was incomplete (n=317), peer review had discontinued (n=36) or the article had been removed by authors (n=3).\" Does \"incomplete\" include both articles that received poor reviews and articles that have received 1 or fewer reviews? Clarify the number of articles that received something equivalent to \"rejections\". I am casually familiar with F1000, perhaps more so than typical but less so than a frequent reader. It is unclear to me if \"peer review ongoing\" could ever revert to another status. Were there any articles that received multiple reviews that indicated poor article quality? Perhaps provide a bit more explanation of the F1000 workflow in the introduction and also define these categories a bit more precisely.\n\n\"29% (n=336), after one revision,\" There appears to be a math error here, probably should be 23%, but please check.\nFigure 2, \"Recommendation\" I think should be revised to \"Recommendation by colleague\", as all the other titles were self-explanatory (to me), but that one I had to open up the survey to understand what it meant (thanks for providing the survey!).\n\"Authors found the process of nominating and reviewers agreeing to review challenging (n=9)\" This sentence seems like you are inferring an opinion by 9 people to the entire author pool. I recommend leading the section \"Experiences of those submitting articles to F1000 Research\" with opinions that can be reliably noted by larger groups from within your sample, and then mention these small N opinions with appropriate caveats (e.g. \"A few authors noted....\"). Giving the Ns means you are not misleading anyone obviously, it just reads a bit odd to present this as a common experience. Likewise with the last sentence in that paragraph, \"A number of authors [had various complains, n=5 and 2].\"\n\n\"The speed and efficiency of publication (n=11) was the main reason that authors felt the experience was positive\" change \"main\" to \"most often noted in the free responses\" (if that is what you mean by \"main\").\n\"The peer review for about 15% of the articles had either been ongoing for more than a year or discontinued completely meaning that many of these articles may have contributed to the vast quantity of inaccessible unpublished literature (and the potential for publication bias) had these articles been subjected to the standard peer review before publication model.\" This sentence implies that the rest of the articles would have been published elsewhere if not submitted to F1000. I think that is overly optimistic. I would describe that 15% as a floor of that estimate. Also, this sentiment is likely to be met by a cynical reader (one who is more skeptical about the value of preprints than I am), that this is a good thing, that many published articles, and certainly many preprints, do not \"deserve\" to be published. I recommend addressing or acknowledging that sentiment here.\n\n\"There was also a general sense that the F1000 Research platform appeared to be ‘modern’ and yet was potentially ‘less attractive to the early career researcher’ because the need to publish in recognised journals with high impact factors is still considered the standard by the vast majority of researchers and institutes to gain promotion and tenure.\"\n‘less attractive to the early career researcher’ Is that a quote from a survey respondent? State that in this part of the discussion, as that could also have come from other sources discussing the pressures on ECRs to publish in more \"prestigious\" venues.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "35534",
"date": "19 Jul 2018",
"name": "Anisa Rowhani-Farid",
"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\nOverall\n\nThis is an excellent meta-research study of the open science platform, F1000 Research, using observational data and survey results. The results of this study contribute to the shaping of journals’ conceptual framework, which influence their publishing practices, such as the adoption of post-publication peer-review, reducing publication costs, implementing open access policies, and reducing time between submission and publication. I approve this study for publication, upon addressing the following points:\n\nTitle\nThe title was a bit confusing grammatically as the “who and why” would require differing conjugation and the “survey of F1000 Research” phrase might need to include “researchers”. Here is a suggestion for the title: A meta-research study of the post-publication platform, F1000 Research – who publishes there and why – findings from a review and survey.\n\nIntroduction\nThis was good and covered the issue of “waste” in research because of the traditional journal publication system. I wonder whether there is a dollar figure that demonstrates this waste?\n\nI think two other articles might be worthwhile citing in this introduction:\n\nAleksic et al1 Tracz et al2\n\nMethods and results\n\nThe methods used are scientifically sound and clear.\n\nI think the ~one-month response period for the survey could be identified as a limitation of the study, especially given that there was only a 25-30% response rate, and after a few reminders too. The emails could have ended up in researchers’ spam/clutter folders too.\n\nI have had a look at the data shared. The excel spread sheet contains the observational data of the articles published with F1000 Research from 13 July 2012 to 30 November 2017. It might be worthwhile to export an anonymised copy of the survey data from REDCap to share as well.\n\nI am wondering why peer review was discontinued for those 36 articles.\nDiscussion\nThis was clear. I have no further comments.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-920
|
https://f1000research.com/articles/7-296/v1
|
08 Mar 18
|
{
"type": "Research Article",
"title": "High monocyte to lymphocyte ratio is associated with impaired protection after subcutaneous administration of BCG in a mouse model of tuberculosis",
"authors": [
"Andrea Zelmer",
"Lisa Stockdale",
"Satria A. Prabowo",
"Felipe Cia",
"Natasha Spink",
"Matthew Gibb",
"Ayad Eddaoudi",
"Helen A. Fletcher",
"Lisa Stockdale",
"Satria A. Prabowo",
"Felipe Cia",
"Natasha Spink",
"Matthew Gibb",
"Ayad Eddaoudi",
"Helen A. Fletcher"
],
"abstract": "Background: The only available tuberculosis (TB) vaccine, Bacillus Calmette-Guérin (BCG), has variable efficacy. New vaccines are therefore urgently needed. Why BCG fails is incompletely understood, and the tools used for early assessment of new vaccine candidates do not account for BCG variability. Taking correlates of risk of TB disease observed in human studies and back-translating them into mice to create models of BCG variability should allow novel vaccine candidates to be tested early in animal models that are more representative of the human populations most at risk. Furthermore, this could help to elucidate the immunological mechanisms leading to BCG failure. We have chosen the monocyte to lymphocyte (ML) ratio as a correlate of risk of TB disease and have back-translated this into a mouse model. Methods: Four commercially available, inbred mouse strains were chosen. We investigated their baseline ML ratio by flow cytometry; extent of BCG-mediated protection from Mtb infection by experimental challenge; vaccine-induced interferon gamma (IFNγ) response by ELISPOT assay; and tissue distribution of BCG by plating tissue homogenates. Results: The ML ratio varied significantly between A/J, DBA/2, C57Bl/6 and 129S2 mice. A/J mice showed the highest BCG-mediated protection and lowest ML ratio, while 129S2 mice showed the lowest protection and higher ML ratio. We also found that A/J mice had a lower antigen specific IFNγ response than 129S2 mice. BCG tissue distribution appeared higher in A/J mice, although this was not statistically significant. Conclusions: These results suggest that the ML ratio has an impact on BCG-mediated protection in mice, in alignment with observations from clinical studies. A/J and 129S2 mice may therefore be useful models of BCG vaccine variability for early TB vaccine testing. We speculate that failure of BCG to protect from TB disease is linked to poor tissue distribution in a ML high immune environment.",
"keywords": [
"Tuberculosis",
"animal models",
"BCG",
"vaccine",
"ML ratio",
"mice"
],
"content": "Introduction\n\nTuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb) is the leading cause of death from a single infectious agent. Multidrug-resistant TB remains a public health crisis, and only one vaccine (the M. bovis-derived Bacillus Calmette-Guérin, BCG) is currently licensed for clinical use. To meet the sustainable development goal of ending the TB epidemic by 2030, new treatments and vaccines are both urgently needed.\n\nBCG efficacy is highly variable1,2. The reasons why BCG protects when it does and why it fails when it doesn’t are incompletely understood, but are crucial to the successful design and testing of new vaccines3. Furthermore, to be able to accurately assess vaccine candidates in models where BCG both protects and does not protect very early on in the vaccine development pipeline would be highly advantageous and de-risk failure in later stage clinical trials. There is however a lack of a broad range of tools that can collectively predict with some confidence whether a vaccine will be protective.\n\nOne of the tools for very early testing of TB vaccine candidates is a mouse model. However, in order to obtain meaningful information, new vaccine candidates should be tested in models that are clinically relevant and reflect the breadth and heterogeneity of immune environments and varying BCG efficacy found in human populations. We propose that this could be achieved by back-translating observations from clinical studies, such as correlates of risk of TB disease, into animal models.\n\nA number of correlates of risk of TB disease have recently been identified, including transcriptomic mRNA signatures in blood, T cell activation, and monocyte to lymphocyte (ML) ratio4–8, while others are being investigated (reviewed in 9). We have chosen the ML ratio to provide proof of principle that correlates of risk can be back-translated into the mouse to develop a model for vaccine testing that better reflects the populations most at risk of BCG failure.\n\nIn this study, we show that inbred, commercially available mouse strains have differing ML ratios, and that a high ML ratio is associated with lower BCG-mediated protection from experimental Mtb challenge. We further suggest that lack of BCG dissemination and/or persistence in ML high mice impairs protection.\n\n\nMethods\n\nAll animal work was carried out in accordance with the Animals (Scientific Procedures) Act 1986 under a license granted by the UK Home Office (PPL 70/8043), and approved locally by the London School of Hygiene and Tropical Medicine Animal Welfare and Ethics Review Body.\n\nThe following mouse strains were used for the experiments reported here (abbreviated names used throughout the manuscript are given in brackets): A/JOlaHsd (A/J); DBA/2OlaHsd (DBA/2); C57BL/6JOlaHsd (C57Bl/6); 129S2/SvHsd (129S2). Female mice were acquired from Envigo UK at 5–7 weeks of age. Animals were housed in specific pathogen-free individually vented cages with environmental enrichment (play tunnel and tapvei block), with 12 hours light / 12 hours dark cycles, at temperatures between 19° – 23°C and relative humidity of 45 – 65%. Mice were fed sterilized diet RM1 and filtered water ad libitum, and were allowed to acclimatize for at least 5 days before the start of any experimental procedure. Mice were allocated to cages as groups of 5 by technical staff not involved in experimental procedures or data analysis, and mice of the same strain were housed together. Each cage was allocated to a treatment in no particular order, but without formal randomisation. Animal welfare was assessed twice every day before and during the study.\n\nTwo independent experiments with separate primary outcomes were carried out to obtain the data described in this report. In Experiment 1, 5 naïve mice of each strain (20 mice total) were culled by anaesthetic overdose, and cardiac blood, lungs and spleens were collected for determination of the ML ratio by flow cytometric analysis (see below for details). The primary outcome for this experiment was the ML ratio (Figure 1). In Experiment 2, parallel groups of mice (n=5 per group) were BCG-immunised or left untreated. These were allocated at the start of the study to either an immunogenicity group (IFNγ response and BCG dissemination; 10 mice per strain; 40 mice total; Figure 3 and Figure 4), or an Mtb challenge group (bacterial burden in lung; 10 mice per strain; 40 mice total; Figure 2). Six weeks after immunisation, mice were infected with Mtb or culled for isolation of cells from lung and spleen. In all, a total of 100 mice (25 per strain) was used to obtain all data presented here.\n\nA Gating strategy for flow cytometric analysis. Cells from naïve animals were fixed, stained and data acquired as described in Materials and Methods. Cell debris was gated out by use of a FSC-SSC gate, followed by gating on single cells (FSC-H and FSC-W). A sequential gating strategy was then applied to determine the frequency of T cells (CD3+), B cells (B220+), neutrophils (CD11b+ Ly6G+) and monocytes/macrophages (CD11b+ CD11clow-int) as a percentage of single cells. Plots shown are from a sample of a C57Bl/6 spleen. B–D The ML ratio was calculated by dividing the percentage of monocytes/macrophages by the sum of the percentages of B and T cells. ML ratio was analysed in blood (B), lung (C) and spleen (D) of four different mouse strains. Each symbol represents one animal; box plots represent the median (middle line), 25th to 75th percentile (box) and minimum to maximum value (error bars). Data sets are presented in order of decreasing protection. p values were determined using ordinary ANOVA with Holm-Sidak test for multiple comparisons. Multiplicity adjusted p values are reported. A p value <0.05 was considered statistically significant.\n\nA Mice of each strain were immunised s.c. with 2x10e5 CFU BCG Pasteur (open circles) or left untreated (filled circle), and infected i.n. with 30 CFU Mtb Erdmann 6 weeks later. Bacterial burden in the lungs of all animals was enumerated 7 weeks after challenge. Each symbol represents one animal. Bacterial numbers are given as log10 CFU per whole organ. No bacteria were detected in one of the C57Bl/6 samples; this value was set to zero. Δ indicates the difference in bacterial burden between naïve and BCG immunised mice. p values were determined by Kruskal-Wallis test with Dunn’s post-test for multiple comparisons between naïve groups, and multiplicity adjusted p values are reported. Individual Mann-Whitney tests to compare naïve with BCG immunised groups of each mouse strain. p<0.05 was considered statistically significant. Error bars represent the median and interquartile range. LoD: limit of detection. B ML ratio in comparison to protection in the lung of ML high and low mice. The median ML ratio in lung (red; as in Figure 1) is plotted together with protection expressed as the median % decrease in CFU (blue; as in A).\n\nSplenocytes were isolated from BCG-immunised or control mice at the time of Mtb challenge (6 weeks after immunisation), and restimulated with PPD. The number of IFNγ producing cells were enumerated using an ELISPOT assay (presented as spot forming units [SFU] per 10e6 cells). Non-specific background measured in unstimulated duplicate wells was removed. Data sets are presented in order of decreasing protection. Each symbol represents one animal. p values were determined by using one-way ANOVA and Dunn’s post test for multiple comparisons. Error bars represent the median and interquartile range. Multiplicity adjusted p values are reported. p<0.05 was considered statistically significant.\n\nApproximately half of each lung (A) and spleen (B) was homogenised at the time of Mtb challenge (6 weeks after immunisation) and plated on 7H11 agar plates to determine viable BCG bacteria. Total CFU per organ are reported. Data sets are presented in order of decreasing protection. Each symbol represents one animal. In some instances, bacteria could not be detected and values were set to zero. p values were determined by using Kruskal-Wallis test and Dunn’s post test for multiple comparisons. Error bars represent the median and interquartile range. Multiplicity adjusted p values are reported. p<0.05 was considered statistically significant.\n\nThe study was not blinded.\n\nThe BCG Pasteur strain was obtained from Aeras (Rockville, MD, USA) as frozen aliquots. These were stored at -80°C until needed. BCG was then thawed at room temperature and diluted to a final concentration of 2×106 CFU/ml in physiological saline solution for irrigation (Baxter Healthcare, Newbury, UK). Each animal received a subcutaneous injection of 100 μl BCG containing 2×105 CFU BCG (vaccinated groups). BCG dose was confirmed by plating of an aliquot of the prepared vaccine suspension on 7H11 agar plates. Colonies were counted after 12–14 days of incubation at 37°C. Animals were then rested for 6 weeks before either infection with Mtb, or sacrifice for cell isolation.\n\nApproximately half of each spleen and the right-hand side lobes of the lungs from mice in the immunogenicity group were used to determine the number of viable BCG bacteria in each organ six weeks after vaccination. Tissues were removed aseptically in a microbiological safety cabinet and placed in sterile 2 ml screwcap vials containing 500 μl PBS + 0.05% Tween80 and Precellys 1.4mm ceramic beads (CK14; Peqlab, Sarisbury Green, UK). A Precellys 24 homogeniser (Peqlab) was used to homogenise tissues for 15 s at 5000 rpm before plating. Each entire homogenate was plated onto two 7H11 agar plates containing 10% OADC supplement (Yorlab, York, UK) and 0.5% glycerol. Colonies were counted after 3 weeks of incubation at 37°C.\n\nMice were infected intranasally with M. tuberculosis Erdman (BEI Resources, Manassas, VA, USA) 6 weeks after BCG immunization and kept in isolators under CL-3 containment. Frozen aliquots of Mtb Erdman were thawed at room temperature, and diluted in saline. Mice were anaesthetized by an intraperitoneal injection of a combination of Ketamine (50 mg/kg; Ketalar Pfizer Itd, Kent, UK) and Xylazine (10 mg/kg; Rompun; Berkshire, UK) in saline. Each animal then received 50 μl of the inoculum, estimated to contain 30 CFU. The number of bacteria in the inoculum was confirmed by plating aliquots on 7H11 agar plates containing 10 % OADC and 0.5 % glycerol.\n\nSeven weeks after infection, animals were killed by cervical dislocation. Lungs and spleens were removed aseptically and homogenized by mechanical disruption in sterile PBS, using the plunger of a 5 ml syringe and a 100 μm cell strainer. A series of 10-fold dilutions of tissue homogenates in PBS with 0.05 % Tween 80 were plated onto 7H11 agar plates with 10 % OADC supplement and 0.5 % glycerol. Plates were incubated at 37 °C and colonies counted after 3 weeks.\n\nProtection is expressed by the percent reduction of median CFU/lung in the BCG group compared to the control group for each mouse strain.\n\nSingle cell suspensions from spleens were prepared in RPMI-1640 media (Sigma-Aldrich, Dorset, UK) containing 10% heat-inactivated FBS (Labtech International Ltd, Uckfield, UK) and 2 mM L-Glutamine (Fisher Scientific, Loughborough, UK) as soon as possible after sacrifice. Spleens were mechanically disrupted by mashing through a 100μm cell strainer using the rubber end of the plunger from a 5ml syringe. Lungs were collected into RPMI-1640 media without FBS and cut into small pieces of approx. 2mm3 before incubation with 0.5mg/ml Liberase TL (Sigma-Aldrich) and 10μg/ml DNAse (Sigma-Aldrich). The enzyme reaction was stopped by adding an equal amount of media containing 10% FBS and the tissue was mashed as above to obtain single cells. Cells were fixed and red blood cells lysed by adding lyse-fix solution (PhosFlow; Becton Dickinson, Oxford, UK). After fixing, cell suspensions were made up in PBS + 1% FBS.\n\nApproximately 106 cells were stained with the following antibody cocktail in BD Brilliant Stain buffer as per manufacturer’s instructions (Becton Dickinson): CD3-APC/Cy7 (clone 17A2, 1:80), B220-BV510 (clone RA3-6B2, 1:40), Ly6G-BV711 (clone 1A8, 1:40), CD11b-BV650 (clone M1/70, 1:60), CD11c-BV605 (clone N418, 1:40). All antibodies were purchased from Biolegend (via Fisher Scientific).\n\nCells from each tissue from one mouse per strain were used as fluorescence minus one (FMO) controls. These were stained with the antibody cocktail as described above, but without one of the antibodies. This was done for each antibody. FMO controls were used to guide gating.\n\nOneComp beads (eBioscience via Fisher Scientific, Loughborough, UK) were stained with single antibodies as per manufacturer’s instructions and used to calculate compensation using the automatic function in FlowJo version 10.4. Compensation matrices were manually checked and adjusted where necessary.\n\nAnalysis was carried out using FlowJo version 10.4.\n\nThe ML ratio was calculated by dividing the percentage of monocytes/macrophages (Mono_Mac) (B220- CD3- Ly6G- CD11b+ CD11clo-int) relative to single cells by the percentage of B cells (B220+) and T cells (CD3+): Mono_Mac (% of single cells) / [B cells (% of single cells) + T cells (% of single cells)]\n\nTo quantify IFN-γ secreting antigen-specific splenocytes, single cell suspensions were prepared by mechanical disruption of the remaining spleen samples from the immunogenicity group through a 100μm cell strainer as soon as possible after sacrifice. After lysis of red blood cells, single cell suspensions were made up in RPMI-1640 media containing 10% heat-inactivated FBS and 2 mM L-Glutamine. 96-well microtiter ELISPOT plates (MAIPS4510, Millipore, Watford, UK) were coated with 10 µg/ml rat anti-mouse IFN-γ (clone AN18, Mabtech, Nacka Strand, Sweden). Free binding sites were blocked with RMPI-1640 supplemented with 10% heat-inactivated FBS and 2 mM L-Glutamine. 2×105 of total splenocytes were added and incubated in duplicate with PPD (10 µg/ml), supplemented RPMI as a negative control, or Phorbol myristate acetate (PMA) (0.1 µg/ml, Sigma-Aldrich) and Phytohemagglutinin (PHA) (1 µg/ml, Sigma-Aldrich) as a positive control. After overnight incubation at 37°C in 5% CO2, IFN-γ was detected with 1 µg/ml biotin labelled rat anti-mouse antibody (clone R4-6A2, Mabtech) and 1 µg/ml alkaline phosphatase-conjugated streptavidin (Mabtech). The enzyme reaction was developed with BCIP/NBT substrate (5-Bromo-4-chloro-3-indolyl phosphate/Nitro blue tetrazolium) (MP Biochemicals, UK) and stopped by washing the plates with tap water when individual spots could be visually detected (up to 5min). ELISPOT plates were analysed using an automatic plate reader. IFN-γ-specific cells are expressed as number of spot-forming units (SFU) per million spleen cells after non-specific background was subtracted using negative control wells.\n\nGroups of animals were compared, and a p value of <0.05 was considered statistically significant. Statistical analysis was carried out using GraphPad Prism version 6. The specific test used for each analysis is described in the figure legends.\n\n\nResults\n\nFour commercially available inbred mouse strains were chosen to investigate the impact of the baseline ML ratio on BCG vaccine efficacy. A/J, DBA/2, C57Bl/6 and 129S2 mice were selected to represent varying monocyte frequencies and a range of BCG-mediated protection based on available data (Jax Phenome Database https://phenome.jax.org;10,11). To allow the direct comparison of the ML ratio between strains, animals were age and sex matched and all samples were processed at the same time, stained with aliquots of the same antibody cocktail, and data acquired as a batch. All animals used in this experiment were included in the analysis.\n\nWe found significant differences between mouse strains in their baseline ML ratio in blood, spleen, and lung (Figure 1). Across all tissues, A/J mice showed the lowest ML ratio (Figure 1 B–D). In spleen and blood, the ML ratio of DBA/2, C57Bl/6 and 129S2 mice was higher than A/J, but only minor differences were found between those three strains (Figure 1 B and D). In the lung, differences between all strains were more apparent (Figure 1C). 129S2 mice had the highest ML ratio, with DBA/2 and C57Bl/6 showing intermediate levels. These differences in ML ratio were mostly driven by differing frequencies in monocytes/macrophages, with minor differences in B and T cell frequencies between mouse strains (Figure. S1–S3).\n\nTo assess the impact of these differences in ML ratio on BCG efficacy in our model, we infected mice of all strains with virulent Mtb Erdmann.\n\nMice of each strain were vaccinated s.c. with BCG, or left untreated, and infected with Mtb Erdmann six weeks later. The extent of protection by BCG was measured by determining the bacterial burden in lung and spleen of vaccinated and control mice (Figure 2). All mice were immunised with the same inoculum of BCG and infected using the same inoculum of Mtb, so that results of the different strains are directly comparable. All animals were included in the analysis and no unexpected adverse events were observed. No more than 10% weight loss was recorded over the duration of the experiment. We found varying degrees of protection (Figure 2A). A/J mice were the most protected with a difference in median CFU (Δ) between vaccinated and control groups of 1.41 log10, followed by DBA/2 (Δ0.81 log10) and C57Bl/6 (Δ0.54 log10), and 129S2 mice were the least protected (Δ0.37 log10). Interestingly, the mouse strain with the highest ML ratio in the lung showed the lowest protection (129S2), and the strain with the lowest ML ratio showed the highest protection (A/J; Figure 2B). Some differences were also observed in the innate susceptibility between strains, most notably C57Bl/6 mice were significantly less susceptible than DBA/2 (Δ1.16 log10 CFU) or A/J mice (Δ1.41 log10 CFU). To investigate whether this differential protection was associated with a differential immune response, we carried out an IFNγ ELISPOT assay.\n\nTo assess the immune response to mycobacterial antigens in the differentially protected mouse strains, splenocytes from mice vaccinated with BCG were stimulated with PPD, and the number of IFNγ-producing cells was measured by ELISPOT (Figure 3). IFNγ is a cytokine associated with control of infection in mice12 and reduced risk of TB disease in humans6. We found that there were significantly more antigen-specific IFNγ producing splenocytes in 129S2 mice than in A/J or DBA/2 mice, both of which were better protected by BCG from Mtb challenge than 129S2 mice.\n\nSince BCG is a live mycobacterium, and monocytes and macrophages are the natural host cells for these organisms, it was plausible that a high ML ratio and high IFNγ production in 129S2 mice impair BCG survival and dissemination to and/or persistence in tissues, and thus prevent the formation of a protective immune response in the lung.\n\nTo determine the extent to which BCG is found in lungs and spleens after s.c. administration, we plated organ homogenates onto 7H11 agar plates and enumerated viable BCG after 3 weeks of culture (Figure 4). We generally found low numbers of bacteria per organ, especially in the lung, where the bacterial count did not exceed 80 CFU/lung in any sample. Moderate differences were detected between mouse strains. Most notably, the most protected ML low A/J mice had the highest number of BCG in the lung, while no BCG was detectable in the lung of the least protected ML high 129S2 mice.\n\nThese data led us to speculate that certain host immune-environments, such as a high ML ratio, can influence BCG dissemination and/or persistence, which in turn could impact BCG efficacy.\n\n\nDiscussion\n\nThe varying efficacy of BCG has long been a concern for the control of the TB epidemic. Numerous factors have been discussed as contributors to this, such as population differences, different BCG strains, vaccination schedules, exposure to environmental mycobacteria, co-infections with viruses and/or parasites, and geographical location1,3,13–16). While BCG works in some populations and protects children from pulmonary and extra-pulmonary disease, new vaccines are urgently needed for the populations where BCG fails to protect, such as adolescents and adults in endemic areas1,2. The incomplete understanding of the mechanisms behind BCG failure, together with limited tools for early assessment of vaccine efficacy, are hampering the development of new vaccines. In order to allow early vaccine testing in a model that better represents the populations most at risk of TB disease, we have taken a correlate of risk observed in human studies and back-translated it into a mouse model. The monocyte to lymphocyte (ML) ratio is a non-specific marker of inflammation and has been shown to be associated with risk of TB disease in several populations, including pregnant HIV infected women, people starting anti-retroviral therapy, BCG vaccinated infants and latently Mtb infected adolescents7,8,17–19). Scriba et al show a cascade of inflammatory events as adolescents progress towards disease, which appears to start with increased Type I/II IFN signalling, followed by an increase in monocytes and a decrease in lymphocytes19. An increased risk of TB disease in BCG vaccinated individuals with a high ML ratio indicates that BCG is failing in those populations.\n\nIn order to provide proof of principle that the ML ratio, and observations from human studies more generally, can be back-translated to develop more clinically relevant mouse models, we chose to use commercially available, genetically tractable inbred mouse strains. We could show that these have highly varying ML ratios in blood, spleen, and lung, and that these differences, particularly in the lung, were associated with varying BCG-mediated protection from Mtb infection. Using mouse strains with genetically different backgrounds means that there are likely other confounding factors that could impact vaccine efficacy. This is exemplified by the DBA/2 mice, which are well protected despite having a relatively high ML ratio across all tissues. This could be due to the fact that they have a higher neutrophil frequency than some of the other strains (data not shown). Although confounding factors may be at play here, this would also be the case in a clinical setting since an altered ML ratio can have different underlying causes. An increased ML ratio is an indicator for inflammation, likely driven by monocytosis during inflammation. Some factors that are thought to influence BCG efficacy also drive inflammation, such as co-infection with viruses or parasites, malnutrition, or exposure to environmental mycobacteria. However, as manipulation of ML ratio in vitro impacts on ability to control mycobacterial growth, it is likely that ML ratio itself is contributing to TB risk, independent of the factor driving inflammation20. It is also possible that host genetic factors determine baseline inflammation and ML ratio21. If this is the case in TB endemic populations, inbred mouse strains with naturally varying ML ratios may be a useful model for BCG vaccine variability.\n\nIt is currently unclear what the relative contributions are of genetic and environmental factors to a change in ML ratio21. Does a naturally high ML ratio in an individual make them more prone to TB disease, or do other factors including Mtb itself drive up the ML ratio, causing BCG to fail? Does a naturally high ML ratio in an individual exacerbate an inflammatory response initiated by environmental factors? There is striking heterogeneity in the immune response to BCG in infants vaccinated at birth, including in the ML ratio and cytokine responses17. This points towards an involvement of host factors, as the influence of environmental factors would be limited so early in life. On the other hand, Scriba et al observe increases in inflammation and ML ratio over time in adolescents as an individual progresses towards active TB disease and diagnosis19. This may point towards the pathogen itself as the cause of an altered immune environment.\n\nInterestingly, we found a higher number of antigen-specific IFNγ-producing splenocytes in unprotected 129S2 mice compared to protected A/J mice at the time of Mtb challenge. This seemed counterintuitive at first, but is in line with recent findings showing that a sub-population of BCG-vaccinated infants with an increased ML ratio showed increased T cell-mediated cytokine production, and this was associated with increased risk of TB disease17. High monocyte count and pro-inflammatory cytokines are indicators for inflammation. It is possible that BCG efficacy is impaired in inflammatory immune environments, which may be caused by a variety of factors, such as chronic viral infection, age, malnutrition, or prolonged contact with environmental mycobacteria, all factors thought to impact BCG efficacy. Ultimately, BCG is administered into host immune environments with a heterogeneity that is not fully characterised or understood, yet may have implications for its efficacy.\n\nThe exact mechanisms behind an altered BCG vaccine efficacy in a ML high immune environment are currently unknown. Our data indicate that impaired dissemination and/or persistence of BCG may play a role in initiating an effective local immune response. A study by Kaveh and colleagues has shown that BCG persists for up to 16 months in lymph nodes and spleen of Balb/c mice and that persistence of live bacilli contributes to optimal protection from M. bovis challenge22. Direct investigation of this phenomenon in humans is not possible; there are however reports that indicate that BCG can disseminate and persist for years in a variety of organs in humans23,24. It is therefore plausible that an immune environment that impairs survival of BCG also impairs vaccine efficacy, although we did not determine in this study whether the absence of live bacilli in lungs of 129S2 mice is due to lack of dissemination or lack of persistence.\n\nThe absolute numbers of BCG might be under-estimated in this study as we only enumerated bacilli that were readily growing on standard 7H11 agar plates. Use of PCR methods or culture with resuscitation promoting factors may increase yield and allow for detection of dormant bacilli and their contribution to protection25,26.\n\nIf impaired dissemination and/or persistence in certain immune environments is indeed reducing BCG vaccine efficacy, it is reasonable to assume that other live mycobacterial vaccines would have similar limitations. Variability in the ML ratio and more general heterogeneity in the host immune environment should thus be taken into account when designing new live TB vaccines or vaccines based on boosting BCG.\n\nTo circumvent potentially impaired dissemination of parenterally administered live vaccines, mucosal administration of live vaccines could be considered. Recent evidence shows that mucosal administration of BCG confers increased protection in animal models, for example by inducing homing of T cells into protected mucosal niches27–29. This agrees with the notion that BCG needs to be present in the lungs to initiate a protective immune response in this organ. Furthermore, subunit vaccines should not show decreased efficacy in ML high immune environments if our hypothesis is correct. Both these possibilities will need careful investigation to shed light on the mechanisms behind BCG failure.\n\nWe did not investigate the phenotype of monocytes/macrophages or T cells in this study, although it is likely that this plays a role. In particular, inflammatory monocytes might lead to killing of BCG. It will be interesting in future to investigate this, as well as comparing the ML ratio and immune cell phenotypes before and after BCG vaccination in the different mouse strains and locations.\n\nWhile our data on the susceptibility of naïve mouse strains to Mtb infection agrees with other studies30, we found some discrepancies between the data presented here and previously published data on BCG-mediated protection. A study using DBA/2 mice has shown that these mice are not protected by s.c. administration of BCG, but control infection with Mtb better when vaccinated i.n.11 This is in contrast to our finding that DBA/2 mice are protected from Mtb challenge after s.c. vaccination. However, there are some differences between these two studies that may explain the different findings: a different source of mice was used; protection was assessed at 8 weeks vs 6 weeks after vaccination; the BCG strain and Mtb strain differed; and the Mtb dose used for challenge was three times higher in the study by Aguilo and colleagues compared to ours. Another report detailing the susceptibility of different mouse strains to Mtb challenge shows that A/J mice are notably less protected than C57Bl/6 mice10, while in our hands the opposite is the case. There are important differences between the two studies, including the age and sex of mice used, the time point after vaccination at which protection was measured, and the use of different BCG and Mtb strains. In combination, these factors may lead to a differing extent of protection. With particular relevance to this study, the immune cell frequency between male and female animals differs in A/J mice. For example, dataset Donahue5 on the Jackson Laboratory Mouse Phenome Database indicates a higher percentage of monocytes in blood for males compared to females, while T cells are less frequent in males, leading to a higher ML ratio in males31.\n\nIn conclusion, the back-translation of a correlate of risk of TB disease, the ML ratio, into an animal model is a step forward for better tools to study the immune mechanisms behind BCG vaccine failure and to test vaccines in a model more relevant to populations most at risk. Hopefully other risk factors can also be back-translated in the future to obtain a more complete picture of vaccine efficacy at very early stages of testing. We would encourage the use of more diverse mouse models in TB vaccine testing and development to reflect the heterogeneity of immune environments found in human populations.\n\n\nData availability\n\nDataset 1: Flow cytometry raw data – This file contains the data underlying the analysis of the ML ratio and cell subsets in blood, spleen, and lung shown in Figure 1, Figure 2B, Figure S1, Figure S2, and Figure S3. 10.5256/f1000research.14239.d19711732\n\nDataset 2: Mtb challenge CFU raw data – This files contains the raw data as colony forming units (CFU) of the graphs in Figure 2. 10.5256/f1000research.14239.d19711833\n\nDataset 3: ELISPOT raw data – This file contains the raw data for the graph shown in Figure 3. 10.5256/f1000research.14239.d19713634\n\nDataset 4: BCG distribution raw data – This file contains the raw data as colony forming units (CFU) for the graph shown in Figure 4. 10.5256/f1000research.14239.d19713735",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe work in this article was funded by a grant to HF from the European Commission HORIZON2020 program (TBVAC2020 grant no. 643381) and an Athena SWAN Career Re-entry award funded by the London School of Hygiene and Tropical Medicine to AZ.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Christopher Sassetti for helpful discussions and selecting the mouse strains, and Stephanie Canning for expert technical assistance with flow cytometry studies.\n\n\nSupplementary material\n\nFigure S1: The frequency of T cells varies in tissues of ML high and ML low mouse strains. T cell (CD3+) frequencies were determined in blood (A), lung (B) and spleen (C) of four different mouse strains using the gating strategy described in Figure 2A. The ML ratio was calculated using these data. Each symbol represents one animal; box plots represent the median (middle line), 25th to 75th percentile (box) and minimum to maximum value (error bars). Data sets are presented in order of decreasing protection. p values were determined using ordinary ANOVA with Holm-Sidak test for multiple comparisons. Multiplicity adjusted p values are reported. A p value <0.05 was considered statistically significant.\n\nClick here to access the data.\n\nFigure S2: The frequency of B cells varies in tissues of ML high and ML low mouse strains. B cell frequencies (B220+) were determined in blood (A), lung (B) and spleen (C) of four different mouse strains using the gating strategy described in Figure 2A. The ML ratio was calculated using these data. Each symbol represents one animal; box plots represent the median (middle line), 25th to 75th percentile (box) and minimum to maximum value (error bars). Data sets are presented in order of decreasing protection. p values were determined using ordinary ANOVA with Holm-Sidak test for multiple comparisons. Multiplicity adjusted p values are reported. A p value <0.05 was considered statistically significant.\n\nClick here to access the data.\n\nFigure S3: The frequency of monocytes/macrophages varies in tissues of ML high and ML low mouse strains. Monocyte/macrophage (CD11b+ CD11clow-int) frequencies were determined in blood (A), lung (B) and spleen (C) of four different mouse strains using the gating strategy described in Figure 2A. The ML ratio was calculated using these data. Each symbol represents one animal; box plots represent the median (middle line), 25th to 75th percentile (box) and minimum to maximum value (error bars). Data sets are presented in order of decreasing protection. p values were determined using ordinary ANOVA with Holm-Sidak test for multiple comparisons. Multiplicity adjusted p values are reported. A p value <0.05 was considered statistically significant.\n\nClick here to access the data.\n\n\nReferences\n\nMangtani P, Abubakar I, Ariti C, et al.: Protection by BCG vaccine against tuberculosis: a systematic review of randomized controlled trials. Clin Infect Dis. Oxford University Press; 2014; 58(4): 470–80. PubMed Abstract | Publisher Full Text\n\nColditz GA, Brewer TF, Berkey CS, et al.: Efficacy of BCG vaccine in the prevention of tuberculosis. Meta-analysis of the published literature. JAMA. 1994; 271(9): 698–702. PubMed Abstract | Publisher Full Text\n\nDockrell HM, Smith SG: What Have We Learnt about BCG Vaccination in the Last 20 Years? Front Immunol. Frontiers; 2017; 8: 1134. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZak DE, Penn-Nicholson A, Scriba TJ, et al.: A blood RNA signature for tuberculosis disease risk: a prospective cohort study. Lancet. 2016; 387(10035): 2312–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerry MP, Graham CM, McNab FW, et al.: An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature. 2010; 466(7309): 973–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFletcher HA, Snowden MA, Landry B, et al.: T-cell activation is an immune correlate of risk in BCG vaccinated infants. Nat Comms. 2016; 7: 11290. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaranbhai V, Hill AV, Abdool Karim SS, et al.: Ratio of monocytes to lymphocytes in peripheral blood identifies adults at risk of incident tuberculosis among HIV-infected adults initiating antiretroviral therapy. J Infect Dis. 2014; 209(4): 500–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaranbhai V, Kim S, Fletcher H, et al.: The association between the ratio of monocytes:lymphocytes at age 3 months and risk of tuberculosis (TB) in the first two years of life. BMC Med. 2014; 12(1): 120. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetruccioli E, Scriba TJ, Petrone L, et al.: Correlates of tuberculosis risk: predictive biomarkers for progression to active tuberculosis. Eur Respir J. European Respiratory Society; 2016; 48(6): 1751–63. PubMed Abstract | Publisher Full Text\n\nSmith CM, Proulx MK, Olive AJ, et al.: Tuberculosis Susceptibility and Vaccine Protection Are Independently Controlled by Host Genotype. mBio. 2016; 7(5): pii: e01516–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAguilo N, Alvarez-Arguedas S, Uranga S, et al.: Pulmonary but Not Subcutaneous Delivery of BCG Vaccine Confers Protection to Tuberculosis-Susceptible Mice by an Interleukin 17-Dependent Mechanism. J Infect Dis. Oxford University Press; 2016; 213(5): 831–9. PubMed Abstract | Publisher Full Text\n\nCooper AM, Dalton DK, Stewart TA, et al.: Disseminated tuberculosis in interferon gamma gene-disrupted mice. J Exp Med. Rockefeller Univ Press; 1993; 178(6): 2243–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrosch R, Gordon SV, Garnier T, et al.: Genome plasticity of BCG and impact on vaccine efficacy. Proc Natl Acad Sci U S A. National Academy of Sciences; 2007; 104(13): 5596–601. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuller J, Matsumiya M, Snowden MA, et al.: Cytomegalovirus infection is a risk factor for TB disease in Infants. bioRxiv. Cold Spring Harbor Laboratory; 2017; 222646. Publisher Full Text\n\nLalor MK, Ben-Smith A, Gorak-Stolinska P, et al.: Population differences in immune responses to Bacille Calmette-Guérin vaccination in infancy. J Infect Dis. Europe PMC Funders; 2009; 199(6): 795–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatterjee N, Natman TB: Helminth-Induced Immune Regulation: Implications for Immune Responses to Tuberculosis. Knoll LJ editor. PLoS Pathog. Public Library of Science; 2015; 11(1): e1004582. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFletcher HA, Filali-Mouhim A, Nemes E, et al.: Human newborn bacille Calmette-Guérin vaccination and risk of tuberculosis disease: a case-control study. BMC Med. BioMed Central; 2016; 14(1): 76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaranbhai V, Moodley D, Chipato T, et al.: The association between the ratio of monocytes: lymphocytes and risk of tuberculosis among HIV-infected postpartum women. J Acquir Immune Defic Syndr. NIH Public Access; 2014; 67(5): 573–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScriba TJ, Penn-Nicholson A, Shankar S, et al.: Sequential inflammatory processes define human progression from M. tuberculosis infection to tuberculosis disease. Sassetti CM editor. PLoS Pathog. Public Library of Science; 2017; 13(11): e1006687. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaranbhai V, Fletcher HA, Tanner R, et al.: Distinct Transcriptional and Anti-Mycobacterial Profiles of Peripheral Blood Monocytes Dependent on the Ratio of Monocytes: Lymphocytes. EBioMedicine. 2015; 2(11): 1619–26. PubMed Abstract | Publisher Full Text\n\nLin BD, Willemsen G, Fedko IO, et al.: Heritability and GWAS Studies for Monocyte-Lymphocyte Ratio. Twin Res Hum Genet. Cambridge University Press; 2017; 20(2): 97–107. PubMed Abstract | Publisher Full Text\n\nKaveh DA, Garcia-Pelayo MC, Hogarth PJ: Persistent BCG bacilli perpetuate CD4 T effector memory and optimal protection against tuberculosis. Vaccine. 2014; 32(51): 6911–8. PubMed Abstract | Publisher Full Text\n\nTajima Y, Takagi R, Nakajima T, et al.: An infant with asymptomatic hepatic granuloma probably caused by bacillus Calmette-Guérin (BCG) vaccination found incidentally at autopsy: a case report. Cases J. BioMed Central; 2008; 1(1): 337. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGormsen H: On The Occurrence Of Epithelioid Cell Granulomas In The Organs Of Bcg-Vaccinated Human Beings. Acta Pathol Microbiol Scand Suppl. Blackwell Publishing Ltd; 1956; 38(Suppl 111): 117–20. PubMed Abstract | Publisher Full Text\n\nLoraine J, Pu F, Turapov O, et al.: Development of an In Vitro Assay for Detection of Drug-Induced Resuscitation-Promoting-Factor-Dependent Mycobacteria. Antimicrob Agents Chemother. American Society for Microbiology; 2016; 60(10): 6227–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin TH, Qu T, Raina A, et al.: Identification of Growth Promoting Effect of rBCG/BCG Culture Supernatant and Its Potential Applications. WJV. Scientific Research Publishing; 2013; 2013(02): 32–8. Publisher Full Text\n\nLai R, Afkhami S, Haddadi S, et al.: Mucosal immunity and novel tuberculosis vaccine strategies: route of immunisation-determined T-cell homing to restricted lung mucosal compartments. Eur Respir Rev. European Respiratory Society; 2015; 24(136): 356–60. PubMed Abstract | Publisher Full Text\n\nPerdomo C, Zedler U, Kühl AA, et al.: Mucosal BCG Vaccination Induces Protective Lung-Resident Memory T Cell Populations against Tuberculosis. mBio. American Society for Microbiology (ASM); 2016; 7(6): pii: e01686–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeyanathan M, Afkhami S, Khera A, et al.: CXCR3 Signaling Is Required for Restricted Homing of Parenteral Tuberculosis Vaccine-Induced T Cells to Both the Lung Parenchyma and Airway. J Immunol. American Association of Immunologists; 2017; 199(7): 2555–69. PubMed Abstract | Publisher Full Text\n\nChackerian AA, Behar SM: Susceptibility to Mycobacterium tuberculosis: lessons from inbred strains of mice. Tuberculosis (Edinb). 2003; 83(5): 279–85. PubMed Abstract | Publisher Full Text\n\nDonahue L, Morgan J: Donahue 5. [Internet]. Mouse Phenome Database. [cited 2018 Feb 9]. Reference Source\n\nZelmer A, Stockdale L, Prabowo S, et al.: Dataset 1 in: High monocyte to lymphocyte ratio is associated with impaired protection after subcutaneous administration of BCG in a mouse model of tuberculosis. F1000Research. 2018. Data Source\n\nZelmer A, Stockdale L, Prabowo S, et al.: Dataset 2 in: High monocyte to lymphocyte ratio is associated with impaired protection after subcutaneous administration of BCG in a mouse model of tuberculosis. F1000Research. 2018. Data Source\n\nZelmer A, Stockdale L, Prabowo S, et al.: Dataset 3 in: High monocyte to lymphocyte ratio is associated with impaired protection after subcutaneous administration of BCG in a mouse model of tuberculosis. F1000Research. 2018. Data Source\n\nZelmer A, Stockdale L, Prabowo S, et al.: Dataset 4 in: High monocyte to lymphocyte ratio is associated with impaired protection after subcutaneous administration of BCG in a mouse model of tuberculosis. F1000Research. 2018. Data Source"
}
|
[
{
"id": "32536",
"date": "09 Apr 2018",
"name": "Nacho Aguilo",
"expertise": [
"Reviewer Expertise tuberculosis vaccines immunology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study by Zelmer et al. evaluates in different mouse genetic backgrounds whether the ratio of monocytes/macrophages vs T lymphocytes could affect BCG-induced protection. This study results are highly interesting as it evaluates in an animal model data derived from a previous vaccine efficacy trial. In addition, it highlights a critical question in the field of vaccine evaluation in mice, which is the genetic heterogeneity among inbred mouse strains. This issue is of particular importance when we are assessing vaccines which aim to be moved to humans. During the reading of the manuscript I have found some minor questions that I think should be amended to improve the quality of the study.\n1. I think that the gating criteria used to enclose the monocyte/macrophage population (CD11b+CD11cint cells) is excluding myeloid populations that could be highly informative for the outcome of the study. This is particularly important in the case of the lungs, where naïve alveolar macrophages have a phenotype CD11bintCD11c+. Alveolar macrophages represent a first-line defense against respiratory pathogens, and therefore results could be very informative to know which is the status of these cells in the different strains used. More even if we consider that authors compare directly data from BCG-induced protection with the ML ratio in the lungs (Figure 2B).\n2. I suggest to make the comparison between ML ratio and protection using a linear regression, so statistical significance of the shown data could be assessed.\n3. I find highly interesting the data showing the differences in BCG persistence among the four strains, more even considering the influence of vaccine persistence in protective efficacy. This result evidences that live vaccine persistence does not only depends on the vaccine itself, but also on the host, and this is a question that should be considered in the future to study novel tuberculosis vaccines. My question is whether the authors could provide some data of BCG persistence in the local draining lymph nodes (in addition to lungs and spleen), as they are the primary sites of BCG biodistribution following subcutaneous/intradermal vaccination.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3745",
"date": "27 Jun 2018",
"name": "Andrea Zelmer",
"role": "Author Response",
"response": "Dear Dr Aguilo Thank you for reviewing our manuscript, and for providing timely and helpful comments and feedback. We discuss your individual points below. The study by Zelmer et al. evaluates in different mouse genetic backgrounds whether the ratio of monocytes/macrophages vs T lymphocytes could affect BCG-induced protection. This study results are highly interesting as it evaluates in an animal model data derived from a previous vaccine efficacy trial. In addition, it highlights a critical question in the field of vaccine evaluation in mice, which is the genetic heterogeneity among inbred mouse strains. This issue is of particular importance when we are assessing vaccines which aim to be moved to humans. During the reading of the manuscript I have found some minor questions that I think should be amended to improve the quality of the study. 1. I think that the gating criteria used to enclose the monocyte/macrophage population (CD11b+CD11cint cells) is excluding myeloid populations that could be highly informative for the outcome of the study. This is particularly important in the case of the lungs, where naïve alveolar macrophages have a phenotype CD11bintCD11c+. Alveolar macrophages represent a first-line defense against respiratory pathogens, and therefore results could be very informative to know which is the status of these cells in the different strains used. More even if we consider that authors compare directly data from BCG-induced protection with the ML ratio in the lungs (Figure 2B). We have analysed the frequency of CD11bintCD11c+cells in the lungs of the four different mouse strains, and have added Fig. 1E, as well as described the data in the Results section and discussed it in the Discussion. In brief, the frequency of this population is significantly higher in C57Bl/6 mice, and seems to be associated with susceptibility of naïve mice, but not BCG-mediated protection. 2. I suggest to make the comparison between ML ratio and protection using a linear regression, so statistical significance of the shown data could be assessed. This would indeed be an informative analysis; however, our experimental design does not allow it here. Since the data shown from flow cytometry and protection studies were acquired from two different sets of mice (see Methods section), the data are not paired, and it would not be appropriate to link them in a linear regression. 3. I find highly interesting the data showing the differences in BCG persistence among the four strains, more even considering the influence of vaccine persistence in protective efficacy. This result evidences that live vaccine persistence does not only depends on the vaccine itself, but also on the host, and this is a question that should be considered in the future to study novel tuberculosis vaccines. My question is whether the authors could provide some data of BCG persistence in the local draining lymph nodes (in addition to lungs and spleen), as they are the primary sites of BCG biodistribution following subcutaneous/intradermal vaccination. This is an important point, and was also raised by Dr Gutierrez, the other reviewer. As you will see from our response to Dr Gutierrez’ comments, we unfortunately do not have the samples available and no additional funding is available at the moment to repeat this study. Furthermore, we believe it would be difficult to justify the use of further animals solely for this purpose; however, we are very interested to investigate the role of BCG distribution in vaccine efficacy in more detail in the future, and thank you for the suggestion."
}
]
},
{
"id": "32475",
"date": "20 Apr 2018",
"name": "Maximiliano G Gutierrez",
"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 from Zelmer and collaborators is very interesting, timely, well executed and clearly written. I anticipate a great interest in the tuberculosis vaccine field.\nI have only minor comments and suggestions:\n\nRegarding introduction and/or discussion, it would be important the authors discuss what are the main genetic differences between the 4 mouse strains selected for this study.\n\nIn Figure 4, it is right to state that colonies vary in tissues but only 2 tissues were analysed here so it would be more appropriate to state \"between lung and spleen\". Moreover, would it be possible to get the data in lymph nodes as well? Lymph nodes are critical for the responses to vaccination.\n\nThe variation in CFU/organ seem to be higher in animals that have high levels of protection in this study (e.g. A/J mice). Could the authors comment on this? Is it possible to know the ML ratio in the 2 mice showing high CFU/lung from Fig 4A?\n\nNot sure if the authors have the samples available, but it would have been interesting to see the profiles of Type I/II IFN in tissue at least by qPCR. This information could be of great interest for the tuberculosis field.\n\nThe authors should discuss the limitations of a relatively low number of biological experiments in this work.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3744",
"date": "27 Jun 2018",
"name": "Andrea Zelmer",
"role": "Author Response",
"response": "Dear Dr Gutierrez We would like to thank you for taking the time to review our manuscript, and for the thoughtful feedback provided. We address your individual comments below. The work from Zelmer and collaborators is very interesting, timely, well executed and clearly written. I anticipate a great interest in the tuberculosis vaccine field. I have only minor comments and suggestions: Regarding introduction and/or discussion, it would be important the authors discuss what are the main genetic differences between the 4 mouse strainsselected for this study. We have now included a paragraph in the Discussion to address this. In Figure 4, it is right to state that colonies vary in tissues but only 2 tissues were analysed here so it would be more appropriate to state \"between lung and spleen\". Moreover, would it be possible to get the data in lymph nodes as well? Lymph nodes are critical for the responses to vaccination. We have now specified “lung and spleen” in the legend to Fig. 4 and also in the sub-heading in the results section. It would be very interesting to look at tissue distribution to lymph nodes, and this is a point both reviewers have brought up. Unfortunately, we do not have the samples available and no additional funding is available at the moment to repeat this study. Furthermore, we believe it would be difficult to justify the use of further animals solely for this purpose; however, we will investigate this in future experiments and thank you for the suggestion. The variation in CFU/organ seem to be higher in animals that have high levels of protection in this study (e.g. A/J mice). Could the authors comment on this? Is it possible to know the ML ratio in the 2 mice showing high CFU/lung from Fig 4A? Since a different set of animals was used for flow cytometry and BCG distribution studies (outlined in the Methods section), it is not possible to directly compare the ML ratio and BCG CFU of individual animals. With regards to variation in Fig. 4, we believe this is likely to be a technical effect – we detected very few bacilli in general, and especially in the lung, we did not detect any bacteria in some samples from lesser protected animals. This does not mean there are non – it just means that we were not able to detect them with the method used. As such, it is difficult to interpret the real variation within those mouse strains. Not sure if the authors have the samples available, but it would have been interesting to see the profiles of Type I/II IFN in tissue at least by qPCR. This information could be of great interest for the tuberculosis field. Unfortunately we do not have those samples available. It is an important point, and would indeed be very relevant. We plan to investigate this in future studies. The authors should discuss the limitations of a relatively low number of biological experiments in this work. We have added a comment to this effect at the end of the results section."
}
]
}
] | 1
|
https://f1000research.com/articles/7-296
|
https://f1000research.com/articles/6-2173/v1
|
22 Dec 17
|
{
"type": "Research Article",
"title": "Transcriptional responses of Anopheles gambiae s.s mosquito larvae to chronic exposure of cadmium heavy metal",
"authors": [
"Catherine N. Muturi",
"Martin K. Rono",
"Daniel K. Masiga",
"Francis N. Wachira",
"Richard Ochieng",
"Paul O. Mireji",
"Martin K. Rono",
"Daniel K. Masiga",
"Francis N. Wachira",
"Richard Ochieng",
"Paul O. Mireji"
],
"abstract": "Background: Anopheles gambiae larvae traditionally thrive in non-polluted environments. We previously documented the presence of the larvae in heavy metal polluted urban aquatic environments and the associated biological cost. The goal of this study was to unravel the molecular dynamics involved in the adaptation of the mosquitoes to the heavy metals. Methods: Total RNA was extracted from third instar larvae of both cadmium treated populations and untreated control populations. The RNA concentrations were normalized and complementary DNAs were prepared. Then annealing control primer (ACP) technology was applied to establish transcriptional responses in An. gambiae larvae following several generational (n=90) chronic exposures to cadmium. Differentially expressed genes were determined by their differential banding patterns on an agarose gel. Gel extraction and purification was then carried out on the DEGs and these were later cloned and sequenced to establish the specific transcripts.\n\nResults: We identified 14 differentially expressed transcripts in response to the cadmium exposure in the larvae. Most (11) of the transcripts were up-regulated in response to the cadmium exposure and were putatively functionally associated with metabolism, transport and protein synthesis processes. The transcripts included ATP-binding cassette transporter, eupolytin, ribosomal RNA, translation initiation factor, THO complex, lysosomal alpha-mannosidase, sodium-independent sulfate anion transporter and myotubularin related protein 2. The down-regulated transcripts were functionally associated with signal transduction and proteolytic activity and included Protein G12, adenylate cyclase and endoplasmic reticulum metallopeptidase. Conclusions: Our findings shed light on pathways functionally associated with the adaptation to heavy metals that can be targeted in integrated vector control programs, and potential An. gambiae larvae biomarkers for assessment of environmental stress or contamination.",
"keywords": [
"Anopheles gambiae larvae",
"differentially expressed genes",
"cadmium",
"heavy metal tolerance"
],
"content": "Introduction\n\nHeavy metal pollution has become a global environmental problem and severely threatens biological diversity and human health. Our studies on adaptation to heavy metals have documented presence of the mosquitoes in polluted habitats (Mireji et al., 2008) with growing evidence that this adaptation comes at a biological cost to the mosquito (Mireji et al., 2010b). Similar biological costs to adaptations have also been observed elsewhere in Culex pipiens L responses to cadmium, copper, lead and mercury (El-Sheikh et al., 2010). To date, molecular dynamics underpinning heavy metal tolerance in insects have been tied to transcripts and genes associated functionally with immunity (Sorvari et al., 2007) and defense and repair mechanisms such as glutathione transferases and heat shock proteins (Liao & Freedman, 1998; Kim et al., 2000; Stohs et al., 2001). We have previously putatively implicated metallothioneins, alpha-tubulin and cytochrome p450 genes associated with defense, repair and pyrethroid metabolism mechanisms in insects with heavy metal tolerance, using single gene assessment approaches with Anopheles gambiae mosquito larvae (Mireji et al., 2010b; Mireji et al., 2006; Musasia et al., 2013). Here, we have emulated ab initio relatively higher throughput annealing control primer (ACP) transcriptional profiling, to identify:\n\n1) Pathways functionally associated with heavy metal adaptation observed in the field and their associated biological costs (Mireji et al., 2008; Mireji et al., 2010b); and\n\n2) Potential An. gambiae larvae biomarkers that can be applied for assessment of environmental stress or contamination.\n\n\nMethods\n\nAnopheles gambiae s.s mosquitoes that had been collected from the Mbita field station (00.025’S, 34.013’E), Homa Bay County in Kenya were used for the study. The colony was kept in the Animal Rearing and Quarantine Unit (ARQU) at the International Centre of Insect Physiology and Ecology (ICIPE), Nairobi, Kenya. Larval stages of Anopheles gambiae s.s. were selected for tolerance to cadmium heavy metal through chronic exposures of Maximun Acceptable Toxicant Concentration (MATC) that had been empirically determined (Mireji et al., 2010a). Cadmium metal tolerant strains and control (untreated) strains of the mosquito were raised separately and in triplicates. All subsequent generations of the mosquito were subjected to chronic exposures of cadmium metal as described in Mireji et al., (2010a). Standard Operating Procedure (SOP) for the rearing of Anopheles mosquitoes was followed for colony maintenance (Ford & Green, 1972). Cadmium used in our study was applied as Cadmium Chloride (CdCl2) 99.99% pure (Fisher Scientific LLC, Fair Lawn, NJ, U.S.A).\n\nTotal RNA was extracted from the third instar larvae of experimental and control An. gambiae populations using Trizol (Invitrogen). Quantification of the extracted RNA was done using the micro-spectrophotometer Genequant pro (Amersham Pharmacia Ltd., Bucks, UK). In addition, DNaseI digestion was carried out to remove any residual DNA that could present in the extracted RNA. Total RNA that was isolated and stored at -80°C.\n\nThe total RNA extracted from experimental and control An. gambiae populations were normalized to same concentrations and directly used for the synthesis of first strand complementary DNA (cDNA) using reverse transcriptase (Hwang et al., 2003). Reverse transcription was carried out in a final reaction volume of 20µl containing 2µg of the purified mRNA at 42°C for 1.5 hours. The components of the reaction were: 4µl of 5X reaction buffer (Promega, Madison, WI, U.S.A), 2µl of 10µmol cDNA synthesis dT-ACP 1 primer (5’- CGTGAATGCTGCGACTACGATIIIII(T)18-3’), 5µl dNTPS- 2mM each, 0.5µl RNase inhibitor(40U/µl, Promega) and 1µl Moloney murine leukemia virus reverse transcriptase (200U/µl, Promega). The synthesized first strand cDNA was diluted by adding 80µl ultra-purified water. Storage was at -20°C awaiting PCR procedure.\n\nAnnealing control primer based PCR using the GeneFishing TM DEG kit from Seegene, Seoul, South Korea (Kim et al., 2004), was used to determine differentially expressed genes in the heavy metal treated group and the control population.\n\nSynthesis of the second strand cDNA and PCR was carried out in a single tube. The second strand was synthesized in one cycle of first stage PCR at 50°C, in a final reaction volume of 20µl. The components in the reaction tubes included 3–5µl of diluted first strand cDNA, 1µl 10Mm dT-ACP2 reverse primer (5’-CTGTGAATGCTGCGACTACGATIIIII(T)15-3’), 10µl 2x master mix (Seegene, Seoul, South Korea) and 1µl 10µM arbitrary ACP (forward primer).\n\nPCR procedures for the synthesis of the second strand were completed in one cycle, at 94°C for 1 min then 50°C for 3min and 72°C for 1 min.\n\nThe second stage of the PCR protocol consisted of 40 cycles at 94°C for 40s, 65°C for 40s, 72°C for 40s and the final extension for 10 min at 72°C. 2% agarose gel electrophoresis with ethidium bromide staining was used for separation of the PCR products.\n\nDifferentially expressed bands in the control and cadmium exposed population subjected to the same primer set were excised from the agarose gel using a scapel under Ultra Violet illumination. The gel slices were then purified using the QIAquick® gel extraction kit (QIAGEN, Inc., Valencia, CA), following the instructions from the manufacturer.\n\nGel-purified PCR products were directly cloned into a pGEMT Easy vector (Invitrogen, Carlsbad, CA, USA), using JM109 competent cells. Colonies were grown at 37°C for 18 hours on Luria broth agar plates, containing ampicillin, X-gal and IPTG for blue/white colony screening. Cloned plasmids were then purified using the GeneJET™ Miniprep kit (Fermentus, Thermo Fisher Scientific Inc.), as per the instructions from the manufacturer.\n\nSequencing was done with ABI PRISM® 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using M13 primers. The sequences were edited using VecScreen and BioEdit software. Edited sequences were analyzed by searching for similarities in VectorBase against the Anopheles gambiae PEST strain transcripts sequences, AgamP4.6 geneset using the BLASTn search program (Altschul et al., 1990)\n\n\nResults\n\nWe successfully implemented the ACP system to identify differentially expressed genes (DEGs) in larvae chronically exposed to cadmium, as previously demonstrated in blastocyst experiments (Cui et al., 2005; Hwang et al., 2004; Hwang et al., 2005). Our differential banding patterns of the cDNA representation of DEGs is summarized in Figure 1. Fourteen DEGs were identified after chronic exposure of An. gambiae larvae to cadmium heavy metal (Table 1). Most (11) of the differentially expressed genes were induced in cadmium exposed relative to the cadmium un-exposed controls. Our BLAST (REF) results revealed that the cadmium induced transcripts were clustered into metabolism (AGAP008584-RA, AGAP001249-RA and AGAP009563-RA), transport (AGAP012302-RA and AGAP002638-RA) and protein synthesis (AGAP028915-RA, AGAP004750-RA, AGAP028391-RA, AGAP003870-RA, AGAP028907-RA, AGAP028818-RA and AGAP028899-RA) processes.\n\nThe arrows indicate the DEGs observed using ACP 75, ACP 76 and ACP 78 primer set. Number 1 represents Cadmium population while 2 represents control population. M= 50bp molecular marker. High intensity of a band represents an up-regulation of a particular gene in cadmium or control population.\n\nSequence data obtained was blasted against Anopheles gambiae PEST strain transcript sequences, AgamP4.6 geneset in May 2017.\n\nThree of the DEGs identified were suppressed in the cadmium exposed larvae and these included AGAP006187-RA, AGAP002262-RA and AGAP003078-RA.\n\n\nDiscussion\n\nWe identified ATP-binding cassette transporters belonging to the superfamily of membrane proteins that are present in all living organisms (Dean & Annilo, 2005). They are normally associated with movement of substrates such as amino acids, peptides, sugars, metals, inorganic ions, lipids, lipopolysaccharides and xenobiotics across biological membranes (Dawson & Locher, 2006; Hollenstein et al., 2007a). The ABC transporters have been shown to affect development, metabolism and insecticide resistance in insects (Borycz et al., 2008; Dow & Davies, 2006; Ricardo & Lehmann, 2009; Vache et al., 2007). The silencing of the ABCH1 gene has been shown to result in the death of larvae and pupae (Guo et al., 2015). Therefore, induction of the ABC transporters may suggest that they are involved in cadmium transport through membranes to reduce toxicity of cadmium metal to the larvae in their environment.\n\nThe induction of the eupolytin gene may have a role in the activation of defense molecules. In a study involving the ground beetle Eupolyphaga sinensis, eupolytin-1 gene encoding a protease was shown to be involved in the activation of plasminogen and hydrolysis of fibrinogen (Yang et al., 2011).\n\nRibosomal genes are involved in protein synthesis and upregulation of the various arrays of ribosomal RNAs in this study, which suggests their role in enhancing the survival of An. gambiae in the heavy metal polluted environment by the transcription and translation of genes which are important in the adaptation of the larvae to xenobiotics.\n\nThe nuclear structure referred to as THO complex is usually conserved in all kingdoms, and it has an important role in the packing of pre-mRNA molecules into RNA-protein assemblies which are termed mRNPs (Köhler & Hurt, 2007). A study carried out recently has shown that the THO complex is required for efficient expression of some genes, ensuring genetic stability thereby preventing transcription-associated recombination (Gewartowski et al., 2012). The expression of the THO complex is suggestive of its role in expressing defense genes that enhance survival of larvae in a Cadmium polluted environment.\n\nSuppression of AGAP006187-RA, AGAP002262-RA and AGAP003078-RA transcripts that included G- Proteins couple receptors to adenylyl cyclase stimulation indicated increasing levels of cAMP and a cascade of events that constitute the signal transduction pathway that drive cellular responses. Metallopeptidases are a ubiquitous and diverse group of enzymes containing both endopeptidases and exopeptidases. Although they vary widely at the sequence, structural, and functional levels, they all require a metal ion for catalytic activity (Rawlings & Salvesen, 2013). The suppression of these important genes involved in signal transduction and proteolytic activity would account for the larval mortality rates that are usually observed in larvae raised in the cadmium heavy metal environment.\n\nOur findings shed light on the adaptation of the An. gambiae mosquito to heavy metals by differentially expressing particular genes in response to a toxicant impact. A study to determine differentially expressed genes in cadmium-exposed sebastes schlegeli unraveled genes related to pathogenesis, extrinsic stresses, and catalytic metabolites (Woo & Yum, 2014). Previous studies have indicated that metallothionein and α-tubulin genes that are present in insects can be used as potential biomarkers (Hare, 1992; Klerks & Weis, 1987; Mattingly et al., 2001; Roesijadi, 1994). Metallothionein was assessed through C. quinquefasciatus mosquito larvae for Copper, Cadmium and Zinc aquatic environmental levels (Sarkar et al., 2004). Therefore, the genes identified might be useful in the development of potential biomarkers that can be used to assess the level of environmental pollution of heavy metal origin in An. gambiae mosquitoes.\n\n\nConclusions\n\nWe were able to identify genes that are differentially expressed due to chronic exposure of An. gambiae larvae to cadmium metal using the ACP-based PCR method. However, application of more sensitive techniques like those used in proteomics would generate more data.\n\n\nData availability\n\nDataset 1: Sequence data obtained after sequence analysis using the BioEdit software. The sequences were subsequently taken through a BLAST search. The results of the sequence analysis are shown on the manuscript. DOI, 10.5256/f1000research.13062.d187045 (Muturi et al., 2017a).\n\nDataset 2: Sample of the colony PCR experiment. The gel photo of a colony PCR of 20 samples that was carried out after blue/white colony screening using M13 primers. DOI, 10.5256/f1000research.13062.d187046 (Muturi et al., 2017b).",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nFunding for this study was provided by the Department of Research and Extension, Egerton University and the DAAD in-country Scholarship.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe hereby wish to acknowledge the following individuals for their contribution to this work:\n\nThe Head of the Capacity Building Department at ICIPE, for granting us permission to carry out this work in their Molecular and Biotechnology unit.\n\nThe Director of the Research and Extension Department at Egerton University.\n\nThe DAAD team for the financial support, which enabled this work to be completed.\n\n\nReferences\n\nAltschul SF, Gish W, Miller W, et al.: Basic local alignment search tool. J Mol Biol. 1990; 215(3): 403–410. PubMed Abstract | Publisher Full Text\n\nBorycz J, Borycz JA, Kubow A, et al.: Drosophila ABC transporter mutants white, brown and scarlet have altered contents and distribution of biogenic amines in the brain. J Exp Biol. 2008; 211(Pt 21): 3454–3466. PubMed Abstract | Publisher Full Text\n\nCui XS, Shin MR, Lee KA, et al.: Identification of differentially expressed genes in murine embryos at the blastocyst stage using annealing control primer system. Mol Reprod Dev. 2005; 70(3): 278–287. PubMed Abstract | Publisher Full Text\n\nDawson RJ, Locher KP: Structure of a bacterial multidrug ABC transporter. Nature. 2006; 443(7108): 180–185. PubMed Abstract | Publisher Full Text\n\nDean M, Annilo T: Evolution of the ATP-binding cassette (ABC) transporter superfamily in vertebrates. Annu Rev Genomics Hum Genet. 2005; 6: 123–142. PubMed Abstract | Publisher Full Text\n\nDow JA, Davies SA: The Malpighian tubule: rapid insights from post-genomic biology. J Insect Physiol. 2006; 52(4): 365–378. PubMed Abstract | Publisher Full Text\n\nEl-Sheikh TMY, Fouda MA, Hassan MI, et al.: Toxicological Effects of Some Heavy Metal Ions on Culex pipiens L. (Diptera: Culicidae). Acad J biolog Sci. 2010; 2(1): 63–76. Reference Source\n\nFord HR, Green E: Laboratory rearing of Anopheles albimanus. Mosq News. 1972; 32: 509–513. Reference Source\n\nGewartowski K, Cuéllar J, Dziembowski A, et al.: The yeast THO complex forms a 5-subunit assembly that directly interacts with active chromatin. Bioarchitecture. 2012; 2(4): 134–137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo Z, Kang S, Zhu X, et al.: The novel ABC transporter ABCH1 is a potential target for RNAi-based insect pest control and resistance management. Sci Rep. 2015; 5: 13728. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHare L: Aquatic insects and trace metals: bioavailability, bioaccumulation, and toxicity. Crit Rev Toxicol. 1992; 22(5–6): 327–369. PubMed Abstract | Publisher Full Text\n\nHollenstein K, Frei DC, Locher KP: Structure of an ABC transporter in complex with its binding protein. Nature. 2007a; 446(7132): 213–216. PubMed Abstract | Publisher Full Text\n\nHwang KC, Cui XS, Park SP, et al.: Identification of differentially regulated genes in bovine blastocysts using an annealing control primer system. Mol Reprod Dev. 2004; 69(1): 43–51. PubMed Abstract | Publisher Full Text\n\nHwang IT, Kim YJ, Kim SH, et al.: Annealing control primer system for improving specificity of PCR amplification. Biotechniques. 2003; 35(6): 1180–1184. PubMed Abstract\n\nHwang KC, Lee HY, Cui XS, et al.: Identification of maternal mRNAs in porcine parthenotes at the 2-cell stage: a comparison with the blastocyst stage. Mol Reprod Dev. 2005; 70(3): 314–323. PubMed Abstract | Publisher Full Text\n\nKim KA, Chakraborti T, Goldstein GW, et al.: Immediate early gene expression in PC12 cells exposed to lead: Requirement for protein kinase C. J Neurochem. 2000; 74(3): 1140–1146. PubMed Abstract | Publisher Full Text\n\nKim YJ, Kwak CI, Gu YY, et al.: Annealing control primer system for identification of differentially expressed genes on agarose gels. Biotechniques. 2004; 36(3): 424–6, 428, 430 passim. PubMed Abstract\n\nKlerks PL, Weis JS: Genetic adaptation to heavy metals in aquatic organisms: a review. Environ Pollut. 1987; 45(3): 173–205. PubMed Abstract | Publisher Full Text\n\nKöhler A, Hurt E: Exporting RNA from the nucleus to the cytoplasm. Nat Rev Mol Cell Biol. 2007; 8(10): 761–73. PubMed Abstract | Publisher Full Text\n\nLiao VH, Freedman JH: Cadmium-regulated genes from the nematode Caenorhabditis elegans. Identification and cloning of new cadmium-responsive genes by differential display. J Biol Chem. 1998; 273(48): 31962–31970. PubMed Abstract | Publisher Full Text\n\nMattingly KS, Beaty BJ, Mackie RS, et al.: Molecular cloning and characterization of a metal responsive Chironomus tentans alpha-tubulin cDNA. Aquat Toxicol. 2001; 54(3–4): 249–260. PubMed Abstract | Publisher Full Text\n\nMireji PO, Keating J, Hassanali A, et al.: Biological cost of tolerance to heavy metals in the mosquito Anopheles gambiae. Med Vet Entomol. 2010b; 24(2): 101–107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMireji PO, Keating J, Hassanali A, et al.: Expression of metallothionein and alpha-tubulin in heavy metal-tolerant Anopheles gambiae sensu stricto (Diptera: Culicidae). Ecotoxicol Environ Saf. 2010a; 73(1): 46–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMireji PO, Keating J, Hassanali A, et al.: Heavy metals in mosquito larval habitats in urban Kisumu and Malindi, Kenya, and their impact. Ecotoxicol Environ Saf. 2008; 70(1): 147–153. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMireji PO, Keating J, Kenya E, et al.: Differential Induction of Proteins in Anopheles gambiae sensu stricto (Diptera: Cullicidae) Larvae in Response to Heavy Metal Selection. Int J Trop Insect Sci. 2006; 26(4): 214–226. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMusasia FK, Isaac AO, Masiga DK, et al.: Sex-specific induction of CYP6 cytochrome P450 genes in cadmium and lead tolerant Anopheles gambiae. Malar J. 2013; 12: 97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuturi CN, Rono MK, Masiga DK, et al.: Dataset 1 in: Transcriptional responses of Anopheles gambiae s.s mosquito larvae to chronic exposure of cadmium heavy metal. F1000Research. 2017a. Data Source\n\nMuturi CN, Rono MK, Masiga DK, et al.: Dataset 2 in: Transcriptional responses of Anopheles gambiae s.s mosquito larvae to chronic exposure of cadmium heavy metal. F1000Research. 2017b. Data Source\n\nRawlings ND, Salvesen GS: Handbook of Proteolytic Enzymes. Elsevier, San Diego, Calif, USA. 2013. Reference Source\n\nRicardo S, Lehmann R: An ABC transporter controls export of a Drosophila germ cell attractant. Science. 2009; 323(5916): 943–946. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoesijadi G: Metallothionein induction as a measure of response to metal exposure in aquatic animals. Environ Health Perspect. 1994; 102(Suppl 12): 91–95. PubMed Abstract | Free Full Text\n\nSarkar S, Duttagupta AK, Mal TK: Effects of heavy metals on population growth and metallothionein gene expression in the mosquito Culex quinquefasciatus, from Calcutta, India. Environ Pollut. 2004; 127(2): 183–193. PubMed Abstract | Publisher Full Text\n\nSorvari J, Rantala LM, Rantala MJ, et al.: Heavy metal pollution disturbs immune response in wild ant populations. Environ Pollut. 2007; 145(1): 324–328. PubMed Abstract | Publisher Full Text\n\nStohs SJ, Bagchi D, Hassoun E, et al.: Oxidative mechanisms in the toxicity of chromium and cadmium ions. J Environ Pathol Toxicol Oncol. 2001; 20(2): 77–88. PubMed Abstract | Publisher Full Text\n\nVache C, Camares O, Cardoso-Ferreira MC, et al.: A potential genomic biomarker for the detection of polycyclic aromatic hydrocarbon pollutants: multidrug resistance gene 49 in Drosophila melanogaster. Environ Toxicol Chem. 2007; 26(7): 1418–1424. PubMed Abstract | Publisher Full Text\n\nWoo S, Yum S: Differentially Expressed Genes in Cadmium-Exposed Sebastes Schlegeli Using Dd-Pcr. App Sci Report. 2014; 6(2): 62–66. Reference Source\n\nYang H, Wang Y, Xiao Y, et al.: A bi-functional anti-thrombosis protein containing both direct-acting fibrin(ogen)olytic and plasminogen-activating activities. PLoS One. 2011; 6(3): e17519. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "29393",
"date": "29 Jan 2018",
"name": "David Essumang",
"expertise": [
"Reviewer Expertise Environmental Scientist"
],
"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 main work is outside my expertise and I was struggling to make some meaningful contribution. The bulk of the work is in molecular experimentation and I have limited knowledge in their methods. However, my major concern with the entomological aspect is that the authors did not show how long (the number of generations) the mosquitoes were exposed to the heavy metal during the selection. This would be helpful if someone wants to repeat the work.\n\nFurthermore, the study did not show the impact of the biomarkers identified in the development, enzyme activities and insecticide resistance of the mosquitoes used for the study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3397",
"date": "02 Feb 2018",
"name": "Catherine Ngambi",
"role": "Author Response",
"response": "Thanks so much for you review on my article. I appreciate. I wanted to respond to the question raised about the number of generations that mosquitoes were exposed to cadmium heavy metal. As stated in the Abstract its for 90 generations. Kindest regards, Catherine."
}
]
},
{
"id": "34199",
"date": "29 May 2018",
"name": "Shüné V. Oliver",
"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 manuscript is technically sound, and presents a worthwhile body of research. I would recommend accepting. However, there is one major correction that must be made before accepting. Please give details about the assessment of RNA integrity. Without this, the other results are not valid, so please detail about your confirmation of RNA integrity post extraction.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3730",
"date": "11 Jun 2018",
"name": "Catherine Ngambi",
"role": "Author Response",
"response": "Thanks so much for your positive review of the manuscript. I am sorry, I had not given the data about the RNA integrity but had only stated that I had quantified the extracted RNA using the microspectrophotometer genequant pro. I have revised the manuscript and have indicated the quantification results and the equipments used. To estimate RNA purity, the ratio of the Absorbance contributed by the RNA to the Absorbance of the contaminants is calculated. The acceptable ratios for purity or typical requirements for A260/A280 ratios are 1.8-2.2. The samples that were use for the downstream application were those that met the purity criteria of 1.8-2.2, that is the ratios for A260/A280. Therefore, the revised document now contains these details. Kindest Regards, Catherine."
}
]
}
] | 1
|
https://f1000research.com/articles/6-2173
|
https://f1000research.com/articles/6-1452/v1
|
14 Aug 17
|
{
"type": "Research Article",
"title": "Study of oral aniracetam in C57BL/6J mice without pre-existing cognitive impairments",
"authors": [
"Conner D. Reynolds",
"Taylor S. Jefferson",
"Meagan Volquardsen",
"Ashvini Pandian",
"Gregory D. Smith",
"Andrew J. Holley",
"Joaquin N. Lugo",
"Conner D. Reynolds",
"Taylor S. Jefferson",
"Meagan Volquardsen",
"Ashvini Pandian",
"Gregory D. Smith",
"Andrew J. Holley"
],
"abstract": "Background: The piracetam analog, aniracetam, has recently received attention for its cognition enhancing potential, with minimal reported side effects. Previous studies report the drug to be effective in both human and non-human models with pre-existing cognitive dysfunction, but few studies have evaluated its efficacy in healthy subjects. A previous study performed in our laboratory found no cognitive enhancing effects of oral aniracetam administration 1-hour prior to behavioral testing in naïve C57BL/6J mice. Methods: The current study aims to further evaluate this drug by administration of aniracetam 30 minutes prior to testing in order to optimize any cognitive enhancing effects. In this study, all naïve C57BL/6J mice were tested in tasks of delayed fear conditioning, novel object recognition, rotarod, open field, elevated plus maze, and marble burying. Results: Across all tasks, animals in the treatment group failed to show enhanced learning when compared to controls. Conclusions: These results provide further evidence suggesting that aniracetam conveys no therapeutic benefit to subjects without pre-existing cognitive dysfunction.",
"keywords": [
"nootropic",
"aniracetam",
"learning",
"memory"
],
"content": "Introduction\n\nIn the 1970s, pharmacologist Cornelius Giurgea coined the term nootropics to describe a novel group of compounds capable of enhancing cognitive processes, intersynaptic communication, and the exchange of information between cerebral hemispheres. These compounds can be divided into five primary categories: cholinergic agonists, psychostimulants, piracetam compounds, hormones & essential nutrients, and agonists of cerebral blood flow1. Initial interest in these compounds was limited to reversing the cognitive impairments in subjects with neurological damage or age-related decline. This investigation led to the development of a variety of neuroenhancing compounds, showing promise for cognitive restoration following epilepsy2, traumatic brain injury3, cerebral vascular accident4, Alzheimer’s disease5, and dementia6. Nootropics have also been investigated in the treatment of many neurodevelopmental disorders, such as autism7, ADHD8, and schizophrenia9.\n\nRecently, there has been increasing prevalence of nootropic use among otherwise healthy subjects aiming to enhance academic performance, particularly college populations. According to recent population-based studies, the overall incidence of non-medicinal prescription psychostimulant use within the college student population is 4.1%–10.8% over the past year, and 6.4%–19.6% during their lifetime10–14. However, misuse of these medications can be dangerous, as psychostimulant toxicity has been linked to cardiac dysrhythmia, myocardial infarction, psychosis, and sudden death15,16.\n\nThe piracetam analog, aniracetam, has recently received attention due to its potential for cognitive enhancement associated with minimal reported side effects17. In previous studies, aniracetam has been shown to enhance excitatory post synaptic potentials18, reduce glutamatergic receptor desensitization18, increase EPSC decay time19, and augment long-term potentiation in the hippocampus20. Although the definitive mechanism of this compound is unclear, some evidence suggests that it acts as a reversible positive allosteric modulator of AMPA receptors21. Aniracetam has proven effective in both human22 and non-human23–28 models of cognitive dysfunction. However, few studies have evaluated its efficacy in healthy subjects without cognitive impairment.\n\nIn a previous study, our laboratory evaluated whether daily oral administration of aniracetam (50 mg/kg) 1-hour prior to testing could improve cognitive performance in naïve C57BL/6J mice29. Through a series of behavioral tasks, we observed that aniracetam did not improve spatial learning, fear learning, or motor learning. Further investigation of aniracetam pharmacokinetics suggested that peak serum levels are achieved approximately 30 minutes following oral administration30. In light of this evidence, the current study aims to further evaluate aniracetam’s effects by administering aniracetam 30 minutes prior to testing, in order to optimize any cognitive enhancing effects. If aniracetam is truly a cognitive enhancer, we hypothesized that treated mice would display significantly greater learning and memory compared to controls.\n\n\nMaterials and methods\n\nTwenty-four C57BL/6J male mice were generated at Baylor University for use in this study. The strain was originally purchased from Jackson Labs and bred at Baylor University. All mice were independently housed in a vivarium, where environmental conditions were controlled to an ambient temperature of 22°C with 12-hour light/12-hour dark diurnal cycles. All mice were also given ad libitum access to food and water. No health concerns were found during the courses of the experiments in this study. There were no adverse effects on the mice during the studies, and every effort was made to ameliorate any discomfort.\n\nAfter reaching approximately 2 months of age, all mice were randomized to receive either one dose of aniracetam (100mg/kg) (1-[4-methoxybenzoyl)]-2-pyrrolidinone) (Shanghai Suyong Biotechnologies Inc., China), or an identical placebo by oral administration in a gelatin-based suspension 30 minutes prior to behavioral testing. Aniracetam or placebo was administered prior to each behavioral test. This route of administration was selected in order to mimic the typical mode of aniracetam consumption used in humans. During the double blind phase, all mice were subjected to a battery of behavioral tests by designated experimenters blinded to treatment group assignments. All procedures were conducted in compliance with Baylor University Institutional Animal Care and Use Committee, as well as the National Institute of Health Guidelines for the Care and Use of Laboratory Animals. All protocols were approved by the Baylor University Animal Care and Use Committee (Animal Assurance Number A3948-01).\n\nA two-day delayed fear conditioning protocol was used to assess amygdala-dependent learning. For this procedure, we used a 26cm x 22cm x 18cm operant chamber, composed of two clear acrylic sides and two metal sides, a metal grid floor capable of receiving an electric shock, an interior light providing constant luminescence (2 lux), and a speaker. The operant chamber was then placed inside of a sound attenuated isolation cubicle (Coulburn Instruments, Allentown, PA, USA) in order to control for external light and sound contamination. During all phases of this task, learning was assessed by the degree of freezing, as it is the most reliable measure of fear memory in mice. All testing was recorded and measured by automatic video tracking software, with visual confirmation of conditioned stimulus (CS) and unconditioned stimulus (US) presentations by the designated experimenter.\n\nFor this procedure, we used a 40cm x 40cm x30cm clear acrylic open top box. This task was performed in an isolated room controlled for light levels, temperature, and background noise. Prior to testing, all mice were habituated to the arena without any objects for 20 minutes. During the first phase of testing, the two identical objects were placed on opposite sides of the apparatus and interactions with each were measured over a 10-minute period. During the second phase of testing, both objects were removed and replaced with the original object and a novel object. These were placed on opposite sides of the arena and interactions with each were measured over a 10-minute period. All trials were video recorded and manually scored by the designated experimenter after testing.\n\nThe rotarod task was used to assess cerebellar motor coordination and learning. For this procedure, we used a rotating rod (Series 8 Rotarod; IITC Inc., Woodland Hills, CA, USA) which gradually accelerated from 5rpm to 40rpm. All mice were subjected to two 5 minute trials, with a 60 minute ITI, across 4 days of testing. The designated experimenter was responsible for monitoring and recording the length of time in which mice could hold onto the rotating rod before falling. This task was performed in an isolated room controlled for light levels, temperature, and background noise.\n\nThe open field task was used to assess locomotion and anxiety. For this procedure, we used a 40cm x 40cm x 30cm clear acrylic box. This task was performed in an isolated room controlling for light levels, temperature, and background noise. All mice were placed in the center of the apparatus and allowed to explore for 10 minutes. Total time spent in the inner and outer regions were recorded and measured via Fusion optical recording system. Time spent in the outer and inner regions of the field was examined. A greater amount of time spent in the outer region is associated with anxious behavior.\n\nThe elevated plus maze task was used to assess levels of anxiety. For this procedure, we used a maze constructed of four white acrylic arms raised 40cm from the floor. All arms were 30cm long x 5cm wide. Two opposing arms were enclosed (walls, 15cm tall) and two opposing arms were left open. During this task, mice were placed in an open arm near the center of the maze and were allowed to explore for 10 minutes. Total distance and time spent in open versus closed arms was recorded by Noldus motion-tracking software (Ethovision, Netherlands). Video recordings were also manually scored by designated experimenters for additional behavioral observations, such as number of rearings in the open versus closed arms and number of head dips in the open arms. A greater amount of time spent in the closed arms versus open arms indicates higher levels of anxiety.\n\nThe marble burying task was used to examine repetitive behavior. For this procedure clean home cages were filled with approximately 2–3cm of bedding and twenty black glass marbles were assembled into four evenly spaced columns of five rows. All mice were then placed into the testing cage in front of the array of marbles for 30 minutes. Several measurements of the percentage of the marble buried (50, 75, 100 and completely buried) was recorded by the designated experimenter. The measurement of 100% refers to a marble that is buried to its entire height with some bedding covering, but still in view of the experimenter, while completely buried marbles refers to those not in view of the experimenter. The increased marble burying reflects a higher tendency towards repetitive behavior.\n\nAll behavioral data with a single measurement was analyzed using an independent samples t-test. The Independent samples t-test was used to analyze behavior in the open field, for day 2 of fear conditioning (fear memory), and for novel object recognition. All behavioral data with repeated measures were analyzed using a two-way Analysis of Variance (ANOVA), with experimental group as the independent factor and the trials or block number as the repeated factor. The two-way analyses were performed on rotarod data and on data from day 1 of fear conditioning (acquisition of fear learning). All data were analyzed using SPSS 20.0 (IBM, USA) or GraphPad Prism 7 software (La Jolla, CA). Values are shown as mean ± S.E.M. for each group.\n\n\nResults\n\nIn the elevated plus maze task, we found no significant differences in time spent in the open t(1,22) = 0.63, p = 0.53; center t(1,22) = 0.04, p = 0.23; or closed arms t(1,22) = 0.42, p = 0.67 (Figure 1A). Similar results were found in the frequency of arm entries, with no difference in number of arm entries into the open arms t(1,22) = 0.69, p = 0.49; center t(1,22) = 0.39, p = 0.69; or closed arms t(1,22) = 0.05, p = 0.96 (Figure 1B).\n\n(A) In the elevated plus maze test, an independent measures t-test revealed no significant differences in time spent in the open arms t(1,22) = 0.63, p = 0.53; center t(1,22) = 0.04, p = 0.23; or closed arms t(1,22) = 0.42, p = 0.67. (B) There were also no significant differences in the number of entries into the open arms t(1,22) = 0.69, p = 0.49; center area t(1,22) = 0.39, p = 0.69; or closed arms t(1,22) = 0.05, p = 0.96. (C) In the open field test, an independent measures t-test revealed no significant differences between groups in total distance moved t(1,22) = 0.90, p = 0.37, (D) or stereotypy time t(1,22) = 1.45, p = 0.16.\n\nIn the open field task, we found no significant differences between the groups in total distance moved in the 10 minute trial t(1,22) = 0.90, p = 0.37 (Figure 1C). There were also no significant differences observed in stereotypy time t(1,22) = 1.45, p = 0.16. (Figure 1D) Together, these results suggest that aniracetam has no effect on locomotion or anxiety.\n\nAcross 8 rotarod trials, we did not observe any main effect between groups (F (1,22) = 0.4073, p = 0.5299) (Figure 2). However, there was a main effect of learning across multiple trials (F (7, 154) = 11.97; p < 0.0001), indicating that motor learning had occurred within both groups. These results suggest that aniracetam has no cognitive enhancing effect on motor learning.\n\nAcross 8 trials, there was no main effect between groups (F (1,22) = 0.4073, p = 0.5299). However, there was a main effect of learning across multiple trials (F (7, 154) = 11.97; p<0.0001), indicating that motor learning had occurred within both groups.\n\nIn the marble burying task, we found no significant differences in performance when measured at: 50% t(1,22) = 1.18, p = 0.24; 75% t(1,22) = 0.76, p = 0.45; 100% t(1,22) = 0.50, p = 0.61; or at the completely buried level t(1,22) = 0.05, p = 0.95 (Figure 3). These results suggest that aniracetam has no effect on repetitive behavior.\n\nIndependent measures t-tests revealed no significant differences in the animal’s performance in marble burying when measured at: 50% t(1,22) = 1.18, p = 0.24; 75% t(1,22) = 0.76, p = 0.45; 100% t(1,22) = 0.50, p = 0.61; or at the level of completely buried t(1,22) = 0.05, p = 0.95.\n\nOn the first day of fear conditioning, mice were placed into an operant chamber where multiple tone and foot shocks were administered. We observed no main effect of group (F (1, 19) = 0.1048; p = 0.7497) or interaction between groups (F (4, 76) = 0.5453; p = 0.7029) (Figure 4A). However, there was a main effect of learning across multiple trials (F (4, 76) = 42.35 p < 0.0001), indicating that fear learning had occurred within both groups. On the second conditioning day mice were placed into the same operant conditioning chamber with novel context. Upon presentation of the tone both groups displayed increased fear, however there was no significant difference in the freezing behavior expressed between treated and control mice t(1,19) = 1.013; p = 0.3238) (Figure 4B). These results suggest that aniracetam treatment has no cognitive enhancing effect on associative fear learning.\n\n(A) On the first day of testing, mice were placed into the operant chamber and 2 minutes of baseline activity levels were recorded. This was followed by a 30 second conditioned stimulus (CS) tone (80dB white noise), a 2 second unconditioned stimulus (US) shock (0.85mA), and a 2-minute inter-trial interval (ITI). Another identical CS-US pairing was then presented and followed with a 30 second ITI. Following multiple foot shock and tone presentations, a two-way ANOVA test indicated no main effect of group (F (1, 19) = 0.1048; p = 0.7497) or interaction between groups (F (4, 76) = 0.5453; p = 0.7029). (B) The second day of testing consisted of two trials. During the first trial, mice were placed back into the original operant chamber for 5 minutes and baseline activity was recorded. Before the second trial the operant chamber was modified with a foam pad under an acrylic square to cover the metal floor grid, an acrylic wall placed diagonally across that halved the space into triangular form, and 1mL of pure vanilla extract (Adam’s Extracts, USA) placed beneath the floor. These changes to the tactile, spatial, and olfactory contexts of the chamber were made in order to prevent context dependent learning. During the second trial, mice were placed into the contextually modified operant chamber and 3 minutes of the trial baseline activity was monitored. This was followed by a 3-minute period of CS tone presentation (80dB white noise). Upon presentation of the tone within novel context, an independent measures t-test indicated no significant difference in the freezing behavior exhibited between groups in novel context t(1,19)=1.013; p=0.3238).\n\nDuring the initial phase of testing, object preference was measured for identical objects. There was no significant preference towards the left or right object between treated t(1,22) = 1.333, p = 0.20 or control group mice t(1,22) = 0.1583, p = 0.88 (Figure 5A). During the second phase of testing, object preference was measured between a familiar and novel object. There was a significant preference towards the novel object in both treated t(1,22) = 4.968, p < 0.0001 and control mice t(1,22) = 3.776, p < 0.001. However, there were no differences in preference between the groups in the novel object condition t(1,22) = 0.6112, p = 0.5474 (Figure 5B). These results suggest that aniracetam treatment has no cognitive enhancing effect on novel object recognition.\n\n(A) Independent measures t-tests indicated no significant preference towards the left or right object between treated t(1,22) = 1.333, p = 0.20 or control group mice t(1,22) = 0.1583, p = 0.88. (B) An independent measures t-test indicated no differences in preference between groups in the novel object condition t(1,22) = 0.6112, p = 0.5474.\n\n\nDiscussion\n\nAlthough significant progress has been made towards understanding the neuroenhancing effects of aniracetam in subjects with cognitive impairment, there has been little investigation into its therapeutic effects on healthy subjects. In a previous study29, our laboratory demonstrated that drug treatment in healthy C57BL/6J mice did not produce any significant effects on learning and memory, anxiety, locomotion, or repetitive behavior. Given the existing body of evidence supporting aniracetam’s cognitive enhancing effects, we elected to investigate this substrate further by using a modified drug treatment schedule to ensure peak serum levels during behavioral testing. Through this follow-up investigation, it was demonstrated that aniracetam conveys no significant cognition enhancing effects in healthy subjects.\n\nOur findings are in contrast to a previous study by Rao et al.26, which demonstrated that intrahippocampal aniracetam infusions significantly improved Y-maze performance in healthy rats. A key difference in experimental design between this study and ours is the route of administration. Intrahippocampal drug infusion provides tightly controlled, localized doses by circumventing first-pass metabolism. This method leads to a more accurate assessment of the drug serum levels necessary to achieve a therapeutic effect, but is restricted specifically to animal studies. Oral drug administration typically has a much lower bioavailablity due to hepatic biotransformation, but provides a higher ecological validity. For the purposes of our study, we elected to administer aniracetam orally, as it closely mimics the route of administration used most commonly in human reports. Although it is theoretically possible that cognitive enhancement could be achieved by aniracetam treatment, the therapeutic dose required to achieve this effect may be unrealistic.\n\nDespite any peer-reviewed data of non-medicinal use in humans, our findings contrast many subjective reports from healthy individuals purporting the cognitive enhancing effects of aniracetam and other piracetam-analogs. In a previous study, Corazza et al.31 performed a multilingual qualitative assessment from a range of available online resources subjectively reporting benefits from piracetam use. These authors found that while the drug is used to improve academic and work-related performance, its use is also associated with side effects, such as hallucinations, dysphoria, fatigue, dizziness, memory loss, and headaches. Their findings indicate these side effects may be dose-dependent; however, because both the drug and their manufacturers are currently unregulated, it is impossible to determine an effective dose or therapeutic index in humans.\n\nTo our knowledge, both the present and previous studies represent the first empirical evidence of aniracetam treatment by oral administration in healthy subjects.\n\nBecause this study closely mimics the drug administration in humans, we can infer that these results should most accurately depict the effects in healthy human subjects. Based on our findings, it can be suggested that non-medicinal and/or recreational use by healthy individuals may have only marginal therapeutic benefit, while the risk of harmful side effects remains.\n\n\nData availability\n\nDataset 1: Raw data for ‘Study of oral aniracetam in C57BL/6J mice without pre-existing cognitive impairments.’ (A) Open field total distance data and stereotypy results for vehicle and aniracetam-treated subjects. (B) Elevated-plus maze mean time and total frequency visits for open, closed, and center arms for vehicle and aniracetam-treated subjects. (C) Marble burying data for marbles buried at 50%, 75%, 100%, and total marbles for vehicle and aniracetam-treated subjects. (D) Delay fear conditioning data for day 1 and day 2 for vehicle and aniracetam-treated subjects. (E) Rotarod data for latency to fall off rotarod for vehicle and aniracetam-treated subjects. (F) Novel object recognition data for phase 1 and phase 2 for vehicle and aniracetam-treated subjects. DOI, 10.5256/f1000research.11023.d17254232",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis research was supported by intramural funds from Baylor University Research Council.\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\nNootropic Drugs: Encyclopedia of Cognitive Science. Wiley. 2005. Reference Source\n\nKitano Y, Komiyama C, Makino M, et al.: Anticonvulsant and neuroprotective effects of the novel nootropic agent nefiracetam on kainic acid-induced seizures in rats. Brain Research. 2005; 1057(1–2): 168–176. PubMed Abstract | Publisher Full Text\n\nChen CC, Wei ST, Tsaia SC, et al.: Cerebrolysin enhances cognitive recovery of mild traumatic brain injury patients: double-blind, placebo-controlled, randomized study. Br J Neurosurg. 2013; 27(6): 803–807. PubMed Abstract | Publisher Full Text\n\nZhao X, Yeh JZ, Narahashi T: Post-stroke dementia. Nootropic drug modulation of neuronal nicotinic acetylcholine receptors. Ann N Y Acad Sci. 2001; 939: 179–186. PubMed Abstract | Publisher Full Text\n\nWei ZH, He QB, Wang H, et al.: Meta-analysis: the efficacy of nootropic agent Cerebrolysin in the treatment of Alzheimer’s disease. J Neural Transm (Vienna). 2007; 114(5): 629–634. PubMed Abstract | Publisher Full Text\n\nTanaka H, Yamazaki K, Hirata K: Effects of nootropic drugs for demented patients—a study using LORETA. Int Congr Ser. 2002; 1232: 605–611. Publisher Full Text\n\nAkhondzadeh S, Tajdar H, Mohammadi MR, et al.: A Double-blind Placebo Controlled Trial of Piracetam Added to Risperidone in Patients with Autistic Disorder. Child Psychiatry Hum Dev. 2008; 39(3): 237–245. PubMed Abstract | Publisher Full Text\n\nBreggin PR: Psychostimulants in the treatment of children diagnosed with ADHD: Risks and mechanism of action. International Journal of Risk & Safety in Medicine. 1999; 12: 3–35. Reference Source\n\nNoorbala AA, Akhondzadeh S, Davari-Ashtiani R, et al.: Piracetam in the treatment of schizophrenia: implications for the glutamate hypothesis of schizophrenia. J Clin Pharm Ther. 1999; 24(5): 369–374. PubMed Abstract | Publisher Full Text\n\nMcCabe SE: Screening for drug abuse among medical and nonmedical users of prescription drugs in a probability sample of college students. Arch Pediatr Adolesc Med. 2008; 162(3): 225–231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTeter CJ, McCabe SE, LaGrange K, et al.: Illicit use of specific prescription stimulants among college students: prevalence, motives, and routes of administration. Pharmacotherapy. 2006; 26(10): 1501–1510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArria AM, O'Grady KE, Caldeira KM, et al.: Nonmedical Use of Prescription Stimulants and Analgesics: Associations with Social and Academic Behaviors among College Students. J Drug Issues. 2008; 38(4): 1045–1060. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCabe SE, Knight JR, Teter CJ, et al.: Non-medical use of prescription stimulants among US college students: prevalence and correlates from a national survey. Addiction. 2005; 100(1): 96–106. PubMed Abstract | Publisher Full Text\n\nMcCabe SE, Teter CJ, Boyd CJ: Medical Use, Illicit Use and Diversion of Prescription Stimulant Medication. J Psychoactive Drugs 2006; 38(1): 43–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLakhan SE, Kirchgessner A: Prescription stimulants in individuals with and without attention deficit hyperactivity disorder: misuse, cognitive impact, and adverse effects. Brain Behav. 2012; 2(5): 661–677. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSichilima T, Rieder MJ: Adderall and cardiovascular risk: A therapeutic dilemma. Paediatr Child Health. 2009; 14(3): 193–195. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGouliaev AH, Senning A: Piracetam and other structurally related nootropics. Brain Res Brain Res Rev. 1994; 19(2): 180–222. PubMed Abstract | Publisher Full Text\n\nIsaacson JS, Nicoll RA: Aniracetam reduces glutamate receptor desensitization and slows the decay of fast excitatory synaptic currents in the hippocampus. Proc Natl Acad Sci U S A. 1991; 88(23): 10936–10940. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKolta A, Lynch G, Ambros-Ingerson J: Effects of aniracetam after LTP induction are suggestive of interactions on the kinetics of the AMPA receptor channel. Brain Res. 1998; 788(1–2): 269–286. PubMed Abstract | Publisher Full Text\n\nSatoh M, Ishihara K, Iwama T, et al.: Aniracetam augments, and midazolam inhibits, the long-term potentiation in guinea-pig hippocampal slices. Neurosci Lett. 1986; 68(2): 216–220. PubMed Abstract | Publisher Full Text\n\nJin R, Clark S, Weeks AM, et al.: Mechanism of positive allosteric modulators acting on AMPA receptors. J Neurosci. 2005; 25(39): 9027–36. PubMed Abstract | Publisher Full Text\n\nKoliaki CC, Messini C, Tsolaki M: Clinical efficacy of aniracetam, either as monotherapy or combined with cholinesterase inhibitors, in patients with cognitive impairment: a comparative open study. CNS Neurosci Ther. 2012; 18(4): 302–312. PubMed Abstract | Publisher Full Text\n\nBartolini L, Casamenti F, Pepeu G: Aniracetam restores object recognition impaired by age, scopolamine, and nucleus basalis lesions. Pharmacol Biochem Behav. 1996; 53(2): 277–283. PubMed Abstract | Publisher Full Text\n\nCumin R, Bandle EF, Gamzu E, et al.: Effects of the novel compound aniracetam (Ro 13-5057) upon impaired learning and memory in rodents. Psychopharmacology (Berl). 1982; 78(2): 104–11. PubMed Abstract | Publisher Full Text\n\nLu Y, Wehner JM: Enhancement of contextual fear-conditioning by putative (+/-)-alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor modulators and N-methyl-D-aspartate (NMDA) receptor antagonists in DBA/2J mice. Brain Res. 1997; 768(1–2): 197–207. PubMed Abstract | Publisher Full Text\n\nRao Y, Xiao P, Xu S: Effects of intrahippocampal aniracetam treatment on Y-maze avoidance learning performance and behavioral long-term potentiation in dentate gyrus in rat. Neurosci Lett. 2001; 298(3): 183–186. PubMed Abstract | Publisher Full Text\n\nMartin JR, Moreau JL, Jenck F: Aniracetam reverses memory impairment in rats. Pharmacol Res. 1995; 31(2): 133–6. PubMed Abstract | Publisher Full Text\n\nVaglenova J, Pandiella N, Wijayawardhane N, et al.: Aniracetam reversed learning and memory deficits following prenatal ethanol exposure by modulating functions of synaptic AMPA receptors. Neuropsychopharmacology. 2008; 33(5): 1071–83. PubMed Abstract | Publisher Full Text\n\nElston TW, Pandian A, Smith GD, et al.: Aniracetam does not alter cognitive and affective behavior in adult C57BL/6J mice. PLoS One. 2014; 9(8): e104443. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOgiso T, Uchiyama K, Suzuki H, et al.: Pharmacokinetics of aniracetam and its metabolites in rat brain. Biol Pharm Bull. 2000; 23(4): 482–6. PubMed Abstract | Publisher Full Text\n\nCorazza O, Bersani FS, Brunoro R, et al.: The diffusion of performance and image-enhancing drugs (PIEDs) on the internet: the abuse of the cognitive enhancer piracetam. Subst Use Misuse. 2014; 49(14): 1849–56. PubMed Abstract | Publisher Full Text\n\nReynolds CD, Jefferson T, Volquardsen M, et al.: Dataset 1 in: Study of oral aniracetam in C57BL/6J mice without pre-existing cognitive impairments. F1000Research. 2017. Data Source"
}
|
[
{
"id": "26005",
"date": "12 Oct 2017",
"name": "Marie-H Monfils",
"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 The study was focused on assessing whether healthy (i.e. non-cognitively impaired) mice would display benefits following the administration of the nootropic drug aniracetam. Mice were run through a battery of tests aimed following oral administration of either a vehicle solution or an aniracetam solution. Though the main focus of the paper seemed to be on identifying whether aniracetam administered mice showed improvement in a variety of learning and memory tasks, mice were also assessed on measures of anxiety and repetitive behavior. The aniracetam treated mice did not show any significant improvements on any of the learning and memory tasks compared to controls. In addition, aniracetam treated mice did not display increased anxiety or repetitive behaviors as compared to control mice.\n\nSuggested Changes\nThe title should be rewritten to include information about the outcome measures that are used. As it is, it is not readily clear from the title that this is a behavioral study.\n\nThe introduction does not provide an explanation for some of the tests that were run. Specifically, there was no mention of anxiety or repetitive behavior being linked with nootropic use. A discussion of this relationship should be included at some point in the introduction.\n\nThe authors do not mention which light phase the mice were run during. Given that all the outcome measures are behavioral and metabolic factors may influence how the mice process the aniracetam, this is a detail that needs to be included.\n\nWhether or not the order of the tests was randomized and the amount of time that was given between each test should be mentioned.\n\nThe write-up of the fear conditioning methodology is missing a number of important details. The following need to be added: (1) A thorough description of the CS in use (e.g. was it a pure tone or white noise? What was the volume?), (2) the number of CS-US pairings, (3) the intensity and duration of the US, (4) the amount of time the mouse was given to habituate to the cage, and (5) the amount of time between CS presentations.\n\nExcitatory post-synaptic current should be written out fully before being abbreviated to EPSC.\n\nA description of Figure 4C needs to be added to the caption of Figure 4.\n\nA significant difference between the freezing displayed by aniracetam treated rats and control rats is indicated in Figure 4B but the statistics backing up this finding are not included in the results or discussed at any point in the paper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3276",
"date": "22 Jun 2018",
"name": "Joaquin Lugo",
"role": "Author Response",
"response": "Dear Drs. Monfils and Agee,We thank you for your input. We have revised the paper with your recommendations in mind. We believe the paper has been improved by their input. We have changed the title to reflect the behavioral outcome measures we used in the study. We have provided more justification in the introduction for the anxiety and repetitive behavior test we used in our experiments.MethodsWe included more information on the light cycle and when the mice were tested.We included the order of the behavioral tests and more detailed information on the fear conditioning tests.ResultsWe revised figure 4 to include figures 4A, 4B, and 4C. We also included all statistics for each portion of the tests. We wrote out a description of figure 4C in the caption.DiscussionWe wrote out the term excitatory post-synaptic current before we abbreviated it to EPSC."
}
]
},
{
"id": "25058",
"date": "17 Oct 2017",
"name": "Stephan G. Anagnostaras",
"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 fairly well-performed study but it has a serious flaw that isn't really addressed, which is that it uses a single dose. Single dose studies are problematic unless that dose is really well established as the appropriate dose, and it's really not very clear here how or why that dose is chosen. Although the dose chosen is maybe in the dose range people tend to take (maybe double the maximum dose people seem to report), pro-cognitive effects are notoriously dose-specific and it is possible they simply have the wrong dose. Rather than the very extensive behavioral approach, why didn't the authors simply run one task (e.g., fear conditioning) using a range of doses? At the very least the authors need to acknowledge this substantial limitation.\n\nThe other issues have been mentioned in the other review; in particular, the fear conditioning parameters and scoring methodology are not adequately described.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "3267",
"date": "15 Dec 2017",
"name": "Joaquin Lugo",
"role": "Author Response",
"response": "We would like to thank Dr. Stephan G. Anagnostaras for his comments on our paper. We have included discussion on the limitation of the use of a single dose. We agree that this is a substantial limitation of our paper and the reviewer provided a great experimental design to determine an effective dose on cognition. We appreciate the input and plan to use this approach in future studies. We have also included more information on the methods of the fear conditioning test."
}
]
}
] | 1
|
https://f1000research.com/articles/6-1452
|
https://f1000research.com/articles/7-563/v1
|
09 May 18
|
{
"type": "Software Tool Article",
"title": "Comparing protein structures with RINspector automation in Cytoscape",
"authors": [
"Guillaume Brysbaert",
"Théo Mauri",
"Marc F. Lensink",
"Théo Mauri",
"Marc F. Lensink"
],
"abstract": "Residue interaction networks (RINs) have been shown to be relevant representations of the tertiary or quaternary structures of proteins, in particular thanks to network centrality analyses. We recently developed the RINspector Cytoscape app, which couples centrality analyses with backbone flexibility predictions. This combined approach permits the identification of crucial residues for the folding or function of the protein that can constitute good targets for mutagenesis experiments. Here we present an application programming interface (API) for RINspector that enables interplay between Cytoscape, RINspector and external languages, such as R or Python. This API provides easy access to batch centrality calculations and flexibility predictions, and allows for the easy comparison of results between different structures. These comparisons can lead to the identification of specific and conserved central residues, and show the impact of mutations to these and other residues on the flexibility of the proteins. We give two use cases to demonstrate the interest of these functionalities and provide the corresponding scripts: the first concerns NMR conformers, the second focuses on mutations in a structure.",
"keywords": [
"rinspector",
"cytoscape",
"protein structure",
"residue interaction network",
"centrality analysis",
"flexibility prediction",
"automation",
"structure ensemble"
],
"content": "Introduction\n\nKnowledge of the structures of proteins is important to understand their function and provide a starting point for further exploration through techniques such as molecular dynamics or docking. These structures can be used directly, in the form of a PDB file, or undergo transformations prior to analysis. Residue interaction networks (RINs) are networks built from a 3D structure, where nodes represent amino acids and edges represent detected interactions between them. These networks can be generated with different tools, like RING21, RINerator2 or Chimera3, and then be imported into network visualization and analysis tools. Depending on the purpose, these network tools can be libraries available for languages like R or Python (e.g iGraph or NetworkX) or software with a graphical interface like the well-known Cytoscape4, a reference for biological network studies. The CyREST technology5, which is now included as a core app of Cytoscape 3.3 (and higher versions), enables further analyses, letting the user the possibility to complement analyses done in Cytoscape with scripting developments through external languages like R or Python.\n\nWe recently developed the RINspector6 app for Cytoscape, which combines centrality analyses of residue interaction networks with flexibility predictions of a protein from its associated sequence through a call to the DynaMine flexibility prediction server7,8. Centrality analyses have been shown to identify residues important for functions, folding or allostery, as well as long range interactions (e.g. Refs 9,10). Coupled with flexibility predictions, the app enables users to highlight a subset of these central residues, which might affect the dynamics upon mutation6.\n\nThe RINspector app is convenient if one works on only a few networks, but certain centrality calculations may be CPU- and memory-intensive and require long execution times for analyses as soon as the number of networks increases. This may be the case for residue centrality analyses (RCAs) calculated on NMR data, when several conformers need to be considered. Another example is the comparison of RINs generated from a wild-type structure and several mutants. Furthermore, the app provides a score for each residue in each network and the export of a recap chart would be useful to compare between RINs. RINspector now provides a documented application programming interface (API) in its 1.1.0 version that provides automation of centrality calculations and flexibility predictions, thereby addressing these issues.\n\nHere we present the automation functionalities of RINspector, which allows for combination of the power of Cytoscape with scripting languages such as R or Python, in order to calculate centralities, predict flexibilities, visualize the networks, and get a recap chart with scores. We present two use cases, one based on NMR data and another one based on a mutated domain structure.\n\n\nMethods\n\nRINspector is implemented in Java as an app for Cytoscape 3. It uses the JFreeChart library (Copyright 2000–2009, Object Renery Limited and Contributors) for graph representations. The DynaMine flexibility predictions are retrieved through a JSON API. Betweenness and closeness centrality values were calculated through the functions developed by Assenov et al.11,12 for the Cytoscape NetworkAnalyzer core tools.\n\nCentrality analyses and flexibility predictions can be performed via the Cytoscape graphical interface through the App/RINspector menu (see Ref. 6 for a description). They can also be called through the Cytoscape command dialog or through the CyREST command API presented here.\n\nThe API consists of two commands, offered with documentation through the CyREST command API in Cytoscape. The commands are:\n\n- “centrality”, which calculates centralities for each residue in the residue interaction network currently selected in Cytoscape. This command needs one parameter, which is the type of centrality calculation to perform, selectable between:\n\n○ RCA (average shortest path length (ASPL) change under removal of individual nodes)\n\n○ Betweenness centrality analysis – BCA\n\n○ Closeness centrality analysis – CCA\n\nRCA corresponds to the calculation proposed by 9. The ASPL of the RIN is first calculated, after which the ASPL is calculated for each network upon removal of individual single nodes. A Z-score is then computed for each node based on the change of ASPL compared to the initial one. BCA and CCA are the classical betweenness and closeness centrality calculations, both of which are followed by the calculation of a Z-score (for more details about the process, see the Supplementary materials of Ref. 6).\n\n- “dynamine”, which queries the DynaMine server7,8 with the sequence of the currently selected RIN. This function requires that the table of nodes in Cytoscape contains three columns: ResType, ResIndex and ResChain. The ResType column should contain the 3-letter code for each residue (e.g. ARG). The ResIndex column should contain the serial number (or index) of each residue (e.g. 153). The ResChain column should contain the chain identifier (e.g. A). These three columns are automatically created if the RIN is generated with Chimera through the structureViz2 app. One parameter has to be specified to the “dynamine” command, namely the chain of the protein, formatted as for the ResChain column. The DynaMine server returns a S² flexibility score for each residue.\n\nEach run of centrality calculation or flexibility prediction returns a score per residue. These scores appear in a dedicated column in the node table in Cytoscape and are returned as a two-columns table (node ID and score) in JSON through the REST service. The visual style of the network is also adapted (see Figure 1 and Ref. 6). The output table and/or the created columns in the node table can be further treated by a third party program written, for example, in R or Python. Table 1 presents the parameters of the POST request that are used in the provided scripts for a call of centrality or flexibility predictions and associated responses. Considering system requirements, use of a computer with at least 16 GB of memory is advised because the RCAs are memory demanding, and the bigger the structure, the more memory required. The running of use cases should not last more than a few minutes. Automation for RINspector requires Cytoscape 3.6.0 (or higher) and RINspector 1.1.0. The scripts we provide were developed and tested for Python 3.5.2 and R 3.2.3, with Cytoscape 3.6.1.\n\n(A) Overlapped structure of the 10 conformers used to generate the residue interaction networks (RINs) for ALG13 (PDB ID 2JZC) in ribbon representation. (B) The RINs on which RCAs were performed; central residues are colored from yellow (Z-score = 2) to red (Z-score ≥ 4), labels and node sizes are adapted as functions of the Z-scores: the higher, the bigger; only the first two RINs of the ten calculated are shown. (C) Final result table, which contains the centralities of all the networks, sorted by descending order of Z-scores of the first conformer; Z-scores considered as relevant (Z-score ≥ 2) have a pink background; only the first 15 residues are displayed.\n\nFor each request to the REST service, the url, body and encoding format are specified. The response in both cases is a table that contains the Z-score (centrality) or the S² score (dynamine) associated to each node ID, in JSON format.\n\n\nUse cases\n\nWe present here two examples that benefit from the API. The first one considers NMR data with 10 conformers of a yeast N-acetylglucosamine transferase, all grouped in the PDB ID 2JZC. The second one is the tetratricopeptide repeat (TPR) domain of a human O-GlcNAc transferase (OGT) (PDB ID 4GYW).\n\nNMR data are well suited for automation treatment because the associated PDB files usually contain multiple conformers for which RINs can be generated. We built the RINs in Cytoscape from Chimera through the structureViz2 app. Once these are created, the RINspector API can be queried to calculate centralities in batch. In this example, we calculate residue centrality scores for each residue and compare them between the conformations. We wrote a script in R and an equivalent one in Python that, starting from a Cytoscape session containing one RIN for each conformer, perform RCA on each RIN. The scripts then gather the Z-scores for each residue in each RIN in a single recap table, allowing for easy comparison. The app also creates a style for each RIN that permits a visual comparison (Figure 1).\n\nHere we compare the centralities of residues when RINs are generated from structures which contain point mutations. We also compare the impact of mutations on flexibility.\n\nWe created six mutants in the tetratrico peptide repeat (TPR) domain of the O-GlcNAc transferase (OGT) by editing the PDB file of the wild type. Five mutants were asparagines to alanines (N322A, N325A, N356A, N390A, N424A), the sixth contained all five point mutations. The OGT is an enzyme that catalyzes the transfer of a single N-acetylglucosamine from UDP-GlcNAc to a serine or threonine amino acid (called O-GlcNAcylation). These five asparagines have been shown to decrease the efficiency of the OGT enzymatic activity on a category of peptides when simultaneously mutated into alanines13. We generated the RINs (without ligands) for the wild-type structure and each of the six variants. We wrote a script that, starting from a Cytoscape session containing these RINs, calculates residue centralities and predicts flexibilities for each structure. As in Use case 1, the results of centralities are gathered in a common table. In addition, the S² flexibility scores are also gathered in a recap table and in a single plot to permit easy comparison (Figure 2). These tables and plot allow for comparison between the centralities, to see which of the mutations have an impact on which centralities compared to the wild type, and to see the impact of each mutation on the backbone flexibility of the TPR domain.\n\n(A) Final result table showing the wild type TPR domain of the human OGT and the 6 mutants N322A, N325A, N342A, N356A, N390A and 5N5A (containing all five mutations), sorted by descending order of Z-scores of the wild type; Z-scores considered as relevant (Z-score ≥ 2) have a pink background; only the first 15 residues are shown. (B) DynaMine flexibility graphs of the wild type (WTLs) and each of the mutants; X=sequence index, Y=S², the higher S² value, the more rigid the backbone is predicted to be; grey horizontal straight lines delimit a context dependent zone; mutated residues are highlighted with red vertical straight lines.\n\n\nDiscussion\n\nComparisons between structures, and more precisely between residues in structures, are useful for identification of one or a subset of amino acids that are crucial for the folding or functions of proteins. The automation functionalities of RINspector provided through CyREST in Cytoscape make these comparisons easier. Indeed, they permit to perform centrality analyses and flexibility predictions of residue interaction networks generated from structures of proteins, in batch mode from an external script. The results can then be exploited using the libraries available for the language of the script (usually R or Python).\n\nIn our cases, we treated at most ten networks at a time, which is reasonable. This number can of course be higher, but depending on the size of the structure/network, it may be difficult to deal with a substantially higher number of networks. Indeed, while BCAs and CCAs are usually performed within a few seconds, RCAs are memory demanding and running many of them may quickly fill the RAM of the system. In such cases we recommend to perform all the calculations through a workflow outside Cytoscape. The CyREST core or command API can nevertheless still be called in the workflow, e.g. for the generation of networks visualizations. RCAs and BCAs usually give many central residues which show various degrees of overlap. Depending on the objective of the analyses, it may be more relevant to consider the results of one centrality, the intersection, or the union of these selections. However, CCAs generally result in few residues with Z-scores ≥ 2. If the user is interested in this specific measurement, we advise to visualize directly the closeness centrality values and not the Z-scores (with NetworkAnalyzer tools in Cytoscape or RINalyzer app2 for instance).\n\nThe RINspector API can be used in conjunction with other apps, especially with the structureViz2 app, which connects Chimera to Cytoscape, because it permits to open a structure in Chimera and work with it. This connection allows for the easy generation of RINs and the possibility of synchronization of colors between network and structure, which is a welcome feature for the generation of images. In this respect we would also like to point out the RINalyzer app, which enables the user to layout the RIN in function of coordinates of residues displayed in Chimera.\n\nIn Use case 2, DynaMine results were exploited to build a chart for comparison of flexibilities between structures. For such analyses the proposed script is particularly interesting, as computation resources are not a limiting factor. In most cases, however, only a few mutations are really of interest and the result panel designed for DynaMine in RINspector may be more convenient to visualize the effect of these mutations directly and select one or several for protein design. The user should also be aware that the sequence sent to the DynaMine server is built from the RIN/structure, which means that missing residues will simply be skipped in the DynaMine flexibility graph.\n\nOther use case examples might be the effect of one or several ligands in a structure by generating RINs with or without these ligands, the comparison of centralities of different structures of the same complexes or the comparison of the flexibility of structures of several orthologs. Our scripts constitute starting points to perform such analyses with recap tables and charts as output.\n\nWe plan to extend the interaction between other languages and Cytoscape complementing batch centrality analyses and flexibility predictions with automatic generation of RINs and integration of conservation data.\n\n\nSummary\n\nThe RINspector API permits scripting of centrality analyses and flexibility predictions. The automation provided bridges between Cytoscape and other languages, such as R or Python, to prepare data, run batch analyses and treat output data. It enlarges the possibilities of treatments that were initially given by the app in particular to compare centralities and flexibilities between multiple structures. We provide R and Python scripts that illustrate two use cases and that are to be seen as starting scripts for more elaborate analyses.\n\n\nData and software availability\n\n1. RINspector is available from the Cytoscape App Store: http://apps.cytoscape.org/apps/rinspector\n\n2. Latest source code is available at: https://sourcesup.renater.fr/scm/?group_id=3888\n\n3. Archived source code as at time of publication: http://doi.org/10.5281/zenodo.124043714.\n\n4. Software license: CeCILL; version 2.1: http://www.cecill.info",
"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\nAcknowledgments\n\nWe thank Wim F. Vranken for providing access to the DynaMine server, the Cytoscape core team for their help in the development of the automation process and the SourceSup platform for hosting the project.\n\n\nReferences\n\nPiovesan D, Minervini G, Tosatto SC: The RING 2.0 web server for high quality residue interaction networks. Nucleic Acids Res. 2016; 44(W1): W367–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoncheva NT, Klein K, Domingues FS, et al.: Analyzing and visualizing residue networks of protein structures. Trends Biochem Sci. 2011; 36(4): 179–82. PubMed Abstract | Publisher Full Text\n\nPettersen EF, Goddard TD, Huang CC, et al.: UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem. 2004; 25(13): 1605–12. PubMed Abstract | Publisher Full Text\n\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\nOno K, Muetze T, Kolishovski G, et al.: CyREST: Turbocharging Cytoscape Access for External Tools via a RESTful API [version 1; referees: 2 approved]. F1000Res. 2015; 4: 478. [cited 2018 Mar 15]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrysbaert G, Lorgouilloux K, Vranken W, et al.: RINspector: a Cytoscape app for centrality analyses and DynaMine flexibility prediction. Bioinformatics. 2018; 34(2): 294–296. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCilia E, Pancsa R, Tompa P, et al.: From protein sequence to dynamics and disorder with DynaMine. Nat Commun. 2013; 4: 2741. PubMed Abstract | Publisher Full Text\n\nCilia E, Pancsa R, Tompa P, et al.: The DynaMine webserver: predicting protein dynamics from sequence. Nucleic Acids Res. 2014; 42(Web Server issue): W264–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\ndel Sol A, Fujihashi H, Amoros D, et al.: Residues crucial for maintaining short paths in network communication mediate signaling in proteins. Mol Syst Biol. 2006; 2: 2006.0019. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu G, Yan W, Zhou J, et al.: Residue interaction network analysis of Dronpa and a DNA clamp. J Theor Biol. 2014; 348: 55–64. PubMed Abstract | Publisher Full Text\n\nAssenov Y, Ramírez F, Schelhorn SE, et al.: Computing topological parameters of biological networks. Bioinformatics. 2008; 24(2): 282–4. PubMed Abstract | Publisher 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\nRafie K, Raimi O, Ferenbach AT, et al.: Recognition of a glycosylation substrate by the O-GlcNAc transferase TPR repeats. Open Biol. 2017; 7(6): pii: 170078. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrysbaert G, Mauri T, Lensink MF: RINspector source code (Version 1.1.0). Zenodo. 2018. Data Source"
}
|
[
{
"id": "33871",
"date": "12 Jun 2018",
"name": "Sjoerd J. de Vries",
"expertise": [
"Reviewer Expertise computational structural biology",
"structural bioinformatics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nBrysbaert et al. describe an API for their RINspector tool for the analysis of residue interaction networks (RINs). This API allows RIN analysis in an automated fashion (batch mode), controlled from any programming language that can perform HTTP requests. The authors provide two use cases as an example. The article is sound and well-written. I have only minor comments pertaining the didactic value of the paper: I feel the authors could lower the barrier for users of the software, by explaining better what it is good for and how to get it running.\nThe running time of the use cases should be reported, to demonstrate the amount of human time saved. The use case example scripts are available from the Git repository, but this is not mentioned, and the repository itself is not easy to browse. Direct links would be better. The software relies heavily on Cytoscape, its app store, and it's cyREST technology, and the authors assume that the reader has good knowledge of these. However, Cytoscape is not much used in the protein structure community. I recommend that the authors include a link to a short install manual for Cytoscape, cyREST and RINspector, plus any other necessary steps that must be performed before the example scripts can be run. It deserves some explanation that the RINspector API is not meant to (and in fact, unable to) provide any kind of web service. The reason why might be well-known within the Cytoscape community, but again, Cytoscape is not very well known among structural biologists, and I have never used it myself. From what I understand, the cyREST technology is very poorly named, since it is not a proper (i.e. stateless) REST protocol at all. Instead, it seems to be a technology to send state-manipulating commands (load data; analyze current data; etc.) over HTTP to a running instance of Cytoscape. The RINspector API follows this state-manipulating paradigm. In addition, it seems that Cytoscape cannot run behind a web server because it always runs a GUI.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3747",
"date": "22 Jun 2018",
"name": "Guillaume Brysbaert",
"role": "Author Response",
"response": "Dear Dr de Vries Thank you for your interest of our manuscript and your helpful remarks. We took them in consideration and updated the manuscript. Here are our replies to your comments: The running time of the use cases should be reported, to demonstrate the amount of human time saved. → Running times were added in the text in the ‘Operation’ part. The use case example scripts are available from the Git repository, but this is not mentioned, and the repository itself is not easy to browse. Direct links would be better. → We added the information that the example files are available in the git repository. We specified the relative folder on the git repository and set a link to the folder in tree visualisation. The software relies heavily on Cytoscape, its app store, and it's cyREST technology, and the authors assume that the reader has good knowledge of these. However, Cytoscape is not much used in the protein structure community. I recommend that the authors include a link to a short install manual for Cytoscape, cyREST and RINspector, plus any other necessary steps that must be performed before the example scripts can be run. → In each use case folder, we included a README.txt file that details the required steps to run the scripts. We added this information in the ‘Use cases’ part of the article. CyREST is included in the last versions of Cytoscape, that’s why the README.txt files don’t contain any mention about CyREST installation. It deserves some explanation that the RINspector API is not meant to (and in fact, unable to) provide any kind of web service. The reason why might be well-known within the Cytoscape community, but again, Cytoscape is not very well known among structural biologists, and I have never used it myself. From what I understand, the cyREST technology is very poorly named, since it is not a proper (i.e. stateless) REST protocol at all. Instead, it seems to be a technology to send state-manipulating commands (load data; analyze current data; etc.) over HTTP to a running instance of Cytoscape. The RINspector API follows this state-manipulating paradigm. In addition, it seems that Cytoscape cannot run behind a web server because it always runs a GUI. → RINspector is an app for the standalone Cytoscape software that uses the REST service provided by the CyREST tool inside Cytoscape. As such, contrarily to the usual REST services designed for a web usage, Cytoscape is a standalone tool with a GUI that has to be run on a single computer to start the REST service that is only accessible on this computer. Requests are done thanks to the HTTP protocol and are formatted in JSON. We added explanations in the ‘Implementation’ part. The REST architecture that is used is stateless in the sense that each request is independent from another and no state is conserved by the local server. Best regards"
}
]
},
{
"id": "33872",
"date": "12 Jun 2018",
"name": "Turkan Haliloglu",
"expertise": [
"Reviewer Expertise Modeling and simulation of macromolecules. Protein dynamics and allostery."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 RINspector app allows network and flexibility analysis of protein structures, which plausibly has the capacity to disclose functionally critical residues. The paper by Brysbaert et al. explains the updated (automated) version of the RINspector app. The new version is stated to work more efficiently than previous one, especially while treating multi-conformers. The paper could further be improved, if the following points are addressed.\n\n- The differences between the previous and current version of the program are not clearly stated. It is confusing what is new and what is updated, especially for someone who is not familiar with the RINspector app. Maybe the authors could mention the version numbers explicitly when they make a comparison. - There is a tutorial provided for the initial version of RINspector. A similar tutorial should also be prepared for the use cases presented in this paper. In this way, the users can exactly reproduce what is stated in the manuscript. - Relevant links to the software, the use case folders and scripts should be cited within the text. In the current version, it is hard to link the source code files with the options/conditions stated in the paper.\nOverall, RINspector app stands a useful resource that would corroborate structural and functional studies in the community.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3748",
"date": "22 Jun 2018",
"name": "Guillaume Brysbaert",
"role": "Author Response",
"response": "Dear Dr Haliloglu and Dr Karaca, Thank you for your interest in our article and your constructive comments. We address them point by point. - The differences between the previous and current version of the program are not clearly stated. It is confusing what is new and what is updated, especially for someone who is not familiar with the RINspector app. Maybe the authors could mention the version numbers explicitly when they make a comparison. → The versions of RINspector were specified in the text to avoid confusion. - There is a tutorial provided for the initial version of RINspector. A similar tutorial should also be prepared for the use cases presented in this paper. In this way, the users can exactly reproduce what is stated in the manuscript. → We included a ‘README.txt’ file in each use case folder to explain how to run the example scripts. The point of the paper being the automation of RINspector, we don’t explain how to generate the RINs but we provide a python script called ‘creation_RIN_NMR.py’ in the use case 1 that permits to generate the scripts with structureViz2/Chimera. - Relevant links to the software, the use case folders and scripts should be cited within the text. In the current version, it is hard to link the source code files with the options/conditions stated in the paper. → In the ‘Use cases’ part, we added the information that the use cases files (inputs/scripts/outputs) are available in the git repository and added a hyperlink to this folder in the git. Best regards"
}
]
},
{
"id": "33873",
"date": "18 Jun 2018",
"name": "Rita Casadio",
"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\nComparing protein structure with RINspector is an intersting application that allows an easy determination of crucial residues in protein structures. It can be useful for pinpointing specific positions in the protein backbone for further experimental investigations.\nThe paper clearly explains the procedure, and clarifies with test cases the relevance of the application. Software is detailed and reproducible. Figures are clear and appropriate.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3749",
"date": "22 Jun 2018",
"name": "Guillaume Brysbaert",
"role": "Author Response",
"response": "Dear Dr Casadio, Thank you for the review and your interest in our work. We are happy to see that you accept the manuscript. Best regards"
}
]
}
] | 1
|
https://f1000research.com/articles/7-563
|
https://f1000research.com/articles/7-878/v1
|
22 Jun 18
|
{
"type": "Research Article",
"title": "Bibliometrics of the 100 most-cited articles on refugee populations",
"authors": [
"Musatafa Khosa",
"Ahmed Waqas",
"Mahnoor Javaid",
"Jessica Singh",
"Sadiq Naveed",
"Salman Majeed",
"Faisal Khosa",
"Musatafa Khosa",
"Mahnoor Javaid",
"Jessica Singh",
"Sadiq Naveed",
"Salman Majeed",
"Faisal Khosa"
],
"abstract": "Background: Bibliometrics is a form of quantitative analysis that employs peer-reviewed research, journal articles and citation counts to examine the content of current literature on a particular topic. The authors aim to identify the major academic disciplines that dominate the landscape of published materials and research endeavors on the topic of refugees. Methods: Using the Web of Science, a database of most-cited articles was created by a team with expertise in bibliometrics. Results: Citations ranged between 1,493 and 105; averaging 203 citations per article. The publications spanned the years from 1973 to 2010. The year 2004 had the highest number of publications. All articles were published by 45 journals. In total, 294 investigators authored these articles. Psychiatry, psychology and public health constituted the top three fields of affiliation, with the most investigated feature being the mental health of refugees. Single investigators authored a quarter of all articles. Conclusion: This bibliometric evaluation allowed a multi-dimensional outlook on the conditions of refugee populations across the globe, through collation of relevant peer-reviewed research journal articles. This specialized form of assessment has resulted in a multi-disciplinary compendium of publications on the subject.",
"keywords": [
"refugee",
"resettlement",
"bibliometric",
"citation analysis",
"mental health"
],
"content": "Introduction\n\nAccording to the Office of the United Nations High Commissioner for Refugees (UNHCR), “a refugee is someone who has been forced to flee his or her country because of persecution, war, or violence”1. Refugees fear persecutions owing to their ethnicity, religious beliefs, and nationality in resettlement areas and foreign lands. Despite these fears, they cannot return to their home-countries due to ongoing wars, ethnic, tirbal and religious violence and conflicts2.\n\nRecent estimates affirm a global refugee population of 21.3 million, with the majority of refugees from Syria, Afghanistan, and Somalia1. Beset by a diverse set of health concerns and diseases, the well-being of refugees is contingent on the quality of the available health care, as mandated by responsible resettlement authorities2. The resettlement of refugees in Europe and North America also drives changes in the sociocultural landscape and epidemiology of diseases requiring an update in healthcare system of the host country, and generation of new policy frameworks3. These emerging social and political issues demand a review of current research initiatives in the literature concerning refugees.\n\nBibliometrics is a form of quantitative analysis that employs peer-reviewed research, journal articles and citation counts to examine the content of current literature on a particular topic4. This form of review evaluates a given topic through three main facets: i) outlining the most and least investigated research areas; ii) providing a summary of the journals containing publications on the topic; and iii) providing a summary of relevant authors on the topic4. In this way, this type of study effectively: i) gauges the dimensions of available literature on the topic; ii) identifies the journals that accommodate publications on the topic; iii) maps out the quality of current research on the topic; and iv) guides and prioritizes forthcoming research4. In addition, bibliometrics can help connect researchers pursuing similar interests, opening up the possibility for joint research ventures.\n\nThe bibliometric assessment performed in the present study explores a diverse set of academic research on refugee populations around the world. Using a method of “citation analysis”5, this study examines this topic through citation counts accompanying peer-reviewed scientific publications. These citation counts provide an appraisal of the current bibliographic data through a calculation of the number of references associated with each article5. Bibliometrics have been published in a number of medical disciplines, and the authors have published on the topic as well6–10. The current study seeks to highlight the diversity of the body of research on refugee communities. In effect, it will identify the major academic disciplines that dominate the landscape of published materials and research endeavors on this topic.\n\n\nMethods\n\nUsing Thomas Reuters’ Web of Science (WOS), a database of peer-reviewed scientific journal articles and citation index, the search item: “refugee” was plugged into the search bar. The option, “All Databases” was selected to ensure a comprehensive result list. Subsequently, the feature, “most cited to least cited,” was used to sort the generated result list in descending order of citation counts.\n\nThis bibliometric assessment was conducted in June 2016. The list of generated results in WOS was reviewed and an itemized list was generated by M.K. and S.N., who had expertise and prior experience in conducting bibliometric analyses. Eligibility of articles for inclusion was assessed by M.K., S.N. independently and any differences were resolved through consensus among all authors. All peer-reviewed scientific research journal articles pertaining to refugee populations across the globe were included in this study. Microsoft Excel 2016 was used to create a list of one hundred approved articles that were rearranged from highest to lowest citation counts.\n\nTwo academic databases included in WOS were utilized to generate data related to top 100 cited articles: i) Web of Science (all databases) and ii) Web of science (core collection). The Web of Science Core Collection consists of ten indexes comprising 20 thousand scholarly journals, books, book series, reports, and conferences across sciences, social sciences, and humanities disciplines11. Whereas WOS (all databases) expands upon core-database collection by including information from more databases including SciELO, Medline and Russian citation index among others11.\n\nThis cascading list recorded the following particulars of each article: i) complete citation; ii) citation counts in WOS “All Databases;” iii) citation counts in the WOS Core Collection; iv) name of journal; v) year of publication; vi) author(s); vii) total number of authors; viii) study design; ix) sample size; x) number of institutional affiliations; xi) fields of affiliations; xii) usage of statistical analyses; xiii) research areas; xiv) disciplines; and xv) places of origin. All particulars were stratified and analyzed separately. No ethical approval was sought from any institution because of the bibliometric study design and lack of human participants in the study.\n\nWhile the research areas were provided by WOS, the disciplines were formulated in this study as broader categories accommodating analogous research areas. For example, the research areas of adult psychiatry, child and adolescent psychiatry, psychology, behavioral sciences, addiction medicine and substance abuse were incorporated into the single discipline of “Mental health and Behavioral and Addiction Medicine.” Further details of the research areas contained within each discipline is given in Table 1.\n\nStatistical analyses of data were performed using continuous variables via SPSS v23.0 (SPSS, Inc., Chicago, IL). Quantitative variables were presented as median (IQR) and categorical variables as frequencies (percentages). Categorical variables were presented as bar graphs and pie charts. Association between number of citations and year of publication of articles was presented as a line graph.\n\n\nResults\n\nDataset 112 displays the finalized catalog of the 100-most cited articles on refugees.\n\nTwo categories of citations were recorded: WOS All Databases and WOS Core Collection. Citations in WOS All Databases ranged between 1,493 and 105; with a median value of 203 citations per article (IQR = 77.25). A total of 76 articles (over three-fourths of all articles) were cited between 100 and 199 times. This was followed by 15% (n=15) of all articles being cited from 200 to 299 times. Only 9 articles (less than 10% of all articles) were cited over 300 times. The most cited article, garnering 1,493 citations, was titled, “Immigration, Acculturation, and Adaptation,” and was published in Applied Psychology: An International Review in the year 1997. The least cited article, with 105 citations, was titled, “A 6-year follow-up study of Cambodian refugee adolescents traumatized as children,” and was published in the Journal of the American Academy of Child & Adolescent Psychiatry in the year 1993.\n\nThe features that differentiate the findings of the WOS All Databases from those of the WOS Core Collection included: i) citations in the WOS Core Collection ranged between 1,462 and 101; averaging at 200 citations per article (IQR = 76.25); and ii) the least cited article, with 101 citations, is titled, “Environmental concerns and international migration,” and was published in the International Migration Review in the year 1996. Figure 1 compares the citation counts of WOS All Databases and WOS Core Collection.\n\nOut of the 100 articles, there were 46 reviews, 44 cross-sectional study designs and 3 meta-analyses. Both the aforementioned most-cited article (in WOS All Databases and WOS Core Collection) and the least-cited article in the WOS Core Collection were review articles. Figure 2 depicts the types of study designs used by the articles.\n\nA total of 57 articles employed multiple forms of quantitative analyses, with the remaining 43 articles not utilizing statistical analyses. As review articles, both the aforementioned most-cited article (in both WOS All Databases and WOS Core Collection) and the least-cited article in the WOS Core Collection, did not employ quantitative analyses.\n\nThe majority of the articles (n=52) had defined sample sizes, which ranged between 3 and 60,000,000. The remaining 48 articles had no defined sample sizes, as they either reviewed literature or detailed the conditions and predicaments of refugees – thus lacking clinical study designs. With the largest sample size of 60,000,000, the retrospective study titled, “Ages, life stages, and generational cohorts: Decomposing the immigrant first and second generations in the United States,” was published in International Migration Review in the year 2004. Marking the smallest sample size, 3 individuals were recruited in a cross-sectional study titled, “Unexplained Rabies in Three Immigrants in the United States a Virologic Investigation.” This study was published in the New England Journal of Medicine in the year 1991. Both the aforementioned most cited article (in WOS All Databases and WOS Core Collection) and the least cited article in the WOS Core Collection, had no defined sample sizes; however, the least cited article in WOS All Databases documented a longitudinal cross-sectional study design with a sample size of 84.\n\nThese 100 publications spanned the years from 1973 to 2010. The year 2004 marked the year with the highest number of publications (n=8); followed by the years 1995 and 1998 that produced 7 publications, each. The years 1973, 1979, 1984, 2007, and 2009 churned out the least number of publications; yielding 1 article during each year. The period from 2000 to 2010 witnessed the production of 46 articles, while the period from 1990 to 1999 marked the production of 45 articles. Between the years 1973 and 1989, only 9 articles were published. Figure 3 portrays the trend of articles published according to years of publication.\n\nThe 100 articles in the list were published by 45 peer-reviewed academic journals. The Journal of the American Medical Association (JAMA) published a total of 15 articles present on this list; followed by the American Journal of Psychiatry, which published 9 articles. Four journals, the International Migration Review, the Journal of the American Academy of Child & Adolescent Psychiatry, the Lancet, and the Social Science & Medicine, issued 5 articles each. Figure 4 presents each journal with their respective numbers of cited articles.\n\nA total of 294 investigators authored these 100 articles. The most cited author is Richard Francis Mollica, with a total of 9 articles credited to his name. A total of 25 articles were authored by only one author; followed by 21 articles and 12 articles that were authored by two authors and four authors, respectively. A total of 38 articles had only one institutional affiliation; followed by 24 articles and 14 articles that had two and three institutional affiliations, respectively. The majority of the articles (n=71) either completely or partly originated within the United States of America, while 23 articles and 14 articles either completely or partly originated from the United Kingdom and Australia, respectively. Furthermore, psychiatry, psychology, and public health constituted the top three fields of affiliation. Figure 5 depicts the association between the most-cited articles and authors’ fields of affiliations.\n\nThe top three research areas include: i) Psychiatry; ii) General and Internal Medicine; and iii) Psychology. Similarly, the discipline, Mental Health and Behavioral and Addiction Medicine was the most frequent, followed by Medical Sciences, and Humanities and Social Sciences. Figure 6 illustrates the number of cited articles and research areas. Figure 7 presents the the number of cited articles according to their academic disciplines.\n\n\nDiscussion\n\nThis bibliometric evaluation allowed a multi-dimensional outlook on the conditions of refugee populations across the globe through the collation of relevant peer-reviewed research journal articles. This specialized form of assessment has resulted in a multi-disciplinary compendium of publications on the subject. Along with studying the various components of this compendium, it was also essential to craft the category of “discipline” by aggregating comparable areas of research. This categorization allowed a better understanding of the focus of research. Moreover, the primacy of psychiatry as the most focused research area, as well as the dominance of the discipline of Mental Health and Behavioral and Addiction Medicine demonstrated the extensive evaluation and reproduction of the psychiatric and psychological aspects of refugee communities; 5 of the top 10 articles on this list related to this discipline. Although refugee resettlement is a complex process, this study revealed that the most thoroughly referenced feature of this process pertains to its mental health implications.\n\nContrary to our initial expectations, the discipline of Public, Environmental and Occupational Health and its subsection of refugee and immigrant health did not rank amongst the top three explored disciplines. Aside from this discipline, the least researched disciplines were the following: i) government, law, and public policy; and ii) natural and biomedical sciences. Thenceforth, it is incumbent on researchers to redirect the focus of research and materialize more publications on these least surveyed disciplines.\n\nAs far as the prevailing trends in the current literature are concerned, it is essential to understand the turnover of research publications over successive time periods. While only 9 articles out of this list were published between the years 1973 and 1989, the two decades 1990–1999; and 2000–2010 produced 45 articles and 46 articles, respectively. Along with denoting a surge in the turnover of publications, these time periods also characterized two decades of stagnancy and lack of turnover growth. Similarly, the most cited article in this study was published in 1997, almost two decades ago. The most recent well-cited article was published in the year 2009. As the ninth most cited article on the list, with over 300 citations, it was titled, “Association of torture and other potentially traumatic events with mental health outcomes among populations exposed to mass conflict and displacement: a systematic review and meta-analysis,” and was published in the Journal of the American Medical Association (JAMA). Two of the most overall recent articles were published in 2010. No articles published between 2011 and 2016 appeared on this list.\n\nAlthough most articles (n=57) were empirical evidence-based investigations that employed quantitative analyses, 43 articles did not utilize any form of statistical analyses and 48 articles had no defined sample sizes. The most well-cited academic review article is also the most-cited article of the entire list. As academic review articles (n=46) represented a large fraction of all most-cited articles, it can be concluded that the refugee populations were examined and inspected through multiple perspectives. The sizable volume of the constituent review articles uncovered recent advances and discoveries, highlighted emerging debates, and identified lacunae in the current literature. Thus, it can be ascertained that prior publications on refugees had been appropriately outlined and contextualized by these review studies.\n\nPsychiatry dominated the fields of affiliations of authors in the most-cited list. The most cited author, with nine publications, is a psychiatrist, Richard Francis Mollica—the Director of the Harvard Program in Refugee Trauma (HPRT) at the Massachusetts General Hospital and Harvard Medical School.\n\nThe primary limitation is associated with the word, “refugee” (the keyword used in the search bar of WOS), since the keyword entered into any search engine determines the validity of the generated list of results. This study’s exclusive focus on refugee populations (rather than immigrants, foreigners, expatriates, or aliens) might not be comprehensive or inclusive enough—it is possible that these distinct terms may or may not have been used to define refugee populations in certain studies. For instance, certain studies may not have specifically studied refugee populations, but rather may have examined them as an aggregate of various populations. Secondly, the inclusivity of the generated list of results is contingent on the proficiency of WOS as a searchable database and citation index.\n\n\nConclusion\n\nThis bibliometric evaluation allowed a multi-dimensional outlook on the conditions of refugee populations across the globe, through collation of relevant peer-reviewed research journal articles. This specialized form of assessment has resulted in a multi-disciplinary compendium of publications on the subject. The top cited articles were published from the developed countries than the low and middle income countries where a high percentage of refugee population has settled.\n\n\nData availability\n\nDataset 1. Bibliometric of 100-most-cited articles on refugees. DOI: http://doi.org/10.5256/f1000research.15106.d20739412.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe authors declared that no grants were involved in supporting this work.\n\n\nReferences\n\nUnited Nations High Commissioner for Refugees: Global Trends: Forced Displacement in 2015. Geneva; 2016. Reference Source\n\nUnited Nations High Commissioner for Refugees: UNHCR - Health. Reference Source\n\nRefugee Studies Center: Annual Report: 2014–2015. Oxford; 2016. Reference Source\n\nResearch@library: Bibliometrics: an overview. Leeds. 2014. Reference Source\n\nUWM Libraries Research and Course Guides: Citation Analysis - Faculty and Staff Guide to the UWM Libraries. Milwaukee, 2016. Reference Source\n\nO’Keeffe ME, Hanna TN, Holmes D, et al.: The 100 most-cited original articles in cardiac computed tomography: A bibliometric analysis. J Cardiovasc Comput Tomogr. 2016; 10(5): 414–23, [cited 2017 May 24]. PubMed Abstract | Publisher Full Text\n\nDolan RS, Hanna TN, Warraich GJ, et al.: The top 100 articles in the radiology of trauma: a bibliometric analysis. Emerg Radiol. 2015; 22(6): 667–75, [cited 2017 May 24]. PubMed Abstract | Publisher Full Text\n\nGong B, Mohammed ME, Nicolaou S, et al.: Diagnostic Imaging in Disasters: A Bibliometric Analysis. Disaster Med Public Heal Prep. 2018; 12(2): 265–277, [cited 2017 Oct 8]. PubMed Abstract | Publisher Full Text\n\nUsman MS, Siddiqi TJ, Khan MS, et al.: A Scientific Analysis of the 100 Citation Classics of Valvular Heart Disease. Am J Cardiol. 2017; 120(8): 1440–9, [cited 2017 Oct 8]. PubMed Abstract | Publisher Full Text\n\nNasir SAR, Gilani JA, Fatima K, et al.: Top 100 Most-Cited Articles on Spontaneous Intracerebral Hemorrhage: A Bibliometric Analysis. World Neurosurg. 2018; 110: 445–449.e6. [cited 2017 Oct 8]. PubMed Abstract | Publisher Full Text\n\nClarivate analytics: Web of Science platform. Reference Source\n\nKhosa M, Waqas A, Javaid M, et al.: Dataset 1 in: Bibliometrics of the 100 most-cited articles on refugee populations. F1000Research. 2018. Data Source"
}
|
[
{
"id": "36652",
"date": "13 Aug 2018",
"name": "Syed Raza Shah",
"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\nAuthors have done a great job in writing this manuscript trying to find published materials and research endeavors on the topic of refugees. However, I have a few suggestions to add:\nNo data to compare with previously done studies on refugee population in the discussion section? It would be highly appreciated if comparable studies on the current subject could be added in a separate paragraph in the discussion section. Any particular reasons why “only 9 articles out of this list were published between the years 1973 and 1989, the two decades 1990–1999; and 2000–2010 produced 45 articles and 46 articles, respectively.” Please explain the reasons you think of this disparity? Inclusion variable should be well explained in the methods section. Minor grammatical errors should be corrected\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "37116",
"date": "20 Aug 2018",
"name": "Samy A Azer",
"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\nI read with great interest the above titled article. However, there are major problems in the submitted manuscript.\n\nAbstract: under background. State that you are searching the top-cited articles (rather than determine …of the published materials and research endeavors on the topic of refugees. Add the rationale of the study, and your research question instead of the long background. Abstract: Method: only stated in 2 lines. You need to add the search words used and how the study was designed and implemented. Focus on important key issues. Results: Results. Is the 203 is the mean? Add the IQR. Needs to be organized. The last sentence in results could be stated earlier Abstract conclusions: Is the research about “conditions” of refugees? Correct grammatical errors. It should be strengthened and linked with the overall concept of refugees included in the study. Introduction: The first two paragraphs should be written in a way that reflects the purpose of searching top-100 cited articles on refugees. Item (i) for example and others, in the third paragraph are not fitted with the purpose of this study. Introduction: The fourth paragraph should include the rationale of the study, why this study is needed, and the research questions. Methods: Explain why the Web of Science was selected as the search engine rather than Scopus or others.\n\n“Methods: Top line in the second column: What are these “All Databases”? I thought you are using Web of Science only, as stated, please explain and amend. The method on top cited articles has been described by Azer SA 2015,2016,2017, and Azer SA and Azer S 2016,2018 and the research of Lin CL in this area. You should acknowledge earlier research in the field of bibliometrics. The search was conducted in June 2016 over 2 years ago. The citation numbers and possible some articles are not among top 100. For example, the first article by Bery JW stated to be cited 1493 times, which current check of the Web of Science by the reviewer showed a total citation of 2,355. Higher by 862. Other changes have been noticed in other articles number of citations. Please review and update your work. Again the items or parameters of search of articles have been described earlier in Azer’s research . It is not clear if the authors included research articles, reviews, articles, letters to the editors, commentaries, editorials, monographs, or limited their findings on original research? The authors must include inclusion and exclusion criteria. Table 1: the title started “research areas”, again, where these all research articles? I doubt. Please be specific. Which of these were research, and which were articles, or reviews or…. These categories should be defined. I suggest that all the top-cited articles should be cited to the reference list of the manuscript. Also add the reference numbers in the Table for each category rather than giving the number of articles only. Results: Is this a catalog of the 100-most cited, or an appendix or a table summarizing the top-cited articles. Include them as references (12-111) and add them to the list of references, if possible. Study design: The title should be under methods rather than under results. How did you agree on research or article or commentary etc.? How these results were reached? Explain this under methods. Did you measure the inter-rater agreement? Figure 1: Is redundant. Not clear why the authors are measuring Core collection vs all databases? Figure 1 shows no differences across all categories. Should be omitted. All databases is fine. Journals: could be listed in a table, and the references numbers added in the second column. The subtitle: “Authors, fields of affiliations, etc.: is too crowded and not clearly written. Figure 4: Not clear. The yellow colour is making it difficult to read. Figure 6 could be omitted or included as a column with the table suggested for journals under item 16. Discussion needs to be strengthened and discussed against other important studies in the literature. Why the year 2004 had a higher number of cited publications. You need to explain in regard to historical events, topics raised in these articles, and the meaning of the findings, rather than the numbers. Conclusions should be strengthened. Figure 7 not needed. Please omit. References: add the references in the table (12-111).\n\nexamples of citations to read: Azer et al, 20161, Azer et al, 20162; Azer et al, 20173; Azer et al, 20184\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-878
|
https://f1000research.com/articles/7-877/v1
|
22 Jun 18
|
{
"type": "Research Article",
"title": "Mothers’ adherence to optimal infant and young child feeding practices in Uganda: a cross-sectional study",
"authors": [
"Zabinah Nabirye",
"Frank Kiwanuka",
"Zainah Nakaye",
"Ivan Kamurasi",
"Agbele Alaba Tolulope",
"Zabinah Nabirye",
"Zainah Nakaye",
"Ivan Kamurasi",
"Agbele Alaba Tolulope"
],
"abstract": "Background: The benefits of adherence to optimal infant and young child feeding (IYCF) to both the mothers and their infants below two years are well documented. However, compliance to optimal IYCF practices has been noted to vary in different settings. This study sought to establish factors influencing mothers’ adherence to optimal infant and young child feeding practices for babies below two years in Mpigi town council- Mpigi District. Methods: The study was a cross-sectional study carried out among 264 mothers of babies between six months to two years of age attending postnatal care units of health facilities in Mpigi town council, Uganda: Mpigi Health Center (HC) IV and Kyaali HC III. Purposive sampling method was used to select the health centers while simple random sampling was then used to select the sample from the selected centers. A self-administered questionnaire was used to collect data. Data entry and analysis was performed using SPSS version 16. Results: 264 participants were invited to participate in the study, 100% of these fully completed the survey. The majority of the mothers were aged 20 to 34 years (80.3%). After scoring each participant using the four characteristics which included: initiation of breastfeeding within the first hour following birth, exclusive breastfeeding up to 6 months followed by continued breastfeeding with appropriate complementary foods upto 2 years and beyond, the majority of the participants were adherent (79.6%) to IYCF practices while 20.4% were non adherent to IYCF practices. Conclusion: A good estimate of adherence to optimal IYCF practices was revealed in this study. Sustaining well-established policies to support IYCF programmes is recommended to maintain optimal IYCF practices.",
"keywords": [
"Adherence",
"Mothers",
"Infant and Young Child Feeding",
"Uganda",
"Africa",
"Low and Middle-Income countries"
],
"content": "Introduction\n\nOptimal Infant and Young Child Feeding (IYCF) is crucial for survival across the entire childhood continuum. The World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) have recommended initiation of exclusive breastfeeding from the time of birth up to 24 months of age1. Optimal feeding of infants from 0 to 5 months has been associated with significantly lower risk of all-cause and related mortality compared to partial or not breastfeeding at all1,2.\n\nEstimates suggest that optimal IYCF practices could prevent approximately 12% of mortality in the under 5’s annually. This accounts for about 800,000 lives in Low and Middle-Income Countries (LMICs)3.\n\nEarly initiation of optimal IYCF within the first hour following birth in the form of exclusive breastfeeding up to 6 months followed by continued breastfeeding for up to 2 years and beyond with appropriate complementary foods after completion of 6 months is the gold standard for optimal IYCF1,4.\n\nSub-Saharan African countries have one of the highest prevalence of breastfeeding at one year, however only 37% of infants less than 6 months are exclusively breastfed3,5. Various factors have been cited for the low optimal IYCF practices, including; socio-economic status6,7, home births, culture and poorly implemented and monitored strategies of regulating the marketing of breast-milk substitutes, feeding bottles and teats6–10. Other cited factors to non-adherence to optimal IYCF include; education levels, health providers’ practices and market pressures to use breast milk substitutes and other nutritional supplements3,11.\n\nThe Uganda demographic and health survey12 noted that only 63% of infants younger than six months are exclusively breast fed and complementary foods are not introduced in a timely fashion for all children; 33% of children under five are stunted, with only 6% of children aged 6–23 months being appropriately fed12.\n\nThis study sought to establish the factors influencing mothers’ adherence to optimal infant feeding practices for infants below the age of two years in Mpigi town council health facilities, Mpigi District, Uganda. Findings from this study provide country specific data on the extent of adherence to optimal IYCF practices. Such evidence is seminal to policy makers to sustain IYCF policies and programmes.\n\n\nMethods\n\nThis was a cross-sectional study carried out among mothers attending postnatal care units of health facilities in Mpigi town council, Uganda from May 2016 to August 2016. Mpigi district is located in the Central region of Uganda, 32 Kilometers West of Kampala Capital City along Masaka road. The study was conducted in two health centers: Mpigi Health Center (HC) IV and Kyaali HC III. Mpigi HC IV is the largest health facility in Mpigi District.\n\nMixed sampling methods were used. Purposive sampling was used to select the health centers. Simple random sampling was then used to select the sample from the selected centers. Mothers and caretakers of babies six months to two years of age were included in the study while as those who were participating in interventional breastfeeding studies were excluded from the study.\n\nData was collected using a structured questionnaire (Supplementary File 113). The questionnaire was developed through literature review. The tool consisted of three sections; section one assess socio-demographic data, section two assessed maternal and infant data, while section three assessed Healthcare related data. The tool was validated by 5 experts. The tool was pre-tested at Mulago National referral Hospital before data collection. The questionnaires were administered by research assistants to participants during Young Child Clinics (YCC) days at the selected centers. Data collection was carried out as the mothers waited to see the Healthcare providers in the queues.\n\nWe defined optimal infant feeding as initiation of optimal IYCF within the first hour following birth in the form of exclusive breastfeeding up to 6 months followed by continued breastfeeding for up to 2 years and beyond with appropriate complementary foods after completion of 6 months is the gold standard for optimal IYCF1,4.\n\nWe established a scale to assess IYCF using four questions that cover the definition of IYCF, these included: (i) Breastfed within the first hour after delivery (ii) Were you able to breastfeed up to 6 months? (iii) Continue to breastfeed after introducing others foods to baby and (iv) for how long? For question i, ii and iii, option yes was score 1 while option no was scored zero. For question 5, option (a) was scored 1, option two was scored 2 while option three was scored 3 (Table 2). This lead to a maximum adherence range of 6 and a minimum of 0. Adherence was then categorized into “non-adherence and adherence” subscales. Non-adherence ranged from 0 to 2 while adherence ranged from 3 to 6.\n\nThe data was entered, cleaned and analyzed by SPSS version 16.0 statistical packages. Descriptive summary statistics were used to describe the respondents’ characteristics while appropriate statistical analysis was performed to assess the factors associated with adherence to infant feeding.\n\n\nResults\n\n264 participants were invited to participate in the study, 100% of these fully completed the survey.\n\nMajority of the mothers were aged 20 to 34 years (80.3%). Most of the mothers had at least 1 to 3 children (70.8) (Table 1) and had currently delivered female babies compared to male ones. The most common level of education attained by the mothers was primary education. With regards to occupation and average monthly income, most mothers were housewives (63.3%)and earned an average monthly income of 100,000 to 300,000 Uganda Shillings (equivalent to $26 to $78 at time of publication) (Table 1).\n\nSource: primary data. BF: Breastfeeding\n\nPractices that reflect mother’s adherence to IYCF are shown in Table 2 below. The majority of the mothers (71.2%) reported that they initiated breastfeeding within the first hour following delivery. Most mothers (80.7%) also reported that they were able to breast upto 6 months (Table 2).\n\nSource: primary data\n\nRegarding breastfeeding alongside complementary feeding after six months, almost all mothers (95.1%) reported that they were able to continue breastfeeding with complementary feeding after six months. Among those that were able to do so (n=251), the majority reported that they were able to breastfeed alongside complementary feeding upto one year (61%), followed by those who reported they were able to breastfeed alongside complementary feeding for two years and beyond.\n\nAfter scoring each participant using the four characteristics in Table 2, the majority of the participants were adherent (n=210, 79.6%) to IYCF practices while 54 (20.4%) were non-adherent to IYCF practices (Table 2).\n\nThe majority of the mothers reported that they were healthy during the antenatal care visits (ANC) period (61%). Reasons reported to enable those who embraced optimal IYCF practices included; had plenty of breast milk (63.5%), enough time with the baby (14.2%) and encouragement from HCWs (22.3%). Those who were unable to abide by optimal IYCF practices (21.6%) reported reasons such as lack of optimal breast milk production (68.7%) and plenty of cow’s milks (31.3%) (Table 3).\n\nOC: Ovarian Cancer; HCWs: Healthcare Worker\n\nRegarding awareness on the time for introduction of complementary feeding, less than a third of the mother reported the optimum ITCF time for starting complementary feeding. Various benefits of optimal IYCF were reported by the respondents, these include; immunity for the baby, proper growth and development, bonding, protection of the mother from ovarian cancer (Table 3). Most mothers reported that it is possible to breastfeed up to 2 years while most mothers were not aware of how to prepare nutritious complementary food for the baby after 6 months. Reasons for failure to prepare nutritious food for the babies include; lack of awareness on how to prepare foods for the baby, lack of resources and some mothers reported that often the babies dislike the food (Table 3).\n\nAll mothers reported that they attended ANC. Information received during ANC visits include; positioning of the baby during breastfeeding, benefits of breastfeeding, how to feed an infant using a cup, importance of Colostrum and time of initiating breastfeeding. Majority of the mothers (53.4%) reported that they did not received practical demonstration during ANC visits. Different forms of support were reported to have been received immediately after birth. These include; initiating breastfeeding within one hour (40.1%), positioning of the newborn during breastfeeding and expressing of breast milk. Mothers reported that they received various health education talks during YCC visits, these included talks on; exclusive breast feeding, duration of breastfeeding and complementary feeding, and the benefits of breastfeeding (Table 4).\n\nSource: primary data. BF: Breastfeeding. CF: Complementary feeding\n\nAt bivariate analysis, age, number of children, sex of the infant, mothers level of education, occupation and average monthly income were significantly associated with optimal infant feeding (P<0.05) (Table 1).\n\nMaternal characteristics such as health status during labor, situations that enable mothers to breastfeed during first hour following delivery, perception of importance of first breast milk, awareness of the right age to introduce feeds, perceived benefits of breastfeeding, and knowing how many times in a day a baby is supposed to breastfeed were significant predictors of adherence to optimal infant feeding (P<0.05) (Table 3).\n\nHealth care related characteristics such as information received on infant feeding during ANC, the kind of information received, practical support received during ANC and counseling of mothers on infant feeding during ANC visits were significant determinants of adherence to optimal infant feeding (p<0.05). While receiving practical support was not significantly associated with optimal feeding (p=0.998) (Table 4).\n\nMultivariate analysis was performed to determine independent predictors of adherence to optimal IYCF. It was revealed that the age of the mother, level of education, average monthly of the family, health condition of the mother during pregnancy, awareness of the right age to introduce complementary food and having received information on IYCF during ANC visits were independent predictors of adherence to optimal IYCF practices.\n\n\nDiscussion\n\nIn this study, majority of the participants were adherent (79.6%) to IYCF practices while 20.4% were non-adherent to IYCF practices. This could be attributed to well-established strategies facilitating optimal IYCF in Uganda. Indeed, numerous campaigns and policies are in place in support of optimal IYCF in Uganda. These include, but are not limited to, activities involving frontline healthcare providers such as nurses and midwives and ANC services. Optimal IYCF was reflected by assessing practices reflecting IYCF practices such as: breastfed within the first hour after delivery, ability to exclusively breastfeed upto 6 months, continue to breastfeed after introducing others foods to baby and recommended duration of breastfeeding post introduction of complementary feeding14. This is also in line with the WHO published set of population-level IYCF indicators of appropriate feeding practices in children ages 6–23 months1.\n\nInfant feeding practices in this study were appropriate with regards to the WHO criteria for optimal IYCF practices. This finding is contrary to that reported in a study in Kenya in Urban informal settlements in Nairobi that reported inappropriate IYCF practices11. Optimal IYCF practices such as breast feeding have been reported to be high in Uganda compared to other Sub-Saharan African Countries like Nigeria and Niger9. This finding supports the findings of this study of a good estimate of optimal IYCF practices. There were significantly more mothers who reported to have initiated BF within an hour compared to less than half reported in a similar study in India15.\n\nIn context of characteristics of the sample, socio-demographic and economic characteristics of the participants in this study were similar to those reported form the recent population census in Uganda12. This makes our findings generalizable to the entire population with minimal limitations. The majority of the mothers were aged 20 to 34 years, this is attributed to the fact that majority of the reproductive population in Uganda lies within this age group. Most of the mothers had at least 1 to 3 children. This finding is in line with the fertility rate reported in Uganda, owing to the fact that on average, each mother in Uganda has on average up to 7 children. There were comparatively more female babies compared to male ones, this is in line to the female to male ratios reported in the recent population of Uganda. The most common level of education attained by the mothers was primary education. Indeed, it has been estimated that at least majority of the Ugandan population has attained primary education12. With regards to occupation and average monthly income; most mothers were housewives and earned an average monthly income of 100,000 to 300,000 Uganda Shillings. This is in line with the economic characteristics of majority of the Ugandan population revealed in the recent census.\n\nThe reported breast feeding rate in this study is higher than that reported in World Breastfeeding Initiative: Uganda Assessment Report that showed that 40% of the children under 6 months of age have not been exclusively breastfed and only 52.5% were breastfed within an hour of birth.\n\nReasons reported to enable those who embraced initiation of BF with the first hour included; had plenty of breast milk production, enough time with the baby and encouragement from HCWs. Indeed, strong individual champions such as HCWs who encourage IYCF such as initiating BF in the first hour have been reported elsewhere as catalysts to optimal IYCF practices14,16. Engagement of the mother by providing enough time with the baby would in turn stimulate milk production. This could in part solve the problem of lack of breast milk production which was one of the barriers to initiating BF in the first hour. In addition, controlling for pain after delivery could go a long way in giving mothers more comfort post-delivery so that they could start enjoying happy times with the new born.\n\nIn context of initiating complementary feeding, few mothers in this study were aware of the right timing. Emphasis on the right timing and preparation of complementary feeds during ANC and PNC visits needs to be hastened in facility and community based programmes. Awareness of the benefits of BF was significant in this study.\n\nAttendance of ANC was overwhelmingly high in this study. This could be attributed to programmes such as the Baby Friendly Hospital Initiative (See World Breastfeeding Trends Initiative, Uganda Report 2015). However; practical demonstrations need to be incorporated in the ANC visits. The majority of the mothers reported that they did not received practical demonstration during ANC visits. These could improve feeding behaviors of mothers and in turn sustain optimal IYCF practices upto a recommended time.\n\nCorrelates to adherence to optimal IYCF included socio-demographic factors such as; age of the mother, level of education and average monthly income of the family. A similar study that assessed the impact of Socio-demographic and health service factors on breastfeeding in SSA using demographic health survey data reported that educational attainment was a significant predictor of IYCF practices such as exclusive breast feeding8. Health condition of the mother during pregnancy, awareness of the right age to introduce complementary food and having received information on IYCF during ANC visits were independent predictors of adherence to optimal IYCF practices. We argue that to promote and sustain IYCF practices, it is imperative to create public awareness about appropriate complementary feeding, its timing and providing IYCF counseling support systems for mothers.\n\n\nLimitations of the study\n\nWhile as this study presents important information on the extent of adherence to IYCF practices in Uganda, the data was obtained using self-report from the mothers. Such information gained from self-report is liable to bias as we cannot ascertain the truthfulness of information provided. In addition, the findings of this study are obtained from a semi-urban setting, as such these finding could not be generalizable to mothers in urban settings who could have different socio-demographic and economic characteristics.\n\n\nConclusion\n\nA significantly good adherence to optimal IYCF practices was revealed in this study. Promotion and sustaining well-established policies to support IYCF programmes is recommended to maintain optimal IYCF practices. In addition, practical demonstrations need to be incorporated in the ANC visits.\n\n\nEthical approval and consent\n\nThe study proposal was approved by the Institute of Public Health and Health Management of Clarke International University. Permission to collect data from health facilities was granted by the District Health Officer (DHO) and the administrator in-charge of the health facilities. It was strictly voluntary to participate in the study and written informed consent was obtained from participants at all times.\n\n\nData availability\n\nThe data underlying this study is available from figshare. Dataset 1: A study on Factors influencing mothers’ adherence to optimal and young child feeding practices for babies below two years in Mpigi town council, Mpigi district, Uganda. https://doi.org/10.6084/m9.figshare.6470705.v117\n\nThis dataset is available under a CC by 4.0 license\n\nData points containing ‘999’ or ‘99’ denote missing variables. Additionally where respondents have answered yes or no, their answer in some cases prevents them from answering the next question and their for the subsequent analysis. ‘999/99’ was used in case also.\n\n\nAuthor endorsement\n\nMasoomeh Imanipour confirms that the author has an appropriate level of expertise to conduct this research, and confirms that the submission is of an acceptable scientific standard. Masoomeh Imanipour declares they have no competing interests. Affiliation: Faculty of Nursing and Midwifery, Tehran University of Medical Sciences (TUMS), Tehran, Iran.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no funding was obtained in support of this work.\n\n\nAcknowledgements\n\nThe authors wish to extend their deep appreciation to Mr. Alege JB, Dean of the Institute of Public Health and Management of Clarke International University.\n\n\nSupplementary material\n\nThe questionnaire used in the study is available from figshare Supplementary File 1: Mothers’ adherence to optimal infant and young child feeding practices in Uganda: a cross-sectional study https://doi.org/10.6084/m9.figshare.6468653.v113\n\nAvailable under a CC by 4.0 license\n\n\nReferences\n\nJones AD, Ickes SB, Smith LE, et al.: World Health Organization infant and young child feeding indicators and their associations with child anthropometry: a synthesis of recent findings. Matern Child Nutr. 2014; 10(1): 1–7. PubMed Abstract | Publisher Full Text\n\nSankar MJ, Sinha B, Chowdhury R, et al.: Optimal breastfeeding practices and infant and child mortality: a systematic review and meta-analysis. Acta Paediatr. 2015; 104(467): 3–13. PubMed Abstract | Publisher Full Text\n\nBlack RE, Victora CG, Walker SP, et al.: Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet. 2013; 382(9890): 427–51. PubMed Abstract | Publisher Full Text\n\nTiwari S, Bharadva K, Yadav B, et al.: Infant and Young Child Feeding Guidelines, 2016. Indian Pediatr. 2016; 53(8): 703–13. PubMed Abstract\n\nVictor R, Baines SK, Agho KE, et al.: Determinants of breastfeeding indicators among children less than 24 months of age in Tanzania: a secondary analysis of the 2010 Tanzania Demographic and Health Survey. BMJ Open. 2013; 3(1): pii: e001529. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIssaka AI, Agho KE, Page AN, et al.: Determinants of suboptimal complementary feeding practices among children aged 6-23 months in seven francophone West African countries. Matern Child Nutr. 2015; 11 Suppl 1: 31–52. PubMed Abstract | Publisher Full Text\n\nOgbo FA, Page A, Agho KE, et al.: Determinants of trends in breast-feeding indicators in Nigeria, 1999-2013. Public Health Nutr. 2015; 18(18): 3287–99. PubMed Abstract | Publisher Full Text\n\nOgbo FA, Page A, Idoko J, et al.: Trends in complementary feeding indicators in Nigeria, 2003-2013. BMJ Open. 2015; 5(10): e008467. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOgbo FA, Agho KE, Page A: Determinants of suboptimal breastfeeding practices in Nigeria: evidence from the 2008 demographic and health survey. BMC Public Health. 2015; 15: 259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNankumbi J, Muliira JK: Barriers to infant and child-feeding practices: a qualitative study of primary caregivers in Rural Uganda. J Health Popul Nutr. 2015; 33(1): 106–16. PubMed Abstract | Free Full Text\n\nMacharia TN, Ochola S, Mutua MK, et al.: Association between household food security and infant feeding practices in urban informal settlements in Nairobi, Kenya. J Dev Orig Health Dis. 2018; 9(1): 20–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUganda Bureau of Statistics: The National Population and Housing Census 2014–Main Report. 2016. Reference Source\n\nFrank K, Zabinah N: Mothers’ adherence to optimal infant and young child feeding practices in Uganda: a cross-sectional study. 2018. Publisher Full Text\n\nPuri S: Transition in Infant and Young Child Feeding Practices in India. Curr Diabetes Rev. 2017; 13(5): 477–481. PubMed Abstract | Publisher Full Text\n\nNguyen PH, Avula R & Menon P: Estimates of child deaths prevented from scaling up infant and young child feeding interventions in India 2016-2025. Matern Child Nutr. 2018; 14(S2).\n\nShaker-Berbari L, Ghattas H, Symon AG, et al.: Infant and young child feeding in emergencies: Organisational policies and activities during the refugee crisis in Lebanon. Matern Child Nutr. 2018. PubMed Abstract | Publisher Full Text\n\nFrank K, Zabinah N: Dataset: A study on Factors influencing mothers’ adherence to optimal and young child feeding practices for babies below two years in Mpigi town council, Mpigi district, Uganda. 2018. Publisher Full Text"
}
|
[
{
"id": "38277",
"date": "24 Sep 2018",
"name": "Seema Puri",
"expertise": [
"Reviewer Expertise Infant and young child nutrition",
"clinical nutrition"
],
"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 article brings out the IYCF practices in a town of Uganda. It reports the profile of the mothers, their access to healthcare as well the adherence to suitable IYCF practices.\n\nWhile the authors have defined optimal IYCF as initiation of breastfeeding within the first hour following birth, exclusive breastfeeding up to 6 months followed by continued breastfeeding for up to 2 years and beyond with appropriate complementary foods after completion of 6 months, their scale to assess IYCF based on 4 questions does not cover these issues. For example, exclusive breastfeeding up to 6 months is not asked; instead the question is \"Were you able to breastfeed up to 6 months?” - the concept of exclusive breastfeeding is lost. Since this is the basis of defining adherence, the whole data categorization becomes faulty, and hence subsequent analysis to study correlates of IYCF also are biased.\n\nWhat do the P values displayed in Table 3 onwards imply? This was not given in the discussion.\n\nAlso in Table 4, many of the questions are multiple responses and therefore cannot add up to 100% nor can any statistical test be applied to them.\n\nNo details of multivariate analysis given.\n\nLanguage errors need to be corrected.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "45357",
"date": "20 Mar 2019",
"name": "Olukemi K. Amodu",
"expertise": [
"Reviewer Expertise Genetics/Public Health"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study was designed to identify the factors that influence mother’s adherence to optimal infant and young feeding (IYCF) practices in Uganda. The authors used four characteristics: initiation of breastfeeding within the first hour following birth; exclusive breastfeeding up to 6 months; continued breastfeeding with appropriate complementary foods up to 2 years and beyond to assess IYCF. The study however has some deficiencies.\nThe question, \"Were you able to breastfeed up to 6 months?” did not take into consideration adherence to breastfeeding for 6 months. Adherence was a key factor in this study and this deficiency weakened the design and results.\n\nThe other question that would have also helped is “when did you introduce complementary feeding?”. This would have helped clarify results.\n\nThe statistics were not extensive. The method section needs further explanation.\n\nThe discussion is weak and not well written. The results were not well discussed. There is need to properly link and discuss the findings from demographic information to other findings. Findings from other studies are not discussed.\n\nThere are grammatical errors in the article. Run-off statements and unclear sentences need editing. Engaging editors’ service will be helpful to the manuscript.\n\nData from 264 respondents cannot be generalized for a country with over 44 million population. There is the need to review this in the discussion.\n\nThe meaning and importance of P-value in each of the tables presented needs explanation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "45356",
"date": "25 Mar 2019",
"name": "Sajid Bashir Soofi",
"expertise": [
"Reviewer Expertise Public Health scientist with special interest in Nutrition"
],
"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 was conducted to determine factors contributing towards mothers’ adherence to optimal infant and young child feeding practices for children under 2 years. The study investigators adopted cross sectional design and purposive sampling methods. The data was collected at randomly-selected health centres. The authors reported baseline SES characteristics of study participants and mothers’ characteristics. They also reported mothers’ practices on few IYCF indicators. They presented simple descriptive analysis.\n\nThe authors did not provide analysis and discussion regarding the main objective of the study: factors influencing mothers’ adherence to optimal IYCF practices for babies below two years. This study has limited scope and interest of audience at a larger scale. Further, the data collection tools used are not appropriate especially the question asked to determine EBF rates. The language needs some revision. The analysis section needs more elaboration.\n\nThis paper wouldn’t be accepted for indexing as there are major issues in the paper and the authors need to do further analysis and then draw conclusions.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-877
|
https://f1000research.com/articles/7-875/v1
|
22 Jun 18
|
{
"type": "Research Article",
"title": "The nociception level index (NOL) response to intubation and incision in patients undergoing video-assisted thoracoscopic surgery (VATS) with and without thoracic epidural analgesia. A pilot study.",
"authors": [
"Laurent A. Bollag",
"Srdjan Jelacic",
"Carlos Delgado Upegui",
"Cynthia Wu",
"Philippe Richebe",
"Srdjan Jelacic",
"Carlos Delgado Upegui",
"Cynthia Wu",
"Philippe Richebe"
],
"abstract": "Background: The PMD100™ (Medasense Biometrics Ltd., Ramat Yishai, Israel) is a novel non-invasive nociception monitor that integrates physiological parameters to compute a real-time nociception level index (NOL) in the anesthetized patients. Thoracic epidural analgesia provides effective analgesia and improves surgical outcomes. Side effects include sympathectomy, hypotension, changes in skin temperature and a decreased cardiac accelerator fiber tone. The purpose of this pilot study was to evaluate changes in NOL values after incision in patients with and without epidural analgesia.\n\nMethods: Half of the patients scheduled for Video-Assisted Thoracoscopic Surgery (VATS) received a thoracic epidural catheter, placed and tested 2h before surgery and activated prior to incision. The other half of the patients received i.v. fentanyl (1 mcg/kg) five minutes before incision. Anesthesia and analgesia were maintained in a standardized manner. NOL and heart rate (HR) were compared before and after the nociceptive stimuli intubation and skin incision. Results: NOL significantly increased in all patients after intubation by 10.2 points (CI: 4.5-16.0; p=0.002) as well as HR by 9 beats per minute after intubation in all patients (CI: 3.3-15.6; p=0.01). After incision, in patients without epidural analgesia the NOL increased by 13.9 points (CI: 7.4-20.3; p=0.0001), compared to 5.4 points (CI: -6.3-17.1; p=0.29) in patients with epidural analgesia. HR did not significantly vary after incision in both groups. The area under the curve of delta NOL and delta HR variations after incision were significantly different (p<0.05) between groups and delta NOL variations were significantly different from baseline values but not the delta HR variations. Conclusions: This pilot study suggests that the PMD100™ Monitor may be a useful tool to evaluate the efficacy of an intraoperative thoracic epidural analgesia. Clinical Trial Registry Number: ClinicalTrials.gov record ID: NCT01978379 registered 10/25/2014.",
"keywords": [
"intraoperative nociception monitoring",
"epidural analgesia",
"multiparametric nociception monitoring",
"lung surgery",
"regional anesthesia",
"general anesthesia"
],
"content": "Introduction\n\nTo date, there are no standards to assess intra-operative nociception in a quantitative manner. The PMD100™ (Medasense Biometrics Ltd., Ramat Gan, Israel) is a non-invasive nociception monitor. The device integrates multiple physiological, parameters, including heart rate (HR), heart rate variability, photo-plethysmogram wave amplitude, skin conductance level, and number of skin conductance fluctuations, movements and their time derivatives to compute a real-time nociception index, called NOL. All data is measured by a single finger-mounted probe. A monitor displays the nociception level (NOL) expressed as an index. The data is computed every five seconds. The index ranges from 0–100: a NOL of 0 suggests a very low sympathetic activation equal to a nociception free state, while a NOL of 100 suggests high sympathetic activation, representing high nociception levels.\n\nThe exact details of the utilized technology can be found elsewhere1.\n\nRecent clinical trials reported that the NOL index might have a higher sensitivity and specificity than heart rate (HR) and mean blood pressure (MBP) to detect painful stimulations such as intubation, incision and tetanic stimulations in patients under general anesthesia2,3.\n\nThoracic epidural catheters and neuraxial administered local anesthetics provide excellent intra- and post-operative analgesia4 and are frequently used in patients receiving thoracic surgery to improve post-operative pain and pulmonary recovery3. However, epidural side effects include a dermatomal reduction of sympathetic tone and could therefore affect NOL measurements, which are based on autonomous nervous system variables.\n\nThe purpose of this pilot study was to evaluate changes in NOL values after incision for trocar insertion for video-assisted thoracoscopic surgery (VATS) in patients with and without epidural analgesia. This would suggest that the NOL index is a reliable parameter to assess the epidural analgesia in the anesthetized.\n\nSecondary aims included changes in delta NOL index, in HR and delta HR values before and after two clinical stimuli, intubation and skin incision.\n\n\nMethods\n\nAfter institutional review board approval (#44750-B) informed consent was obtained from 20 eligible women and men, 18 years of age or older and of American Society of Anesthesiology Physical Status I-III, scheduled for video-assisted thoracoscopic surgery (VATS) at the University of Washington Medical Center in Seattle, USA. The trial was registered with ClinicalTrials.gov on the 11/07/2013 record ID: NCT01978379, the study period was from January till December 2014.\n\nNon-inclusion criteria included refusal, non-English speaking, chronic use of opioid analgesics, severe psychiatric disorder, history of previous thoracotomy, Body Mass Index (BMI)>40, and current beta-adrenergic blocking agent treatment. Inability to measure the NOL or HR led to exclusion from the data analysis.\n\nPatients in the epidural group (n = 10) received a mid-thoracic epidural catheter at either T7/8 or T6/7 dermatomal level, per surgical request. All other patients (n = 10) were considered as the no epidural group, again per surgical request. Randomization, due to surgical preference for epidural analgesia, was not possible.\n\nAfter successful epidural placement, at least 2 hours before surgery, all catheters were tested with 3ml 1.5% lidocaine (45mg) with 15µg epinephrine added, to confirm epidural catheter tip location.\n\nAll patients received 1–2 mg intravenous (i.v.) midazolam for anxiolysis before being transferred to the operating room where standard anesthesia monitoring was applied, including a five-lead electrocardiography, non-invasive arterial blood pressure, continuous pulse oximetry, and a bispectral index monitor (BIS) (Medtronic, Mansfield, MA, USA. Philips Bispectral Index (BIS®), BISx Power Link™, IntelliVue MP70, Philips, Netherland).\n\nAdditionally, the PMD100™ nociception monitor finger probe was connected to the middle finger on the blood pressure cuff free arm. NOL values were displayed following a 30 second calibration phase. The NOL data was recorded on a laptop using the Medasense biometric software. For each case, laptop times were adjusted to the time of the Anesthesia Information Management System (DocuSafe Version 7.2, Merge Healthcare, Chicago, IL, USA).\n\nGeneral anesthesia was standardized and induced with intravenous 1.5mg/kg lidocaine, 2 mcg/kg fentanyl, 1–2 mg/kg propofol, and 0.5mg/kg rocuronium. The time between i.v. administration of fentanyl and intubation was standardized for all patients and set at 5 minutes. Intubation was performed with a double lumen 37 or 39 Fr tube. Hypnotic depth was maintained with 1–1.5% end-expiratory sevoflurane concentration and was adapted to the age adjusted minimum alveolar concentration (MAC) at 1 to 1.2 to achieve target BIS values between 40–60. 100% inspired oxygen was used throughout the study period. Positive pressure ventilation mode was used in all patients. Hypotension, defined as a 20% decrease in systolic or diastolic blood pressure from the first measured values in the operating room or a mean arterial pressure below 65 mmHg, was treated with 100 µg phenylephrine bolus as needed. Then patients were repositioned laterally to facilitate the surgical approach to the lung cavity.\n\nTen minutes before skin-incision patients in the epidural group received an epidural bolus of 5ml 2% lidocaine (100mg), while patients in the no epidural group were administered an additional 1 mcg/kg fentanyl bolus, five minutes before incision. The study period ended five minutes after skin incision.\n\nNOL and HR were recorded every five seconds during the study period. Presented values span 90 seconds before and 190 seconds after intubation and incision, respectively.\n\nFor “Pre-stimulation” NOL and HR values, the values at “-10 seconds” before an event were used, after an Anova one-way analysis showed that pre-stimulation NOL and HR values were stable and did not vary during the preceding 90 seconds period (see Figure 1 and Figure 2).\n\nA) NOL variations after tracheal double-lumen intubation. There was no variation over time for the baseline NOL values prior to intubation (One Way ANOVA repeated measures, p=0.2312, F=1.523). There was a statistical significant variation in NOL absolute values after the tracheal intubation (One Way ANOVA repeated measures, p=0.0031, F=4,982). Dunnett’s multiple comparisons was used to compare each value to the control value at T minus 10sec. B) Delta NOL Baseline: There was no variation over time for the baseline delta NOL values prior to intubation (One Way ANOVA repeated measures: p=0.2312, F=1.523). There was a statistical significant variation in NOL absolute values after the tracheal intubation (One Way ANOVA repeated measures: p=0.0031, F=4.982). Dunnett’s multiple comparisons was used to compare each value to the control value at T minus 10sec. C) Heart Rate: Baseline without variation (One Way ANOVA repeated measures: p=0.5807, F=0.5187). After Intubation: significant variation in HR (One Way ANOVA repeated measures: p=0.0056, F=4,846). Dunnett’s multiple comparisons was used to compare each value to the control value at T minus 10sec. D) Delta Heart Rate: Baseline without variation (One Way ANOVA repeated measures: p=0.5807, F=0.5187). After Intubation: significant variation in delta HR (One Way ANOVA repeated measures: p=0.0056, F=4.846). Dunnett’s multiple comparisons was used to compare each value to the control value at T minus 10sec. For A, B, C and D significant differences with baseline values of each parameter are shown by: #: p<0.005, ##: p<0.001, ###: p<0.0001.\n\nA) NOL variations after incision. The baseline values (before incision) were analyzed using One Way ANOVA and showed no intra-group time effect (p=0.8148), meaning the NOL baseline values for each group were stable. Baseline NOL values were significantly different between groups before incision (p=0.0354, *). After incision: One Way ANOVA for repeated measures showed no significant difference in NOL values after incision in the epidural group. In the no-epidural group, one way Anova analysis and Dunnett’s multiple comparisons showed a significant increase of NOL values after incision (#, ##). Two Ways Anova (Time x Epidural) showed an Interaction (p=0.0015) and Bonferroni’s multiple comparisons showed a significant difference between epidural versus no- epidural at 10sec, 30sec and 50sec after incision (p<0.05). B) Delta NOL baseline values showed no difference between groups and were stable Two Ways Anova (Time x Epidural): Interaction p=0.1048, Epidural p=0.5891, Time p=0.4010). After incision, delta NOL increased significantly in the no-epidural group (Dunnett’s multiple comparison, p<0.05; #, ##) but not in the epidural group. Bonferroni’s multiple comparison between epidural versus no-epidural showed no statistical significant difference. C and D) HR or Delta HR baselines were stable prior to incision (One Way Anova, no time effect in each group) and not significantly different between groups. After incision, HR did not significantly change in both groups when compared to baseline values (One Way ANOVA, Dunnett’s multiple comparison for time, no time effect in each group). Bonferroni’s multiple comparison comparing epidural versus without epidural showed no significant difference between two group after incision for HR and delta HR values. For A, B, C and D significant differences with baseline values of each parameter are shown by: #: p<0.005, ##: p<0.001.\n\nFor “Post-Stimulation” values NOL or HR were averaged every 20 seconds for 180 seconds after an event. In Figure 1 and Figure 2, the dots represent values averaged over 20-second periods. NOL and HR values are presented as means and standard deviations.\n\nPre and post-stimulation data was compared using Anova two-ways analysis and post-hoc analysis, Dunnet’s test was performed to evaluate the changes over time for each parameter, as well as the difference between groups (Figure 2).\n\nDelta NOL/HR signals were calculated by subtracting the baseline NOL and HR values from the signal. The baseline value was defined as the average NOL or HR value over the last 60 seconds before an event (-60 to 0 sec).\n\nFor delta NOL and delta HR, the area under the curve (AUC) was calculated in the time window from the stimulus, at 0 sec, to 180 seconds and compared with the Student t-Test.\n\nA p-value of less than 0.05 was considered significant to reject the null hypothesis, which was that average NOL and HR values before and after the two events as well as their changes in patients with and without thoracic epidurals would not change.\n\nFor this pilot study, no power calculation was made to determine the number of subjects to be included. This number of subjects per group was arbitrarily set at 10.\n\nThe Statistical analysis was performed using IBM SPSS Statistics for Mac version 24.0 (IBM Corp., USA).\n\n\nResults\n\nAfter consent was obtained, 20 subjects were included into this pilot study. Due to a technical fault with the study computer and broken cable, only the data of 16 patients could be analyzed; 8 with and 8 without epidural catheters. Demographic data are presented in Table 1.\n\nBMI – Body mass index, ASA – American Society of Anesthesiologists physical status score.\n\nDuring the study periods (see methods), no vasoactive drugs such as phenylephrine were administered. After intubation, the NOL and HR increased significantly in all patients by 11.3 points (CI: 2.7–19.9; p=0.013) and 9.4 bpm (CI: 3.3–15.6; p=0.0105), respectively in all patients compared to the baseline value (average of values over 60 seconds before event). NOL and delta NOL significantly increased after intubation for 90 seconds (Figure 1 A, B). After intubation, NOL values increased 90% compared to baseline values. HR and delta HR also significantly increased after intubation for 130 seconds (Figure 1 C, D). HR increased only by 12% after intubation when compared to baseline values.\n\nAfter skin incision mean NOL values in the no-epidural group increased by 13.9 points (CI: 7.4–20.3; p=0.001) compared to 5.4 points (CI: -6.3–17.1; p=0.29) in the epidural group. The mean difference between no-epidural and epidural groups was 8.4 points (CI: -3.7–20.6; p=0.15). After the incision, NOL and delta NOL values significantly increased until 190 seconds in the no-epidural group, in the epidural group no significant change was observed (Figure 2A, B).\n\nThe skin incision stimulus did not increase HR and delta HR significantly in both groups and the mean difference in HR increase between groups was only 0.8 (CI: -7.6–9.2; p=0.84) (Figure 2C, D).\n\nThe areas under the curve, calculated for delta NOL and delta HR after the incision, showed a significant lower delta NOL AUC in the epidural group than the no-epidural group (Table 2). AUC calculated for delta HR after incision did not show any significant difference between the groups.\n\n* represents p≤0.05, Student t-Test.\n\n\nDiscussion\n\nThis is the first study assessing the feasibility of NOL to evaluate the quality of epidural analgesia during a nociceptive stimulation under general anesthesia.\n\nIn this pilot study, the level of nociception (NOL) assessed with the PMD100™ pain monitor as well as HR, significantly increased in all our patients after intubation (Figure 1) as previously reported by other authors2,3. This significant increase in NOL, Delta NOL, HR and Delta HR lasted for 1.5 to 2 minutes after the stimulus, despite a moderate amount of i.v. fentanyl dose (1mcg/kg) was administered 5 minutes before intubation. This demonstrates that intubation remains a strong, detectable nociceptive stimulus under general anesthesia with appropriate fentanyl analgesia.\n\nIn our study, NOL index increased by 90% after intubation whereas HR increased by only 12% for the same stimulus. This finding aligns with previous reports that the NOL index might have a better sensitivity in detecting noxious stimulus such as an intubation2,3.\n\nSkin incision followed by the first trocar insertion and endoscope placement for the VATS procedure caused a significant NOL index increase in patients without epidurals, despite the standardized fentanyl administration prior to incision. In patients with a prior to incision activated epidural catheter in place, NOL values did not significantly increase after incision (Figure 2A, B), while the HR did not significantly vary in both groups after incision (Figure 2C, D) emphasizing the NOL’s higher sensitivity and specificity to detect nociceptive stimuli2,3. The observed smaller variations of the NOL index or the delta NOL after incision in the epidural group is likely caused by effective epidural analgesia and successful attenuation of the nociceptive autonomous response caused by the skin incision5, hence they might be a good quantitative parameter to assess the quality of analgesia provided by an epidural analgesia.\n\nMean NOL values before skin incision were different between groups. A previous study found the threshold for nociception to be around a NOL index of 12 while values around 20 were associated with mild pain. The authors suggested a NOL value of 16 to be the threshold for pain detection under general anesthesia2. In our study, patients were subjected to many types of stimuli after the induction of general anesthesia: manual ventilation, intubation, lateral positioning which all together can induce a small amount of pain or discomfort and might explain why the basal threshold of NOL was higher (20.3 +/-18 in the epidural group) compared to values reported in previous studies2,3 when patients were left at rest under general anesthesia.\n\nThis is particularly true in the epidural group as they did not receive supplementary doses of fentanyl as opposed to the no-epidural group (see methods). Because the epidural catheter was placed 2 hours before surgery and only a small (test) dose was injected, no effect of the epidural could be expected before intubation.\n\nWe found no difference in NOL values between the epidural and no-epidural group before intubation (Dataset 16). The lower NOL values of the no-epidural group before skin incision are likely caused by an analgesic effect of the intravenous fentanyl bolus (1 mcg/kg), given 5 minutes prior, questioning the predictive value of the monitor. Fentanyl yields a rapid onset of systemic analgesia, whereas epidural analgesia is dermatomal, only.\n\nMonitors of nociception assess single or multiple changes of the autonomous nervous system, including HR and its variability, skin vasomotor reflex and conductance, and photoplethysmogram7–10. Some literature suggests that multi parametric indices are more sensitive compared to single parameter devices to detect mild and moderate noxious stimulation1,11,12. There is no literature to date evaluating the effect of regional anesthesia on the variation of indexes after painful stimulus offered by these monitors.\n\nIn an obstetric population, lumbar epidurals were found to increase HR variability due to an increase in parasympathetic activity after epidural analgesia13. Another study found that a neuraxial blockade reduced low-frequency power and high-frequency power of HR variability, suggesting a total decrease in autonomic activity14.\n\nEpidural autonomous nervous system blockade effects include dermatomal sympathectomy, hypotension, changes in skin temperature regulation15,16, decrease of cardiac accelerator fibers tone, and a slight reduction in heart rate17,18. Theoretically, all these changes could affect nociception measurements, such as the NOL.\n\nMid-thoracic epidural analgesia (T5-T11) inhibits efferent sympathetic preganglionic outflow, causing vasodilatation of the highly compliant splanchnic bed in a dose dependent manner that leads to a decrease of systemic arterial pressure because of venous pooling of blood in this region13,19. Additionally, the relative hypovolemia, secondary to the epidural sympathectomy-mediated vasodilatation might cause a physiological tachycardic response, potentially increasing NOL values in the epidural group; we did not observe this in our study.\n\nWe recognize several limitations in our study: a double blind randomized study design might have been a better choice, but this would have added complexity. This pilot study aimed at whether epidural analgesia can be detected by NOL in patients under general anesthesia in order to design stronger studies. Further, all data were electronically recorded, hence NOL values could not be influenced by the research, anesthesia or surgical teams. We are presenting the results of small pilot and feasibility study, future larger studies are warranted to evaluate if the NOL is useful for titration of epidural local anesthetics during combined general-epidural anesthesia.\n\nIn summary, this is the first study looking at the feasibility of assessing the NOL index in patients under general anesthesia and thoracic epidural analgesia and its ability to assess intra-operative epidural analgesia.\n\n\nEthical standards\n\nAll procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the local ethical committee on the 24th of June 2014 (Human Subjects Division, University of Washington, Seattle, WA USA. Chairperson: Jane Hitti, MD. Approval # 44750)\n\n\nData availability\n\nDataset 1: Nociception level index (NOL) Dataset 10.5256/f1000research.15279.d2071646",
"appendix": "Competing interests\n\n\n\nPhilippe Richebé, MD, PhD, is part of the scientific advisory board of the company Medasense LTD that makes the PMD100TM which was used in this study to provide the NOL index. As such he received honorarium as a consultant for this company.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe PMD100™ device and the finger electrodes supplies have been offered by the company Medasense Biometrics Ltd for this study.\n\n\nFunding disclosures\n\nThe PMD100™ device and the finger electrodes supplies have been offered by the company Medasense Biometrics Ltd., Ramat Yishai, Israel, for the purpose of this study.\n\n\nReferences\n\nBen-Israel N, Kliger M, Zuckerman G, et al.: Monitoring the nociception level: a multi-parameter approach. J Clin Monit Comput. 2013; 27(6): 659–68. PubMed Abstract | Publisher Full Text\n\nMartini CH, Boon M, Broens SJ, et al.: Ability of the nociception level, a multiparameter composite of autonomic signals, to detect noxious stimuli during propofol-remifentanil anesthesia. Anesthesiology. 2015; 123(3): 524–34. PubMed Abstract | Publisher Full Text\n\nEdry R, Recea V, Dikust Y, et al.: Preliminary Intraoperative Validation of the Nociception Level Index: A Noninvasive Nociception Monitor. Anesthesiology. 2016; 125(1): 193–203. PubMed Abstract | Publisher Full Text\n\nLirk P, Hollmann MW: Outcome after regional anesthesia: weighing risks and benefits. Minerva Anestesiol. 2014; 80(5): 610–8. PubMed Abstract\n\nJänig W: Autonomic reactions in pain. Pain. 2012; 153(4): 733–5. PubMed Abstract | Publisher Full Text\n\nBollag L, Jelacic S, Delgado Upegui C, et al.: Dataset 1 in: The nociception level index (NOL) response to intubation and incision in patients undergoing video-assisted thoracoscopic surgery (VATS) with and without thoracic epidural analgesia. A pilot study. F1000Research. 2018. Data Source\n\nIkuta Y, Shimoda O, Ushijima K, et al.: Skin vasomotor reflex as an objective indicator to assess the level of regional anesthesia. Anesth Analg. 1998; 86(2): 336–40. PubMed Abstract | Publisher Full Text\n\nShimoda O, Ikuta Y, Nishi M, et al.: Magnitude of skin vasomotor reflex represents the intensity of nociception under general anesthesia. J Auton Nerv Syst. 1998; 71(2–3): 183–9. PubMed Abstract | Publisher Full Text\n\nStorm H: Changes in skin conductance as a tool to monitor nociceptive stimulation and pain. Curr Opin Anaesthesiol. 2008; 21(6): 796–804. PubMed Abstract | Publisher Full Text\n\nLoggia ML, Juneau M, Bushnell MC: Autonomic responses to heat pain: Heart rate, skin conductance, and their relation to verbal ratings and stimulus intensity. Pain. 2011; 152(3): 592–8. PubMed Abstract | Publisher Full Text\n\nTreister R, Kliger M, Zuckerman G, et al.: Differentiating between heat pain intensities: the combined effect of multiple autonomic parameters. Pain. 2012; 153(9): 1807–14. PubMed Abstract | Publisher Full Text\n\nRantanen M, Yli-Hankala A, van Gils M, et al.: Novel multiparameter approach for measurement of nociception at skin incision during general anaesthesia. Br J Anaesth. 2006; 96(3): 367–76. PubMed Abstract | Publisher Full Text\n\nDeschamps A, Kaufman I, Backman SB, et al.: Autonomic nervous system response to epidural analgesia in laboring patients by wavelet transform of heart rate and blood pressure variability. Anesthesiology. 2004; 101(1): 21–7. PubMed Abstract | Publisher Full Text\n\nLandry DP, Bennett FM, Oriol NE: Analysis of heart rate dynamics as a measure of autonomic tone in obstetrical patients undergoing epidural or spinal anesthesia. Reg Anesth. 1994; 19(3): 189–95. PubMed Abstract\n\nFreise H, Meissner A, Lauer S, et al.: Thoracic epidural analgesia with low concentration of bupivacaine induces thoracic and lumbar sympathetic block: a randomized, double-blind clinical trial. Anesthesiology. 2008; 109(6): 1107–12. PubMed Abstract | Publisher Full Text\n\nFreise H, Van Aken HK: Risks and benefits of thoracic epidural anaesthesia. Br J Anaesth. 2011; 107(6): 859–68. PubMed Abstract | Publisher Full Text\n\nGoertz AW, Seeling W, Heinrich H, et al.: Influence of high thoracic epidural anesthesia on left ventricular contractility assessed using the end-systolic pressure-length relationship. Acta Anaesthesiol Scand. 1993; 37(1): 38–44. PubMed Abstract | Publisher Full Text\n\nTanaka K, Harada T, Dan K: Low-dose thoracic epidural anesthesia induces discrete thoracic anesthesia without reduction in cardiac output. Reg Anesth. 1991; 16(6): 318–21. PubMed Abstract\n\nClemente A, Carli F: The physiological effects of thoracic epidural anesthesia and analgesia on the cardiovascular, respiratory and gastrointestinal systems. Minerva Anestesiol. 2008; 74(10): 549–63. PubMed Abstract"
}
|
[
{
"id": "35364",
"date": "27 Jun 2018",
"name": "Sarah Saxena",
"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\nSummary:\nThis pilot study evaluated the efficacy of an intraoperative thoracic epidural analgesia by using a multi-parameter device, the nociception level index (NOL).\n\nThe authors offer an interesting perspective on the use of a multi-parameter intra-operative nociception device in patients undergoing VATS.\n\nComments:\nAbstract: please mention in the conclusion that more studies are needed.\nIntroduction: please mention that NOL is the only nociception monitor currently available combining several parameters (in contrast to for example ANI; SPI)\nDiscussion: In the limitations, you state that randomization would have added complexity. Can you please explain this?\nPlease mention in the conclusion that future large studies are warranted.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "35368",
"date": "02 Jul 2018",
"name": "Jean Pierre Estebe",
"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 provide pilot study which try to evaluate the interest to monitor the anesthesia and surgical stresses by the NOL index (multiple physiological parameters). It was a prospective observational study with patients scheduled for video-assisted thoracoscopic surgery with or without thoracic epidural analgesia (function of patient preference). The comparison was made between the NOL and the usual cardiovascular response (HR, and MBP). During the first step the authors compare the variations during and just after the tracheal intubation under regular anesthesia protocol (lidocaine, fentanyl, propofol then sevoflurane, and rocuronium) under BIS monitoring. During the second step; similar evaluation was performed during the surgical incision (epidural 5mL lidocaine 2% Vs. bolus 1mcg/mL fentanyl). The authors reported a more sensitive and reliable NOL variations than HR variations (with or without sympathetic protection; i.e. intubation Vs. thoracic surgical incision).\n\nMy major remark is just based on semantic reflection. Instead of a discussion of intra-operative nociception it could be better to discuss the cardiovascular response to the stress, or in a more general way to the sympathetic / para sympathetic balance assessment to the anesthetic stress (intubation) and surgical stress (skin incision in this study).\nMinor Points:\nThe authors must clarify the procedure used to control the epidural catheter tip location (ephedrine is just for the accidental IV administration) and his true efficacy (postoperative pain evaluation; not only by the lack of supplementary intraoperative doses of fentanyl). Table 1 results must be expressed with variations (i.e. +/- SD). It is a pity that the postoperative pain evaluation was not reported, and correlate or not, with NOL variations (or with the AUC of NOL).\n\nDespite several limitations in the study reported by the authors (not double-blind randomization); the concept proof of reliable monitoring by the NOL index is confirmed in this pilot study. Unfortunately, the authors did not try to correlate these NOL variations with the level of postoperative pain (or analgesic consumption).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "35367",
"date": "09 Jul 2018",
"name": "Emmanuel Boselli",
"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 have performed a study aiming at assessing NOL response to intubation and incision in patients undergoing video-assisted thoracoscopic surgery with or without epidural analgesia. The study followed a quasiexperimental non-randomized design. In total, 20 patients were included, and 16 were analyzed (8 in each group). They observed that NOL, ∆NOL, HR and ∆HR increased significantly in all patients after intubation. After skin incision, NOL and ∆NOL increased significantly in the epidural group, whereas no difference was observed for HR and ∆HR. The AUCs for ∆NOL and ∆HR after stimulus during 180 s were calculated. A significant difference was observed for ∆NOL AUC in patients with or without epidural analgesia. This was not observed for ∆HR. The authors conclude that NOL may be useful to evaluate the efficacy of an intraoperative thoracic epidural analgesia.\nAlthough of interest, there are some concerns regarding this study.\nMajor Concerns:\nThe authors should emphasize the fact that this study was a non-randomized quasiexperimental study performed on a small amount of patients (8 per group). Although some variations were observed, the results should be considered with caution and confirmed by larger trials.\n\nCould the authors explain why ANOVA for repeated measures (and corresponding post hoc tests) were not performed for all time by group interactions ? It is unclear to me why no intergroup analysis was assessed (epidural vs no epidural) during time for intubation.\n\nThe AUC is somewhat confusing and should be better described (and referenced). What is exactly the information provided by this parameter? Moreover, this study doesn't provide any threshold for NOL or ∆NOL variation to assess any outcome. Could the author better explain how these variations may be useful to modify anything in the anesthesia protocol?\nMinor Concerns:\nMicrogram should be labeled µg , not mcg, in the entire manuscript. Please correct.\n\nSeconds should be labeled s, not sec, in the entire manuscript and figures. Please correct.\n\nTable 2: please provide estimates of dispersion for AUC (95% CI, SD,...) and present data with groups in columns and values (∆NOL and ∆HR in rows).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "35366",
"date": "09 Jul 2018",
"name": "Georges Daccache",
"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 observational pilot study was designed to evaluate the feasibility of monitoring the autonomic nervous system response to intraoperative stress (intubation and skin incision), by a novel and unique multi-parameter device, the nociception level index (NOL) compared to the heart rate response (HR) under sevoflurane general anesthesia for VATS.\n\nFor the skin incision, the authors separated the patients in 2 groups of 8 each: 100 mg lidocaine bolus through an epidural catheter versus 1 mcg/kg intravenous fentanyl. The authors showed that NOL variations were more sensitive than HR variations in detecting the skin incision stress.\nThis pilot study confirmed the reliability of monitoring by the NOL the autonomic response after the two most common intraoperative stresses (intubation and skin incision).\n\nDespite several limitations that have been well-discussed by the authors (not randomized, high basal NOL values in the epidural group), these encouraging results have to be confirmed by future larger studies and compared to other available nociception monitoring devices (SPI, Analgesia Nociception Index) in various clinical settings.\n\nFuture studies should also focus on the clinical impact of monitoring nociception by the NOL in terms of postoperative pain and analgesic consumption.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-875
|
https://f1000research.com/articles/7-430/v1
|
06 Apr 18
|
{
"type": "Research Article",
"title": "Spectrum of Sellar and Parasellar Region Lesions: A retrospective study from Basrah, Iraq",
"authors": [
"Abbas Ali Mansour",
"Ali Hussain Ali Alhamza",
"Ammar Mohammed Saeed Abdullah Almomin",
"Ibrahim Abbood Zaboon",
"Nassar Taha Yaseen Alibrahim",
"Rudha Naser Hussein",
"Muayad Baheer Kadhim",
"Haider Ayad Yassin Alidrisi",
"Hussein Ali Nwayyir",
"Adel Gassab Mohammed",
"Dheyaa Kadhim Al-Waeli",
"Ibrahim Hani Hussein",
"Ali Hussain Ali Alhamza",
"Ammar Mohammed Saeed Abdullah Almomin",
"Ibrahim Abbood Zaboon",
"Nassar Taha Yaseen Alibrahim",
"Rudha Naser Hussein",
"Muayad Baheer Kadhim",
"Haider Ayad Yassin Alidrisi",
"Hussein Ali Nwayyir",
"Adel Gassab Mohammed",
"Dheyaa Kadhim Al-Waeli",
"Ibrahim Hani Hussein"
],
"abstract": "Background: Sellar and parasellar region lesions spectrum includes a wide variety of diseases. This study aimed at providing a comprehensive overview of such lesions in patients from Faiha Specialized Diabetes, Endocrine and Metabolism Center (FDEMC) in Basrah (Southern Iraq). Methods: Retrospective data analysis from FDEMC for the period January 2012 through June 2017. We included all patients with sellar and parasellar region lesions who received a MRI scan on their pituitary region Results: The total enrolled patients were 232 (84 men and 148 women),with age range 15-75 years.Pituitary disease and adenoma were more common among women. Those with macroadenoma were older than those with microadenoma, with nearly equal gender prevalence of macroadenoma. Pituitary adenoma constituted the bulk of pituitary disease in this setting (67.2%). Growth hormone secreting adenoma were the most common adenoma seen in 41.0%, followed by clinically non-functioning pituitary adenoma (NFPA) in 31.4% and prolactinoma in 26.9%. About 64.8% of pituitary adenoma was macroadenoma. Macroadenoma was seen in 73.4% of growth hormone secreting adenoma (acromegaly), 61.2% in NFPA and 62.0% of prolactinom a(of them six were giant prolactinoma). Conclusion: Pituitary adenoma constituted the bulk of sellar and parasellar region lesions, growth hormone secreting adenoma is the the most common adenoma followed by NFPA and prolactinoma due to referral bias. A change in practice of adenoma treatment is needed.",
"keywords": [
"Sellar and parasellar region lesions",
"pituitary disease",
"pituitary adenoma",
"classification",
"epidemiology."
],
"content": "Introduction\n\nSellar and parasellar region lesions spectrum includes a wide variety of conditions ranging from adenoma to empty sella syndrome, apoplexy, congenital or acquired condition1–4. Other than adenoma, genetic causes of pituitary disease are increasingly recognized3.\n\nPituitary adenomas are not rare and account for 20% all intracranial tumors5,6. Half of these secrete hormones, and half are microadenoma2. Clinically non-functioning adenomas (NFPA) constitute 15–54% of all adenomas. Prolactinomas accounts for 32–66%, growth hormone secreting adenoma (acromegaly) account for 8–16%, adrenocorticotropic hormone (ACTH)-secreting adenoma (Cushing's disease) forms 2–6%, and TSHoma accounts for less than 1%2,7. These pituitary adenomas behave as typical or have a more aggressive to malignant behavior6,8. They can cause mass effect, in addition to hypersecretion or hypopituitarism7,9.\n\nAdvances in neuroradiology have opened the door for earlier and easier diagnosis of pituitary disease and other sellar and suprasellar lesions10.\n\nThe Faiha Specialized Diabetes, Endocrine, and Metabolism Center (FDEMC) in Basrah is a tertiary referral center receiving patients with pituitary diseases from most of Southern Iraq. The FDEMC is trying to adapt the three mission criteria of the pituitary center of excellence, which includes care and support for patients, fellowship training and contribution to pituitary disease research11. To our knowledge, there are no studies on sellar and parasellar region lesions in Iraq.\n\nThis study aimed at providing a comprehensive overview of sellar and parasellar region lesions for patients from FDEMC in Basrah (Southern Iraq).\n\n\nMethods\n\nRetrospective data analysis of the FDEMC database for the period January 2012 through June 2017.\n\nInclusion criteria: We included all patients,age range 15–75 years with sellar and parasellar region lesions who have received a MRI scan on their pituitary region.\n\nExclusion criteria: patients with sellar and parasellar region lesions who did not receive a MRI scan.\n\n\nDefinition of variables\n\nSequences of pituitary MRI imaging were classified according to the international standard12. Adenomas were classified as macroadenoma if these were 10 mm or more in size, while microadenoma if less than 10 mm and giant prolactinoma if these were 4 cm and above2.\n\nPituitary adenoma, NFPA, prolactinoma, growth hormone secreting adenoma (acromegaly), and ACTH-secreting adenoma were defined according to the usual criteria2,8,12.\n\nHypopituitarism, whether postoperative or in those with or without adenoma, was considered according to the hormonal assessment with basal and dynamic hormonal tests13.\n\nEmpty sella syndrome, whether primary or secondary to surgery or apoplexy, were considered based on MRI findings14.\n\nCraniopharyngioma diagnosis was based on clinical behavior with MRI and pathological diagnosis.\n\n\nData analysis\n\nAnalysis was done in July 2017. All patients with labeling diagnosis of pituitary disease were included. Data were included on an Excel spreadsheet and transferred to SPSS for Windows, Version 23.0 (SPSS Inc., Chicago, USA).\n\nContinuous variables were summeried as number and percentage and dichotomous varibales as mean ±SD.\n\nThe ethics committee of the Medical College in Basrah University approved the study design and the Center authorities agreed to review the patients data. At the time of registeration in the Center, all patients included in this study approved the use of their clinical information for research purposes.\n\n\nResults\n\nA total of 232 patients were included in this study. Pituitary disease and adenoma were more common among women (Table 1). Those with macroadenoma were older than those with microadenoma with nearly equal gender prevalence of macroadenoma. Four patients died; two with growth hormone secreting adenoma (acromegaly) and advanced cardiovascular disease, and two with prolactinoma that caused hypopituitarism and adrenal failure.\n\n*Of 156 pituitary adenoma\n\n**Of 101 macroadenoma\n\nTable 2 shows that pituitary adenoma constituted the bulk of pituitary disease in this registry (67.2%). Growth hormone secreting adenoma (acromegaly) were the commonest adenoma seen in 41.0% followed by NFPA in 31.4% than prolactinoma in 26.9%. Hypopituitarism due to various causes was observed in 24.5% in this series. Empty sella syndrome, whether primary or secondary, were seen in 9.4%. Craniopharyngioma and Sheehan syndrome were seen in 3.9% each. Meningioma based on MRI finding was been observed in 4 patients (1.7%).\n\n*GH-secreting adenoma,2 of them stain on biopsy for lactotroph cell\n\n**Acromegaly in 4\n\nIn this study, 64.8% of pituitary adenoma were macroadenomas (Table 3). Macroadenoma was seen in 73.4% of acromegaly, 61.2% in NFPA and 62.0% of prolactinoma (of them six were giant prolactinoma).\n\n*Of them six giant Prolactinoma\n\nIn Table 4 we see hypophysectomy whether transsphenoidal or transcranial or both was performed in 45 patients with pituitary adenoma (28.8%). Stereotactic radiosurgery is done in 5 patients (3.2%) with pituitary adenoma. Growth hormone secreting adenoma (acromegaly) and prolactinomas were treated primarily with medical therapy (71.4% and 76.1% respectively).\n\n*Of patients with acromegaly\n\n**Of patients with prolactinoma\n\n\nDiscussion\n\nAll pituitary disease and adenoma were more common among women in this study. The gender predominance among patients with pituitary adenoma is variable in the literature depending on hormone secretion and age of the patients, the size of the tumor and female dominance is not clear15,16. However, female dominance has been seen in Saudi Arabia17 and one series from Argentina18. Those with macroadenoma tend to be older in age with no difference in the prevalence between men or women.\n\nSeen in about two-thirds of patients, pituitary adenoma constituted the main bulk of pituitary disease in this study, which is compatible with reports in the literature16.\n\nThe commonest pituitary adenoma was growth hormone secreting adenoma (acromegaly), followed by NFPA and prolactinoma. This is entirely different from the literature on the prevalence of pituitary adenoma2,16,17,19. This could be attributed to selection bias because only growth hormone secreting adenoma (acromegaly) patients are being referred, while NFPA and prolactinoma were treated by different specialties, such as neurosurgeons or gynecologists, without referral to a specialized Center like FDEMC. In Basrah, most cases of hyperprolactinemia were seen by a gynecologist because of amenorrhea and infertility, and the neurosurgeon follows patients with NFPA without referring them.\n\nHypopituitarism is prevalent in a quarter of this pituitary centre, from different causes, ranging from macroadenoma to hypophysectomy. Evaluation for hypopituitarism remains an integral part of the workup for any pituitary lesions because missing such diagnosis could be catastrophic9,13. This figure is far higher than that of Saudi Arabia, which was 1.2%17.\n\nEmpty sella syndrome was seen in 9.4% of patients in this study, which can be primary or secondary to surgery or apoplexy. Empty sella syndrome needs an extensive workup to assess pituitary function20.\n\nCraniopharyngioma and Sheehan syndrome are two diseases with a different spectrum of age distribution, but they were seen at the same frequency in this cohort. Craniopharyngioma is a disease of childhood and adolescence21. Sheehan syndrome is supposed to be rare in developed countries, but is still seen in developing countries22.\n\nLess than two thirds of adenoma in this study were macroadenomas. While in most series macroadenomas constitute 50% of the pituitary adenomas2; however in Canada, a similar finding has been seen compared with this study23. Again this could be explained by referral bias in this study. In Saudi Arabia, microadenomas were more prevelant17.\n\nFor growth hormone secreting adenoma (acromegaly), more than two thirds were macroadenomas, which is established fact for all acromegaly2,24,25.\n\nNFPA was a macroadenoma and seen at around 60% in this study. A similar finding was seen in a previous series2.\n\nProlactinomas were macroadenoma in around 60% of cases in this study. This differs from the literature, where more than 90% of prolactinomas were microadenomas2,18.\n\nHypophysectomy-transsphenoidal as surgical treatment was done in one third of pituitary adenomas, while transcranial approach or stereotactic radiosurgery was contemplated in the minority. This is a typical approach for most of the pituitary adenomas2,26. For growth hormone secreting adenoma (acromegaly), the primary treatment in this study was medical treatment in about two thirds of individuals. This is contrary to literature where surgery is the main mode of therapy26. The explanation is that we are just building a new neurosurgery unit for pituitary glands over the last few years, and in the future, surgery of pituitary is supposed to improve, and early referral will be the best.\n\nFor prolactinoma, primary medical treatment was done in two thirds of patients, while it should be the main treatment of choice in more than 90%, as seen in previous literature26.\n\nMalignant disease metastasizing to the pituitary is not observed in this study because they are not referred from Oncology Center in Basrah.\n\nThis study supposes to involve most of the pituitary disease patients in Basrah because the Center is a tertiary referral center. However, due to referral bias among some neurosurgeons and gynecologologists, we cannot guarantee that the data includes all patients with this condition in Basrah.\n\n\nConclusion\n\nPituitary adenomas constituted the bulk of pituitary disease in patients treated at the FDEMC, Basrah. Growth hormone secreting adenoma (acromegaly) is the most frequent adenoma followed by NFPA and prolactinoma due to referral bias. A change in the practice of adenoma treatment is needed.\n\n\nData availability\n\nDataset 1: Description of patients included in the study 10.5256/f1000research.13632.d19743927",
"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\nAcknowledgments\n\nAll authors are thankful to the medical staff of FDEMC for their contribution and support.\n\n\nReferences\n\nGlezer A, Bronstein MD: Pituitary apoplexy: pathophysiology, diagnosis and management. Arch Endocrinol Metab. 2015; 59(3): 259–64. PubMed Abstract | Publisher Full Text\n\nMolitch ME: Diagnosis and Treatment of Pituitary Adenomas: A Review. Jama. 2017; 317(5): 516–24. PubMed Abstract | Publisher Full Text\n\nLarson A, Nokoff NJ, Meeks NJ, et al.: Genetic causes of pituitary hormone deficiencies. Discov Med. 2015; 19(104): 175–83. PubMed Abstract\n\nMesserer M, Dubourg J, Raverot G, et al.: Non-functioning pituitary macro-incidentalomas benefit from early surgery before becoming symptomatic. Clin Neurol Neurosurg. 2013; 115(12): 2514–20. PubMed Abstract | Publisher Full Text\n\nMehta GU, Lonser RR: Management of hormone-secreting pituitary adenomas. Neuro Oncol. 2017; 19(6): 762–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatzellis E, Alexandraki KI, Androulakis II, et al.: Aggressive pituitary tumors. Neuroendocrinology. 2015; 101(2): 87–104. PubMed Abstract | Publisher Full Text\n\nNarendra BS, Dharmalingam M, Kalra P: Acromegaloidism Associated with Pituitary Incidentaloma. J Assoc Physicians India. 2015; 63(6): 79–82. PubMed Abstract\n\nDai C, Feng M, Liu X, et al.: Refractory pituitary adenoma: a novel classification for pituitary tumors. Oncotarget. 2016; 7(50): 83657–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHigham CE, Johannsson G, Shalet SM: Hypopituitarism. Lancet. 2016; 388(10058): 2403–15. PubMed Abstract | Publisher Full Text\n\nFabian UA, Charles-Davies MA, Fasanmade AA, et al.: Male Sexual Dysfunction, Leptin, Pituitary and Gonadal Hormones in Nigerian Males with Metabolic Syndrome and Type 2 Diabetes Mellitus. J Reprod Infertil. 2016; 17(1): 17–25. PubMed Abstract | Free Full Text\n\nMcLaughlin N, Laws ER, Oyesiku NM, et al.: Pituitary centers of excellence. Neurosurgery. 2012; 71(5): 916–24; discussion 24–6. PubMed Abstract | Publisher Full Text\n\nChanson P, Raverot G, Castinetti F, et al.: Management of clinically non-functioning pituitary adenoma. Ann Endocrinol (Paris). 2015; 76(3): 239–47. PubMed Abstract | Publisher Full Text\n\nKim SY: Diagnosis and Treatment of Hypopituitarism. Endocrinol Metab (Seoul). 2015; 30(4): 443–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpaziante R, Zona G, Testa V: Primary empty sella syndrome. Surg Neurol. 2003; 60(2): 177–8; author reply 178. PubMed Abstract | Publisher Full Text\n\nMcDowell BD, Wallace RB, Carnahan RM, et al.: Demographic differences in incidence for pituitary adenoma. Pituitary. 2011; 14(1): 23–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDrange MR, Fram NR, Herman-Bonert V, et al.: Pituitary tumor registry: a novel clinical resource. J Clin Endocrinol Metab. 2000; 85(1): 168–74. PubMed Abstract | Publisher Full Text\n\nAljabri KS, Bokhari SA, Assiri FY, et al.: The epidemiology of pituitary adenomas in a community-based hospital: a retrospective single center study in Saudi Arabia. Ann Saudi Med. 2016; 36(5): 341–5. PubMed Abstract | Publisher Full Text\n\nDay PF, Loto MG, Glerean M, et al.: Incidence and prevalence of clinically relevant pituitary adenomas: retrospective cohort study in a Health Management Organization in Buenos Aires, Argentina. Arch Endocrinol Metab. 2016; 60(6): 554–61. PubMed Abstract | Publisher Full Text\n\nLake MG, Krook LS, Cruz SV: Pituitary adenomas: an overview. Am Fam Physician. 2013; 88(5): 319–27. PubMed Abstract\n\nGhatnatti V, Sarma D, Saikia U: Empty sella syndrome - beyond being an incidental finding. Indian J Endocrinol Metab. 2012; 16(Suppl 2): S321–3. PubMed Abstract | Free Full Text\n\nMüller HL: Craniopharyngioma. Handb Clin Neurol. 2014; 124: 235–53. PubMed Abstract | Publisher Full Text\n\nKaraca Z, Laway BA, Dokmetas HS, et al.: Sheehan syndrome. Nat Rev Dis Primers. 2016; 2: 16092. PubMed Abstract | Publisher Full Text\n\nImran SA, Yip CE, Papneja N, et al.: Analysis and natural history of pituitary incidentalomas. Eur J Endocrinol. 2016; 175(1): 1–9. PubMed Abstract | Publisher Full Text\n\nLavrentaki A, Paluzzi A, Wass JA, et al.: Epidemiology of acromegaly: review of population studies. Pituitary. 2017; 20(1): 4–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPortocarrero-Ortiz LA, Vergara-Lopez A, Vidrio-Velazquez M, et al.: The Mexican Acromegaly Registry: Clinical and Biochemical Characteristics at Diagnosis and Therapeutic Outcomes. J Clin Endocrinol Metab. 2016; 101(11): 3997–4004. PubMed Abstract | Publisher Full Text\n\nAl-Dahmani K, Mohammad S, Imran F, et al.: Sellar Masses: An Epidemiological Study. Can J Neurol Sci. 2016; 43(2): 291–7. PubMed Abstract | Publisher Full Text\n\nMansour AA, Alhamza AH, Almomin AM, et al.: Dataset 1 in: Spectrum of Sellar and Parasellar Region Lesions: A retrospective study from Basrah, Iraq. F1000Research. 2018. Data Source"
}
|
[
{
"id": "32887",
"date": "30 Apr 2018",
"name": "Abdul Al-Toma",
"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. Mansour and co-authors investigated the Spectrum of pituitary disease (Sellar and Parasellar Region Lesions) in a large cohort of patients from a tertiary referral center in the southern region of Iraq. This center receives patients with pituitary diseases from most of Southern Iraq, a population of 6-8 million. The authors provided a well written retrospective analysis of the study population. The manuscript is focused on providing an attractive epidemiological description of the studied patients.\n\nAbstract section: concise text and states clearly the objective of the report. In their conclusion they stated that a change in practice of adenoma treatment is needed. It would be more informative if they provide some data on how they reached to this conclusion. A minor correction: the following 2 sentences need to be separated by full stop mark instead of comma (Pituitary adenoma constituted the bulk of sellar and parasellar region lesions, growth hormone secreting adenoma is the most common adenoma followed by NFPA and prolactinoma due to referral bias.)\nThe following sections are well written and data were clearly presented:\nMethods and results. In addition they included an appropriate Statistical analysis. However, minor English language edition is needed.\nThe discussion section described the study results in comparison with recent literature. Recent literature has been adequately addressed. The authors stated the limitation f their manuscript clearly. It would be interesting to provide some information on the prognosis of their study patients, both for surgically or medically treated patients.\nThe references are clearly presented and conform the current standards.\nThe tables are clearly written.\nMinor comments: although the manuscript is well written, however some revision of the English text needs to done.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "34531",
"date": "14 Jun 2018",
"name": "Khaled Mohammed Al-Dahmani",
"expertise": [
"Reviewer Expertise Pituitary",
"thyroid"
],
"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 Mansour AA et al. is an important contribution to the limited pituitary research in the middle East and North Africa (MENA) region.\nThe following points need further modifications/clarifications;\nSellar and parasellar region lesions usually refers to structural abnormalities and do not necessarily include other pituitary pathologies like hyperprolactinemia and hypopituitarism. Therefore, I suggest using the term ''pituitary disorders'' instead; more comprehensive.\n\nThe inclusion of patients needs further clarifications in the ''Method'' section. what diagnoses where looked for?\n\nPituitary adenoma includes all functioning and non-functioning pituitary tumors. In some part of the text, it was used separately from its subtype (see paragraph; Definition of variables). Maybe just typo.\n\nTable (2) needs to be simplified. The total number of patients should equal clearly equal 232. I suggest using 3-4 categories only; Sellar masses/abnormalities, Hyperprolactinemia, hypopituitarism and others.\n\nThe \"miscellaneous\" cases need to be mentioned; what diagnoses are here?\n\nTable 3 the sum of patient with pituitary adenoma subtypes are not equal to the total number of cases.\n\nIn the conclusion, ''A change in the practice of adenoma treatment is needed'' is very general. Based on the presented data, change in the management of GH secreting adenomas is more specific and relevant.\n\nAdditional language editing will further enhance the manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3742",
"date": "22 Jun 2018",
"name": "Abbas Mansour",
"role": "Author Response",
"response": "The study by Mansour AA et al. is an important contribution to the limited pituitary research in the middle East and North Africa (MENA) region.The following points need further modifications/clarifications; Sellar and parasellar region lesions usually refers to structural abnormalities and do not necessarily include other pituitary pathologies like hyperprolactinemia and hypopituitarism. Therefore, I suggest using the term ''pituitary disorders'' instead; more comprehensive. Done The inclusion of patients needs further clarifications in the ''Method'' section. what diagnoses where looked for? Retrospective data analysis of FDEMC database for the period from January 2012 through June 2017.Inclusion criteria:We included all patients with pituitary disorders who have MRI pitutiary regardles the age . Pituitary adenoma includes all functioning and non-functioning pituitary tumors. In some part of the text, it was used separately from its subtype (see paragraph; Definition of variables). Maybe just typo. Done Table (2) needs to be simplified. The total number of patients should equal clearly equal 232. I suggest using 3-4 categories only; Sellar masses/abnormalities, Hyperprolactinemia, hypopituitarism and others. Its cant be 232 because some may have more than one diagnosis The \"miscellaneous\" cases need to be mentioned; what diagnoses are here? Miscellaneous includes galactorrhea, hypogonadotropic hypogonadism, and acromegaloidism Table 3 the sum of patient with pituitary adenoma subtypes are not equal to the total number of cases. GH-secreting adenoma,2 of them stain on biopsy for lactotroph cellGH-secreting adenoma with hyperprolactinemia in 5 patients In the conclusion, ''A change in the practice of adenoma treatment is needed'' is very general. Based on the presented data, change in the management of GH secreting adenomas is more specific and relevant. A change in the practice of pituitary adenoma treatment is needed.All adenomas treatment needed reevaluation . Additional language editing will further enhance the manuscript. Done"
}
]
}
] | 1
|
https://f1000research.com/articles/7-430
|
https://f1000research.com/articles/7-824/v1
|
21 Jun 18
|
{
"type": "Research Article",
"title": "The Texas Conservation Plan has not slowed oil and gas well development in dunes sagebrush lizard habitat",
"authors": [
"Jacob Malcom",
"Matthew Moskwik",
"Matthew Moskwik"
],
"abstract": "Background: The dunes sagebrush lizard (Sceloporus arenicolus) is an imperiled species that is restricted to shinnery oak (Quercus havardii) sand dune habitats in southeastern New Mexico and West Texas, USA. This region is also a hotspot of oil and gas development that is a major threat to the species. Methods: Here we use well data and a natural experiment to test the effectiveness of voluntary conservation agreements for slowing or stopping oil and gas well approval in the lizard’s habitat in New Mexico and Texas. Results: We show that the Candidate Conservation Agreement (CCA) and CCA with Assurances in New Mexico, both of which contain strong avoidance mechanisms, are associated with a steep decline in oil and gas well approval in the New Mexico portion of the lizard’s range, but not outside the lizard’s range. By contrast, the Texas Conservation Plan (TCP), which does not include mandatory avoidance, is not associated with any decline of oil and gas well approval in the lizard's Texas range relative to the broader landscape. Conclusions: These results indicate that the TCP is insufficient to conserve the lizard in Texas, thereby jeopardizing genetic and geographic representation across the range of the species.",
"keywords": [
"Endangered Species Act",
"dunes sagebrush lizard",
"Sceloporus arenicolus",
"oil and gas development",
"voluntary conservation",
"Candidate Conservation Agreement"
],
"content": "Introduction\n\nThe dunes sagebrush lizard (DSL; Sceloporus arenicolus) is an imperiled species whose distribution is restricted to shinnery oak (Quercus havardii) sand dunes in the Mescalero Sandhills of eastern New Mexico and the Monahans Sandhills of West Texas, USA (Degenhardt et al., 1996; Fitzgerald & Painter, 2009). This area is within the Permian Basin, which is the focus of extensive and intensive oil, gas, and infrastructure development that degrades or destroys the species’ habitat (Sias & Snell, 1998; U.S. Fish and Wildlife Service, 2010). The species’ perilous conservation status has been recognized since at least 1982, when the U.S. Fish and Wildlife Service (FWS) first established S. arenicolus as a candidate for listing under the U.S. Endangered Species Act (ESA) (U.S. Fish and Wildlife Service, 2010). After episodes in and out of candidate status, the FWS proposed to list the species as endangered in 2010 (U.S. Fish and Wildlife Service, 2010). That proposal was withdrawn in 2012, in large part because of voluntary conservation agreements in New Mexico and Texas that the FWS believed offered adequate protections for the species (U.S. Fish and Wildlife Service, 2012).\n\nThe agreements for New Mexico and Texas are structured very differently, but both are based on the voluntary conservation component of section 10(a)(1)(A) of the ESA. For non-listed species, these include Candidate Conservation Agreements (CCAs) and CCAs with Assurances (CCAAs; for brevity, we refer to both agreements as the “CCA/As”). CCAs are agreements between FWS and one or more public or private parties that stipulate the actions enrollees will take or will avoid to conserve a species, which may preclude the need for listing under the ESA. CCAAs are similar to CCAs, but apply only to non-federal parties and include assurances that enrollees will not face ESA restrictions beyond those described in the CCAA if the covered species is listed in the future. (Federal agencies are not eligible for CCAAs because they cannot be exempted from the duty to avoid jeopardizing listed species under section 7(a)(2).) Parties in New Mexico drafted a CCA (for federal entities) and complementary CCAA (for non-federal entities) to protect lesser prairie-chicken (Tympanuchus pallidicinctus) and S. arenicolus habitat in December, 2008. These CCA/As include strong requirements to avoid the lizard’s shinnery oak sand dune habitats (U.S. Fish and Wildlife Service et al., 2008; U.S. Fish and Wildlife Service and the Center for Excellence in Hazardous Materials Management, 2008), which reflect the practices of the Bureau of Land Management in their Range Management Plan for the species (BLM, 2008). However, the CCA/As also direct oil and gas wells into interstitial habitats (between the large sand dune blowouts that the species uses) that provide connectivity among core dunes habitats.\n\nIn contrast to the New Mexico CCA/As, the Texas Conservation Plan (TCP) for the DSL—which is a CCAA with a tailored name—does not include avoidance requirements (Texas Comptroller of Public Accounts, 2011). Instead, the TCP offers only guidance to attempt to avoid habitat; there is no requirement for enrollees to avoid developing oil and gas wells in lizard habitat. Even though the same legal instrument underlies the agreement of each states—section 10(a)(1)(B) of the ESA—the differences in the details means we expect different conservation outcomes for S. arenicolus.\n\nThe objective of this study was to test whether voluntary conservation agreements for the DSL in New Mexico and Texas may have been effective at reducing oil and gas development in the species’ habitat. We hypothesized that the New Mexico CCA/As have produced a noticeable reduction of new oil and gas wells approval in DSL habitat, but that the TCP did not produce such a reduction. Our predictions were:\n\n1. The rate of new well approval through time is approximately the same inside and outside of DSL habitat before the CCA/As (2009) (in New Mexico) and before the TCP (2012) (in Texas);\n\n2. The rate of new well approval in New Mexico is lower inside of DSL habitat than outside of DSL habitat after the CCA/As were adopted; and\n\n3. The rate of new well approval in Texas is not different inside and outside of DSL habitat after the TCP was adopted.\n\n\nMethods\n\nWe downloaded all oil and gas well data for New Mexico from the state’s Oil Conservation Division site on 03 April 2018. The Texas Railroad Commission makes its oil and gas well data available through a separate provider, http://www.texas-drilling.com/, from which we downloaded the data on 03 April 2018. We filtered out well approvals that were marked as canceled in the datasets from both states. We defined the range of S. arenicolus in New Mexico as the boundaries recognized in 2008, at the time the CCA/As were developed and adopted. We defined the species’ range in Texas as the boundaries of the “Hibbitts Map” of suitable habitat (from low to very high quality; Fitzgerald et al., 2011). In New Mexico, the area outside of the species range included oil and gas well data from the Permian basin excluding the 2008 range boundaries. In Texas, the area outside included the five counties (i.e., Andrews, Crane, Ector, Ward, and Winkler) that encompass the species’ range, excluding the “Hibbitts Map.”\n\nTo test our hypotheses and determine the rate of oil and gas well expansion, we counted the number of wells approved each year since 1990 inside and outside of S. arenicolus habitat. Because this scenario is an intervention experiment with a before-after-control-impact design, we fit log-link Poisson generalized linear models (McCullagh & Nelder, 1999), with terms for time period and in/out of habitat, for statistical inference. We fit separate linear models, of the form number_wells ~ year + in_CCAA_area, for New Mexico and Texas data to plot the trends in/out of habitat and before/after the agreements were approved. All code and the data needed to replicate our results is available in the Open Science Foundation repository at https://doi.org/10.17605/OSF.IO/HKVSU (Malcom, 2018).\n\n\nResults\n\nThe well data supported our predictions. We observed that the rate of new well approval was much lower within the DSL habitat than outside of the DSL habitat after the adoption of the CCA/As in New Mexico (Figure 1); however, the rate of new well approvals was no different inside versus outside of DSL habitat after the adoption of the TCP in Texas (Figure 2). The trends visible in the plots are supported by the generalized linear model statistics (Table 1).\n\nThe plot shows the number of wells approved by the State of New Mexico per year (dots), inside and outside of the lizard’s habitat (yellow and purple, respectively). Fitted lines are from a simple least-squares model of the form number_wells ~ year + in_CCAA_area, split by pre- and post-CCA/A. Figure CC-BY Defenders of Wildlife 2018, available at https://doi.org/10.6084/m9.figshare.6226721 (Malcom & Moskwik, 2018).\n\nThe plot shows the number of wells approved by the Texas Railroad Commission per year (dots), inside and outside of the lizard’s habitat (yellow and purple, respectively). Fitted lines are from a simple least-squares model of the form number_wells ~ year + in_TCP_area, split by pre- and post-TCP. CC-BY Defenders of Wildlife 2018, available at https://doi.org/10.6084/m9.figshare.6226964.v2 (Malcom & Moskwik, 2018).\n\nse, standard error.\n\n\nDiscussion\n\nConserving the DSL requires protecting its remaining habitat in both New Mexico and Texas: if the species is lost from either state then representation (Shaffer & Stein, 2000)—both in terms of unique genetic contributions (Chan et al., 2009; Chan et al., 2013) and geographical distribution—will be lost. Our analyses indicate that the CCA/As in New Mexico have significantly reduced oil and gas development, one of the most notable direct threats to the DSL and its habitat. In contrast to New Mexico, the data show that the TCP has had no effect on the rate of new well approval inside DSL habitat in Texas. This research highlights how the details of voluntary conservation agreements, even those authorized under the exact same provision of law, can lead to markedly different outcomes.\n\nThe decline in the number of new oil and gas wells approved each year in New Mexico after the CCA/As were adopted reflect the avoidance requirements in those agreements. The number of new wells approved in the DSL’s range in New Mexico is not zero since the CCA/As were enacted because the agreements allow well siting in interstitial habitat. While this reduces the direct effects of development, it likely harms connectivity (U.S. Fish and Wildlife Service, 2008) and may have secondary effects on landscape characteristics that influence DSL life history (Ryberg et al., 2015).\n\nBecause the TCP does not require avoidance of DSL habitat loss, i.e., there is no mechanism protecting the habitat, we expected to not see any effect of the TCP. The data supported our prediction, and even hint that the rate outside DSL habitat decreased faster than inside. Had the State of Texas incorporated strong avoidance requirements of the New Mexico CCA/As into the TCP, then our analysis may have shown that voluntary conservation efforts were sufficient to protect the DSL.\n\n\nData availability\n\nIn addition to obtaining the data as described in the manuscript, the raw oil and gas well data associated with this article can also be found on OSF: https://osf.io/hkvsu/ (Malcom, 2018).\n\n\nSoftware availability\n\nSoftware available from: https://github.com/jacob-ogre/DSL_well_approvals.\n\nArchived software at time of publication: https://osf.io/hkvsu/ (Malcom, 2018).\n\nLicense: BSD 2-Clause \"Simplified\" License.",
"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\nWe thank Y.-W. Li and R. Dreher for their helpful comments on the manuscript.\n\n\nReferences\n\nBureau of Land Management: Special status species Record of Decision and approved resource management plan amendment. Pecos District Office, Bureau of Land Management, Roswell, New Mexico. 2008; 110. Reference Source\n\nChan LM, Archie JW, Yoder AD, et al.: Review of the systematic status of Sceloporus arenicolus Degenhardt and Jones, 1972 with an estimate of divergence time. Zootaxa. 2013; 3664(3): 312–320. PubMed Abstract | Publisher Full Text\n\nChan LM, Fitzgerald LA, Zamudio KR: The scale of genetic differentiation in the dunes sagebrush lizard (Sceloporus arenicolus), an endemic habitat specialist. Conserv Genet. 2009; 10(1): 131–142. Publisher Full Text\n\nDegenhardt WG, Painter CW, Price AH: Amphibians and Reptiles of New Mexico. University of New Mexico Press; 1996. Reference Source\n\nFitzgerald LA, Painter CW: Dunes sagebrush lizard (Sceloporus arenicolus). In L.L.C. Jones and R.E. Lovich (editors). Lizards of the American Southwest: a photographic field guide. Rio Nuevo Publishers, Tucson, Arizona. 2009; 198–220.\n\nFitzgerald LA, Painter CW, Hibbitts TJ, et al.: Final report on the range and distribution of Sceloporus arenicolus in Texas: Results of surveys conducted 8-15 June 2011. Unpublished report. 2011. Publisher Full Text\n\nMalcom JW: Dunes Sagebrush Lizard: Oil and Gas Well Approvals and Voluntary Conservation Agreements. Open Science Framework. 2018. Data Source\n\nMalcom JW, Moskwik M: Effect of New Mexico conservation agreements for the dunes sagebrush lizard on oil and gas well approval rates. figshare. Figure. 2018. Publisher Full Text\n\nMalcom JW, Moskwik M: No effect of the Texas Conservation Plan on oil and gas well approval in dunes sagebrush lizard habitat in Texas. figshare. Figure. 2018. Publisher Full Text\n\nMcCullagh P, Nelder JA: Generalized linear models. Chapman & Hall, CRC. 1999. Reference Source\n\nRyberg WA, Hill MT, Painter CW, et al.: Linking irreplaceable landforms in a self-organizing landscape to sensitivity of population vital rates for an ecological specialist. Conserv Biol. 2015; 29(3): 888–898. PubMed Abstract | Publisher Full Text\n\nShaffer M, Stein B: Safeguarding our Precious Heritage. In Stein, B.A., L.S. Kutner, and J.S. Adams, editors. Precious Heritage: The Status of Biodiversity in the United States. Oxford University Press. New York. 2000; 301–322. Reference Source\n\nSias DS, Snell HL: The sand dune lizard Sceloporus arenicolus and oil and gas in southwestern New Mexico. Final report of field studies 1995-1997. Submitted to the New Mexico Department of Game and Fish. Department of Biology, University of New Mexico, Albuquerque, New Mexico. 1998. Reference Source\n\nTexas Comptroller of Public Accounts: Texas Plan for the dunes sagebrush lizard (Sceloporus arenicolus). Developed in consultation with the U.S. Fish and Wildlife Service, U.S. Department of Agriculture – Natural Resources Conservation Service, Texas A&M University, Texas Endangered Species Task Force, Texas Department of Agriculture, Texas Parks and Wildlife, Texas Railroad Commission, University of Texas Systems University lands, Texas Farm Bureau, Texas Oil and gas Association, Texas Royalty Council, Texas and Southwestern Cattle Raisers Association, Texas Wildlife Association, and Texas Association of Business. Texas Comptroller of Public Accounts, Austin, Texas. 2011; 150. Reference Source\n\nU.S. Fish and Wildlife Service: Intra-Service Section 7 Conference Opinion on the proposed Issuance of a Section 10(a)(1)(A) Enhancement of Survival Permit for Lesser Prairie-Chicken (LPC) and Sand Dune Lizard (SDL) to The Center of Excellence for Hazardous Materials Management (CEHMM) and proposed Implementation of the BLM Candidate Conservation Agreement. New Mexico Ecological Services Office, Albuquerque, New Mexico. 2008; 25. Reference Source\n\nU.S. Fish and Wildlife Service: Endangered and threatened wildlife and plants; endangered status for dunes sagebrush lizard. Fed Regist. 2010; 75: 77801–77817. Reference Source\n\nU.S. Fish and Wildlife Service: Endangered and threatened wildlife and plants; withdrawal of the proposed rule to list dunes sagebrush lizard. Fed Regist. 2012; 77: 36872–36899. Reference Source\n\nU.S. Fish and Wildlife Service, U.S. Bureau of Land Management, and Center for Excellence for Hazardous Materials Management: Candidate conservation agreement for the lesser prairie-chicken (Tympanuchus pallidicinctus) and sand dune lizard (Sceloporus arenicolus) in New Mexico. 2008. Reference Source\n\nU.S. Fish and Wildlife Service and Center for Excellence for Hazardous Materials Management: Candidate conservation agreement with assurances for the lesser prairie-chicken (Tympanuchus pallidicinctus) and sand dune lizard (Sceloporus arenicolus) in New Mexico. 2008. Reference Source"
}
|
[
{
"id": "35353",
"date": "02 Jul 2018",
"name": "Howard L. Snell",
"expertise": [
"Reviewer Expertise Conservation biology",
"herpetology",
"functional & evolutionary ecology."
],
"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\nArticle provides a potential analysis of the efficacy of various types of agreements among parties to protect species that could have been listed as endangered by the US Fish & Wildlife Service. A comparison between the pattern of permits issued by New Mexico or Texas within and outside habitat of the Dunes Sagebrush Lizard before and after the agreements were finalized appears to suggest that New Mexico's agreements have a greater effect that those of Texas. That conclusion may be supported, but additional analyses could clarify the actual situation.\nBasically, previous research has shown that the effects of oil and gas development on populations of the Dunes Sagebrush Lizard (DSL) vary with the density of wells and their associated infrastructure. However, the current analyses compare total numbers permits for wells and not the densities that would result from those wells if constructed. This may lead to some confusion. Both states permit considerably fewer wells within DSL habitat than outside - both before and after the agreements were finalized. However, without knowing the relative amounts of habitat it is hard to interpret the pattern. Texas appears to permit fewer wells after the agreements - both within and outside DSL habitat, while New Mexico appears to increase permitting outside of DSL habitat after the permits and reduce permitting within. Adjusting the permitting data for area could clarify if differences between the states' patterns exist and, perhaps more importantly, are the apparent differences likely to impact populations of the lizard.\nThe use of explicit hypotheses and predictions is clear. However, it might appear directed to demonstrate less efficacy in Texas than in New Mexico due to the \"one-tailed\" nature of the directional predictions. Perhaps broadening the hypotheses and predictions into a \"two-tailed\" format could increase objectivity. Basically \"The agreements of the two states differ. We tested whether or not those differences translate into lower or higher densities of wells in the future\".\nFinally, if the permit data were adjusted to densities, it should be densities of existing wells and future permitted wells. That could then be compared to the threshold densities proposed by Sias & Snell 19981.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "40292",
"date": "21 Jan 2019",
"name": "Philip N. Smith",
"expertise": [
"Reviewer Expertise Ecological risk assessment",
"toxicology",
"wildlife"
],
"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\nIntroduction The sentence beginning “After episodes in and out of candidate status,…”is nebulous and uninformative. What are the episodes that are alluded to? In order to provide necessary context, it would appear that those episodes should be described to some extent.\n\nPertaining to the section describing the Texas CCA/As: a search of the referenced TCP document reveals no specific wording as it is characterized here. No specific language related to “attempt to avoid habitat” is included in the referenced document. Why is the word “attempt” emphasized in italics? It is not verbatim from the reference. Clearly, there are differences in the NM and TX plans, but this passage characterizes the Texas document as intentionally or tacitly promoting disruptive oil and gas development. Another interpretation of the TCP would be that it allowed greater flexibility to enrollees as to conservation strategies. This section should be revised to more objectively and thoroughly present the differences between the TCP and NM CCA.\n\nWhy isn’t the objective of the paper to evaluate actual protection of the DSL? That would seem to be a much more direct and appropriate test of the process and allow for comparison of each state’s effectiveness in carrying out the process.\n\nAs written, outcomes of the predictions (results) appear to be solely affected by the CCA/As; as if no other factors drive oil and gas development approvals. Myriad other factors influence these outcomes, but none are listed here (and should be). This paper lacks acknowledgement and review of mitigating/confounding factors.\n\nMethods Data Identification Selection of DSL range and thus lands to be included in these analyses is critical. Justification for the selection procedure, or land masses included in analyses are not well documented in this manuscript. As for the Texas DSL range, the Hibbitts map is somewhat controversial and perhaps not well suited (in its entirety) for these analyses. Habitat designated as “low quality” likely does not and may have never supported viable populations of DSLs. Inclusion of those areas in these analyses is questionable and potentially confounding. What is the reference for the NM DSL range? It is difficult to determine if the authors are comparing apples and apples, or apples and oranges. Detailed maps and much stronger justification for inclusion of land in these analyses would be most helpful.\n\nData Analysis This seems like an overly complicated way to evaluate some very simple questions. Why not simply use Chi-square analysis to determine application approval rate differences inside and out, before and after? For that matter, is the focus on approval rates or simply the numbers of new wells? The introduction speaks of approval rates, not numbers of new wells. Data presented only depict new wells, not applications that were approved or rejected.\n\nResults The authors switch back and forth between new well approvals and numbers of new wells. They are not the same. What about wells that were not approved? The figures indicate a reduction in number of wells in NM after implementation of the CCA, and no decrease in areas deemed “outside DSL range”. Clearly the delineation of inside versus outside seriously influences these models. The figure depicting new wells in Texas DLS habitat also show reductions in well development both inside and outside DSL ranges. Why is that? The slopes of these lines are not presented for evaluation. Also, the table of statistics is vague and confusing. What is the term “Year:in-habitat”? Is that an interaction term? Unclear as presented.\n\nWith regard to the figure legends: there is insufficient documentation of the “strong avoidance requirements” in the NM plan as well as no substantiation of the “lacks strong avoidance requirements” claims about the TCP. As such, this manuscript appears, to some degree, to lack objectivity.\n\nDiscussion These results indicate that (at best) oil and gas development has slowed in some areas in both states since about 2008. There is no definitive proof that the CCA/As were the cause. This is perhaps simply due to coincidence. These results could also demonstrate how comparing apples to oranges can lead one to whatever conclusion an author wishes to demonstrate. The objectivity and methods of this research are questionable.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-824
|
https://f1000research.com/articles/7-462/v1
|
13 Apr 18
|
{
"type": "Software Tool Article",
"title": "Clinotator: analyzing ClinVar variation reports to prioritize reclassification efforts",
"authors": [
"Robert R. Butler III",
"Pablo V. Gejman",
"Pablo V. Gejman"
],
"abstract": "While ClinVar has become an indispensable resource for clinical variant interpretation, its sophisticated structure provides it with a daunting learning curve. Often the sheer depth of types of information provided can make it difficult to analyze variant information with high throughput. Clinotator is a fast and lightweight tool to extract important aspects of criteria-based clinical assertions; it uses that information to generate several metrics to assess the strength and consistency of the evidence supporting the variant clinical significance. Clinical assertions are weighted by significance type, age of submission and submitter expertise category to filter outdated or incomplete assertions that otherwise confound interpretation. This can be accomplished in batches: either lists of Variation IDs or dbSNP rsIDs, or with vcf files that are additionally annotated. Using sample sets ranging from 15,000–50,000 variants, we slice out problem variants in minutes without extensive computational effort (using only a personal computer) and corroborate recently reported trends of discordance hiding amongst the curated masses. With the rapidly growing body of variant evidence, most submitters and researchers have limited resources to devote to variant curation. Clinotator provides efficient, systematic prioritization of discordant variants in need of reclassification. The hope is that this tool can inform ClinVar curation and encourage submitters to keep their clinical assertions current by focusing their efforts. Additionally, researchers can utilize new metrics to analyze variants of interest in pursuit of new insights into pathogenicity.",
"keywords": [
"ClinVar",
"variation",
"clinical variant",
"pathogenic",
"benign",
"variant interpretation",
"variant reclassification",
"pathogenicity"
],
"content": "Introduction\n\nThe dbSNP database1 currently contains over 300 million RefSNPs, and dbVar2 adds over 5 million variant regions to the documented plasticity of the human genome. ClinVar3 is small by comparison, documenting the clinical impact of 300,000 variants. This may seem like a far simpler task; however, the substantial impact of these clinical variants on the lives of patients places a heavier burden on the level of evidence gathering required. Add to this the fragmented nature of the evidence—spread out across publications, databases, predictive software analysis and in individual health records—meaning each of these ClinVar records becomes its own meta-data analysis4. ClinVar, ClinGen5, and the American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology (AMP)4 have done an excellent job formulating assertion criteria that allows for a comprehensive analysis of all available data, collating them into a standardized classification. While this became the minimum standard upon its inception, there is still a backlog of older assertions with ill-defined criteria or those missing a specification altogether. Many of these would benefit from submitter reclassification based on the more recent standards.\n\nGiven the inconsistent amounts of variant data across the genome and the rapid generation of new studies, the significance of variants also changes at an accelerated pace6. Put in statistical terms, the ClinVar clinical significance represents an estimate of the true population significance, and current estimates are based on limited, often private datasets. In many of the instances of discrepancies in assertion, consensus has been achieved simply by sharing evidence previously unavailable to one party7–9. Harrison et al. found that 87.2% of discordant variants were resolved by reassessment and data sharing8. Several initiatives5,10–14 have had success in encouraging the public sharing of datasets and new studies, but for the foreseeable future, private data will continue to be a challenge to achieving consensus. Clinical assertions based on insufficient evidence can persist in public databases and consequently seed misinformation into future interpretations9,15. Recently, in the field of cardiovascular disease, there have been several high-profile instances of cardiovascular variants deemed to be highly pathogenic, yet not segregating with disease9,16,17. This unfortunate outcome is inevitable owing to the aforementioned reasons and illustrates a key issue: the continual need to share and reconcile new information with old data and reclassify clinical assertions as appropriate on a regular basis6,8.\n\nAs a clinician or researcher looking to utilize this information, the depth and sophistication of ClinVar presents a daunting learning curve. This is necessary, as ClinVar houses not only assertions, but evidence, literature, and an impressive amount of cross-reference material3. As Yang et al.18 have suggested, this makes the process of evidence interpretation challenging on an individual variant level and the batch processing of variants even more so. ClinVar itself has provided a utile web interface and simplified data structures for programmatic use19. To the same end, other tools have been developed to address both aims: to easily browse variations and compare curations20,21, or import and manipulate flattened ClinVar data for variant analysis22,23. While the browsing tools allow for user-friendly and web-hosted comparison, they do not provide the throughput to analyze large datasets. Conversely, the local database tools allow for deep analysis on large variant sets, but require a significant amount of programming experience and local computational resources to access and operate.\n\nClinotator is unique in that it provides largescale batch analysis without necessitating a large local computational resource or deep programming knowledge. It can quickly generate simple annotation tables, annotate vcf files or be integrated into annotation pipelines with little overhead. The goals were two-fold: (i) deliver filtered ClinVar information for each variant, focusing on clinical assertions being made about the variant; and (ii) generate several metrics by which the robustness and consistency of the evidence can be gauged for the overall clinical assertion. Clinotator’s quantification of assertion evidence takes into account significance type, submission age and submitter expertise category for a standardized scoring of clinical impact based on the five ACMG/AMP descriptors of Mendelian disorders: Benign (B), Likely Benign (LB), Uncertain Significance (US), Likely Pathogenic (LP) and Pathogenic (P)4.\n\nOur aim is for Clinotator to be useful in a number of capacities, including prioritizing variants that need reclassification, guiding submitter reconciliation efforts or simply identifying discordant variants for future research targets. To demonstrate its utility, we examined test sets of three-star variants (per ClinVar’s review status star ratings) and variants in ClinVar with “Conflicting Interpretations of pathogenicity” (CI). Clinotator was able to confirm recently published concordance trends6,8,18, and identify several groups of discordant variants for further investigation. It accomplished this efficiently, using a large-scale systematic approach with a minimal computational effort.\n\n\nMethods\n\nMetric calculation. Clinotator collects a variety of characteristics from ClinVar and generates four additional metrics (Table 1). The core component of these metrics is the Clinotator raw score (CTRS), generated as the sum of a variant’s weighted individual clinical assertions (i):\n\n\n\n\n\nThe assertion weight factor (xi) was chosen such that a certain multiple of the next lowest priority significance would be less than or equal to the value of the current significance. The distance between US and LB was then a multiple of 10, and the value of B was a multiple of two higher than LB. No multiple of US could attain LP, which was set as the equivalent positive value to LB; and P was a multiple of two higher than LP. The age of the assertion factor (di) reduces the assertion weight over time after a buffer. For the first 2 years, there is no penalty, then there is a 10% reduction gradation in weight per year through 6 years, at which point the penalty stays at a static 50% reduction thereafter. The submitter class factor (si) is weighted based on ClinVar submitter category as curated by ClinGen, with regular clinical assertions by genetic testing laboratories unweighted at 1.00, expert reviewers receiving a 1.10 and practice guidelines receiving a score of 1.25. Literature-only submissions are filtered out and submissions without assertion criteria or with incomplete data (for example no assertion date provided) are also omitted.\n\nThe Clinotator average assertion age (CTAA) is the mean age (in years) of valid clinical assertions. Variant submissions without an assertion age are omitted.\n\nThe Clinotator predicted significance (CTPS) is a predicted clinical significance based on the CTRS scores of variants in ClinVar with two or more valid clinical assertions. A dataset of all variants that score two stars in ClinVar and have a Mendelian significance was used as a calibration for the category ranges. For the purposes of this calibration, variants with a Pathogenic/Likely pathogenic (PLP) or Benign/Likely benign (BLB) overall significance were excluded as they could not definitively be placed in either category. Assertions with incomplete assertion data were also excluded. Finally, two-star variants with fewer than two clinical assertions with assertion criteria were excluded. Using this filtered calibration dataset, the bounded regions for each CTPS category were set based on a combination of ClinVar star criteria and non-parametric prediction intervals (PI). The lower bound of each range was set at 2[(assertion weight) * (0.7) * (1.0)]; namely, the minimum ClinVar qualification for two stars with both assertions being no more than 5 years old. The quantiles of each distribution as well as the PIs were examined for a range of confidences. The PI for each clinical significance was chosen as the highest possible confidence that aligned to the above established lower bound. Calculations were conducted in R24, and the non-parametric PIs were defined as the cth and rth values in each distribution, where25,26:\n\n\n\nThe Clinotator reclassification recommendation (CTRR) is a ranked reclassification priority based on the absolute difference between the ClinVar clinical significance (CVCS) and the CTPS. This field uses the seven values of clinical significance associated with Mendelian diseases (B, BLB, LB, US/CI, LP, PLP, P), valued one through seven. For the purposes of reclassification, CI is scored the same as US. Each shift along the scale increases the rank by one, and a transition between overall zones (all benign ⇔ US/CI ⇔ all pathogenic) adds an additional point.\n\n\n\nThe total number of points is capped at three. Rankings range from zero to three, in escalating degree of inconsistency.\n\nSoftware structure. The functional components of Clinotator are contained in four modules and a global variables file. The main program, clinotator.py, handles the I/O, errors and options for various file types.\n\nThe getncbi.py module handles querying of the E-utilities database servers27. It splits the input list into batches if necessary (default eLink batch size of 1000) and posts to the Entrez history server. It then fetches xml records in batches (default eFetch batch size of 4500). It handles some minor connection interruptions and gives three retries per batch before giving up. Returned batches are added to a list of xml objects to be handled by variation.py.\n\nThe variation.py module defines the VariationClass object, and its methods parse ClinVar xml records and calculate the scoring metrics, which are then stored as instance variables. clinotator.py then utilizes pandas to collect and organize tabled data for output. As the ClinVar xml format is highly sophisticated, it does not frequently lend itself to flattening without considerable database structure. The construction of variation.py will allow for future modification, and storage of additional ClinVar xml data as class attributes, allowing for significant backend manipulation with a minimal footprint on the local machine.\n\nThe vcf.py module is dedicated to the handling of vcf as an input type. It stores the header and adds the new INFO field definitions for the new annotation in the output file. The rsIDs in the ID column of the vcf are then sent through the rsID input method. After the annotation table has been created in clinotator.py, vcf.py matches annotations to vcf calls by rsID and alt allele combination. Alt alleles are handled as lists (and ClinVar haplotypes are handled as list instance objects), so multi-allelic loci are correctly labeled with their appropriate ClinVar report. Haplotypes are identified as such, but the ‘vcf_match’ field (Table 1) is omitted from the vcf annotation. The other 12 fields are added to the INFO field as outlined in the vcf version 4.3 standards28.\n\nThe global_vars.py file supplies a location for most static variables in the program, including several dictionaries of calibration values. Most of these values do not need any modification, but can be; for instance, download batch sizes from NCBI. If the default values result in frequent http errors, the batch size can be reduced. The maximum eLink batch size (for rsID and vcf types) is 1000 ids, while the maximum eFetch batch size is theoretically 10,000 ids. Both are set to lower levels to reduce the incidence of http errors and can be throttled based on available bandwidth.\n\nClinotator was designed in a Linux environment and implemented in Python (2.7 or ≥3.4)29, and can run in similar OSX and Windows Python environments. The required modules are pandas (0.20.0 minimum, 0.22.0 recommended)30 and biopython (1.70)31. It can be run on a personal computer with relatively modest system requirements; a minimum of 2 GB available RAM. The command line interface requires three pieces of information: (i) the type of input file, (ii) the file itself, and (iii) your email address. The input file can be one of three types: an rsID list using dbSNP identifiers; a Variation ID (VID) list using ClinVar identifiers; or a vcf file. In each case, multiple files can be included and will be processed in batches. If using a list type file, it should be a plain text file with a list of identifiers, one per line. The email is required by NCBI/biopython.\n\nAdditionally, the user can specify several options. A highly recommended log option (--log) generates a text file with the warnings from the run. A more extensive long log file (--long-log) can be specified for annotation details. Both log files override the terminal annotation warnings that occur when Clinotator finds missing xml data in ClinVar records in the default (no log) mode. The log files are written in append mode, so batch runs or multiple runs of Clinotator in the same folder can generate a significantly large log file. Users can also specify the output file prefix (the default is “clinotator”), which will label the output “tsv”, “anno.vcf”, and log files.\n\nIn all cases, a tab-delimited table file will be produced. The columns will be the VID, RSID and the fields in Table 1. If a vcf file is selected, Clinotator will generate an additional annotated output vcf file. Annotations are concatenated in the INFO field, including the VID and Table 1. Multi-allelic input variants will include comma-separated values for each minor allele. For further details about installation and usage, see the github repository for this project (Data and software availability section).\n\nAll VID lists used in analysis were generated at ClinVar, using the search filters and downloading a UI list in text format. The set of all variants with at least two stars was generated February 24th, 2018, and the set of all CI variants was generated on February 27th, 2018. Both sets of variants were analyzed with Clinotator, and split into two-star, three-star, four-star and CI sets. Additional computational analysis was done using dplyr, ggplot2, ggExtra, gridExtra, and RColorBrewer R packages32–36.\n\n\nResults\n\nA test set of 10,000 VIDs, was run on a system with a single core from an i7-4770 CPU with 16 GB of available memory. Clinotator averaged 1.79 min to complete, 87% of which comprised the NCBI query and download time. The greatest limitation to run time is the bandwidth of the connection to the NCBI databases. When running the list of all variants with at least two stars in ClinVar (>50,000), the run time never exceeded 15 min, with a post-download parsing time of around 90 s. As Clinotator keeps the NCBI xml results in memory, there can be a substantial memory usage. At the time of writing, the entire ClinVar xml set is approaching 6 GB. Loading the entire set into memory is doable with at least 8 GB of memory, though it is recommended that you batch your queries in this rare case. More typical usage for subsets of ClinVar or batch vcf annotations should not pose a memory issue.\n\nBatch annotation of vcf files is similarly efficient, working on single or multi-sample vcfs. Given the set of seven multi-sample, exonic vcf files available at the 1000 Genomes project, Clinotator was able to generate a variant table and annotate output vcf files for all seven files (15,171 total rsIDs) in an average of 3.94 min, 68% of which was NCBI query and download time. A potential speed limitation to vcf-based annotation is that NCBI is queried for each input vcf file, resulting in duplicate queries of common variants, but the tradeoff is not having to create a local query storage file that may potentially become very large if hundreds or thousands of vcf files are being analyzed in a pipeline. If higher throughput is required, it may be more efficient to consider a variant database structure which can return a non-redundant list of total database rsIDs and utilize the list rsID method to generate a reference table.\n\nA total of 48,483 variants were identified with two or more stars in ClinVar and at least two clinical assertions, with 23 four-star, 5,743 three-star and 42,717 two-star variants. Grouping by stars, there is a consistent average CTAA and CTNA across the various types, around three assertions per variant, and an average of about 1.5 years old. This points to a general continuity in ClinVar, encouraging for previous reports of concordance between different clinical labs and expert review panels18. The only exception is the outlier case of the four-star variant set, which averages about six assertions and an average assertion age of almost four. These variants are a particular group of well-documented CFTR variants, though the practice guideline assertion has not been reevaluated since 2004.\n\nThe two-star variants are graphed by CTRS in Figure 1A. The distributions of CTRS widely overlap and significantly skew towards overlarge outliers. The US group is the exception, with a leptokurtic distribution. Notably, despite the weighting of B and P assertion types by twice as much as their “likely” counterparts, the distributions of variants of each zone remain resolutely overlapped. The BLB distribution in particular seems both the largest and the most far ranging, extending beyond the B group. While the P group is slightly more distributed above its family members, the LP and PLP distributions, the PLP distribution still spreads over almost the entire positive side of the spectrum. As the PLP or BLB rating in ClinVar is based on a single piece of each type of evidence, there is not a quantification of how much P/B and how much LP/LB evidence is factored into each assessment.\n\n(A) All two-star variants, plotted according to CTRS, and colored based on the seven ClinVar clinical significance designations: Benign (B), Benign/Likely benign (BLB), Likely benign (LB), Uncertain Significance (US), Likely pathogenic (LP) Pathogenic/Likely pathogenic (PLP) and Pathogenic (P). (B) Prediction intervals for the five primary Mendelian clinical significances (B, LB, US, LP and P). Intervals plotted by CTRS value, using five different interval confidences (vertical axis). The optimal confidence interval for each clinical significance is marked with an asterisk. (C) All two-star variants plotted according to Clinotator Raw Score, and colored based on the seven Clinotator predicted significance ranges (B, BLB, LB, US, LP, PLP and P) after calibration with prediction intervals. Dashed lines denote prediction interval boundaries from (B).\n\nA total of 29,252 variants were two-star variants that qualified to be in the five control groups (Table 2). These variants were used to calculate the five prediction intervals depicted in Figure 1B. For each range, the quantiles and prediction intervals were chosen as described above. Given the fixed lower bounds defined by two-star status in ClinVar, the confidence of every prediction interval exceeds the similarly bounded median-centered quantile range, excepting the US category (the center 99.8% of the US distribution is larger than the 99.5% confidence prediction interval). As the US category has no lower bound, its upper bounds are defined by the lower bounds of LB and LP categories, which are outside the entire US control distribution, resulting in a prediction interval confidence greater than 99.99%, still not covering the full width between LB and LP. The likelihood of a US variant falling outside the chosen range is small.\n\nCTRR, Clinotator Reclassification Recommendation. B, Benign; LB, Likely Benign; US, Uncertain Significance; LP, Likely Pathogenic; P, Pathogenic; PI, prediction interval.\n\nThe resulting CTPS intervals are shown in Figure 1C, with the BLB and PLP intervals defined by the overlapping prediction intervals. It is worth noting that the overlap between B and LB is much wider than that between P and LP, which reflects the greater overlap of control B and LB distributions. This overlap disparity is observable for all fixed PI confidences individually and in the mixed-confidence PI model used in the final calibration (Figure 1B). Defining BLB and PLP groups with this approach has the advantage of classifying the BLB and PLP quantitatively in a range that cannot be called as either classification by the given confidence, with both classifications exceeding 95% confidence. For this purpose, the overlapping regions of PLP and BLB exist—not as yet another classification bin—as a measure of plasticity of borderline assertions.\n\nA potential concern for non-parametric prediction intervals is that they are inaccurate for values outside the control distribution. However, as the lower bounds are not defined by the prediction interval, only the upper bounds are vulnerable to extreme outliers. Regardless of how far outside the upper interval boundary a given variant may fall, the CTPS determination remains the same, limiting the potential for outliers to impact the reclassification score.\n\nReclassification recommendations for all of the two-star variants in Figure 2 largely confirm most variants shift by only a single position, if at all (see also Table 2). Only 16 variants are reclassified from an LB, LP, or PLP to a US classification. Most shifts occurred between the overlap categories and their immediate neighbors, likely resulting from the altered definition of the overlap category. Both of these results support previous research showing a fairly high general concordance in ClinVar18,37,38.\n\n(A) A schematic of the CTRR scoring workflow. ClinVar Clinical Significance (CVCS) is used as a starting point, and each rectangle passed to arrive at the Clinotator Predicted Significance (CTPS) counts as a point. Transitioning a significance family boundary adds an extra point (i.e., moving from Uncertain Significance (US) to Benign/Likely benign (BLB) scores a CTRR = 2 + 1 = 3). (B) A heat map of variant counts with each CVCS and CTPS combination. Darker squares correspond to higher numbers of variants. Blue represents Benign (B), BLB and Likely benign (LB); yellow represents US and red represents Likely pathogenic (LP), Pathogenic/Likely pathogenic (PLP) and Pathogenic (P).\n\nOf variants with at least two stars, all but one of the high priority for reclassification (CTRR = 3) were in the three-star group, and were further investigated. All 56 of these three-star, CTRR rank three variants were of the US classification. As Figure 3A suggests, these variants are primarily predicted to be in the benign family (41 BLB, 1 B). In total, 14 are predicted to be medically significant, belonging to the pathogenic family (9 PLP, 5 P). The submitters with assertion criteria for these 14 are examined in greater detail in Table 3. In 10/14 cases, the expert assertion is the oldest, eight of which are approaching 5 years of age. Additionally, there is a high level of consensus among the three most represented clinical laboratories, with at least two asserting a P or LP in 12/14. It is also notable that 13 of the variants are associated with cancer (Variation ID 42965 is associated with hypertrophic cardiomyopathy). Yang et al. previously found similar trends in clinical lab concordance, age-related discordance and highest concordance among hereditary cancer genes18. As the expert review overrides the other reviews, the tiered system likely disadvantages these variants, making them ideal candidates for reclassification. The full list of 57 variants with a CTRR score of three is available in Supplementary Table 1.\n\n(A) All three-star uncertain significance variants plotted on the left by CVNA and CTRS. On the right is a matched frequency distribution of CVNA. Values are colored according to their Clinotator reclassification recommendation (CTRR): blue represents a score of zero, yellow a score of two and red a score of three. (B) All conflicting interpretations of pathogenicity variants, plotted in the same manner and coloring scheme as (A).\n\nCTPS, Clinotator predicted significance; CTRS, Clinotator raw score; CVNA, ClinVar number of clinical associations; CTAA, average clinical assertion age; LB, Likely Benign; US, Uncertain Significance; LP, Likely Pathogenic; P, Pathogenic; PI, prediction interval.\n\nOne-star variants with CI status comprise a set of 13,880 variants. These variants are shown in Figure 3B, and show a similar trend to the Figure 3A distribution of three-star US variants: primarily benign, and increasing CTRR with a larger number of assertions. Looking at the distribution of CI variants with a CTRR of three (Figure 4), there are a number of potential reclassifications, which is unsurprising given their conflicted status. To sample what constitutes a minimum amount of evidence for a CTRR of three, the medically significant variants with only two criteria-based clinical assertions are provided in Table 4 (14 variants of PLP significance). Unlike the variants in Table 3, the majority of these variants are not associated with cancer. Instead, they are associated with cardiovascular diseases, metabolic diseases, and Rett syndrome. Despite not being cancer-focused, there is still a fair amount of concordance among clinical lab assertions. In most cases, the reason for conflict is a single significance provided without assertion criteria, substantially older than the two valid assertions. Given the ages of the conflicting assertions, and the lack of assertion criteria, inviting the submitters to re-evaluate their submissions would most likely reconcile the discrepancies. The full set of CI variants with a CTRR of three are available in Supplementary Table 2.\n\nCI variants with a Clinotator reclassification recommendation (CTRR) of three, counted by CVNA and colored by Clinotator predicted significance (CTPS). Blue represents Benign (B), yellow represents Benign/Likely benign (BLB), orange represents Pathogenic/Likely pathogenic (PLP) and red is Pathogenic (P). The asterisk denotes the column of 14 variants examined in Table 4.\n\nB, Benign; LB, Likely Benign; US, Uncertain Significance; LP, Likely Pathogenic; P, Pathogenic; PLP, Pathogenic/Likely pathogenic.\n\n* variant has somatic variant interpretations, 18 LP and 1 US\n\nNotably, one of the variants in this list, Variation ID 161516, had a CI significance based on one P, one LP, 18 LP (somatic) and one US (somatic) assertions. The literature has largely not addressed how to reconcile somatic and germline assertions, and the ACMG/AMP guidelines explicitly state they are “not intended for the interpretation of somatic variation”4.\n\n\nDiscussion\n\nAs shown above, Clinotator can be a useful tool for identifying discrepant records amongst a large and complex database. With limited resources, submitters and curators alike can utilize Clinotator metrics for prioritization of reclassifications and research. Additionally, Clinotator can be used to obtain ClinVar information in batch annotations, providing a convenient method to rapidly obtain some simple ClinVar metrics and Clinotator metrics with minimal computational effort. It can be readily integrated into existing pipelines or stand alone as a quick reference.\n\nClinotator’s ability to identify and filter missing data fields can also be leveraged to clean up older or incomplete submissions in ClinVar. For instance, the list of variations with at least two stars returned over 9000 assertions with a blank ‘Date Last Evaluated’ field, which has become a required field for current submissions. Submitters can check their own assertions to identify their submissions that lack an assertion date.\n\nIt should also be noted that Clinotator should not be used as a tool for directly determining clinical significance. Although Clinotator does develop a predicted significance, this is not through the use of primary evidence. The predictive range generated is for rating evidence strength and reclassification impact. Reclassification should always be done using the ACMG/AMP guidelines and assessing all primary evidence available to the researcher.\n\nTo compare/analyze variation report quality (a secondary analysis), Clinotator attempts to establish some common criteria. How to combine independent analyses is a particular problem, as these are not individual data points, but professional judgements using a coordinated guideline and overlapping evidence. It has been previously noted that there will always be some level of professional judgement that results in incongruous assertions8, but ultimately this needs to be reconciled to arrive at an overall interpretation by consensus. Mean or median assertion values will not account for the total body of assertions, falling prey to skew or omission, respectively. This is particularly so when there are multiple weighting factors modulating assertion values, thus an aggregated score can better express the total volume of assertions. Clinotator utilizes its raw score, which is an aggregate of these weighted clinical assertions.\n\nA potential issue that arises out of an aggregate model is that lower-level assertions made in a larger volume might artificially inflate the overall value of a variant. For instance, five LP assertions may give a variant the P status, despite no one submitter having enough evidence for the P category. However, while individual assertions utilize overlapping data, each one likely possesses additional private data as evidence. Thus each LP assertion does provide an additive value in terms of overall pathogenicity. We should therefore consider the five ‘Likely pathogenic’ assertions as more likely ‘Pathogenic’. Clinotator highlights this hypothetical five LP variant as a prime candidate for data sharing and reconciliation between the submitters to meet concordance. Clinotator is calibrated on the current, unambiguous two-star data in ClinVar and will be recalibrated on a regular basis to ensure that these boundaries: (i) change with richer information being submitted to ClinVar, and (ii) honor the intent of the ClinVar starring system when possible. In the ideal case, all of the submitters to ClinVar would have all of the data available and the resources to analyse all variants in ClinVar on a regular basis. In such an optimistic context, Clinotator would likely switch to a mean/median model.\n\nAssigning assertion weights to significance types is unfortunately a subjective process. There is not a universal, objective measure of quantity of pathogenicity available in ClinVar, or, arguably, in the literature. In lieu of a more objective metric, a range of assertion weights were tested and the control two-star distributions were examined, as was the set of all two-star variants (42,717 variants; Figure 1). This allowed for the analysis of variants with mixed assertion types, including analysis of CI variants and variants with mixed submitter expertise categories. For the assertion weights we tried, the relative shapes and overlaps of the five control distributions were largely consistent with the final values. Larger assertion weights primarily expanded the tail skew and overall CTRS values, while smaller weights lowered the distance between US and the other classes, shrinking all CTRS values. Expanding the distance between “Likely” and full class members similarly modified overall overlap CTRS values, but the relative overlap trend (that BLB carried a wider range than PLP) did not change. Mixed assertions can never be separated. Ultimately, Clinotator’s assertion weights are relative to class control distributions, so the current values were spread enough to observe comparative differences in overlap, while delineating pathogenicity families with a high degree of confidence. Future versions of Clinotator will need to be periodically recalibrated on current ClinVar distributions, and ultimately may weight assertion types differently if a more objective standard becomes available.\n\nThe PI ranges themselves are defined more objectively. As the control distributions are non-normal in several respects, a ranked non-parametric PI is most appropriate26, relying on the fairly large cross sections of the total variant sets in ClinVar (Table 2). Simply setting a static confidence level for the PI would be preferred, but as the lower bound is set by the ClinVar two-star criterion, scaling the whole range is far better than modifying it. As a result, there is a higher confidence in the predictions of some classes than others, but all are at least 95% confident (Figure 1B). As the goal of these prediction ranges is to assess evidential disparities and not to definitively classify variants, having conservatively wide ranges ensures a higher specificity for the CTRR statistic.\n\nThe age of the assertion matters. This has been previously identified as an issue6,8,18, but as both the test cases demonstrate, outdated assertions often fail to take into account new evidence and negatively impact classification. One of the key benefits of the current ACMG/AMP criteria is that any assertion must review all previous evidence and existing data available4. Thus while old data never loses its value, old assertions do; particularly if they were made prior to the establishment of the current standards. Reclassification on a regular interval should be a goal for submitters to ClinVar. Counter to the concept of clinical significance as a static value, it is intrinsically dynamic based on the limited availability of evidence. Thus Clinotator weights against the age of assertions after a grace period to ensure that current literature and data are more effectively integrated into the variation report. The maximum threshold of the age weight was set to 0.5 so that a P or B (±6) assertion at or beyond the limit is effectively downgraded to the associated “Likely” category (±3). LB or LP assertions cannot be downgraded to US significance, but are similarly halved in strength.\n\nFinally, the submitter expertise category is a continued confounder in ClinVar. It has become essential for experts in individual conditions to become involved in classification, as different conditions have nuanced profiles of pathogenicity8,9. However, as seen in our test cases, having the expert reviewers supersede all other clinical assertions results in a masking of assertion data. This complication is exacerbated by the age of assertion issue, but more frequent reclassification wouldn’t address the tiered nature of system. Clinotator’s solution is to weight by reviewer status, giving expert reviews a louder voice without drowning out the clinical significance conversation. The specific weights for expertise are subjective, owing to the absence of an objective submission quality metric on which to rank submitter expertise.\n\nNext steps for development involve a variety of fine tuning work. As its metrics are used for analysis, their effectiveness can be assessed and modified, particularly those with subjective elements. An ideal scenario for assertion type weighting would be for submitters to declare the evidence types they utilized, and whether that came from a private resource (i.e. PS4, private data; or PM2, ExAC data). This would allow for assertion type scoring based on an aggregate of evidence without overlap. Similarly, as variant annotations are tracked over time, the submitter expertise category can be calibrated to reflect the total body of experience that a submitter has, or the relative rates of reclassification in the different review status tiers.\n\nThe above examples only begin to describe Clinotator’s applications. Clinotator presents a framework for quantitatively assessing ClinVar evidence, and exploration of variants that have unusual Clinotator metrics. Clinotator can also incorporate new utilities to improve its data parsing sophistication, and additional metrics can be included, potentially incorporating new factors such as somatic mutation. Hopefully, it will become a useful tool for curation of ClinVar, and can be integrated with other tools, allowing for the improved classification of variants.\n\n\nData and software availability\n\nRRID: SCR_016054.\n\nClinotator source code available from: https://github.com/rbuteriii/clinotator.\n\nArchived source code at the time of publication: https://doi.org/10.5281/zenodo.121020439.\n\nSoftware license: GNU General Public License v3\n\nRaw data and analysis is available at: https://doi.org/10.5281/zenodo.121027340.",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nInstitutional funding from NorthShore University HealthSystem was used to carry out this research.\n\n\nAcknowledgements\n\nWe would like to thank Dr. Jubao Duan and Dr. Alan R. Sanders for research advice and editing assistance, and Mr. Sean McCarthy for statistical insights. We would also like to thank Mr. Alex Henrie, Ms. Sarah Hemphill, Dr. Karen Eilbeck and Dr. Heidi Rehm for advanced access to their upcoming publication on the ClinVar Miner software and web utility.\n\n\nSupplementary material\n\nSupplementary Table 1. ClinVar three-star variants of high priority for reclassification. A table of the output annotations for ClinVar three-star variants with a Clinotator reclassification recommendation score of 3.\n\nClick here to access the data.\n\nSupplementary Table 2. ClinVar conflicting interpretation variants of high priority for reclassification. A table of the output annotations for ClinVar ‘conflicting interpretation of pathogenicity’ variants with a Clinotator reclassification recommendation score of 3.\n\nClick here to access the data.\n\n\nReferences\n\nSherry ST, Ward MH, Kholodov M, et al.: dbSNP: the NCBI database of genetic variation. Nucleic Acids Res. 2001; 29(1): 308–311. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLappalainen I, Lopez J, Skipper L, et al.: DbVar and DGVa: public archives for genomic structural variation. Nucleic Acids Res. 2013; 41(Database issue): D936–941. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandrum MJ, Lee JM, Benson M, et al.: ClinVar: public archive of interpretations of clinically relevant variants. Nucleic Acids Res. 2016; 44(D1): D862–868. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichards S, Aziz N, Bale S, et al.: Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015; 17(5): 405–424. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRehm HL, Berg JS, Brooks LD, et al.: ClinGen--the Clinical Genome Resource. N Engl J Med. 2015; 372(23): 2235–2242. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith ED, Radtke K, Rossi M, et al.: Classification of Genes: Standardized Clinical Validity Assessment of Gene-Disease Associations Aids Diagnostic Exome Analysis and Reclassifications. Hum Mutat. 2017; 38(5): 600–608. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmendola LM, Jarvik GP, Leo MC, et al.: Performance of ACMG-AMP Variant-Interpretation Guidelines among Nine Laboratories in the Clinical Sequencing Exploratory Research Consortium. Am J Hum Genet. 2016; 98(6): 1067–1076. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarrison SM, Dolinsky JS, Knight Johnson AE, et al.: Clinical laboratories collaborate to resolve differences in variant interpretations submitted to ClinVar. Genet Med. 2017; 19(10): 1096–1104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhouse J, Skov MW, Bigseth RS, et al.: Distinguishing pathogenic mutations from background genetic noise in cardiology: The use of large genome databases for genetic interpretation. Clin Genet. 2018; 93(3): 459–466. PubMed Abstract | Publisher Full Text\n\nNussbaum RL: Sharing Clinical Reports Project. 2018. Reference Source\n\nStenson PD, Ball EV, Mort M, et al.: Human Gene Mutation Database (HGMD): 2003 update. Hum Mutat. 2003; 21(6): 577–581. PubMed Abstract | Publisher Full Text\n\nGreen RC, Goddard KA, Jarvik GP, et al.: Clinical Sequencing Exploratory Research Consortium: Accelerating Evidence-Based Practice of Genomic Medicine. Am J Hum Genet. 2016; 98(6): 1051–1066. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalmaña J, Digiovanni L, Gaddam P, et al.: Conflicting Interpretation of Genetic Variants and Cancer Risk by Commercial Laboratories as Assessed by the Prospective Registry of Multiplex Testing. J Clin Oncol. 2016; 34(34): 4071–4078. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLek M, Karczewski KJ, Minikel EV, et al.: Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016; 536(7616): 285–291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMacArthur DG, Manolio TA, Dimmock DP, et al.: Guidelines for investigating causality of sequence variants in human disease. Nature. 2014; 508(7497): 469–476. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaggerty CM, James CA, Calkins H, et al.: Electronic health record phenotype in subjects with genetic variants associated with arrhythmogenic right ventricular cardiomyopathy: a study of 30,716 subjects with exome sequencing. Genet Med. 2017; 19(11): 1245–1252. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Driest SL, Wells QS, Stallings S, et al.: Association of Arrhythmia-Related Genetic Variants With Phenotypes Documented in Electronic Medical Records. JAMA. 2016; 315(1): 47–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang S, Lincoln SE, Kobayashi Y, et al.: Sources of discordance among germ-line variant classifications in ClinVar. Genet Med. 2017; 19(10): 1118–1126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandrum MJ, Lee JM, Benson M, et al.: ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 2018; 46(D1): D1062–D1067. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q, Wang K: InterVar: Clinical Interpretation of Genetic Variants by the 2015 ACMG-AMP Guidelines. Am J Hum Genet. 2017; 100(2): 267–280. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenrie A, Eilbeck K: ClinVar Miner. 2018. Reference Source\n\nZhang X, Minikel EV, O'Donnell-Luria AH, et al.: ClinVar data parsing [version 1; referees: 2 approved]. Wellcome Open Res. 2017; 2: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXin J, Mark A, Afrasiabi C, et al.: High-performance web services for querying gene and variant annotation. Genome Biol. 2016; 17(1): 91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR: A language and environment for statistical computing. (R Foundation for Statistical Computing, Vienna, Austria). 2016. Reference Source\n\nWilks SS: Determination of Sample Sizes for Setting Tolerance Limits. Ann Math Stat. 1941; 12(1): 91–96.\n\nFligner MA, Wolfe DA: Nonparametric Prediction Intervals for a Future Sample Median. J Am Stat Assoc. 1979; 74(366): 453–456. Publisher Full Text\n\nSayers EW: A General Introduction to the E-utilities. In Entrez Programming Utilities Help [Internet]. Ch. 4, National Center for Biotechnology Information, 2009. Reference Source\n\nThe Variant Call Format Specification v. VCFv4.3 and BCFv2.2. 2017. Reference Source\n\nPython Language Reference v. 3.5.2. Samurai Media Limited, United Kingdom, 2015. Reference Source\n\nMcKinney W: Data Structures for Statistical Computing in Python. In Proceedings of the 9th Python in Science Conference. (AQR Capital Management, LLC) 2010; 51–56. Reference Source\n\nCock PJ, Antao T, Chang JT, et al.: Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics. 2009; 25(11): 1422–1423. PubMed Abstract | Publisher Full Text | Free Full Text\n\ndplyr: A Grammar of Data Manipulation v. R package version 0.7.4. 2017. Reference Source\n\nRColorBrewer: ColorBrewer Palettes v. R package version 1.1-2. 2014. Reference Source\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. (Springer-Verlag), 2009. Publisher Full Text\n\ngridExtra: Miscellaneous Functions for \"Grid\" Graphics v. R package version 2.3. 2017. Reference Source\n\nggExtra: Add Marginal Histograms to 'ggplot2', and More 'ggplot2' Enhancements. v. R package version 0.7 2017.\n\nMaxwell KN, Hart SN, Vijai J, et al.: Evaluation of ACMG-Guideline-Based Variant Classification of Cancer Susceptibility and Non-Cancer-Associated Genes in Families Affected by Breast Cancer. Am J Hum Genet. 2016; 98(5): 801–817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNussbaum RL, Yang S, Lincoln SE: Clinical Genetics Testing Laboratories Have a Remarkably Low Rate of Clinically Significant Discordance When Interpreting Variants in Hereditary Cancer Syndrome Genes. J Clin Oncol. 2017; 35(11): 1259–1261. PubMed Abstract | Publisher Full Text\n\nButler RR, Gejman PV: rbutleriii/Clinotator: Manuscript Public Release (Version v1.0.0). Zenodo. 2018. Data Source\n\nButler RR, Gejman PV: Clinotator Raw Data [Data set]. Zenodo. 2018. Data Source"
}
|
[
{
"id": "33507",
"date": "02 May 2018",
"name": "Kai 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\nIn this study, the authors introduced a new method named Clinotator to help researchers easily find the discrepancies on categorizations of clinical impacts among different submitters or curators. Besides providing re-classification prioritization scores, this study also provided new metrics (CTRS, CTAA, CTPS, and CTRR) for the interpretation ClinVar variants, which will be helpful for users of ClinVar data. Some additional improvements on description of the methods and interpretation of results can be made as described below.\n\nMajor:\nA brief description about the selection of the values for the assertion weight factor can be added to the methods. Although there is a discussion about this in the discussion part, it is not present in the Methods itself. Based on Figure 2A, it is still unclear how the CTRR (Clinotator reclassification recommendations) system works. The legend says “scoring workflow”, but this is not really a workflow figure. Maybe a flowchart with a simple example will help (for example, to show how moving from Uncertain Significance (US) to Benign/Likely benign (BLB) scores works for a particular variant based on a specific reason). For the definition of CTRR in page 4, is it suitable to define both the cases with “insufficient information” and “consistent identity” to be 0? It seems that the cases with “insufficient information” have a good chance to be re-classified in the future when more information is available. In the discussion part, the authors mentioned that Clinotator might classify the variant with five LP assertions as P status, and this issue was further explained by the authors through the fact that those LP assertions do provide additive values for overall pathogenicity. However, in the further discussion, the authors mentioned that “One of the key benefits of the current ACMG/AMP criteria is that any assertion must review all previous evidence and existing data available”, which means that the latest assertions were dependent on the previous ones. Will this affect the additive assumption? Maybe a discussion about this can help. rsID does not uniquely identify a variant. It is merely a locus identifier telling people an approximate genomic location. In general it is a bad idea to use rsID to denote a variant, especially since many variants can be at the exactly same locus (3 different SNPs, and many different indels can all be located at exactly the same genomic position and be represented by exactly the same rsID). ClinVar does assign its own ID, and it is a better idea to just use ClinVar’s own ID system when describing variants. One important discussion point is the use of phenotype information in the clinical interpretation of genetic variants. It can be incorporated in ACMG guidelines, but here for ClinVar, there is no phenotype information associated with a variant ID, so a pathogenic variant for one phenotype could simply be benign for another phenotype. It is something that needs to be discussed.\n\nMinor:\nThe word “Benign (B)” is in bold font in both of Figure 1 and Figure 2 legend. This is very confusing, because there is also an explanation of the panel (A), (B), etc. Too many abbreviations were used in the manuscript. It would be ideal to make a list to explain each of them? For example, in Table 3, there is no further explanation for VID, clinsig, even though there are explanations for other abbreviations. Both “rsID” and “RSID” are used in the manuscript, a more consistent naming scheme is needed. How is the “age of assertion factor” defined? Is it the current date (i.e. the date of preparing the manuscript, Feb 2018) minus the date of assertion? Maybe a clear description about this will help readers get a better understanding. In Table 3, it is better to use “Ambry Genetics” rather than “Ambry” In the “Data and software availability” part, should the Clinotator source code be https://github.com/rbutleriii/Clinotator rather than “https://github.com/rbuteriii/clinotator”?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "3721",
"date": "20 Jun 2018",
"name": "Robert Butler",
"role": "Author Response",
"response": "Thank you very much for the review. It was a very helpful and I went through and made changes to the manuscript and program, reflected below. The result is a new version of Clinotator, and a rerun of the original variant set, with updated analysis and figures to reflect the changes. Major 1. A brief description about the selection of the values for the assertion weight factor can be added to the methods. Although there is a discussion about this in the discussion part, it is not present in the Methods itself. I added an explanation of assertion weights to the methods, referencing the later Discussion section as well: “Initial values of US were tried from -0.5 to -0.1 in increments of 0.1, ultimately defined at -0.3. The assertion weight factor for LB was tried as several multiples of US (4, 5, 6, 10, 12, and 20). A 10-fold multiple, with LB equal to -3, eliminated overlap between US and LB distributions. The value of B was tested at a range of multiples of LB (3/2, 5/3, 2, 3, and 4), and fixed at a 2-fold value, with a B assertion weight of -6. No multiple of US could attain LP, which was set as the equivalent positive value to LB; and P was set 2-fold higher than LP (after trying a range of multiples as with B). Further rationale of weights is continued in the Results and Discussion sections.” 2. Based on Figure 2A, it is still unclear how the CTRR (Clinotator reclassification recommendations) system works. The legend says “scoring workflow”, but this is not really a workflow figure. Maybe a flowchart with a simple example will help (for example, to show how moving from Uncertain Significance (US) to Benign/Likely benign (BLB) scores works for a particular variant based on a specific reason). Figure 2 was reworked to better depict the CTRR classification and reclassifications by CVSZ. See figure. 3. For the definition of CTRR in page 4, is it suitable to define both the cases with “insufficient information” and “consistent identity” to be 0? It seems that the cases with “insufficient information” have a good chance to be re-classified in the future when more information is available. The reviewer makes an excellent point, I modified the software to separate CTRR=0 as consistent identity only. Insufficient information variants now receive a CTRR= “.”. Version 1.1.0 onward should reflect this change. 4. In the discussion part, the authors mentioned that Clinotator might classify the variant with five LP assertions as P status, and this issue was further explained by the authors through the fact that those LP assertions do provide additive values for overall pathogenicity. However, in the further discussion, the authors mentioned that “One of the key benefits of the current ACMG/AMP criteria is that any assertion must review all previous evidence and existing data available”, which means that the latest assertions were dependent on the previous ones. Will this affect the additive assumption? Maybe a discussion about this can help. I rephrased the section of the discussion to make this clear. “One of the key benefits of the current ACMG/AMP criteria is that any assertion must review all previous evidence and existing data available 4 . Note that this should not include old assertions, only evidence. Thus while old data never loses its value, old assertions do; particularly if they were made prior to the establishment of the current standards.” Additionally, in the Methods section: “Note that the CTRS metric only includes clinical assertions where the submitter has published a defined assertion criteria on the ClinVar website. Literature-only submissions such as those from OMIM are filtered out as they are a type of evidence, and not a clinical assertion. Assertions made without assertion criteria or with incomplete data are also omitted, as the reliability of these assertions is unknown.” 5. rsID does not uniquely identify a variant. It is merely a locus identifier telling people an approximate genomic location. In general it is a bad idea to use rsID to denote a variant, especially since many variants can be at the exactly same locus (3 different SNPs, and many different indels can all be located at exactly the same genomic position and be represented by exactly the same rsID). ClinVar does assign its own ID, and it is a better idea to just use ClinVar’s own ID system when describing variants. A fair consideration about rsIDs. In the Operation section I added a paragraph on data types to clarify the related concerns: “The preferred input file types are a VID list or a vcf file. The rsID list alone is inherently ambiguous, as multi-allelic rsIDs can have several VIDs associated, one with each allele. The rsID to VID conversion is not 1:1, so the table file generated will return rows for all possible VIDs associated with the rsID. Thus an rsID generated table may require additional matching using the alternate allele column (CVAL). However, vcf files will only be annotated with the correct rsID/alternate allele combination, preventing a mix-up for the vcf input type. Conversely some VIDs have multiple rsIDs, either because they are a haplotype variant, or due to other complications with rsID curation. The ‘vcf_match’ field addresses this reverse situation by identifying all rsIDs associated with a VID and its haplotype allele status.” I also added the alternate allele “CVAL” field description (formerly CVMA) to Table 1. It was already in the Clinotator output, and used in the backend of the vcf module, but overlooked in the paper methods. 6. One important discussion point is the use of phenotype information in the clinical interpretation of genetic variants. It can be incorporated in ACMG guidelines, but here for ClinVar, there is no phenotype information associated with a variant ID, so a pathogenic variant for one phenotype could simply be benign for another phenotype. It is something that needs to be discussed. A section has been added to the Discussion talk about this: “Some phenotype information is reported by Clinotator: the conditions associated with the submitter’s assertion. While it is possible to split assertions by condition and develop a clinical significance for each, this is currently too problematic to implement. For example, VID 9 has entries associated with “Hereditary hemochromatosis,” “Hemochromatosis type 1,” “Hemochromatosis type 1 (Autosomal recessive inheritance),” “Hemochromatosis juvenile digenic,” and “not provided” all with varying or absent identifiers in phenotype databases (MedGen, Human Phenotype Ontology, OMIM). As these are all potentially ambiguous classifications of the same or similar conditions, it would difficult to effectively group them without more comprehensive standardization.\" \"Additionally, some submitters provide a single assertion with multiple conditions associated, while others provide multiple assertions per variant, one for each condition. And some variants have two assertions total (as with VID 267572), which differently describe the same condition (hereditary breast and ovarian cancer). If these were split, there would not be enough information for Clinotator to calculate metrics on either. Of the variants examined here, 19,249 have only two valid clinical assertions with differing conditions for each; potentially excluding almost a third of ClinVar variants with multiple assertions. As ClinVar evidence grows, and phenotype ontologies become more sophisticated, it will become more feasible to split variant assertions by phenotype.” Additionally, the CVDS reporting has been modified to reflect the clinical significance associated with each condition assertion. This will hopefully integrate disease specific assertions downstream and in future versions. Minor 1. The word “Benign (B)” is in bold font in both of Figure 1 and Figure 2 legend. This is very confusing, because there is also an explanation of the panel (A), (B), etc. Fixed the bolding in the figure legend. 2. Too many abbreviations were used in the manuscript. It would be ideal to make a list to explain each of them? For example, in Table 3, there is no further explanation for VID, clinsig, even though there are explanations for other abbreviations. The VID is defined in the text body, but I also added it to Table 1 with rsID for more clarity. I additionally added both to all of the appropriate online table legends, as well as the “clinsig” definition. 3. Both “rsID” and “RSID” are used in the manuscript, a more consistent naming scheme is needed. Converted all the instances of RSID to the more consistent rsID. 4. How is the “age of assertion factor” defined? Is it the current date (i.e. the date of preparing the manuscript, Feb 2018) minus the date of assertion? Maybe a clear description about this will help readers get a better understanding. Description of age calculation added to the Metric Calculations section: “The Clinotator average assertion age (CTAA) is the mean age (in years) of valid clinical assertions. Each assertion’s age is calculated at the time of Clinotator script execution as the number of years since the clinical significance last evaluation date. Assertions without a last evaluation date are omitted.” I also added a “run on” date to the terminal/logfile, and the annotated vcf has a metadata field describing the Clinotator version and date. 5. In Table 3, it is better to use “Ambry Genetics” rather than “Ambry” Company name corrected. 6. In the “Data and software availability” part, should the Clinotator source code be https://github.com/rbutleriii/Clinotator rather than “https://github.com/rbuteriii/clinotator”? Url typos corrected."
}
]
},
{
"id": "33596",
"date": "08 May 2018",
"name": "Amalio Telenti",
"expertise": [
"Reviewer Expertise Genomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGenetic testing is increasingly considered in clinical grounds, and may accelerate as payers approve specific applications. The clinical validity is built on substantial evidence for pathogenicity of individual variants, in particular those that have been vetted by expert panels, and exceptionally, endorsed by guidelines. This parsimonious implementation of genetic information in the clinics (evidence, expert support, guidelines and reimbursement) is difficult because current techniques provide large amount of information that is incompletely assessed. Thus, only a small proportion of known variants are confidently classified and the community and corporate efforts aim at creating standards at the same time that a legacy of genetic variant interpretation is being updated. Prime among the efforts is ClinVar, a public database with information on over 300,000 variants. ClinVar uses various approaches to classify variants on the basis of the quality of evidence; and by doing so, exposes the number of entries that are conflicting, or just of unknown significance. ClinVar creates a platform that supports the ongoing classification and correction of errors.\n\nClinotator, described in this present publication, aims at highlighting entries that may need reappraisal. It does it primarily for batch analysis of variants or genes. This is achieved by building weights – mostly through heuristics – around key factors that contribute to misclassification: the age of the submission and last review of evidence, the standing of the submitter. The program results in a metric of pathogenicity and an alert for the need of reclassification. Calibration of the metrics and weights is achieved by using subsets of ClinVar variants with established criteria.\nClinotator adds to other efforts to improve the quality of annotation, most notably ClinVar Miner (https://clinvarminer.genetics.utah.edu). However it is important to underscore that other criteria, beyond the age of record, review of evidence and submitter, are important in the overall effort – in particular the use of disease prevalence1, and the consideration for penetrance of variants2. Given the current emphasis on machine learning in genetics and genomics, it is conceivable that a more comprehensive modeling of evidence and of biological basis of deleteriousness (eg, pathogenicity scores such as CADD3) may contribute increasingly accurate ascertainment of variants.\nGiven the considerations above, there are two aspects of the work that merit attention: (i) a more complete representation of the contribution of Clinotator in the context of other efforts, and (ii) a discussion of the likelihood that the current algorithm simply reclassifies variants between the overlap categories and their immediate neighbors. It would be useful to evaluate the accuracy of the reclassification by an external expert panel.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3720",
"date": "20 Jun 2018",
"name": "Robert Butler",
"role": "Author Response",
"response": "Thank you very much for the review, and the perceptive considerations posed. I reworked the analysis and paper to highlight the topics suggested by the reviewer and include the recommended references. Below are specific considerations: (i) a more complete representation of the contribution of Clinotator in the context of other efforts I reworked several elements to better define the context for Clinotator. Particularly in the Introduction and Discussion, I spell out the role of Clinotator as a secondary analysis tool for comparing primary analyses: individual clinical assertions. “Harrison et al. found that 87.2% of discordant variants were resolved by reassessment and data sharing 8 . New public data has recently been leveraged with private datasets to identify misclassified variants on the basis of variant penetrance given disease prevalence [ins cit] . However the majority of these reclassification efforts still rely on access to private data, which will continue to be an unavailable to most researchers for the foreseeable future.” “Since it is based entirely on data available in ClinVar, it requires no private dataset or access to external resources. To demonstrate its utility, we examined test sets of two-star, three-star, and four-star variants (per ClinVar’s review status star ratings) and variants in ClinVar with “Conflicting Interpretations of pathogenicity” (CI). Clinotator was able to confirm recently published concordance trends 6, 8, 18 , and identify several groups of discordant variants for further investigation. It accomplished this efficiently, using a large-scale systematic approach with a minimal computational effort.” “As shown above, Clinotator is a useful secondary analysis tool for identifying discrepant records amongst the large and complex ClinVar database. With limited resources, submitters and curators alike can utilize Clinotator metrics for prioritization of reclassifications and research.” Part of the usefulness of Clinotator as a secondary analysis is its lightweight nature, not requiring the user to have their own cohort data. If we modify Clinotator to include primary evidence in its analysis—including penetrance, prevalence, and prediction software—that runs the risk of a circular set of self-reinforcing assertions, built on other assertions which may already include those evidence categories. Still, those primary evidence types above are very valuable, and should be factored into new clinical significance assertions (with criteria provided). The evidential papers themselves can be entered into ClinVar as “Literature Reviews” evidence type for associated variants. Ideally then, that information can be readily incorporated into all of the subsequent assertions as they become more frequently updated. Tools like those mentioned in the introduction (InterVar, ClinVar Miner, MyVariant.info, HGMD, SCRP, etc.) and various annotation software are then the appropriate resources for generating updated primary assertions. (ii) a discussion of the likelihood that the current algorithm simply reclassifies variants between the overlap categories and their immediate neighbors. The reviewer is correct in noticing that the majority of CTRR=1 variants are likely shuffling between overlap categories and their neighbors. To make this stand out better in the paper, I expanded the analysis from the two use case scenarios (3-star uncertain and Conflicting Interpretation) to an overall assessment of variant classifications in two-star, three-star, four-star and conflict categories. Updating figures 2, 3, the Results and Discussion to this end (with software modifications suggested by Reviewer 1, the entire analysis was redone). Some relevant modifications: “For this purpose, the overlapping regions of PLP and BLB exist—not as yet another classification bin—as a measure of plasticity of borderline assertions. The quantitative nature of the CTRS also allows a given variant to transition out of the overlap should enough additional assertions arise, or if sparse limited assertions are not updated.” “The schematic in Figure 2A gives a demonstration of CTRR outcomes, depicted for two-star and three-star variants in Figure 2B and Figure 2C. Reclassification recommendations for all of the two-star variants in Figure 2B largely confirm that most variants shift by only a single position, if at all (see also Table 2). The most common shifts occurred between the overlap categories (BLB and PLP) and their immediate neighbors. This is likely the result of the altered definition of the overlap category, as opposed to a genuine reclassification recommendation. In Figure 3A, most two-star variants with a CTRR of zero follow one of three identifiable linear correlations between CTRS and CVNA. Given that only 212 out of 47,854 two-star variants demonstrate a CTRR more than 2, these results support previous research showing a fairly high general concordance in ClinVar 18, 37, 38 .” “The 5,807 three-star variant reclassification recommendations are depicted in Figure 2C. This distribution is notably different than two-star variants; expected given an overall CVCS based on the—in most cases single—expert review assertion. There are no overlap variants in BLB, and only three in PLP. The majority of variants still have CTRR values of zero or one, but more three-star variants had a CTRR of two, 4.2% versus 0.4% of two-star variants. All but one of the high priority for reclassification variants (CTRR = 3) were in the three-star group, and these stand out noticeably in the comparison of Figure 2.” These are some of the modifications regarding the overlapping categories, in a reworked analysis. This also makes the direction of the analysis clearer, and allowed me to flesh out discussion of the overlap categories and the benefit of redefining overlaps with a quantitative scale. I also then highlighted that conflicting interpretations and older 3-star assertions are key focal points for Clinotator. It would be useful to evaluate the accuracy of the reclassification by an external expert panel. Any submitter that properly conforms to the ACMG/AMP guidelines for an assertion is theoretically giving an expert opinion, but as Harrison, et al (2017) point out, there will always be disagreement as long as some level of professional judgement is involved in the determination. Thus, in the absence of an absolute correct standard, a consensus of professional judgements is the aim for Clinotator. I further detailed the Future Directions section to describe our next steps in terms of developing a longitudinal model for submitter accuracy in terms of historical classification: “…or the relative rates of reclassification in the different review status tiers. With more longitudinal data on variants as ClinVar grows, it may become possible to establish a submitter expertise structure based on number of assertions submitted, relative reassessment rate and/or number of misclassified variants.”"
}
]
}
] | 1
|
https://f1000research.com/articles/7-462
|
https://f1000research.com/articles/6-1976/v1
|
08 Nov 17
|
{
"type": "Software Tool Article",
"title": "bcbioRNASeq: R package for bcbio RNA-seq analysis",
"authors": [
"Michael J. Steinbaugh",
"Lorena Pantano",
"Rory D. Kirchner",
"Victor Barrera",
"Brad A. Chapman",
"Mary E. Piper",
"Meeta Mistry",
"Radhika S. Khetani",
"Kayleigh D. Rutherford",
"Oliver Hofmann",
"John N. Hutchinson",
"Shannan Ho Sui",
"Rory D. Kirchner",
"Victor Barrera",
"Brad A. Chapman",
"Mary E. Piper",
"Meeta Mistry",
"Radhika S. Khetani",
"Kayleigh D. Rutherford",
"Oliver Hofmann",
"John N. Hutchinson",
"Shannan Ho Sui"
],
"abstract": "RNA-seq analysis involves multiple steps from processing raw sequencing data to identifying, organizing, annotating, and reporting differentially expressed genes. bcbio is an open source, community-maintained framework providing automated and scalable RNA-seq methods for identifying gene abundance counts. We have developed bcbioRNASeq, a Bioconductor package that provides ready-to-render templates and wrapper functions to post-process bcbio output data. bcbioRNASeq automates the generation of high-level RNA-seq reports, including identification of differentially expressed genes, functional enrichment analysis and quality control analysis.",
"keywords": [
"RNA-seq",
"pipeline",
"quality metrics",
"differential expression",
"functional analysis",
"RMarkdown",
"report"
],
"content": "Introduction\n\nRNA sequencing (RNA-seq) analysis seeks to identify differential expression among groups of samples, providing insights into the underlying biology of a system under study1. Automating a full analysis from raw sequence data to functionally annotated gene results requires the coordination of multiple steps and tools. From the first data processing steps to quantify gene expression, to the data quality checks necessary for identification of differentially expressed genes2 and functionally enriched categories, RNA-seq analysis involves the repetition of commands using various tools. This is done on a per-sample basis, and each step can require varying degrees of user intervention. As a bioinformatics core facility that processes a large number of RNA-seq datasets, we have developed a Bioconductor (BioC)3 package called bcbioRNASeq to aggregate and automate the execution of tools for RNA-seq quality control (QC), differential expression and functional enrichment analysis as much as possible, while still retaining full, flexible control of critical parameters.\n\nThis package relies on the output of bcbio, a python framework that implements best-practice pipelines for fully automated high-throughput sequencing analysis (including RNA-seq, variant discovery, and ChIP-seq). bcbio is a community driven resource that handles the data processing component of sequencing analysis, providing researchers with more time to focus on the downstream biology. For RNA-seq data, bcbio generates QC and gene abundance information compatible with multiple downstream BioC packages. We briefly describe some of the tools included in the bcbio RNA-seq pipeline to help our users understand the outputs of bcbio that are used in the bcbioRNASeq package.\n\nTo ensure that the library generation and sequencing quality are suitable for further analysis, tools like FastQC4 examine the raw reads for quality issues. Cutadapt5 can optionally be used to trim reads for adapter sequences, along with other contaminant sequences such as polyA tails and low quality sequences with PHRED6,7 quality scores less than five. Salmon8 generates abundance estimates for known splice isoforms. In parallel, STAR9 aligns the reads to the specified reference genome, and featureCounts10 generates counts associated with known genes. bcbio assesses the complexity and quality of the RNA-seq data by quantifying rRNA content and the genomic context of alignments in known transcripts and introns using a combination of custom tools and Qualimap11. Finally, MultiQC12 generates an interactive HTML report in which the metrics from all tools used during the analysis are combined into a single dynamic file. bcbio handles these first stages of RNA-seq data processing with little user intervention.\n\nThe next stages of an RNA-seq analysis include assessing read and alignment qualities, identifying outlier samples, clustering samples, assessing model fit, choosing cutoffs and finally, identifying differentially expressed genes. These steps often occur in multiple iterations, and require more active analyst involvement to integrate multiple tools that accept input data with incompatible formats and properties (see Use Case section). For example, the featureCounts gene counts from STAR-based alignments (a simple matrix) are useful for quality control, providing many more quality metrics than the quasi-alignments from Salmon. However, the quasi-alignments from Salmon (which are imported by tximport into a list of matrices) have been shown to be more accurate when testing for differential gene expression13,14. Managing these disparate data types and tools can make analyses unnecessarily time consuming, and increases the risk of inconsistency between analyses. Given the complexity of the analysis, it is essential to report the final parameters and associated results in a cohesive, reproducible manner.\n\nbcbioRNASeq was developed to address these issues and ease the process of documentation and report generation. The package offers multiple R Markdown templates that are ready-to-render after configuration of a few parameters and include example text and code for quality control metrics, differential expression, and functional enrichment analyses. Although other packages have been developed to solve similar issues, bcbioRNASeq allows for tight integration with the bcbio framework, and provides a unified package with objects, functions and pre-made templates for fast and simple RNA-seq analysis and reporting.\n\n\nMethods\n\nAs noted, bcbio runs a number of tools to generate QC metrics and compute gene counts from RNA-seq data. Additional information on the bcbio RNA-seq pipeline is available on readthedocs). At the end of a bcbio run, the most important files are stored in a separate directory specified by the user in the bcbio configuration YAML file under the upload: parameter; this directory is called \"final\" by default. Within this directory there is a dated project directory containing quality metrics, provenance information, and data derived from the analysis that have been aggregated across all samples, e.g. count files. In addition, there is a directory corresponding to each sample that contains the binary alignment map (BAM) files and Salmon count data for that sample.\n\n\n\nThe final upload directory generated by bcbio is used as the input for bcbioRNASeq. Once the bcbio run is complete, you can open an R session and load the bcbioRNASeq library (available from our GitHub repository). Use the loadRNASeq() function to create a structured S4 object (see below) that contains all of the necessary information for downstream analysis. The only required argument when creating this object is the full path to the final directory. We also recommend that you use the interestingGroups argument to indicate variables that are present in your metadata that are of interest for the analysis. Note that bcbioRNASeq will transform all metadata column headings to lowerCamelCase format without spaces, dashes, periods or underscores; therefore interestingGroups should be specified in the same format. Once the S4 object is created, use the saveData() function to save it as an RData (.rda) file.\n\n\n\nThis S4 object is unique to the bcbioRNASeq package, as it contains all of the necessary data from the bcbio run required for analysis. From here, you can use various functions in bcbioRNASeq to perform analysis, make figures, and generate data tables and results files as we describe in later sections. This object is also used as the input for the R Markdown templates for report generation. First, we begin by describing the object in more detail.\n\nWe have designed a new S4 class named bcbioRNASeq, which is an extension of SummarizedExperiment15. Our S4 class adds a slot to SummarizedExperiment named bcbio that facilitates the inclusion of additional objects related to the experiment that cannot be contained in a regular SummarizedExperiment. The bcbio slot allows the incorporation of three additional data structures: the Salmon quasi-alignment data for differential expression analyses from tximport13, an automatically generated DESeqDataSet to provide support for quality control plots, and alternative counts generated by featureCounts. To see all available slots in the bcbioRNASeq object listed by name, you can use slotNames(bcb). Each of the slots are described in more detail below:\n\n@assays: ShallowSimpleListAssays containing count matrices derived from Salmon quasi-aligned counts imported with tximport. This slot is accessible with assays().\n\n– raw: raw counts, generated by Salmon and imported with tximport. These are the primary counts and can be accessed with assay().\n\n– normalized: Normalized counts, with DESeq2 sizeFactors applied.\n\n– tpm: transcripts per million (TPM), calculated by tximport.\n\n– tmm: trimmed mean of M-values, calculated by edgeR.\n\n– rlog: regularized log transformation, calculated by DESeq2.\n\n– vst: variance stabilizing transformation, calculated by DESeq2.\n\n@colData: DataFrame describing the columns (samples) of the count matrices slotted in assays(). This slot is accessible with colData().\n\n@elementMetadata: DataFrame describing the rows (genes) of the count matrices slotted in assays(). This slot is accessible with rowData().\n\n@NAMES: Ensembl gene identifiers; rownames of the matrices slotted in assays(), used in conjunction with rowData().\n\n@metadata: SimpleList containing any metadata relevant to the dataset and the information pertaining to/generated from previous steps in the workflow. This slot is accessible with metadata().\n\n– version: Version of bcbioRNASeq package used to generate the object.\n\n– uploadDir: Path to bcbio final upload directory.\n\n– sampleDirs: Paths of sample directories contained in bcbio upload.\n\n– projectDir: Path to project directory in bcbio upload.\n\n– template: Name of YAML file used to configure bcbio run.\n\n– runDate: Date of bcbio run completion.\n\n– interestingGroups: Groups of interest to use by default for quality control plot coloring.\n\n– organism: Latin species name (e.g. \"Homo sapiens\").\n\n– genomeBuild: Genome build (e.g. \"hg38\" or \"mm10\").\n\n– ensemblVersion: Ensembl annotation version (e.g. \"Ensembl Genes 90\"). Defaults to \"current\".\n\n– annotable: Ensembl annotations obtained from AnnotationHub with ensembldb16,17.\n\n– tx2gene: Transcript to gene identifier mappings.\n\n– lanes: Number of flow cell lanes used during sequencing.\n\n– yaml: bcbio run YAML containing summary statistics and sample metadata saved during configuration.\n\n– metrics: Sample quality metrics from bcbio analysis, generated from aligned counts produced by STAR and featureCounts.\n\n– sampleMetadataFile: Path to custom sample metadata file, used to override metadata saved in run YAML.\n\n– dataVersions: Genome versions used by bcbio.\n\n– programs: Program versions used by bcbio.\n\n– bcbioLog: bcbio run log.\n\n– bcbioCommandsLog: bcbio commands log.\n\n– allSamples: Whether the object contains all samples from the run.\n\n– date: Date the bcbio run was loaded into R with loadRNASeq().\n\n– wd: Working directory.\n\n– utilsSessionInfo: utils::sessionInfo() output.\n\n– devtoolsSessionInfo: devtools::sessionInfo() output.\n\n– unannotatedGenes: Character vector of gene identifiers present in the RNA-seq counts matrix (assays()) that are missing from the internal Ensembl annotations data.frame. This includes gene identifiers that are now deprecated on Ensembl or FASTA spike-ins.\n\n@bcbio: SimpleList used to store different R objects which are computed once and used as input for different plots or to other R packages functions. To access these secondary objects use bcbio().\n\n– tximport: tximport list of Salmon counts, to be used in conjunction with DESeq2.\n\n– DESeqDataSet: DESeq2 dataset generated from featureCounts derived data using an empty design formula. Used for quality control report.\n\n– featureCounts: aligned counts generated by STAR aligner and featureCounts. Used to generate quality metrics summary.\n\n\nUse case\n\nTo demonstrate the functionality and configuration of the package, we have taken an experiment from the Gene Expression Omnibus (GEO) public repository of expression data to use as an example use case. The RNA-seq data is from a study of acute kidney injury in a mouse model (GSE65267)18. The study aims to identify differentially expressed genes in progressive kidney fibrosis and contains samples from mouse kidneys at several time points (n = 3, per time point) after folic acid treatment. From this dataset, we are using a subset of the samples for our use case: before folic acid treatment, and 1, 3, 7 days after treatment.\n\nA pre-computed version of the example bcbioRNASeq object used in this workflow (bcb.rda) and the code for reproduction are available at the f1000v1 branch of our package repository. Alternatively, a minimal version of this example dataset is automatically loaded with the package library, and is accessible using data(bcb).\n\nFirst, load the bcbioRNASeq object and a few other libraries to demonstrate how to access the different types of information contained in the object.\n\n\n\nThe counts() function returns the abundance estimates generated by Salmon. Read counts for each sample in the dataset are aggregated into a matrix, in which columns correspond to samples and rows represent genes. Multiple normalized counts matrices are saved in the bcbioRNASeq object, and are accessible with normalized argument:\n\n1. Raw counts (normalized = FALSE); default.\n\n2. DESeq2 normalized counts (normalized = TRUE).\n\n3. Transcripts per million (normalized = \"tpm\").\n\n4. Trimmed mean of M-values normalization method (normalized = \"tmm\"). Also accessible with tpm().\n\n5. Regularized log transformation (normalized = \"rlog\").\n\n6. Variance stabilization transformation (normalized = \"vst\").\n\nFunctions exist to extract expression abundances in the various formats. Outlined below are the steps to save these counts external to the bcbioRNASeq object. These steps utilize functions from the DESeq2, edgeR and tximport packages both directly as well as within wrapper functions. For discussions on RNA-seq data normalization methods and count formats see 19; we typically save at least the DESeq2 normalized counts (library size adjusted) and transcripts per million counts (gene length adjusted) for further analyses.\n\n\n\nNote, the raw counts generated by featureCounts are different from the Salmon-derived counts slotted in assays(), as described above. This alternative counts matrix will have a different number of rows, and is accessible using bcbio(bcb, \"featureCounts\").\n\nA typical RNA-seq analysis requires multiple quality control assessments at the read, alignment, sample and model level. Most of the data required to make these assessments is automatically generated by bcbio; the bcbioRNASeq package makes it easier for users to access it. For instance, the Qualimap tool run as part of the bcbio pipeline generates various metrics that can be used to assess the quality of the data and consistency across samples. The output of Qualimap is stored in the bcbioRNASeq object, and the package has several functions to visualize this output in a graphical format. These plots can be used to check data quality. Visual thresholds appear in many plots to help the user to assess quality. For example, a vertical burgundy line is used as a warning threshold, indicating that any samples with values below that require close attention. A vertical black line threshold represents the optimal value. The default cutoffs for these thresholds can be easily changed using function arguments. Using metrics(bcb) allows the user to extract a data.frame with all metrics information used by the functions in the QC report. In this way, custom figures can also be easily created using the same data but with user-preferred packages.\n\nBelow, we provide several examples of recommended QC steps for RNA-seq data with a short explanation outlining their usefulness.\n\nTotal reads per sample and mapping rate are metrics that can help identify imbalances in sequencing depth or failures among the samples in a dataset (Figure 1A–B). Generally, we expect to see a similar sequencing depth across all samples in a dataset and mapping rates greater than 75%. Low genomic mapping rates are indicative of sample contamination, poor sequencing quality or other artifacts.\n\n\n\nSample classes, as defined by the interestingGroups argument, are represented by the different colors as defined in the legend of each plot. Vertical black lines indicate optimal values while vertical burgundy lines indicate cutoffs for substandard values. The total reads plot (A) indicates the total number of reads sequenced per sample and the mapping rate plot (B) shows the percentage of reads mapping to the reference genome. Here, all samples are well within recommended ranges, having well over 25 million reads and almost 90% of reads mapping. The exonic and intronic mapping rate plots (C and D) indicate the percentage of reads mapping to exons or introns, respectively. Here, all samples are within recommended ranges, with samples from day 3 and day 7 showing higher proportions of reads mapping to intronic versus exonic regions as compared to the day 1 and normal sample classes. The genes detected plot (E) indicates the total number of genes for each sample with a least one mapped read. Optimal gene detection values will vary based on an organism’s transcriptome size. The gene detection saturation plot (F) shows the relationship between the number of reads mapped and the number of genes detected. If this trend is not linear, it indicates that the sequencing may have been saturated in terms of detecting gene expression.\n\nFor RNA-seq, the majority of reads should map to exons and not introns. Ideally, at least 60% of total reads should map to exons. High levels of intronic mapping may indicate high proportions of nuclear RNA or DNA contamination. Samples should also have ribosomal RNA (rRNA) contamination rates below 10% (not shown) (Figure 1C–D).\n\n\n\nDetermining how many genes are detected relative to the number of mapped reads is another good way to assess the sample quality (Figure 1E–F). Ideally, all samples will have similar numbers for genes detected, and samples with higher number of mapped reads will have more genes detected. Large differences in gene detection numbers between samples can introduce biases and should be monitored at later steps for potential influence on sample clustering.\n\n\n\nComparing the distribution of normalized gene counts across samples is one way to assess sample similarity within a dataset. We would expect similar count distributions for all genes across the samples unless the library sizes or total RNA expression are different (Figure 2). The plotCountsPerGene() and plotCountDensity() functions provide two ways to visualize this comparison.\n\n\n\nNormalized count distributions are displayed as boxplots (A) and density plots (B). The log10 TMM-normalized counts per gene normalization method equates the overall expression levels of genes between samples under the assumption that a majority of them are not differentially expressed20. Therefore, by normalizing for total RNA expression by sample, we expect the spread of the log10 TMM-normalized counts per gene to be similar for every sample. Sample classes (as set with the interestingGroups argument) are highlighted in different colors. Here, there is high similarity among the samples.\n\nIt is important to explore the fit of the model for a given dataset before performing differential expression analysis. The normalized and transformed data can be used to assess the variance-expression level relationship in the data, to identify which method is best at stabilizing the variance across the mean. The plotMeanSD() function wraps the output of different variance stabilizing methods (including the vst() and rlog() transformations from the DESeq2 package) and plots them with the vsn package’s meanSdplot() function21 (Figure 3).\n\n\n\nPlots show the standard deviation of normalized counts (normalizedCounts) using log2() (A), rlog() (B), and variance stabilizing (vst()) (C) transformations by rank(mean). The red line shows the running median estimator. As (B) and (C) show, the transformations greatly reduce the standard deviation, and the rlog() transformation is most effective at stabilizing the variance across the mean.\n\nAnother plot that is important to evaluate when performing QC on RNA-seq data is the plot of dispersion versus the mean of normalized counts. For a good dataset, we expect the dispersion to decrease as the mean of normalized counts increases for each gene. The plotDispEsts() function provides easy access to model information stored in the bcbioRNASeq object, using the plotting code provided in the DESeq2 library (Figure 4).\n\n\n\nThe QC metrics assessed up to this point are performed to get a global assessment across the dataset and look for similar trends across all samples . However, often we have a dataset in which samples can be classified into groups and it is common to interrogate how similar replicates are to each other within those groups, and the relationship between groups. To this end, bcbioRNASeq provides functions to perform this level of QC with Inter-Correlation Analyses (ICA) and Principal Components Analyses (PCA) between samples. Furthermore, we can use the results of the PCA to identify covariates that correlate with principal components.\n\nSince these analyses are based on variance measures, it is recommended that the variance stabilized rlog transformed counts be used. Using these transformed counts minimizes large differences in sequencing depth and helps normalize all samples to a similar dynamic range . Simple logarithmic transformations of normalized count values tend to generate a lot of noise for low expressor genes, which can consequently dominate the calculations in the similarity analysis. An rlog transformation will shrink the values of low counts towards the genes’ averages across samples, without affecting the high expression genes.\n\nInter-Correlation analysis allows us to look at how closely samples are related to each other by first computing pair-wise correlations between expression profiles of all samples in the dataset and then clustering based on those correlation values. Samples that are similar to one another will be highly correlated and will cluster together. We expect samples from the same group to cluster together (Figure 5), although this is not always the case. We can also identify potential outlier samples using ICA, if there are samples that show low correlation with all other samples in the dataset. For more control over the graphing parameters of the ICA heatmap, other packages can be used to generate these plots by using the normalized data, accessed with counts (bcb, normalized = \"rlog\"), as input. One example is the pheatmap () function from the pheatmap package22, which underlies this plot.\n\n\n\nAll pairwise sample Pearson correlations are shown. Correlations are clustered by both row and column, with sample classes (as set with the interestingGroups argument) highlighted across the top of the heatmap. Here, the sample classes cluster well, with the normal and day 1 samples showing the highest intra-group correlations.\n\nPCA is a multivariate technique that allows us to summarize the systematic patterns of variations in the data23. PCA takes the expression levels for genes and transforms them in principal component space, reducing each sample to a single point. It allows us to separate samples by expression variation , and identify potential sample outliers. The PCA plot is a great way to explore both inter- and intra-group clustering (Figure 6). As with the ICA plots, other packages can be used to generate these kinds of plots from the normalized data, which can be accessed with counts(bcb, normalized = \"rlog\").\n\n\n\nThe first two principal components of the gene expression dataset are plotted here for each of the samples. Sample classes (as set with the interestingGroups argument) are highlighted in different colors. Alternatively, sample labels can be added with the label = TRUE argument (not shown) to identify individual samples, which is particularly useful for identifying outliers. Here, we see good clustering of the samples by group with no apparent outliers.\n\nWhen there are multiple factors that can influence the results of a given experiment, it is useful to assess which of them is responsible for the most variance as determined by PCA (Figure 7). The plotPCACovariates() function passes transformed count data and metadata from the bcbioRNASeq object to the degCovariates() function of the DEGreport package24. This method adapts the method described by Daily et al. for which they integrated a method to correlate covariates with principal components values to determine the importance of each factor25 (Figure 7).\n\n\n\nPCA is performed on transformed counts (rlog by default; but vst is also supported) and correlations between principal components and the different metadata covariates are computed. Significant correlations (FDR < 0.1) are shaded from blue (anti-correlated) to orange (correlated), with non-significant correlations shaded in gray. Here, none of variables show a significant correlation with any principal component of the data. Asterisks indicates p-value < 0.05.\n\nOnce the QC is complete and the dataset looks good, the next step is to identify differentially expressed genes (DEG). For this part of the workflow, we follow instructions and guidelines from the DESeq2 vignette, using Salmon-derived abundance estimates imported with tximport. As previously noted, a Differential Expression R Markdown template is available with the bcbioRNASeq package for these steps.\n\nThe first step is to define the factors to include in the statistical analysis as a design formula. We chose to study the difference between the normal group and the day 7 group in our dataset. In our example, we have only one variable of interest; however, DESeq2 is able to model additional covariates. When additional variables are included, the last variable entered in the design formula should generally be the main condition of interest. More detailed instructions and examples are available in the DESeq2 vignette.\n\n\n\nThe results from DESeq2 include a column for the P values associated with each gene/test as well as a column containing P values that have been corrected for multiple testing (i.e. false discovery rate values). The multiple test correction method performed by default is the Benjamini Hochberg (BH) method26. Since it can be difficult to arbitrarily select an adjusted P value cutoff, the alphaSummary() function is useful for summarizing results for multiple adjusted P value or FDR cutoff values (Table 1).\n\n\n\nUse the results() function to generate a DESeqResults object containing the output of the differential expression analysis. The desired BH-adjusted P value cutoff value is specified here with the alpha argument (< 0.05 shown).\n\n\n\nThe plotMA() function plots the mean of the normalized counts versus the log2 fold changes for all genes tested (Figure 8).\n\n\n\nEach point represents one gene, with mean expression levels across all samples plotted on the x-axis and the log2 fold change observed in the contrast of interest on the y-axis. Significant differentially expressed genes (with an adjusted P value less than the adjusted cutoff P value we chose earlier) are colored red27.\n\nThe plotVolcano() function produces a volcano plot comparing significance (here the BH-adjusted P value) for each gene against the fold change (here on a log2 scale) observed in the contrast of interest28 (Figure 9).\n\n\n\nThis plot compares the amount of gene expression change (not strictly interpretable here as these fold changes were derived from an LRT test) to the significance of that change (here plotted as the -log10 transformation of the multiple test adjusted P value), with each grey point representing a single gene. Options are available to highlight top gene candidates by shading (here, genes in the green shaded areas are bounded by a minimal fold change and -log10 adjusted P value cutoffs) and by text labeling. The two (optional) marginal plots showing the distributions of the log2 fold changes and negative log10 adjusted P values are useful in assessing cutoff choices and trade-offs.\n\nThe plotDEGHeatmap() function produces a gene expression heatmap useful for visualizing the expression of differentially expressed (DE) genes across samples. The heatmap shows only DE genes on a per-sample basis, using an additional log2 fold change cutoff. By default, this plot is scaled by row and uses the ward.D2 method for clustering29. The gene expression heatmap is a nice way to explore the consistency in expression across replicates or differences in expression between sample groups for each of the DE genes (Figure 10).\n\n\n\nThe heatmap shows only DE genes on a per-sample basis, using an additional log2 fold change cutoff. By default, this plot is scaled by row (centered and scaled) and uses the ward.D2 method for clustering29, with red and blue colors denoting higher and lower expression levels respectively. Results are clustered by both row and column. Heatmaps are drawn with the pheatmap() function of the pheatmap package22.\n\nIn addition to looking at the overall results from the differential expression analysis, it is useful to plot the expression differences for a handful of the top differentially expressed genes. This helps to check the quality of the analysis, by validating the expression for genes that are identified as significant. It is also helpful to visualize trends in expression change across the various sample groups (Figure 11). As bcbioRNASeq is integrated with the DEGreport package24, we can use DEGreport’s degPlot() function to view the expression of individual DE genes.\n\n\n\nGene expression patterns are shown for the top 3 (as selected in the function options) differentially expressed genes (by BH-adjusted P value). Normalized, transformed counts are shown for each replicate from each sample group; groups are set within the function options. Interestingly, even though our analysis compared day 7 to normal, all 3 genes show their greatest increases in gene expression at day 3, with some leveling off or relative decreases in expression at day 7.\n\nThe full set of example data is from a time course experiment, as described previously. Up to this point, we have only compared gene expression between two time points (normal and day 7), but we can also analyze the whole dataset to identify genes that show any change in expression across the different time points. As recommended by DESeq2, the best approach for this type of experimental design is to perform a likelihood ratio test (LRT) to test for differences in gene expression between any of the sample groups in the context of the time course. More information about time-course experiments and LRT is available in the DESeq2 vignette. This approach will yield a list of differentially expressed genes, but will not report how the expression is changing. Visualizing patterns of expression change amongst the significant genes is helpful in identifying groups of genes that have similar trends, which in turn can help determine a biological reason for the changes we observe (Figure 12). The DEGreport package includes the degPatterns() function, which is designed to extract and plot genes that have a similar trend across the various time points. More information about this function can be found in the DEGreport package24. Note that this function works only with significant genes; significants() returns the significant genes based on log2FoldChange and padj values (0 and 0.05 respectively, by default).\n\n\n\nClustered and scaled expression patterns for the top 500 differentially expressed genes. Each group represents a gene expression pattern shared among different DE genes, with the number of groups determined by the expression correlation patterns of the groups. Boxplots are shown for the expression patterns of each gene within the group to give a better idea of how well the groupings fit the expression data.\n\nThe output of the degPatterns() function is the plot, as well as a list object that contains a data.frame with two columns - the \"genes\" and the corresponding \"cluster\" number. To extract genes from a specific cluster for further analysis, the base function subset() can be utilized as follows to obtain the list of genes from cluster 8:\n\n\n\nFinally, at the end of the analysis a results table can be extracted containing the DE genes at a specified log fold change threshold. These results can be written to files in the specified output folder. In addition to the DE genes, a detailed summary and description of results and the output files are generated along with the cutoffs used to identify the significant genes.\n\n\n\nOnce the results table object is ready, the top up- and down-regulated genes (sorted by log2 fold change) can now be displayed. Here, we output the top 5 DE genes in each direction (Table 2).\n\n\n\nOutput files from this use case include the following gene counts files (output at the count normalization step):\n\nnormalized_counts.csv.gz: Use to evaluate individual genes and/or generate plots. These counts are normalized for the variation in sequencing depth across samples.\n\ntpm.csv.gz: Transcripts per million, scaled by length and also suitable for plotting.\n\nraw_counts.csv.gz: Only use to perform a new differential expression analysis. These counts will vary across samples due to differences in sequencing depth, and have not been normalized. Do not use this file for plotting genes.\n\nIf desired, rlog variance stabilized counts can also be output for future use in variance based plotting methods such as PCA. DEG tables containing the differential expression results from the analysis summary step are sorted by BH-adjusted P value, and contain the following columns:\n\nensgene: Ensembl gene identifier.\n\nbaseMean: Mean of the normalized counts per gene for all samples.\n\nlog2FoldChange: log2 fold change.\n\nlfcSE: log2 standard error.\n\nstat: Wald statistic.\n\npvalue: Wald test P value.\n\npadj: BH-adjusted Wald test P value (corrected for multiple comparisons; false discovery rate).\n\nsymbol: Ensembl gene name.\n\ndescription: Ensembl description.\n\nbiotype: Ensembl biotype (e.g. protein_coding).\n\nbroadClass: Broad class definition, based on the biotype (e.g. coding).\n\nTo gain greater biological insight into the list of DE genes, it is helpful to perform functional analysis. We provide the Functional Analysis R Markdown template, which contains code from the clusterProfiler package30, to identify potentially enriched biological processes among the DE genes. We use this tool to take the significant gene list and the background gene list (all genes tested) as input to perform statistical enrichment analysis of gene ontology (GO) terms using hypergeometric testing. The template includes a table of GO terms that are significantly enriched among the DE genes and a variety of plots summarizing the significantly enriched GO processes. The plots included are:\n\ndotplot: shows the number of genes associated with the top 25 most enriched terms (size) and the p-adjusted values for these terms (color).\n\nenrichMap: shows the relationship between the top 25 most significantly enriched GO terms, by grouping similar terms together. The color represents the P values relative to the other displayed terms (brighter red is more significant) and the size of the terms represents the number of genes that are significant from the significant genes list.\n\ncnetplot: shows the relationships between the genes associated with the top five most significant GO terms and the fold changes of the significant genes associated with these terms (color). The size of the GO terms reflects the P values of the terms, with the more significant terms being larger.\n\n\nR Markdown Templates\n\nTo facilitate analyses and compile results into a report format, we have created easy-to-use R Markdown templates that are accessible in RStudio31,32. Once the bcbioRNASeq package library has been installed, you can find these templates under FILE -> NEW FILE -> R MARKDOWN... -> FROM TEMPLATE. There are three main templates for RNA-seq: (1) Quality Control, (2) Differential Expression, and (3) Functional Analysis. You may need to restart Rstudio to see the templates. It is recommended that users run the reports in the order described above, as there may be functions that depend on data generated from the previous report. If you are not using RStudio, you can create new documents based on the templates using the rmarkdown::draft() function:\n\n\n\nThe instructions above will create an R Markdown (.Rmd) file from each of the templates. Each file begins with a YAML header, followed by sub-sections containing code chunks and some relevant text and/or a sub-heading to describe that step of the analysis. Each R Markdown file takes as input the bcbioRNASeq object, such that various functions from the package can be run on the data stored within the object to output figures, tables and carefully formatted results.\n\nNote you can add more text, headings and code chunks to the body of the R Markdown files to customize the reports as desired. Before rendering the file into a report you will want to run the prepareRNASeqTemplate() function in order to obtain the accessory files necessary for a fully working template.\n\n\n\nFinally, the main analysis parameters need to be specified in the YAML section on the top part of each document. For this example we assume R objects are stored in the folder data. Edit the following parameters in the Quality Control template:\n\n\n\nbcb.rda refers to the bcbioRNASeq object.\n\nIn the YAML section on the top of the Differential Expression template, modify the following parameters:\n\n\n\nIn the YAML section on the top of the Functional Analysis template, modify the following parameters (clusterProfiler and pathview have to be installed previously using this report):\n\n\n\nres.rda refers to a DESeqResults object from DESeq2 package.\n\nThe downloaded accessory files will be saved to your current working directory and should be kept together with your main analysis R Markdown files that were generated from the templates. These accessory files include: a) _output.yaml, to specify the R Markdown render format; b) _header.Rmd and c) _footer.Rmd, to add header/footer sections to the report; d) bibliography.bib, BibTex file for citations; and e) setup.R, to fill in some of the parameters related to figures and rendering format.\n\n\nConclusions\n\nHere we describe bcbioRNASeq, a Bioconductor package that provides functionality for quality assessment and differential expression analysis of RNA-seq experiments. This package supplements the bcbio community project, as it takes the output from automated bcbio RNA-seq runs as input, allowing for the data to be stored in a structured S4 object that can easily be accessed for various steps of the RNA-seq workflow downstream of bcbio. Built as an open source project, bcbio is a well-supported and documented platform for effectively using current state-of-the-art RNA-seq methods. Taken together, bcbio and bcbioRNASeq provide a full framework for rapidly and accurately processing RNA-seq data. The package also provides a set of configurable templates to generate comprehensive HTML reports suitable for biological researchers. With the use of R Markdown, all steps of the analysis are fully configurable and traceable.\n\nWe provide a full set of instructions for using bcbioRNASeq, including an example use case that demonstrates all of the main functionality. Quality control (pre- and post-quantification), model fitting, differential expression, and functional analysis provide a comprehensive set of metrics for evaluating the robustness of the RNA-seq results. Methods such as hierarchical clustering, principal components analysis, and time point analysis allow for an interactive examination of the data’s structure. Collectively, the workflow we describe can help researchers to identify true biological signal from technical noise and batch effects when analyzing RNA-seq experiments.\n\n\nSoftware and data availability\n\nCurrent source code: https://github.com/hbc/bcbioRNASeq\n\nWorkflow code: https://github.com/hbc/bcbioRNASeq/tree/f1000v1\n\nArchived source code (v0.1.1) as at time of publication: http://doi.org/10.5281/zenodo.103743933\n\nLicense: MIT\n\nThe data used in the use case can be accessed from NCBI GEO using accession GSE65267.",
"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\nAcknowledgments\n\nThanks to Mike Love of the Department of Biostatistics and Department of Genetics at the University of North Carolina at Chapel Hill for providing advice regarding RNA-seq differential expression and the use of the DESeq2 and tximport packages.\n\n\nReferences\n\nWang Z, Gerstein M, Snyder M: RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009; 10(1): 57–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove MI, Anders S, Kim V, et al.: RNA-Seq workflow: gene-level exploratory analysis and differential expression [version 2; referees: 2 approved] . F1000Res. 2016; 4: 1070. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, VJ Carey, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndrews S: FastQC: a quality control tool for high throughput sequence data. 2010. Reference Source\n\nMartin M: Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet Journal. 2011; 17(1): 10–12. Publisher Full Text\n\nEwing B, Hillier L, Wendl MC, et al.: Base-calling of automated sequencer traces using phred. i. Accuracy assessment. Genome Res. 1998; 8(3): 175–185. PubMed Abstract | Publisher Full Text\n\nEwing B, Green P: Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res. 1998; 8(3): 186–194. PubMed Abstract | Publisher Full Text\n\nPatro R, Duggal G, Love MI, et al.: Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017; 14(4): 417–419. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: featurecounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014; 30(7): 923–930. PubMed Abstract | Publisher Full Text\n\nOkonechnikov K, Conesa A, García-Alcalde F: Qualimap 2: advanced multi-sample quality control for high-throughput sequencing data. Bioinformatics. 2016; 32(2): 292–294. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEwels P, Magnusson M, Lundin S, et al.: MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics. 2016; 32(19): 3047–3048. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoneson C, Love MI, Robinson MD: Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences [version 2; referees: 2 approved]. F1000Res. 2016; 4: 1521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobert C, Watson M: Errors in RNA-Seq quantification affect genes of relevance to human disease. Genome Biol. 2015; 16(1): 177. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan M, Obenchain V, Hester J, et al.: SummarizedExperiment: SummarizedExperiment container. 2017. Publisher Full Text\n\nMorgan M: AnnotationHub: Client to access AnnotationHub resources. 2017. Publisher Full Text\n\nRainer J: ensembldb: Utilities to create and use ensembl-based annotation databases. 2017. Publisher Full Text\n\nCraciun FL, Bijol V, Ajay AK, et al.: RNA Sequencing Identifies Novel Translational Biomarkers of Kidney Fibrosis. J Am Soc Nephrol. 2016; 27(6): 1702–1713. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi P, Piao Y, Shon HS, et al.: Comparing the normalization methods for the differential analysis of illumina high-throughput RNA-Seq data. BMC Bioinformatics. 2015; 16: 347. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, Oshlack A: A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010; 11(3): R25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, von Heydebreck A, Sültmann H, et al.: Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics. 2002; 18 Suppl 1: S96–104. PubMed Abstract | Publisher Full Text\n\nKolde R: pheatmap: Pretty Heatmaps. 2015. Reference Source\n\nJolliffe IT: Principal component analysis. Wiley Online Library, 2002. Publisher Full Text\n\nPantano L: DEGreport: Report of DEG analysis. 2017. Publisher Full Text\n\nDaily K, Ho Sui SJ, Schriml LM, et al.: Molecular, phenotypic, and sample-associated data to describe pluripotent stem cell lines and derivatives. Sci Data. 2017; 4: 170030. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenjamini Y, Hochberg Y: Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 1995; 57(1): 289–300. Reference Source\n\nDudoit S, Yang YH, Callow MJ, et al.: Statistical methods for identifying differentially expressed genes in replicated cDNA microarray experiments. Stat Sin. 2002; 12(1): 111–139. Reference Source\n\nCui X, Churchill GA: Statistical tests for differential expression in cDNA microarray experiments. Genome Biol. 2003; 4(4): 210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWard JH Jr: Hierarchical grouping to optimize an objective function. J Am Stat Assoc. 1963; 58(301): 236–244. Publisher Full Text\n\nYu G, Wang LG, Han Y, et al.: clusterprofiler: an R package for comparing biological themes among gene clusters. OMICS. 2012; 16(5): 284–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAllaire JJ, Cheng J, Xie Y, et al.: rmarkdown: Dynamic Documents for R. 2017. Reference Source\n\nRStudio Team: RStudio: Integrated Development Environment for R. RStudio, Inc., Boston, MA, 2016. Reference Source\n\nSteinbaugh M, Pantano L, Barrera V, et al.: hbc/bcbioRNASeq: v0.1.1. Zenodo. 2017. Data Source"
}
|
[
{
"id": "27761",
"date": "20 Nov 2017",
"name": "Charlotte Soneson",
"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 describes bcbioRNASeq, an R package for analysis of RNA-seq data for which gene expression estimation and quality assessment have been done via the bcbio pipeline. The package contains functions for generating plots and performing downstream analysis as well as Rmarkdown templates for generating stand-alone reports of the quality control, differential expression analysis and functional analysis steps.\n\nIn general, I find the article well written and easy to follow. The included functionality covers the most important parts of a typical RNA-seq analysis, and the package is likely to be a useful tool for bcbio users. Also, the provided templates are easy to extend with additional analyses if necessary. Following are some suggestions for improvement of specific parts of the article.\n1. It would be good to indicate any dependencies on particular versions of R/Bioconductor/specific packages, and preferably also give the session info for the session with which the manuscript was generated. I ran the code with R v3.4.2 (Bioconductor v3.6, DESeq2 v1.18.1, bcbioRNASeq v0.1.2), and while most of the code executes correctly, there are some lines that do not. In particular: > normalized <- counts(dds, \"normalized\")\n\n## dds doesn't exist > tpm <- tpm(txi)\n\n## txi doesn't exist > writeCounts()(raw, normalized, rlog, tpm)\n\n## formatting error > resTbl <- resultsTables(res, lfc = 1, write = TRUE, dir = deDir)\n\n## deDir doesn't exist\n\nOther lines only execute properly with a functional internet connection, which could perhaps be indicated in the text: > plotVolcano(res) > resTbl <- resultsTables(res, lfc = 1, write = TRUE, dir = \".\")\n\nFinally, in some places, executing the provided code does not seem to generate the same results as in the article. More precisely: > plotPCACovariates(bcb, fdr = 0.1) does not generate Figure 7 (the asterisks are missing). > alphaSummary(dds, contrast = c(factor = \"group\", numerator = \"day7\", denominator = \"normal\")) does not generate the numbers in Table 1. > resPatterns <- degPatterns(counts(bcb, \"rlog\")[significants(res, ], metadata = colData(bcb), time = \"group\", col = NULL) does not generate Figure 12. As a consequence, subsetting to cluster 8 also doesn't work. > topTables(resTbl, n = 5) does not generate the numbers in Table 2.\n2. If I read correctly, the \"tximport\" slot of the bcbio object contains length-scaled TPMs, not aggregated transcript counts from tximport. This should be made clearer in the article. Is there an explicit choice in the bcbio pipeline that determines the type of count-scale abundances that are generated?\n\n3. It would be useful to indicate in the beginning of the article where the metadata is stored in the output from bcbio. I.e., where should one look for the values available to supply to the \"interestingGroups\" argument of loadRNASeq()?\n4. In the object description, it would be worth explaining a bit more clearly how the values contained in the slot \"-tmm: trimmed mean of M-values, calculated by edgeR\" were calculated.\n5. In the object description, devtools::sessionInfo() should be devtools::session_info()\n6. In the Use case, \"Also accessible with tpm()\" should presumably be under point 3.\n7. Regarding the visual thresholds in the plots (warning thresholds and optimal values), how are the default values determined? Are they fixed, or do they depend on some characteristics of the data? And are they particularly suitable for data generated under specific conditions, in specific organisms or with particular protocols? I am also wondering whether the use of the word \"optimal\" to designate one threshold may cause confusion. For example, if the \"optimal\" total number of reads is ~20M, and the \"optimal\" mapping rate is ~90%, it may not be immediately clear how one should interpret values exceeding (and potentially far away from) these \"optimal\" values. Finally, is there a reason for only having one line in the exonic and intronic mapping rate plots, but both lines in the other plots?\n\n8. In the \"Model fitting\" section, it is suggested that it is important to evaluate the variance stabilizing performance of different transformations before the differential expression analysis. However, the transformed data are never actually used for the DE analysis (which is performed with DESeq2). Thus, it should be clarified how the results obtained here are used to inform the downstream analysis.\n\n9. For the QC and differential expression analysis, the article outlines the analysis steps in detail. However, the functional analysis is only described through the existence of an Rmarkdown template. It would be nice to have at least part of the functional analysis also explained and written out in the article.\n\n10. In resPatterns[[\"plot\"]], the x-axis labels are not centered under the respective boxplots.\n\n11. The package is referred to as a \"Bioconductor package\", but as far as I can see it is not (yet) in Bioconductor.\n\n12. It seems that zero counts are excluded from the plots in Figure 2. This could be clarified in the text.\n\n13. In some places, \"library\" is used in the place of \"package\".\n14. For the degPatterns() call, it is indicated that it is \"CPU intensive\". It might be useful to indicate approximately *how* CPU/time intensive, since all other steps in the workflow execute quickly.\n15. In the code blocks, it would be easier if non-code characters like > and + were removed, so that the code could be directly copied into an R session.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "27759",
"date": "12 Dec 2017",
"name": "Davide Risso",
"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 bcbioRNASeq, an R package for the QC and differential expression analysis of RNA-seq data. The package takes as input the output of the bcbio software, not presented in this work.\nbcbio is a community driven resource that handles the data processing of several high-throughput sequencing applications, ranging from variant calling to ChIP-seq and RNA-seq analyses. The bcbioRNASeq package focuses on RNA-seq.\nOverall, I enjoyed the article and I think it represents a nice resource for practitioners looking to perform standardized RNA-seq analyses of several datasets and for bcbio users that want to perform high-level statistical analyses after RNA-seq preprocessing.\nThe article is well written and provides a reproducible example that walks the reader through the proposed pipelines. I am glad to report that (except for some minor issues reported below) I was able to fully reproduce the analysis. The following points will hopefully help the authors improve the manuscript.\nCurrently, the package does not appear to be available through Bioconductor, but only through the authors' Github. Are the authors planning to submit it to Bioconductor? I definitely encourage them to do so, as a way to manage package versions and dependencies (see next point). If not, I would ask the authors to refer to bcbioRNASeq as an R package rather than a Bioconductor package.\n\nThe authors do not specify the versions of the packages needed for their workflow to work. Although the DESCRIPTION file of the package provides such information, adding it to the manuscript would allow readers to reproduce the workflow example. This would be automatically taken care of if the package was part of a Bioconductor release. In addition, at the beginning of the paper, the authors load the packages \"DESeq2\" and \"DEGreport\". Aren't these packages in the Import: field of the DESCRIPTION file of bcbioRNASeq?\n\nS4 object. Is it really needed to store all the normalized data in the S4 object? This could lead to a huge object when the analysis is run on hundreds of samples. Since scaling normalization is very fast wouldn’t it be better to compute normalized data on the fly and only store the raw and tpm data computed by tximport and featureCounts? On a related note, wouldn’t it be better for the authors to store the featureCounts data in an additional element of the assays() slot and provide coercion methods from their object to the DESeqDataSet and tximport objects? Is it really needed to store both tximport results in the assays slot and in the bcbio slot? Overall, I have the feeling that the object is needlessly big and this could lead to a big memory footprint.\n\nAre the plots based on raw or normalized data? If the latter, which normalization / transformation is used by default? How does the user change it?\n\nInterpretation of the plots. Although the authors describe the plots in generic terms, it would be useful to explain them more specifically referring to the actual example analysis. For instance, which of the three transformation of Figure 3 is best for the example data? Is Figure 1F a typical pattern or does it uncover something unusual with the data? Same for Figure 7.\n\nA better metric to highlight the difference in distribution among samples (Figure 2) is the Relative Log Expression (RLE) plot. The authors might want to include such plot to their already excellent array of QC plots.\n\nWhat to do if the data fail the QC step? The authors present all their QC plots but then move on by simply stating \"Once the QC is complete and the dataset looks good [...]\". What if the data do not look good? It would be good to advice on what to do (just as a discussion perhaps). E.g., there could be outlying samples to be removed or batch effects to be accounted for in the model.\n\nMinor issues:\nThe last sentence of the first paragraph of the Introduction seems to indicate that the R package actually runs the QC tools, while these are run in bcbio and the results are loaded in the R package for exploration. First chunk of R code (page 4 of the PDF version of the paper): at my first read I was wondering how to get the data to run this command. It should be made clearer that this is only meant to show the syntax and is not part of the runnable example. The link to the bcb.rda file is broken. `normalized <- counts(dds, \"normalized\") ` This line doesn’t work. Did the author mean normalized=TRUE? `tpm <- tpm(txi) ` This line doesn’t work. Did the author mean `tpm <- counts(dds, normalized=“tpm”)`? Please describe writeCounts(). Please consider removing the \"+\" in the R chunks (e.g., at page 13 of the PDF) so that readers could run the code by copying and pasting into an R session. In statistics, ICA is often used to refer to Independent Component Analysis, so the authors may want to avoid this acronym for the correlation analysis to avoid confusion. In my RStudio session the plotPCACovariates() plot did not work (I couldn't see any points but just a gray background). YAML parameters for the Functional Analysis Rmarkdown: I believe that the line \"res\" should be \"resFile\".\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/6-1976
|
https://f1000research.com/articles/7-803/v1
|
20 Jun 18
|
{
"type": "Opinion Article",
"title": "A new paradigm for the scientific enterprise: nurturing the ecosystem",
"authors": [
"Alexander K. Lancaster",
"Anne E. Thessen",
"Arika Virapongse"
],
"abstract": "The institutions of science are in a state of flux. Declining public funding for basic science, the increasingly corporatized administration of universities, increasing “adjunctification” of the professoriate and poor academic career prospects for postdoctoral scientists indicate a significant mismatch between the reality of the market economy and expectations in higher education for science. Solutions to these issues typically revolve around the idea of fixing the career \"pipeline\", which is envisioned as being a pathway from higher-education training to a coveted permanent position, and then up a career ladder until retirement. In this paper, we propose and describe the term “ecosystem” as a more appropriate way to conceptualize today’s scientific training and the professional landscape of the scientific enterprise. First, we highlight the issues around the concept of “fixing the pipeline”. Then, we articulate our ecosystem metaphor by describing a series of concrete design patterns that draw on peer-to-peer, decentralized, cooperative, and commons-based approaches for creating a new dynamic scientific enterprise.",
"keywords": [
"academia",
"higher education",
"independent scholarship",
"careers",
"science studies",
"politics of science",
"systems-thinking",
"peer-to-peer science",
"collaboration"
],
"content": "Introduction\n\nThe institutions of science are in a state of flux. Declining public funding for basic science1,2 has led academic institutions to change their business models3,4; the administration of universities is becoming increasingly corporatized5. With increasing “adjunctification” of the professoriate6, continued use of outdated funding models for research science in academia, and dwindling academic career prospects for postdoctoral scientists7–10, it is clear that there is a mismatch between the reality of the market economy and expectations in higher education.\n\nThe evolving funding landscape at academic and research institutions has had a major impact on career opportunities for scientists, particularly those who are early-career. As a result of grant dollars being increasingly awarded to a disproportionately small number of established investigators and institutes11, intellectual discovery has become captured by a privileged few12, leading to greater bias in scientific research, diminished scientific productivity13, and less potential for breakthrough discoveries14,15. Such a lack of social diversity and equity is a major challenge in science, technology, engineering, and mathematics (STEM)16,17. Solutions are often sought out by proposing adjustments to the “career pipeline”, but these issues in STEM continue to be unresolved.\n\nThe career pipeline envisions a straight career path, from higher-education training to a coveted permanent position, and then up a career ladder until retirement (Figure 1). While such a direct path to success may be optimal for some, it does not reflect the reality of typical career development (Box 1). In today’s economy, permanent positions are becoming rarer across all industries, including universities, which are as more contract positions that are short-term and require no employee benefits are offered18.\n\nThe pipeline includes formal scientific training and different scientific career paths. The pipeline is characterized as a set of distinct streams with little flow between each stream, and a career ladder within each sector.\n\nThe alarming disproportion between the number of people with PhDs and the number of university-based academic positions available for PhDs has become a major preoccupation in the trade science press and beyond25. Indeed, training models for graduate students, and particularly for PhDs, in STEM often focus on delivering them to a tenure-track faculty position. Decades-long reliance on graduate students and post-doctoral researchers as cheap labor contributes to today’s unsustainable academic models and underemployment of academic scientists. Such a system has become a poverty trap for many graduate students and early PhDs, as they work long hours for low wages under the expectation that their participation in the pipeline will eventually lead to a permanent position in academia. The reality is that only 8% of postdocs are able to land a tenure-track job within 6 years of being awarded a PhD26.\n\nMuch of the discussion around career prospects for PhDs assumes that they must find a traditional position in a university in order to continue pursuing their scientific goals27. Funding changes have also produced an academic structure that provides limited prospects for early-career scientists to advance their careers within academia9. Postdoctoral training periods also continue to expand. While the increasing complexity of research may require longer training periods, it is unlikely that longer postdoc positions result in better researchers; many postdocs rarely get the appropriate direct training and mentoring to start an independent lab28. It is unknown how many promising, early-career scientists become trapped in postdoctoral limbo, as the morass of titles given to postdocs disguises the scale of the scientific workforce that exists in this state29. The ever-lengthening apprentice time for scientists has created a Red Queen's Race: scientists must run twice as fast to stay in the same place with their point of “independence” postponed almost indefinitely30.\n\nAdditionally, an outdated mentality still persists that the path to faculty tenureship requires putting science ahead of all of life’s other priorities, and this can have a severely negative effect on the mental health of those who try to conform (e.g., 31). While this model may have worked decades ago for those (mostly male) scientists that could rely on compliant spouses to raise families and perform domestic duties, it does not work in today's world. By presenting it as such, the pool of tenured faculty is limited to those who have the means to commit to such a lifestyle: typically young, male, unencumbered with children and geographically unconstrained. This demographic is steadily decreasing proportionally across the whole scientific research community, so career advice solely targeting this group is increasingly irrelevant.\n\nEven after gaining a tenure-track position, the mechanics of gaining tenure can be just as rigid and unforgiving32. The intense pressure on faculty to be “productive” (i.e., larger grants, publications in “high impact” journals), as well as the race to achieve institutional measures of “excellence”33, all divert attention from scholarship2,28,34,35 and undermine the creativity that nourishes scientific progress36–38. There have been serious concerns that such hypercompetition has led to a drop in scientific quality, especially in the biomedical field, as labs vie to publish their papers in the limited number of “slots” in high profile journals39,40. Indeed, the existence of fairly accurate predictive models41 of one’s chances of “becoming a principal investigator” based on affiliation, publication, and grant metrics, is itself a symptom of a profoundly narrow view of “success” in science.\n\nThe limitations of the pipeline as a conceptual model for education and careers is being recognized in both the tech industry19 and science20. The consequences of continuing to apply this outdated model is stalled career development in science, underemployment of some of the most highly educated people in our society, and overall loss of STEM professionals as they seek out career alternatives21–23. In 2016 $3 billion were invested into federal agencies to support STEM education programs24. Considering the governmental and individual investment that is made into higher STEM education each year, this is not just an academic conundrum—it is a societal problem.\n\nBy persisting with the assumption of the pipeline, we also miss engaging in conversations that address the fundamental cultural change that is occurring in science today. A new conceptual model is needed to help guide both early-career scientists and those who care about the scientific enterprise towards a more sustainable and resilient professional future. In this paper, we propose and describe the term “ecosystem” as a more appropriate way to conceptualize today’s scientific training and the professional landscape of the scientific enterprise. First, we highlight the issues around the concept of “fixing the pipeline”. Then, we articulate our ecosystem metaphor by describing a series of design patterns that draw on peer-to-peer, decentralized, cooperative, and commons-based approaches to science. We finish by describing the related cultural shifts underway that can hasten a more diverse and fluid scientific enterprise into the 21st century.\n\n\nFixing the pipeline\n\nMuch effort has gone into improving the efficiency of the scientific “career pipeline” (Figure 1). Proposed solutions tend to fall into one of three categories: adjusting the rates of flow in the pipeline, modifying oneself to fit the pipeline, or switching to an “alternative” pipeline. These solutions focus primarily on accessing tenure track faculty positions, as this issue has received the most attention in literature regarding the challenges facing the “pipeline” for scientists (Box 1). Here, we highlight some of the main issues facing these proposed solutions.\n\nAlberts et al.35 proposed that the supply of the pipeline (i.e., the number of trainees) should be adjusted to match the demand (i.e., the number of tenure-track positions). Increasing the number of tenure-track-style positions, as desirable as that might be, seems unlikely given the current trends in science funding. By addressing the demand end, it has also been suggested that the size of individual labs should be decreased to reduce the number of trainees that must be moved into faculty positions10,42. While these proposed reforms are thoughtful and well-intentioned, they do not address the problems of an oversupply of talented researchers and funding models that rely on large numbers of low-paid trainees to get work done. Such a solution could certainly lead to fewer scientists with more career stability, but is this the future that we envision for science and our society?\n\nToday’s search for a tenure track position is often a multi-year process as many open positions are saturated with candidates43. Many candidates must go to great lengths to make themselves more competitive. Recognizing this need, academic departments, as well as for-profit companies44, now assist with academic “survivorship” to help potential candidates prepare CVs, develop interview skills, and write research statements. This is leading to a system that is even less accessible to those with lesser means45 and may contribute even further to a bias towards hiring candidates from elite institutions46. It also contributes to the ever-lengthening time period before a scientist can “fledge” as a professional in academia. Much of the energy, currently being devoted to preparing individuals to adapt to the system, could be redirected towards more collaborative and collective solutions.\n\nThe paucity of traditional academic jobs has led to increased career advice about alternatives to academic careers, and there are encouraging indications that many PhDs are finding employment outside academia47,48. Although most of these efforts are entirely well-meaning, they sometimes have the unintended consequence of reinforcing the same tendencies as “pipeline-thinking”. For example, a recent Nature editorial discussed the ability of PhDs or postdocs to “stay in touch with science” by working in a “related function such as administration, outreach or publishing”48. While these career options suit many people, the possibility of doing any future self-directed science outside of an academic or research position is rarely considered. Such a narrative is disempowering overall, as junior researchers are encouraged to “take a hard look at their job prospects” rather than question the nature of institutions or the pipeline25. Other discussions of the STEM pipeline are framed by discussions of “workforce needs” and the production of workers that are “globally competitive”49,50, reinforcing a business approach to science51, rather than on doing science as a public good or for its own sake52.\n\nMuch of the career advice on how to be successful in the academic science pipeline reflects the values and dynamics present in the job market when many senior scientists obtained their first positions (e.g., that outlined by Diamandis53), instead of the realities of today54,55. As many of these scientists’ career experiences are often limited to academia, their perspectives may not be relevant to early-career scientists who are open to other career options. Academics often fail to recognize the broad applicability and value of PhD degrees, or encourage their trainees to work outside the traditional academic pipeline.\n\nBy framing solutions in terms of “fixing the pipeline”, the underlying career structures for scientists remain largely unchallenged. As such, early-career scientists occupy a passive role, waiting for change to come from the top, such as through institutional reform driven by senior leaders. Likewise, the scarcity of research positions is accepted as a given, limiting how much science can be done. There is, no limit on society’s need to address complex challenges, the number of research questions that can be asked, or the amount of scientific work that can be done. New models are needed to help identify different ways for scientists to continue their work outside of a standard academic or agency job.\n\n\nThe science ecosystem\n\nWe propose an ecosystem as a conceptual model that is relevant both to the training of a scientist and their role as a professional (Figure 2). The two most inner circles in the Figure depict the basic necessities, training, and professionalism of science. Here, traditional scientific labs may still have a role, but the networks of peer-to-peer collaborators that span both within and outside of institutions are emphasized. The two outermost circles are the impetus behind the changing context of science today. It is becoming more evident that a new systems-based approach is needed to allow science to adapt more quickly to the complex socio-political and biophysical context of today (the outermost circle). There are, however, now new resources, tools, and infrastructure (courtesy of STEM advances), such as lab space, journal access, and high-performance computing, either publicly available, or available for rent, that allow science to thrive outside of traditional institutions (the orange, next outermost circle)56. In addition, bottom-up changes are already being driven by early career scientists themselves in many different ways57–60.\n\nThe inner circle (beige) represents the basic necessities needed to be a functioning member of society, as well as a scientist. The next circle (purple) shows the different groups that are often involved in the pursuit of knowledge and scientific progress. Because the circle can be rotated there is no ‘up’ in this representation; no one type of institution is privileged in this representation and there is exchange in all directions. In addition, the borders between the different institutes are highly porous—there is collaboration, reflection, and sharing of resources between them. The next circle (orange) represents different kinds of resources and infrastructure needed to support science. The outermost circle (light blue) represents the environmental context, including biophysical limitations, and the socio-political and economic landscape, that science and scientists must function within, adapt to, and seek to understand and affect.\n\nMany postdocs and adjunct scientists already have the majority of tools that they need to do independent science, such as deep training and understanding of their field, a body of work that demonstrates their scientific ability, pre-existing networks of colleagues with similar intellectual interests, and the Internet to collaborate and share. By moving beyond the existing pipeline model of academic science, the ecosystem vision provides the space, flexibility, and diversity that science needs to be more responsive to both local and broader complex scales affecting science.\n\nTo demonstrate how an ecosystem model would work in practice, we present a set of conceptual design patterns loosely inspired by commons-based approaches61–63, systems-thinking approaches64, and the sustainable livelihoods framework65. We acknowledge specific social movements and grassroots changes that are occurring today, and demonstrate how science now has the means to be more egalitarian, inclusive, and diverse by being less dependent on their institutional settings. We recognize, that major institutional reforms are needed to realize this vision to its fullest, so we also address the changing role of institutions within this vision. We deliberately chose the term “ecosystem” to resonate with many of the phenomena that exist in biological ecosystems: diversity, resilience, multiplicity of scales, dynamic feedback loops, etc., and we use some of these concepts when framing each of the design patterns. That said, we do not claim a one-to-one correspondence with biological ecosystems.\n\nBasic necessities (i.e., Maslow’s hierarchy of needs) are fundamental to any human livelihood, and certainly for a scientist to be able to flourish. To truly allow independent scientists to develop, a strong set of progressive social policies, such as universal health care, basic income, and high-quality free education, are needed to strengthen the core of the ecosystem66,67. The ecosystem concept recognizes that the journey of a scientist through training is often an indirect path with many more career development influences than the direct path that a “pipeline” implies (Figure 1). Instead, an individual learns foundational knowledge, explores ideas, and gathers experience through a journey that is influenced by a broad range of interests, a balance of personal and professional goals, and adaptation to the challenges of life overall.\n\nSuch a student might attend the traditional classes expected in their field, explore other fields of interest (e.g., fine arts, social activism), and gather experience through interactions with others, work, internships, and volunteering, both within their field and outside of it. Along the way, they might explore other career (or life) choices, and perhaps return to academia completely, or explore specific scientific questions from a new perspective in another career choice outside of traditional academic institutions. Overall, the ecosystem model emphasizes that there is no right way to become a scientist. The diversity of experiences and perspectives are key to advancing STEM development in novel and more inclusive ways.\n\nMost importantly, the ecosystem model recognizes that every scientist is a person, meaning that people are more than their jobs and must balance a myriad of responsibilities, goals, and limitations that change as they move through life. While the conventional academic scientist pipeline assumes that individuals are functioning within a protected static environment (i.e., male scientists who serve as the breadwinner for their family in a generally unchanging environment), the ecosystem model encompasses the diverse and dynamic roles that individuals (both men and women) take on, particularly as they move through the household lifecycle (i.e., people’s needs and resources change throughout their life and most notably when they raise a family68). Indeed, sustainable livelihood strategies65 further emphasize this point by recognizing that people must be constantly making decisions to most efficiently use their resources (human, natural, financial, physical, and social capital) to meet their livelihood needs, and such decisions are often made within the context of the changing biophysical and socio-political conditions of the system that they live in.\n\nFurthermore, people must balance both non-monetary activities (i.e., child-rearing, house-keeping) with income earning activities, and both are equally important to almost all individuals. There is a lack of recognition of the importance of non-monetary activities in making livelihood decisions within conventional career models, as well as limited supporting economic and political structures to support these activities69. The pipeline model simply doesn’t consider both career and life balance in such a dynamic environment, nor does it present any opportunities to leverage the social diversity that occurs through such processes. By contrast, the ecosystem model not only presents a flexible model that encompasses the dynamism of the system, it also thrives on social, economic, and experiential diversity.\n\nIncome-generating activities in the ecosystem approach can be diverse different, contrasting with the expectation of there being a sole niche, such as tenure-track employment in university settings. Some of these new niches include part-time, or “fractional scholarship”57, through virtual institutes such as the Ronin Institute70 (see Box 2), small business or consulting companies (either partly or wholly focused on scientific research), and non-traditional start-ups. For example, some independent scholars run consultancies involving their scientific expertise in a commercial setting, but reserve time to pursue their own research; their research and consulting activities help inform the other, resulting in more grounded research and science-informed solutions, respectively. At a larger scale, some independent scientists have obtained venture capital funding to pursue biomedical research71,72, such as Perlara in San Francisco, which operates as a public benefit “B-corporation”.\n\nOrganizations of independent scholars, such as the non-profit Ronin Institute for Independent Scholarship (of which the authors are all members), the National Coalition of Independent Scholars, the Institute for Globally Distributed Open Research and Education, CORES Science and Engineering, Neurolinx Research Institute and research consortia such as the Complex Biological Systems Alliance2 enable highly trained individuals to contribute to basic science outside the traditional academic setting. Independent labs focusing on more specific research questions or subject areas have also emerged, such as the Orthogonal Lab. The Ronin Institute, in particular, provides support and infrastructure for submitting grants to existing funding agencies, but crucially also provides a home for scientists who may have either a part-time or full-time job, but wish to pursue their scholarship on a part-time or “fractional” basis57. Many such scholars also retain joint or visiting status with traditional universities, demonstrating the porousness between institutes as part of the overall scientific ecosystem. Through seminars, virtual meet-ups, and in-person unconferences, the Ronin Institute provides an essential community for independent scholars to trade ideas and identify new collaborations, so that they are not operating within a vacuum. For example, PhD graduates who work in private companies, government agencies, or the non-profit sector do not have to trade their scientific career for a profession. In addition, independent scholars can side-step some of the bureaucracy of the university, while maintaining their scientific identity that can be lost while working in full-time industry jobs.\n\nAcademics often view the abandonment of the search for a permanent tenured university position as a signal that a person is “leaving science”, but we argue that this should not necessarily be a one-way valve. Thus, another step towards building the open ecosystem is to normalize the movement into and out of traditional university positions. The formal system for scientific training must value students, postdocs, or other researchers who leave and re-enter programs or jobs for their professional and diverse experiences, and the unique network of colleagues that they bring to programs or jobs. Such a change will reduce the fear that scientists have in diversifying their career experience. We expect that for some kinds of science (i.e., those that require the use of expensive equipment), employment in a traditional university-based setting may continue to be the most appropriate, but other types of science (e.g., theoretical and computational sciences) can easily be practiced outside traditional academic settings (Box 2). Normalizing these movements as one of many flows within the overall scientific ecosystem would be a big step in the right direction (Box 3) for both broadening and diversifying science, and creating new career opportunities for scientists.\n\nA tenure-track job is still the dominant yardstick of legitimacy for a scientist27, and such a lack of vision for scientific careers makes institutional and cultural change in science difficult. Benderly (2015) offers one example where non-tenure-track early career scientists have been dismissed in biomedicine77. Unfortunately not all in positions of power are good-faith participants in this conversation: beneficiaries of a system built on implicit assumptions of zero-sum competition for attention and prestige are unlikely to welcome change. However, many senior academics recognize the unsustainability of the current system78. There are many steps that such sympathetic senior academics can take to support the ecosystem view. Here are just some: (i) Use language more carefully: don’t refer to scientists who do not secure a traditional academic job as “leaving the field”79; (ii) More visibly reject the journal impact factor prestige system and embrace preprints and other forms of open research80,81; and (iii) Include collaborators beyond those who are university-based, when possible. These small shifts will add up, especially if they originate from well-respected senior academics.\n\nThe increasingly all-consuming competitive nature of academic life often discourages speculation, innovation, and collaboration73–75. Little time and energy is left for the reflection needed to develop original ideas35. Interesting and creative science does not necessarily require this intense pace and may even be inhibited by it. The “slow” science movement encapsulates a more deliberative and conscious approach to science (Box 4). Smaller-scale research projects can have quite modest budgets, and crowdfunding sites such as Kickstarter and Experiment.com have supported many such interesting scientific projects76. Relative to standard grants, funds raised via crowdfunding might be considered tiny, but with the large overhead required by brick and mortar academic institutes (often over 50% of a grant), such funds can often go a long way in a research budget. Such monies have been fairly modest to date, but these approaches are still in their infancy and have much more potential for growth. The scientific ecosystem concept explicitly recognizes the value of, and enables, such scientific work at multiple niches and scales.\n\nToday, many grant dollars are awarded to a disproportionately small number of established investigators11, due in part to an increasing emphasis on high-budget, complex science. The rise of “Big Institute” science is channelling both philanthropic and federal dollars towards specific industries (e.g., biomedical research, specific diseases) and biotech hubs82, leading to fewer large labs working in a limited number of lucrative research areas. Traditional labs in university settings are incentivized to grow in size, partly because the overhead earned through grants helps fund the institution as a whole42. “High profile” journals also tend to favor work done at larger, and therefore more expensive, scales. These two factors bend institutional incentives towards funding expensive research questions and hiring faculty with publications in prestigious journals who do such work.\n\nBy contrast, the ecosystem model explicitly acknowledges the value of “slower” approaches to scholarship and science, and recognizes that scientific progress is enabled by scientists pursuing questions at all different scales and pace83–85. While expensive, big science certainly has an important place, its prioritization over smaller-scale investigator-driven science tacitly and severely constrains the scope and types of questions that can be asked82,86,87. The slow science movement emphasizes a more deliberative, less publish-or-perish pace to provide time for scientists to build trust, create more effective and durable collaborations, and invest in more collective approaches for doing science. The presence of the collaborative and reflective personalities required to address complex problems with team-based solutions88–91 tend to be weeded out in the current system. Moreover, even in traditional academic positions, the conventional narrative of faculty publishing productivity—a peak of publications in early career with a gradual decline—doesn’t fit the actual publication records of many scientists. In computer science, for example, it only describes about one-fifth of tenure-track careers92, even though this conventional narrative has a great influence on hiring and tenure decisions.\n\nResearchers in traditional settings operate in a highly hierarchical system, where the usual benchmarks of “success” are controlled by a relatively small number of people at each level. A large number of apprentices are under the control of a small number of masters93; this structure can make it difficult for new ideas to gain a toehold94,95 and can also lead to the exploitation of apprentices96,97. While senior investigators occupy important roles in identifying promising research avenues, providing synoptic views and institutional memories of their field, and many are tireless in promoting their trainees and helping them succeed, this success generally assumes continuing within the “pipeline” model. In contrast, we believe that moving towards less hierarchical peer-to-peer ecosystem models, which emphasize more democratic decision-making, cooperativity, and solidarity, will lead to a more dynamic and creative scientific enterprise overall (Box 5).\n\nThe model of academic science that early career investigators are taught to aspire towards (a “principal investigator” directing a large number of apprentices) is essentially feudal in structure; it is a historical product of the structure of academic institutions. This predominant structure dates roughly from Vannevar Bush’s famous 1945 memo, “Science: the endless frontier”104, and is not intrinsic to the discovery process itself. The cultural movement of increasing equity and access, which is driven by the ubiquitous presence of the Internet, the rise of peer-to-peer, commons-based production62,105,106, and crowdsourcing107, challenges this feudal/industrial model of the production of cultural goods105,108, including science. This looser, self-organized approach to intellectual production is demonstrated by the free, libre, or open-source software (FLOSS) movement (which has produced operating systems like GNU/Linux), as well as Wikipedia109.\n\nContributors to FLOSS projects also occupy different niches: some developers of open-source software are employed by companies (e.g. Red Hat), others contribute only one or two lines of code as volunteers, and most contribute at levels in between. Some scientific projects are run in similar ways110, and there is natural fit with idea of fractional scholarship introduced in our second pattern on the multiplicity of niches (subheading 2): important scientific contributions can be made at whatever time and energy level that an individual can provide. Open-source projects, and peer-to-peer production more generally, are not without structure62. Projects still develop leadership and lines of authority, and the need for mentorship does not disappear111, but leadership is fluid, often chosen by consensus, and based more on time, energy, and enthusiasm rather than an academic title or pedigree112. Consistent with our ecosystem approach, we imagine a multiplicity of solutions whereby the traditional PI structure is but one model within a larger spectrum of more peer-to-peer structures.\n\nThe burgeoning open science movement15,98 is a key enabler of the scientific ecosystem by bringing data, models, and resources out from behind institutional walls. Initiatives, such as the Center for Open Science99 and Sage Bionetworks100, have developed pioneering tools to enable large-scale sharing, and thus mining of open biological data by any interested parties without the need for the original infrastructure that generated the data. Access to shared resources, such as lab space, are also part of this shift (Box 6). In addition, the citizen science101 and indigenous research102 movements reach out to bright and creative individuals outside the academic system who are eager to contribute skill and time to advance science. The growth of open-access publishing will especially benefit the peer-to-peer model proposed here (Box 6).\n\nThe DIYbiology movement has championed an extremely low-tech, low-cost approach to experimental molecular biology. The emergence of community labs and commons-based co-working spaces (e.g., manylabs.org) are making work possible outside university settings113,114. Facilities for genome sequencing115 and “rent-by-the-bench” lab space, such as QB3@953 in San Francisco, CA and LabCentral in Cambridge, MA116, can enable lower-cost, lab-based research that might be completely off the radar from more traditional academic labs or large biotech companies. Mathematical and computationally based research is also now well within the reach of many independent scientists. Cloud-based computing servers offered via Google Cloud and Amazon Web Services117,118 can be done at a fraction of the cost of running a large university-based high-performance computing (HPC) cluster.\n\nOpen-access is another strand in an ecosystem model of science. Notably, open-access information now allows many “researched” communities to finally have access to information about their own communities, and take action from a grassroots position (e.g., members of developing countries are now more empowered to take part in science). However, the “author-pays Article Processing Charges” model of many open-access journals will require some rethinking in the absence of institutional support. The incredible uptake of preprints, especially by younger scientists, as a way around the artificial scarcity of the journal prestige system is an encouraging first step119. The growth of low-cost non-commercial models of publishing based on “platform co-operativism” principles120,121 that are owned and run by scientists are likely to be more equitable and sustainable in the long-run for scholars than venture-capital backed experiments122 www.scholarlyhub.org;). Open science and open access platforms should be focused on the goal of improving communication, scholarship, and learning, rather than being simply a way to extract commercial value from scientist’s labor123,124. New conferences like OpenCon and FORCE11 are leading the conversation in this area.\n\nSelf-funding of research can sometimes be sufficient, since many studies, especially in computational, mathematical, and social science fields cost little beyond the time required. Free of arbitrary institutional expectations of “bringing in grant money”, this can be quite liberating. However, many other kinds of research, especially wet lab biology, can be expensive and labor-intensive. The benefit of joining institutes like the Ronin Institute, allows scholars to apply directly for traditional federal grants (e.g., NIH or NSF) with reduced institutional overhead, leading to more efficient use of money for research. However, existing funding agencies, especially federal, are largely geared to favour the already well-funded or those who are working on whatever scientific questions they may be prioritizing that year. More balance is needed to ensure that scientific questions that are valued by society are also represented. Funding solutions outside of the traditional federal agencies and more distributive approaches, whether federally funded or otherwise, must be considered to help fill this gap (Box 7).\n\nVarious federal agencies have mechanisms to compensate for the bias towards large institutions and senior, male scientists. For example, the R15 grant mechanism of the National Institutes of Health is restricted towards institutions that receive less than an overall amount per year. The National Science Foundation has grants specifically for postdocs and early career faculty, and they may withdraw or extend a deadline if gender is not well-represented among the applicants. These programs help to correct bias, but still assume participation in the pipeline model. In addition, the massive amount of bureaucracy involved in the submission and reporting of such grants is a barrier to applicants with minimal grant administration support. Attempts to be more explicitly redistributive, such as a proposed cap of three concurrent R01 grants, have been met with fierce resistance125.\n\nFundamental rethinking of funding to address the current concentration of resources is needed. Here, we present two ideas. First, we propose the “fail fast” model, which comes from the tech start-up world. In this model, many smaller projects receive 6–12 months of funding to pursue an idea. If the idea works, the next phase receives funding. If the idea does not work, the researchers move on to the next idea. The “fail fast” model would support many smaller groups for a shorter period of time. It also addresses the reluctance of many funding agencies to fund riskier research as well as the need to have nearly all of the proposed work complete before writing a grant to get the funds to do said work. This model is unlikely to support students, however, as they often need 3–4 years of stable funding.\n\nSecond, we support experimenting with “collective allocation” models, where each qualified scientist is given a fixed basic grant (somewhat analogous to a “basic income”) and also receives additional funds from other scientists who think they would make good use of the grant money126. Each such scientist would also be required to “pay forward” a fraction of the previous years’ grant money to other scientists in the same manner, thus increasing the overall flow of funds through the scientific community. One benefit is that it reduces the time-consuming and costly bureaucratic infrastructure of the grant review system, while still maintaining the positive influence of peer-review. In addition, there could be rules to minimize “gaming” the system by preventing paying it forward to immediate co-authors127. Related ideas include allocating a portion of funding via lottery128. The new head of Science Europe has expressed interest in trialing these some of these new mechanisms129, indicating a growing frustration with the existing grant process and a willingness to experiment with more radical models.\n\nWhile we have emphasized existing grassroots movements and trends, this is not because we do not need reform of our institutional practice of science—we certainly do. However, we believe that social and structural changes are often initiated outside institutions, and these efforts can catalyze internal reforms. That said, many of the design patterns have clear institutional analogs. The aforementioned peer-to-peer approaches discussed under subheading 5 can be implemented within institutions as well, particularly at the level of individual labs, by giving postdocs greater autonomy outside of individual projects or grants103. Some institutions, such as the Santa Fe Institute, have postdocs that are tied directly to the institution, rather than an individual professor. Furthermore, institutions should recognize that researchers work in different styles and at a diversity of scales, as discussed under subheadings 2 and 4, respectively, and avoid monolithic ranking and assessing of scholars by external metrics.\n\nFinally, the ecosystem approach of providing multiple on-ramps and off-ramps to academia (subheading 3) together with distributing resources more broadly (subheading 7), point to a reform of the tenure system itself. The tenure system could be made more fluid, and the benefits of tenure (stability, security) extended to more people, rather than a lucky few. A progressive commons-based economy would be one way to realize this vision (Box 8). Organizations such as Future of Research, which are leading the way in engaging institutions to encourage open-science practices (subheading 6) and shape a more equitable job and funding landscape (subheading 7), should also be vigorously supported58.\n\nAs we noted under subheading 1, basic stability and security is a necessary condition for people to flourish. One of the continuing appeals of university-based, tenure-track positions, especially in the United States, is the job security and health-care benefits promised with tenure. Unfortunately in many industries, including academia, job security is going by the wayside; thus, many new progressive economic models are switching from a focus on job security to a focus on income security, and this takes the form of such initiatives as universal health-care and universal basic income66,67. With this kind of security in place, the developing flexible “gig economy” in science132 is more likely to fully enable scientists to pursue their work in whichever part of the ecosystem that fits them best, rather than in predefined and allocated roles in the pipeline. The dark side of exploitation and insecurity in much of the existing mainstream cultural and gig economy133–135 is very real, so we advocate for a truly progressive economic system that protects and creates security for all. The benefits of the new flexible peer-to-peer approaches to work must be widely distributed and not concentrated into a small number of hands62,136.\n\n\nWriting a new cultural narrative\n\nOur paper proposes a fundamental change in the way that scientific training and its professional landscape are viewed. In the spirit of our ecosystem concept, we hope to spark discussion, debate, and action rather than offer a complete “turn-key” solution to the current state of careers in science (if such a thing were ever possible). Increased participation and mobilization of scientists and their partners towards such a concept will ultimately determine how the concept evolves in a bottom-up manner. We particularly hope to reach early-career scientists who are making career decisions and developing professional identities, since they will be most actively involved in co-creating this new ecosystem. We envision that scientists at all levels that are seeking to address the challenges faced by scientific institutions may also find inspiration in this paper to help them catalyze change from the inside out.\n\nUp until this point we have focused on the economic, social, logistic, and professional challenges of a new way of conceiving scientific work and careers. These are critical and necessary steps towards creating a more equitable scientific ecosystem. However, even if we demonstrate economically viable and sustainable ways for more scientists to contribute high quality research outside of traditional institutions, we must still overcome a potentially larger hurdle: the dominant cultural narrative of scientific success based on scarcity, especially of certain kinds of jobs (Box 3). The set of strategies we propose, while highly disparate and pluralistic in approach, are all characterized by a view that success does not need to be a zero-sum game.\n\nThe language we use to frame our new narrative is therefore key to cultural change. As Lakoff and Johnson130 point out, metaphors frame our thinking, as well as shape and constrain the cultural narratives of success. The ecosystem metaphor is our attempt to reframe and expand the discussion of STEM careers and science, beyond what has often become a sterile and arid debate about competition and scarcity within academia, by introducing the concepts of open flows, resilience, diversity, and feed-back loops of ecological systems. To this end, in Table 1, we provide a lexicon that highlights the non-zero-sum and egalitarian aspects of our ecosystem model, illustrating the contrasting language between the pipeline and the ecosystem. We do not assume that the pursuit of prestige or financial rewards will wither away; convivial competition clearly has a proper place in science131 and will always co-exist with cooperation (existing somewhere on the spectrum between the two columns in Table 1). But we strongly believe that making science better is not just about “creating better incentives”, but a collective cultural shift beyond viewing competition and individualistic success as the sole defining feature of science (i.e., the pipeline model).\n\nThe existing socialization process of traditional academic science vastly over values signals of academic capital (title, rank, and institutional affiliation) and economic metrics of productivity (amount of grant funding, high-impact papers, and h-index); scholars who do not meet these criteria are often disregarded, limiting diversity in science. The most important part of the shift toward an ecosystem model is cultural and psychological: the essential spirit of science must be re-captured by emphasizing that there are no gatekeepers to the scientific development of knowledge. Academia doesn't “own” science, any more than museums, art schools, or galleries own art. Just as with visual or performing arts or music, the open-ended exploration of scientific ideas is something worth celebrating in itself, regardless of the nature or scale of the question.\n\nChanging the cultures of research careers and the scientific enterprise is an experiment itself: actively practicing new a scientific culture can encourage others to be even bolder in their experimentation. The existing institutions that are tasked with supporting basic curiosity-driven inquiry need to be reformed and strengthened, but that alone is insufficient. We must build new structures that are informed by an ecosystem view from conception. The beauty is that science can be made available to everyone and our technologies are making it increasingly so. It is not a scarce resource: we should build our new ecosystem to recognize this truth.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Competing interests\n\n\n\nAlexander Lancaster is an owner of Amber Biology; Anne E. Thessen is the owner of The Data Detektiv; Arika Virapongse is the owner of Middle Path EcoSolutions.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThanks to M. J. Hickerson, J. Wilmer and L. Wishart for early discussions and to R. Fagen, J. Garb, J. Paulas, and J. F. Wilkins, for valuable comments and feedback. The University of Florida and the University of Sydney generously provided access to library resources.\n\n\nReferences\n\nBourne HR, Lively MO: Iceberg alert for NIH. Science. 2012; 337(6093): 390, [cited 2015 May 18]. PubMed Abstract | Publisher Full Text\n\nBalch C, Arias-Pulido H, Banerjee S, et al.: Science and technology consortia in U.S. biomedical research: a paradigm shift in response to unsustainable academic growth. BioEssays. 2015; 37(2): 119–22, [cited 2014 Nov 12]. PubMed Abstract | Publisher Full Text\n\nArchibald RB, Feldman DH: State higher education spending and the tax revolt. J High Educ. 2006; 77(4): 618–644. Publisher Full Text\n\nBarr A, Turner SE: Expanding enrollments and contracting state budgets: The effect of the Great Recession on higher education. Ann Am Acad Pol Soc Sci. 2013; 650(1): 168–193. Publisher Full Text\n\nBordelon D: How Do We Stop It? Strategies for Pushing Back Corporate U. THOUGHT ACTION. 2014; 91, [cited 2015 Apr 21]. Reference Source\n\nDunn S: An ‘Alarming Snapshot’ of Adjunct Labor. Chronicle Vitae. 2014; [cited 2015 Aug 6]. Reference Source\n\nRohn J: Give postdocs a career, not empty promises. Nature. 2011; 471(7336): 7, [cited 2013 Mar 25]. PubMed Abstract | Publisher Full Text\n\nBourne H: The postdoc holding tank (TRR-VI). BiomedWatch. 2012; [cited 2015 May 7]. Reference Source\n\nBourne HR: A fair deal for PhD students and postdocs. eLife. 2013; 2: e01139, [cited 2015 May 18]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPowell K: The future of the postdoc. Nature. 2015; 520(7546): 144–7, [cited 2015 Apr 24]. PubMed Abstract | Publisher Full Text\n\nLarivière V, Macaluso B, Archambault É, et al.: Which scientific elites? On the concentration of research funds, publications and citations. Res Eval. 2010; 19(1): 45–53, [cited 2015 Jun 3]. Publisher Full Text\n\nParker JN, Lortie C, Allesina S: Characterizing a scientific elite: the social characteristics of the most highly cited scientists in environmental science and ecology. Scientometrics. 2010; 85(1): 129–43, [cited 2015 Jul 6]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarjak F, Robinson S: International collaboration, mobility and team diversity in the life sciences: impact on research performance. Soc Geogr. 2008; 3(1): 23–36. Publisher Full Text\n\nFleming L: Breakthroughs and the \"long tail\" of innovation. MIT Sloan Manag Rev. 2007; 49(1): 69–74. Reference Source\n\nNielsen M: Reinventing discovery: the new era of networked science. Princeton University Press; 2012; [cited 2015 Apr 24]. Reference Source\n\nSheltzer JM, Smith JC: Elite male faculty in the life sciences employ fewer women. Proc Natl Acad Sci U S A. 2014; 111(28): 10107–12, [cited 2018 Apr 23]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNCSES: Women, Minorities, and Persons with Disabilities in Science and Engineering. National Science Foundation; Report No.: nsf17310. 2017; [cited 2018 May 22]. Reference Source\n\nWeber L: The end of employees. The Wall Street Journal. 2017. Reference Source\n\nGregg M: The Deficiencies of Tech’s “Pipeline” Metaphor. The Atlantic. 2015; [cited 2017 Oct 3]. Reference Source\n\nGarbee E: The Problem With the “Pipeline.”. Slate. 2017; [cited 2018 Apr 13]. Reference Source\n\nXue Y, Larson RC: STEM crisis or STEM surplus? Yes and yes. Mon Labor Rev. 2015; [cited 2018 May 22]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcKenna L: Should Professors Be Fired for Damaging a College’s Reputation? The Atlantic. 2016; [cited 2016 Mar 2]. Reference Source\n\nRobinson JA: The Academic Job Market Is Tottering, But Nobody’s Telling Graduate Students. The James G. Martin Center for Academic Renewal. 2016; [cited 2018 May 22]. Reference Source\n\nHandelsman J, Smith M: STEM for All. Obama White House. 2016; [cited 2018 May 22]. Reference Source\n\nMany junior scientists need to take a hard look at their job prospects. Nature. 2017; 550(7677): 429, [cited 2017 Nov 3]. PubMed Abstract | Publisher Full Text\n\nJaschik S: Tenure Track as Alt-Ac. Inside HigherEd. 2013, [cited 2015 Apr 24]. Reference Source\n\nLewontin RC: Epilogue: Legitimation Is the Name of the Game. In Rebels, Mavericks, and Heretics In Biology. 2008; 372. Reference Source\n\nCallier V, Vanderford NL: Ailing academia needs culture change. Science. 2014; 345(6199): 885, [cited 2014 Aug 25]. PubMed Abstract | Publisher Full Text\n\nSchaller MD, McDowell G, Porter A, et al.: What's in a name? eLife. 2017; 6: pii: e32437, [cited 2017 Nov 3]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeo S, Wei Y, Daunert S: Probing a myth: does the younger generation of scientists have it easier? Anal Bioanal Chem. 2012; 403(8): 2065–7, [cited 2015 Aug 7]. PubMed Abstract | Publisher Full Text\n\nEtchells P: The human cost of the pressures of postdoctoral research. The Guardian. 2017; [cited 2018 May 22]. Reference Source\n\nCowen T: The Tenure Track Is Too Rigid to Help Diversity. Bloomberg.com. 2017; [cited 2017 Nov 3]. Reference Source\n\nMoore S, Neylon C, Eve MP, et al.: “Excellence R Us”: university research and the fetishisation of excellence. Palgrave Commun. 2017; 3: 16105, [cited 2017 Feb 21]. Publisher Full Text\n\nCasadevall A, Fang FC: Reforming science: methodological and cultural reforms. Infect Immun. 2012; 80(3): 891–6, [cited 2018 May 4]. PubMed Abstract | Publisher Full Text http://www.ncbi.nlm.nih.gov/pmc/articles/3294645\n\nAlberts B, Kirschner MW, Tilghman S, et al.: Rescuing US biomedical research from its systemic flaws. Proc Natl Acad Sci U S A. 2014; 111(16): 5773–7, [cited 2014 Apr 16]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGraeber D: Of Flying Cars and the Declining Rate of Profit. The Baffler. 2012; (19): [cited 2015 Apr 24]. Reference Source\n\nSouthwick F: Opinion: Academia Suppresses Creativity. The Scientist. 2012; [cited 2017 Nov 3]. Reference Source\n\nNess R: The Creativity Crisis: Reinventing Science to Unleash Possibility. Oxford University Press; 2014. Reference Source\n\nSmaldino PE, McElreath R: The natural selection of bad science. R Soc Open Sci. 2016; 3(9): 160384, [cited 2016 Jul 28]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNissen SB, Magidson T, Gross K, et al.: Publication bias and the canonization of false facts. eLife. 2016; 5: pii: e21451, [cited 2017 Mar 28]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Dijk D, Manor O, Carey LB: Publication metrics and success on the academic job market. Curr Biol. 2014; 24(11): R516–7, [cited 2017 Oct 25]. PubMed Abstract | Publisher Full Text\n\nMartinson BC: Universities and the money fix. Nature. 2007; 449(7159): 141–2, [cited 2015 Apr 24]. PubMed Abstract | Publisher Full Text\n\nYoder J: I Found a Tenure-Track Job. Here’s What it Took. ChronicleVitae. 2017; [cited 2017 Nov 3]. Reference Source\n\nSchuman R: A Kind of Grad-Student Gaslighting. Slate. 2015; [cited 2015 Aug 27]. Reference Source\n\nKendzior S: The Paradoxical Success of The Professor Is In. Chronicle Vitae. 2015; [cited 2015 Aug 27]. Reference Source\n\nClauset A, Arbesman S, Larremore DB: Systematic inequality and hierarchy in faculty hiring networks. Sci Adv. 2015; 1(1): e1400005. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMervis J: “Employment crisis” for new Ph.D.s is an illusion. Science. 2016; 352(6288): 880. PubMed Abstract | Publisher Full Text\n\nYoung researchers thrive in life after academia. Nat News. 2016; 537(7622): 585, [cited 2017 Nov 3]. PubMed Abstract | Publisher Full Text\n\nShapard C: Growing the bioscience career pipeline. Nat Biotechnol. 2008; 26(9): 1047–8, [cited 2017 Nov 3]. PubMed Abstract | Publisher Full Text\n\nNational Science Foundation: Revisiting the STEM workforce NSF - National Science Foundation. National Science Foundation; 2015; [cited 2017 Nov 3]. Reference Source\n\nLazebnik Y: Are scientists a workforce? - Or, how Dr. Frankenstein made biomedical research sick: A proposed plan to rescue US biomedical research from its current 'malaise' will not be effective as it misdiagnoses the root cause of the disease. EMBO Rep. 2015; 16(12): 1592–600, [cited 2018 May 4]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards MA, Roy S: Academic Research in the 21st Century: Maintaining Scientific Integrity in a Climate of Perverse Incentives and Hypercompetition. Environ Eng Sci. 2017; 34(1): 51–61, [cited 2017 Dec 5]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiamandis EP: Getting noticed is half the battle. Science. 2015; 349(6244): 206, [cited 2015 Jul 23]. PubMed Abstract | Publisher Full Text\n\nFeltman R: Hundreds of scientists ask Science to stop publishing a smorgasbord of stereotypes. Washington Post. 2015; [cited 2015 Jul 23]. Reference Source\n\nGaensler B: Workaholism isn’t a valid requirement for advancing in science. The Conversation. 2015; [cited 2015 Jul 23]. Reference Source\n\nVasbinder JW: What if there were no universities? PsyCh J. 2017; 6(4): 316–325, [cited 2017 Dec 21]. PubMed Abstract | Publisher Full Text\n\nArbesman S, Wilkins J: The Rise of Fractional Scholarship. Rochester, NY: Social Science Research Network; 2012; [cited 2015 Feb 9]. Report No.: ID 2048780. Reference Source\n\nMcDowell GS, Gunsalus KT, MacKellar DC, et al.: Shaping the Future of Research: a perspective from junior scientists [version 2; referees: 2 approved]. F1000 Res. 2014; 3: 291, [cited 2014 Dec 5]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNicholson J: Why science needs more people to quit the lab. STAT. 2015; [cited 2016 Aug 23]. Reference Source\n\nHansen WD, Scholl JP, Sorensen AE, et al.: How do we ensure the future of our discipline is vibrant? Student reflections on careers and culture of ecology. Ecosphere. 2018; 9(2): e02099. Publisher Full Text\n\nOstrom E: Governing the commons: the evolution of institutions for collective action. Cambridge, UK: Cambridge University Press; 1990. Reference Source\n\nBauwens M: The political economy of peer production. CTheory. 2005. Reference Source\n\nBollier D, Helfrich S: The wealth of the commons: A world beyond market and state. Levellers Press; 2014. Reference Source\n\nMeadows DH: Thinking in Systems. White River Junction VT Chelsea Green Publ. 2008; 77–78. Reference Source\n\nDepartment for International Development: DFID sustainable livelihoods guidance sheets. United Kingdom: Department for International Development; 1999; [cited 2018 May 22]. Reference Source\n\nLehdonvirta V: Can universal basic income counter the ill-effects of the gig economy? The Conversation. 2017; [cited 2017 Nov 3]. Reference Source\n\nStanding G: Basic Income: A Guide for the Open-Minded. Yale University Press; 2017. Reference Source\n\nPerz SG: Household demographic factors as life cycle determinants of land use in the Amazon. Popul Res Policy Rev. 2001; 20(3): 159–186. Publisher Full Text\n\nOlivetti C, Petrongolo B: The Economic Consequences of Family Policies: Lessons from a Century of Legislation in High-Income Countries. J Econ Perspect. 2017; 31(1): 205–30. PubMed Abstract | Publisher Full Text\n\nO’Brien K: The Ronin Institute for wayward academics. The Boston Globe. 2012; [cited 2012 Sep 15]. Reference Source\n\nTansey B: The Experiment Begins: Y Combinator Admits First Biotech Startups. Xconomy. 2014; [cited 2015 Jun 3]. Reference Source\n\nTansey B: Y Combinator, Move Over For IndieBio: A Second Biotech Accelerator Xconomy. Xconomy. 2014; [cited 2015 Jun 3]. Reference Source\n\nHagstrom WO: Competition in science. Am Sociol Rev. 1974; 39(1): 1–18. Publisher Full Text\n\nCabota JB, Grimaldo F, Squazzoni F: When Competition Is Pushed Too Hard. An Agent-Based Model Of Strategic Behaviour Of Referees In Peer Review. In: 27th European Simulation and Modelling Conference. Lancaster University, Lancaster UK: ECMS; 2013; 881–7, [cited 2015 Jun 29]. Publisher Full Text\n\nFang FC, Casadevall A: Competitive Science: Is Competition Ruining Science? Infect Immun. 2015; 83(4): 1229–33, [cited 2015 Jun 9]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWheat RE, Wang Y, Byrnes JE, et al.: Raising money for scientific research through crowdfunding. Trends Ecol Evol. 2013; 28(2): 71–2, [cited 2015 Jun 3]. PubMed Abstract | Publisher Full Text\n\nBenderly BL: What well-dressed riffraff are wearing this season. Science. 2015; [cited 2015 Aug 27]. Publisher Full Text\n\nAitkenhead D: Peter Higgs: I wouldn’t be productive enough for today’s academic system. The Guardian. 2013; [cited 2018 May 22]. Reference Source\n\nPaulas J: To a Revolutionary Degree. Public Seminar. 2017; [cited 2018 Apr 13]. Reference Source\n\nEisen MB: The Mission Bay Manifesto on Science Publishing. it is NOT junk. 2015; [cited 2015 Oct 2]. Reference Source\n\nEisen M, Vosshall L: Coupling Pre-Prints and Post-Publication Peer Review for Fast, Cheap, Fair, and Effective Science Publishing. 2016; [cited 2016 Aug 24]. Reference Source\n\nKozubek J: The Trouble with Big Science. Los Angeles Review of Books. 2017; [cited 2017 Nov 30]. Reference Source\n\nAlleva L: Taking time to savour the rewards of slow science. Nature. 2006; 443(7109): 271, [cited 2017 Nov 10]. PubMed Abstract | Publisher Full Text\n\nMountz A, Bonds A, Mansfield B, et al.: For slow scholarship: A feminist politics of resistance through collective action in the neoliberal university. ACME Int J Crit Geogr. 2015; 14(4): 1235–1259. Reference Source\n\nStengers I: Another Science is Possible: A Manifesto for Slow Science. John Wiley & Sons; 2018. Reference Source\n\nDalrymple GB: The importance of “small” science. Eos Trans Am Geophys Union. 1991; 72(1): 1–4, [cited 2015 Jun 3]. Reference Source\n\nO’Hara B; GrrlScientist: New Breakthrough Prize in Life Sciences is misguided. The Guardian. 2013; [cited 2013 Feb 22]. Reference Source\n\nLawrence PA: The heart of research is sick. Lab Times. 2011; 2: 24–31, [cited 2015 Apr 6]. Reference Source\n\nHall KL, Stokols D, Stipelman BA, et al.: Assessing the value of team science: a study comparing center- and investigator-initiated grants. Am J Prev Med. 2012; 42(2): 157–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrant A, Dutton J: Beneficiary or benefactor: are people more prosocial when they reflect on receiving or giving? Psychol Sci. 2012; 23(9): 1033–9, [cited 2015 Aug 27]. PubMed Abstract | Publisher Full Text\n\nGrant A: Give and take: A revolutionary approach to success. Hachette UK; 2013. Reference Source\n\nWay SF, Morgan AC, Clauset A, et al.: The misleading narrative of the canonical faculty productivity trajectory. Proc Natl Acad Sci U S A. 2017; 114(44): E9216–E9223. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchultheiss K: Ghost Advising. The Chronicle of Higher Education. 2018; [cited 2018 Mar 12]. Reference Source\n\nBenderly BL: How scientific culture discourages new ideas. Science. 2016; [cited 2017 Nov 30]. Reference Source\n\nAzoulay P, Fons-Rosen C, Zivin JSG: Does Science Advance One Funeral at a Time? National Bureau of Economic Research; 2015; [cited 2017 Nov 30]. Report No.: 21788. Reference Source\n\nMartin B: Academic exploitation. In: Martin B, Baker CMA, Manwell C, Pugh C, editors. Intellectual Suppression: Australian Case Histories, Analysis and Responses. Sydney: Angus & Robertson; 1986; 59–62. Reference Source\n\nReay D: Cultural capitalists and academic habitus: Classed and gendered labour in UK higher education. In: Womens Stud Int Forum. Elsevier; 2004; 27(1): 31–39. Publisher Full Text\n\nFecher B, Friesike S: Open Science: One Term, Five Schools of Thought. In: Opening Science. Springer, Cham; 2014; 17–47, [cited 2018 Mar 14]. Publisher Full Text\n\nMiguel E, Camerer C, Casey K, et al.: Social science. Promoting transparency in social science research. Science. 2014; 343(6166): 30–1, [cited 2015 Apr 24]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaiser J: Profile: Stephen Friend. The visionary. Science. 2012; 335(6069): 651–3, [cited 2015 Jun 4]. PubMed Abstract | Publisher Full Text\n\nSilvertown J: A new dawn for citizen science. Trends Ecol Evol. 2009; 24(9): 467–471, [cited 2015 Jun 4]. PubMed Abstract | Publisher Full Text\n\nSmith LT: Decolonizing methodologies: Research and indigenous peoples. Zed Books Ltd.; 2013. Reference Source\n\nCallier V, Polka J: Fellowships are the future. Nature. 2015; 528(7580): 155–6, [cited 2017 Dec 5]. Publisher Full Text\n\nBush V: Science: the endless frontier: a report to the president on a program for postwar scientific research. Office of Scientific Research & Development; 1945; [cited 2016 Nov 22]. Reference Source\n\nBenkler Y: The wealth of networks: How social production transforms markets and freedom. Yale University Press; 2006. Reference Source\n\nWilbanks J, Boyle J: Introduction to Science Commons. 2006; 2010, [cited 2015 Jun 4]. Reference Source\n\nShirky C: Here comes everybody: The power of organizing without organizations. Penguin; 2008. Reference Source\n\nHeimans J, Timms H: Understanding “New Power”. Harv Bus Rev. 2014; [cited 2015 Jul 23]. Reference Source\n\nReagle JM Jr: Good faith collaboration: The culture of Wikipedia. MIT Press; 2010. Reference Source\n\nWillinsky J: The unacknowledged convergence of open source, open access, and open science. First Monday. 2005; 10(8), [cited 2015 Jun 4]. Publisher Full Text\n\nCanfora G, Di Penta M, Oliveto R, et al.: Who is going to mentor newcomers in open source projects? In: Proceedings of the ACM SIGSOFT 20th International Symposium on the Foundations of Software Engineering. ACM; 2012; 44. Publisher Full Text\n\nWeber S: The success of open source. Harvard University Press; 2004. Reference Source\n\nWohlsen M: Biopunk: Solving Biotech’s Biggest Problems in Kitchens and Garages. Reprint edition. New York: Current; 2012; 272. Reference Source\n\nDelfanti A: Biohackers: the politics of open science. Pluto Press; 2014; [cited 2015 Apr 24]. Publisher Full Text\n\nPichler FB, Turner SJ: The power and pitfalls of outsourcing. Nat Biotechnol. 2007; 25(10): 1093–6, [cited 2015 Jun 4]. PubMed Abstract | Publisher Full Text\n\nKirsner S: LabCentral, a new hatchery for science-oriented startups, is seeking space in Kendall Square. Boston.com. 2012; [cited 2015 Jun 4]. Reference Source\n\nFusaro VA, Patil P, Gafni E, et al.: Biomedical cloud computing with Amazon Web Services. PLoS Comput Biol. 2011; 7(8): e1002147, [cited 2014 May 18]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSouilmi Y, Lancaster AK, Jung JY, et al.: Scalable and cost-effective NGS genotyping in the cloud. BMC Med Genomics. 2015; 8(1): 64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCallaway E, Powell K: Biologists urged to hug a preprint. Nature. 2016; 530(7590): 265, [cited 2017 Nov 3]. PubMed Abstract | Publisher Full Text\n\nScholz T: Platform cooperativism. Challenging the corporate sharing economy. 2016. Reference Source\n\nScholz T, Schneider N: Ours to hack and to own: The rise of platform cooperativism, a new vision for the future of work and a fairer Internet. OR books; 2016. Reference Source\n\nWillinsky J, Rusk M: If funders and libraries subscribed to open access: The case of eLife, PLOS, and BioOne. PeerJ Preprints. 2017; 5: e3392v1, [cited 2017 Nov 10]. Publisher Full Text\n\nKansa E: It’s the Neoliberalism, Stupid: Why instrumentalist arguments for Open Access, Open Data, and Open Science are not enough. Impact of Social Sciences. 2014; [cited 2016 Mar 10]. Reference Source\n\nLancaster AK: How do we build a human-centered open science? The Winnower. 2016; 5: e147624.46708, [cited 2017 Nov 14]. Publisher Full Text\n\nReardon S: NIH scraps plans for cap on research grants. Nat News. 2017; [cited 2017 Nov 3]. Publisher Full Text\n\nBollen J, Crandall D, Junk D, et al.: From funding agencies to scientific agency: Collective allocation of science funding as an alternative to peer review. EMBO Rep. 2014; 15(2): 131–3, [cited 2015 Jun 4]. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Vrieze J: With this new system, scientists never have to write a grant application again. Science. 2017; [cited 2018 Apr 13]. Publisher Full Text\n\nGross K, Bergstrom CT: Contest models highlight inefficiencies of scientific funding. ArXiv180403732 Phys Stat. 2018; [cited 2018 Apr 13]. Reference Source\n\nMatthews D: Radical ideas required to cut research grant waste, funders told. Times Higher Education (THE). 2018; [cited 2018 Mar 14]. Reference Source\n\nLakoff G, Johnson M: Metaphors We Live By. University of Chicago Press; 1980. Reference Source\n\nHarvie D: Commons and communities in the university: Some notes and some examples. The Commoner. 2004; 8: 1–10, [cited 2015 Apr 24]. Reference Source\n\nKwok R: Flexible working: Science in the gig economy. Nature. 2017; 550(7676): 419–21, [cited 2017 Nov 3]. Publisher Full Text\n\nTaylor A: The People’s Platform: Taking Back Power and Culture in the Digital Age. Random House Canada; 2014; [cited 2015 Jul 23]. Reference Source\n\nSchor J: Debating the sharing economy. Gt Transit Initiat. 2014; [cited 2015 Aug 3]. Reference Source\n\nSlee T: What’s yours is mine: Against the Sharing Economy. Or Books; 2016. Publisher Full Text\n\nIDE M: MIT Students Propose Policy Ideas for Gig- Economy Workers. Medium. 2018; [cited 2018 Apr 18]. Reference Source"
}
|
[
{
"id": "35295",
"date": "25 Jun 2018",
"name": "Jonathan P. Tennant",
"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\nHerein is my response to the manuscript entitled “A new paradigm for the scientific enterprise: nurturing the ecosystem”, by Alexander Lancaster and colleagues to F1000 Research.\nThe authors present a criticism of the concept of the ‘academic pipeline’, and instead introduce a new ecosystem model that includes more dimensions and scope for understanding the complexity of academic culture. My relevant ‘expertise’ in reviewing this manuscript comes from being a researcher interested in developments in scholarly communication. As this is an opinion piece, the only real substance I can offer is that of my own opinion, as well as attempt to help improve the clarity and argumentation style of the article. I hope the authors find the comments useful and know that this was an enjoyable and provocative read for me.\nFigures Figure 1. The arrows in the green academia arrow seem to be misaligned. Yes, I am making this comment. Do you by any chance have statistics that could help to quantify this flow diagram? I know the Royal Society have a very similar one for the UK, but would be cool (but not obligatory) to get some numbers on this. Great figure though, and very easy to interpret. Figure 2. Given that this is an open peer review, it should perhaps come as no surprise that I think Open Access should be changed to Open Science. So I found this a little annoying flicking between the caption and the image to interpret it. Is it possible to put a very simply coloured legend on the side as well to help maintain focus?\nAbstract The abstract is concise, interesting and impactful, and onveys the context and main findings of the article.\nIntroduction The first paragraph of the Introduction is basically identical to the Abstract, which seems a bit odd to me. Can one be rephrased, or is there a different way of introducing the issue? Second paragraph, can early career be defined. And does every country use dollars? Here, I think it would be worth mentioning that it is not just a funding issue, but that this has been exacerbated by the incentive/reward system, which has become virtually divorced from the issues raised here (i.e., anything vaguely scientific). The final sentence of this paragraph needs a couple of citations to support too, or to be stitched to the beginning of the following paragraph. Third paragraph. Do you think there are phases between training and permanent positions? I’m thinking of short-term contract postdocs, for example.\nBox 1. Can it be mentioned here that, as a PhD student/ECR, you are often almost guilted into this state, as leaving for a job outside of academia is perceived as ‘failure’ by those within? As well as this, many bright young students come into this system, full of so much potential and energy, and then become disillusioned by it and leave as a result. Which basically creates a bias in which it is those who are conformist, the privileged, or have more support (or just a stronger desire) who end up staying. In the paragraph on families, I would make it clear that this present state is “actively discriminating” against those who cannot conform to that lifestyle. True, and impactful. The content of this box though is very well written and conveys many of the large issues in academia very well.\nParagraph 5. What is the fundamental cultural change?\nThe introduction is well-written, provides sufficient context and background for the piece, and finishes with a nice summary of the work within.\nFixing the pipeline First paragraph. Should make it clear here, I think, that by framing it as a pipeline it constrains the potential scope for solutions.\nAdjusting the flow. I think it is imperative to make the point here that adjustments at this end do not impact all demographics equally. For example, by raising costs to reduce student numbers, this discriminates against different groups based on their financial status. Here, could one or two specific examples of changes to the supply side be provided, and what their impact was. Adapting to the pipeline. Here, I think a comment is needed about the divergence between what it means to be a good academic in order to conform to the system and advance one’s career, and the fundamentals of research. For example, this has repercussions on how research is performed and communicated.\nFinding another pipeline. In the final sentence, do you think this goes wider than just science though, and is an outcome of an increasingly neoliberal academy?\nThinking outside the pipeline. So, once you’re ‘in the pipeline’, is it very difficult to see outside the pipeline, do you think? And do you think reform is so slow because we are waiting for it to come from the top down? “I made it here, so why can’t you. Nothing needs to change.”\nThe science ecosystem Personal preference, but I would love it if somewhere here the UN Sustainable Development Goals could be mentioned (https://sustainabledevelopment.un.org/). Do you feel that, within the rapidly developing Open Science ‘movement’, researchers are being taught the core skills and competencies they need to maintain their careers? See the EU report on this from 2017: https://publications.europa.eu/en/publication-detail/-/publication/3b4e1847-c9ca-11e7-8e69-01aa75ed71a1/language-en.\nFundamental development of the scientist. For me, journey implies pathway, which is kind of similar to a pipeline. I think the key here is that development is part of an immensely complex network, where every interaction, every thought, every event has some sort of untraceable impact. You could say, to use your own analogy, that each individual has their own ecological niche within a hierarchical network, and that niche is defined by an incredibly multi-dimensional series of developmental traits. I like this analogy. Does an increased diversity of developmental traits decrease the chance of extinction?\nMultiplicity of niches. A key here is the recognition that an individual is the product of their ecosystem. Therefore, it takes into account developmental history (point 1), as well as external factors over which they have no control. As well as genetic heritability, which I think in academia is manifested as privilege. Is there any data out there on the proportion of researchers who have secondary jobs or income-generating activities? Thinking things like teaching, writing, public talks etc.\nOpen ecosystem flows. As someone who is independently pursuing research beyond academia, I think that this section is critical. Academia is vastly prohibitive to numerous dimensions of social mobility, and allowing greater physical flow is also important for researchers to take advantage of the power of networked worlds. Research does not need always need a desk.\nBox 3: As well as what is already mentioned, how about support fairer research evaluation initiatives like DORA (https://sfdora.org/), as well as organisations like Future of Research (http://futureofresearch.org/). There could be whole papers written about this, so appreciate if the authors don’t want to list all possibilities here!\nDiversity of scales and budget. I would mention here the impact that this can have on research. Salami slicing for publications, claims of data ownership and generally selfish attitudes, data doctoring and manipulation, fraud, other questionable research practices – all driven by hyper-competition and the modes of research evaluation. Box 4 might be better for this though. Box 4. ““High profile” journals also tend to favor work done at larger, and therefore more expensive, scales.” – citation needed.\nPeer to peer networks. It’s a shame the way the present power dynamic works. Even if such a research apprentice should be fortunate enough to sit on a high-level council, they might not be granted the rank of master. This can have disastrous consequences for the Galaxy. Box 5. I like this idea of sort of decentralised power structures within academia. Are there any examples you can think of where Wiki- or FLOSS-like communities exist within research specifically? A recent study by Hartgerink and van Zelst1 on peer to peer networks might be interesting to mention here too. What impact do these approaches have on inclusivity, as well as prestige or reputation for belonging to such a community?\nFluid communication networks and shared access to resources. So, this is about exposing just the outputs of research, then? What about wider engagement in science, especially from citizens? And more transparency into the processes, to help enable verification, trust, and collaboration in research? Note that this aspect also includes factors such as open evaluation, and therefore incentive structures within academia.\nBox 6. Should probably mention something about the principles of the scholarly commons (https://zenodo.org/record/569952#.Wy0XHIjRBPY, and other links). I would perhaps mention some of the causes for this open science movement, such as negative feelings towards the publishing industry, non-scientific evaluation criteria, reproducibility issues, a demand for greater transparency and public accountability etc. If you really wanted, you could discuss some of the wider impacts of Open Access too2. Might also be worth mentioning that commercial publishers are rapidly moving into this area too, such as the recent launch of the EU Open Access Monitor, which was sub-contracted to Elsevier, one of the scholarly publishers with the worst reputation (which I’m sure I don’t have to tell you; https://ec.europa.eu/info/research-and-innovation/strategy/goals-research-and-innovation-policy/open-science/open-science-monitor/about-open-science-monitor_en). Broader distribution of resources. How many researchers does Ronin currently support? And how do we work out which questions are more valued by society?\nInstitutional change. How about institutes recognizing different forms of scholarship, beyond just research papers? So, training, software, code, tech support, educational resources, among other things. I think that it’s important to mention the link between institutional change and incentives.\nWriting a new cultural narrative This sums up the content of the work nicely, and makes it clear what the intention of it was. I wonder, just to strengthen it, could you perhaps provide just several simple, concrete steps that the different stakeholder groups can take towards reaching such a new culture? Otherwise, I find this section very inspirational!\nCongratulations to the authors on a great piece of work. I think that this could be very impactful and spark substantial new discussion and action in this area.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "35293",
"date": "21 Sep 2018",
"name": "Gary S. McDowell",
"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\nOverall this is an extremely interesting thesis and lays out a lot of the current issues with “the pipeline” succinctly. I have made a number of suggestions that I hope the authors find interesting or useful in revising the paper, and a request for citations. I have marked the paper as approved but am hopeful the authors are receptive to revising based on comments below.\nMajor comments:\nI would clarify when you are talking about a) biomedicine vs all fields and b) issues in the U.S. vs issues globally, where possible. For example, in the introduction you state: “Declining public funding for basic science” and cite examples that are particularly focused on biomedicine in the U.S. Checking throughout the manuscript to ensure that geographical/field limitations of data cited are highlighted and differentiated would be very helpful to see where issues can be generalized or not, and may led to further interesting insights where differences do or do not arise. While the biomedical enterprise in the U.S. is certainly useful as the “canary in the coalmine”, it would be interesting to set this in context.\nI would suggest looking into the work of Dr. Wayne Wahls1-2 and Dr. Kenneth Gibbs Jr.3-7 and also Levecque et al. 20178 and Piefer 20179 and considering citing/including any reactions to these in your discussions.\nIn particular the work of Dr. Gibbs may also help inform a discussion throughout of diversity in terms of gender and race/ethnicity; a key strength of the ideas proposed in this paper is that it can tie into making an enterprise that would help people align their interests with the reasons to stay in various parts of the scientific ecosystem, and in particular diversifying academic faculty through a more fluid scientific ecosystem would be very interesting to explore here.\nMinor comments: Introduction: Suggest deleting “which are” in last sentence of third paragraph in introduction.\n“In 2016 $3 billion were invested into federal agencies to support STEM education programs” - another interesting number would be an estimated amount that is spent by the taxpayer on training writ broadly - i.e. not just training mechanisms and education programs, but also considering that 85% of grads and postdocs funded by NIH (including all foreigners) are not funded from NRSA training mechanisms. Or could it be approximated what proportion of NIH’s budget is linked to trainees in some way? It may also be worth mentioning the previous interest Congressionally in the sustainability of the training enterprise, which led to the creation of the NRSA (see p6 in Addressing the Nation's Changing Needs Report10) - which contains, again, only 15% of its trainees.\nBox 1: “Postdoctoral training periods also continue to expand” - please provide a citation, or points of comparison i.e. specific lengths in specific years, also again specifying countr(y/ies) and field(s). One place to look for data that addresses that talking points may be the recent National Postdoc Survey11.\nIn the discussion of the default faculty path it might be useful to include Dr. Paula Stephan’s analysis of NSF data showing the proportions of U.S. biological sciences PhDs entering particular job sectors 5 years post-PhD (only published in The Atlantic in https://www.theatlantic.com/business/archive/2013/02/the-phd-bust-americas-awful-market-for-young-scientists-in-7-charts/273339/) and also Figure 2-1 in the NASEM “Breaking Through” report could be of interest too12. Dr. Stephan’s analysis in particular shows that tenure-track faculty positions have not been the principal (let alone majority) destination for PhDs for the lifetime of most Millenials, which allows you to qualify how outdated this mentality is.\n“unencumbered with children” - it may be worth looking further into this - in our analysis with the U.S. Census Bureau in http://sjscience.org/article?id=570 and the related Nature Comment piece13, The New Face of US science - the latter in particular points out specifically that men in their 30s are 7x more likely to have a non-working spouse, and many who are married tend to have children, and in fact there may be some discussion to be had about whether being married is beneficial. Men are more likely to be married than women, but women are more likely than men to be married to someone who is *also in the biomedical workforce*. Combining these factors with the advantage that tenure clock interventions designed for women actually confer on men (http://sjscience.org/article?id=570) it may be that the “unencumbered with children” comment requires qualifying, or at least some rethinking/citations to show that being childless does confer an advantage. Again the National Postdoc Survey could be useful to check into here.\nFigure 1: I would suggest with industry also breaking out not only into academic/non-academic but also research/non-research. In preparing the National Academies report, Breaking Through you see in Figure 2-1 a representation of the phenomenon that many industry jobs are not bench-based, but are sales or managerial positions. Consultancy should also be a large group. This would be useful to illustrate the need for the ecosystem even more, as even “industry” is complicated. It might also be worth looking at the taxonomy work that has been going on nationally14.\nFixing the pipeline: I was somewhat confused by the angle the authors were coming from in the last sentence in the “flow in the pipeline” discussion. Why would fewer scientists with more career stability be a bad thing in the tenure-track path? Are they proposing instead more scientists with more career stability, in a diverse range of careers? The authors should be careful not to fall into the “more scientists are better for society” argument, as data about the labor market and PhD outcomes is scarce but points to a massive gap in the labor market with supply far exceeding demand, across all jobs15. If however they are trying to make the argument that fixing flow is too restrictive a view/the issues that arise in focusing only on flow, I believe all but the last sentence support this view. Some expansion here to clarify what is meant would be helpful. Having read through the rest of the Fixing the Pipeline section I am satisfied that clarification is all that is needed here, because I get a strong sense of the line of argument in the other sections.\nThe adapting to the pipeline argument is very well made. It could also be worth pointing out that this is causing pressure to extend the length of postdoc positions, because it is taking longer to get a position, rather than changing the system so that postdoc positions are not so long i.e. by evaluating metrics other than high IF publications in competition with other applicants. I am glad to see the argument being made about for-profit services, it might be worth considering also asking why institutions are not providing such services when they receive federal funds to train these trainees.\nIn the thinking outside the pipeline section it could also be worth pointing out the lack of access to alumni networks, or the poor communication between academics and non-academics, and also the simple inability of mentors, no matter how well-meaning, to advise on careers they themselves do not know about. It is also worth considering that many trainees do not feel comfortable bringing up that they may wish to leave academia, even though the PI may be receptive to this idea. This is largely due to over-arching cultural expectations, which also include the idea that “to leave is to fail”, which could also be mentioned somewhere in here.\nBox 3 - It may also be worth suggesting a vital role professional societies can play in ensuring that “the field” is not synonymous with academia.\nThere was a lot of excellent content in the description of the scientific ecosystem: I would suggest considering a way of placing key elements of this discussion in a figure, as there was a lot of information, and it was sometimes difficult to appreciate all the points being made. Some way of parsing out the information in a slightly more digestible format would be very helpful, to really draw in readers and highlight some of the excellent points made.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "38436",
"date": "01 Oct 2018",
"name": "Cameron Neylon",
"expertise": [
"Reviewer Expertise I work on the culture of researchers and research organisations. I have expertise in research evaluation",
"open research",
"and communication for change."
],
"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 offers a way of thinking about the environment of career structures in science. The other two reviewers have noted a series of small issues which I concur would be good to deal with. The paper overall offers some good critical thinking and summary of issues with the way we think about career structures and job roles and provides advocacy for new ways of thinking. It is generally well written and well argued.\nI am going to raise two broader issues which I think should be addressed before finalising for indexing:\nIs the 'ecosystem' concept well defined and expressed? Is this a 'new paradigm'?\n\nThere has been much criticism of ecosystem as a metaphor, primarily from ecologists who object that the term is used imprecisely. I think here there is a risk of getting muddled. Words like 'ecosystem' and 'landscape' are used in opposition to the traditional concept of a 'pipeline' but are not explicitly defined or drawn as analogies to true ecosystems.\nI was left wondering whether the 'ecosystem' referred to an ecology of job roles, of workers within those roles, organisations that support them or something else? Figure 2 to me does not evoke an ecosystem but a set of roles and organisations. Rather some elements seem to be drawn from Ostrom's Institutional Analysis and Design model, along with specific instances of change, then roles then knowledge (from outside to in). From a different angle it also looks like a social learning model (moving from centre outwards, Lave and Wenger, Situated Learning). In passing I note a tension between the claim the circle 'has no top' with the line above the figure that '....bottom-up changes are already being driven by early career scientists.\nThe model/figure also does not seem to be used further in the examples below. I think this is a sign that the model and narrative is not quite right. If I understand correctly the goal is to use the ecosystem metaphor to emphasise interdependence of actors and the ability to work effectively together (adopting from commons language throughout). However the pipeline model refers to development of those actors. My suggestion would be to consider more explicitly how the pipeline model presumes hierachical institutional forms and how the new model (community/social learning in communities of practice?) would support new, networked and flexible institutional forms.\nThis happens towards the end in 'Writing a new cultural narrative' to a certain extent but I would like to see this made much stronger. I think the relationship between rhetoric, narrative, social learning/acculturation and institutional forms could be made much earlier, ideally in 'The science ecosystem' as an answer to the deficiencies in the pipeline model that are ably described prior to this. This could draw on a range of previous work.\nThis brings me to the second point. The claim is made that this is a new model. I am not convinced this is accurate. Elements of the ecosystem metaphor, or at least the division of labour and necessary organisation to create value, are found in Merton (many of the papers in Sociology of Science), Lave and Wenger (Situated Learning), Ravetz (Scientific Knowledge and its Social Problems), Latour (Politics of Nature) and in others discussing the broad area of Science and Technology Studies (Collins on expertise, Stilgoe on public policy development, Stirling on engagement and diversity etc etc).\nMore directly in STEM debates around the formation and solidification of Bioinformatics as a discipline involving new technical workers and skills (and more recently Digital Humanities - I think it would be good for the application of the model to humanities and social sciences to be at least gestured at), the formation of research institutes, and massive authorship in astronomy, high energy physics, and some areas of biology have all pointed in the direction of an 'ecosystem' of interdependent roles to support modern large-scale science.\nMy primary suggestion is that referring to this earlier literature will strengthen this paper and aid in the development of a stronger narrative, which will in turn hopefully improve the impact of the article in achieving the authors' goals. I would consider re-working Figures 1 and 2:\nFigure 1 emphasises that the rhetoric of a pipeline leads to a hierarchical structuring of institutions and a necessary conservatism as a result. You might consider adding an 'average age' of each transition, job titles etc to emphasise this. This can then be picked up more strongly in Box 1 as to how the presumed path and hierarchy leads to the negative results described. In Figure 2 I'd invert the sense of the diagram. Start at the outside and show many spiral paths inwards to a range of roles and organisations (can the relationship between organisations and roles be shown). This echoes Lave and Wenger's concept of learning as 'legitimate peripheral participation'. It then becomes important to emphasise that there are many such target diagrams and that various paths can encompass several of them.\nI think these two modifications make it easier to refer back to the figures in the examples and cases described in the numbered sections and boxes. That in turn will make the whole narrative stronger.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-803
|
https://f1000research.com/articles/7-189/v1
|
14 Feb 18
|
{
"type": "Research Article",
"title": "Unraveling endometriosis-associated ovarian carcinomas using integrative proteomics",
"authors": [
"Felix Leung",
"Marcus Q. Bernardini",
"Kun Liang",
"Ihor Batruch",
"Marjan Rouzbahman",
"Eleftherios P. Diamandis",
"Vathany Kulasingam",
"Felix Leung",
"Marcus Q. Bernardini",
"Kun Liang",
"Ihor Batruch",
"Marjan Rouzbahman",
"Eleftherios P. Diamandis"
],
"abstract": "Background: To elucidate potential markers of endometriosis and endometriosis-associated endometrioid and clear cell ovarian carcinomas using mass spectrometry-based proteomics. Methods: A total of 21 fresh, frozen tissues from patients diagnosed with clear cell carcinoma, endometrioid carcinoma, endometriosis and benign endometrium were subjected to an in-depth liquid chromatography-tandem mass spectrometry analysis on the Q-Exactive Plus. Protein identification and quantification were performed using MaxQuant, while downstream analyses were performed using Perseus and various bioinformatics databases. Results: Approximately 9000 proteins were identified in total, representing the first in-depth proteomic investigation of endometriosis and its associated cancers. This proteomic data was shown to be biologically sound, with minimal variation within patient cohorts and recapitulation of known markers. While moderate concordance with genomic data was observed, it was shown that such data are limited in their abilities to represent tumours on the protein level and to distinguish tumours from their benign precursors. Conclusions: The proteomic data suggests that distinct markers may differentiate endometrioid and clear cell carcinoma from endometriosis. These markers may be indicators of pathobiology but will need to be further investigated. Ultimately, this dataset may serve as a basis to unravel the underlying biology of the endometrioid and clear cell cancers with respect to their endometriotic origins.",
"keywords": [
"ovarian cancer",
"endometriosis",
"clear cell carcinoma",
"endometrioid carcinoma",
"proteomics",
"bioinformatics"
],
"content": "Introduction\n\nOvarian cancer (OvCa) is not a single disease but is made up of several distinct subtypes, including serous, endometrioid, clear cell and mucinous. It is now accepted that the majority of serous OvCa arise from carcinomas of the fallopian tube secretory epithelium due to histopathological evidence (Kurman & Shih, 2016). Unfortunately, the origins of the endometrioid, clear cell and mucinous subtypes are not well-delineated and these subtypes remain poorly understood. Genomic and morphologic studies have identified links between endometriosis lesions that progress to endometrioid and clear cell OvCa (Prowse et al., 2006; Wang et al., 2015). Of note, ARID1A and PTEN/PIK3CA mutations have been identified as hallmark features of endometriosis-associated OvCa (Jones et al., 2010; Kuo et al., 2009; Wiegand et al., 2014); however, none of these associations have been characterized at the proteomic level and the mechanisms driving tumourigenesis in these precursors have yet to be identified. As such, proteomic profiling may aid in substantiating these purported precursors of non-serous OvCa, as well as in revealing the underlying biology behind why these seemingly distinct diseases converge on the ovaries upon metastasis and clinical presentation.\n\nProteomic profiling of OvCa has mainly revolved around mass spectrometry (MS)-based analyses. The study of protein expression in OvCa has been increasingly important due to the central role of proteins in all biological processes. Moreover, the proteome integrates the cellular genetic information and environmental influences. As such, MS has been increasingly implemented as it allows for simultaneous examination of thousands of proteins in biospecimens. With respect to endometriosis-associated OvCa, there exist limited studies investigating the diseases on a proteomic level. One recent study utilized proteomic approaches to characterize ARID1A and PIK3CA mutations in endometriosis-associated clear cell and endometrioid OvCa (Wiegand et al., 2014), but currently, there are no studies aimed at comprehensive proteomic profiling of these cancers and their suspected endometriotic lesions.\n\nTo this end, an in-depth proteomic analysis was performed on gynaecological tissue specimens using a label-free, liquid chromatography-tandem mass spectrometry (LC/MS-MS) method. Specifically, tissue proteomes of the following specimens were delineated: clear cell ovarian carcinoma (CC); endometrioid ovarian carcinoma (EC); endometriosis (EMT); and benign endometrium (END). This exercise has identified approximately 9000 unique proteins and represents one of the most comprehensive proteomes of endometriosis and its associated cancers to date. These discovery data may serve as the basis to identifying markers of disease as well as understanding the pathobiology for these endometriosis-associated ovarian cancers.\n\n\nMethods\n\nGynaecological tissue samples from a total of 21 patients were retrospectively identified and selected for proteomic analysis. All samples were collected at University Health Network, Toronto, Canada (UHN REB Number 13-6360-CE ) and immediately stored in liquid nitrogen until retrieval. Approximately 1 cm3 of each fresh, frozen sample was retrieved for proteomic analysis. All samples were histopathologically confirmed for their diagnoses and tissue purity (at least 70%) by a gynaecologic pathologist using matched formalin-fixed, paraffin-embedded tissue slides. With respect to the histopathological diagnoses: six cases of CC, seven cases of EC, three cases of EMT and five cases of END were identified. A detailed description of clinical and histological characteristics of the patients can be found in Table 1.\n\nFresh, frozen samples were first grinded with 0.05% RapiGest (Waters, MA, USA) in 50 mM ammonium bicarbonate. The tissue mixtures were then homogenized and sonicated in order to disrupt cell membranes. This was followed by two rounds of centrifugation at 13000 rpm for 30 minutes at 4°C and collection of the supernatant. Total protein concentration of the supernatant was determined using the Bradford protein assay. After adjusting for 1 mg of total protein content, each sample was subjected to reduction with 15 mM dithiothreitol (Sigma-Aldrich, ON, Canada) in 50 mM ammonium bicarbonate at 60°C for 30 minutes, followed by alkylation with 15 mM iodoacetamide (Sigma-Aldrich, ON, Canada) in 50 mM ammonium bicarbonate for 45 minutes in the dark at room temperature. Protein digestion was carried out with trypsin (Sigma-Aldrich, ON, Canada) in 50 mM ammonium bicarbonate (1:50 trypsin to total protein ratio) overnight at 37°C. RapiGest and trypsin digestion were stopped with the addition of 1% trifluoroacetic acid followed by centrifugation at 13000 rpm for 30 minutes at 4°C. Digested samples were immediately frozen at -80°C until all samples were ready for fractionation via high-performance liquid chromatography (HPLC) – using strong cation exchange (SCX) columns – to reduce peptide complexity.\n\nTrypsinized samples were diluted two-fold in mobile phase A (0.26 M formic acid, 5% acetonitrile, pH 2–3) and loaded directly onto a 500 μL loop connected to a PolySULFOETHYL A Column (2.1 mm × 200 mm; 5 μ; 200 Å; The Nest Group, Inc., MA, USA), containing a silica-based hydrophilic, anionic polymer (poly-2-sulfoethyl aspartamide). The Agilent 1100 HPLC system (Agilent Technologies, Germany) was used for SCX peptide fractionation. A 60-minute gradient was employed with a gradual increase of mobile phase B (0.26 M formic acid, 5% acetonitrile, 1 M ammonium formate, pH 4–5) starting at 30 minutes (30–40 minutes 20% mobile phase B; 40–55 minutes 100% mobile phase B) for the elution of peptides at a flow rate of 200 μL/minute. The eluent was monitored at a wavelength of 280 nm and fractions were collected every three minutes from 28 to 55 minutes resulting in a total of 9 fractions per sample. SCX column and system performance were assessed by running a quality control peptide mixture consisting of 1 μg/μL alpha bag cell peptide, 1 μg/μL fibrinogen fragment, 5 μg/μL human adrenocorticotropic hormone, and 5 μg/μL angiotensin-converting enzyme inhibitor (American Protein Company, CA) after every biological sample.\n\nThe SCX fractions were purified through C-18 OMIX Pipette Tips (Agilent Technologies, Germany) to remove impurities and salts as well as to resuspend the tryptic peptides in a buffer compatible with the mass spectrometer. The fractions were eluted in 5 μL of 65% MS buffer B (90% acetonitrile, 0.1% formic acid, 10% water, 0.02% trifluoroacetic acid) and 35% MS buffer A (5% acetonitrile, 0.1% formic acid, 95% water, 0.02% trifluoroacetic acid). Using an auto-sampler, 18 µL of each sample were injected into an in-house packed 3.3 cm trap pre-column (5 μm C18 particle, column inner diameter 150 μm) and peptides were eluted from the 15 cm analytical column (3 μm C18 particle, inner diameter 75 μm, tip diameter 8 μm). The LC EASY-nLC system (Thermo Fisher, Denmark) was coupled online to the Q-Exactive Plus (Thermo Fischer, CA, USA) mass spectrometer with a nanoelectrospray ionization source. A 60 min LC gradient was applied with an increasing percentage of MS buffer B for peptide elution at a flow rate of 300 nL/min. Full MS1 scan was acquired from a scan range of 400–1500 m/z in the Orbitrap at a resolution of 70000, followed by the MS2 scans for the top 12 precursor ions at a resolution of 17500 in a data-dependent acquisition mode and isolation window of 1.6 m/z. The dynamic exclusion was enabled for 45 seconds and unassigned charge, as well as charge states +1 and +4 to ≥8 were omitted from MS2 fragmentation. Each biological sample was separated into nine fractions with each fraction being subjected to a 60 minute LC gradient. To minimize for instrumentation bias, samples were run in batches with each batch containing approximately one patient from each disease cohort. Quality controls were run between each biological sample (before and after all nine fractions from a sample) as well as between batches to ensure consistent machine performance for all samples.\n\nRAW files were uploaded into MaxQuant ver. 1.5.2.8 (Cox & Mann, 2008) and searched with Andromeda (built into MaxQuant) against the human UniProtKB/Swiss-Prot database (January 2015 release; 550299 sequence entries). Search parameters included a fixed carbamidomethylation of cysteines and variable modifications of methionine oxidation and N-terminal acetylation. Data was initially searched against a “human first search” database with a parent tolerance of 20 ppm and a fragment tolerance of 0.5 Da in order to calculate and adjust the correct parent tolerance to 5 ppm for the search against the UniProtKB/Swiss-Prot database. During the search, the UniProtKB/Swiss-Prot database was randomized and false detection rate was set to 1% at the peptide and protein levels. Data was analyzed using “Label-free quantification” checked, and the “Match between runs” interval was set to 2 minutes. Proteins were identified with a minimum of two unique peptides. “LFQ Intensity” columns corresponding to the extracted ion current value of each protein were used for further statistical analyses to determine overexpressed proteins.\n\nThe produced MaxQuant output matrix was loaded onto the Perseus software (ver. 1.5.2.6) to perform statistical and bioinformatics analyses (Tyanova et al., 2016). Specifically, LFQ data were logarithmically-transformed and imputed by creating a Gaussian distribution of random numbers with a standard deviation of 30% relative to the standard deviation of the measured values and one standard deviation down-shift of the mean to simulate the distribution of low signal values. Hierarchical clustering of proteins was performed on logarithmized intensities after z-score normalization of the data, using Euclidean distances. PCA was performed on logarithmized values using singular value decomposition in order to find the principal components. Gene Ontology, the Protein Analysis Through Evolutionary Relationships (PANTHER) Classification System (Mi et al., 2013), and ConsensusPathDB-human (Kamburov et al., 2009) were utilized to retrieve additional annotations.\n\nQuantitative protein data were log-transformed before statistical analyses. Proteins with missing values (‘0’ normalized values) were imputed using the Perseus software. The elevated protein levels were detected using LIMMA, an empirical Bayes method proposed by Smyth (Smyth, 2004). The posterior probability of being elevated for each protein was computed using the two-group model (Efron & Tibshirani, 2002) with true null probability estimated by the right-boundary procedure (Liang & Nettleton, 2012). Protein levels in control tissues were assumed to follow normal distributions and estimated their means and variances conservatively.\n\n\nResults\n\nTo decipher the in-depth proteome of the tissues, we utilized an offline two-dimensional LC-MS/MS workflow amenable to label-free quantification (LFQ) with minimal amount of protein required upfront (1mg total protein). As seen in Figure 1A, lysis, reduction and alkylation, and trypsin digestion were performed in a single-tube manner prior to SCX-HPLC. The additional offline SCX-HPLC coupled with online C18 reverse phase chromatography fractionation and MS/MS analysis on the Q-Exactive Plus ensured comprehensive proteome coverage and high mass accuracy.\n\n(A) Overview of the label-free LC-MS/MS workflow developed for proteomic analysis of tissue specimens. (B) Summary of number of proteins identified overall and within the patient cohorts. The error bars represent standard deviation. (C) Interpatient correlation of the protein expression profiles.\n\nOverall, 8793 unique proteins were identified across the 21 patient samples with roughly 5000 unique proteins identified within each patient cohort, as seen in Figure 1B. In terms of overlap, 5379 proteins were identified in all four cohorts (Supplementary Figure 1). Approximately 90% of all proteins were identified with at least 2 or more peptides (Supplementary Figure 2). With respect to protein expression patterns, strong correlations between samples within the same patient cohort were observed. As seen in Figure 1C, protein expression was relatively consistent when comparing patients with the same diagnosis with Pearson correlation coefficients (PCC) above approximately 0.80 for CC, EC, EMT and END. A notable exception was patient EC 18 who displayed decreased correlation with the other EC patients with PCC ≤0.72. Interestingly, subsequent re-examination of Patient 18 revealed more mucinous-like histological features and it was concluded that the final diagnosis of EC (as opposed to mucinous carcinoma) was due to the lack of metastatic gastrointestinal involvement (data not shown). Patient 18 was removed from further analyses given its mucinous-like features, thus making it less likely to be an endometriosis-associated cancer. The biological soundness of the proteomic data was further demonstrated by the stable expression of housekeeping proteins, such as ribosomal biogenesis and assembly proteins, across all samples (Supplementary Figure 3). Collectively, the minimal variation with respect to patient cohort and housekeeping protein expression (despite being processed in different batches) suggests that any differences observed in this dataset are due to true biological differences and not from technical artefacts.\n\nAs an initial assessment of the accuracy of proteomic profiling, we investigated the expression of various immunohistochemical (IHC) markers used in histopathological analysis of endometriosis-associated ovarian cancers (Kalloger et al., 2011; Köbel et al., 2008; Köbel et al., 2014). A spectrum of markers including general epithelial ovarian carcinoma markers and markers specific to various subtypes (serous, endometrioid, clear cell, and mucinous) were identified with the full panel of markers summarized in Figure 2. Overall, the proteomic data across the six CC and six EC samples correlated well with IHC expression based on literature evidence. Serous-specific markers such as Wilm’s Tumour 1 (WT1), p53 (TP53), cytokeratin 7 (CK7) and cytokeratin 20 (CK20) and mucinous-specific markers such as carcinoembryonic antigen (CEA) and mucin 2 (MUC2) were found to be expressed in almost none of the CC or EC samples. Meanwhile, EC-specific markers such as estrogen receptor (ER), progesterone receptor (PR), p16 (CDKN2A), trefoil factor 3 (TFF3), Dickkopf-related protein 1 (DKK1), and matrix metalloproteinase 7 (MMP7) and the CC-specific marker hepatocyte nuclear factor 1β (HNF1B) displayed near exclusive expression in their respective subtypes. Expression of these markers in endometriotic and benign endometrial tissues was mostly low with the exception of ER and PR which were found to be constitutively expressed in the benign endometrial tissues (data available in “MaxQuant Analysis” dataset).\n\nAn expression matrix demonstrating the correlation of proteomic data with known IHC markers. Label-free quantification values (in arbitrary units represented by the purple gradient scale) were generated from MaxQuant and known IHC markers across the various subtypes were based on collating literature evidence. The expression of the markers across the serous (SC), clear cell (CC), endometrioid (EC) and mucinous (MC) subtypes are denoted by the red, green, blue and indigo lines, respectively.\n\nInterestingly, mucin 16 (MUC16) and WAF four-disulfide core domain protein 2 (WFDC2) displayed high overall expression in the EC cohort and variable expression in the CC cohort. Mucin 16 (also known as CA125) and WAF four-disulfide core domain protein 2 (also known as HE4) are the only clinically-approved serum markers of ovarian cancer and have been shown to perform better for serous OvCa compared to other subtypes (Leung et al., 2016; Shen et al., 2017). Here, our data suggests that serum levels of such tumour markers (Table 1) do not entirely capture the biology of the actual tumours and may actually be dependent on tumour release of the markers into circulation rather than tumour production.\n\nFurthermore, protein expression of ARID1A, PTEN, and PIK3CA was investigated as mutations in these genes are often prevalent in endometriosis-associated cancers and have been suggested to be involved with genomic instability and tumourigenesis (Jones et al., 2010; Kuo et al., 2009; McConechy et al., 2014; Wiegand et al., 2014; Yamamoto et al., 2012). Despite genomic data that suggests concurrent loss of tumour suppressors ARID1A and PTEN with activating mutations of PIK3CA are involved in the pathogenesis of CC and EC, our data demonstrates that these genomic findings do not necessarily translate at the protein level. Indeed, variable expression of the three proteins is observed across the CC and EC cohorts. However, it is important to note that expression of these proteins does not imply proper functionality and thus, they may still contribute to tumourigenesis due to deleterious mutations.\n\nTo evaluate how biological differences between tumours (CC or EC) and their precursors (EMT and END) were reflected at the proteomic level, unsupervised clustering was performed using the entire proteomic dataset without a priori enrichment. Using normalized log2 expression values, samples were analyzed using principal component analysis (PCA) to visualize the clustering of samples from the same cohorts as well as the separation between the cohorts. As expected, PCA revealed that in both comparisons of CC to EMT and END, and EC to EMT and END, samples within a cohort clustered together to produce clear separation between the cohorts with respect to overall protein expression (Figures 3A). The strong segregation between the END, EMT and CC/EC cohorts suggested, once again, that differences between the cohorts are due to biological variation and that inspection of the proteomic landscape recapitulated these differences very well.\n\nTo elucidate the biological mechanisms contributing to the differences between tumours and their precursors, proteins demonstrating the highest variances between the cohorts had to be enriched for first. Using an ANOVA test with a Bonferroni-Hochberg FDR of 0.01, 127 proteins for the comparison of CC versus EMT/END and 119 proteins for the comparison of EC versus EMT/END were identified as having the strongest differential expression between the cohorts (Supplementary Table 1). Subsequently, unsupervised hierarchical clustering was performed using these differentially-expressed proteins to verify the ability to distinguish between the cohorts, as well as to identify any notable clusters of differential expression patterns as seen in Figures 3B and 3C. As expected, the differential proteins were able to accurately separate all of the cohorts and produced distinct patterns of differential expression between tumours and their precursors. To identify overrepresented ontologies within these patterns of differential expression, enrichment analysis for Gene Ontology annotations was performed using the PANTHER Classification System. For CC, two major clusters were found to be characterized by predominantly muscle- and cytoskeletal-related processes (Figure 3B), while for EC, two major clusters were found to be characterized by cell junction- and extracellular matrix-related processes (Figure 3C). A condensed list of overrepresented annotations in both cancers are displayed in Supplementary Table 2.\n\n(A) Principal component analysis of entire proteomic dataset without enrichment. Hierarchical clustering of differentially-expressed proteins between CC, EMT and END (B) and EC, EMT and END (C) with overrepresented GO annotations identified through enrichment analysis.\n\nTo assess concordance with existing genomic data, the proteomic dataset was compared against reported ‘subtype-specific gene signatures’ for CC and EC in previous RNA expression studies. For CC, a 113-gene signature that was reported to be able to differentiate between CC and high-grade serous OvCa was used (Hughes et al., 2016) while for EC, two studies comparing EC against high-grade serous OvCa were used after identifying 15 underexpressed genes and 40 overexpressed genes in EC (Banz et al., 2010; Uehara et al., 2015). Of the 113 CC-specific genes, 11/34 of the underexpressed genes were downregulated at the proteomic level while 40/79 overexpressed genes were upregulated at the proteomic level (Figure 4A). It is important to note that 10 of the 40 concordant genes were significantly upregulated in our data (FDR = 0.05, S0 = 1) with many of them being known markers of CC including napsin A (NAPSA), annexin 4 (ANXA4) and hepatocyte nuclear factor 1-beta (HNF1B). Furthermore, cystathionine gamma-lyase (CTH) and interleukin-6 receptor subunit beta (IL6RB) – markers that have been associated with CC (Cochrane et al., 2017; Hughes et al., 2016; Yanaihara et al., 2016) – were found to be elevated in CC compared to EMT samples. CTH was found to be significantly elevated in CC compared to EMT (p=0.01) while IL6RB was undetectable in all EMT samples but expressed in 4/6 of the CC samples. For EC, 3/15 of the underexpressed genes were downregulated at the proteomic level while 21/40 overexpressed genes were upregulated at the proteomic level (Figure 4B). Unlike the observations in CC, only one of the concordant proteins (desmoplakin or DSP) was significantly upregulated. Interestingly, the concordant genes for both CC and EC were able to discriminate cancer from control patients at the proteomic level but could not distinguish EMT from END patients (Supplementary Figure 4).\n\nVolcano plot of CC versus EMT proteomes (A) and EC versus EMT proteomes (B) overlaid with concordant genomic features. The black lines denote statistical significance. (C) depicts overrepresented pathways identified in the ‘disease signature’ derived for CC.\n\nTo further assess how well existing genomic data translates at the proteomic level, ARID1A and PIK3CA/PTEN-related proteins were inspected for their expression levels as these pathways are often perturbed in EC and CC. Overall, 21/37 ARID1A components and 49/254 PIK3CA/PTEN components were identified in our proteomic data (components identified using Reactome; http://www.reactome.org/). Unfortunately, these pathway-related proteins were poor discriminators and did not show obvious differential expression between the cancer and control cohorts (Supplementary Figure 5), further highlighting that fact that genomic features do not always translate at the proteomic level.\n\nIn order to identify potential disease markers on a proteomic level, proteins with progressively increased expression from END to EMT to CC/EC were identified using LIMMA (see Materials & Methods for details). Using the following criteria: (A) progressively increased expression from END to EMT to EC/CC; (B) only increased expression from END to EMT; and (C) only increased expression from EMT to EC/CC, 25 proteins for EC and 252 proteins for CC were identified as ‘disease markers’ (Supplementary Table 3–Supplementary Table 4). These disease marker signatures were able to discriminate between the different disease cohorts more accurately than the genomic signatures, especially with respect to differentiating between EMT and END patients (Supplementary Figure 6). Further analysis of the CC signature revealed that there was an overrepresentation of pathways involving hepatocyte growth factor receptor (MET), α6β4 integrin, and retinoic acid (Figure 4C). Pathway analysis was not performed for the EC signature due to a low number of proteins.\n\n\nDiscussion\n\nIn this study, a comprehensive proteomic analysis of 21 tissues from CC and EC tumours, as well as benign endometriotic precursors was performed. The identification of almost 9000 unique proteins represents the most in-depth proteomic profiles of these gynaecological specimens to date. In fact, the workflow used has generated one of the largest datasets to date with the Q-Exactive Plus. As a comparison, in a study delineating the proteomes of pancreatic tissue specimens from controls and Type 1 diabetic patients, approximately 5500 unique proteins were identified using a similar workflow on the Q-Exactive Plus (Liu et al., 2016).\n\nAlthough isotope-based labeling methods are the gold standard for quantitative proteomics, LFQ is becoming increasingly utilized due to its simplicity and practicality. Unlike labeling methods, LFQ avoids extra pre-analytical steps and can be applied to any type of sample. This is especially relevant in clinical samples that cannot be metabolically labeled using standard labeling techniques. Furthermore, recent advances with the MaxLFQ algorithm in the MaxQuant software have greatly increased the accuracy and robustness of LFQ (Cox et al., 2014). In this study, quality assessment demonstrated that proteomic data was consistent across all samples with respect to intracohort variation and housekeeping protein expression. Furthermore, correlation and clustering analysis without enrichment demonstrated that samples separated into relatively distinct clusters within their respective cohorts.\n\nThe ability of label-free quantitative proteomics to accurately recapitulate CC and EC was further highlighted by the correlation of the proteomic data with clinical IHC markers. Whereas known markers specific to the serous and mucinous subtypes were not identified almost none of the CC and EC samples, markers specific to CC and EC were identified almost exclusively in their respective subtypes. A caveat here is that our proteomic analysis will not capture specific expression patterns (such as focal or diffuse) that are commonly used in IHC-based differential diagnosis of ovarian tumours (Kaspar & Crum, 2015). On the other hand, proteomic profiling may reveal insights that IHC-based diagnoses are unable to as the former considers protein expression across the entirety of the tumour while the latter interrogated localized areas. For example, the expression of HNF1B in one of the EC samples (sample EC4) may indicate a mixed CC/EC origin in that tumour due to the fact that HNF1B is often used to rule in a diagnosis of CC. Retrospective analysis of such ‘discordant’ tumours would be useful to determine if these tumours contained any areas of CC-like histology. Additionally, re-examination of sample EC18 via histopathology revealed more mucinous-like features which was reflected in its proteomic profile being discordant with the other EC samples. In these regards, proteomic profiling may thus enhance IHC-based diagnoses by offering a macroscopic view of ovarian tumours that would otherwise be missed.\n\nFinally, bioinformatic analyses revealed that the CC and EC proteomes are potentially more informative than genomic data with regards to markers potentially implicated in disease pathogenesis. The CC proteome was characterized by muscle- and cytoskeleton-related processes, while the EC proteome was represented by cell junction- and extracellular matrix-related processes. Further analysis with the disease signatures revealed novel contributions from MET, α6β4 integrin and retinoic acid pathways for CC. This is the first study to identify these associations as possible avenues of pathogenesis from endometriosis, but further functional studies will need to be performed to elaborate their true roles in tumourigenesis. Nevertheless, such distinctions may be indicative of the different underlying biology and mechanisms that contribute to development of CC and EC from EMT. As a result, each subtype could be targeted for their key pathogenic mechanisms instead of the standard platinum-based chemotherapy administered to all OvCa patients. Future studies should thus expand on the use of proteomic-based profiling of endometriosis-associated cancers in order to provide novel insights into the etiology and pathogenesis of the diseases, which in turn, will affect diagnosis and treatment of these cancers.\n\nIn summary, this study has generated proteomic data for endometriosis and its associated (ovarian) cancers. Overall, we have demonstrated that not only is our dataset robust and comprehensive, but it is also reflective of the molecular profiles of the various diseases. Clustering analysis revealed unique expression patterns within the cohorts and that proteomic profiling may serve as a more accurate representation than genomic profiling. As well, disease signature analyses have demonstrated that each cancer subtype is characterized by distinct markers which can be exploited for further insight into the etiology of each subtype as well as identification of novel therapeutic targets.\n\n\nData availability\n\nThe output of the MaxQuant analyses is available in the spreadsheet ‘MaxQuant Analysis’ along with an interpretive guide (‘MaxQuant Analysis Guide’). Additional files or raw data can be provided upon reasonable request.\n\nDataset 1: MaxQuant Analysis Guide – An interpretive guide for the spreadsheet output from the MaxQuant analysis. 10.5256/f1000research.13863.d193746 (Leung et al., 2018a)\n\nDataset 2: MaxQuant Analysis – Spreadsheet output from the label-free quantification analysis using the MaxQuant software. The spreadsheet contains the details of the identification and quantitation of all proteins across the 21 biological samples. 10.5256/f1000research.13863.d193821 (Leung et al., 2018b)",
"appendix": "Competing interests\n\n\n\nThe authors declare they have no competing interests.\n\n\nGrant information\n\nThis work was funded by a grant to EPD from the Early Detection Research Network of the NIH, USA (Grant #5U01CA152755).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Dr. Blaise Clarke for his help in reviewing the pathology specimens.\n\n\nSupplementary material\n\nSupplementary Table 1 – Differential proteins identified in CC and EC compared to EMT and END.\n\nClick here to access the data.\n\nSupplementary Table 2 – Overrepresented ontologies identified with enrichment analysis of Gene Ontology (GO) annotations.\n\nClick here to access the data.\n\nSupplementary Table 3 – Differential proteins identified with significant increased expression from END to EMT to EC.\n\nClick here to access the data.\n\nSupplementary Table 4 – Differential proteins identified with significant increased expression from END to EMT to CC.\n\nClick here to access the data.\n\nSupplementary Figure 1 – Venn diagram displaying overlap of proteins between the four patient cohorts.\n\nClick here to access the data.\n\nSupplementary Figure 2 – Identified proteins categorized according to the number of peptide hits associated with each protein identification. An identified protein was defined as any protein with a non-zero normalized LFQ value in at least one patient sample.\n\nClick here to access the data.\n\nSupplementary Figure 3 – Conservation of levels of ribosomal biogenesis and assembly proteins across the 21 biological replicates. The protein intensity levels are derived from the label-free quantification intensities generated from MaxQuant.\n\nClick here to access the data.\n\nSupplementary Figure 4 – Clustering analysis of concordant proteins across the cancer and control cohorts.\n\nClick here to access the data.\n\nSupplementary Figure 5 – Expression levels of ARID1A-related proteins (A) and PIK3CA/PTEN-related proteins (B) across the cancer and control cohorts.\n\nClick here to access the data.\n\nSupplementary Figure 6 – ‘Disease signatures’ for CC (A) and EC (B) derived from identifying proteins with progressively increased expression from END to EMT to EC/CC.\n\nClick here to access the data.\n\n\nReferences\n\nBanz C, Ungethuem U, Kuban RJ, et al.: The molecular signature of endometriosis-associated endometrioid ovarian cancer differs significantly from endometriosis-independent endometrioid ovarian cancer. Fertil Steril. 2010; 94(4): 1212–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCochrane DR, Tessier-Cloutier B, Lawrence KM, et al.: Clear cell and endometrioid carcinomas: are their differences attributable to distinct cells of origin? J Pathol. 2017; 243(1): 26–36. PubMed Abstract | Publisher Full Text\n\nCox J, Mann M: MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008; 26(12): 1367–1372. PubMed Abstract | Publisher Full Text\n\nCox J, Hein MY, Luber CA, et al.: Accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction, termed MaxLFQ. Mol Cell Proteomics. 2014; 13(9): 2513–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEfron B, Tibshirani R: Empirical bayes methods and false discovery rates for microarrays. Genet Epidemiol. 2002; 23(1): 70–86. PubMed Abstract | Publisher Full Text\n\nHughes CS, McConechy MK, Cochrane DR, et al.: Quantitative profiling of single formalin fixed tumour sections: proteomics for translational research. Sci Rep. 2016; 6: 34949. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJones S, Wang TL, Shih IeM, et al.: Frequent mutations of chromatin remodeling gene ARID1A in ovarian clear cell carcinoma. Science. 2010; 330(6001): 228–231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalloger SE, Köbel M, Leung S, et al.: Calculator for ovarian carcinoma subtype prediction. Mod Pathol. 2011; 24(4): 512–521. PubMed Abstract | Publisher Full Text\n\nKamburov A, Wierling C, Lehrach H, et al.: ConsensusPathDB--a database for integrating human functional interaction networks. Nucleic Acids Res. 2009; 37(Database issue): D623–628. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaspar HG, Crum CP: The utility of immunohistochemistry in the differential diagnosis of gynecologic disorders. Arch Pathol Lab Med. 2015; 139(1): 39–54. PubMed Abstract | Publisher Full Text\n\nKöbel M, Bak J, Bertelsen BI, et al.: Ovarian carcinoma histotype determination is highly reproducible, and is improved through the use of immunohistochemistry. Histopathology. 2014; 64(7): 1004–1013. PubMed Abstract | Publisher Full Text\n\nKöbel M, Kalloger SE, Boyd N, et al.: Ovarian carcinoma subtypes are different diseases: implications for biomarker studies. PLoS Med. 2008; 5(12): e232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuo KT, Mao TL, Jones S, et al.: Frequent activating mutations of PIK3CA in ovarian clear cell carcinoma. Am J Pathol. 2009; 174(5): 1597–1601. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKurman RJ, Shih IeM: The Dualistic Model of Ovarian Carcinogenesis: Revisited, Revised, and Expanded. Am J Pathol. 2016; 186(4): 733–47. PubMed Abstract | Publisher Full Text\n\nLeung F, Bernardini MQ, Brown MD, et al.: Validation of a novel biomarker panel for the detection of ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2016; 25(9): 1333–1340. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeung F, Bernardini MQ, Liang K, et al.: Dataset 1 in: Unraveling endometriosis-associated ovarian carcinomas using integrative proteomics. F1000Research. 2018a. Data Source\n\nLeung F, Bernardini MQ, Liang K, et al.: Dataset 2 in: Unraveling endometriosis-associated ovarian carcinomas using integrative proteomics. F1000Research. 2018b. Data Source\n\nLiang K, Nettleton D: Adaptive and dynamic adaptive procedures for false discovery rate control and estimation. J R Stat Soc Series B Stat Methodol. 2012; 74(1): 163–82. Publisher Full Text\n\nLiu CW, Atkinson MA, Zhang Q: Type 1 diabetes cadaveric human pancreata exhibit a unique exocrine tissue proteomic profile. Proteomics. 2016; 16(9): 1432–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcConechy MK, Ding J, Senz J, et al.: Ovarian and endometrial endometrioid carcinomas have distinct CTNNB1 and PTEN mutation profiles. Mod Pathol. 2014; 27(1): 128–134. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMi H, Muruganujan A, Casagrande JT, et al.: Large-scale gene function analysis with the PANTHER classification system. Nat Protoc. 2013; 8(8): 1551–66. PubMed Abstract | Publisher Full Text\n\nProwse AH, Manek S, Varma R, et al.: Molecular genetic evidence that endometriosis is a precursor of ovarian cancer. Int J Cancer. 2006; 119(3): 556–62. PubMed Abstract | Publisher Full Text\n\nShen F, Lu S, Peng Y, et al.: Performance of ROMA based on Architect CA 125 II and HE4 values in Chinese women presenting with a pelvic mass: A multicenter prospective study. Clin Chim Acta. 2017; 471: 119–125. PubMed Abstract | Publisher 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(1): Article3. PubMed Abstract | Publisher Full Text\n\nTyanova S, Temu T, Sinitcyn P, et al.: The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat Methods. 2016; 13(9): 731–740. PubMed Abstract | Publisher Full Text\n\nUehara Y, Oda K, Ikeda Y, et al.: Integrated copy number and expression analysis identifies profiles of whole-arm chromosomal alterations and subgroups with favorable outcome in ovarian clear cell carcinomas. PLoS One. 2015; 10(6): e0128066. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Mang M, Wang Y, et al.: Tubal origin of ovarian endometriosis and clear cell and endometrioid carcinoma. Am J Cancer Res. 2015; 5(3): 869–79. PubMed Abstract | 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\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\nYanaihara N, Hirata Y, Yamaguchi N, et al.: Antitumor effects of interleukin-6 (IL-6)/interleukin-6 receptor (IL-6R) signaling pathway inhibition in clear cell carcinoma of the ovary. Mol Carcinog. 2016; 55(5): 832–41. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "31989",
"date": "03 Apr 2018",
"name": "Nathalie Lepage",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors of the manuscript evaluated protein content and quantitation of tissues collected from 21 patients with a range of ovarian diseases (endometriosis, endometrioid and clear cell ovarian carcinomas). They used a well-designed proteomics approach. Their findings were unique and did not exactly correlated with current knowledge from genomic analysis. Clearly additional work will be required to confirm and expand on their current findings.\n\nMinor comments:\nThere are some proteins that were identified only using 1 peptide (>500 proteins). The authors should elaborate of the expected protein size for these proteins and confirm that the identification based on 1 peptide is sufficient.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3724",
"date": "20 Jun 2018",
"name": "Vathany Kulasingam",
"role": "Author Response",
"response": "Thank you for raising this point. We would like to clarify that although the dataset contains proteins identified only using 1 peptide, these proteins were not selected for further analyses (only those with a minimum of 2 unique peptides were investigated further). We have modified our manuscript to reflect this point. We are confident that through setting the false discovery rate (FDR) at 1% for both the peptide and protein levels, we were able to minimize false identifications even for proteins of higher molecular weight. This has been shown to be fairly resistant to false identifications even with 1 peptide identification (Gupta N & Pevzner PA. J Proteome Res 8:4173-4181 2009). The expected sizes of the proteins (identified using 1 peptide) range from approximately 10-670 kDa."
}
]
},
{
"id": "32818",
"date": "23 Apr 2018",
"name": "Rajeevan Selvaratnam",
"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\nLeung et al., has applied a proteomic profiling approach to solubilized gynecological tissue samples to delineate protein markers differentiating subtypes of non-serous ovarian cancer. Notably, these sub-types include protein markers that may differentiate endometrioid and clear cell carcinoma. Using 21 patient samples, of which 6 patients were histopathologically diagnosed with clear cell carcinoma and another 6 samples with endometrioid carcinoma, the authors have identified over 5000 proteins in the small cohorts using mass spectrometry based label-free quantitative proteomics approach. In this regard, I found Supplementary Figure 1 with the Venn diagram to be more informative than Figure 1b. It may be worthwhile to consider swapping Figure 1b with supplementary Figure 1. In general, this manuscript has many figures that nicely complement each other, including the covariance analysis presented in Figure 1c and the principal component analysis in Figure 3a.\nThe experimental procedures applied here are sound. However, I am curious to know if the 1 cm3 of the tissue is a randomly selected segment or purposely segmented region of the tissue that define the tumor stage. This may be important for the samples in the early tumor stages and in general where there is heterogeneity as may be the case for sample 18. Nonetheless, it appears the approach is to be a macroscopic view as the authors note and the results are encouraging. I found the strength of the study was in the ability to show the contrast between previous genomic data and the author’s proteomic findings (notably for the expression of ARI1DA, PTEN, and PIK3CA). However, the authors also indicate some closeness of their findings to previous genomic data on markers associated with clear cell carcinoma.\nAlthough the size of samples or cohorts may be small, these initial findings are easily appreciated and much needed to pursue further directions toward understanding non-serous ovarian cancer subtypes. Given the novelty of potential pathways implicated in clear cell carcinoma from the findings and need for further evaluation on a clinical and scientific basis, I believe this manuscript needs only minor revisions and should be accepted for indexing.\nMinor comments:\nReport centrifugation speeds in x g unit, as the RPM is not indicative of the force and varies with radius of the centrifuge. Figure 1b caption. Authors note the error bars are the standard deviation. Briefly elaborate on the standard deviation or spread of which data (e.g. within each data set based on 2 minimum peptides, or software determined, etc).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3723",
"date": "20 Jun 2018",
"name": "Vathany Kulasingam",
"role": "Author Response",
"response": "Thank you for raising these important points. The caption for Figure 1b has been modified to explain that the standard deviations were calculated within each data set based on protein identifications with a minimum of 2 unique peptides. As well, Figure 1b has been switched with Supplementary Figure 1. Dr. Selvaratnam’s point of heterogeneity within the tissue sample is valid. It was decided that macroscopic dissection of the frozen tissue block combined with histopathological review for tumour purity of that section (>70% tumour tissue with minimal necrotic/benign tissue) was sufficient for the purposes of this study. Some degree of heterogeneity is unavoidable but as the reviewer has noted, our proteomic data does show concordance with some of the more prominent genomic markers for both endometrioid and clear cell carcinomas. The centrifugation units have been modified as per the reviewer’s comments."
}
]
}
] | 1
|
https://f1000research.com/articles/7-189
|
https://f1000research.com/articles/7-770/v1
|
19 Jun 18
|
{
"type": "Review",
"title": "Inflatable penile prosthesis in the radical prostatectomy patient: a review",
"authors": [
"Nelson Bennett",
"I-shen Huang",
"I-shen Huang"
],
"abstract": "In the population of patients with prostate cancer, survivorship has come to the forefront of continuity-of-care. In addition to urinary control, erectile function is a significant issue after radical pelvic surgery. Penile prosthesis surgery remains an excellent option for restoring erectile function to those for whom more conservative measures have failed. This review article outlines the anatomical, surgical and post-operative consideration involved in the placement of a penile prosthesis in this special patient population.",
"keywords": [
"erectile dysfunction",
"penile prosthesis",
"prostate surgery"
],
"content": "Erectile dysfunction (ED) rates after radical prostatectomy (RP)\n\nPenile erection is the culmination of complex series of highly integrated phenomena involving the central nervous system, peripheral nervous system, endocrine system and the vascular system. These systems must be working in concert and at a high level in order for full erection to occur. ED may occur when there is an impairment or derangement to any of these systems. ED has been defined as the inability of a man to achieve or maintain an erection sufficient for satisfactory penetrative sexual intercourse1.\n\nPost-RP ED may be neurogenic, venogenic, arteriogenic or a combination of these etiologies. In all cases, injury of the cavernous nerves occurs during dissection of the prostate. The injury, whether caused by contact, traction, electro-cautery or transection, initiates a cascade of events that culminates in ED. Microscopically, cavernous nerve-fiber injury initiates Wallerian degeneration that will incapacitate the axon back to the cell body, typically at the level of the spinal cord. The lack of nervous input at the end-organ (cavernous muscle) is a major contributor to cavernosal tissue degeneration and atrophy2,3. Venous leakage is another underlying mechanism responsible for ED after prostatectomy. Bilateral cavernosal nerve injury has been shown to induce cavernosal smooth muscle death which will lead to veno-occlusive dysfunction3. Chronic hypoxia, denervation, and activation of the TGF-β cascade are believed to initiate the apoptosis process, and increase the deposition of collagen-laden scar tissue4–7.\n\nThe initiating factor in the development of arteriogenic ED as a result of radical pelvic surgery is transection of the accessory pudendal arteries. These arteries arise from the peri-prostatic vasculature and course toward the penis and providing a significant portion of the arterial inflow required for normal erectile function8,9.\n\nThe incidence of post-prostatectomy ED has been reported to be between 29-88%10–16. This wide range of values can largely be attributed to failure to control various confounding factors including age, degree of nerve sparing, different definition of potency and preoperative ED. The CaPSURE study revealed that only 20% of patients returned to their preoperative baseline potency levels 1 year after RP12. In the Prostate Cancer Outcome Study (PCOS), 59.9% of men self-reported having ED following RP14. Similar results were found by the Memorial Sloan-Kettering Cancer Center group17–19. According to The Scandinavian Prostate Cancer Group, men who chose watchful waiting had a 45% incidence of ED. Those that selected RP as a treatment option experienced an incidence of ED at 80%20.\n\n\nInflatable penile prosthesis (IPP) utilization after RP\n\nPhosphodiesterase-5 inhibitors (PDE5Is) initially were the first-line treatment for ED secondary to RP. However, they are not uniformly effective. In a subset of patients who initially respond, the response deteriorates over time as progressive cavernosal tissue damage occurs and subsequent venous leak develops18,21,22. Penile implant surgery is a viable treatment option in patients in whom nonsurgical ED treatments are unsatisfactory or are associated with adverse effects23. The utilization rate of penile implants after RP varies from 0.8 to 1.9%24,25. These reported rates have been obtained from analysis of the Surveillance Epidemiology and End Results (SEER) cancer registry database. The higher rate of 1.9% in a study by Stephenson et al.25 is probably due to a younger cohort, as 45% of subjects were younger than 65 years. There are numerous reasons that could be postulated for low utilization of penile implants. Firstly, prostate cancer treatment modalities have improved, thereby decreasing the incidence of ED that is unresponsive to nonsurgical intervention. Secondly, effective nonsurgical treatment modalities have been developed as alternatives to surgical treatment, predominantly PDE5Is. Meta-analysis of contemporary publications by Tal et al. revealed an overall erectile function recovery rate of 58% among men younger than 60 years after RP26. The study acknowledged that the definitions of ED were different in each member study and some men used of PDE5i’s for erectile rigidity26. Nevertheless, Stanley et al. found that there was no significant change in total number of penile implant procedures over a 10-year period before and after the introduction of sildenafil citrate27.\n\n\nSurgical considerations\n\nTechnique of dilation of fibrotic corpora. Corporal crossover and urethral perforation are more likely to occur during dilation of fibrotic corpora28. After placing the stretched penis into anatomical position (urethral meatus pointing in a superior direction), a small dilating instrument is placed into the corporotomy and slowly advanced in a latero-superior direction until it reaches the mid-glans penis. Sequential dilation is needed until the cavernosa accepts a 11-12 Fr. dilator. The most important caveat to remember is to orient the dilating instrument in a latero-superior direction when advancing the dilator within the corporal space28. This will prevent corporal crossover, as well as, provide a visual representation of the location of the dilator within the corpora. Tools used for dilation may include Metzenbaum scissors, Hagar dilators, Brooks dilators, Rossello cavernotomes, Mooreville cavernotomes, and the dialmezinsert dilator29.\n\nTechnique of corporal measurement. After corporotomy, PDS 2-0 stay sutures of are placed at the corporotomy edge. The sutures are used for traction, as well as, for closure of the corporotomy after insertion of the prosthesis. Historically, corporal dilation would then ensue with Brooks or Hegar dilators. After ensuring the corporal space was dilated to 14 mm, the corporal length was then measured and the prosthesis placed. In the contemporary setting, many implanters first measure the length of the corpora with the Furlow30. This narrow device provides enough passive dilation to place an inflatable prosthesis, especially if the corpora are non-fibrotic. If the corpora are fibrotic, the implanter would then dilate the corpora to ensure smooth insertion of the prosthesis31. When dilating or measuring the corpora, it is imperative to direct any instrumentation laterally to avoid urethral injury or corporal crossover32.\n\nTo measure the length of the corpora, gently advance the cylindrical measuring device proximally within the corporal space. When the bottom of the corpora cavernosa is reached, a measurement is recorded. Next, the penis is placed and securely held in “anatomical” position. The measuring instrument is passed distally towards the glans penis while angling the instrument laterally. A distal measurement is recorded and added to the proximal measurement. There should be no more than a 1-cm discrepancy between the right and left corpora. A >1 cm difference suggests incomplete dilation, corporal crossover, crural perforation or urethral perforation.\n\nBoth proximal and distal corporal crossover can happen during dilation, measurement, or cylinder placement33. In addition, the initial correct distal tunneling technique using laterally directed dilators will help avoid crossover28. Side-by-side placement of the Brooks or Hagar dilators in each corpus to check for symmetry and proper positioning is the best way to check for proximal or distal crossover34. If a crossover is detected, the dilator may simply be redirected with the contralateral dilator left in place to prevent repeat crossover28.\n\nProximal crural perforation is suspected when there is asymmetry of proximally positioned dilators or a significant length differential. Gentle dilation and corporal measurement can prevent this manageable complication. In the event of proximal perforation, management may take the form of direct repair of perforation, creation and placement of windsock using an “off-the-shelf” implantable graft, creation of a hammock using a rear-tip-extender, or anchoring of cylinder tubing to tunica35,36. A simple ‘U-type’ suture will prevent proximal cylinder migration.\n\nDistal corporal and urethral perforation requires termination of the procedure, especially if distal perforation occurs during dilation of the first side37. If a second side is perforated after successful cylinder placement of the contralateral side, the single cylinder may be left on the non-perforated side38. Urethral tear may be repaired or, if very small, left to heal over the catheter39. Many surgeons will abandon the case during urethral injury in fear of prosthesis infection.\n\n\nReservoir placement\n\nThe reservoir is normally placed in space of Retzius. This is done to reduce the creation of inguinal floor weakness and to reduce the potential risk of visceral injury. After ensuring complete bladder drainage, the index finger is placed through the IPP incision and advanced to the medial aspect of the external inguinal ring. Using firm pressure, the finger is advanced in a posterior direction, piercing transversalis fascia. If finger pressure is inadequate, the fascia can be perforated with the tip of an instrument (scissors or clamp). This action will create a rent large enough to insinuate an index finger into the space of Retzius. Alternatively, a long-bladed nasal speculum is useful in expanding the retroperitoneal space.\n\nIf the space of Retzius is obliterated due to previous pelvic surgery, ectopic placement of the reservoir should be considered40,41. The reservoir may be placed in the deep to the abdominal musculature superior or posterior to transversalis fascia42. A Foerster or Debakey clamp may be used to advance the deflated reservoir to its ectopic position. Stember et al reported the outcomes of 2687 men who underwent ectopic reservoir placement42. In total, 83%, of men had reservoirs placed posterior to the transversalis fascia. The remainder had reservoirs placed in the anterior transversalis space. No injuries to the bowel or major blood vessels occurred with initial insertion of the reservoir, however two patients experienced bladder injury. Eight patients required reservoir revision secondary to herniation42.\n\nThe most serious intraoperative complications of penile prosthesis insertion occur during reservoir placement43. Traditionally, the reservoir is placed blindly in a retrograde fashion into the space of Retzius through a penoscrotal incision. The serious potential complications include vascular injury, bowel perforation and bladder perforation44.\n\nVascular injury (arterial or venous avulsion) may occur during overly aggressive finger or instrument dilation of the inguinal ring. In the event of brisk bleeding, tapenade with an index finger or sponge stick is advised45. Direct access into the space of Retzius is then accomplished through an inguinal incision. Meticulous inspection of the pelvic sidewall will frequently localize the avulsed venous vessel. In the event of vascular injury of the major pelvic vessels, consultation from a vascular surgeon is recommended.\n\nBladder injury is a complication that should be recognized and managed immediately. Prior pelvic surgery or radiation may result in adherence or fixation to the pelvic sidewall. Bladder perforation can happen while piercing the transversalis fascia46. Emptying the bladder prior to placing the reservoir can decrease the incidence of these injuries. Bladder injury is noted when gross blood is seen in the urine or the observation of urine emanating through the IPP incision31. The injury can be confirmed via flexible cystoscopy or by an on-table cystogram. In case of bladder injury due to scissors, the reservoir should be removed and placed on the contralateral side. The bladder should be drained for 7–10 days. Cystogram should be done prior to catheter removal.\n\nThe bowel may be damaged in a similar mechanism to bladder injury during reservoir placement31. Upon recognition bowel injury (succus entericus in the wound), a general surgeon should be consulted for repair and the prosthesis removed.\n\n\nSatisfaction rates after penile prosthesis implantation\n\nSerial reports regarding of penile prosthesis surgery outcomes demonstrate excellent long-term mechanical reliability of contemporary prosthesis models; satisfaction is superior when compared to PDE5Is and injections. Carson et al. performed a retrospective long-term multicenter study on 372 patients who underwent penile prosthesis implant and focused on the longevity, morbidity and patient satisfaction47. More than 80% of patients were satisfied with the function of the device, the ease of inflation, and level of rigidity. Steege et al. reported that patient satisfaction with semi-rigid prostheses was higher than 90%, however, inflatable devices enjoy a slightly higher satisfaction48. Holloway and Farah reported that the AMS 700 Ultrex prosthesis had a 86% patient and 76% partner satisfaction at a mean postoperative follow-up of 42 months49. In a study by Rajpurkar et al., the investigators demonstrated significantly enhanced erectile function and sexual satisfaction when compared to those receiving sildenafil and intracavernosal prostaglandin50.\n\nThe psychosexual adaptation to penile implant may take up to 6 months. The patients experience a marked enhancement in erectile function with elevation of libido. Apprehension regarding the maintenance of an erection during intercourse is markedly assuaged. In addition to an upsurge in the regularity of sexual activity, a decrease in feeling of sadness, depression, anxiety and an improvement in sexual satisfaction has also been noted51.\n\nTwo major factors contributing to high level of satisfaction are rapid generation of erection and consistently excellent rigidity. Other factors, such as degree of postoperative pain and swelling, postoperative complications, ease of concealment, cosmetic outcome, device function, ease of use and partner acceptance, are critical in determining the patient satisfaction. Potential predictors of patient dissatisfaction with penile prosthesis include Peyronie’s disease, a body mass index >30, or previous RP52.\n\n\nPredictors of requiring a penile prosthesis\n\nThe natural recovery time of erectile function may be up to 24 months after radical pelvic surgery; however, resultant penile rigidity may be maximized by early treatment with intracorporal injection therapy53–55. Various different symptomatic treatments are available for patients who fail to regain a natural erection. Sildenafil becomes effective in the late recovery phase as the nerves recover from intraoperative injury54. At 2 or more years from surgery, the recovery of natural function and improved response from other therapies is unlikely and implantation of penile prosthesis is warranted54. A study by Tal et al. found that men who had surgery as an initial treatment (versus radiotherapy), were of a younger age, were of African/American/Hispanic race, were unmarried, and were living in a geographic region other than the North-east were more likely to utilize penile implants25. Similarly, in a retrospective analysis of claims data from Medicare & Commercial databases of 3928 men undergoing penile prosthesis, Segal and Burnett elucidated the factors with the greatest predictive strength of penile prosthesis implantation, which included a diagnosis of prostate cancer, a diagnosis of diabetes mellitus and previous treatment with first-line ED therapy56.\n\n\nConclusion\n\nIn the population of patients with prostate cancer, the concept of survivorship has become a central tenet of patient care. To that end, quality of sexual life, and especially erectile function has become a significant issue. Penile prosthesis surgery remains an excellent option for restoring erectile function to those who fail more conservative measures. IPP implantation enjoys high satisfaction rates for patients and their partners. Intraoperative complications can be distressing, but with prompt recognition, most of these complications can be navigated with excellent postoperative results.\n\n\nData availability\n\nNo data are associated with this article.",
"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\nNIH Consensus Conference. Impotence. NIH Consensus Development Panel on Impotence. JAMA. 1993; 270(1): 83–90. PubMed Abstract | Publisher Full Text\n\nKüry P, Stoll G, Müller HW: Molecular mechanisms of cellular interactions in peripheral nerve regeneration. Curr Opin Neurol. 2001; 14(5): 635–9. PubMed Abstract | Publisher Full Text\n\nUser HM, Hairston JH, Zelner DJ, et al.: Penile weight and cell subtype specific changes in a post-radical prostatectomy model of erectile dysfunction. J Urol. 2003; 169(3): 1175–9. PubMed Abstract | Publisher Full Text\n\nKlein LT, Miller MI, Buttyan R, et al.: Apoptosis in the rat penis after penile denervation. J Urol. 1997; 158(2): 626–30. PubMed Abstract | Publisher Full Text\n\nYamanaka M, Shirai M, Shiina H, et al.: Loss of anti-apoptotic genes in aging rat crura. J Urol. 2002; 168(5): 2296–300. PubMed Abstract | Publisher Full Text\n\nYao KS, Clayton M, O'Dwyer PJ: Apoptosis in human adenocarcinoma HT29 cells induced by exposure to hypoxia. J Natl Cancer Inst. 1995; 87(2): 117–22. PubMed Abstract\n\nZhou F, Li GY, Gao ZZ, et al.: The TGF-β1/Smad/CTGF pathway and corpus cavernosum fibrous-muscular alterations in rats with streptozotocin-induced diabetes. J Androl. 2012; 33(4): 651–9. PubMed Abstract | Publisher Full Text\n\nAboseif S, Shinohara K, Breza J, et al.: Role of penile vascular injury in erectile dysfunction after radical prostatectomy. Br J Urol. 1994; 73(1): 75–82. PubMed Abstract | Publisher Full Text\n\nDroupy S, Hessel A, Benoît G, et al.: Assessment of the functional role of accessory pudendal arteries in erection by transrectal color Doppler ultrasound. J Urol. 1999; 162(6): 1987–91. PubMed Abstract | Publisher Full Text\n\nCatalona WJ, Bigg SW: Nerve-sparing radical prostatectomy: evaluation of results after 250 patients. J Urol. 1990; 143(3): 538–43; discussion 544. PubMed Abstract | Publisher Full Text\n\nKao TC, Cruess DF, Garner D, et al.: Multicenter patient self-reporting questionnaire on impotence, incontinence and stricture after radical prostatectomy. J Urol. 2000; 163(3): 858–64. PubMed Abstract | Publisher Full Text\n\nLitwin MS, Flanders SC, Pasta DJ, et al.: Sexual function and bother after radical prostatectomy or radiation for prostate cancer: multivariate quality-of-life analysis from CaPSURE. Cancer of the Prostate Strategic Urologic Research Endeavor. Urology. 1999; 54(3): 503–8. PubMed Abstract | Publisher Full Text\n\nSchover LR: Sexual rehabilitation after treatment for prostate cancer. Cancer. 1993; 71(3 Suppl): 1024–30. PubMed Abstract\n\nStanford JL, Feng Z, Hamilton AS, et al.: Urinary and sexual function after radical prostatectomy for clinically localized prostate cancer: the Prostate Cancer Outcomes Study. JAMA. 2000; 283(3): 354–60. PubMed Abstract | Publisher Full Text\n\nWalsh PC: Patient-reported impotence and incontinence after nerve-sparing radical prostatectomy. J Urol. 1998; 159(1): 308–9. PubMed Abstract\n\nWalsh PC, Schlegel PN: Radical pelvic surgery with preservation of sexual function. Ann Surg. 1988; 208(4): 391–400. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMazzola C, Mulhall JP: Penile rehabilitation after prostate cancer treatment: outcomes and practical algorithm. Urol Clin North Am. 2011; 38(2): 105–18. PubMed Abstract | Publisher Full Text\n\nMulhall JP, Slovick R, Hotaling J, et al.: Erectile dysfunction after radical prostatectomy: hemodynamic profiles and their correlation with the recovery of erectile function. J Urol. 2002; 167(3): 1371–5. PubMed Abstract | Publisher Full Text\n\nNelson CJ, Scardino PT, Eastham JA, et al.: Back to baseline: erectile function recovery after radical prostatectomy from the patients' perspective. J Sex Med. 2013; 10(6): 1636–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHolmberg L, Bill-Axelson A, Steineck G, et al.: Results from the Scandinavian Prostate Cancer Group Trial Number 4: a randomized controlled trial of radical prostatectomy versus watchful waiting. J Natl Cancer Inst Monogr. 2012; 2012(45): 230–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOhebshalom M, Parker M, Guhring P, et al.: The efficacy of sildenafil citrate following radiation therapy for prostate cancer: temporal considerations. J Urol. 2005; 174(1): 258–62; discussion 262. PubMed Abstract | Publisher Full Text\n\nTeloken PE, Parker M, Mohideen N, et al.: Predictors of response to sildenafil citrate following radiation therapy for prostate cancer. J Sex Med. 2009; 6(4): 1135–40. PubMed Abstract | Publisher Full Text\n\nMulcahy JJ: Erectile function after radical prostatectomy. Semin Urol Oncol. 2000; 18(1): 71–5. PubMed Abstract\n\nStephenson RA, Mori M, Hsieh YC, et al.: Treatment of erectile dysfunction following therapy for clinically localized prostate cancer: patient reported use and outcomes from the Surveillance, Epidemiology, and End Results Prostate Cancer Outcomes Study. J Urol. 2005; 174(2): 646–50; discussion 650. PubMed Abstract | Publisher Full Text\n\nTal R, Jacks LM, Elkin E, et al.: Penile implant utilization following treatment for prostate cancer: analysis of the SEER-Medicare database. J Sex Med. 2011; 8(6): 1797–804. PubMed Abstract | Publisher Full Text\n\nTal R, Alphs HH, Krebs P, et al.: Erectile function recovery rate after radical prostatectomy: a meta-analysis. J Sex Med. 2009; 6(9): 2538–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStanley GE, Bivalacqua TJ, Hellstrom WJ: Penile prosthetic trends in the era of effective oral erectogenic agents. South Med J. 2000; 93(12): 1153–6. PubMed Abstract\n\nKarpman E: Management of Distal & Proximal Penile Prosthesis Crossover. J Sex Med. 2016; 13(6): 1008–12. PubMed Abstract | Publisher Full Text\n\nBrooks MB: Penile prosthesis: a new device for corpus cavernosal dilation. Urology. 1989; 34(4): 225–6. PubMed Abstract | Publisher Full Text\n\nHenry G, Houghton L, Culkin D, et al.: Comparison of a new length measurement technique for inflatable penile prosthesis implantation to standard techniques: outcomes and patient satisfaction. J Sex Med. 2011; 8(9): 2640–6. PubMed Abstract | Publisher Full Text\n\nMulcahy JJ: Surgical management of penile prosthesis complications. Int J Impot Res. 2000; 12 Suppl 4: S108–11. PubMed Abstract | Publisher Full Text\n\nMoncada I, Martínez-Salamanca JI, Jara J, et al.: Inflatable penile prosthesis implantation without corporeal dilation: a cavernous tissue sparing technique. J Urol. 2010; 183(3): 1123–6. PubMed Abstract | Publisher Full Text\n\nAntonini G, Busetto GM, De Berardinis E, et al.: Minimally invasive infrapubic inflatable penile prosthesis implant for erectile dysfunction: evaluation of efficacy, satisfaction profile and complications. Int J Impot Res. 2016; 28(1): 4–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSharma D, Smith RP: Troubleshooting intraoperative complications of penile prosthesis placement. Transl Androl Urol. 2017; 6(Suppl 5): S892–S97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFishman IJ: Corporeal reconstruction procedures for complicated penile implants. Urol Clin North Am. 1989; 16(1): 73–90. PubMed Abstract\n\nQuinn AD, Das S: Proximal extrusion of semirigid penile prosthesis. Scand J Urol Nephrol. 1989; 23(3): 239–40. PubMed Abstract | Publisher Full Text\n\nArrabal-Polo MÁ, López-Carmona Pintado F, González-Torres S: Transurethral extrusion of penile prosthesis. Urol J. 2013; 10(3): 941. PubMed Abstract | Publisher Full Text\n\nMulcahy JJ: Distal corporoplasty for lateral extrusion of penile prosthesis cylinders. J Urol. 1999; 161(1): 193–5. PubMed Abstract | Publisher Full Text\n\nShaeer O: Management of distal extrusion of penile prosthesis: partial disassembly and tip reinforcement by double breasting or grafting. J Sex Med. 2008; 5(5): 1257–62. PubMed Abstract | Publisher Full Text\n\nKarpman E, Sadeghi-Nejad H, Henry G, et al.: Current opinions on alternative reservoir placement for inflatable penile prosthesis among members of the Sexual Medicine Society of North America. J Sex Med. 2013; 10(8): 2115–20. PubMed Abstract | Publisher Full Text\n\nHenry G, Hsiao W, Karpman E, et al.: A guide for inflatable penile prosthesis reservoir placement: pertinent anatomical measurements of the retropubic space. J Sex Med. 2014; 11(1): 273–8. PubMed Abstract | Publisher Full Text\n\nStember DS, Garber BB, Perito PE: Outcomes of abdominal wall reservoir placement in inflatable penile prosthesis implantation: a safe and efficacious alternative to the space of Retzius. J Sex Med. 2014; 11(2): 605–12. PubMed Abstract | Publisher Full Text\n\nShelling RH, Maxted WC: Major complications of silicone penile prosthesis: predisposing clinical situations. Urology. 1980; 15(2): 131–3. PubMed Abstract | Publisher Full Text\n\nWilson SK, Delk JR 2nd: Prevention and treatment of complications of inflatable penile prosthesis surgery: a review article. Arch Esp Urol. 1996; 49(3): 306–11. PubMed Abstract\n\nKaufman JJ, Lindner A, Raz S: Complications of penile prosthesis surgery for impotence. J Urol. 1982; 128(6): 1192–4. PubMed Abstract | Publisher Full Text\n\nMorgenstern JH, Stein BS, Kendall AR: Complications in patient with triple penile prosthesis. Urology. 1982; 20(5): 530–1. PubMed Abstract | Publisher Full Text\n\nCarson CC, Mulcahy JJ, Govier FE: Efficacy, safety and patient satisfaction outcomes of the AMS 700CX inflatable penile prosthesis: results of a long-term multicenter study. AMS 700CX Study Group. J Urol. 2000; 164(2): 376–80. PubMed Abstract | Publisher Full Text\n\nSteege JF, Stout AL, Carson CC: Patient satisfaction in Scott and Small-Carrion penile implant recipients: a study of 52 patients. Arch Sex Behav. 1986; 15(5): 393–9. PubMed Abstract | Publisher Full Text\n\nHolloway FB, Farah RN: Intermediate term assessment of the reliability, function and patient satisfaction with the AMS700 Ultrex penile prosthesis. J Urol. 1997; 157(5): 1687–91. PubMed Abstract | Publisher Full Text\n\nRajpurkar A, Dhabuwala CB: Comparison of satisfaction rates and erectile function in patients treated with sildenafil, intracavernous prostaglandin E1 and penile implant surgery for erectile dysfunction in urology practice. J Urol. 2003; 170(1): 159–63. PubMed Abstract | Publisher Full Text\n\nTefilli MV, Dubocq F, Rajpurkar A, et al.: Assessment of psychosexual adjustment after insertion of inflatable penile prosthesis. Urology. 1998; 52(6): 1106–12. PubMed Abstract | Publisher Full Text\n\nAkin-Olugbade O, Parker M, Guhring P, et al.: Determinants of patient satisfaction following penile prosthesis surgery. J Sex Med. 2006; 3(4): 743–8. PubMed Abstract | Publisher Full Text\n\nMulhall J, Land S, Parker M, et al.: The use of an erectogenic pharmacotherapy regimen following radical prostatectomy improves recovery of spontaneous erectile function. J Sex Med. 2005; 2(4): 532–40; discussion 540–2. PubMed Abstract | Publisher Full Text\n\nKatz D, Bennett NE, Stasi J, et al.: Chronology of erectile function in patients with early functional erections following radical prostatectomy. J Sex Med. 2010; 7(2 Pt 1): 803–9. PubMed Abstract | Publisher Full Text\n\nMüller A, Parker M, Waters BW, et al.: Penile rehabilitation following radical prostatectomy: predicting success. J Sex Med. 2009; 6(10): 2806–12. PubMed Abstract | Publisher Full Text\n\nSegal RL, Camper SB, Ma L, et al.: Prediction model for penile prosthesis implantation for erectile dysfunction management. Curr Med Res Opin. 2014; 30(10): 2131–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "35266",
"date": "27 Jun 2018",
"name": "Rafael E Carrion",
"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 provide a review manuscript highlighting the role of a specific surgical modality in the realm of prostate cancer survivorship. They initially start by providing a concise review of ED after RP.\nUnder \"Surgical Considerations,\" they begin to discuss specific steps of the penile implantation surgical process. I felt they provided a good review on the separate domains of penile implant surgery. However, I would avoid specific surgeon bias specific steps such as the first sentence under \"Technique of corporal measurement,\" where the author's state, \"After corporotomy, PDS 2-0 stay sutures...\" This is NOT a surgical techniques manuscript, it is a review of the topic at hand.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
},
{
"id": "35273",
"date": "06 Jul 2018",
"name": "Lawrence Jenkins",
"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\nWell put together manuscript reviewing the literature in penile prosthesis placement after radical prostatectomy. I commend the authors on their thorough summary. There are several points I would recommend clarified. 1.\n\nPage 2 paragraph 3 first sentence – “The initiating factor in the development…” This singular cause is debatable. There was a paper by Box et al.1 that showed no difference with preservation of the APAs. I would recommend softening the tone. 2.\n\nPage 2 paragraph 4 – when discussing the potency rates I would recommend stating that those numbers include patients with and without the use of PDE5Is. 3.\n\nPage 2 – surgical considerations – 1st paragraph, last sentence – spell check – Dilamezinsert is the name listed in the Cooper Surgical catalog. 4.\n\nPage 3 – 3rd paragraph, prevention and management of perforation – Please clarify how the ‘U-type’ suture is used to prevent cylinder migration. 5.\n\nPage 3 satisfaction rates after penile prosthesis implantation, first sentence - Check grammar, “of” can be removed as the 4th word.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-770
|
https://f1000research.com/articles/7-759/v1
|
18 Jun 18
|
{
"type": "Research Article",
"title": "A meta-analysis of narrow-band imaging for the diagnosis of primary nasopharyngeal carcinoma",
"authors": [
"David CM Yeung",
"Alexander C Vlantis",
"Eddy WY Wong",
"Michael CF Tong",
"Jason YK Chan",
"David CM Yeung",
"Alexander C Vlantis",
"Eddy WY Wong",
"Michael CF Tong"
],
"abstract": "Background: Narrow band imaging (NBI), an endoscopic technique featuring an augmented definition of microvasculature and mucosal patterns. NBI is increasingly advocated as a tool to characterize neoplasia and intestinal metaplasia in endoscopic standards, such as for colorectal polyps and tumors. Recently NBI has also been studied in the detection of Nasopharyngeal Carcinoma (NPC). Here we aimed to assess the diagnostic utility of NBI for the diagnosis of NPC. Methods: A meta-analysis of studies comparing narrow-band imaging and white light endoscopy in the diagnosis of primary nasopharyngeal carcinoma was performed. The review process involved two independent investigators. The databases used were MEDLINE, PubMed, the Cochrane library, Embase, and the Web of Science. Statistical analysis was performed with OpenMetaAnalyst, MetaDiSc version 1.4, and Medcalc version 17.9.7. Results: Five studies including 2480 patients were included. The sensitivity and specificity for narrow-band imaging were 0.90 (0.73-0.97) and 0.95 (0.81-0.99) respectively. The positive likelihood ratio and negative likelihood ratio were 18.82 (0.31-82.1) and 0.08 (0.02-0.31). For white light endoscopy, the sensitivity and specificity were 0.77 (0.58-0.89) and 0.91 (0.79-0.96). The positive likelihood ratio was 7.61 (3.61-16.04), and the negative likelihood ratio was 0.21 (0.11-0.39). The odds ratio for detection rates between narrow-band imaging and white light endoscopy was 4.29 (0.56-33.03, p = 0.16). Area under the curve for narrow-band imaging was 0.98 (SE: 0.02), and for white light it was 0.93 (SE: 0.03). There was no significant difference in the receiver operating characteristic curves between the two modalities (p = 0.14). Conclusion: Narrow-band imaging showed a higher sensitivity and positive likelihood ratio for the diagnosis of nasopharyngeal carcinoma. However, there was no significant difference in detection rates compared to white light endoscopy. Further investigation with a uniform diagnostic criteria and terminology is needed for narrow-band imaging in the diagnosis of nasopharyngeal carcinoma.",
"keywords": [
"Nasopharyngeal carcinoma",
"narrow-band imaging",
"endoscopy",
"meta-analysis"
],
"content": "Introduction\n\nNasopharyngeal carcinoma (NPC) is a common head and neck cancer in the southeast Asia1. The age-standardized incidence rate in Hong Kong is 12.6 per 100,000 for males and 3.9 per 100,000 for females2. The current standard for NPC diagnosis is histological from a white light endoscopy (WL) directed biopsy3. Large tumors are easy to identify. Early and small tumors might be impossible to differentiate from adenoidal tissue or normal nasopharyngeal mucosa4.\n\nNarrow-band imaging (NBI) is an imaging technique that uses two specific wavelengths of light that are strongly absorbed by hemoglobin, allowing improved visualization and delineation of mucosal microvascular patterns5. This technique, which has been used for the detection of adenomas in the gastrointestinal tract, has the potential to reduce the false negative rates associated with conventional white light endoscopy6. If the sensitivity of abnormal vasculature with the assumed overlying mucosal malignancy seen on NBI was able to surpass that of abnormal morphology of the nasopharynx seen on WL, the false negative findings would be reduced and unnecessary biopsies and their potential complications avoided7.\n\nNBI has been described in the early detection of other head and neck cancers, including squamous cell carcinomas (SCC) of the larynx, floor of mouth8, oropharynx, and hypopharynx9. Among these studies, the finding of brownish spots was the most common descriptive morphology followed by irregular vascular patterns. Similar NBI abnormalities have been adapted to identify primary NPC. The aim of this study was to use a meta-analysis to evaluate the diagnostic utility of NBI compared to conventional WL for the detection and diagnosis of NPC.\n\n\nMethods\n\nWe included all prospective studies detecting NPC by using NBI compared with standard WL. Excluded studies were reviews, data reported only as abstracts, non-diagnostic studies, those that did not include histological confirmation or extractable raw data, and retrospective studies. The publications, their relevance, and eligibility were determined independently by DCMY and JYKC. Application of the inclusion and exclusion criteria was undertaken independently by both reviewers, and any difference of opinion was resolved by discussion between the reviewers. Data extraction was done by DCMY and JYKC. Included studies were assessed for quality. The PRISMA diagram is shown in Figure 1. The study was exempt from Institutional Review Board approval as no patient identifiable data was utilized.\n\nMEDLINE, PubMed, the Cochrane library, Embase, and the Web of Science were searched to identify studies in which narrow band imaging endoscopy was used to look for nasopharyngeal carcinoma compared with white light endoscopy. We used the search terms ‘narrow band imaging,’ ‘narrow band imaging vs white light imaging,’ and ‘nasopharyngeal carcinoma’. As an example, for MEDLINE, we searched the terms “Narrow Band Imaging” and “Nasopharyngeal Neoplasms” separately. We subsequently combined them as an “AND” search, yielding six articles for that specific database. We only included prospective trials of NBI versus standard WL. Only articles in English were included. Reviewers were not blinded to the names of authors, institutions, or journals. The reference lists of these articles were searched for additional relevant articles.\n\nA DerSimonian-Laird diagnostic random effects model was adopted for statistical analysis of sensitivity, specificity, positive and negative likelihood ratios for NBI and WL respectively. Detection rates, defined by true positives divided by sample size, were analyzed and compared between NBI and WL using a binary random effects model. Receiver operating characteristic (ROC) curves were constructed and compared with the Hanley and McNeil approach. Funnel plots were not constructed as the relatively small number of primary studies available for this meta-analysis would make it difficult to interpret10. Statistical analysis was performed with OpenMetaAnalyst version 12.11.14; ROC curves and meta-regression were performed using MetaDiSc version 1.4; ROC curve comparison analysis was performed with Medcalc version 17.9.7.\n\n\nResults\n\nA total of 2480 patients, 61% male and 39% female, were included in our meta-analysis. The mean patient age was 49.5 years. No range was calculated for age and sex as not all studies had included them. Basic demographics are listed in Table 1. The indications for nasoendoscopy in the studies are shown in Table 2. Details of endoscopic examination specifics of the included studies are listed in Table 2. A total of 191 patients were diagnosed with NPC. NBI and WL successfully detected 191 and 163 of these cases respectively.\n\nNBI – Narrow-band imaging\n\nThe pooled sensitivity and specificity for NBI was 0.90 (0.73–0.97) and 0.95 (0.81–0.99) respectively as shown in Figure 2. The ROC curve is shown in Figure 4 and has a calculated area under the curve (AUC) of 0.98 (SE: 0.02). The pooled positive likelihood ratio and negative likelihood ratio was 18.82 (4.31–82.06) and 0.08 (0.02–0.31). The pooled diagnostic odds ratio for NBI was 200.13 (32.56–1230.33, p < 0.001) with tau^2 3.34, Q(df=4) 23.90, hetergeneity p-value < 0.001, and I^2 being 83.26.\n\nArea under the curve and standard error was calculated.\n\nFor WL, the pooled sensitivity and specificity was 0.77 (0.58–0.89) and 0.91 (0.79–0.96) as shown in Figure 3 respectively. The ROC curve is shown in Figure 5, and the AUC calculated as 0.93 (SE: 0.03) The pooled positive likelihood ratio is 7.61 (3.61–16.04) and the negative likelihood ratio is 0.21 (0.11–0.39). The pooled diagnostic odds ratio is 34.00 (15.58–74.21, p < 0.001) for WL, with tau^2 0.45, Q(df=4) 9.67, hetergeneity p-value: 0.046, I^2 being 58.63. A summary of pooled statistics and analyses is depicted in Table 3.\n\nArea under the curve and standard error was calculated.\n\nNBI - Narrow-band imaging. WL - white light endoscopy. AUC – area under the curve. SE – standard error.\n\nFor heterogeneity analysis, meta-regression was performed to identify the source of heterogeneity for the following factors: number of patients, percentage of males or females and mean age. However, none of them accounted for heterogeneity in either group.\n\nIn the analysis of detection rates between NBI and WL, the odds ratio was 4.29 (0.56–33.03, p = 0.16), the Tau^2 was 4.35, Q(df=4) was 25.39, heterogeneity p value was <0.001, and I^2 was 84.24. There was no significant difference between detection rates of NBI and WL. Comparing the ROC curve between and NBI and WL, there was no significant difference (p = 0.14).\n\n\nDiscussion\n\nIn this meta-analysis comparing NBI to WL for the detection and diagnosis of primary nasopharyngeal carcinoma, our study found that NBI had a higher specificity, sensitivity, and positive likelihood ratio. However contrary to previous studies, there were no significant differences between NBI and WL for sensitivity analyses and detection rates. Both tests had similar accuracies as indicated by an AUC approaching the value of 1. This likely reflects the fact that WL is an established examination to evaluate the nasopharynx, and that there are no significant advantages of using current otolaryngological NBI systems to detect NPC, perhaps also indicative of the lack of magnification that is available with larger diameter gastrointestinal endoscopes but not with the smaller nasopharyngeal endoscopes.\n\nEarly detection of NPC is important given the differences in treatment regimens and prognoses for early versus late NPC. Modalities useful in the screening, diagnosis and staging of primary NPC that supplement nasoendoscopy including MRI, CT, PET-CT, and plasma Epstein-Barr virus (EBV) DNA11. However, one or more of these may not always be readily available, may be time consuming, and may be costly in the routine diagnosis of NPC. Plasma EBV DNA has recently been shown to be a highly sensitive and specific screening tool for NPC1, but again the technology to assess plasma EBV DNA has not been standardized to make this a definitively useful investigation. For these reasons, NBI has the potential to be useful by improving the endoscopic detection of primary NPC.\n\nEndoscopes used in the examination of the nasopharynx are usually 4mm in diameter, unlike gastrointestinal endoscopes which are 9 to 12mm in diameter. As current NBI endoscopes are distal sensing endoscopes, the smaller diameter limits the size of the distal sensing chip at the tip of the endoscope, thus limiting the pixel density and resolution and thus the ability to detect smaller lesions12. The endoscopes used in this study might not have had sufficient magnification to observe the microvascular patterns of the nasopharynx in sufficient detail when compared to gastrointestinal endoscopy13. With the advance of ultra-high definition distal chips now offering a resolution of up to 4k, and with 8k resolution under development14, the utility of NBI in the detection and diagnosis of primary NPC may improve significantly.\n\nA further potential issue with NBI being used as a screening tool for the detection of NPC is that NBI endoscopy requires specific training and there is a learning curve. NBI images are initially exceptionally difficult to interpret, and without uniform diagnostic criteria, are not particularly helpful. The interpretation of abnormal features such as vascular tufts or tortuous vessels could theoretically affect accuracy. One concern is that NBI might lead to an increased number of unjustified biopsies due to false positive findings of NBI abnormalities15. NBI was however shown to have a high specificity of 0.95 in our study. This could be either due to the fact that the endoscopists included in this study were already well trained and experienced, or that the learning curve was less of a problem than was postulated.\n\nMost of the papers included in this study primarily focused on what they termed brownish spots as the predominant NBI detected abnormality, which was felt to represent a macroscopic focal increase in subepithelial microvascular architectural density16–18. Terms of vascular patterns such as vessel tortuosity, dilation, and irregularity followed. The utilization of other mucosal surface structural abnormalities in the epithelial layer was only mentioned in one study which included light crests and side morphological differences detected by NBI19. In our colorectal counterparts, a universal NBI magnifying endoscopic classification of colorectal tumors based on objective grounds using a modified Delphi method, followed a proposal by the Japanese NBI Expert Team. They classified abnormal NBI findings into four categories based on the vascular pattern. Mucosal surface patterns were included in this classification: dark or white spots; tubular, branched, and papillary; irregular or obscure; and amorphous areas20. Mucosal surface patterns of oval, tubular, papillary, and destructive were described in histological confirmed gastric carcinomas21. One example of the utilization of mucosa surface structural abnormalities in the head and neck region was in a study of NBI on laryngeal squamous cell carcinoma9. The sensitivity and specificity of NBI was described to be both 0.91 respectively. Mucosal abnormalities detected with NBI were demarcated brownish areas with scattered brown spots in the lesion on the epiglottis. In the nasopharynx, the most common type of epithelial malignancy is a non-keratinizing undifferentiated carcinoma. Although the sensitivity and specificity was 0.90 and 0.95 in our meta-analysis, adopting a uniform epithelial abnormality classification similar to colorectal and upper gastrointestinal diagnostics might be a suitable step in optimizing NBI for the detection of NPC.\n\nFurthermore, solely using vascular patterns to differentiate malignant from benign lesions may be difficult in practice. In a paper investigating the difference between benign basal cell hyperplasia (BCH) and head and neck SCC, BCH was described as having a regular distribution of capillary loops and preserved intervascular transparency compared to SCC. However, no significant differences were detected in the sharpness of the lesion border, nor in the dilatation and tortuosity of the capillary loops22. If every lesion showing dilatation and tortuosity of capillary loops were to be biopsied, it would defeat one aim of NBI and that is to decrease the number of unjustified biopsies.\n\nLimitations of the current analysis include the heterogeneity between primary studies that limit the accuracy of this meta-analysis. These variations include inclusion and exclusion criteria, indications for nasoendoscopy, operator experience, interpretation of endoscopic findings, and diagnostic thresholds. Convenience samples of available examiners and power calculations were not included in any of the studies to calculate the number of examiners needed to detect significant differences. Only studies written in English were included in this meta-analysis. Other languages may offer primary studies with larger sample sizes. Finally, the inclusion of examiners with high baseline detection rates but with little potential to improve may also have limited the effect sizes.\n\n\nConclusion\n\nFor the detection of primary nasopharyngeal carcinoma, narrow-band imaging has not been shown to have significant advantage over white light endoscopy in this meta-analysis, which may be related to the heterogeneity of studies analyzed. Detection may be improved with uniform diagnostic criteria and the inclusion of additional definitions and patterns of mucosal microstructures and submucosal microvascular abnormalities.\n\nThis work was previously presented at the IFOS World Congress of Otorhinolaryngology and Head and Neck Surgery on 28 June 2017, Paris, France.\n\n\nData availability\n\nDataset 1: OpenMetaAnalyst file contain data analysis performed in this study. 10.5256/f1000research.15183.d20697723",
"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\nSupplementary material\n\nSupplementary File 1 – Completed PRISMA checklist.\n\nClick here to access the data.\n\n\nReferences\n\nLin YC, Wang WH, Tsai WC, et al.: Predicting the early invasiveness of nasopharyngeal mucosal neoplasia after radiotherapy by narrow-band imaging: a pilot study. Head Neck. 2013; 35(1): 46–51. PubMed Abstract | Publisher Full Text\n\nCentre for Health Protection: Nasopharyngeal cancer. 2017; January 5, 2018. Reference Source\n\nTay JK, Lim MY, Kanagalingam J: Screening in nasopharyngeal carcinoma: current strategies and future directions. Curr Otorhinolaryngol Rep. 2014; 2(1): 1–7. Publisher Full Text\n\nThong JF, Loke D, Karumathil Sivasankarannair R, et al.: Use of narrow-band imaging in detection of nasopharyngeal carcinoma. J Laryngol Otol. 2013; 127(2): 163–169. PubMed Abstract | Publisher Full Text\n\nPiazza C, Dessouky O, Peretti G, et al.: Narrow-band imaging: a new tool for evaluation of head and neck squamous cell carcinomas. Review of the literature. Acta Otorhinolaryngol Ital. 2008; 28(2): 49–54. PubMed Abstract | Free Full Text\n\nDinesen L, Chua TJ, Kaffes AJ: Meta-analysis of narrow-band imaging versus conventional colonoscopy for adenoma detection. Gastrointest Endosc. 2012; 75(3): 604–611. PubMed Abstract | Publisher Full Text\n\nVlantis AC, Woo JK, Tong MC, et al.: Narrow band imaging endoscopy of the nasopharynx is not more useful than white light endoscopy for suspected nasopharyngeal carcinoma. Eur Arch Otorhinolaryngol. 2016; 273(10): 3363–3369. PubMed Abstract | Publisher Full Text\n\nGono K, Yamazaki K, Doguchi N, et al.: Endoscopic observation of tissue by narrowband illumination. Opt Rev. 2003; 10(4): 211–215. Publisher Full Text\n\nWatanabe A, Taniguchi M, Tsujie H, et al.: The value of narrow band imaging for early detection of laryngeal cancer. Eur Arch Otorhinolaryngol. 2009; 266(7): 1017–1023. PubMed Abstract | Publisher Full Text\n\nSong F, Khan KS, Dinnes J, et al.: Asymmetric funnel plots and publication bias in meta-analyses of diagnostic accuracy. Int J Epidemiol. 2002; 31(1): 88–95. PubMed Abstract | Publisher Full Text\n\nRaab-Traub N: Epstein-Barr virus in the pathogenesis of NPC. Semin Cancer Biol. 2002; 12(6): 431–441. PubMed Abstract | Publisher Full Text\n\nBruno MJ: Magnification endoscopy, high resolution endoscopy, and chromoscopy; towards a better optical diagnosis. Gut. 2003; 52(suppl 4): iv7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNi XG, Wang GQ: The Role of Narrow Band Imaging in Head and Neck Cancers. Curr Oncol Rep. 2016; 18(2): 10. PubMed Abstract | Publisher Full Text\n\nYamashita H, Aoki H, Tanioka K, et al.: Ultra-high definition (8K UHD) endoscope: our first clinical success. Springerplus. 2016; 5(1): 1445. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPiazza C, Del Bon F, Peretti G, et al.: Narrow band imaging in endoscopic evaluation of the larynx. Curr Opin Otolaryngol Head Neck Surg. 2012; 20(6): 472–476. PubMed Abstract | Publisher Full Text\n\nWang WH, Lin YC, Chen WC, et al.: Detection of mucosal recurrent nasopharyngeal carcinomas after radiotherapy with narrow-band imaging endoscopy. Int J Radiat Oncol Biol Phys. 2012; 83(4): 1213–1219. PubMed Abstract | Publisher Full Text\n\nWen YH, Zhu XL, Lei WB, et al.: Narrow-band imaging: a novel screening tool for early nasopharyngeal carcinoma. Arch Otolaryngol Head Neck Surg. 2012; 138(2): 183–188. PubMed Abstract | Publisher Full Text\n\nYang H, Zheng Y, Chen Q, et al.: The diagnostic value of narrow-band imaging for the detection of nasopharyngeal carcinoma. ORL J Otorhinolaryngol Relat Spec. 2012; 74(5): 235–239. PubMed Abstract | Publisher Full Text\n\nWang WH, Lin YC, Lee KF, et al.: Nasopharyngeal carcinoma detected by narrow-band imaging endoscopy. Oral Oncol. 2011; 47(8): 736–741. PubMed Abstract | Publisher Full Text\n\nSano Y, Tanaka S, Kudo SE, et al.: Narrow-band imaging (NBI) magnifying endoscopic classification of colorectal tumors proposed by the Japan NBI Expert Team. Dig Endosc. 2016; 28(5): 526–533. PubMed Abstract | Publisher Full Text\n\nOk KS, Kim GH, Park do Y, et al.: Magnifying Endoscopy with Narrow Band Imaging of Early Gastric Cancer: Correlation with Histopathology and Mucin Phenotype. Gut Liver. 2016; 10(4): 532–541. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYagishita A, Fujii S, Yano T, et al.: Endoscopic findings using narrow-band imaging to distinguish between basal cell hyperplasia and carcinoma of the pharynx. Cancer Sci. 2014; 105(7): 857–861. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYeung DC, Vlantis AC, Wong EW, et al.: Dataset 1 in: A meta-analysis of narrow-band imaging for the diagnosis of primary nasopharyngeal carcinoma. F1000Research. 2018. Data Source"
}
|
[
{
"id": "35284",
"date": "09 Jul 2018",
"name": "Chwee Ming Lim",
"expertise": [
"Reviewer Expertise head and neck cancer",
"onco-immunology",
"robotic surgery"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 comprehensive review/meta-analysis on the topic of NBI in diagnosis of NPC. Several comments that i have are\n1) PRISMA chart - Figure 1 The authors should expand on the reasons of the exclusion of the papers. Having 13 papers excluded after title screen, one paper excluded after data collection without clarification may leave the readers wondering on the nature of the exclusion. 2) The heterogeneity analysis is an inherent weakness of this study which the authors have addressed in the discussion. Having NBI in separate clinical context ( ie screening and if so, in specific high risk cohorts versus general population; and surveillance cohort should be explored). Can a separate robust analysis be performed on the use of NBI in these specific clinical scenarios with the existing dataset? 3) A recent meta-analysis published in Otolaryngology-Head and Neck Surgery on this subject yielded similar conclusions. The authors may want to highlight some similarities or differences in their analysis and approach.\nOverall, an informative article on the subject matter and reviewing the existing evidence of NBI in diagnosis of NPC.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "35278",
"date": "10 Jul 2018",
"name": "Daniel Clayburgh",
"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 manuscript, Yeung et al perform a meta-analysis of narrow band imaging versus white light endoscopy for the diagnosis nasopharyngeal carcinoma.A few areas that could be improved in this report:\nIntroduction Explain more about how narrow band imaging works and detects mucosal change.\nMethods Included studies inclusion criteria are very loosely explained (discussed between DCMY and JYKC), No specific criteria other than prospective studies that compare NBI to WL. Please provide more detail on study protocols that were used. Provide additional detail on imaging characteristics that were concerning for neoplastic change and warranted biopsy.\nResults: Table 1 and 2 could easily be combined in to a \"study characteristics\" table.\n\nDiscussion One confounding factor would be in the indications for imaging and/or biopsy. Mucosal appearance of screening patients and surveillance patients are likely very different aka post-treatment/ radiated mucosa.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-759
|
https://f1000research.com/articles/7-757/v1
|
18 Jun 18
|
{
"type": "Software Tool Article",
"title": "Oncoyeasti: a web-based application to translate data obtained from Saccharomyces cerevisiae high-throughput drug screens into cancer therapeutics",
"authors": [
"Ruby Gupta",
"Samir Cayenne",
"Madhu Dyavaiah",
"Pragnya Srinivas",
"David Otohinoyi",
"Debjyoti Talukdar",
"Moheem Halari",
"Chidambra Halari",
"Ashok Ramani",
"Joshua Yusuf",
"Khushdeep Chahal",
"Rupinder Kaur",
"Ankit Patel",
"Avaniben Patel",
"Ravindrasingh Rajput",
"Harish Siddaiah",
"Shilpadevi Patil",
"Ashish Patil",
"Nikhilesh Anand",
"Ruby Gupta",
"Samir Cayenne",
"Madhu Dyavaiah",
"Pragnya Srinivas",
"David Otohinoyi",
"Debjyoti Talukdar",
"Moheem Halari",
"Chidambra Halari",
"Ashok Ramani",
"Joshua Yusuf",
"Khushdeep Chahal",
"Rupinder Kaur",
"Ankit Patel",
"Avaniben Patel",
"Ravindrasingh Rajput",
"Harish Siddaiah",
"Nikhilesh Anand"
],
"abstract": "The budding yeast (Saccharomyces cerevisiae) gene deletion library consists of a collection of more than 6,000 gene-deletion mutants and is useful for high-throughput screening of anti-cancer drugs. Because of the shorter doubling time and the significant homology the budding yeast\n\nshares with human cells, using a high-throughput chemical screen of budding yeast gene deletion library, one can rapidly identify various genetic targets of anti-cancer drugs. But analyzing the data derived from a yeast library screen to identify corresponding human homologs and their status in various cancers is a cumbersome process. We have developed a web-based app, Oncoyeasti, which enables the researcher to automatically identify the corresponding human homologs of S. cerevisiae and the status of these homologs genes in tumor samples from The Cancer Genome Atlas Database and cell line samples from the Cancer Cell Line Encyclopedia. This would enable the scientists to identify the tumors and choose cell lines to work on and thus serve as an indispensable tool to translate their research into human cancers.",
"keywords": [
"saccharomyces cerevisiae",
"Cancer",
"drug screening",
"toxicity screen",
"human homologs",
"cancer cell lines",
"TCGA",
"CCLE",
"translational research"
],
"content": "Introduction\n\nCancer chemotherapies, the backbone of cancer, treatment have multiple shortcomings including significant toxicities, narrow therapeutic indexes, and the emergence of resistance. Now, a better understanding of cancer pathogenesis and genomics has given rise to new treatment options in terms of targeted therapies1,2. Targeted approaches aim to modulate the molecular pathways that are crucial for tumor growth and maintenance, without causing significant systemic toxicities. This leads to the selective killing of cancer cells with more precision without affecting normal cells, thus generating fewer side effects2. For example, imatinib is commonly used for the treatment of the chronic myeloid leukemia (CML), targeting fusion protein BCR-ABL, and in the treatment of gastrointestinal stromal tumors (GIST) that harbor cKIT and PDGFRA mutations3–5. In addition, there are case reports that have documented the regression of cutaneous squamous cell cancer in patients with CML treated with imatinib6. Another example is rituximab, one of the earliest targeted therapies which are antibodies against cell surface marker CD20 used in patients with leukemia and lymphoma7,8.\n\nAlthough this type of molecular profiling led to tailored therapy for cancer patients, it did not take a long time to discover that it has its own shortcomings. It was found that although these therapeutic agents produce dramatic effects in most patients, some did not respond at all and others showed only a partial response to the treatment1,9,10. One of the mechanisms attributed to the development of resistance is intra-tumor heterogeneity9–11.\n\nDiversity within cells of a single neoplastic lesion is called intra-tumor heterogeneity. This is due to genome instability attributed by the pool of endogenous and exogenous mutations in a given microenvironmental context (selection pressure), which is known to alter the evolutionary trajectory of a tumor leading to clonal and subclonal expansions of cancer cells. This ultimately results in a genetically diversified tumor12–14. For example, in a study of a patient with malignant melanoma who initially responded to the targeted BRAF inhibitor vemurafenib, but later progressed and developed five metastatic lesions, it was found that resistance to vemurafenib was developed via several different, independent mechanisms. BRAFV600E amplification was found in three resistant metastases, which likely occurred as independent events15,16. The fourth drug-resistant metastatic lesion harbored an aberrant form of BRAF, whereas, in yet another metastasis, a new activating in-frame insertion in MEK was identified, thus highlighting the extent of inter- and intra-tumoral heterogeneity and its impact on clinical outcomes17.\n\nStudies have shown that intra-tumoral heterogeneity is the key mechanism underlying tumor progression and the frequent lack of therapeutic responses17. Intra-tumor variability has not only posed a significant challenge in predicting the behavior of the neoplasm, but has also resulted in the emergence of neoplastic clone’s resistance to a given therapy12,17. It is critical to explore novel methodologies to gain a better insight into the biological and therapeutic impact of intra-tumor genetic heterogeneity for the improvement of existing therapies in cancer management.\n\nStudies have shown that two tumors that are histologically similar may behave and respond differently to the same treatment due to being genetically different—a phenomenon that can now be explained by the concept of inter-tumor heterogeneity18. Recent studies that sequenced the genomes of breast and colorectal cancers have shown us that the tumor evolution is a complicated multi-step process caused by a combination of series of mutations intermixed in a vast genomic landscape, confirming tumor heterogeneity19,20. Inter-tumor heterogeneity has significant clinical implications. For example, among patients with colon cancer there has been an attempt to classify the disease with respect to genetic defects. For example, activating mutations of the BRAF oncogene are associated with microsatellite instability and mutations in TP53 or KRAS are associated with chromosomal instability21. This understanding of inter-tumor heterogeneity in colorectal cancers has significant clinical implications, not only for predicting the prognosis of the patient but also for tailoring the targeted therapy to be used for individual tumor type. Approximately 60% of colon cancer patients have KRAS wild-type mutations and are being treated with EGFR inhibitors, but only 30% benefit from the treatment. Non-response to EGFR inhibitors in the remaining 30% of patients has been attributed to other genetic biomarkers, such as BRAF mutations (8–10%), PTEN null mutations (10–15%) and PI3KC mutations (10–12%)18,19. It has been proposed that these patients with different genetic profiles may benefit from therapies targeted towards individual biomarker thus reminding us the clinical and therapeutic implications of inter-tumoral heterogeneity15,16,18.\n\nThe Cancer Genome Atlas (TCGA) is an ever-expanding catalog of genetic mutations identified in tumors in different patients that are responsible for causing cancer. So far it has cataloged and characterized genomic changes including copy number variations, mutations, and RNA expression in around 33 cancer types in more than 45,000 patients, including for 10 rare cancers22,23. TCGA Data Portal enables the researcher to download copy number variations, mutations and mRNA expression profiles of all the genes in a particular type of tumor24,25. However, if the researcher is interested in studying a particular type of genetic alteration across a set of different cancer types and across a set of patients then they will have to use interactive exploratory tools like the cBio Portal to perform data analysis across TCGA Database26.\n\nTCGA also contains the dataset from the Cancer Cell Line Encyclopedia (CCLE) project, which provides information concerning DNA copy number variation, mRNA expression profile and, mutational data of genes in more than 1000 cancer cell lines, providing an indispensable tool for scientists conducting pharmacogenomics and pharmaco-therapeutics studies in oncology. The CCLE has been successfully used in the genetic prediction of genomic profile-based drug responses in the preclinical setting and its incorporation into cancer clinical trial design has helped in the emergence of “personalized” therapeutic regimens27.\n\nUsing the cBioPortal to interactively explore cancer genomics datasets, the researcher can access data from more than 45,000 tumor samples from cancer studies in TCGA database. This enables the scientists to choose cancer cell lines and type of tumor models to pursue their research of targeted cancer therapy26.\n\nThe principal objective of cancer genomics research has been to identify alterations in gene expression, provide information on mutations and copy number variations, the sequences of cancer genes and data on the mutational processes that occur are functioning in tumors, and to develop genetic profile-based targeted therapies against these tumors during cancer evolution. One of the challenges in doing so is finding a method wherein genetic studies can be performed conveniently and rapidly28. Many practical and ethical obstacles severely limit the scope of experiments using human cells in genetic studies. One of the biggest limitations is that the human cell lines are diploid and take a longer time to divide than yeast or bacterial cells, making it difficult to perform manipulation of genes in them. In addition, the human cell lines take a long time to divide, thus limiting the ability to use human cell lines in high-throughput screening of drugs and gene manipulation studies in humans. To overcome this limitation, model organisms have been used, in which most of the fundamental biological mechanisms and pathways that control development and survival have been evolutionarily conserved between species. These organisms were found to have genes that have structural and functional homology with human genes29.\n\nBudding yeast (Saccharomyces cerevisiae) is one of the model organism used to analyze phenomena that involves important basic eukaryotic cell functions, such as metabolism, regulation of the cell cycle, membrane targeting and dynamics, protein folding, and DNA repair. The S. cerevisiae genome was first fully sequenced in 1996, which opened the gateway to various researchers as it provided a treasure of information on genome organization and evolution30.\n\nNow, S. cerevisiae has also been considered a principal model organism for conducting research owing to the ease with which gene manipulation can be done, short doubling time harbored by these cells in and significant genotypic and phenotypic homology shared with human cells. In addition, the complete mapping of the yeast genome and creation of the yeast gene deletion collection (library of deletion mutants lacking all non-essential genes) has resulted in a shift in research focus from individual genes to a more global view of genetic networks. This has resulted in the development of high-throughput screens to identify multiple genetic targets of a drug using the S. cerevisiae gene deletion library (Figure 1)31. Data from these screens helps researchers to identify novel drug candidates\n\nThe clones with gene deletions that lead to drug sensitivity will not grow or grow slowly compared to the wild-type or non-sensitive clones after 3 days of incubation and will not form colonies. The YPD-agar plates are scanned to identify genes deletion leading to sensitivity for a given anti-cancer drug tested.\n\nHigh-throughput screening using the yeast gene deletion library enables researchers to study small molecules in chemical libraries that induce a phenotype in the clones that are engineered to reflect specific genetic changes in human cancers or obtain essential information for further drug design31,32. This eukaryotic microorganism shares a number of fundamental cellular and molecular properties with human cells. This homology has been useful in the study of gene structure and function in S. cerevisiae and the findings have been applied to human biology30,33. Although the entire genetic concentration in a yeast packs into 16 chromosomes, containing only 10% of the DNA of human chromosomes, the packaging of the genetic material in this micro-organism is much more compressed and dense than the human genome, making it fit for use in studies related to human biology34.\n\nOnce a researcher identifies the sensitive S. cerevisiae target genes from the high-throughput drug screen, bioinformatics tools can be used to identify corresponding human homologs and then browse through TCGA using gateways like cBioPortal to identify the status of these homologs in different tumors and cancer cell lines. This process is time-consuming and there is an increased chance of human errors, especially when a researcher gets a list of many genes from the screen.\n\nIn our study, we have developed a web-based application, Oncoyeasti, that matches genes S. cerevisiae genes with corresponding human homologs and browses through the TCGA and cBioPortal database to show the status of these matched human homolog genes in various cancers. Our web app not only saves time and resources the researcher needs to commit in order to analyze the data obtained from the drug screen, it also reduces chances of human errors because the entire data-analysis is streamlined.\n\n\nMethods\n\nThe main objective of Oncoyeasti is to enable the oncologist to use S. cerevisiae as a screening tool to rapidly identify genetic targets of anti-cancer drugs and to establish the status of the human homologs of the identified targets in the tumor database of TCGA. It is a cumbersome task to translate the results obtained from a S. cerevisiae high-throughput drug screen into application-oriented cancer research into cancerous tumors and cancerous cell lines, especially when the drug screen results in the identification of multiple genetic targets. We have developed Oncoyeasti to serve as a platform that analyzes the data obtained from the yeast screen, to identify the corresponding human homologs and status of these homologs in different cancers and cancer cell lines data deposited in TCGA (Figure 2). Oncoyeasti has been programmed to browse TCGA database by automatically using the interactive data exploration and visualization.\n\nOnce the researcher input the Saccharomyces cerevisiae genes, Oncoyeasti matches the S. cerevisiae genes with the corresponding human homologs and generates a link for those homologs to the corresponding cBioPortal webpage describing the status of these homologs in tumors sequence data in The Cancer Genome Atlas (TCGA) database and cancer cell line genomic alteration data in the Cancer Cell Line Encyclopedia (CCLE) database.\n\ncBioPortal provides a visual output of the status of human homologs of S. cerevisiae genes in various cancers. At the time of writing, the TCGA database had a genetic profile of more than 45,000 tumor samples and more than a 1000 cancer cell lines. The sample size is expected to continuously increase with time by updating the database with the sequencing of newer tumors.\n\n\nSystem implementation\n\nIn order to have a platform that provides scientist the human humologs of yeast genes for easy anticancer drug screening, we used the server-side scripting language PHP for web development platform along with MYSQL interface for matching the input S. cerevisae genes with the corresponding human homologs. The URL for the cross cancer summary of the matched human homologs in tumors present in TCGA database, and navigated using cBioPortal interface, was generated using the MSQL interface. Using the MSQL interface and cBioPortal the URL for the status of the matched human homologs in the tumor cell lines data present in the CCLE was generated. We used a simple approach to build the algorithm for the search program which has the capability to analyze up to 100 S. cerevisiae genes at once. The algorithm functions as follows. First, it takes the input in form of S. cerevisiae standard or systematic gene names and starts searching the human homologs dataset for a possible match by comparing the input text with S. cerevisiae standard and systematic gene names. Once the matched human homologs are identified, the cBioPortal URL for the status of the corresponding homolog in various cancer datasets is automatically generated. Using this technique, we are able to display multiple search results for corresponding human homolog gene data stored in the database table along with URL links to the cBioPortal/TCGA database. The design of the web based application Oncoyeasti is unique in terms of identification of human homologue S. cerevisiae genes. It uses feature rich, fast JQuery API which works across multitude of browsers. The project features improved page start-up performance with modified properties and methods which can load JavaScript events and data from the server and return the matched element. Apart from that, the w3 container class uses bootstrap grid system that allows us to organize the data in grid system where classes can be combined in a more responsive, flexible and dynamic layouts.\n\nOncoyeasti is implemented in Bluehost, a highly versatile and reliable web hosting company. Their platform can efficiently support large volumes of traffic at remarkable speed. As a result, the time lapse from the dataset search of Oncoyeasti database to the automatic retrieval of corresponding links from the cBioPortal/TCGA database is estimated to be 761 milliseconds, depending on the internet service provider.\n\nThe usability of oncoyeasti is observed to have efficient function among old and new operating system and browsers. Oncoyeasti has a minimal requirement of Internet Explorer 9, Firefox v60, Safari (all versions), Opera v51, Chrome v66, iOS (all versions), and Android (all versions).\n\n\nUse cases\n\nOnce you visit the portal (http://www.oncoyeasti.org/), you will be prompted to enter S. cerevisiae gene names (Figure 3). You can enter up to 100 gene names, with 1 gene per line. The gene names can be either in standard or systematic gene name format. For example, the RAD9 gene has the standard name RAD9 and systematic name YDR217C, you can either enter RAD9 or YDR217C and click submit.\n\nIn this case, RAD9 is the input.\n\nOnce you click submit, Oncoyeasti will process your entry and match your S. cerevisiae gene entry with the corresponding human homolog. In the case of RAD9 it is TP53BP1 (Figure 4). It will also automatically generate a link to http://www.cbioportal.org/ that will provide information of the status of TP53BP1 in more than 45,000 tumor samples, in the TCGA database. This link is provided under the section cross cancer summary of the human homolog in tumors present in TCGA database. In addition, generate a link to check for the status of the TP53BP1 in the cancer cell lines sequence data present in the CCLE database.\n\nIn order to check the status of TP53BP1 in TCGA database, one needs to click on the Cross Cancer Summary Tab. Upon clicking on the corresponding cross cancer summary tab for RAD9, we will reach the cBioPortal page visually depicting the cross-cancer alteration summary for TP53BP1 in tumor profiles present in 147 different cancer studies (Figure 5). The visual depiction of the data is in a histogram form, showing a percentage of various types of chromosomal instabilities likes homozygous deletion indicated by blue, homozygous amplification indicated by red and mutations indicated by green.\n\nThe link directs the researcher to the cBioPortal page for the cross cancer summary of the corresponding human homolog (TP53BP1 in this case), where in one gets a histogram depicting cross-cancer alteration summary for TP53BP1 in approximately 147 studies in The Cancer Genome Atlas database. Image obtained from cBioPortal26,35.\n\nYou can click on the histogram bars for the respective cancer study to get an “oncoprint”, which is the visualization of distinct genomic alterations in different patients in a given study, such as copy number alterations, somatic mutations, and mRNA expression. Individual genes are represented as rows, and individual cases or patients are represented as columns. For example, if we click on the uterine corpus endometrial carcinoma (provisional TCGA database of 242 samples), we will be directed to the oncoprint of TP53BP1 gene status in 242 uterine corpus endometrial carcinoma samples (Figure 6). The tumors with homozygous amplification of the given gene in an oncoprint (TP53BP1 in this case) is indicated by the red bar and homozygous deletion is indicated by the blue bar.\n\nIn an oncoprint, each tumor is indicated by a bar. Red color indicates homozygous amplification for the given gene and blue color indicates homozygous deletion for the gene of interest. Image obtained from cBioPortal26,35.\n\nOne can download the copy number alterations by clicking on the download tab and export it into an excel form to perform a copy number analysis for the gene. Once exported into the excel, -2 indicates homozygous deletion, -1 heterozygous deletion, 0 is diploid, 1 indicates a heterozygous amplification and 2 a homozygous amplification\n\nSimilarly, if the researcher wants to pick a cancer cell line, depending on the yeast phenotypic screen data, he can click link generated by Oncoyeasti under the CCLE column. The link will direct the researcher to cbioportal CCLE section and provide him with an oncoprint for the gene of interest (TP53BP1, a RAD9 homolog in this case). This will enable the researcher to visualize the copy number alterations of the gene of interest in more than 1000 different types of cancer cell lines, perform copy number alteration analysis in excel as described previously and thus select appropriate cancer cells lines depending on their S. cerevisiae genetic screen results.\n\n\nDiscussion\n\nS. cerevisiae is one of the most widely used eukaryotic model organisms36. It has been used to study biological processes including cell cycle, gene regulation, signal transduction, cell cycle, apoptosis, aging and neurodegenerative disorders29,36. Up to 30% of genes implicated in human disease may have orthologs in the yeast37. Out of an estimated 2,271 known disease-associated genes in humans, 526 (~23%) have a close ortholog in the yeast genome. Similarities in the gene sequence and function have made this organism useful in elucidating biological pathways in humans. Yeast genomic methodologies have been applied in drug development studies including drug target identification, identification of targeted genes and pathways of therapeutic drugs38. The genomic methodologies include use of deletion mutants for drug-target identification, cDNA microarray analysis of drug-target action, complete synthetic screens, fitness profiling using barcoded mutants and analysis of cellular responses to external stimuli were developed in yeast38,39. In addition, individual pathways such as cell cycle regulators and histone deacetylase inhibitors have been used in the identification of new anticancer drug targets39. From all these yeast genomic technologies, a list of genes that respond to a given drug will be available. For example, yeast gene deletion library screening to a particular drug, we get a list of genes whose deletion resulted in cell sensitivity or resistant to a particular drug40. In other yeast technologies, such as cDNA microarray analysis of drug-target action and fitness profiling using barcoded mutants, experiments also resulted in a list of genes up-regulated or down-regulated in response to the drug41. The ultimate aim of the yeast model in either drug discovery or the study of pathways is to apply to humans and thus finding the drug responsive yeast genes in a human would be difficult for a large number of genes. Therefore, Oncoyeasti, a web-based application represents a useful tool for researchers who are using yeast phenotypic or genotypic screens to rapidly identify genetic targets of anticancer drugs. It enables them to identify the corresponding human homologs of the S. cerevisiae genes of interest and automatically generates a link to cBioPortal to visualize the genetic alterations of the corresponding homologs in more than 45,000 different tumor sample datasets present in the TCGA database. As shown in Figure 4, Oncoyeasti identifies the sequential human homolog (TP53BP1) of the corresponding S. cerevisiae gene RAD9 and the corresponding human homologs in tumor samples of TCGA database and cancer cell lines. In addition, Oncoyeasti generates a corresponding link to cBioPortal showing histogram depicting cross-cancer alteration summary of TP53BP1 in 147 studies in TCGA and CCLE (Figure 5). By clicking on the bar of CCLE dataset in the histogram, researchers can navigate CCLE datasets to identify the genetic alterations of their genes in interest (TP53BP1 in this case) in the cancer cell lines datasets present in the CCLE database. Thus it can serve as a useful tool for the researchers to predict different cancerous tumors; the data derived from the yeast screen can be used and serve as a framework to select tumors or cancer subtypes in which their research would be useful. It also serves as a tool for researchers to select appropriate cancer cell lines for their further studies and thus translate their research from budding yeast to humans.\n\nOncoyeasti identifies human target genes and the tumors in which these genes are perturbed based on a sequential homology with S. cerevisiae. But homolog variants identified using sequential homology from a model organism like S. cerevisiae are assumed to be well tolerated in humans, but in certain cases may not perform the same function in humans and vice versa, in functional complementation studies42. Hamza et al. showed that approximately 60% of S. cerevisiae null mutant clones can be rescued by functional complementation by corresponding human homolog genes, indicating that utility of the cross-species platform to screen and identify targets based on sequential homology43. Since 60% of budding yeast homologs show functional complementation, the human cancer cell lines with genetic targets homologous to the yeast mutants identified in the chemical screen should behave similarly to the corresponding yeast mutants identified in the high-throughput chemical screen (the chemical used in this case may be a potential therapeutic drug). Hence Oncoyeasti, when used in synchrony with yeast gene deletion drug screen can serve as an indispensable tool for scientists to rapidly identify target tumors and cancer cell lines that are sensitive or resistant to a given drug and translate yeast research into human therapeutics.\n\n\nConclusion\n\nThe ultimate aim of understanding cancer biology using model organisms is to apply this knowledge to human cancer and cell lines of interest for drug discovery. Oncoyeasti serves as a useful tool for scientists who are working with the yeast model to understand cancer gene functions, pathways and drug discovery.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and source code, and no additional source data are required.\n\n\nSoftware availability\n\nOncoyeasti available from: http://www.oncoyeasti.org/\n\nSource code available from: https://github.com/oncoadmn/oncoyeasti.org.\n\nArchived source code as at time of publication: https://doi.org/10.5281/zenodo.125771735.\n\nLicense: MIT License.",
"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\nSpecial thanks to Dr. Stacia R. Engel from the Saccharomyces Genome Database (Department of Genetics, Stanford University) for providing us curated human homolog gene list of Saccharomyces cerevisiae.\n\n\nReferences\n\nAparicio S, Caldas C: The implications of clonal genome evolution for cancer medicine. N Engl J Med. 2013; 368(9): 842–51. PubMed Abstract | Publisher Full Text\n\nSaijo N: Progress in cancer chemotherapy with special stress on molecular-targeted therapy. Jpn J Clin Oncol. 2010; 40(9): 855–62. PubMed Abstract | Publisher Full Text\n\nDin OS, Woll PJ: Treatment of gastrointestinal stromal tumor: focus on imatinib mesylate. Ther Clin Risk Manag. 2008; 4(1): 149–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaleddu A, Pantaleo MA, Nannini M, et al.: The role of mutational analysis of KIT and PDGFRA in gastrointestinal stromal tumors in a clinical setting. J Transl Med. 2011; 9: 75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHochhaus A, Larson RA, Guilhot F, et al.: Long-Term Outcomes of Imatinib Treatment for Chronic Myeloid Leukemia. N Engl J Med. 2017; 376(10): 917–927. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKawakami Y, Nakamura K, Nishibu A, et al.: Regression of cutaneous squamous cell carcinoma in a patient with chronic myeloid leukaemia on imatinib mesylate treatment. Acta Derm Venereol. 2008; 88(2): 185–6. PubMed Abstract | Publisher Full Text\n\nMaury S, Chevret S, Thomas X, et al.: Rituximab in B-Lineage Adult Acute Lymphoblastic Leukemia. N Engl J Med. 2016; 375(11): 1044–53. PubMed Abstract | Publisher Full Text\n\nCheson BD, Leonard JP: Monoclonal antibody therapy for B-cell non-Hodgkin's lymphoma. N Engl J Med. 2008; 359(6): 613–26. PubMed Abstract | Publisher Full Text\n\nRivera G, Wakelee HA: Resistance to Therapy. Cancer Treat Res. 2016; 170: 183–202. PubMed Abstract | Publisher Full Text\n\nSuda K, Mitsudomi T: Successes and limitations of targeted cancer therapy in lung cancer. Prog Tumor Res. 2014; 41: 62–77. PubMed Abstract | Publisher Full Text\n\nSun XX, Yu Q: Intra-tumor heterogeneity of cancer cells and its implications for cancer treatment. Acta Pharmacol Sin. 2015; 36(10): 1219–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcGranahan N, Swanton C: Biological and therapeutic impact of intratumor heterogeneity in cancer evolution. Cancer Cell. 2015; 27(1): 15–26. PubMed Abstract | Publisher Full Text\n\nIwasa Y, Michor F: Evolutionary dynamics of intratumor heterogeneity. PLoS One. 2011; 6(3): e17866. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSottoriva A, Spiteri I, Piccirillo SG, et al.: Intratumor heterogeneity in human glioblastoma reflects cancer evolutionary dynamics. Proc Natl Acad Sci U S A. 2013; 110(10): 4009–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrimaldi AM, Simeone E, Festino L, et al.: Novel mechanisms and therapeutic approaches in melanoma: targeting the MAPK pathway. Discov Med. 2015; 19(107): 455–61. PubMed Abstract\n\nHolderfield M, Deuker MM, McCormick F, et al.: Targeting RAF kinases for cancer therapy: BRAF-mutated melanoma and beyond. Nat Rev Cancer. 2014; 14(7): 455–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurrell RA, Swanton C: Tumour heterogeneity and the evolution of polyclonal drug resistance. Mol Oncol. 2014; 8(6): 1095–111. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCusnir M, Cavalcante L: Inter-tumor heterogeneity. Hum Vaccin Immunother. 2012; 8(8): 1143–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Sousa EMF, Vermeulen L, Fessler E, et al.: Cancer heterogeneity--a multifaceted view. EMBO Rep. 2013; 14(8): 686–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPolyak K: Heterogeneity in breast cancer. J Clin Invest. 2011; 121(10): 3786–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarkowitz SD, Bertagnolli MM: Molecular origins of cancer: Molecular basis of colorectal cancer. N Engl J Med. 2009; 361(25): 2449–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Z, Jensen MA, Zenklusen JC: A Practical Guide to The Cancer Genome Atlas (TCGA). Methods Mol Biol. 2016; 1418: 111–41. PubMed Abstract | Publisher Full Text\n\nLee JS: Exploring cancer genomic data from the cancer genome atlas project. BMB Rep. 2016; 49(11): 607–611. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCain J: The cancer genome atlas: new weapon in old war? Biotechnol Healthc. 2006; 3(2): 46–51B. PubMed Abstract | Free Full Text\n\nCancer Genome Atlas Research Network, Weinstein JN, Collisson EA, et al.: The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. 2013; 45(10): 1113–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGao J, Aksoy BA, Dogrusoz U, et al.: Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013; 6(269): pl1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarretina J, Caponigro G, Stransky N, et al.: The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature. 2012; 483(7391): 603–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHunter P: The paradox of model organisms. The use of model organisms in research will continue despite their shortcomings. EMBO Rep. 2008; 9(8): 717–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuina AA, Miller ME, Keeney JB: Budding yeast for budding geneticists: a primer on the Saccharomyces cerevisiae model system. Genetics. 2014; 197(1): 33–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEngel SR, Dietrich FS, Fisk DG, et al.: The reference genome sequence of Saccharomyces cerevisiae: then and now. G3 (Bethesda). 2014; 4(3): 389–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang S, O'Shea EK: A systematic high-throughput screen of a yeast deletion collection for mutants defective in PHO5 regulation. Genetics. 2005; 169(4): 1859–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHammonds TR, Maxwell A, Jenkins JR: Use of a rapid throughput in vivo screen to investigate inhibitors of eukaryotic topoisomerase II enzymes. Antimicrob Agents Chemother. 1998; 42(4): 889–94. PubMed Abstract | Free Full Text\n\nBotstein D, Chervitz SA, Cherry JM: Yeast as a model organism. Science. 1997; 277(5330): 1259–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBotstein D: Why yeast? Hosp Pract (Off Ed). 1991; 26(10): 157–61, 164. PubMed Abstract\n\nCerami E, Gao J, Dogrusoz U, et al.: The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics Data. Cancer Discov. 2012; 2(5): 401–404. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKarathia H, Vilaprinyo E, Sorribas A, et al.: Saccharomyces cerevisiae as a model organism: a comparative study. PLoS One. 2011; 6(2): e16015. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGalao RP, Scheller N, Alves-Rodrigues I, et al.: Saccharomyces cerevisiae: a versatile eukaryotic system in virology. Microb Cell Fact. 2007; 6: 32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBharucha N, Kumar A: Yeast genomics and drug target identification. Comb Chem High Throughput Screen. 2007; 10(8): 618–34. PubMed Abstract | Publisher Full Text\n\nAlamgir M, Erukova V, Jessulat M, et al.: Chemical-genetic profile analysis of five inhibitory compounds in yeast. BMC Chem Biol. 2010; 10: 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiaever G, Nislow C: The yeast deletion collection: a decade of functional genomics. Genetics. 2014; 197(2): 451–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLum PY, Armour CD, Stepaniants SB, et al.: Discovering modes of action for therapeutic compounds using a genome-wide screen of yeast heterozygotes. Cell. 2004; 116(1): 121–37. PubMed Abstract | Publisher Full Text\n\nMarini NJ, Thomas PD, Rine J: The use of orthologous sequences to predict the impact of amino acid substitutions on protein function. PLoS Genet. 2010; 6(5):e1000968. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamza A, Tammpere E, Kofoed M: Complementation of Yeast Genes with Human Genes as an Experimental Platform for Functional Testing of Human Genetic Variants. Genetics. 2015; 201(3): 1263–1274. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "35163",
"date": "19 Jun 2018",
"name": "Ayaz Najafov",
"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 have developed a web tool called Oncoyeasti, which bridges the gap between yeast and human ortholog gene symbols, and allows for easier submission of the queried genes to the TCGA/cBioportal database search. While this software can be useful, the following issues have to be addressed:\n\nMajor points:\nOncoyeasti should be able to retrieve the full list of human homologs that match to the yeast orthologs. For instance, if Sch9 is an ortholog of the Akt family, the output should be Akt1, Akt2, Akt3, rather than Akt3 only.\n\nOncoyeasti does not seem to retrieve some human orthologs accurately. For example, when TOR1 and TOR2 are queried, TOR1 matches to SMG1, while TOR2 matches to mTOR. In reality, both yeast TOR1 and yeast TOR2 should match to human mTOR.\nAdditionally, PKC1, YPK1, and YPK2 all match to Akt3, but they should match to PRKCE and SGK2 instead.\nThus, the Oncoyeasti database needs to be improved for accurate yeast-to-human gene symbol matching. The NCBI’s Homologene database can be employed for this purpose.\n\nIf a certain yeast gene does not have a human ortholog, this should be reported in the output of Oncoyeasti (instead of only reporting those that have a human ortholog and skipping those that don’t have a human ortholog).\n\nMinor points:\nThe article would benefit from grammatical and typographical improvement. This is especially important where such mistakes may lead to misunderstanding of the authors’ message by the readers. Some examples:\na) “the backbone of cancer, treatment” should be “the backbone of cancer treatment”. b) “BRAFV600E” should be “BRAF-V600E” or “BRAFV600E”. c) “BRAFV600E amplification” should be “BRAF-V600E mutation”, since the V600E mutation is not an amplification. d) “containing only 10% of the DNA of human chromosomes” should be rephrased. e) “humologs” should be “homologs”. f) “MSQL” should be either “MySQL” or “mSQL”. g) “feature rich” should be “feature-rich”. h) “cross cancer” should be “cross-cancer”.\n\nThe introduction part of the manuscript should be as concise and as relevant as possible. Since Oncoyeasti does not address the cancer heterogeneity directly, that part of the introduction seems unimportant for the manuscript.\n\nThe presentation of the output can be improved. For instance, the following sentence occupies a large area in the output, but the actual link-containing columns are squeezed to the right end of the page: “THE NEXT COLUMNS CONTAINS ONCOPRINT OF TUMOR GROUPS PRESENT IN CROSS CANCER SUMMARY STUDIES OF THE PATIENT TUMOR SETS PRESENT IN TCGA DATABASE”.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? No",
"responses": [
{
"c_id": "3791",
"date": "03 Jul 2018",
"name": "Ashish Patil",
"role": "Author Response",
"response": "The following were the major points Major points which had raised: 1) Oncoyeasti should be able to retrieve the full list of human homologs that match to the yeast orthologs. For instance, if Sch9 is an ortholog of the Akt family, the output should be Akt1, Akt2, Akt3, rather than Akt3 only. 2) Oncoyeasti does not seem to retrieve some human orthologs accurately. For example, when TOR1 and TOR2 are queried, TOR1 matches to SMG1, while TOR2 matches to mTOR. In reality, both yeast TOR1 and yeast TOR2 should match to human mTOR. Additionally, PKC1, YPK1, and YPK2 all match to Akt3, but they should match to PRKCE and SGK2 instead. Thus, the Oncoyeasti database needs to be improved for accurate yeast-to-human gene symbol matching. The NCBI’s Homologene database can be employed for this purpose. Response: Initially we had used a homolog gene list which was provided to us, by www.yeastgenome.org (Stanford database), but we did not know that this list was not complete and was missing significant number of human homologs. Thank you for bringing this to our notice. Now we used http://useast.ensembl.org/biomart/martview/ba508c9593727fcc0111904d444c5296 and generated the list for human homologs for S Cerevisiae. This is the list we generated http://useast.ensembl.org/biomart/martresults/148?file=martquery_0621224014_815.xls.gz Then we compared ensemble list with our stanford database list, to do corrections and also to add to our database, the additional s cerevisiae genes and their human homologs, which were missing in our list, that we got from yeast genome.org. Using ensemble biomart, we have added 4157 additional human homologs to our existing list which we got from yeastgenome.org and thus making oncoyeasti much more comprehensive."
}
]
},
{
"id": "35158",
"date": "28 Jun 2018",
"name": "Alvaro Galli",
"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 web-based application, described in the manuscript, could be relevant in terms of applicability in cancer therapy because has the potentiality to find new targets or new factors affecting drug response.\n\nThe starting point is homology between yeast and human. The authors only showed the RAD9 as example and reported that the human counterpart is the TP53BP1 gene. Where did they find this information? To my knowledge, yeast RAD9 has no human homolog; I have checked in the yeast genome database (https://www.yeastgenome.org/) and I could not find any homology. TP53BP1 appers to be RAD9 homolog only if you find homoly with no filted application ( PANTHER method only giving this result , 1 out of 10 ) I guess that the authors has to state how and where the human homologous to the yeast gene are found and show a little more examples.\nMoreover, the application only shows data from cBioPortal with no statistical evaluation on how it would be relevant to translate study form yeast to human. What I mean is that there is no functional evaluation of the gene, no data about the level of conservation between human and yeast. What I think is that a high level of conservation should correspond a high degree of reliability.\nRecently, Mercatanti et al (FEMS Yeast Yesearch, Dec. 2017)1 published a web tool where yeast strains could be \"humanized\" and evaluated the reliability score to make the functional assay more relevant for cancer risk evaluation.\nI do not think that the manuscript is not acceptable in the present form.\n\nIs the rationale for developing the new software tool clearly explained? No\n\nIs the description of the software tool technically sound? No\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? No\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? No",
"responses": []
},
{
"id": "35343",
"date": "10 Jul 2018",
"name": "Robert J. D. Reid",
"expertise": [
"Reviewer Expertise Genetics",
"Molecular Biology",
"Yeast",
"DNA repair"
],
"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\nGupta et al. report a web-based application, Onco-Yeasti, to identify human homologs of yeast genes and link to information from The Cancer Genome Atlas through cBio portal. While the concept of the software is reasonable, there are serious limitations to the approach. The core function of the web app is to provide links to information in cBIO portal based on a homologous yeast gene. There are significant issues with how the homology with human genes is determined and it is unclear whether this direct linking to cBIO portal is an advantage compared to initiating a query directly in the cBIO portal interface.\n\nThe main problem with the approach taken in Oncoyeasti is that gene orthology between organisms as diverged as yeast and humans is difficult and there does not seem to be a systematic approach to this core function of the software. There is not a single way to accomplish the task of identifying orthologs as evidenced by the multiple algorithms and databases available, such as InParanoid, Orthofinder, Ensemble compara, etc. There are distinctions between orthologs (same gene in a different organism), paralogs (duplicated gene either within or between organisms), and functional orthologs (different gene that plays an orthologous role in another organism). It is not at all clear how these issues are addressed by Oncoyeasti. My understanding from the manuscript is that the homologs are taken from a static list provided by the Saccharomyces Genome Database. If so, then the details of how that list is derived and curated needs to be stated. On the other hand, based on the comments section of F1000, it seems as though homology is now acquired from ENSEMBL. Details of this approach would need to be similarly defined and cited in the manuscript. In any case, there are some strange calls of orthologs. For instance, Oncoyeasti shows that the yeast genes RAS1 and RAS2 have the Human GTPase REM2 as the ortholog. How was this call made? I doubt that REM2 is a better choice of ortholog than H-RAS, K-RAS, N-RAS, RRAS or RRAS2 - especially considering that H-RAS is capable of complementing the inviability of a ras1 ras2 double mutant yeast 1. Nevertheless, there are already good web resources for identifying orthologs. Queries of yeast genes in the Alliance of Genome Resources web site (Stanford) compares multiple orthology methods such as PANTHER, InParanoid, etc and ranks orthologs in multiple species based on an aggregate of the methods. A useful extension of this approach might include functional orthologs as well as evidence of complementation between human and yeast genes 2,3 as part of the scoring metric.\n\nThe second issue with Oncoyeasti is whether the links it generates to cBIO portal are more useful than querying the cBIO portal website directly. Part of this is a design issue, the row and column labels for the the genes and cancer studies scroll off the page and the site becomes difficult to use. More importantly, I am not sure that a direct link to the cBIO portal oncolinks page is the best approach. For instance, the yeast MMS2 gene is correctly identified as human UBE2V2 and following the link for breast invasive carcinoma calls up the cBIO portal oncoprint showing the modifications in the different breast carcinoma studies. What this does not indicate is that there is significant sample overlap in these studies and the resulting value for percent altered does not automatically remove duplicated samples possibly leading to erroneous conclusions. Making the same query in cBIO portal warns the user of overlapping studies up front. This does not happen in Oncoyeasti.\n\nIn summary, since the core function of the software is identifying orthologs, and it does not seem to do this in a systematic way or offer an improvement over existing methods, I cannot approve this manuscript.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? No\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-757
|
https://f1000research.com/articles/7-752/v1
|
18 Jun 18
|
{
"type": "Opinion Article",
"title": "How to incorporate patient and public perspectives into the design and conduct of research",
"authors": [
"Pat Hoddinott",
"Alex Pollock",
"Alicia O'Cathain",
"Isabel Boyer",
"Jane Taylor",
"Chris MacDonald",
"Sandy Oliver",
"Jenny L. Donovan",
"Alex Pollock",
"Alicia O'Cathain",
"Isabel Boyer",
"Jane Taylor",
"Chris MacDonald",
"Sandy Oliver",
"Jenny L. Donovan"
],
"abstract": "International government guidance recommends patient and public involvement (PPI) to improve the relevance and quality of research. PPI is defined as research being carried out ‘with’ or ‘by’ patients and members of the public rather than ‘to’, ‘about’ or ‘for’ them (http://www.invo.org.uk/). Patient involvement is different from collecting data from patients as participants. Ethical considerations also differ. PPI is about patients actively contributing through discussion to decisions about research design, acceptability, relevance, conduct and governance from study conception to dissemination. Occasionally patients lead or do research. The research methods of PPI range from informal discussions to partnership research approaches such as action research, co-production and co-learning. This article discusses how researchers can involve patients when they are applying for research funding and considers some opportunities and pitfalls. It reviews research funder requirements, draws on the literature and our collective experiences as clinicians, patients, academics and members of UK funding panels.",
"keywords": [
"Public and Patient Involvement",
"Public Engagement",
"Qualitative research",
"Research Methods",
"Co-production",
"Partnership approaches"
],
"content": "Introduction\n\nPatient and public involvement (PPI) is recommended from the earliest research stages through to dissemination of the findings1–6. In the UK, INVOLVE3 states that research should be done with and by patients, but what does this mean for researchers and patient partners when starting a study? International resources are available (Box 1) and six UK PPI standards are being tested to see if they work in practice7. Table 1 summarises on-line guidance for research applications to international government funding programmes that endorse involving patients and the public. Language varies internationally and is evolving as patients take a more central role in deciding what research is done and how. Box 2 provides some definitions which are derived from the INVOLVE jargon buster8 and international resources (Table 1). PPI includes patients, potential patients, families, carers, patient groups and members of the public who use or have access to health and social care services3. We refer to this broad group as ‘patients’ to distinguish them from clinicians and academics. This is consistent with Canadian guidance, which defines ‘patients’ as ‘an overarching term inclusive of individuals with personal experience of a health issue and informal caregivers, including family and friends’9. However, ‘patients’ may include people who do not describe themselves in this way. People may self-care for their condition and general public contributions can add value to research questions. Other relevant international terms, for example stakeholder involvement, consumer involvement, knowledge user engagement and patient orientated research are described in Supplementary File 1, Section A.\n\nInternational endorsement of public and patient involvement in research\n\nAustralian Government National Health and Medical Research Council: Consumer and Community Involvement: https://www.nhmrc.gov.au/research/consumer-and-community-involvement\n\nAustralian Government National Health and Medical Research Council: Statement on Consumer and Community Participation in Health and Medical Research (the Statement on Participation): https://www.nhmrc.gov.au/guidelines-publications/r22\n\nCanadian Institutes of Health Research, Strategy for Patient-Oriented Research (SPOR): A coalition dedicated to the integration of research into care: http://www.cihr-irsc.gc.ca/e/41204.html\n\nCochrane Consumer Network. Statement of Principles for Consumer Involvement in Cochrane: http://consumers.cochrane.org/news/statement-principles-consumer-involvement-cochrane\n\nCochrane Training. Involving People learning resource relating to systematic reviews, developed by the ACTIVE (Authors and Consumers Together Impacting on eVidencE) project: http://training.cochrane.org/ACTIVE\n\nEuropean Patient Academy (EUPATI) is a network of European National Platforms which supports the integration of patient involvement across the entire process of medicines research and provides training. This includes the pharmaceutical industry and regulatory agencies. https://www.eupati.eu/\n\nEuropean Health 2020 Strategy calls for civil society engagement to improve health: http://www.euro.who.int/en/publications/abstracts/health-2020-a-european-policy-framework-supporting-action-across-government-and-society-for-health-and-well-being\n\nGordon and Betty Moore Foundation and The American Institutes for Research: A roadmap for patient and family engagement in health and research: http://patientfamilyengagement.org/\n\nHealth Technology Assessment International Patient and Citizen Involvement: www.htai.org/interest-groups/patient-and-citizen-involvement/\n\nNuffield Council on Bioethics (2012) Emerging biotechnologies: technology, choice and the public good: http://nuffieldbioethics.org/project/emerging-biotechnologies/\n\nOttawa Charter for Health Promotion. World Health Organisation 1986: (http://www.who.int/healthpromotion/conferences/previous/ottawa/en/index1.html)\n\nPatient Centered Outcomes Research Institute (PCORI) standards: https://www.pcori.org/research-results/about-our-research/research-methodology\n\nPCORI Engagement Rubric: https://www.pcori.org/sites/default/files/Engagement-Rubric.pdf\n\nThe European Group on Ethics in Science and New Technologies. (2015). Opinion on the ethical implications of new health technologies and citizen participation. Europa: https://ec.europa.eu/research/ege/pdf/opinion-29_ege_executive-summary-recommendations.pdf\n\nUS Department of Health and Human Services: Public Involvement with the National Institutes of Health: https://www.nih.gov/about-nih/what-we-do/get-involved-nih/public-involvement-nih\n\nKey UK-based resources and organisations\n\nHealthtalk.org. Patient and public involvement in research: personal stories of patient involvement in research: http://www.healthtalk.org/peoples-experiences/medical-research/patient-and-public-involvement-research/topics\n\nINVOLVE: supports active public involvement in NHS, public and social care research. Funded by NIHR. http://www.invo.org.uk/. There are useful pages on ‘Budgeting for Involvement Guidance’: http://www.invo.org.uk/wp-content/uploads/2014/11/10002-INVOLVE-Budgeting-Tool-Publication-WEB.pdf and an Involvement Cost Calculator: http://www.invo.org.uk/resource-centre/payment-and-recognition-for-public-involvement/involvement-cost-calculator/\n\ninvoDIRECT is a directory of organisations, networks and groups that support active public involvement in research and helps people to identify activity in their area of interest. http://www.invo.org.uk/communities/invodirect/.\n\ninvoNET is a network of people who are building the evidence knowledge and learning about public involvement in research: http://www.invo.org.uk/communities/invonet/.\n\nJames Lind Alliance: bring patients, carers and clinicians together in Priority Setting Partnerships to identify and prioritise the top uncertainties, or unanswered questions, about the effects of treatments: http://www.jla.nihr.ac.uk/\n\nNational Co-ordinating Centre for Public Engagement (NCCPE) has sections for researchers (and others) to explore, support, plan and do public engagement. It runs training courses and helps Universities to engage with the public: https://www.publicengagement.ac.uk/\n\nNHS Health Research Authority: protects and promotes the interests of patients and the public in health and social care research and has top tips on public involvement in grant applications: https://www.hra.nhs.uk/planning-and-improving-research/research-planning/public-involvement/\n\nNICE’s approach to public involvement in guidance and standards: a practical guide (2015): https://www.nice.org.uk/media/default/About/NICE-Communities/Public-involvement/Public-involvement-programme/PIP-process-guide-apr-2015.pdf\n\nNIHR Going the Extra Mile strategy: https://www.nihr.ac.uk/patients-and-public/documents/Going-the-Extra-Mile.pdf\n\nNIHR Patient and Public involvement in research. https://www.nihr.ac.uk/patients-and-public/\n\nNIHR Public involvement standards development: a project aiming to improve the quality and consistency of public involvement (PI) in research through the development and introduction of national standards: https://sites.google.com/nihr.ac.uk/pi-standards/home\n\nNIHR Research design service: provides support to health and social care researchers across England on all aspects of developing a grant application including, research design, research methods, identifying funding sources and involving patients and the public. Their advice is confidential and free of charge: https://www.nihr.ac.uk/about-us/how-we-are-managed/our-structure/research/research-design-service/\n\nPatients active in research: A website promoting partnership between patient, carers, members of the public and medical researchers, including case studies of patient involvement in research and opportunities to take part in medical research: https://patientsactiveinresearch.org.uk/\n\nPatients included charters: provide entities with a means of demonstrating their commitment to incorporating the experience and insight of patients into their organisations by ensuring that they are neither excluded, nor exploited: https://patientsincluded.org/\n\nPeople in research: helps researchers and research organisations to find patients to work with and advertises opportunities for public involvement in NHS, public health and social care research: https://www.peopleinresearch.org/\n\nResearch Councils UK Concordat for Engaging the Public with Research: https://www.ukri.org/public-engagement/\n\n*Websites for international research guidance, definitions of terms and additional information are provided in Box 1, Box 2 and Supplementary File 1\n\nSome acronyms for involving people in research\n\nPPI –Patient and Public Involvement: http://www.ukcrc.org/patients-and-public/. In the European Research Commission, PPI means Public Procurement of Innovative Solutions.\n\nPPIE – Public Patient Involvement and Engagement: https://www.nihr.ac.uk/about-us/documents/PPIE-Leadership/NIHR-PPIE-Strategy_2018-19.pdf\n\nPIA – Public Involvement Activities17\n\nPCORI – USA Patient-Centered Outcomes Research Institute http://www.pcori.org/program/engagement\n\nNIP – National Involvement Partnership which includes the 4PI – Principles, Purpose, Presence, Process, Impact which are national involvement standards https://www.nsun.org.uk/FAQs/4pi-national-involvement-standards\n\nDefinitions\n\nDefinitions are derived from the INVOLVE jargon buster8 and international resources in Table 1 and Box 1.\n\nParticipating in research describes people who have consented to provide data for analysis to further knowledge (participants). Historically participants were referred to as ‘subjects’ of research. ‘Participatory research approaches’18 is used as an umbrella term which covers ‘participatory action research’19,20, co-design21,22 and co-production of research23,24. In our opinion, a more suitable umbrella term is ‘partnership approaches’.\n\nInvolving. INVOLVE3 defines public involvement in research as research being carried out ‘with’ or ‘by’ members of the public rather than ’to’, ‘about’ or ‘for’ them and states that the term ‘public’ includes:\n\n• Patients, potential patients, carers\n\n• People who use health and social care services\n\n• People from organisations that use services\n\nINVOLVE makes a distinction between the ‘public’ and people who have a professional role in health and social care.\n\nThe European Union (EU) website refers to ‘citizen Involvement’ which includes upstream priority setting, influencing funding decisions to a more direct downstream involvement of citizens and patients in the use and application of medical knowledge and information. It covers both active citizens who engage from a position of agency as well as those unaware of their contribution25, ‘Citizen Science’ is used as an EU umbrella term which is envisioned as various forms of public engagement with science as a way to promote responsible research and innovation.\n\nPartnership is when people who get actively involved in research have a relationship that involves mutual respect and have an equal voice. This contrasts to someone who is consulted occasionally. PCORI consider that the principle is demonstrated when time and contributions of patients and stakeholder partners are valued and demonstrated in fair financial compensation, as well as in reasonable and thoughtful requests for time commitment. When PCORI studies include priority populations, the research team is committed to diversity across all project activities and demonstrates cultural competency, including people with disabilities, when appropriate.\n\nReciprocal Relationships is one of six PCORI engagement principles. They are demonstrated when the roles and decision-making authority of all research partners, including patients, are defined collaboratively and clearly stated.\n\nCollaborating is active, on-going involvement in the research process; however, responsibilities are not equally shared like they are in partnerships. Patients may be co-applicants on a grant application, take part in an advisory group or work with researchers to design, undertake and/or disseminate the results of a research project.\n\nEngaging is a term used in the USA by PCORI and the Canadian Institutes of Health Research. PCORI define engagement as meaningful involvement of patients, carers, citizens, clinicians and other healthcare stakeholders in the topic selection, design, conduct and dissemination of research findings. There are six PCORI patient engagement principles: reciprocal relationships; co-learning, partnerships, transparency, honesty and trust. The UK National Co-ordinating Centre for Public Engagement in Research (https://www.publicengagement.ac.uk/do-it) defines public engagement as: ‘the myriad of ways in which the activity and benefits of higher education and research can be shared with the public. Engagement is by definition a two-way process, involving interaction and listening, with the goal of generating mutual benefit’.\n\nDevolving is to place decision making in the hands of patients or communities, for example, a community development approach26.\n\nConsulting is gaining feedback from patients and communities through e.g. meetings, on-line fora, workshops. The role is considered to be relatively passive when compared to ‘engagement’\n\nAction Research brings about improvement or practical change. A group of people who know about a problem work together in a ‘partnership’ to develop an idea about how it might be resolved. They then go and test this idea. The people who take part in the testing provide feedback on their experiences. It has key tenets20:\n\n- Flexible planning – the detailed content and the direction of the research are not determined at the outset\n\n- Iterative cycles with all involved to i) decide what the problem is, ii) decide an action iii) take action iv) learn the lessons from the action v) reconsider the problem and repeat the cycle\n\n- Subjective meanings of those involved determine the content, direction and measures of success of the research\n\n- The research simultaneously improves the situation\n\n- The unique and ever changing social context is taken into account\n\nCo-production means people who use services, members of the public and professionals working together in a ‘partnership’ to produce research or service improvement. It is an umbrella term for a concept that means coming together to find a shared solution. ‘Co-‘ can be put before specific research tasks like ‘co-design’, ‘co-build’ and ‘co-construct’. Co-production27 covers the whole research process from idea to dissemination of findings in order to change practice.\n\nCo-learning is a term used by PCORI, where the goal is to help patients or other partners to understand research processes. The goal is not to turn patient partners into researchers. PCORI use the term in the context of ‘reciprocal relationships’, where all research partners including patients learn collaboratively. https://www.pcori.org/sites/default/files/Engagement-Rubric.pdf\n\nPPI is put into practice through patients discussing, helping to make decisions and occasionally doing research in order to enhance study relevance, design, conduct and governance. There is no ‘one-size-fits-all’ approach. Flexibility is required to tailor patient involvement to the topic, research question, methods and resources available. This article describes steps that researchers and patient partners can follow when preparing a research funding application (Box 3). We refer throughout the article to an illustrative example of a researcher who wants to do a study to improve outcomes for patients with migraine, and we provide examples from the literature and authors’ experiences.\n\nA clinician wants to involve patients in a trial of treatment for migraine. Here are steps for involving patients when preparing a research funding application.\n\n1. Understand what patient and public involvement (PPI) is and the different approaches\n\ni. refer to research funder guidance about public and patient involvement because it varies internationally and is rapidly evolving\n\nii. understand how patient involvement differs from patients participating in research\n\niii. use language precisely because it varies internationally\n\n2. Find out what research questions are priorities for patients\n\ni. search the internet for existing work on patient priorities and ask patient organisations\n\nii. if patient priorities are unknown, discuss this with your proposed funder and consider how you might fill the gap to progress your research\n\niii. prioritise patient-centered outcome measures and find acceptable research methods\n\n3. Identify patients (not your own), charities and/or patient groups to potentially involve as early as possible\n\ni. consider identifying a professional or lay link worker, perhaps through a charity or a university or hospital patient advisory group\n\n4. Select patients and/or patient groups to be involved in your study\n\ni. consider equity of opportunity, unheard perspectives and health inequalities\n\nii. consider the potential for bias and conflicts of interest\n\n5. Negotiate and agree an approach, tasks and responsibilities at an early stage\n\ni. consider which approach will add value and rigour to your research\n\n6. Negotiate appropriate funding to pay patients, reimburse expenses, fund activities and staff time to facilitate patient involvement\n\n7. Consider whether training will be required for the proposed roles and responsibilities\n\n8. Consider whether patients or patient groups will ‘do’ any research\n\ni. do they have appropriate skills?\n\nii. how will they add value and are there risks?\n\niii. will they be employed?\n\niv. who will mentor and provide supervision?\n\n9. Consider the ethical and research governance implications for involving patients in your study\n\n10. Involve patients in writing the grant application\n\n11. Involve patients to plan future reporting and dissemination of your research\n\n\nSteps for how to involve patients and the public when applying for research funding\n\nAt the outset, it is important to understand the theory underpinning PPI. In depth reviews and discussion of theory are available1,4,10–13 and suggest that, depending on the circumstances, PPI will:\n\n- ensure that the research questions and outcomes really matter to patients\n\n- provide perspectives that complement or challenge those of researchers and clinicians\n\n- make research more relevant to the people whom it is designed to benefit\n\n- ensure that proposed research will be acceptable to patients so that they will be willing to participate\n\n- improve the quality of research\n\n- offer lay knowledge that is either independent for the purpose of governance, or specific to the focus of study to enhance its design or conduct\n\n- make research more equitable and ethical, particularly when publicly funded\n\n- improve dissemination to reach wider lay audiences\n\n- increase the likelihood that research will be implemented into everyday practice and impact on patient care\n\n- enable patients to feel that their voice matters.\n\nAll of the above could reduce research waste14–16 if PPI is put into practice in ways that ensure that research is meaningful, acceptable, ethical and useful.\n\nPatient perspectives can be sought through patient involvement and through patients participating in surveys, interviews or focus groups to provide data for others to analyse, interpret and act on. The authors have observed that in grant applications and study protocols, PPI is often conflated with qualitative research or patient opinion surveys. Collecting data from patients can be important to gain diverse or representative views, but it is different from PPI and both are often needed (Table 2). Discussion with patients at a workshop can seem similar to collecting data in a focus group, because both involve listening to patients’ perspectives, but the context and outcomes from listening differ. PPI means that researchers are in a continuing and reciprocal relationship with patients and make decisions with them about the research. In qualitative research, researchers listen to patients in order to improve their understanding of a topic. Focus group discussions or qualitative interviews are audio-recorded and transcribed. Researchers collate, analyse and interpret text data from carefully sampled patients to produce valid new knowledge and generate hypotheses. Qualitative and survey research have systematic methodological quality standards. However, the researcher holds the power and patients may express strong views which may not be reported. In any research, the PPI and the data collection to gain wider patient perspectives can be separate, combined or overlap in some study phases, or they can be completely integrated throughout (Figure 1). Any combination is possible (Supplementary File 2, Example 1). They are often combined and integrated in equitable partnership research methods like action research, and ‘co-‘ prefixes to research terms, e.g. co-learning and co-production (Box 2).\n\nAction research historically precedes co-production and gathered momentum in the 1940’s as a community-led action in research initiative19,20,30. The UK National Institute of Health Research (NIHR) who fund research advocate co-production27 as a method of involving patients meaningfully from start to finish of the research process. Differences in definitions (Box 2) are subtle, vary internationally and researchers may apply the approaches flexibly in practice. ‘Partnership approaches’ is used in this article as an umbrella term because it acknowledges the changing roles of patients beyond being ‘participants’ or ‘subjects’. Partnership research methods involve patients, clinicians, academics and other relevant stakeholders as equal mutually respected partners in the research team. Being a patient partner implies equal opportunity and equal voice. Equal power in decision-making is sometimes implied, however there are structural and economic power differentials between different types of partner in terms of pay, employment contracts, status and workplace environments. As language is evolving internationally it is more helpful to describe actual patient roles, tasks and responsibilities explicitly rather than use a label for an approach that is open to misinterpretation. For example, co-production27 may mean consulting patients regularly or patients may actively collect and interpret research data. Terms like ‘Participatory Action Research’ confuse because the definition of ‘participation’ in a study means to contribute data, rather than active involvement in research decisions. Partnership research teams decide who has access to participant level data, how to share data securely and how decisions will be made collectively. Partnership approaches can be resource intensive require leadership skills to balance equity of decision-making with a strong scientific rationale. Negotiation skills are required to accommodate different perspectives in order to reach consensus in a timely manner. An important limitation to consider is how the partnership approach is interacting with the intervention: for example action research can become an active intervention component (Supplementary File 2, Example 2).\n\nWhen starting to design a study about migraine, understand how PPI will add value to the research and which uncertainties about patient perspectives might benefit from additional analysis of patient data from a survey or qualitative interviews.\n\nMany funders require researchers to justify that their research question addresses what is important to patients31–33. If a research question is of low priority to the people affected by the condition, or important outcomes are not considered, and/or the intervention in question is considered unacceptable to patients, then further research is wasteful11.\n\nA starting point for researchers is to find out if patients’ priorities already exist for their topic. Many national and international organisations involve patients to identify and publish research priorities specific to a healthcare condition. In the UK, the James Lind Alliance (JLA) specifically identifies and prioritises research questions for funders and there is a register34,35. JLA establish Priority Setting Partnerships which involve collaborations between patients, carers and clinicians. The NIHR funds JLA advisors and the infrastructure, but a Priority Setting Partnership is responsible for its own funding. The JLA has a guidebook which provides step by step processes to identify research uncertainties and prioritise a top 10 list for different conditions36. Researchers are advised to evaluate how priorities were established and the rigour of processes, as priorities can change with time and some groups may not have had an opportunity to be involved.\n\nIf patients’ priorities are unknown, and a Priority Setting Partnership is not available, contacting the potential funder to discuss options may be helpful. When researchers plan bespoke methods to prioritise research, it is important to find patients as soon as possible to identify the topic and refine the research question to ensure relevance. For example: work with a charity or a research organisation to conduct an on-line survey (Supplementary File 2, Example 3); advertise and run open public workshops with patients to rank research priorities; or ask participants in qualitative interviews what would make a difference, then construct research scenarios for them to ‘think aloud’ which one they would prioritise. Once patients have prioritised the research topic and questions, the next step is to prioritise the outcomes that matter, patient-centered outcome measures and identify acceptable research methods.\n\nA first step for a researcher is to search the internet for key organisations and guidelines to see if patient research priorities for migraine are available. If not, a researcher can contact migraine charities and talk to a potential funder to seek their advice.\n\nResearchers are advised to find people to involve and to plan potential roles, responsibilities and tasks for their study as early as possible. Research teams may approach patients through formal patient groups, charities, community groups, University or Health and Social Care patient advisory panels, national directories such as ‘People in Research’37, invoDIRECT38, patients who are involved in producing guidelines like The National Institute of Health and Care Excellence39, or through personal recommendation or advertisement. See Supplementary File 2, Example 4. It is usually not considered appropriate to involve patients that members of the research team are currently providing clinical care to40. In the UK, InvoDIRECT38 provides an A-Z on-line resource of organisations, networks and groups that support PPI in health and social care research (Box 1).\n\nLay or professional coordinators or link people may help and different sources of patients may be used for different purposes. For example, a head office of a patient charity may be invited to nominate a person to join a study steering committee, whereas a local patient group may help to make recruitment materials appealing and easily understood. Participants in a preparatory survey, focus group or qualitative interview may be invited to volunteer for patient involvement in future research. The qualitative research and PPI then become synergistic.\n\nA researcher wanting to study migraine could contact a charity, their University or Health Service patient advisory panel or consult directories of patients who are interested in being involved in research. Invite a patient link worker to join the team who will co-ordinate wider patient involvement.\n\nAs with any appointment, selection criteria for patients based on the research plan are useful to inform decisions. Deciding the number of patients to involve in a study requires careful consideration. Two is the minimum number recommended by INVOLVE3, however international guidance is less specific. The patient characteristics, skills and numbers will vary according to:\n\n- the study design, e.g. several patients with diverse personal experiences of a health condition may be consulted about which outcomes will be measured in a trial41. Co-authors Arthritis Research UK expect patients to be involved in all applications including lab-based early phase research to develop new treatments\n\n- the prevalence of the condition, e.g. it may be challenging to identify two or more patients with rare conditions\n\n- the relevance and reach of a new intervention, e.g. adverts on Facebook for selected postcodes can identify rural and under-privileged urban perspectives\n\n- how much personal tailoring and choice is possible in the design of the research, e.g. two closely involved patients may advise the research team at meetings for a Cochrane Systematic Review, whereas many diverse patient groups may be consulted when prioritising research questions to improve migraine outcomes.\n\nEquity of opportunity for patients to be involved in research underpins UK guidance. The NIHR standards for PPI7 provide practical examples for how researchers can offer inclusive opportunities and sustain respectful, productive relationships. There is a danger that patient contributors are atypical, as the more confident and financially secure are more likely to volunteer. It can be easier to involve older, white and educated people, which can marginalise other perspectives. Health inequalities and equity are important when making research decisions42. Aim to find patients who represent the demographic of those affected by the condition. It can be challenging to access ‘typical’ members of the target population for the specific research question42–44. See Supplementary File 2, Example 5. An adult or child may be selected to represent their own views45 or, when the research involves children, vulnerable patients or patients with cognitive impairment, then a guardian, relative or carer may represent the patient’s views. A lack of resources can hinder recruiting some patients, such as those from ethnic minorities, the less privileged and less literate. Yet this is important because they tend to experience lower health status and poorer access to services. For these patients it can feel intimidating to meet researchers and attend meetings in a University. Alternative strategies include researchers going out into the community in order to build rapport and trust with patients on their own turf, which can then lead to discussions about research (Supplementary File 2, Example 1)46,47. An outreach model for patient involvement via a link coordinator (professional or lay) can help to access less heard perspectives (Supplementary File 2, Example 5)48. A useful guide for getting started and arranging a meeting with patients is available on the INVOLVE website49.\n\nA charity partner might help a researcher to plan how patients on low-income or from ethnic minorities can contribute to a research study on migraine. Adverts, social media and attractive visual information in local newspapers and chemist shops may help.\n\nPPI in research and political lobbying can co-occur and introduce conflicts of interest with the potential to influence research decisions in ways that have been under-researched50. Researchers are advised to consider sources of funding and affiliations of patient contributors, and to re-assess any arising conflicts of interest during their study.\n\nPatients can work with research teams over many years, attend training courses and become a ‘PPI methodologist’ or expert individual or group. This has advantages and risks. Experienced patients can have an overview of a particular health condition that is invaluable. However, becoming embedded in a research team or an organisation can risk losing the ‘eye of the public’51. Researchers are advised to consider whether bias due to ‘group think’52 is possible. This is a risk in any established team, for either researchers or patients to become so familiar with the group or clinical area that they lose sight of fresh perspectives. Selecting new untrained patients for a study can highlight researchers’ preconceptions and assumptions. However, this also has limitations, as it can be difficult for patients to understand, question and challenge researchers when the language and culture are unfamiliar. Patients who have benefited from or experienced adverse events from a particular treatment can introduce bias. Select patients to balance views, for example patients who have positive and negative outcomes from a new procedure or treatment. It may add rigour to include qualitative or survey research to gain diverse and/or representative patient perspectives.\n\nThroughout all stages of a study, researchers and patients make decisions that need to balance and prioritise evidence, personal experiences and competing values held at the individual, family, organisational, political, cultural and environmental levels. Rigour and quality standards for PPI in research are important to counter critics, as there is still some resistance to implementing PPI53.\n\nA researcher is advised to consider conflicts of interest and sources of bias, for example links to industry or private companies. Seek to balance positive and negative patient experiences relevant to the study.\n\nOnce patients are involved, it is advisable to agree clear boundaries about the scope of the role, specific tasks and responsibilities. Some flexibility is desirable to accommodate unexpected issues that can arise in research and there are grey areas. See Supplementary File 2, Example 6. The approach can be bespoke for each study or for each phase within a study12,17,44,48,54 and can vary in the level of patient engagement, responsibility and control. Patients can contribute to three key functions: research decision-making; enhancing understanding of patient experience; and advising how to capture knowledge from other patients. For each function, a question to ask is: which method for involving people will add value and rigour? Example 7 (Supplementary File 2) draws on the work of Gamble and Colleagues who have produced a useful list of tips for patient roles in clinical trials derived from a cohort study of 111 funded trials28.\n\nBe realistic about what will be possible to achieve and the resources required3. A template for Terms of Reference is available on the INVOLVE website49. Terms of Reference acknowledge the importance of mutual respect, practical communication issues and can be reviewed as the research progresses. Researchers may invite patients to propose ground rules for the length of time required to read and respond to emails and comment on documents, for mutual agreement. It is important for researchers to remember that patients may be managing ongoing health conditions which can be unpredictable. Patients value individual constructive and honest feedback about their contributions in order to learn, gain confidence and maintain motivation7.\n\nAt an early stage a researcher is advised to discuss roles and tasks involved in the migraine study. For example: help to design an appealing patient leaflet, recruit patients, attend project management meetings, interpret findings and present them to lay audiences.\n\nInternational arrangements for supporting patient involvement in research vary according to the funding opportunity. It is important for researchers to check current guidance for the funding call they are applying to and budgeting guidance is usually available (Table 1). Negotiate with patients the costs: payment for patient time, any special needs (e.g. childcare, hearing impairment, translation services), training, reimbursement of travel and subsistence expenses. In addition, include costs for staff time to co-ordinate, support, train and facilitate patient involvement. Researchers are advised to spell out to patients the best case and worst case scenarios (e.g. delays to study start and finish), and what contributing to the study would and could involve. Some patients prefer to volunteer, others prefer cash payment or vouchers. Consider patients who are less financially secure. Patients may rely on benefits, part time work or retirement pensions, therefore consider how difficult it is to pay upfront for travel, to scan travel tickets in order to claim research expenses or to have access to computers or printers to access documents for a meeting.\n\nPreparatory PPI activity prior to submitting the grant application can pose a problem for researchers because funding for this is seldom available prior to a grant. Yet this is precisely when patients can have important impact on the study research question, design and plan. In England, the NIHR Research Design Service will provide small amounts of money to cover PPI at the design stage55. Some Universities fund generic patient partnership panels (e.g.56) to work with researchers who are seeking funding and larger charities can often help57.\n\nWhen costing a study about migraine, negotiate sufficient funds to pay for the planned PPI activities, be realistic about the workload and the resources required and consider special needs.\n\nProviding or offering training may or may not be appropriate depending on the patient role and the purpose of training. Training may be desirable in order to undertake highly skilled roles like reviewing grant applications or sitting on independent trial steering committees. In particular, training in the principles of evidence based medicine, with consideration of where and how patient stories fit in evidence hierarchies may be useful. Example 8 (see Supplementary File 2) provides some training programmes that support patient involvement in research. For patients new to a PPI role, support to develop their abilities and confidence to express their views and question researchers may be relevant. Many universities, research funders and charities provide learning and support activities.\n\nThere are many PPI tasks where training is not necessary, where a different perspective is what really matters and patient experience of a healthcare condition is the required expertise. For example, when helping to choose important outcomes or advising on patient information or recruitment strategies, ‘untrained’ patients may make particularly valuable contributions.\n\nTraditionally, academics with qualifications, experience and recognised research skills collect and analyse data. However, increasingly patients are helping to recruit participants, collect or analyse data and some UK grant application forms ask about this (Supplementary File 1, Section B). Such questions arguably prime researchers to think that all boxes should be ticked, without considering the implications. Only appropriately trained patients or lay people should undertake research. Shared experiences of a condition can build trust, empathy and a bond which may help to recruit difficult to engage groups, for example children in care45. However, attention is required to individual expertise, training requirements, supervision and the scientific rigour necessary to execute high quality research. Patients may do research alongside researchers in partnership research methods58 and a paradigm of patient-led research is emerging facilitated by social media and digital technologies59. INVOLVE has a Patient-Led-Research-Hub to support patients who want to pursue their own research ideas38.\n\nIn the UK, any researcher accessing study participants who are NHS patients or staff requires a letter of access, sometimes referred to as a ‘research passport’, obtained from the NHS Research and Development offices (Supplementary File 2, Example 9)60. If patients or lay people help to recruit participants to research, gain informed consent or collect, share or analyse data from individual or group discussions, qualitative research or surveys, then they are ‘doing research' and there are potential governance implications for the sponsor of the research in terms of employment law, ethics, leave entitlement and indemnity. Researchers should not encourage patients to do research because it requires less resource, or because it obviates the need for relatively costly skilled researchers whilst simultaneously bypassing regulatory hurdles. Rather, researchers and patient partners can decide together whether patient researchers are appropriate and beneficial to specific research projects.\n\nResearchers wanting to study migraine may consider the pros and cons of patients doing aspects of the research and the governance issues.\n\nConsider how to work with patients ethically. PPI can be empowering for individuals and communities, but there are tensions and risks, including exploitation25, and the burden and resource implications can be considerable10. Some ethical principles for researchers to consider when involving patients in research include:\n\n- avoiding discrimination, undue persuasion, excessive burden or creating a sense of obligation to be involved in the study\n\n- the distribution of power in research\n\n- valuing patient contributions and fair financial compensation\n\n- conflicts of interest, research integrity and respect for intellectual property\n\n- the confidentiality of data and protecting anonymity of research participants\n\n- advancing science through honest and accurate reporting.\n\nINVOLVE3 states that UK ethics committee approval is not required when patients advise research teams, prioritise research questions, make choices relating to design, share decision-making or disseminate research findings. However, there can be grey areas particularly in relation to defining ‘data collection’. NHS or University Ethics committee approval is required in the UK if personal information, i.e. data as defined in the Data Protection Act61, is collected, shared and stored for future analysis and reporting. For iterative partnership research approaches like co-production, the current ethics committee processes create many challenges62. Researchers can request informed consent from participants to share anonymised data with patient partners, so that they can be involved in analysis and interpretation as members of the study team.\n\nThere are international differences in requirements for research ethical and governance approvals, and particular challenges with digital health research25 which are beyond the remit of this article. New EU General Data Protection Regulation63 commenced in May 2018, and requires transparency about the source of personal data, the purpose and who data will be shared with.\n\nAudio-recording of PPI meetings in order to write accurate but not verbatim minutes, does not require ethics committee approval. However, it does require at least verbal consent from all present at the start of the meeting and the recording should be destroyed as soon as the minutes are agreed. People should receive forewarning of the intention to audio-record, know the purpose, what will happen to the recording and to the content, and be able to object or withdraw. If audio-recordings are stored for longer than is necessary, transcribed verbatim or if there is an intention to report or publish potentially identifiable quotations or content arising from PPI activities, then ethics committee approval is required. Ethics committees have lay committee members, who consider the ethical issues relating to patient involvement.\n\nA researcher wanting to study migraine should consider the ethical issues when involving patients in the design and conduct of their study. Consider patient burden, equity and power, fair and respectful arrangements, confidentiality and the purpose, processes and consequences of any data collected or stored.\n\nPatients sit on research prioritisation committees and funding panels, alongside clinicians and academics, to decide which research is commissioned and which grants are awarded. See Supplementary File 2, Example 10. Many UK funding panels expect to read convincing and meaningful accounts of how patients have had an impact at key stages: preparatory work to inform the planned research; writing the application form particularly the lay summary; and the proposed PPI activity during the study. Expect to be challenged if PPI appears tokenistic. It is important to consider the trade-offs between specifying a plan for PPI in a research protocol and building in some flexibility for change as the research progresses. This may be challenging in countries where regulatory approvals for amending protocols is time consuming.\n\nPatients can help researchers to write the whole grant application in an engaging, easy to understand language. The lay summary is often one of the first sections in a grant application that funding committee members read to gain an overview of the study. Reviewers like to understand exactly what study participants will experience from start to finish. Describe PPI clearly so that the reader understands who, why, how many, how often, what methods and what impact patients have already had on the grant application and will have in contributing to future research decisions. For example, decisions about recruitment methods, intervention delivery or components, which outcomes will be primary or secondary and how to collect data. It helps to use language precisely and to understand how involving, participating, collaborating, consulting and engaging with patients in research differ (Box 2).\n\nA patient helping to write and edit a grant application can make it clear what will happen to patients who participate and how patients will be involved from study conception to dissemination of findings.\n\nPatients can advise on how research might have an impact on health and health care beyond an academic audience. They often have in-depth knowledge of their condition and of on-line sources of information beyond that of academics and clinicians. They can help to write reports, blogs or summaries of findings creatively. See Supplementary File 2, Example 11. Offering participants a lay summary of the research findings is good practice. ‘Patients Included Charters’ provide accreditation for involving patients in conferences and in journal publications64 and GRIPP2 PPI reporting guidelines29,65 are available. Involvement of patients and the public is a critical component in successful implementation of research findings into healthcare, although evidence for best practice is limited3,66,67.\n\nThe grant application for a study about migraine may propose a public event with a charity to present the results of the study. Researchers and patient partners may give joint talks. Small group discussions with migraine patients can suggest ways to spread the news and change care.\n\n\nConclusion\n\nThis article provides a starting point for researchers and patient partners who are planning to seek funding for research. There is no current international consensus on best practice or terminology and guidance is evolving across countries and research disciplines. A crucial distinction when gaining patient perspectives is between patient involvement in research and patients participating by providing data in surveys, qualitative interviews or group discussions. The ethical governance implications differ particularly regarding data protection.\n\nResearchers and patient partners can choose a wide range of different approaches to PPI and each study will require consideration of the optimal approach. Rigour is needed because patients’ lived experience and persuasive narratives can influence important research decisions and the outcomes are not always predictable. Evidence is needed about how different methods of involving people can improve research decisions, healthcare outcomes and impact. A more collaborative and reciprocal partnership approach with patients has the potential to ensure that research undertaken matters to a wider tranche of society and involves those who stand most to benefit from it.\n\n\nKey messages\n\nImportant questions for researchers about including PPI in their research:\n\n• How can I find people in society (patients, patient groups, carers, the taxpaying public, lay organisations) who can make important contributions to research design, conduct and dissemination?\n\n• How will PPI help me to access the perspectives of those who the research potentially will impact on?\n\n• How can different approaches to involving patients as consultants, collaborators or partners improve the relevance, quality, future implementation and sustainability of research?\n\n• How can patients contribute to three key functions: research decision-making; enhance researchers’ understanding of different perspectives; and knowledge capture?\n\n• How can PPI, qualitative research and surveys of patient opinion be optimally combined?\n\n\nData availability\n\nNo data are associated with this article.\n\n\nAuthor information\n\nPH wrote the first draft. All authors have contributed to and approved the final version. PH and IB are members of the NIHR/HTA Commissioning Board and General Board respectively and AOC was a member of the NIHR Programme Grants for Applied Research panel 2007–2017. JLD is an emerita NIHR Senior Investigator and was previously on the NIHR HTA Commissioning Board, NIHR HSR Board and CRUK Population Health Board. JT has lived with rheumatoid arthritis for over 30 years and is also a carer for a brother with schizophrenia. She has been involved in PPI for 7 years covering the whole spectrum from basic science to applied health services research. Her background is in higher education and she works part time for the Open University. CM is the research manager for Arthritis Research UK, a registered charity in England and Wales no. 207711, Scotland no. SC041156. PH and AP have worked as a General Practitioner and as a Physiotherapist respectively. AP is an associate editor with Cochrane Stroke and has received funding from Cochrane Training to synthesise evidence relating to PPI in systematic reviews68. PH has been a chair and deputy chair of a research ethics committee. PH is guarantor and affirms that the manuscript is an honest, accurate, and transparent account of the analysis reported.",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nNo external funding was sought for this work. JLD, PH and AOC are members of the MRC ConDuCT II Hub for Trials Methodology Research (MR/K025643/1). JLD is supported by the NIHR Collaboration for Leadership in Applied Health Research and Care West at University Hospitals Bristol NHS Trust. AOC and PH are undertaking a MRC funded study (MR/N015339/1) identifying and critiquing different methods of complex intervention development. PH and AP work at the Nursing Midwifery and Allied Health Professions Research Unit which receives core funding from the Chief Scientist Office of the Scottish Government Health and Social Care Directorates.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThis paper originated from discussions between the authors and observations from our work. In particular observations of conflation between PPI and qualitative research in grant applications and confusion about how they differ. We would like to acknowledge and thank:\n\n• Peer reviewers of three earlier versions of this paper, and the editors and editorial boards of a leading peer reviewed UK medical journal. Their comments have helped to improve this article. Post-publication peer review of this article provides them with the option of sharing their comments openly.\n\n• Co-authors from two articles on which this article builds, in particular Heather Morgan, Gill Thomson, Nicola Crossland and PPI groups involved in the BIBS study47,69 which combined PPI with qualitative research; and the co-authors of a paper considering how qualitative research contributes to feasibility and pilot studies70.\n\n\nSupplementary material\n\nSupplementary File 1: Section A, Additional terminology and international information relevant for patient and public involvement in research at the funding application stage; Section B, Questions asked by UK National Institute for Health Research (NIHR) on funding application forms.\n\nClick here to access the data.\n\nSupplementary File 2: Examples of different approaches to incorporating patient and public perspectives into research design and conduct.\n\nClick here to access the data.\n\n\nReferences\n\nBoote J, Wong R, Booth A: ‘Talking the talk or walking the walk?’ A bibliometric review of the literature on public involvement in health research published between 1995 and 2009. Health Expect. 2015; 18(1): 44–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCIHR: Canadian Institutes of Health Research (CIHR): Strategy for patient-oriented research (SPOR) - patient engagement. Ottawa, Canada. 2014. Reference Source\n\nINVOLVE: Briefing notes for researchers: involving the public in NHS, public health and social care research. Eastleigh 2012. Reference Source\n\nINVOLVE: Exploring the impact of public involvement on the quality of research: examples. Eastleigh, 2013. Reference Source\n\nNHMRC: National Health and Medical Research Council (NHMRC): A model framework for consumer and community participation in health and medical research. Canberra, Australia. 2005. Reference Source\n\nNIHR: Patient and public involvement in research. Last accessed 25th May 2018. Reference Source\n\nNIHR: Developing national standards for public involvement in research. 2018. Last accessed 25th May 2018. Reference Source\n\nINVOLVE: Jargon buster. 2017. Last accessed 25th May 2018. Reference Source\n\nCIHR: Strategy for patient-oriented research - patient engagement framework. Canadian Institutes of Health Research, 2014. Last accessed 25th May 2018. Reference Source\n\nBrett J, Staniszewska S, Mockford C, et al.: Mapping the impact of patient and public involvement on health and social care research: a systematic review. Health Expect. 2014; 17(5): 637–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChalmers I, Glasziou P: Avoidable waste in the production and reporting of research evidence. Lancet. 2009; 374(9683): 86–89. PubMed Abstract | Publisher Full Text\n\nSmith E, Ross F, Donovan S, et al.: Service user involvement in nursing, midwifery and health visiting research: a review of evidence and practice. Int J Nurs Stud. 2008; 45(2): 298–315. PubMed Abstract | Publisher Full Text\n\nStaley K: Exploring Impact: Public involvement in NHS, public health and social care research. Eastleigh. 2009. Reference Source\n\nChalmers I, Bracken MB, Djulbegovic B, et al.: How to increase value and reduce waste when research priorities are set. Lancet. 2014; 383(9912): 156–65. PubMed Abstract | Publisher Full Text\n\nIoannidis JP, Greenland S, Hlatky MA, et al.: Increasing value and reducing waste in research design, conduct, and analysis. Lancet. 2014; 383(9912): 166–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Shahi Salman R, Beller E, Kagan J, et al.: Increasing value and reducing waste in biomedical research regulation and management. Lancet. 2014; 383(9912): 176–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLander J, Hainz T, Hirschberg I, et al.: Current practice of public involvement activities in biomedical research and innovation: a systematic qualitative review. PLoS One. 2014; 9(12): e113274. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinkler M, Wallerstein N: Community-based participatory research for health: From process to outcomes. 2nd ed. San Francisco, CA: John Wiley & Sons. 2017. Reference Source\n\nMeyer J: Qualitative research in health care. Using qualitative methods in health related action research. BMJ. 2000; 320(7228): 178–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorrison B, Lilford R: How can action research apply to health services? Qual Health Res. 2001; 11(4): 436–49. PubMed Abstract | Publisher Full Text\n\nBate P, Robert G: Experience-based design: from redesigning the system around the patient to co-designing services with the patient. Qual Saf Health Care. 2006; 15(5): 307–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobert G, Cornwell J, Locock L, et al.: Patients and staff as codesigners of healthcare services. BMJ. 2015; 350: g7714. PubMed Abstract | Publisher Full Text\n\nJasanoff S: States of knowledge: the co-production of science and the social order. 3rd ed. London: Routledge, 2006. Reference Source\n\nRycroft-Malone J, Burton CR, Bucknall T, et al.: Collaboration and Co-Production of Knowledge in Healthcare: Opportunities and Challenges. Int J Health Policy Manag. 2016; 5(4): 221–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEuropean Group on Ethics in Science and New Technologies: Opinion on the ethical implications of new health technologies and citizen participation: European Commission. 2015; Last accessed 25th May 2018. Reference Source\n\nMoule P, Davies R: A devolved model for public involvement in the field of mental health research: case study learning. Health Expect. 2016; 19(6): 1302–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNIHR: Guidance on co-producing a research project. National Institute for Health Research, 2018; Last accessed May 2018. Reference Source\n\nGamble C, Dudley L, Allam A, et al.: An evidence base to optimise methods for involving patient and public contributors in clinical trials: a mixed-methods study. Southampton, UK: NIHR Journals Library. 2015; 3(39). PubMed Abstract | Publisher Full Text\n\nStaniszewska S, Brett J, Simera I, et al.: GRIPP2 reporting checklists: tools to improve reporting of patient and public involvement in research. BMJ. 2017; 358: j3453. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMertens DM: Transformative Paradigm. J Mix Methods Res. 2007; 1(3): 212–25. Publisher Full Text\n\nElwyn G, Crowe S, Fenton M, et al.: Identifying and prioritizing uncertainties: patient and clinician engagement in the identification of research questions. J Eval Clin Pract. 2010; 16(3): 627–31. PubMed Abstract | Publisher Full Text\n\nOliver SR: How can health service users contribute to the NHS research and development programme? BMJ. 1995; 310(6990): 1318–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPartridge N, Scadding J: The James Lind Alliance: patients and clinicians should jointly identify their priorities for clinical trials. Lancet. 2004; 364(9449): 1923–4. PubMed Abstract | Publisher Full Text\n\nChalmers I, Atkinson P, Fenton M, et al.: Tackling treatment uncertainties together: the evolution of the James Lind Initiative, 2003–2013. J R Soc Med. 2013; 106(12): 482–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetit-Zeman S, Firkins L, Scadding JW: The James Lind Alliance: tackling research mismatches. Lancet. 2010; 376(9742): 667–69. PubMed Abstract | Publisher Full Text\n\nAlliance JL: JLA Guidebook. In: National Institute for Health Research, ed., 2016. Reference Source\n\nPeople In Research: Opportunities for public involvement in NHS, public health and social care research. Last accessed 25th May 2018. Reference Source\n\nINVOLVE: invoDIRECT. Last accessed 25th May 2018. Reference Source\n\nNICE: NICE’s approach to public involvement in guidance and standards: a practical guide. In: National Institute for Health and Care Excellence, ed., 2015. Reference Source\n\nBMJ: Patient partnership. Last accessed 25th May 2018. Reference Source\n\nCrowe S, Regan S, Daly A: Public engagement in outcomes development - three degrees of separation. Trials. 2015; 16(Suppl 3): O3. Publisher Full Text\n\nBeresford P: Beyond the usual suspects. 2013; Last accessed 25th May 2018. Reference Source\n\nBoote J, Baird W, Beecroft C: Public involvement at the design stage of primary health research: a narrative review of case examples. Health Policy. 2010; 95(1): 10–23. PubMed Abstract | Publisher Full Text\n\nOliver S, Liabo K, Stewart R, et al.: Public involvement in research: making sense of the diversity. J Health Serv Res Policy. 2015; 20(1): 45–51. PubMed Abstract | Publisher Full Text\n\nTaylor J, Bradbury-Jones C, Hunter H, et al.: Young People's Experiences of Going Missing From Care: A Qualitative Investigation using Peer Researchers. Child Abuse Rev. 2014; 23(6): 387–401. Publisher Full Text\n\nPollock A, St George B, Fenton M, et al.: Development of a new model to engage patients and clinicians in setting research priorities. J Health Serv Res Policy. 2014; 19(1): 12–18. PubMed Abstract | Publisher Full Text\n\nMorgan H, Thomson G, Crossland N, et al.: Combining PPI with qualitative research to engage ‘harder-to-reach’ populations: service user groups as co-applicants on a platform study for a trial. Res Involv Engagem. 2016; 2(1): 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilson P, Mathie E, Keenan J, et al.: ReseArch with Patient and Public invOlvement: a RealisT evaluation – the RAPPORT study. Health Services and Delivery Research. Southampton, UK: NIHR Journals Library. 2015; 3(38). PubMed Abstract | Publisher Full Text\n\nINVOLVE: Briefing note eight: Getting started. 2012; Last accessed 25th May 2018. Reference Source\n\nTaylor J, Denegri S: Industry links with patient organisations. BMJ. 2017; 356: j1251. PubMed Abstract | Publisher Full Text\n\nIves J, Damery S, Redwod S: PPI, paradoxes and Plato: who's sailing the ship? J Med Ethics. 2013; 39(3): 181–85. PubMed Abstract | Publisher Full Text\n\nJanis IL: Groupthink: Psychological studies of policy decisions and fiascoes. 1st ed. Boston: Houghton Mifflin, 1982. Reference Source\n\nMadden M, Speed E: Beware Zombies and Unicorns: Toward Critical Patient and Public Involvement in Health Research in a Neoliberal Context. Frontiers in Sociology. 2017; 2: 7. Publisher Full Text\n\nEdelman N, Barron D: Evaluation of public involvement in research: time for a major re-think? J Health Serv Res Policy. 2016; 21(3): 209–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNIHR: Research Design Service. Last accessed 25th May 2018. Reference Source\n\nImperial College London: Patient experience research centre: patient and public involvement. Last accessed 25th May 2018. Reference Source\n\nAndrews LM, Allen H, Sheppard ZA, et al.: More than just ticking a box…how patient and public involvement improved the research design and funding application for a project to evaluate a cycling intervention for hip osteoarthritis. Res Involv Engagem. 2015; 1(1): 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJagosh J, Macaulay AC, Pluye P, et al.: Uncovering the benefits of participatory research: implications of a realist review for health research and practice. Milbank Q. 2012; 90(2): 311–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVayena E, Tasioulas J: Adapting standards: ethical oversight of participant-led health research. PLoS Med. 2013; 10(3): e1001402. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNHS: Health Research Authority. Last accessed 25th May 2018. Reference Source\n\nInformation Commissioner's Office: The Guide to Data Protection.2018; Last accessed 25th May 2018. Reference Source\n\nGoodyear-Smith F, Jackson C, Greenhalgh T: Co-design and implementation research: challenges and solutions for ethics committees. BMC Med Ethics. 2015; 16(1): 78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHealth Research Authority Guidance on the General Data Protection Regulation (GDPR). 2018; Last accessed May 2018. Reference Source\n\nPatients Included: Patients included charters. Last accessed 25th May 2018. Reference Source\n\nNIHR: Reporting public and patient involvement. Last accessed 25th May 2018. Reference Source\n\nEuropean Science Foundation: Implementation of medical research in clinical practice. In: European Science Foundation, ed. France, 2012. Reference Source\n\nKreis J, Puhan MA, Schünemann HJ, et al.: Consumer involvement in systematic reviews of comparative effectiveness research. Health Expect. 2013; 16(4): 323–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPollock A, Campbell P, Struthers C, et al.: Stakeholder involvement in systematic reviews: a protocol for a systematic review of methods, outcomes and effects. Res Involv Engagem. 2017; 3(1): 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan H, Hoddinott P, Thomson G, et al.: Benefits of Incentives for Breastfeeding and Smoking cessation in pregnancy (BIBS): a mixed-methods study to inform trial design. Health Technol Assess. 2015; 19(30): 1–522, vii-viii. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Cathain A, Hoddinott P, Lewin S, et al.: Maximising the impact of qualitative research in feasibility studies for randomised controlled trials: guidance for researchers. Pilot Feasibility Stud. 2015; 1(1): 32. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "35144",
"date": "25 Jun 2018",
"name": "Gary Hickey",
"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 helpful article. A few recommendations that will help improve the accuracy and help avoid any confusion.\nOne step that is missing I think is 'Preparing the research team for PPI'. Support and training are mentioned in relation to the public but this also applies to the research team.\np1. 'The research methods of PPI...' I don't think 'methods' is the right word (eg 'informal discussions' is not a research method). Perhaps 'techniques' or 'The ways in which patients and public are involved in research...'\np3. I think the sentence in the first para beginning 'In the UK, INVOLVE states...' should come at the end of that para.\np3. 'Steps for how etc'. Are these steps? Or are they 'Issues to consider'?\np3. 'Understand what patient and public involvement is'. A more accurate heading would be 'Understand your rationale for patient and public involvement.' And in the next two sentences you mention 'theory'. I don't think these are theories - I think it is about 'understanding the rationale or motivation for patient and public involvement.'\np3 'How does patient involvement differ from patient participation?' The only step that is posed as a question - I'd change to 'Understanding how patient involvement differs etc'. And it would be worth adding in again to this section (I know it's already in elsewhere) INVOLVE's distinction between involvement and participation.\np7 The list of acronyms includes both terminology and organisations. Confusing. Take out the organisations - if they're in the main body of the article then they should be written in full anyway.\np11 'A first step....'. I'd reword to make less like an instruction and consistent with the rest of the article. So 'A researcher could search the internet etc. Another approach would be to contact migraine charities etc.'\np11 Para beginning 'Lay or professional coordinators or link people...' Consider explaining what these terms mean. And replace 'whereas' with 'and'. Lose the sentence 'The qualitative research and PPI become synergistic' - it confuses the point being made in this section and I'm nort sure it's accurate.\np11 'Decide who and how many patients to involve' - we've moved from people to patients. Needs to be consistent throughout. Also add in something about why you might want to consider having more than one person ie a) public can support each other b) helps redress the power balance in the room and c) the public can not always make a meeting and so, if you have more than one person, it reduced the likelihood that the public voice will be absent at any given time.\np11 Replace the sentence 'Aim to find patients who represent the demogaphics etc' with 'Consider the demographic of those affected etc'. Some readers will take the first sentence to the extreme and it may become a barrier to involving people. I would also suggest losing the sentence 'It can be challenging to access 'typical' members etc' - I'm not sure what is meant by 'typical' here. Need to give some consideration here to the issues of 'representativeness' - when you have only one or two people involved in your research it is unlikely that can be 'the' voice of everyone but they can be 'a' voice. If you want something more representative then surveys etc might be a more appropriate answer.\np11 The authors say that 'the more confident and financially secure are more likely to volunteer'. Need to add in something about researchers have struggled to access certain groups. The sentence 'A lack of resources' - not sure that 'less privileged' is a phrase I would use . Perhaps 'less well off' or something similar?\n\np11 Need to be careful with the sentence 'Yet this is important..' - some ethnic minority groups might be offended that you are asserting that they tend' to experience lower health status' . Perhaps 'some groups tend to experience etc'\np12 'Patients can contribute to three key functions etc'. Lose this sentence. I'm not sure that it's true. For example they can also be included in data collection (also applies to the penultimate bullet point in 'key messages' on page 15).\np12 'AT an early stage a researcher is advised to discuss roles and task etc' - I would also add in here behaviours or responsibilities.\np13 'INVOLVE has a Patient-Led-Research-Hub to support patients who want to puruse their own research.' Are you referring to INVODirect? This is more a list of organisations who support active PPI in research.\np13 'Researchers wanting to study migraine etc'. Add on to the end of the sentence 'this entails.' (I read it at first as the patients doing aspects of the governance issues).\np15 Either lose the final bullet point on the key messages - I found this confusing - or add in something like 'to ensure you get the public view'.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "3761",
"date": "28 Jun 2018",
"name": "Pat Hoddinott",
"role": "Author Response",
"response": "Thank you Gary for these very helpful suggestions, in particular the sentences where we could confuse the reader. Pat Hoddinott"
}
]
},
{
"id": "35143",
"date": "02 Jul 2018",
"name": "Kristin Liabo",
"expertise": [
"Reviewer Expertise Patient and public involvement in research"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article gives a comprehensive overview of the involvement of patients, carers and members of the public in health research. It is well written and addresses some important thorny issues in regards to involvement. The article is likely to be useful for researchers new to involvement. Also, as someone who has been facilitating involvement for some time it is very useful to see all this information pulled together. Partly due to the comprehensiveness, I suggest some of the information in boxes and supplements could be cut down. The sheer volume of information might confuse novice readers.\nFor example, the authors point to the many different and overlapping terms used in this field. For a new author it might be more helpful if the information about all these terms was reduced so that the article focuses on the substantial differences rather than the terms used. I found boxes 1 and 2 quite overwhelming to read through at the beginning of the article. Perhaps it would help if they were supplementary files instead of in-text?\nAnother example is the definition of ‘participating’ – I found the introduction to participatory action research here quite confusing because this can be a study design where involvement and participation happen in tandem or are intersected.\nOverall, I would have liked to see less of the detailed information on various websites and terms, and more incorporation of the thorny issues, e.g. where research and involvement intersect, and discussion about what we can do about this. It is these aspect of the article that are most interesting, in my view. But the attention to this would depend on the purpose of the article.\nSome other comments: In Box 3, which gives an overview of the involvement process, I would suggest not using the term bias, because this is commonly used for samples. I agree with the point made – and it is important as often ignored in involvement guidelines – but I think it would be better not to use bias and instead say something about whether the topic of the research is contentious amongst different groups of patients (or something to that effect).\nAlso in Box 3, bullet number 7 mentions roles and responsibilities. I would suggest the agreement of these needs to come much earlier in the process. For example, this could come under point 1 – when the researcher familiarises themselves with involvement. What roles would they like patients, carers or members of the public to have?\nI really like Tables 1 and 2 – these are super useful overviews. I am not clear what purpose Supplement 1 has, this relates to my previous point on the brevity and detail of information.\nIn Supplement 2 I find some of the examples lacking in purpose and clarity beyond saying that involvement can happen. I don’t find that most of the examples provide new information that isn’t available elsewhere in many different formats. However, some examples are very interesting and could merit more space. These are: Example 2 about the cautionary tale – this is a new point that I have not heard before; Example 6 which gives a very interesting example of when qualitative research and involvement overlap (again, would merit more discussion); Example 9 which is very brief but points to how protectionist policies can exclude people from participating (a common reason for not involving children, young people, people with disabilities, the frail elderly and other vulnerable populations) – I don’t think this has been considered in-depth by policies intended to increase participation; Example 10 is good on details on how researchers can work collaboratively with patients/carers/members of the public and will be of interest to people looking for new involvement ideas.\nYou describe patients as primarily fulfilling three functions when they are involved in research: research decision-making, enhancing the patient perspective, how to capture knowledge. I suspect many involved patients/public advisors will object to this as being too limiting. In my own work I have seen at least three additional kinds of input: 1) public advisors helping researchers plan involvement in their research, 2) public advisors helping with dissemination and collaboration strategies (recently a public partner pointed out that the researchers had not asked for a letter of support to a very key national charity which could really help with dissemination), and 3) what I’d call ‘hidden or obvious talents’. Examples of the latter are patients drawing on their previous careers or hobbies, or other talents for seeing new aspects of an established research method. A parent carer we work with decided to use VideoScribe to disseminate some research she had initiated, she then presented about this software at a research seminar, and the research programme bought the software as a direct result of hearing her talk about it. In this example the function of the involved parent was to influence the dissemination and communication strategy of a whole programme of work.\nI hope these comments are useful.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "39334",
"date": "16 Oct 2018",
"name": "Sally Crowe",
"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\nThank you for the opportunity to review this article – which I think is an important mile stone publication in the business of patient and public involvement in the activity that prepares and contributes to research applications. This is a timely and useful review of the literature and the experiences of the research authors, one day we might look back and realise that investing in this stage of research yields benefits and savings in research down the line. I was interested in the reasons for the authors writing this article and was not surprised to see that it was about the conflation of qualitative research and patient and public involvement in research. This article helpfully, and in a practical way helps to ‘de couple’ the most pertinent issues in this domain. The use of an illustrative example in migraine research helps to bring the review of evidence and discussion in each section to a useful and practical conclusion for the reader, and could be easily re interpreted for their own research context. It is also a strength in this article that research is considered as both primary studies and research synthesis and reviews, both of which require careful patient and public involvement Page 3 – in the section How does patient involvement differ from patient participation in research? “the context and outcomes from listening (to patients) differ. PPI means that researchers are in a continuing and reciprocal relationship with patients and make decisions with them about the research. In qualitative research, researchers listen to patients in order to improve their understanding of a topic”. This is probably the most helpful sentence I have read in a while! The further discussions about the choices of researchers that may or may not include qualitative perspectives in their analysis further underlines that in PPI the power dynamic is different and it is an important difference. Figure 1 is helpful I have a problem with the word iterative – and would suggest a plainer language option especially as the rest of the language used in the diagram is of the non-research variety. Table 2\nI struggled with this table maybe because I am not a researcher – I would prefer to see the differences between the two rather than the strengths and limitations of each but can appreciate that for researchers making choices this might be very helpful information and analysis. Boxes 1 and 2\nFor an article such as this I think that the contents of Box 2 are more helpful for readers untangling what is meant by PPI in pre funded research and would put Box 1 as a supplementary file – this information is more easily found for a curious researcher and Box 2 really adds value to the article as a whole as authors have collected the (sadly) rather large collection of terminology used in PPI and sought to differentiate it. I like the fact that the authors have stated and used their preferred terminology and articulated the reasons for their choice (partnership approaches).\n\nTable 1\nI imagine this could be immensely useful to researchers and it is a useful comparison tool but positioning in the middle of the narrative is a shame I think as it breaks up the flow of reading.\n\nFind out what research questions are priorities for patients I think that there is a step before initiating priority setting exercises that encompass PPI – increasingly there are published accounts of priority setting that may or may not include the perspectives of patients and the public. There is also an emerging checklist to support the quality assessment of these accounts and specifically the degree of PPI in them\nTong A., Sautenet B., Chapman JR., Appraisal checklist used in a systematic review of priority setting partnerships in health research (Currently being reworked as a priority setting reporting checklist and will be renamed REPRISE Checklist).\n\nConsider equity of opportunity, unheard perspectives and health inequalities\n\nThis is a particularly important and pertinent section and offers practical advice and ideas for researchers, especially the use of outreach models that offer more scope for addressing these issues more directly, but may put researchers out of their comfort zones.\n\nConsider the potential for bias and conflicts of interest\n\nThis feels an underwritten first paragraph. It feels more important to instigate transparency and declaration of financial and other interests for all research participants – than the more in-depth description of the ‘experienced patient?’. Additionally some text about how to manage these conflicts of interests in patients/public may really help readers address this issue.\n\nAgree appropriate funding for patient involvement\n\nThere are important considerations here especially around equity of involvement – its good to see these spelled out.\n\nTraining for patients involved in research I was very pleased to see this in the text for this section “For patients new to a PPI role, support to develop their abilities and confidence to express their views and question researchers may be relevant.” A much under-appreciated aspect of PPI, I would suggest that it doesn’t just apply to those new to a role who may find it harder (as am embedded part of the research team) to challenge the research orthodoxy…\n\nAlso the training needs to be two way – for research teams as well as involved patients and the public.\n\nKey messages Important questions\n\nI like these but I would include a challenging first question to researchers – ‘Why do they want to do PPI?’ I think that there is an important aspect of self-discovery in PPI in research and it helps to understand self/organisational motivations to doing this preparatory work. These reasons may encompass rational or outcome-based reasons (which the current list addresses) but it is also important to understand motivations on a human and relational level.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "39231",
"date": "02 Nov 2018",
"name": "Kathryn M Sibley",
"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 the opportunity to review this manuscript. This article gives an overview of steps research teams and patient partners can take when working together to prepare a research funding application. We believe a guide for inclusive and meaningful involvement of patients and public partners, particularly in the planning stages of research, is a highly useful and timely resource for health researchers. We thought it was appropriate that the authors included a variety of resources from different continents, and appreciated the attempt to summarize the varied (and often confusing) terms and concepts that exist around patient and public involvement. It was great to see involvement of patient partners in authorship of the paper. We also wonder if patient partners have been involved in providing review comments as well.\n\nWe did struggle with the overall flow of the article, due to its non traditional structure, thereby making it challenging to follow for the reader. Despite the valiant attempt to describe differences in terminology around patient and public involvement in research and key concepts that need to be considered for meaningful patient and public involvement, terms are at times dispersed, seem to be used interchangeably throughout the article, and do not follow a step-by-step order that would be expected from an overview of ‘steps’ in a process. In addition, many key considerations that should be made prior to engaging patients or members of the public are left to the end of the article (such as training for patients involved in research, or working together ethically). In general, the article operationalizes PPI into a series of steps and although mentioned at some points later in the article, the theory and principles of equity and social justice that are critical to meaningful and inclusive engagement are presented later in the article. However, to ensure researchers are incorporating these principles, we suggest they be introduced sooner and have a more central role as the lens through which PPI is viewed. We feel it would be more beneficial to begin with terminology and key concepts, and then provide an overview of the subsequent steps involved in engagement.\nGiven that the article appears to be an attempt to synthesize information and provide an overview of the steps to take when involving patients and the public in preparing a research grant application, a clear description of the methods used to gather, sort, and summarize the sources and information contained in the article is missing. Such information is critical for demonstrating the strength and rigor of the recommendations. Although we recognize this is an opinion article that is not intended as a formal synthesis, we feel that including a description of the literature review methods would provide a landmark for others who may seek to undertake a full synthesis in the future.\n\nThe tables and boxes included are excellent resources for researchers seeking to learn more about PPI and how it is done in different contexts and regions. However, aside from the definitions, their placement may be better suited to the supplementary material, rather than embedded within the article itself (especially Table 1, as it is in landscape orientation making it difficult to read in digital form).\n\nAs we have heard from our stakeholders and patient and public partners, language is a critical aspect of ensuring that PPI is successful, inclusive, and respectful. Many of the decisions that should be made through dialogue with patient partners are presented as decisions for the research team to make and share with patient partners. We recommend directing researchers to dialogue with patient and public partners about topics such as ground rules, preferences for compensation, preferences for feedback, and other matters in which patient partners are involved, in order to ensure that decision-making power in engagement activities is equally shared. A specific example is on page 12 under the heading “Agree appropriate funding for patient involvement”, where the authors state that researchers “Negotiate with patients the costs: payment for patient time, any special needs…” and “Researchers are advised to spell out to patients the best case and worst case scenarios…” – we caution to consider carefully how language can challenge or perpetuate power differentials that often tend to place the researcher above the patient as having more expertise and education. We suggest that terms like ‘discuss’ and ‘dialogue’, in place of ‘negotiate’ and ‘spell out’, can help work towards equality and mutual respect in the research partnership. Some specific points the authors may want to address are detailed below.\n\nPage 3: Given that the concept of PPI is central to the article, a definition (or a discussion of the variability between definitions) should be one of the first points of discussion. The authors state that “PPI includes patients, potential patients, families, carers, patient groups and members of the public who use or have access to health and social care services”. We would suggest PPI also includes those who may be currently unable to use or access health and social care services.\nThe authors discuss ‘patients’ including “people who do not describe themselves in this way”, an important consideration for ensuring inclusive and meaningful engagement – however, we feel the discussion of why people may not describe themselves as ‘patients’ (for example, medicalization of people with disabilities, stigma attached to living with a mental health issue). Box 1: It would be of great interest to us to know how the resources listed in this box were chosen and why. We also believe that given the iterative nature of many online resources and the frequency with which URLs change, it may be best to avoid including hyperlinks for specific documents, many of which will likely become defunct within a year’s time. Box 2: Under ‘some acronyms for involving people in research’, it is unclear why this particular handful of terms was chosen, yet we do not see common acronyms such as PE (patient engagement; https://bcsupportunit.ca/patient-engagement-methods-cluster) and PAR (participatory action research1) . The section of definitions appears to lack structure or hierarchy and is confusing in its presentation. Rather than clarify terms, this box seems to further complicate the differences in language used. Specifically, ‘partnership’ in engagement should also include participatory action research and community-based participatory research2; devolving would also seem to fall under ‘partnership’; ‘consulting’ is described as “relatively passive when compared to ‘engagement’”, but still falls within the realm of ‘engagement’.\n\nBox 3: This overview is an excellent concept and will be highly useful to research teams looking to engage with patients in their work. However, we feel that there may be some information not found here that may be useful. This includes a step for assessing capacity of both research teams and potential patient and public partners to meaningfully engage – for example, determining the willingness of the research team to change directions based on patient and public partner input; determining what resources are available to support engagement activities; and determining the underlying purpose for engaging.\n\nPoint 3 indicates researchers should identify patients to involve “as early as possible” – this may be unclear to those unfamiliar with the PPI process, and should be more explicit – i.e. engaging before the research questions and methods have been determined – something we do not see mentioned elsewhere in this article.\n\nUnder 3.i, it is suggested that researchers consider identifying a “professional or lay link worker”, however many research teams will not (for whatever reason) choose to hire an outside professional to assist with engagement activities, and will embark on it themselves. For this reason, we feel it is critical to mention the need for reflexive practice and trauma-informed approaches for those who are not familiar with these approaches and their need when engaging with people with lived experience of health issues, and the potential retraumatization associated with sharing those experiences.\n\nUnder 5, there is a minor grammar error – “agree an approach” should read “agree on an approach”.\n\nUnder 6, we believe the use of language such as ‘negotiate’ may perpetuate power imbalance between research teams and patient and public partners. This language posits researchers and patients on opposing sides, when perhaps what the authors intended to suggest with ‘negotiate’ was a method of working together to determine appropriate funding and compensation – with which we agree. Perhaps using a term such as ‘researchers and patient partners should work together’ instead of ‘negotiate’ would be more appropriate. In relation to this point, it seems that a key component missing from this article’s discussion of compensation for patient and public partners is determining the preference of patients and actually asking themwhat they want in terms of level of involvement, supports, payment, etc. Similarly, under 8, the authors suggest that researchers “consider whether patients or patient groups will ‘do’ any research” and subsequently assess their skills, ability to add value, status as employees, and mentorship/supervision – we again suggest that these considerations should be made with patient and public partners, rather than decided for them.\n\nPage 10: From the title of the section “Action research, co-design, co-learning and co-production”, it seemed as though the authors would be providing a discussion of these different principles, but instead was a discussion of the differences between definitions that should probably have been included at the beginning of the article. We would suggest moving this information to before the ‘steps’ of involvement are discussed, and potentially changing the heading since it is somewhat misleading given the content of the section that follows. Additionally, the first sentence discusses the historical context of ‘action research’, yet this is the only mention of the chronology of terms and seems somewhat out of place. Without discussion of the origin of other terms mentioned in the article, we would suggest that this is unnecessary to the reader’s understanding of concepts.\n\nUnder “Action research, co-design, co-learning and co-production”, there is a minor grammatical error – in the sentence “Equal power in decision-making is sometimes implied, however there are structural and economic power differentials between different types of partner in terms of pay…”, the word “partner” should be pluralized to read “partners”.\n\nThe authors have brought up a critical consideration regarding power differentials between patients, however the discussion of these and other systemic power imbalances seems to be somewhat lacking in the article, particularly regarding the dynamics of relationship-building in research partnerships, where researchers may hold more ‘power’ and need to engage in reflexive practice to understand how this impacts the process of engagement. We also recommend a discussion of safe spaces, a key consideration for addressing power imbalances in research partnerships. As a reference, we would suggest reviewing the 2017 BMC Health Services Research article by Shimmin et al.3.\n\nIn the second column on this page, the sentence “Partnership approaches can be resource intensive require leadership skills…” appears to contain a grammatical error and should perhaps read “Partnership approaches can be resource intensive and require leadership skills” or “Partnership approaches can be resource intensive, requiring leadership skills”.\n\nThe authors mention “negotiation skills are required to accommodate different perspectives in order to reach consensus in a timely manner”, however we are not sure that this aligns with the principles of meaningful and inclusive PPI in research – though consensus and timeliness are obviously of importance to researchers, we are not convinced that these are central tenets of engagement, and would argue that respect for patients’ stories and lived experiences should be valued over ‘timeliness’.\n\nUnder “Find out what research questions are priorities for patients”, the authors state that “If a research question is of low priority to the people affected by the condition, or important outcomes are not considered, and/or the intervention in question is considered unacceptable to patients, then further research is wasteful”. Although we do not disagree, this is a bold statement that we would suggest tempering, given that basic, fundamental research is often of low priority to patients, and may not have a direct intervention for treating disease or improving health, though it often serves as critical foundations for future discoveries.\n\nFigure 1: The figure demonstrates the people involved in different types of activities. It seem as though the integrated partnership approaches are intended to be positioned as the ‘overlap’ between PPI and qualitative and survey research. Generally, we found this figure is somewhat difficult to interpret and aesthetically unpleasing. We also question the exclusion of health and social care staff from PPI, and wonder why people in these roles would not be able to participate in PPI.\n\nPage 11: In determining patient priorities, we are pleased to see mention of JLA, an important methodology. We also think this would be an excellent opportunity to introduce the Patient-Led Research Hub, mentioned later in the article.\n\nThe use of survey methodology as an example of identifying patient priorities seems overused – there is a lack of discussion of other potential priority-setting methods (such as patient journey mapping and digital storytelling). This is particularly important for researchers who are new to the concept of PPI who would benefit greatly from information about alternatives to the survey methodology.\n\nMention of hard-to-reach groups should be mentioned initially under “Identify patients and/or patient groups to involve as early as possible”, to ensure it is at the forefront of researchers’ minds when considering their recruitment strategy.\n\nPage 12: Under “Consider the potential for bias and conflicts of interest”, we would recommend mentioning that this should be a consideration for discussing ground rules and/or terms of reference. This would also be a good time to mention that the research team and patient partners should discuss together what to do in cases of potential conflicts of interest, and how those will be addressed in the context of the research partnership.\n\nWithin the heading “Negotiate and agree an approach” there appears to be a word missing, such that the heading should read “Negotiate and agree on an approach”. Additionally, in the first sentence of the paragraph under this heading, “Once patients are involved, it is advisable to agree clear boundaries…” should read “Once patients are involved, it is advisable to agree upon clear boundaries…”.\n\nIn discussing the ‘scope of the role’, it is also important to note that the role of the researcher in the partnership should also be discussed, and this should be indicated in the article.\n\nThe authors state that “Patients can contribute to three key functions” – we would argue that patients often contribute to other functions such as collecting data, interpreting results, and informing knowledge translation activities, to name a few.\n\nThe authors state that “Patients value individual constructive and honest feedback about their contributions in order to learn, gain confidence and maintain motivation.” Although this statement may be true, the indicated reference is a link to a NIHR site declaring the development of national standards for PPI (from March 2017). This statement in particular should be supported by a reference, or replaced with the suggestion that research teams discuss with patient and public partners how they want to receive feedback and what they need to learn, gain confidence, and maintain motivation.\n\nThe heading “Agree appropriate funding for patient involvement” seems to be missing a word, and should probably read “Agree on appropriate funding for patient involvement.”\n\nThe sentence “Negotiate with patients the costs: payment for patient time” should also include expertise, a major contribution of patients to the research process, such that this part of the sentence should read “payment for patient time and expertise.”\n\nPage 13: The inclusion of consideration for patients not having to pay upfront for travel and having access to computers and printers is critical and we commend the authors for including it.\n\nIn discussing the steps for PPI in grant development, discussing how to budget for these early activities would be appropriate and useful for those reading this article. At the very least, a reference to an existing budget tool (such as INVOLVE’s cost calculator - http://www.invo.org.uk/resource-centre/payment-and-recognition-for-public-involvement/involvement-cost-calculator/) would be appropriate and highly useful.\n\nUnder “Working together ethically”, it would be helpful to include references for researchers to explore these concepts in more detail – how to ensure distribution of power in research, perhaps guides to reflexive practice questions (such as Shimmin et al.’s article), guidelines for valuing patient contributions and fair compensation (such as those produce by the SPOR National Disease Networks).\n\nPage 14: The authors mention “Audio-recording… does not require ethics committee approval” – we are unsure whether this advice is accurate for all potential jurisdictions, and would suggest including a note about inquiring with the readers’ own regulatory bodies would be appropriate here.\n\nThe authors warn “Expect to be challenged if PPI appears tokenistic”, but provides no guidance for what this means and how to avoid it. We would suggest including more detail and/or a reference.\n\nThe authors state that “It helps to use language precisely and to understand how involving, participating, collaborating, consulting and engaging with patients in research differ” - this statement gives the impression that these concepts are mutually exclusive, when in fact they are overlapping and intertwined. Perhaps this could be restated as “how involving, participating, collaborating, consulting and engaging with patients in research differ, and what the overlap between these concepts is.”\n\nIn the Conclusion, the authors call for evidence about the engagement process, and we would suggest they may also want to touch on the potential to have patient and public partners included as participants (in evaluating the engagement process), and should discuss this when issues around ethics are discussed (page 13) and/or when comparing partners and participants (page 3).\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "39232",
"date": "12 Nov 2018",
"name": "Clare Jinks",
"expertise": [
"Reviewer Expertise I am an applied health services researcher (musculoskeletal conditions and long term term conditions) with an interest in PPI practice and research"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for the opportunity to review this paper. The aim of this paper is to discuss how researchers can involve patients when applying for research funding, and to outline opportunities and pitfalls. It uses a combination of knowledge about this process from literature and the authors' wide ranging experience.\nThis article provides a general background to patient and public involvement in research. The article very usefully includes in Box 1 and Table 1 current resources and guidelines that are available. The information is very comprehensive and relevant to the aims of the article - which is how to to incorporate PPI into the design and conduct of research. This overview is unique and in my view will be widely used. However can it be moved to later in the article? The tables do break up the flow of the paper and break up what I think is an important section about the difference between PPI and patient participation in research.\n\nThe section on the difference between patient involvement and patient participation in research is much needed as PPI is often conflated with qualitative research. This is still a common problem. I wondered if Figure 1 could come earlier, or if Box 2 can show the differences more explicitly instead of providing a list of definitions, can the differences be displayed side by side, using maybe column headings of differences and similarities. This will enable the reader to access the important distinctions more easily.\nOn page 10 in the section about action research, co-design and co-learning there is a sentence about how a partnership approach may interact with an intervention. I think this is the first time intervention studies are mentioned as previous text has stated that patient participation in research may be through interviews, focus groups and surveys. I wondered if this could be expanded to include that these can be undertaken in the context of trials, as then the text would align better with the material in the supplementary documents.\n\nThere are a lot of examples in the supplementary files and this can be quite difficult to navigate.\n\nI have an observation on how the steps are characterised. At first I thought the text in italics are the recommended steps for researchers? For example on page 11 the first step for researchers outlined is to search the internet for organisations and guidelines to see if research priorities have been established in their area of interest. There is however some advice given earlier (written in italics) that recommends understanding what PPI will add and which uncertainties about patient perspectives may benefit from more exploration.\nThen I realised that the steps are characterised in the blue subheadings. These are a mix of 1. advice (e.g. understand what PPI is (this is a conceptual issue)), 2. questions (e.g. how does PPI differ from patient participation?) and then 3. labels for different issues (e.g. patients doing research, which is a practical issue). Could there be more consistency across the paper? I wondered if it would work if they were all translated into specific statements / recommendations particularly as the aim is to provide guidance on how to incorporate PPI.\nRelated to the above - I wondered whether the section headings of \"how does PPI differ from research participation\" and action research, co-design, co-learning and co-production\" should be at the same level as the others. They are very important - but I think are fundamentally about understanding what PPI is.\n\nThe text in italics is also a mix I think of statements of principles - e.g. seek to balance positive and negative views, rather than a statement of how to do that (the title of the paper is how to incorporate PPI....)\n\nThe section on how many patients to involve is again really helpful as this is the question that people often ask. This is usually asked in the context of \"representativeness\" of the patients' views. Can this section be expanded - with some examples of what teams can say when the \"representativeness\" of their patient advisors is questioned. It links to a later section on bias and conflict of interest.\n\nThe authors recommend that researchers select patients with balanced views. It is difficult to know what view people hold at the outset. Can this recommendation again be expanded upon - with ideas about how you may do this, or how you may get fresh perspectives over time.\nTowards the end there is a heading about involving patients in writing a grant application. This made me realise that the advice spans both the process of PPI at pre-award and then outlining what to do post award. I wondered therefore whether this section should come earlier. PPI advisors will have an input into the design of the PPI throughout and the research design throughout, and therefore impact on decision making about earlier topics in the article.\n\nI think this is a really valuable paper in providing such a comprehensive overview of resources and guidelines for PPI and in highlighting the key challenges of good PPI practice. My suggestions are really about structuring the content to make it more accessible to the reader by highlighting maybe two aspects. First the need to understand what PPI is and is not. Second the practical issues - how to do it.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-752
|
https://f1000research.com/articles/7-213/v1
|
22 Feb 18
|
{
"type": "Software Tool Article",
"title": "shinySISPA: A web tool for defining sample groups using gene sets from multiple-omics data",
"authors": [
"Bhakti Dwivedi",
"Jeanne Kowalski",
"Bhakti Dwivedi"
],
"abstract": "As opposed to genome-wide testing of several hundreds of thousands of genes on very few samples, gene panels target as few as tens of genes and enable the simultaneous testing of many samples. For example, some cancer gene panels test for 50 genes that can affect tumor growth and potentially identify treatment options directed against the genetic mutation. The increasing popularity of gene panel testing has spurred the technological development of panels that test for diverse data types such as expression and mutation.\n\nOnce samples are tested, there is the desire to examine clinical associations based on the panel and for this purpose, one would like to identify, among the samples tested, which show support for a molecular profile (e.g., presence of mutation with increased expression) versus those samples that do not among the genes tested. With user-specified molecular profile of interest, and gene panel data matrices (e.g., gene expression, variants, etc.) that define the profile, shinySISPA (Sample Integrated Set Profile Analysis) is a web-based shiny tool to define two sample groups with and without profile support based on our previously published method from which clinical associations may be readily examined. The shinySISPA can be accessed from http://shinygispa.winship.emory.edu/shinySISPA/.",
"keywords": [
"Sample profile",
"Gene set profile",
"Genomics",
"Integrated analysis"
],
"content": "Introduction\n\nUnlike gene set profiling, sample profiling is a challenge due to the heterogeneity between, and within the tumor patient samples. Identification of homogenous groups of samples or molecular subtypes is commonly approached using clustering methods (e.g., Handl et al., 2005; Kowalski et al., 2016; Monti et al., 2003; Șenbabaoğlu et al., 2014; Verhaak et al., 2010). Whether or not the sample groups are meaningful and the clustering stable, requires additional testing, is highly subjective, limited to examining changes in a single data type, and often require removal of genes or samples to obtain the desired results. While clustering tools such as TNBC subtyping (Chen et al., 2012; Lehmann et al., 2011) are convenient for subtype discovery and sample classification, they are restricted to studying a specific cancer tissue and data type, within the context of established expression signature profiles. Although these methods may prove useful in certain case, there is a need for a basic tool that can identify sample groups using any combination of genomic data types based on a gene or gene set and molecular profile of interest. Some examples of gene sets may be derived from a specific biological process, network, gene enrichment analysis, a gene panel, etc. A molecular profile is a series of increasing or decreasing changes among diverse data types operating on a given gene set. For example, a gene mutation with expression is a molecular profile of increased variant support with increased levels of expression. The shinySISPA is a web tool developed to implement the novel method, SISPA (Kowalski et al., 2016), for defining samples with similar molecular profiles based on a user input gene set and data types. SISPA does not impose analytical distribution assumptions on the data, and is scalable to define samples that support a general profile defined by any combination of genomic data types applied to any number of genes.\n\n\nMethods\n\nSISPA is written in the R programming language (R project) and the shiny web application framework is implemented using the Shiny R package (Chang et al., 2016).\n\nThe tool is hosted on a 64bit CentOS 6 server (http://shinygispa.winship.emory.edu/shinySISPA/) running the Shiny Server program designed to host R Shiny applications. This tool has been extensively tested on Windows 7 and Mac Pro 10 operating system with firefox and google chrome browser. Given a dataset of 377 samples and 16 genes under the two-feature analysis, it took three seconds to obtain shinySISPA defined sample groups and less than a second to generate the waterfall plot. The time it takes to generate sample profile diagnostic plots depends on the number of genes in a set; it took less than 10 seconds for 16 genes in both sets of a two-feature analysis. As a note, speed at which results are generated is also dependent on the internet connectivity.\n\nThe tool workflow consists of four basic inputs as shown in Figure 1:\n\nHere, we define samples supporting the molecular profile of decreased gene expression and copy loss. The tool requires user selection of analysis type, user upload of data types on samples and gene sets, and specification of a profile to output the samples supporting that profile (shown in grey). The samples are selected based on a change point model applied to composite (among features and genes), within-sample z-scores. A waterfall plot of profile activity is output with samples selected in orange as showing the most support for the profile.\n\n(1) Selecting the analysis type. User selects a single-, two-, or three-feature analysis, where a feature corresponds to a specific data type (e.g., expression, methylation, mutation, copy number variation) and thus, a single-feature analysis refers to use of a single data type, while a two-feature uses a combination of two data types and so forth.\n\n(2) Uploading the data. User inputs the data for each feature containing the genes and samples of interest. The same samples are required for each feature, though the gene sets may differ between features.\n\n(3) Specifying a molecular profile. A molecular profile is a series of increasing (“up”) or decreasing (“down”) genomic changes within each feature. In Figure 1, a profile of decreased expression with decreased copy number is input.\n\n(4) Selecting the number of breaks to define sample groups. User can specify the change point detection method (Killick et al., 2016) for finding optimal break points in the distribution of computed composite (among features) z-scores within samples (Kowalski et al., 2016; see Supplementary File 1).\n\nThe results are output in four separate tabs:\n\n(1) Input Data. Summarizes the user input data in terms of the input number of genes, number of samples, and box plot distribution by data type.\n\n(2) SISPA Results. Outputs the table of defined sample groups with their gene set enrichment score for the selected analysis type and molecular profile of interest. The scatter plot on the right displays all the change points detected within the data-set, samples falling in the topmost change point are the samples with the profile activity. The frequency plot at the rightmost bottom represents the distribution of the number of samples with and without the profile activity.\n\n(3) Waterfall Plot. To visualize the sample groups that correlate with the profile of interest. Samples with the profile activity have the highest score and are shown in orange filled bars, while samples without the profile activity are shown with grey-filled bars.\n\n(4) Sample Profile. Represents the diagnostic plots to visualize the distribution of the user-input data overall by the identified sample groups. It also allows the users to view data distribution for a selected gene in the set within each data type to assess what genes in particular satisfy the profile versus samples without profile.\n\nAll results generated during the process are directly downloadable on the user’s local computer. A detailed manual with tool settings are provided in the Supplementary File 1. Upon forming such sample groups, one may readily examine the effect of a profile on various clinical and biological clinical outcomes.\n\n\nUse case\n\nWe applied shinySISPA to profile newly diagnosed multiple myeloma (MM) patients for decreased gene expression and copy number based on a GISPA (Gene Integrated Set Profile Analysis)-derived gene set characterizing the IgH translocation in the MM cell lines (Kowalski et al., 2016). The t(14;16) translocation is known to be associated with poor prognosis in MM. By applying the shinySISPA tool to the t(14;16) characterized gene set profile, we were able to translate cell line profiles to patient profiles. Using the IA6 release of Multiple Myeloma Research Foundation (MMRF) CoMMpass study, we downloaded data from 377 newly diagnosed patients at pre-treatment with available clinical outcomes, RNA-Seq expression, and DNA-copy number variations from the MMRF Research Gateway portal. Based on our two-feature analysis, 7 of the 300 MM patients were defined with profile activity (Figure 1) by identifying changes in variance using change point v2.2.2. Furthermore, we used CASAS (Rupji et al., 2017) to compare survival curves of the identified two sample groups for downstream clinical interpretation. We found seven samples with profile activity to be significantly (P<0.0001) associated with poor survival as compared to the 300 samples without the profile activity (HR = 9.81; 95% CI = (3.39, 28.37)).\n\n\nConclusion\n\nWe have demonstrated the utility of our shinySISPA tool in translating cell line characterized gene sets molecular profile to patient profiling (Kowalski et al., 2016); however, one can use any a priori-defined gene sets with any combination of molecular data for identifying samples with a similar gene set profile. The introduction of a change point model to select samples with profile support offers a more objective approach than with clustering methods. With only a gene set and a combination of data types from the same samples, the tool is widely applicable to many settings. For example, shinySISPA may be used to define patients based on known drug targets and pathways, or to identify patients that may be at risk for poor prognosis based on known prognostic markers.\n\n\nData and software availability\n\nThe example sample data used to demonstrate shinySISPA workflow is available on the web-version and with the package source code at: https://github.com/BhaktiDwivedi/shinySISPA.\n\nThe shinySISPA software is available at http://shinygispa.winship.emory.edu/shinySISPA/\n\nThe stand-alone version of SISPA is available at https://www.bioconductor.org/packages/SISPA/.\n\nArchived source code as at the time of publication: https://doi.org/10.5281/zenodo.1164284 (Dwivedi & Kowalski, 2018)\n\nLicense: shinySISPA is available under the GNU public license (GPL-3)",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work is funded by the Georgia Research Alliance Scientist Award (Jeanne Kowalski); Biostatistics and Bioinformatics Shared Resource of Winship Cancer Institute of Emory University and NIH/NCI [Award number P30CA138292, in part]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank the Cancer Informatics Core of the Winship Cancer Institute of Emory University for supporting the CentOS server. Special thanks to Kenneth Buck for his assistance with building and configuring the server for hosting shinySISPA. Special thanks to Manali Rupji for her assistance on clinical survival analysis.\n\n\nSupplementary material\n\nSupplementary File 1: The shinySISPA manual\n\nClick here to access the data.\n\n\nReferences\n\nChang W, et al.: shiny: Web Application Framework for R. R package version 0.13.2. 2016.\n\nChen X, Li J, Gray WH, et al.: TNBCtype: A Subtyping Tool for Triple-Negative Breast Cancer. Cancer Inform. 2012; 11: 147–156. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDwivedi B, Kowalski J: shinySISPA: A web tool for defining sample groups using gene sets from multiple omics data (Version 1.0). Zenodo. 2018. Data Source\n\nHandl J, Knowles J, Kell DB: Computational cluster validation in post-genomic data analysis. Bioinformatics. 2005; 21(15): 3201–3212. PubMed Abstract | Publisher Full Text\n\nKillick R, Haynes K, IA E: changepoint: An R package for changepoint analysis. R package version 2.2.1. 2016.\n\nKowalski J, Dwivedi B, Newman S, et al.: Gene integrated set profile analysis: a context-based approach for inferring biological endpoints. Nucleic Acids Res. 2016; 44(7): e69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLehmann BD, Bauer JA, Chen X, et al.: Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011; 121(7): 2750–2767. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMonti S, Tamayo P, Mesirov JP, et al.: Consensus clustering: A resampling-based method for class discovery and visualization of gene expression microarray data. Mach Learn. 2003; 52(1–2): 91–118. Publisher Full Text\n\nRupji M, Zhang X, Kowalski J: CASAS: Cancer Survival Analysis Suite, a web based application [version 2; referees: 2 approved]. F1000Res. 2017; 6: 919. PubMed Abstract | Publisher Full Text | Free Full Text\n\nȘenbabaoğlu Y, Michailidis G, Li JZ: Critical limitations of consensus clustering in class discovery. Sci Rep. 2014; 4: 6207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerhaak RG, Hoadley KA, Purdom E, et al.: Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010; 17(1): 98–110. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "32660",
"date": "16 Apr 2018",
"name": "Yun Zhang",
"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 introduce a Shiny application that are developed as a user-friendly tool for conducting omics-data analysis with their published methodology work. The article provides detailed description of the inputs and outputs for the shiny application, and shows working example for the use of this application. This article is index-able with some necessary modifications. The following are my major comments.\nBefore introducing the shiny application, the authors should provide a brief summary of their methodology work, including clear definitions of important terminologies and a description of their model. It would greatly help the readers to understand the contents followed subsequently.\nWith a glimpse on the methodology paper, I found the terms such as “molecular profile”, “feature” are defined misleadingly in this article. What is a change point model?\n\nFor the input element (4), where to select the number of breaks? Is it the max Q allowed? What is the max Q? What are the options for “Changes Using” in the bottom right of Figure 1? For the output element (2), it is better to include a figure with the description. In the “Use case” section,\nWhy GISPA is used? What’s the relation of GISPA and SISPA? How many genes are used? Is it 16? Where are the 7 patients in Figure 1? Are they the orange bars? What is “using change point v2.2.2”? “We found seven samples with profile activity to be significantly (P<0.0001) associated with poor survival as compared to the 300 samples without the profile activity (HR = 9.81; 95% CI = (3.39, 28.37)).” Previously mentioned, there are 377 patients. Why are 70 samples missing?\n\nFor the reproducibility of the work, please upload a default dataset in the shiny application, so that the readers/users can replicated the analysis for the first time.\nAlso, I provide my minor comments below.\nPlease consider a thorough revision of the English writing for this scientific article. There are a number of grammatical errors, and sentences do not flow smoothly. Please avoid using colloquial words in scientific writing. Please modify the legend of Figure 1. In the brackets “(shown in grey)”, it is hard to locate where it is referring to since there are many grey colors in the figure. Also, please add labels to the subfigures that are referred in the text.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "3695",
"date": "15 Jun 2018",
"name": "Bhakti Dwivedi",
"role": "Author Response",
"response": "Referee Report 16 Apr 2018Yun Zhang, Department of Biostatistics and Computational Biology, University of Rochester, Rochester, NY, USA Approved with ReservationsThe authors introduce a Shiny application that are developed as a user-friendly tool for conducting omics-data analysis with their published methodology work. The article provides detailed description of the inputs and outputs for the shiny application, and shows working example for the use of this application. This article is index-able with some necessary modifications.The following are my major comments.1) Before introducing the shiny application, the authors should provide a brief summary of their methodology work, including clear definitions of important terminologies and a description of their model. It would greatly help the readers to understand the contents followed subsequently.Response: We have provided a reference to our methods paper that includes detailed information on the SISPA approach. In this paper, our focus is upon introducing the application of the method in terms of tool development and implementation by providing detailed examples and information on data input and output/results. Considering this focus, along with space constraints, we have opted to not repeat the already published method description and instead, reference it. With a glimpse on the methodology paper, I found the terms such as “molecular profile”, “feature” are defined misleadingly in this article. Response: We have intentionally used the terms “molecular profile” and “feature” in the same context as in the published methods papers, whereby “molecular profile” refers to change of either increase (“up”) or decrease (“down”) and “feature” refers to a specific data type (e.g., expression, methylation, copy number change). We have also provided examples of the term “molecular profile” to further clarify the context.In this paper: “A “feature” corresponds to a specific data type (e.g., expression, methylation, mutation, copy number variation) and thus, a single-feature analysis refers to use of a single data type, while a two-feature uses a combination of two data types and so forth.” (pg# 3 last paragraph, under Selecting the analysis type) “A “molecular profile” is a series of increasing (“up”) or decreasing (“down”) genomic changes within each feature…” (pg# 4 second paragraph, under Specifying a molecular profile)In the published methodology paper: “The Cancer Genome Atlas (TCGA) nomenclature (https://tcga-data.nci.nih.gov/tcga/dataAccessMatrix.htm) that references a specific data type (RNA-Seq expression, DNA CpG methylation, etc.) as a feature.” “A profile is defined by specifying a priori, a change of either increase or decrease within each of the features..” 2. What is a change point model?Response: A changepoint model is a method for identifying changepoints within data. It is a published method. Please see below references to the changepoint method and the changepoint R package for details: Killick R and Eckley IA (2014). “changepoint: An R Package for Changepoint Analysis.” Journal of Statistical Software, 58(3), pp. 1–19. http://www.jstatsoft.org/v58/i03/. Killick R, Haynes K and Eckley IA (2016). changepoint: An R package for changepoint analysis. R package version 2.2.2, https://CRAN.R-project.org/package=changepoint. 2) For the input element (4), where to select the number of breaks? Is it the max Q allowed? What is the max Q? What are the options for “Changes Using” in the bottom right of Figure 1?Response: Yes, the allotted maximum number of breaks is specified using the “Max Q Allowed”.“In input element (4), users can modify the Changepoint Input (Killick R, et al., 2016) to find the optimal break points within the estimated profile sample score (Kowalski, et al., 2016). The changes can be found using mean(“mean”), variance (“var”) or both (“meanvar”) with the user-specified changepoint method (“AMOC”, “BinSeg”, “PELT”, or “SeqNeigh”) given the allotted maximum number of change points (“Max Q allowed”). Note that the number of change points identified may differ for the same dataset depending on the change point R package version installed on the system. Currently we are running changepoint version 2.2.2 on our hosting server...”Please see pg# 6 of the supplementary file 1 for the details.3) For the output element (2), it is better to include a figure with the description.Response: The output elements (“Input Data” “SISPA results”, “Waterfall Plot” and “Sample Profile”) screenshot including the figures are explained in detail in the supplementary file 1. Please see pg# 7-11 under “Result” Section.4) In the “Use case” section, Why GISPA is used? What’s the relation of GISPA and SISPA? Response: GISPA (Gene Integrated Set Profile Analysis) is a method designed to define gene sets with similar, a priori specified molecular profile. While, SISPA (Sample Integrated Set Profile Analysis) is a method designed to define sample groups with similar gene set a priori specified molecular profile. Both GISPA and SISPA method are published in Nucleic Acid Research (Kowalski et al., 2016; PMID: 26826710). GISPA was used to identify genes with decreased expression and decreased copy change molecular profile in a multiple myeloma cell line with IgH translocation. This gene set is published in the methodology paper Nucleic Acid Research (Kowalski et al., 2016; PMID: 26826710). Here, we extracted RNA-seq expression, and copy number change data for GISPA derived gene set characterizing the IgH translocation on 377 newly diagnosed patients enrolled in the coMMpass clinical trial to define samples with a similar gene set profile, i.e., decreased expression with copy loss. This example data is provided with this paper. Pg# 4 last paragraph under “Use case” describes the use of application of SISPA using GISPA derived gene sets. 2. How many genes are used? Is it 16?Response: Yes. The number of genes in the expression and copy number variation data is 16. 3. Where are the 7 patients in Figure 1? Are they the orange bars?Response: The patients with and without profile activity are highlighted in “Samples Supporting the Gene Set Profile” labeled section of the Figure 1. Yes, the 7 patients are highlighted in orange-filled bars. 4. What is “using change point v2.2.2”?Response: “using change point v2.2.2”? means that we have used changepoint R package version 2.2.2. 5. “We found seven samples with profile activity to be significantly (P<0.0001) associated with poor survival as compared to the 300 samples without the profile activity (HR = 9.81; 95% CI = (3.39, 28.37)).” Previously mentioned, there are 377 patients. Why are 70 samples missing?Response: We have corrected the typo, please see page# 4 and 5, last paragraph. It is 7 of 370 patients.“Based on our two-feature analysis, 7 of the 370 MM patients were defined with profile activity ( Figure 1) by identifying changes in variance using change point v2.2.2. Furthermore, we used CASAS ( Rupji et al., 2017) to compare survival curves of the identified two sample groups for downstream clinical interpretation. We found seven samples with profile activity to be significantly (P<0.0001) associated with poor survival as compared to the 370 samples without the profile activity (HR = 9.81; 95% CI = (3.39, 28.37)).”5) For the reproducibility of the work, please upload a default dataset in the shiny application, so that the readers/users can replicated the analysis for the first time.Response: All users can access and analyze the example dataset (i.e., default dataset) used in the paper by choosing the “Example data” from the Upload Input option on the web-interface. The data is also available to download from GitHub (https://github.com/BhaktiDwivedi/shinySISPA). Users are able to obtain the same exact results, i.e., samples with and without profile activity using the current default settings implemented in shinySISPA web tool. Also, I provide my minor comments below.1. Please consider a thorough revision of the English writing for this scientific article. There are a number of grammatical errors, and sentences do not flow smoothly. Please avoid using colloquial words in scientific writing.Response: We have reviewed the manuscript and do not identify any such problem. If the reviewer still feels strongly about it, we kindly request specific examples to be cited from the text.2. Please modify the legend of Figure 1. In the brackets “(shown in grey)”, it is hard to locate where it is referring to since there are many grey colors in the figure. Also, please add labels to the subfigures that are referred in the text.Response: We have addressed and incorporated these changes. Please see updated Figure 1 and Figure legend. Is the rationale for developing the new software tool clearly explained? Yes Is the description of the software tool technically sound? Partly Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly Are the conclusions about the tool and its performance adequately supported by the findings presented in the article? PartlyCompeting Interests: No competing interests were disclosed.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, however I have significant reservations, as outlined above."
}
]
}
] | 1
|
https://f1000research.com/articles/7-213
|
https://f1000research.com/articles/7-328/v1
|
15 Mar 18
|
{
"type": "Study Protocol",
"title": "Obstructive sleep apnea as an independent predictor of postoperative delirium and pain: protocol for an observational study of a surgical cohort",
"authors": [
"Patricia Strutz",
"William Tzeng",
"Brianna Arrington",
"Vanessa Kronzer",
"Sherry McKinnon",
"Arbi Ben Abdallah",
"Simon Haroutounian",
"Michael S. Avidan",
"Patricia Strutz",
"William Tzeng",
"Brianna Arrington",
"Vanessa Kronzer",
"Sherry McKinnon",
"Arbi Ben Abdallah",
"Simon Haroutounian"
],
"abstract": "Introduction: Postoperative delirium and pain are common complications in adults, and are difficult both to prevent and treat. Obstructive sleep apnea (OSA) is prevalent in surgical patients, and has been suggested to be a risk factor for postoperative delirium and pain. OSA also might impact pain perception, and alter pain medication requirements. This protocol describes an observational study, with the primary aim of testing whether OSA is an independent predictor of postoperative complications, focusing on (i) postoperative incident delirium and (ii) acute postoperative pain severity. We secondarily hypothesize that compliance with prescribed treatment for OSA (typically continuous positive airway pressure or CPAP) might decrease the risk of delirium and the severity of pain. Methods and analysis: We will include data from patients who have been enrolled into three prospective studies: ENGAGES, PODCAST, and SATISFY-SOS. All participants underwent general anesthesia for a non-neurosurgical inpatient operation, and had a postoperative hospital stay of at least one day at Barnes Jewish Hospital in St. Louis, Missouri, from February 2013 to December 2017. Patients included in this study have been assessed for postoperative delirium and pain severity as part of the parent studies. In the current study, determination of delirium diagnosis will be based on the 3-minute Diagnostic Confusion Assessment Method, and the Visual Analogue Pain Scale will be used for pain severity. Data on OSA diagnosis, OSA risk and compliance with treatment will be obtained from the preoperative assessment record. Other variables that are candidate risk factors for delirium and pain will also be extracted from this record. We will use logistic regression to test whether OSA independently predicts postoperative delirium and linear regression to assess OSAs relationship to acute pain severity. We will conduct secondary analyses with subgroups to explore whether these relationships are modified by compliance with OSA treatment.",
"keywords": [
"Obstructive Sleep Apnea",
"Postoperative Delirium",
"Postoperative Pain"
],
"content": "Introduction\n\nObstructive sleep apnea (OSA) is the most common form of sleep-disordered breathing. OSA is characterized by repetitive, functional collapse of the airway leading to cyclical decrements or cessations of airflow during sleep1. It is estimated that 20% of the general population suffers from OSA2,3, and among adults with OSA, up to 75% are unaware of the diagnosis4,5. Of relevance to perioperative medicine, there is also a high OSA prevalence in surgical patients3. In common with the general population, many of these patients are unaware they have OSA6,7. Also of note, prevalence of sleep apnea often varies by type of surgery; for example, prevalence in the bariatric surgery population is estimated to be 70%8,9. OSA prevalence combined with ignorance of diagnosis is cause for concern given the wide range of health consequences. OSA has been causally implicated in an assortment of both acute and chronic disorders. Acutely, OSA has been associated with disrupted sleep, tiredness, and episodic hypoxia and hypercapnia during sleep10,11. Chronically, OSA has been linked to a multitude of co-morbidities, including ischemic heart disease and stroke12, hypertension13,14, arrhythmias15,16, aortic dissection17,18, chronic fatigue19, pulmonary hypertension20,21, diabetes22, and respiratory acidosis with compensatory metabolic alkalosis23,24.\n\nOSA is becoming a growing concern in the perioperative period, as there is increasing evidence linking OSA to adverse postoperative outcomes25,26. For example, following various surgical procedures, patients with OSA probably have more respiratory, cardiac, and neurologic complications27–30, as well as increased postoperative infections31. Unsurprisingly surgical patients with OSA therefore have a higher transfer rate to the ICU28, increased stay in the ICU31, and increased overall length of hospital stay27,28.\n\nOf particular relevance to the research focus of this protocol, certain aspects of OSA such as recurrent hypoxemia, systemic inflammation, and sleep disruption have been associated with altered pain processing and incident delirium32–34. A causal link between OSA and delirium would be clinically important given the negative outcomes associated with postoperative delirium. In the DSM-5, delirium is defined as a disturbance in attention, awareness, and cognition that develops over a short period of time and over the course of a day, fluctuates in severity35. In older adults, the incidence of postoperative delirium ranges from 10–70%, depending on the type of surgery36. Patients who experience postoperative delirium often require an extended stay in the intensive care unit37, subsequently report decreased quality of life38, and might be at increased risk for accidental falls, long-term cognitive decline and death after hospital discharge39. Thus, postoperative delirium is associated with a considerable burden on patients and their families, and an increase to society in the overall cost of healthcare40,41.\n\nThe reported impact of OSA on postoperative pain and pain perception poses further challenges to clinicians and patients. Adequate postoperative analgesia is an important component of recovery, and pain negatively impacts quality of life. Mechanistic evidence in various populations suggests that sleep deprivation promotes up-regulation of cytokines42–47, including interleukin-1β, interleukin-6, and tumor necrosis factor, all of which might induce excessive sensitivity to pain45,48. Consistent with these studies clinical evidence, including compelling data from burn victims, suggests that interrupted and inadequate sleep promotes hyperalgesia32–34,49. Furthermore, Khalid et al. showed that treatment with continuous positive airway pressure (CPAP) in adults with OSA dampened their sensitivity to painful stimuli50. Thus, whether or not people with OSA have increased pain sensitivity might to some extent depend on how effectively they are treated. To complicate matters further, people with OSA, especially if they experience episodic hypoxemia during sleep, reportedly have increased susceptibility to the respiratory depressant effects of opioid medications51,52. Thus, since opioids are the mainstay of therapy for severe postoperative pain, it can be especially difficult to provide safe and adequate analgesia to surgical patients with OSA.\n\nThe objectives of this study are to investigate further the relationships between OSA on the one hand, and common postoperative complications such as pain and delirium on the other hand. We hypothesize that patients with OSA experience more severe postoperative pain and have a higher incidence of postoperative delirium. We further hypothesize these negative outcomes might be mitigated by compliance with OSA treatment.\n\n\nProtocol\n\nThis protocol describes an observational study, investigating the relationship between OSA as a risk factor, and postoperative delirium and acute postsurgical pain severity as adverse outcomes. The three parent studies from which the data are being obtained for the current study have all been approved by the Human Research Protection Office (HRPO) at Washington University, and patients enrolled in all three studies provided written informed consent. The HRPO has also provided approval for this current study. Data will be aggregated from the Systematic Assessment and Targeted Improvement of Services Following Yearlong Surgical Outcomes Surveys Study (SATISFY-SOS, NCT02032030); the Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes study (ENGAGES, NCT02241655); and the Prevention of Delirium and Complications Associated with Surgical Treatments study (PODCAST, NCT01690988). For greater detail regarding the three parent studies, please review previously published protocols and literature53–57.\n\nPatients ≥ 18 years who underwent general anesthesia for a non-neurosurgical inpatient operation at Barnes Jewish Hospital in St. Louis, Missouri, from February 2013 to December 2017, will be included in our analysis. Patients had a postoperative hospital stay of at least one day. The main outcomes of interest will include postoperative delirium and pain, assessed daily until postoperative day 3. The primary aim of this study is to investigate whether OSA is an independent predictor of postoperative delirium and acute postsurgical pain severity. We will conduct secondary analyses with subgroups to explore whether these associations are modified by compliance with OSA treatment.\n\nThis protocol is compliant with published guidelines for observational study protocols, and the conduct and reporting of this study will adhere to the RECORD and STROBE guidelines for observational studies58–60.\n\nInclusion criteria:\n\n(i) Enrollment in the SATISFY-SOS, ENGAGES, or PODCAST study;\n\n(ii) Postoperative stay of at least 1 day following surgery at Barnes Jewish Hospital\n\n(iii) General anesthesia for elective surgical procedures\n\nExclusion criteria:\n\n(i) Neurosurgery\n\n(ii) Age <18\n\ni. Baseline Data. Patients undergoing elective surgery are routinely screened at the Center for Preoperative Assessment and Planning at Barnes Jewish Hospital in St. Louis, Missouri, where detailed medical history is collected and screening tests are administered, including the STOP-BANG (Snoring, Tiredness, Observed Apnea, High Blood Pressure, Body Mass Index>35kg/m2, Age >50, Neck circumference, male Gender) test for OSA risk. Baseline characteristics will be extracted via electronic chart review and will include but are not limited to: age, sex, race, ethnicity, smoking history, alcohol use (average per week), STOP-BANG criteria, OSA status, and pre-existing medical conditions.\n\nii. Delirium assessment method. Delirium is one of the primary outcomes of this study, and will be determined using the 3D-Confusion Assessment Method (3D-CAM), a validated, abbreviated assessment derived from the Confusion Assessment Method (CAM)61. The CAM, a delirium assessment instrument used primarily by non-psychiatrists, typically takes between 15 and 30 minutes to complete62. The 3D-CAM was developed as a method to more efficiently screen patients for delirium. It consists of a subset of the questions used in the CAM, as well as CAM scoring items that are based on patient behavior (10 cognitive testing items, 10 interviewer observations). With this approach, the 3D-CAM is intended to only take 3 minutes.\n\nDelirium in two of the parent studies, ENGAGES and PODCAST, was assessed using either the long form of the CAM or an abbreviated CAM designed and validated63 for critically ill patients, often found in the intensive care unit (CAM-ICU). For this sub-study, pertinent 3D-CAM data from the long CAM assessments will be extracted for our analysis. In the third parent study (SATISFY-SOS), delirium was assessed with the 3D-CAM. The presence of delirium will be defined as a positive 3D-CAM or CAM-ICU during any postoperative assessment through postoperative day 3. In order to qualify for a diagnosis of delirium with the 3D-CAM, the following three criteria must be met: 1) either acute onset OR a fluctuating course; 2) inattention; and 3) either disorganized thinking OR altered level of consciousness. A patient will be considered positive for delirium if the patient is recorded to have had a single instance of delirium during their postoperative stay.\n\niii. Pain Assessment Method. Pain during hospital stay will be assessed using the Visual Analogue Scale (VAS), a validated pain assessment instrument that has been widely used in adult populations64,65. Patients are asked to indicate on a line 100mm in length the severity of their pain in three different situations: 1) at rest, 2) taking a deep breath or coughing, and 3) moving (sitting up, walking, or moving extremities). The patient’s mark is then measured with a ruler and recorded in mm. For our analysis, we will incorporate the highest pain score recorded on any postoperative assessment as our value of interest. As postsurgical pain is often dependent on the type of surgery, we will adjust for type of surgery in our statistical model, as well as other confounding variables described in the methods below.\n\niv. OSA Classification For the primary analysis (Figure 1), patients will be grouped into one of three categories: high risk of OSA (HR-OSA), intermediate risk of OSA (IR-OSA), and low risk of OSA (LR-OSA). Patients with a history of a positive polysomnography test will be classified as HR-OSA, whereas patients with a history of negative polysomnography will be classified as LR- OSA. Patients with no history of polysomnography testing will be classified into one of the three categories based on STOP-BANG screening status. The STOP-Bang questionnaire classifies patients into three commonly accepted categories based on scoring: 0–2 indicates low risk of OSA; 3–4 indicates intermediate risk; 5–8 indicates high risk66. We will follow these guidelines for classifying patients as HR-OSA, IR-OSA, or LR-OSA for our primary analysis, and thus likely demonstrate important trends between and among groups.\n\nOf note, current literature classifies, often for simplicity, a STOP-Bang score of ≥3 as high risk for OSA. However, this can obscure analysis, potentially resulting in a falsely weaker association between OSA risk and risk of postoperative adverse outcomes. Therefore, we will not group intermediate risk of OSA with high risk of OSA. Also, some literature incorporates bicarbonate levels to help determine OSA risk. As baseline laboratory values are not available for each participant, we will not include this component for classifying OSA risk.\n\nFor secondary analysis (Figure 2), we will analyze delirium incidence and pain severity among five patient groupings: confirmed OSA + report using prescribed CPAP, confirmed OSA + report not using prescribed CPAP, high risk for OSA (STOP-Bang 5–8), intermediate risk for OSA (STOP-Bang 3–4), low risk for OSA (STOP-Bang <3). Thus, secondary analysis will likely demonstrate if reported CPAP adherence mitigates these adverse outcomes.\n\nv. Sample Size. We estimate that we will have data with complete outcomes (pain severity and incident delirium) and information on OSA status for approximately 1,300 patients. We estimate that 260 (~20%) of these patients will have incident postoperative delirium. We will have patient reported pain outcomes for all participants. We will use logistic and linear regression, including potential confounder variables, to test for an independent association between OSA as a risk factor and postoperative delirium and pain severity as outcomes of interest. We estimate that it will be appropriate to include up to 25 variables in each of the regression models.\n\nAll electronic data collected during this study, as well as the SATISFY-SOS, ENGAGES, and PODCAST databases, are hosted on a firewall-secured network server. This server is managed and maintained by the IT team of the Department of Anesthesiology, and is securely housed behind two locked doors in the departmental offices. The Project Informaticist, Data Manager, and Director(s) are the only individuals with full access to these password-protected and encrypted databases. Delirium and pain assessments are first completed using paper surveys, which are then securely stored in locked cabinets within the department. Results are entered into a Research Electronic Data Capture (REDCap) tool hosted at Washington University School of Medicine in St. Louis67. REDCap is a secure, web-based application designed to support data capture for research studies, providing: 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for importing data from external sources.\n\ni. Statistical considerations. Continuous variables will be graphically evaluated with histograms, boxplots, and q-q plots, and numerically with measures of skewness, kurtosis, and Kolmogorov-Smirnov tests. Outliers will be excluded, and approximate normality will be ensured before parametric statistics are applied. Perioperative variables will be described with mean ± SD, median [IQR], and numbers/proportions, as appropriate. Differences in patient and other perioperative factors between groups will be evaluated with chi-squared, t-tests, ANOVA, Kruskal-Wallis, and/or Wilcoxon-Mann-Whitney tests, as appropriate. Participants missing outcome data will be excluded from analysis.\n\nii. Delirium. Logistic Regression will be used to assess the relationship between OSA as a risk factor and incident postoperative delirium as an outcome. For our analyses, we will include no more than 1 variable for every 8 outcomes. With an estimated incident postoperative delirium rate of 20%, we plan to include up to 25 pre-specified candidate predictor variables in the primary regression models, including the most clinically relevant interaction terms. Variables for our primary analysis have been selected based on existing evidence, and will likely include: OSA status; Age; Sex; Type of Surgery; Charlson Comorbidity Index; Procedural Cardiac Risk; ASA physical status; Alcohol use. We will also include a history of any of the following comorbidities: Previous Surgeries; Stroke; Dementia or Mild Cognitive Impairment; Visual Impairment; Depression; Anxiety; Chronic Pain; and Diabetes Mellitus. We hope to include BMI and age independently of the OSA risk classification since they are continuous variables, and their inclusion in the regressions might improve the models. We also hope to include the variable ‘tiredness’ in the models since this particular symptom could plausibly independently predict both delirium and pain.\n\niii. Pain. Linear Regression will be used to examine OSA’s potential relationship to postoperative pain. For this analysis, the outcome is continuous rather than binary, and will apply to all 1,300 patients. It will be reasonable to include up to 25 pre-specified candidate predictor variables in the linear regression models, including interaction terms. As risk factors for delirium and pain are overlapping, the same candidate predictor variables will be used in this regression. Sensitivity analyses will be conducted to address limitations regarding pain. Since it is important to consider delirious patients might be unable to report pain accurately, we plan to conduct a sensitivity analysis with pain as the outcome, excluding all the patients who were diagnosed with postoperative delirium. Additionally, since our primary analysis will not consider duration of severe pain or distinguish between rest and provoked pain, we plan to conduct a sensitivity analysis with median provoked pain during hospital stay (up to postoperative day 3) as the outcome. The responses to two VAS questions (pain when (i) taking a deep breath or coughing, and (ii) moving (sitting up, walking, or moving extremities)) will be compiled to represent provoked pain during hospital stay.\n\nWe expect that patients with OSA will experience greater postoperative pain severity, and have a higher risk for postoperative delirium following surgical procedures. For our secondary analyses, we propose that these adverse outcomes might be modified by compliance with CPAP treatment. We predict patients with diagnosed OSA who do not use prescribed CPAP will experience a higher incidence of delirium and increased pain. We also expect a step-wise increase in these adverse outcomes (delirium incidence and pain severity) when analyzing patients based on their STOP-Bang assessment groups (high risk vs. intermediate risk vs. low risk).\n\n\nDiscussion\n\nOSA is a common and frequently undiagnosed perioperative problem. This observational study will help to clarify whether or not OSA is an independent predictor of postoperative incident delirium and acute postoperative pain. Secondary analyses may show if these adverse outcomes might be modified by compliance with OSA treatment.\n\nIn this study, we will attempt to replicate the reported finding showing that OSA is an independent predictor of postoperative delirium and acute postsurgical pain severity32–34. This study will have important strengths compared to the existing literature; most notably the database including routine structured preoperative screening for OSA, and postoperative delirium and pain assessments on a broad surgical population. The researchers who collected data for this study were all expertly trained in administering delirium and pain assessments. In an effort to improve methodological rigor, we have pre-specified independent variables for regression models, and have described our statistical analyses.\n\nThis study will also have important limitations. Although we will have thorough medical histories routinely collected from preoperative clinic assessments, we will not know severity of OSA or other comorbidities. In common with any observational study, this study will be unable to distinguish association from causation. In particular, if we do find in this study that OSA is associated with either increased delirium incidence or pain severity, we will not be able to determine (i) whether OSA is causally implicated or (ii) whether there is another explanatory factor associated with both OSA and these outcomes. Regarding the outcome of delirium, this study will address on the crude association with incident delirium as a binary outcome. It might be more important to focus on either the duration or severity of delirium. Regarding pain, it is important to consider that delirious patients might be unable to report pain accurately. This limitation is common to all studies evaluating postoperative pain. To mitigate this to an extent, we plan to conduct a sensitivity analysis with highest VAS pain score as the outcome, excluding all the patients who were diagnosed with postoperative delirium. Also in relation to pain, our primary outcome will be most severe pain reported in postoperative days 1–3. This approach will not consider duration of severe pain or distinguish between rest and provoked pain. To mitigate this to an extent, we plan to conduct a sensitivity analysis exploring median provoked pain through postoperative day 3 as the outcome. Additionally, it will be important to include analgesic medication as potential confounders in the regression analyses, and accurate data on these might not be available.\n\nIn conclusion, while likely providing stronger evidence regarding the impact of OSA on postoperative delirium and pain, this study might also discern interventional strategies for treatment and prevention. For example, in relation to delirium, we could test perioperative delirium prevention bundles in patients with OSA or we could investigate whether preoperative initiation of CPAP treatment decreases this complication. The role of CPAP therapy in relation to improved analgesia should also be clarified. Regarding pain, we could further develop analgesic plans especially for surgical patients with OSA, such as emphasizing regional analgesia or non-opioid analgesics. We could also implement procedures intended to improve the safety of patients with OSA receiving respiratory depressant medications in the perioperative period. With emerging knowledge about biased signaling with opioids68, it is possible that certain opioids (e.g. morphine) are safer than others (e.g., Fentanyl) for patients with OSA in terms of their propensity to provide analgesia rather than to cause respiratory depression. We hope to use the foundational work proposed in this observational study to guide the design of such trials and clinical plans, with the goals of reducing postoperative delirium and acute postoperative pain severity for the large number of patients at risk due to OSA.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Competing interests\n\n\n\nThe authors report no conflicts of interest in conducting this study.\n\n\nGrant information\n\nResearch reported in this publication was supported by the National Center For Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH) under Award Number TL1TR002344. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This research was also supported by grants from the National Heart, Lung, and Blood Institute of the NIH under Award Numbers 5R21HL123666 and 5T35HL007815, as well as the 2017-Washington University School of Medicine Meharry Summer Research Program, Stipend Name: Lilly.\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 and thank other experts and advisors involved in the study including: Angela Mickle, M.S.; Hannah Maybrier, B.S.; Thaddeus Budelier, M.D.; Jamila Burton, B.S.; Jordan Oberhaus, B.S.; Daniel Park, B.S.; Amrita Aranake-Chrisinger, M.D., MSCI; Bradley Fritz, M.D., MSCI; and Mark Willingham, M.D., MSCI; of Washington University School of Medicine, Department of Anesthesiology (660 S Euclid Ave Campus Box 8054, St. Louis, MO, 63110).\n\n\nCompliance and Ethical considerations\n\nAs this study is an observational data analysis of patients enrolled in Satisfy-SOS, ENGAGES, and PODCAST there is no direct burden placed upon the patients in the study and procedures for monitoring exposure compliance are not necessary. There is low risk of breach of confidential health information. However, all data are hosted on a firewall-secured network server, managed by the Department of Anesthesiology, which is securely stored behind two locked doors within the departmental office suite. Only Health Insurance Portability Accountability Act (HIPPA)-trained employees of the Department of Anesthesia or Barnes Jewish Healthcare have access to resources on the private network server. The three parent studies from which the data are being obtained for the current study have all been approved by the Human Research Protection Office (HRPO) at Washington University, and patients enrolled in all three studies provided written informed consent. The HRPO has also provided approval for this current study.\n\n\nReferences\n\nDempsey JA, Veasey SC, Morgan BJ, et al.: Pathophysiology of sleep apnea. Physiol Rev. 2010; 90(1): 47–112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYoung T, Peppard PE, Gottlieb DJ: Epidemiology of obstructive sleep apnea: A population health perspective. Am J Respir Crit Care Med. 2002; 165(9): 1217–1239. PubMed Abstract | Publisher Full Text\n\nPeppard PE, Young T, Barnet JH, et al.: Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013; 177(9): 1006–1014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimpson L, Hillman DR, Cooper MN, et al.: High prevalence of undiagnosed obstructive sleep apnoea in the general population and methods for screening for representative controls. Sleep Breath. 2013; 17(3): 967–973. PubMed Abstract | Publisher Full Text\n\nLee W, Nagubadi S, Kryger MH, et al.: Epidemiology of obstructive sleep apnea: A population-based perspective. Expert Rev Respir Med. 2008; 2(3): 349–364. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFinkel KJ, Searleman AC, Tymkew H, et al.: Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center. Sleep Med. 2009; 10(7): 753–758. PubMed Abstract | Publisher Full Text\n\nLockhart EM, Willingham MD, Abdallah AB, et al.: Obstructive sleep apnea screening and postoperative mortality in a large surgical cohort. Sleep Med. 2013; 14(5): 407–415. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrey WC, Pilcher J: Obstructive sleep-related breathing disorders in patients evaluated for bariatric surgery. Obes Surg. 2003; 13(5): 676–683. PubMed Abstract | Publisher Full Text\n\nO'Keeffe T, Patterson EJ: Evidence supporting routine polysomnography before bariatric surgery. Obes Surg. 2004; 14(1): 23–26. PubMed Abstract | Publisher Full Text\n\nBrown KA: Intermittent hypoxia and the practice of anesthesia. Anesthesiology. 2009; 110(4): 922–927. PubMed Abstract | Publisher Full Text\n\nFassbender P, Herbstreit F, Eikermann M, et al.: Obstructive Sleep Apnea-a Perioperative Risk Factor. Dtsch Arztebl Int. 2016; 113(27–28): 463–469. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYaggi HK, Concato J, Kernan WN, et al.: Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med. 2005; 353(19): 2034–2041. PubMed Abstract | Publisher Full Text\n\nPeppard PE, Young T, Palta M, et al.: Prospective study of the association between sleep-disordered breathing and hypertension. N Engl J Med. 2000; 342(19): 1378–1384. PubMed Abstract | Publisher Full Text\n\nPedrosa RP, Barros IML, Drager LF, et al.: OSA is common and independently associated with hypertension and increased arterial stiffness in consecutive perimenopausal women. Chest. 2014; 146(1): 66–72. PubMed Abstract | Publisher Full Text\n\nTung P, Levitzky YS, Wang R, et al.: Obstructive and central sleep apnea and the risk of incident atrial fibrillation in a community cohort of men and women. J Am Heart Assoc. 2017; 6(7): pii: e004500. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCadby G, McArdle N, Briffa T, et al.: Severity of OSA is an independent predictor of incident atrial fibrillation hospitalization in a large sleep-clinic cohort. Chest. 2015; 148(4): 945–952. PubMed Abstract | Publisher Full Text\n\nSaruhara H, Takata Y, Usui Y, et al.: Obstructive sleep apnea as a potential risk factor for aortic disease. Heart Vessels. 2012; 27(2): 166–173. PubMed Abstract | Publisher Full Text\n\nNaito R, Sakakura K, Kasai T, et al.: Aortic dissection is associated with intermittent hypoxia and re-oxygenation. Heart Vessels. 2012; 27(3): 265–270. PubMed Abstract | Publisher Full Text\n\nMariman A, Delesie L, Tobback E, et al.: Undiagnosed and comorbid disorders in patients with presumed chronic fatigue syndrome. J Psychosom Res. 2013; 75(5): 491–496. PubMed Abstract | Publisher Full Text\n\nIsmail K, Roberts K, Manning P, et al.: OSA and pulmonary hypertension: Time for a new look. Chest. 2015; 147(3): 847–861. PubMed Abstract | Publisher Full Text\n\nJavaheri S, Javaheri S, Javaheri A: Sleep apnea, heart failure, and pulmonary hypertension. Curr Heart Fail Rep. 2013; 10(4): 315–320. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPunjabi NM, Beamer BA: Alterations in glucose disposal in sleep-disordered breathing. Am J Respir Crit Care Med. 2009; 179(3): 235–240. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang T, Eskandari D, Zou D, et al.: Increased carbonic anhydrase activity is associated with sleep apnea severity and related hypoxemia. Sleep. 2015; 38(7): 1067–1073. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYuan H, Pinto SJ, Huang J, et al.: Ventilatory responses to hypercapnia during wakefulness and sleep in obese adolescents with and without obstructive sleep apnea syndrome. Sleep. 2012; 35(9): 1257–1267. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLam KK, Kunder S, Wong J, et al.: Obstructive sleep apnea, pain, and opioids: Is the riddle solved? Curr Opin Anaesthesiol. 2016; 29(1): 134–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVasu TS, Grewal R, Doghramji K: Obstructive sleep apnea syndrome and perioperative complications: A systematic review of the literature. J Clin Sleep Med. 2012; 8(2): 199–207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta RM, Parvizi J, Hanssen AD, et al.: Postoperative complications in patients with obstructive sleep apnea syndrome undergoing hip or knee replacement: A case-control study. Mayo Clin Proc. 2001; 76(9): 897–905. PubMed Abstract | Publisher Full Text\n\nKaw R, Pasupuleti V, Walker E, et al.: Postoperative complications in patients with obstructive sleep apnea. Chest. 2012; 141(2): 436–441. PubMed Abstract | Publisher Full Text\n\nHwang D, Shakir N, Limann B, et al.: Association of sleep-disordered breathing with postoperative complications. Chest. 2008; 133(5): 1128–1134. PubMed Abstract | Publisher Full Text\n\nKaw R, Golish J, Ghamande S, et al.: Incremental risk of obstructive sleep apnea on cardiac surgical outcomes. J Cardiovasc Surg (Torino). 2006; 47(6): 683–689. PubMed Abstract\n\nLiao P, Yegneswaran B, Vairavanathan S, et al.: Postoperative complications in patients with obstructive sleep apnea: a retrospective matched cohort study. Can J Anaesth. 2009; 56(11): 819–828. PubMed Abstract | Publisher Full Text\n\nHaack M, Lee E, Cohen DA, et al.: Activation of the prostaglandin system in response to sleep loss in healthy humans: potential mediator of increased spontaneous pain. Pain. 2009; 145(1–2): 136–141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith MT, Klick B, Kozachik S, et al.: Sleep onset insomnia symptoms during hospitalization for major burn injury predict chronic pain. Pain. 2008; 138(3): 497–506. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoehrs T, Hyde M, Blaisdell B, et al.: Sleep loss and REM sleep loss are hyperalgesic. Sleep. 2006; 29(2): 145–151. PubMed Abstract | Publisher Full Text\n\nMeagher DJ, Morandi A, Inouye SK, et al.: Concordance between DSM-IV and DSM-5 criteria for delirium diagnosis in a pooled database of 768 prospectively evaluated patients using the delirium rating scale-revised-98. BMC Med. 2014; 12: 164. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWhitlock EL, Vannucci A, Avidan MS: Postoperative delirium. Minerva Anestesiol. 2011; 77(4): 448–456. PubMed Abstract | Free Full Text\n\nInouye SK: Delirium in older persons. N Engl J Med. 2006; 354(11): 1157–1165. PubMed Abstract | Publisher Full Text\n\nKoster S, Hensens AG, Schuurmans MJ, et al.: Consequences of delirium after cardiac operations. Ann Thorac Surg. 2012; 93(3): 705–711. PubMed Abstract | Publisher Full Text\n\nScholz AF, Oldroyd C, McCarthy K, et al.: Systematic review and meta-analysis of risk factors for postoperative delirium among older patients undergoing gastrointestinal surgery. Br J Surg. 2016; 103(2): e21–8. PubMed Abstract | Publisher Full Text\n\nFlink BJ, Rivelli SK, Cox EA, et al.: Obstructive sleep apnea and incidence of postoperative delirium after elective knee replacement in the nondemented elderly. Anesthesiology. 2012; 116(4): 788–796. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMirrakhimov AE, Yen T, Kwatra MM: Delirium after cardiac surgery: Have we overlooked obstructive sleep apnea?. Med Hypotheses. 2013; 81(1): 15–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrueger JM, Clinton JM, Winters BD, et al.: Involvement of cytokines in slow wave sleep. Prog Brain Res. 2011; 193: 39–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChennaoui M, Sauvet F, Drogou C, et al.: Effect of one night of sleep loss on changes in tumor necrosis factor alpha (TNF-α) levels in healthy men. Cytokine. 2011; 56(2): 318–324. PubMed Abstract | Publisher Full Text\n\nKrueger JM: The role of cytokines in sleep regulation. Curr Pharm Des. 2008; 14(32): 3408–3416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKawasaki Y, Zhang L, Cheng JK, et al.: Cytokine mechanisms of central sensitization: Distinct and overlapping role of interleukin-1beta, interleukin-6, and tumor necrosis factor-alpha in regulating synaptic and neuronal activity in the superficial spinal cord. J Neurosci. 2008; 28(20): 5189–5194. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIrwin MR, Wang M, Campomayor CO, et al.: Sleep deprivation and activation of morning levels of cellular and genomic markers of inflammation. Arch Intern Med. 2006; 166(16): 1756–1762. PubMed Abstract | Publisher Full Text\n\nVgontzas AN, Zoumakis E, Bixler EO, et al.: Adverse effects of modest sleep restriction on sleepiness, performance, and inflammatory cytokines. J Clin Endocrinol Metab. 2004; 89(5): 2119–2126. PubMed Abstract | Publisher Full Text\n\nRen K, Dubner R: Interactions between the immune and nervous systems in pain. Nat Med. 2010; 16(11): 1267–1276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith MT, Edwards RR, McCann UD, et al.: The effects of sleep deprivation on pain inhibition and spontaneous pain in women. Sleep. 2007; 30(4): 494–505. PubMed Abstract | Publisher Full Text\n\nKhalid I, Roehrs TA, Hudgel DW, et al.: Continuous positive airway pressure in severe obstructive sleep apnea reduces pain sensitivity. Sleep. 2011; 34(12): 1687–1691. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoufas AG, Tian L, Padrez KA, et al.: Experimental pain and opioid analgesia in volunteers at high risk for obstructive sleep apnea. PLoS One. 2013; 8(1): e54807. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrown KA, Laferrière A, Lakheeram I, et al.: Recurrent hypoxemia in children is associated with increased analgesic sensitivity to opiates. Anesthesiology. 2006; 105(4): 665–669. PubMed Abstract | Publisher Full Text\n\nAvidan MS, Fritz BA, Maybrier HR, et al.: The prevention of delirium and complications associated with surgical treatments (PODCAST) study: Protocol for an international multicentre randomised controlled trial. BMJ Open. 2014; 4(9): e005651. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWildes TS, Winter AC, Maybrier HR, et al.: Protocol for the electroencephalography guidance of anesthesia to alleviate geriatric syndromes (ENGAGES) study: A pragmatic, randomised clinical trial. BMJ Open. 2016; 6(6): e011505. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHelsten DL, Ben Abdallah A, Avidan MS, et al.: Methodologic considerations for collecting patient-reported outcomes from unselected surgical patients. Anesthesiology. 2016; 125(3): 495–504. PubMed Abstract | Publisher Full Text\n\nKronzer VL, Ben Abdallah A, McKinnon SL, et al.: Ability of preoperative falls to predict postsurgical outcomes in non-selected patients undergoing elective surgery at an academic medical centre: Protocol for a prospective cohort study. BMJ Open. 2016; 6(9): e011570. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKronzer VL, Jerry MR, Avidan MS: Assessing change in patient-reported quality of life after elective surgery: Protocol for an observational comparison study. F1000Res. 2016; 5: 976. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenchimol EI, Smeeth L, Guttmann A, et al.: The REporting of studies conducted using observational routinely-collected health data (RECORD) statement. Z Evid Fortbild Qual Gesundhwes. 2016; 115–116: 33–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvon Elm E, Altman DG, Egger M, et al.: The strengthening the reporting of observational studies in epidemiology (STROBE) statement: Guidelines for reporting observational studies. Int J Surg. 2014; 12(12): 1495–1499. PubMed Abstract | Publisher Full Text\n\nPLOS Medicine Editors: Observational studies: Getting clear about transparency. PLoS Med. 2014; 11(8): e1001711. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarcantonio ER, Ngo LH, O'Connor M, et al.: 3D-CAM: Derivation and validation of a 3-minute diagnostic interview for CAM-defined delirium: a cross-sectional diagnostic test study. Ann Intern Med. 2014; 161(8): 554–561. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarcantonio ER, Goldman L, Mangione CM, et al.: A clinical prediction rule for delirium after elective noncardiac surgery. JAMA. 1994; 271(2): 134–139. PubMed Abstract | Publisher Full Text\n\nEly EW, Margolin R, Francis J, et al.: Evaluation of delirium in critically ill patients: validation of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). Crit Care Med. 2001; 29(7): 1370–1379. PubMed Abstract | Publisher Full Text\n\nHawker GA, Mian S, Kendzerska T, et al.: Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP)). Arthritis Care Res (Hoboken). 2011; 63(Suppl 11): S240–52. PubMed Abstract | Publisher Full Text\n\nFerreira-Valente MA, Pais-Ribeiro JL, Jensen MP: Validity of four pain intensity rating scales. Pain. 2011; 152(10): 2399–2404. PubMed Abstract | Publisher Full Text\n\nChung SA, Yuan H, Chung F: A systemic review of obstructive sleep apnea and its implications for anesthesiologists. Anesth Analg. 2008; 107(5): 1543–1563. PubMed Abstract | Publisher 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\n\nSchmid CL, Kennedy NM, Ross NC, et al.: Bias Factor and Therapeutic Window Correlate to Predict Safer Opioid Analgesics. Cell. 2017; 171(5): 1165–1175.e13. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "32028",
"date": "29 Mar 2018",
"name": "Jean Wong",
"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 study proposes to retrospectively examine data previously collected from 3 prospective studies to test whether OSA is an independent predictor of postoperative delirium and acute postoperative pain severity.\n\nThe study is described as ‘observational’ however, the study examines data that was already collected, and so I believe the title should be adjusted to reflect this a retrospective study.\n\nThe incidence of delirium varies depending on the type of surgery and is lower than 20% for some elective surgeries. The type of surgery included should be mentioned.\n\nWhether the patients may have a history of chronic pain should be included. As well, although the patients had a history of using CPAP, ideally, whether the patients were compliant with the use of CPAP while in hospital and the number of hours of CPAP use should be reported.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "3704",
"date": "15 Jun 2018",
"name": "Patricia Strutz",
"role": "Author Response",
"response": "Thank you, Dr. Wong and Dr. Chung, for your feedback on our study protocol. We sincerely appreciate your time and thoughtful suggestions. Along with submitting our revised protocol, we would like to directly address some suggestions. We will be including type of surgery as an independent variable in our regression models, and we will also be adjusting our analysis for a history of chronic pain. We agree with your suggestion to report hospital CPAP use. Unfortunately, we do not have that information available to us and are unable to provide a more robust measure of CPAP adherence- a limitation we would discuss in our manuscript. Again, thank you for your time."
}
]
},
{
"id": "33092",
"date": "03 May 2018",
"name": "Matthias Eikermann",
"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 submitting this interesting study protocol aiming to investigate the relationship between OSA and postoperative delirium as well as pain severity. This protocol targets an important and clinically meaningful research question, and is overall well written and designed. We invite the authors to consider the following questions and suggestions:\n\nPlease address the question as to whether the patients are studied from a historical cohort. If so, how do you control for the effects of previously studied interventions (e.g. effect of anesthetic protocols on postoperative delirium in the ENGAGES study; effects of ketamine treatment on postoperative delirium and pain (PODCAST study)) ?\n\nHow do you discriminate between lingering medication effects and delirium early on during postoperative day 1?\n\nWe suggest considering to additionally use the SPOSA score which was recently published in order to classify patients according to OSA risk (Shin C et al. BMC Anesthesiology 2017). Advantage is you can apply this to bigger observational cohorts without measuring neck circumference etc.\n\nMaybe you can better justify the patient flow. The SATISFY-SOS study aims to enroll 36000 patients. Please provide information that helps understand as to why you expect to have complete data for ~1300 patients only. Looks like you also want to use data from two other studies additionally with enrollment targets of additional 600 and 1200 patients?\n\nHow do you account for patients with undiagnosed OSA that may have been treated with CPAP for other respiratory diseases such as COPD, acute lung injury, neuromuscular disorders etc.?\n\nStatistical models:\nConfounder control: Please consider including comorbidities such as COPD, asthma and neuromuscular disorders in the context of higher risk for intraoperative hypoxemia or respiratory failure; duration of surgery/time under anesthesia in addition to surgery type and some type of measure for procedural complexity such as a risk quantification score (see PSS for Morbidity or Mortality) into your statistical model For sure you should control for age and BMI Consider including pre-prescribed drugs especially neuroleptics/antipsychotics if you can get these data Perhaps also consider accounting for different anesthetic protocols used during surgery such as TIVAs, volatile anesthetics, opioid use, ketamine use Did you consider using multiple imputation methods for missing data? We like the suggested sensitivity analyses.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": [
{
"c_id": "3705",
"date": "15 Jun 2018",
"name": "Patricia Strutz",
"role": "Author Response",
"response": "Thank you, Dr. Eikermann and Dr. Timm, for your feedback on our study protocol. We sincerely appreciate your time and thoughtful suggestions. Along with submitting our revised protocol, we directly address your suggestions and questions below. Please address the question as to whether the patients are studied from a historical cohort. If so, how do you control for the effects of previously studied interventions (e.g. effect of anesthetic protocols on postoperative delirium in the ENGAGES study; effects of ketamine treatment on postoperative delirium and pain (PODCAST study)) ? We have included more details about our cohort under the “Study Design” section. Briefly, yes the patients are studied from a historical cohort, and we have improved our model to adjust for randomization group from the previously studied interventions. Although the PODCAST trial did not have significant findings, we agree that controlling for group allocation provides a more robust analysis. How do you discriminate between lingering medication effects and delirium early on during postoperative day 1? We have added details regarding our delirium assessments under “Data collection: Delirium assessment method.” All POD 1 assessments were administered in the afternoon between 1pm and 8pm. Patients were assessed on POD 0; however we are excluding these assessments because of potential lingering general anesthetic effects. We are also now adjusting our model for the use of certain medications, such as preoperative midazolam, median volatile anesthetic concentration (converted to minimum alveolar concentration [MAC] equivalents), intraoperative ketamine, and intraoperative opioids (converted to morphine equivalents in mg). We suggest considering to additionally use the SPOSA score which was recently published in order to classify patients according to OSA risk (Shin C et al. BMC Anesthesiology 2017). Advantage is you can apply this to bigger observational cohorts without measuring neck circumference etc. Thank you for this suggestion! We are unable to incorporate the SPOSA score for this cohort because of limiting factors with electronic medical data. However, we are excited about this new model and the ability to classify OSA risk in larger cohorts, especially when neck circumference is unavailable. Maybe you can better justify the patient flow. The SATISFY-SOS study aims to enroll 36000 patients. Please provide information that helps understand as to why you expect to have complete data for ~1300 patients only. Looks like you also want to use data from two other studies additionally with enrollment targets of additional 600 and 1200 patients? We have included more details about patient flow under the “Study Design” section. We now hope to include all 1200 patients from the ENGAGES trial. Of the ~600 patients in the PODCAST study, we will include those patients recruited through Washington University in St. Louis (roughly 100 patients). Although SATISFY-SOS aims to enroll a large number of patients, only 200 of those patients received daily inpatient delirium and pain assessments. How do you account for patients with undiagnosed OSA that may have been treated with CPAP for other respiratory diseases such as COPD, acute lung injury, neuromuscular disorders etc.? The data to classify patient reported CPAP use comes from routine questions that are part of our OSA screening done in our pre-operative clinic. Specifically, we ask patients who have been diagnosed with OSA if they are prescribed CPAP (or an alternative OSA PAP treatment) and if they use their CPAP. Unfortunately in this study, we are unable to account for patients with undiagnosed OSA who may have been treated with CPAP for other respiratory diseases. This has the potential to decrease any difference in outcomes we may see between the study groups and bias our findings towards the null hypothesis (H0= no difference in outcomes between the groups). This would be a limitation of our secondary sub-group analysis, which we would discuss in our manuscript. Statistical models: Confounder control: Please consider including comorbidities such as COPD, asthma and neuromuscular disorders in the context of higher risk for intraoperative hypoxemia or respiratory failure; duration of surgery/time under anesthesia in addition to surgery type and some type of measure for procedural complexity such as a risk quantification score (see PSS for Morbidity or Mortality) into your statistical model For sure you should control for age and BMI Consider including pre-prescribed drugs especially neuroleptics/antipsychotics if you can get these data Perhaps also consider accounting for different anesthetic protocols used during surgery such as TIVAs, volatile anesthetics, opioid use, ketamine use Did you consider using multiple imputation methods for missing data? We like the suggested sensitivity analyses. Thank you for these suggestions- they have helped us improve our statistical models. We will include COPD, asthma, and anesthesia time in minutes; however we will not be able to include history of neuromuscular disorders or pre-prescribed drugs. Additionally, we will adjust our models for medications used during surgery. We are still hopeful to include age and BMI, but collinearity may prevent us from incorporating these variables independently. We will explain any statistical limitations in the manuscript. Again, thank you for your time and your suggestions!"
}
]
}
] | 1
|
https://f1000research.com/articles/7-328
|
https://f1000research.com/articles/7-233/v1
|
27 Feb 18
|
{
"type": "Research Article",
"title": "Predicting ionizing radiation exposure using biochemically-inspired genomic machine learning",
"authors": [
"Jonathan Z.L. Zhao",
"Eliseos J. Mucaki",
"Peter K. Rogan",
"Jonathan Z.L. Zhao",
"Eliseos J. Mucaki"
],
"abstract": "Background: Gene signatures derived from transcriptomic data using machine learning methods have shown promise for biodosimetry testing. These signatures may not be sufficiently robust for large scale testing, as their performance has not been adequately validated on external, independent datasets. The present study develops human and murine signatures with biochemically-inspired machine learning that are strictly validated using k-fold and traditional approaches. Methods: Gene Expression Omnibus (GEO) datasets of exposed human and murine lymphocytes were preprocessed via nearest neighbor imputation and expression of genes implicated in the literature to be responsive to radiation exposure (n=998) were then ranked by Minimum Redundancy Maximum Relevance (mRMR). Optimal signatures were derived by backward, complete, and forward sequential feature selection using Support Vector Machines (SVM), and validated using k-fold or traditional validation on independent datasets. Results: The best human signatures we derived exhibit k-fold validation accuracies of up to 98% (DDB2, PRKDC, TPP2, PTPRE, and GADD45A) when validated over 209 samples and traditional validation accuracies of up to 92% (DDB2, CD8A, TALDO1, PCNA, EIF4G2, LCN2, CDKN1A, PRKCH, ENO1, and PPM1D) when validated over 85 samples. Some human signatures are specific enough to differentiate between chemotherapy and radiotherapy. Certain multi-class murine signatures have sufficient granularity in dose estimation to inform eligibility for cytokine therapy (assuming these signatures could be translated to humans). We compiled a list of the most frequently appearing genes in the top 20 human and mouse signatures. More frequently appearing genes among an ensemble of signatures may indicate greater impact of these genes on the performance of individual signatures. Several genes in the signatures we derived are present in previously proposed signatures. Conclusions: Gene signatures for ionizing radiation exposure derived by machine learning have low error rates in externally validated, independent datasets, and exhibit high specificity and granularity for dose estimation.",
"keywords": [
"Ionizing Radiation Exposure",
"Machine Learning",
"Gene Signatures",
"Molecular Diagnostics",
"Validation",
"Biodosimetry",
"Support Vector Machine",
"Minimum Redundancy Maximum Relevance"
],
"content": "Introduction\n\nPotential radiation exposures from industrial nuclear accidents, military incidents, or terrorism are threats to public health1. There is a need for large scale biodosimetry testing, which requires efficient screening techniques to differentiate exposed individuals from non-exposed individuals and to determine the severity of exposure2. Current diagnostic techniques, including the cytogenetic gold standard3–6, may require several days to provide accurate dose estimates1,7 of large cohorts. To address the need for faster diagnostic techniques that accurately measure radiation exposures, gene signatures based on transcriptomic data have been introduced7–10. Probit regression models of radiation response using 25 probes on peripheral blood samples achieved up to 90% accuracy for distinguishing between irradiated blood samples and unirradiated controls9. A 74-gene classifier based on nearest centroid expression levels was 98% accurate in distinguishing four levels of irradiation from controls10. This level of performance implies that samples exposed to different levels of radiation may be distinguishable based on mRNA expression levels of different genes. While this suggests the feasibility of transcriptional modeling of radiation responses, validation with external datasets is required to establish its reliability for rapid diagnostics. A caveat of these signatures is that they have not all been externally validated on datasets independent of the source data used for model development. A 29-gene signature modelled using a support vector machine (SVM) was externally validated on such a dataset, resulting in 80% accuracy in distinguishing higher (≥8Gy) from lower dose (≤2Gy) radiation exposure in novel samples7. The present study derives signatures with improved performance on externally validated samples by employing a different selection of modelling techniques. The machine learning pipeline used here addresses some of the previous limitations through a more rigorous feature selection process and stricter validation procedures.\n\nPreviously, the Student’s t-test7, the F-test10, and correlation coefficients9 were used to identify potential radiation biomarker genes. Although statistical criteria can distinguish genes that are differentially expressed upon radiation exposure, they do not eliminate expressed genes with redundant responses to radiation exposure. Redundancy increases the possibility of overfitting, thereby reducing the generalizability of these models to predict responses in independent datasets. We address this limitation with the information theory-based criterion for gene selection known as minimum redundancy maximum relevance (mRMR)11–13, which ranks genes according to shared mutual information between expression levels and radiation dose (relevance), and by minimizing mutual information shared by expression values of these and other genes (redundancy)11,12. mRMR outperforms ranking criteria based solely on maximizing relevance12. In contrast with heuristic approaches like differential expression, we only consider genes with evidence of a relationship to radiation response, which significantly limits the number of model features. Biochemically-inspired genomic machine learning (ML) has been used to derive high performing gene signatures that predict chemotherapy and hormone therapy responses13–15. From an initial set of mRMR-derived biochemically relevant genes, wrapper approaches for feature selection16 are used to find an optimal set of genes that predict exposure to radiation.\n\nIt can be challenging to obtain highly accurate models that perform well on externally validated samples for several reasons. Aside from biases in training data, batch effects and lack of reproducibility may introduce systematic and random sources of variability into gene expression microarray data. Different source datasets can impact data normalization, reducing model performance. We utilize two validation procedures. The first is a signature-centric approach that mirrors external k-fold validation7. The limitation of signature-centric validation is that, while signatures allow for the identification of important genes associated with radiation response, a tangible model is required to generate actual diagnostic predictions. To address this limitation, we also use a second model-centric approach, which we term “traditional validation”. This procedure applies quantile normalization to training and test data before a model is fitted to the training data. This quantile method has been shown to be more effective than scaling, loess, contrast, and non-linear methods in reducing variation between microarray data17. Model validation was not expected to perform as well as signature validation, because quantile normalization is not always successful in eliminating variation between microarray datasets, whereas k-fold validation is independent of this source of variation. This study shows that robust model validation is a critical step in reproducibly predicting which individuals have been exposed to significant levels of radiation.\n\n\nMethods\n\nMurine gene expression datasets18 were obtained from peripheral blood (PB) mononuclear cell samples of ten-week old C57B16 mice that either received total body radiation at 50 cGy, 200 cGy, or 1000 cGy or were not exposed. Post-exposure, total RNA was isolated after 6 hours and expression was determined by microarray analysis using Operon Mouse V3.0.1 (Gene Expression Omnibus (GEO): GPL4783 from GSE10640[GPL4783])19 and Operon Mouse V4.0 arrays (GEO: GPL6524 from GSE10640[GPL6524])19. Similar analyses were performed with human expression microarrays18, including datasets GEO: GSE6874[GPL4782]9, GSE10640[GPL6522]19, GSE172520, and GSE70121. GSE6874 and GSE10640 consist of PB samples collected 6 hours post-exposure from healthy donors and patients undergoing total body irradiation at 150–200 cGy analyzed with Operon Human V3.0.2 (GEO: GPL4782) and Operon Human V4.0 (GEO: GPL6522) microarrays. GSE10640[GPL6522] consists of 32 patients treated with alkylator-based chemotherapy without radiation. GSE1725 contains lymphoblastoid cell line samples derived from 57 subjects treated with 500 cGy. RNA was extracted 4 hours after exposure. Expression was measured using Affymetrix Human Genome U95 Version 2 Array (GEO: GPL8300). GSE701 contains lymphoblastoid cell lines from Fondation Jean Dausset-CEPH which were irradiated at 300 cGy or 1000 cGy and extracted 1–24 hours after exposure. Expression was measured using the Affymetrix Human Genome U95A Array (GEO: GPL91).\n\nRows and columns of microarray data that are less than 95% complete were removed and any remaining missing values were imputed using the nearest-neighbor algorithm. Only genes that are common across all datasets have been retained. Expression values of each probe were transformed to z-scores and the mean expression value of probes for the same gene have been assigned as the expression of each gene.\n\nIn panel (v), k-fold validation splits data into k sections, where each section acts as a test set in turn while the remaining sections act as a training set. Panel (v) depicts k-fold validation for k = 3. Coloured circles represent the samples in a dataset where different colours represent different radiation doses. In panel (vi), quantile normalization forces data into the same distribution. To demonstrate this, thirty random genes were chosen to form a signature. The histograms on the left represent the distributions of expression levels of these genes in the pre-normalized datasets GSE1725 and GSE6874[GPL4782]. The histograms on the right represent the distribution of expression levels of the same genes post-normalization.\n\nA literature search has been conducted to identify genes implicated in radiation response using the search queries “radiation genes,” “radiation response genes,” and “radiation signatures” on PubMed. Cited genes comprise those differentially expressed after radiation exposure, genes present in DNA repair databases and other radiation signatures, and evolutionarily conserved genes that were highly expressed in radio-resistant species. A list of 998 genes was compiled22–35, Supplementary Table X) for deriving signatures.\n\nRank is assigned by incremental selection of genes based on the mutual information difference (MID) criterion11,12. Highly ranked genes have expression information that shares mutual information with radiation exposure and shares little information with expression of other genes. The MID criterion used to select the next ranked gene is maxi∈Ω[I(i,h)−1|S|∑j∈SI(i,j)], where i is a gene selected from Ω, the total gene space, S is the set of genes selected before i, |S| is the number of genes selected before i, I(i,h) is the mutual information between expression of gene i and radiation dose (h), and I(i,j) is the mutual information between expression of gene i and expression of gene j.\n\nSVM models are classifiers that use hyperplane boundaries to separate samples into exposure classes by maximizing the distance between the separating hyperplanes and samples of each class. The fitcecoc function of MATLAB 2017a’s Statistics and Machine Learning Toolbox36 with a SVM template was used to fit SVM models to training data. The fitcecoc function was used because it allows the fitting of multiclass models, which was required for analysis of murine samples that were irradiated at four different exposure levels. The SVM models use the Gaussian radial basis function kernel and a range of selected box-constraint and kernel-scale parameters. The box-constraint, denoted by the variable C, determines how severely misclassifications are penalized during training. The kernel-scale, denoted σ, represents the width of the Gaussian radial basis function. These parameters collectively control the tradeoff between underfitting and overfitting37. After feature selection, a grid search is performed to determine the optimal (C,σ) combination for values of C and σ between 1 and 100000 (inclusive) by powers of 10 such that C ≥ σ.\n\nGreedy feature selection was used to derive signatures. Complete sequential feature selection (CSFS) sequentially adds genes to an initially empty base set. The added gene is the highest mRMR-ranked gene that is not already included. This is repeated until all genes have been evaluated and the best performing subset of genes is identified. Forward sequential feature selection (FSFS) sequentially adds genes from the top 50 mRMR ranked genes to an initially empty base set. The added gene is the one whose addition improves the model by the greatest margin. Backward sequential feature selection (BSFS) sequentially removes genes from the top 30 mRMR ranked genes. The gene removed is the one whose removal causes the greatest improvement in the model. For BSFS and FSFS, we measure model improvement using misclassification or log loss during k-fold validation (see Performance metrics section below). Genes are added or removed until model performance plateaus. During feature selection, C and σ parameters need to be chosen for SVM learning (see SVM Learning section above). Thus, each signature is characterized by the feature selection algorithm used, the dataset used to derive it, and the C-σ combination used for its SVM models during feature selection. This leads to a large number of possible signatures (see Supplementary Files Y1–Y7). Supplementary Files Y1–Y3 and Supplementary Files Y6–Y7 contain k-fold validation results from which the top 20 signatures (evaluated using average validation log loss), in particular, were analyzed (Figure 2, Figure 3, Figure 6, Figure 7).\n\nGene frequency values are first scaled within datasets and then scaled across datasets to ensure values between 0 and 1.\n\nThe size of each circle is proportional to the frequency at which the gene appears in the top 20 murine signatures ranked by log loss averaged over GSE6874[GPL4783] and GSE10640[GPL6524]. The genes presented match those of the Figure 2.\n\nStratified k-fold validation was used to validate signatures. Samples of the validation dataset were partitioned into k sets, comprised of an approximately equal distribution of radiation levels. For validation, each set was used to test a model trained on the remaining sets, resulting in predictions for all samples in the dataset. Advantages of this approach are that variation between datasets is not pertinent and that signatures can be validated on differently labeled datasets (with samples irradiated at different levels).\n\nModel validation requires separate training and test datasets (the training set is often used for FS). Genes from the signature are extracted from the training and test sets and their expression values are quantile normalized by sample. An important distinction between our approach and a previous study7 is that quantile normalization is applied immediately before validation, so expression of only the genes present in the signature being validated have been normalized. By contrast, previous approaches perform quantile normalization over entire datasets; while this reduces variability in expression values within datasets, it also suppresses the dynamic range, with potential consequential effects on the prognostic value of expression data. After normalization, an SVM model was fit to training datasets and used to generate predictions from the test dataset.\n\nPerformance was determined by comparing predicted radiation doses with actual radiation exposures of each sample. Metrics included misclassification error rate, goodness-of-fit, and multi-class log loss. Misclassification is the percentage of samples that were incorrectly classified, goodness-of-fit is the average absolute value difference between predicted radiation exposure and actual radiation exposure, and multi-class log loss is −1N∑i=1N∑j=1Myijlnpij where N is the number of samples, M is the number of class labels, pij is the predicted probability that observation i is in class j, and yij is an indicator variable equal to 1 if sample i is in class j and 0 otherwise.\n\n\nResults\n\nWe discovered radiation gene signatures using the microarray data of human and mouse peripheral blood samples and human lymphoblastoid cell lines, which were validated either according to signature (Figure 1, panel v) or with the respective model (Figure 1, panel vi). The murine data were obtained from a wider range of radiation exposure levels (0 cGy, 50 cGy, 200 cGy, 1000 cGy) than the human whole body radiation datasets, which were binary comparisons of radiation effects (0 cGy vs. 150-200 cGy, 0 cGy vs. 500 cGy, or 300 cGy vs. 700 cGy). This made possible the discovery of murine gene signatures with finer granularity for discriminating individuals exposed to different exposure levels, which is not currently feasible with the human samples.\n\n\nMurine gene signatures\n\nTable 1 displays the murine signatures derived using our pipeline which had the best performance metrics during k-fold validation on an independent dataset. In addition to the signature information, we report the feature selection algorithm (FS Algorithm) used to discover the signature, the internal validation performance metrics (FS Misclassification fraction and FS Log Loss function). Validation performance metrics on external dataset(s) are indicated by the Validation Misclassification fraction, Validation Log Loss function, and Validation goodness of fit or GoF). In the FS Misclass. and FS Log Loss columns, one value is always N/A because signatures are derived by optimizing either misclassification or log loss, but never both. The remaining murine signatures are presented in Supplementary Files Y6 and Supplementary Files Y7.\n\n1FS: Feature Selection 2GoF: Goodness of Fit.\n\nA list of the most consistently appearing genes in the best performing signatures were obtained by pooling the top 20 murine signatures (assessed by validation log loss) from GSE6874[GPL4783] and GSE10640[GPL6524], and respectively collating the top 17 and 19 most frequent genes. The union of these two sets comprises 33 genes displayed within a heat map based on the frequencies of each gene (Figure 2). Surprisingly, the compositions of signatures derived from both datasets are not as similar as one may expect. The genes that appear more frequently in signatures derived from one dataset infrequently appear in the other even though both datasets consisted of the same types of samples irradiated at the same exposure levels.\n\nThe shared mutual information of these expressed genes with radiation dose (Figure 3) indicates whether only high mutual information genes appear in the best signatures or whether some lower mutual information genes may also be selected by our feature selection algorithms. The frequency of each gene among these signatures (represented by diameter of the circle) correlates with the mutual information between expression and radiation dose (ρ = 0.8016). However, it would be an oversimplification to create signatures based solely upon mutual information, since some genes in lower performing signatures exhibit higher mutual information content. Development of accurate signatures requires more than a collection of gene features whose individual expression values share information with radiation dose, since many of these genes may reveal similar information, and redundant machine learning model features. For instance, Bax and Blnk are both common among the best murine signatures, even though Blnk shares much more mutual information with radiation dose than Bax expression. Since Blnk and Bax are involved in completely different pathways – Bax is an inducer of apoptosis38 whereas Blnk is involved in a B-cell antigen receptor signaling pathway required for optimal B-cell development39, they provide different types of information to the overall model. Conversely, we also observe that genes with high information content, such as Ms4a1, may appear less frequently than genes with lower information content, such as Glipr2.\n\nAlthough mRMR prioritizes genes with non-redundant, complementary contributions, subsequent wrapper steps of forward and backward sequential feature selection occur independently of the mRMR ranking. mRMR reduces the list of features considered by these algorithms, but it is possible for only high mutual information genes to be selected for the final signature. Thus, the inclusion of lower mutual information genes, such as Ube2v1 and Urod, reinforces the effectiveness of the mRMR method.\n\nThe cellular roles of these protein products (Figure 2 and Figure 3) demonstrate a variety of pathways and functions (Figure 4), some of which have previously discussed40. These include DNA repair genes (Polk23 and Pold126), inducers of apoptosis (Ei2430, Bax30, and Phlda330), chaperonins (Cct322 and Cct722), cell cycle regulators (Ccng127 and Cdkn1a30), B-cell development genes (Cd79b19 and Blnk19), B-cell antigens (Cd729 and Ms4a119), and a stress-response kinase that inhibits protein synthesis globally (Eif2ak425).\n\nOne of the best murine signatures derived from GSE10640[GPL4783]: Phlda3, Blnk, Bax, Cdkn1a, Cct3, Pold1, Cd79b, Ei24, Eif2ak4, Ccng1, Glipr2, Hexb, Pou2af1, Swap70, Apex1, Ptpn1, Mdm2, Tpst1, Ly6e, Sdcbp consistently achieved <10% misclassification error with SVM parameters C = 10, σ = 10. However, for samples that are incorrectly classified according to this signature, the misclassification percentage does not reveal the actual deviation from the correct dose. The confusion matrix visualizes the prediction accuracy of this signature on GEO: GSE10640[GPL6524] (Figure 5). Indeed, the performance of the matrix shows that the predicted errors for a small fraction of samples deviate from the actual exposures by no more than a single adjacent exposure level. Although the predictions presented in the confusion matrix come from a single iteration of k-fold validation, the standard error associated with misclassification for this signature is extremely low (0.0013) so this confusion matrix is representative of nearly all possible iterations of k-fold validation.\n\nNumerators represent the number of samples in each category while denominators represent the total number of samples that were irradiated at a given exposure level (i.e. is the sum of the number of samples in each row).\n\n\nHuman gene signatures\n\nThe best performing signatures obtained from each human dataset, assessed by k-fold validation, are presented in Table 2. Although four human radiation datasets were available, GSE701 contained only 10 samples, which was insufficient for derivation of a unique gene signature. While k-fold validation removes the requirement for inter-dataset normalization, it assesses the ability of signatures (genes) to predict radiation exposure without tying the signatures to corresponding models. Each signature is characterized by the feature selection algorithm and its validation statistics, which have been averaged over the 3 independent datasets that were excluded from the original data used to derive the signature.\n\nSince traditional validation typically requires separate training and test sets that feature samples irradiated at the same exposure levels, only signatures derived from GEO: GSE6874[GPL4782] and GEO: GSE10640[GPL6522] could be analyzed. Table 3 presents the best human signatures according to this validation approach. This type of external validation is the most challenging due to the variability associated with different microarray experiments and batch effects of different platforms. This potentially explains the lower performance obtained by traditional validation (Table 3) compared with k-fold validation on the same datasets (Table 2). The remaining human signatures are described in Supplementary Files Y1–Y5.\n\nTo determine which human genes are most consistently selected, the most frequently appearing genes (11 or 12 depending on number of equally prevalent genes in different signatures) were compiled from the top 20 human signatures (assessed by lowest average log loss during k-fold validation) from GSE10640[GPL6522], GSE6874[GPL4782], and GSE1725. The union of these three lists indicates the relative frequencies of each gene (Figure 6). Figure 7 visualizes the mutual information of gene expression (Figure 6) shared with radiation dose.\n\nFrequencies are first scaled within and then between datasets to ensure values between 0 and 1.\n\nThe size of each circle is proportional to the frequency at which the gene appears in the top 20 human signatures ranked by average validation log loss from GSE10640[GPL6522], GSE6874[GPL4782], and GSE1725. The genes shown are also are the same as those indicated in Figure 6.\n\nWhile most genes have similar representation in signatures derived from different datasets, GADD45A and DDB2, in particular, are significantly more frequent in those derived from GSE1725 and GSE10640[GPL6522]. GADD45A and DDB2 are present in signatures derived from samples irradiated at different exposures (GADD45A – 500 cGy, DDB2 – 150-200 cGy). This raises questions as to whether these genes have a larger influence on the accuracy of individual signatures and whether their expression is calibrated to radiation exposure levels. Removal of these gene features was performed to address their impact. Genes of interest have been removed from each of the top 20 human signatures derived from various datasets and then the signatures were revalidated excluding these features (Table 4). The difference between the validation metrics preceding and following removal of a gene represents the weight of the gene within a signature. ΔMC, ΔLL, and ΔGoF represent the changes in misclassification, log loss, and goodness of fit, respectively.\n\n*∆GoF is always N/A for the dataset used to derive signatures because GoF is never used as the optimized metric during signature development (see Feature Selection Algorithms section under Methods).\n\n**Unavailable because the top 20 human signatures derived from GSE1725 were all obtained by optimizing log loss rather than misclassification.\n\nGADD45A appears in 14 of the top 20 signatures derived from GSE1725. Of the 14 signatures, 10 were single gene signatures, as GADD45A alone was expected to sufficiently distinguish irradiated from unirradiated samples. In these cases, it was assumed that a null signature would perform as well as a predictor that randomly draws predictions from a uniform distribution of doses. Removal of GADD45A from these 14 signatures, results in an average increase in misclassification, log loss, and goodness of fit by 0.319, 0.368, and 109 cGy, respectively (see Table 4a). In contrast, elimination of BAX, which only appears in 2 of the top 20 signatures derived from GSE1725 and results in an average increase in misclassification, log loss, and goodness of fit by 0.018, 0.147, and 2.95 cGy respectively (Table 4e). Comparing the effects of removing DDB2 (Table 4c) and PRKAB1 (Table 4f) from the top 20 GSE10640[GPL6522] signatures confirms the impact of genes that frequently occur within the most accurate gene signatures.\n\nHowever, the diagnostic contributions of GADD45A and DDB2 expression to the radiation levels at which samples were exposed (500 cGy and 150-200 cGy respectively) are confounding. The effects on model performance resulting from removal of GADD45A from the GSE10640[GPL6522] signatures (Table 4b) versus the GSE1725 signatures (Table 4a) are discordant. ΔMC is higher when GADD45A is removed from GSE1725, but ΔLL is higher when GADD45A is removed from GSE10640[GPL6522]. ΔLL is large when GADD45A is removed from both datasets, which is consistent with the importance of GADD45A at both radiation doses. Indeed, GADD45A expression has been demonstrated to be rapidly induced by radiation levels as low as 2 Gy41. Similar discordance was observed in the feature removal experiments of DDB2 (Table 4c, 4d).\n\nAs was the case with murine signatures, genes appearing in the best human signatures do not necessarily share high mutual information with radiation dose. However, the compositions of the human signatures are dominated by four genes, DDB2, GADD45A, PCNA, and PPM1D, which all share a lot of information with radiation dose (DDB2: 0.55, GADD45A: 0.39, PCNA: 0.51, PPM1D: 0.46). The functions associated with these and less frequently appearing genes are depicted in Figure 8. The pathways and functions represented include keratinocyte differentiation (PRKCH9), induction of apoptosis (BCL2L31 and BAX30), DNA repair (TP53BP123, RAD1724, DDB219, PRKDC23, and PCNA27), actin nucleation (ARPC1B22), and regulation of JNK-p38 (MAPK14) signalling (GADD45A27 and PPMD127). The four common genes belong to the DNA repair and regulating JNK-p38 (MAPK14) pathways, which may imply particular significance to these functions in human response to radiation exposure. Interestingly, GADD45A and PPMD1 are antagonistic, that is, GADD45A activates while PPMD1 inhibits p38.\n\n\nDiscussion\n\nBiochemically inspired genomic signatures of human and murine radiation response exhibit high accuracies in validating independent datasets (98% in k-fold validation, 92% by traditional methods). Some of the human signatures exhibit among the highest specificities reported (e.g. the signature DDB2, CD8A, TALDO1, PCNA, EIF4G2, LCN2, CDKN1A, PRKCH, ENO1, PPM1D) exhibited 92% accuracy when validated on GSE10640[GPL6522]. This dataset contains both radiation therapy patients (150–200 cGy) and controls (0 cGy) which include healthy donors and chemotherapy patients treated with alkylators9. Thus, the signature distinguished radiation-induced and chemotherapy-associated DNA damage.\n\nSome of the best performing signatures consisted of one to three gene features. The first signature in Table 2 contains GADD45A and DDB2, and exhibits a misclassification error rate of 7%. These relatively short signatures have advantages over longer signatures with similar performance. It is more likely that the model can be generalized to a wider spectrum of data, when fewer features are required, and from a practical standpoint, diagnostic tests based on fewer gene expression measurements are less susceptible to experimental error.\n\nBAX, an inducer of apoptosis, was the single gene shared among those frequently appearing in both murine and human signatures. One possible explanation for this is that the mouse datasets featured samples irradiated at four levels while human datasets contained samples irradiated at two levels. Genes selected by multi-class model algorithms may better discriminate radiation dose. Nonetheless, the radiation response pathways of mice are not necessarily similar to those of humans. In fact, Lucas et al. have shown that the murine signatures they developed are not translatable to human samples42. Furthermore, only two genes, including BAX, are shared by the human and murine signatures derived by Dressman et al.42.\n\nNone of the samples exposed to ≥200 cGy are misclassified below this radiation dose based on the multi-class murine signatures (Figure 5). In the future, a similar analyses could be performed in clinical studies of human subjects exposed to different radiation levels, which might prove useful for determining treatment eligibility after exposure to high levels of myelosuppressive radiation43.\n\nA comparison of the most frequently appearing genes in the optimal human (Figure 6) and mouse signatures (Figure 2) with signatures previously derived in other studies reveals little overlap (Table 5). The compositional differences can be attributed to types of samples used for model training, microarray platforms used, and feature selection techniques used in deriving signatures. However, genes consistently selected in optimized signatures in at least three independent studies include BAX, DDB2, GADD45A, LY9, and TRIM22. Expression of these genes is indeed predictive of radiation dose and not a result of noise in individual datasets. An ensemble signature consisting of these genes achieves up to 92.3% accuracy in k-fold validation over 277 samples and up to 81.2% accuracy in traditional validation over 78 samples. The quality of the gene signature is largely determined by the quality and amount of training data used to fit the SVM model. Thus, this level of accuracy is not the upper bound on the performance of an SVM of the ensemble signature. Additional data at exposures with fixed levels of radiation in matched training and testing samples could improve model performance.\n\nEnsemble models should be considered which combine genes discovered in different well-performing signatures. Although the most frequently represented human and murine genes were compiled, genes common to one dataset did not appear equally frequently in signatures from the other. This discordance may possibly result of noise in the different datasets, or perhaps to intrinsic differences between them. Compilation of frequently appearing genes in different datasets may be useful for discovery of consistently represented genes that are incorporated into high-performance signatures.\n\nThe types of data available for this study and the analytical approaches we used potentially limited the interpretation of these gene signatures. Blood samples of mouse and human datasets were all collected within 24 hours of exposure. Thus, signatures derived on these datasets may only be valid in white blood cells with a limited time window (<24 hours). Additionally, one of the datasets we used to derive signatures, GSE6874, appears to have been a particularly noisy dataset, based on the average misclassification rates on GSE10640, GSE1725, and GSE6874 of 0.03, 0.02, and 0.11, respectively. Assuming that it is possible to differentiate samples irradiated at different levels of exposure using expression data, the feature selection misclassification metric estimates the theoretical limit of how well differentially irradiated samples can be separated based on expression. The surprisingly high feature selection misclassification values obtained from GSE6874 may therefore be indicative of greater levels of noise in the data. Lastly, the greedy feature selection algorithms used to derive signatures cannot guarantee optimal results, that is, we cannot confirm that we have found the best possible signatures from each dataset for predicting radiation exposure. This potentially explains the discordance in gene composition between murine datasets (Figure 2).\n\nNevertheless, the validation performance of radiation signatures is significantly improved (Table 5). The signatures that were externally k-fold validated achieved nearly 100% accuracy. Some of our human signature models are also externally validated in the traditional sense (i.e. using a single model). This validation method, which is representative of an actual scenario, achieves >90% accuracy, and is directly relevant to creating a routine, efficient and highly accurate expression-based radiation prognostic assay.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nZENODO: Matlab code for “Predicting Exposure to Ionizing Radiation by Biochemically-Inspired Genomic Machine Learning”, http://doi.org/10.5281/zenodo.117057244\n\nCode is available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Competing interests\n\n\n\nPKR cofounded CytoGnomix Inc. A patent application on biochemically inspired gene signatures derived by machine learning is pending (US Pat. App. Ser. No. 62/202,796).\n\n\nGrant information\n\nNatural Sciences and Engineering Research Council of Canada (NSERC Discovery Grant RGPIN-2015-06290); the Canadian Foundation for Innovation; Canada Research Chairs, and CytoGnomix Inc.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File X: This spreadsheet lists all the genes found from our literature search (see Methods) that were considered during feature selection. For each gene, we report the reason for inclusion and a link to the paper containing the supporting evidence.\n\nClick here to access the data.\n\nSupplementary Files Y1–Y7: These files contain information concerning all the signatures derived for this paper. Each file contains the validation results of signatures derived from a particular dataset. Files Y1-Y3 contain the k-fold validation results of human signatures derived from GSE1725, GSE6874, and GSE10640, respectively, while Y4-Y5 contain the traditional validation results of human signatures derived from GSE6874 and GSE10640, respectively. Files Y6-Y7 contain the k-fold validation results of mouse signatures derived from GSE10640[GPL4783] and GSE10640[GPL6524], respectively. Each supplementary file contains the following columns: Signature, FS Algorithm, C, sigma, FS Misclassification, FS Log Loss, K, Misclassification, Misclassification Error, Log Loss, Log Loss Error, Goodness of Fit, and Goodness of Fit Error. These headings are described in the tab titled “Legend” in Files Y1-Y7. In addition, Files Y1-Y3 have three extra columns: Average Misclassification, Average Log Loss, and Average Goodness of Fit, which represent the misclassification, log loss, and goodness of fit, respectively, averaged over all validation sets.\n\nClick here to access the data.\n\n\nReferences\n\nPandey BN, Kumar A, Tiwari P, et al.: Radiobiological basis in management of accidental radiation exposure. Int J Radiat Biol. 2010; 86(8): 613–35. PubMed Abstract | Publisher Full Text\n\nSproull MT, Camphausen KA, Koblentz GD: Biodosimetry: A Future Tool for Medical Management of Radiological Emergencies. Health Secur. 2017; 15(6): 599–610. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu J, Li Y, Wilkins R, et al.: Accurate cytogenetic biodosimetry through automated dicentric chromosome curation and metaphase cell selection [version 1; referees: 2 approved]. F1000Res. 2017; 6: 1396. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRogan PK, Li Y, Wilkins RC, et al.: Radiation Dose Estimation by Automated Cytogenetic Biodosimetry. Radiat Prot Dosimetry. 2016; 172(1–3): 207–17. PubMed Abstract | Publisher Full Text\n\nRogan PK, Li Y, Wickramasinghe A, et al.: Automating dicentric chromosome detection from cytogenetic biodosimetry data. Radiat Prot Dosimetry. 2014; 159(1–4): 95–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShirley B, Li Y, Knoll JHM, et al.: Expedited Radiation Biodosimetry by Automated Dicentric Chromosome Identification (ADCI) and Dose Estimation. J Vis Exp. 2017; (127): e56245. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu TP, Hsu YY, Lai LC, et al.: Identification of gene expression biomarkers for predicting radiation exposure. Sci Rep. 2014; 4: 6293. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing LH, Park S, Peyton M, et al.: Distinct transcriptome profiles identified in normal human bronchial epithelial cells after exposure to γ-rays and different elemental particles of high Z and energy. BMC Genomics. 2013; 14: 372. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDressman HK, Muramoto GG, Chao NJ, et al.: Gene expression signatures that predict radiation exposure in mice and humans. PLoS Med. 2007; 4(4): e106. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaul S, Amundson SA: Development of gene expression signatures for practical radiation biodosimetry. Int J Radiat Oncol Biol Phys. 2008; 71(4): 1236–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing C, Peng H: Minimum redundancy feature selection from microarray gene expression data. J Bioinform Comput Biol. 2005; 3(2): 185–205. PubMed Abstract | Publisher Full Text\n\nPeng H, Long F, Ding C: Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans Pattern Anal Mach Intell. 2005; 27(8): 1226–38. PubMed Abstract | Publisher Full Text\n\nMucaki EJ, Baranova K, Pham HQ, et al.: Predicting Outcomes of Hormone and Chemotherapy in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) Study by Biochemically-inspired Machine Learning [version 3; referees: 2 approved]. F1000Res. 2016; 5: 2124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDorman SN, Baranova K, Knoll JHM, et al.: Genomic signatures for paclitaxel and gemcitabine resistance in breast cancer derived by machine learning. Mol Oncol. 2016; 10(1): 85–100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMucaki EJ, Zhao JZL, Lizotte D, et al.: Predicting Response to Platin Chemotherapy Agents with Biochemically-inspired Machine Learning. bioRxiv. 2017; 231712. Publisher Full Text\n\nGuyon I, Elisseeff A: An Introduction to Variable and Feature Selection. J Mach Learn Res. 2003; 3(Mar): 1157–82. Reference Source\n\nBolstad BM, Irizarry RA, Astrand M, et al.: A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinforma Oxf Engl. 2003; 19(2): 185–93. PubMed Abstract | Publisher Full Text\n\nEdgar R, Domrachev M, Lash AE: Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002; 30(1): 207–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeadows SK, Dressman HK, Muramoto GG, et al.: Gene expression signatures of radiation response are specific, durable and accurate in mice and humans. PLoS One. 2008; 3(4): e1912. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRieger KE, Hong WJ, Tusher VG, et al.: Toxicity from radiation therapy associated with abnormal transcriptional responses to DNA damage. Proc Natl Acad Sci U S A. 2004; 101(17): 6635–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJen KY, Cheung VG: Transcriptional response of lymphoblastoid cells to ionizing radiation. Genome Res. 2003; 13(9): 2092–100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrynberg P, Passos-Silva DG, Mourão Mde M, et al.: Trypanosoma cruzi gene expression in response to gamma radiation. PLoS One. 2012; 7(1): e29596. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWood RD, Mitchell M, Sgouros J, et al.: Human DNA repair genes. Science. 2001; 291(5507): 1284–9. PubMed Abstract | Publisher Full Text\n\nBirrell GW, Giaever G, Chu AM, et al.: A genome-wide screen in Saccharomyces cerevisiae for genes affecting UV radiation sensitivity. Proc Natl Acad Sci U S A. 2001; 98(22): 12608–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKarlin S, Mrazek J: Predicted highly expressed and putative alien genes of Deinococcus radiodurans and implications for resistance to ionizing radiation damage. Proc Natl Acad Sci U S A. 2001; 98(9): 5240–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChistiakov DA, Voronova NV, Chistiakov PA: Genetic variations in DNA repair genes, radiosensitivity to cancer and susceptibility to acute tissue reactions in radiotherapy-treated cancer patients. Acta Oncol. 2008; 47(5): 809–24. PubMed Abstract | Publisher Full Text\n\nKabacik S, Mackay A, Tamber N, et al.: Gene expression following ionising radiation: identification of biomarkers for dose estimation and prediction of individual response. Int J Radiat Biol. 2011; 87(2): 115–29. PubMed Abstract | Publisher Full Text\n\nZhou LJ, Zhu ZH, Liu ZX, et al.: Identification and transcriptional profiling of differentially expressed genes associated with response to UVA radiation in Drosophila melanogaster (Diptera: Drosophilidae). Environ Entomol. 2013; 42(5): 1110–7. PubMed Abstract | Publisher Full Text\n\nWang LJ, Zhou LJ, Zhu ZH, et al.: Differential temporal expression profiles of heat shock protein genes in Drosophila melanogaster (Diptera: Drosophilidae) under ultraviolet A radiation stress. Environ Entomol. 2014; 43(5): 1427–34. PubMed Abstract | Publisher Full Text\n\nChauhan V, Howland M, Wilkins R: Identification of gene-based responses in human blood cells exposed to alpha particle radiation. BMC Med Genomics. 2014; 7: 43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDom G, Tarabichi M, Unger K, et al.: A gene expression signature distinguishes normal tissues of sporadic and radiation-induced papillary thyroid carcinomas. Br J Cancer. 2012; 107(6): 994–1000. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilanowska K, Krwawicz J, Papaj G, et al.: REPAIRtoire--a database of DNA repair pathways. Nucleic Acids Res. 2011; 39(Database issue): D788–792. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTarrade S, Bhardwaj T, Flegal M, et al.: Histone H2AX Is Involved in FoxO3a-Mediated Transcriptional Responses to Ionizing Radiation to Maintain Genome Stability. Int J Mol Sci. 2015; 16(12): 29996–30014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMothersill C, O’Malley K, Harney J, et al.: Further investigation of the response of human uroepithelium to low doses of cobalt-60 gamma radiation. Radiat Res. 1997; 147(2): 156–65. PubMed Abstract | Publisher Full Text\n\nLin JY, Mühlmann-Diaz MC, Stackhouse MA, et al.: An ionizing radiation-sensitive CHO mutant cell line: irs-20. IV. Genetic complementation, V(D)J recombination and the scid phenotype. Radiat Res. 1997; 147(2): 166–71. PubMed Abstract | Publisher Full Text\n\nMATLAB: Statistics and Machine Learning Toolbox [Internet]. [cited 2018 Jan 12]. Reference Source\n\nEitrich T, Lang B: Efficient optimization of support vector machine learning parameters for unbalanced datasets. J Comput Appl Math. 2006; 196(2): 425–36. Publisher Full Text\n\nPawlowski J, Kraft AS: Bax-induced apoptotic cell death. Proc Natl Acad Sci U S A. 2000; 97(2): 529–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin G, Hamaguchi Y, Matsushita T, et al.: B-cell linker protein expression contributes to controlling allergic and autoimmune diseases by mediating IL-10 production in regulatory B cells. J Allergy Clin Immunol. 2013; 131(6): 1674–82. PubMed Abstract | Publisher Full Text\n\nChauhan V, Kuo B, McNamee JP, et al.: Transcriptional benchmark dose modeling: Exploring how advances in chemical risk assessment may be applied to the radiation field. Environ Mol Mutagen. 2016; 57(8): 589–604. PubMed Abstract | Publisher Full Text\n\nPapathanasiou MA, Kerr NC, Robbins JH, et al.: Induction by ionizing radiation of the gadd45 gene in cultured human cells: lack of mediation by protein kinase C. Mol Cell Biol. 1991; 11(2): 1009–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLucas J, Dressman HK, Suchindran S, et al.: A translatable predictor of human radiation exposure. PLoS One. 2014; 9(9): e107897. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMyeloid Cytokines for Acute Exposure to Myelosuppressive Doses of Radiation (Hematopoietic Subsyndrome of ARS), Cytokine - Radiation Emergency Medical Management [Internet]. [cited 2018 Jan 12]. Reference Source\n\nZhao JZL, Mucaki EJ, Rogan PK: Matlab Code for “Predicting Exposure to Ionizing Radiation by Biochemically-Inspired Genomic Machine Learning” [Internet]. Zenodo. 2018; [cited 2018 Feb 9]. Data Source"
}
|
[
{
"id": "33291",
"date": "14 May 2018",
"name": "Daniel Oh",
"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\nWell written article. Good analysis of available gene expression data to create gene signatures for ionizing radiation exposure. Good discussion of the identified genes' functions and roles in radiation response. Conclusions are well supported by the results. Hopefully the analysis and genes identified in this study will be incorporated and/or validated in future studies examining prediction of radiation exposure.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3728",
"date": "11 Jun 2018",
"name": "Peter Rogan",
"role": "Author Response",
"response": "Thank you for your kind comments. We agree that the approach and software should be useful for future studies of human radiation exposures. We are particularly motivated to apply multiclass SVMs, which were highly accurate in the murine dataset, to the analysis of a large set of radiation oncology patients exposed to different radiation doses."
}
]
},
{
"id": "31994",
"date": "30 May 2018",
"name": "Michael D. Story",
"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 absolutely correct that in the event of a radiological accident blood is likely to be the most likely source for analytical materials that would reflect a radiation exposure. Given the analytical approach taken we tested this using two sets of data from normal human lung epithelial cells (HBECs) irradiated at multiple doses by γ-rays but also with Fe particles such as those found in the deep space environment. (Reference 8 in this manuscript)\nWe were interested in two things: 1) Would the results from lymphoid cells translate to epithelial cells? There is sufficient evidence in the literature to suggest signatures created with lymphoid cells are poor at predicting radio response in cells from tissue. 2) Would the results from γ-ray exposures translate to something more exotic like Fe particles which have discriminating gene sets that are both common to γ-ray and Fe particles as well as unique to the radiation type.\nWe chose the 10 gene signature derived from GSE6874 and validated against GSE10640. Given the limited size of our sample set, we were surprised to see Prediction Accuracies of 73% for γ-rays and 83% for Fe particle irradiation when examining the normal HBEC cells; and 85 and 76%, respectively, for a genetically manipulated HBEC (p53 knockdown, KRAS mutant over-expressing) cell line.\nWe look forward to testing this approach in genetically diverse tumor cells.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3727",
"date": "11 Jun 2018",
"name": "Peter Rogan",
"role": "Author Response",
"response": "Thank you for your efforts to review our article and evaluate our software. We were excited about the significant results you obtained using the human signature on expression data from irradiated lung epithelial cells, and using the models to detect evidence of Fe particle radiation."
}
]
},
{
"id": "33717",
"date": "30 May 2018",
"name": "Roel Quintens",
"expertise": [
"Reviewer Expertise Radiobiology",
"Biodosimetry",
"Molecular Biology",
"Developmental Neuroscience"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary In this study, the authors have identified gene signatures for radiation dose prediction using machine learning methodologies based on publically available microarray results from human and murine samples (mostly lymphocytes) exposed to ionizing radiation. Their signatures have been independently validated showing a high specificity for dose estimation. The authors have used a novel method, based on the concept of minimum redundancy maximum relevance. This generated signatures which often contained genes that had not previously been identified as potential radiation biomarkers. In all, this is a well-conducted study with relevance to the field. However, we do have some comments/questions/remarks, as outlined below.\n\nIntroduction\nSeveral other studies have applied machine learning methodologies to identify predictive radiation exposure biomarkers. Some of these have been reviewed in Hall et al., Mut Res 2017, Supplementary Table 3.5.1.1.\n\nAnother important aspect of gene expression is alternative splicing, which also occurs in response to ionizing radiation (e.g. Sprung et al., PLoS ONE 2011; Forrester et al., PLoS ONE 2012; Macaeva et al., Sci Rep 2016). The latter study also showed for the first time the suitability of exon signatures as sensitive radiation biomarkers, and highlights the importance of prior knowledge at the exon level for subsequent primer- or probe-based assays (e.g. qRT-PCR). This may be discussed.\n\nMethods\nIn the data Pre-processing “Rows and columns of microarray data that are less than 95% complete were removed and any remaining missing values were imputed using the nearest-neighbor algorithm” How many rows and columns were removed, and on which basis was the 95% threshold selected. Also, what is the effect of the nearest-neighbor algorithm on the data \"over-fitting\". Is it possible to perform PCA on the data after removing any row/columns with less than 100% completeness and compare to the currently presented approach (95% removal and filling the remainder of the missing data)? This would allow the visualization of the effect of the proposed methodology on the segregation between the various records.\n\nOnly genes common to all datasets were retained. Does that mean common between mouse and human datasets? How were aliases identified?\n\nThe second step in the process is the selection of genes based on a non-exhaustive list of publications. Why was this necessary if the mRMR method for feature selection was applied?\n\nI particularly like the idea of performing quantile normalization after feature selection. Is this something that has been published before? Can the authors speculate (or maybe even compare) about the performance of their method on pre-normalized datasets?\n\nConcerning the method used for the “Validation of models”, I would think this approach would be more vulnerable towards the test/training dataset. What would occur to the accuracy when doing the normalization over-all of the data? Would the accuracy change drastically? Is it possible to extend the testing to cover additional data?\n\nMany datasets exist on human PBMCs/whole blood irradiated with a range of doses. Why were these datasets not considered for this study while lymphoblastoid cell lines were?\n\nIt would be helpful to have a comparison of model performance with that of “traditional” machine learning methods, as used in some of the indicated references.\n\nResults\n\n“We discovered radiation gene signatures using the microarray data of human and mouse peripheral blood samples and human lymphoblastoid cell lines, which were validated either according to signature.” Were the human lymphoblastoid cell line and prepheral blood samples grouped together in one model? If so, would it be possible to visualize how the expression data of the shortlisted genes for each data type separately (using PCA for example)?\n\nCan the authors comment on their observation that signatures derived from both murine datasets are not very similar? Apart from “noise, or intrinsic differences in the datasets”, could it possibly also be a consequence of the method used, i.e. mRMR in which low mutual information genes are selected? Based on Fig. 2 and 6 it seems that genes with higher mutual information in general have higher frequencies. Which seems logical.\n\nThe authors state that Ms4a1 appears less frequently than Glipr2. However, from the sizes of the circles, Ms4a1 seems to appear more frequently than Glipr2. Please verify this statement.\n\nAre genes in Tables 1 and 2 ranked according to their frequency, importance,…?\n\nHow do the authors explain the low frequencies of human signature genes (Fig. 6), compared with murine (Fig. 2)?\n\nLikewise, can the authors explain the large number of genes with low mutual information in the human signature (23 out of 26 <0.4), compared with the murine signature (4 out of 33 <0.4).\n\nAlthough I like the idea of mRMR, it is somewhat counterintuitive to have genes with little mutual information to be important for dose prediction. This seems to be confirmed by the fact that the compositions of human signatures are dominated by genes with high mutual information (in fact, these are all well known p53-dependent genes which appear in a high number of published radiation signatures).\n\nDiscussion\nI understand the advantage of small signatures in terms of practicality. However, in case of a real emergency, in which individuals have been irradiated without good knowledge about the exact time since exposure, larger gene signatures may provide the additional benefit of having different dynamics per gene. This may help to also predict not only the dose, but also the time since exposure. Furthermore, one-gene signatures may suffer from higher variability among the population compared to larger gene signatures.\n\nI believe results from other, similar studies may be briefly situated in the introduction/discussion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "3726",
"date": "11 Jun 2018",
"name": "Peter Rogan",
"role": "Author Response",
"response": "Regarding other studies that identify predictive radiation exposure biomarkers, we have now added text citing these and other studies in paragraph 1 of the introduction of version 2 of our paper. We investigated whether alternative splicing in response to ionizing radiation might affect the expression values that were used in training or validation of machine learning models we derived. These values could theoretically be distorted if the hybridization probes used to quantify gene expression in these microarray data were predominantly located in cassette exons. We carefully analyzed the probes on one of the most prominent genes in our signatures: DDB2, which consists of 10 exons in total. Out of the three human datasets used for deriving signatures, GSE6874[GPL4782] lacked information about the location of its DDB2 probe. However, the DDB2 probes in GSE10640[GPL6522] and GSE1725 were located in exons 8-10. According to UCSC Genome browser, the only transcript variants (NM_001300734.1, mRNA AB107039, and mRNA BC050455) involved skipping or fusing exons 3-7. Thus, in this case, alternative splicing does not affect our results, since the DDB2 probes avoid alternatively spliced exons. In general, probes seem to be designed to avoid alternatively spliced regions. Although we have not verified this for all the genes in our signatures, we speculate that taking an average over multiple probes reduces any potential affect of alternative splicing. The nearest neighbor analysis was performed to avoid inclusion of genes or individuals with sparse data in the analysis. We conservatively selected a threshold of 95% completeness to ensure that the original source data were reliable. At this threshold, none of the rows or columns or each of the datasets we used for deriving signatures (GSE1725, GSE6874[GPL4782], GSE10640[GPL6522], GSE10640[GPL4783], GSE10640[GPL6524]) were removed (based on genes common among all datasets). The effect of nearest neighbours on overfitting was minimal. Upon restriction to the set of genes available for feature selection, there were no missing values in GSE1725 and GSE6874[GPL4782], a single missing value in GSE10640[GPL6524] and GSE10640[GPL6522], and 62 missing values in GSE10640[GPL4783]. However, in GSE10640[GPL4783], of the genes available for feature selection, only four genes contained at least one missing expression value: Rad51, Ptpre, Gadd45a, and Pola1. None of these genes are among the 33 most frequently appearing genes of our murine signatures (see Figure 2). (Recall that GSE10640[GPL4783] is a murine dataset). Additionally, overfitting is mitigated by validation using independent datasets. The nearest neighbor source data occurs before model development, so the model is not fit at that point. Regarding the questions about genes common to all datasets being retained: 1) Mouse and human signatures were derived separately. So only genes common to all human datasets were available for selection in deriving human signatures and only genes common to all mice datasets were available for selection in deriving mouse signatures. 2) Our scripts did not take gene name aliases into account. It is therefore possible that genes that were left out of the analysis because they were indicated by different names in different datasets. The initial list of publications we consulted may not have been complete, but it was the result of an extensive search. We were concerned that mRMR method applied to all genes without independent experimental support would result in Type II errors. We wanted to identify the key genes from the large volume of peer-reviewed work implicating various genes in radiation response. This hypothesis based study was not designed to discover novel genes, whose putative role in radiation response was unproven or unknown. We have automated the initial feature selection procedures using expression of all (including non-coding) genes, but this is beyond the scope of our efforts. The discovery strategy would require statistical correction for the likelihood of incorrectly rejecting a null hypothesis due to multiple comparison testing. Regarding quantile normalization after feature selection, this is the first time that we have used this approach. We attempted this because our initial efforts to derive signatures from pre-normalized data were unsuccessful (poor performance). Since we do not use the expression values for the majority of genes, there was no compelling reason to normalize across the entire set of expressed genes. We are not aware of other previous efforts, but cannot exclude this possible. We did not perform an exhaustive literature search for the method that we employed, since the universe of potential applications is nearly limitless. We speculate that if quantile normalization is done before feature selection, then the reduction in dynamic range in the transformed data may lead to the derivation of poorer signatures.To give a more quantitative answer, we took the top performing signature in the Y5 Supplementary file with respect to log loss (DDB2 GTF3A TNFRSF10B) and re-validated with normalization over all data instead of just over genes in the signature; misclassification error was 2% higher and the goodness of fit was 4 cGy higher (log loss remained approximately the same). Interestingly, normalization performed over all of the data does not appear to have significant effects on the performance of our signatures. To “extend the testing to cover additional data,” it would be necessary to renormalize the initial set of genes to include these “additional genes”. If no other genes are to be included, only additional samples, then the expression values of the additional samples would need to be renormalized. Regarding other human datasets of irradiated PBMCs, in the feature selection stage, we specifically required datasets with large numbers of samples. The three human datasets we used for feature selection (GSE6874, GSE10640, GSE1725) were the largest ones we found deposited in GEO. In particular, GSE1725, the dataset with cell lines derived from patients, was the largest dataset available to us, containing 110 samples (171 samples in total, but 61 are UV irradiated). At the earliest stage of our project, we specifically chose datasets that maximized the number of common genes represented among them. This requirement enabled us to validate our signatures by either traditional or k-fold approaches. Genes available for feature selection must be present in all datasets in order for validation to work. We were surprised to find many datasets where key genes in our models were missing from expression data (see the last section on partial body irradiation data analysis in the results section of the revised version 2 of this paper). Regarding comparison of model performance with other traditional machine learning methods, we have used these methods (eg. random forest, SVMs, decision trees) in previous gene expression studies (see reference 13 of version 1 which is reference 18 of version 2). The improved performance we describe here cannot be attributed to the specific model building approaches. Regarding grouping of lymphoblastoid and PBL samples, models were always trained on one dataset at a time; they were not combined. The dissimilarity of the murine datasets may be related to their use of different microarray platforms and were collected at different times. It is difficult to tease out the intrinsic differences of the datasets from the performance of the methods. Regarding Ms4a1 and Glipr2, we verify the reviewers' observation. We corrected the density plot during preparation of the paper, but inadvertently neglected to make the corresponding change to the text. We have replaced Glipr2 with Ccng1 or Eif2ak4. Regarding ranking of genes in Tables 1 and 2, signatures derived using BSFS and CSFS list the genes according to mRMR rank. For signatures derived using FSFS, the genes would be listed according to the order in which they were selected. So in this sense, they are indeed ranked according to importance. Regarding the low frequencies of human vs murine signature genes, in the descriptions of these figures, we mention that “Frequencies are first scaled within and then between datasets to ensure values between 0 and 1.” Human signature gene frequencies appear suppressed because DDB2 and GADD45A in particular were represented more frequently than any other gene by a large margin. Regarding the numbers of genes with low mutual information in the human vs murine signatures, it may be relevant that expression data are not adjusted for either white blood cell count or body mass. We speculate that the gene expression response in the human samples reflects lower numbers of radiation exposed cells. This could dampen the signals and mutual information with radiation dose compared to the murine response. Regarding the idea that it would be counterintuitive to have genes with low mutual information as important for dose prediction, it is true that signatures are dominated by genes whose expression values share high mutual information with dose. The purpose of mRMR is to make sure that we do not overlook the genes whose expression values may encode information that is not present in the genes whose expression values have high mutual information with dose, which is why sometimes genes with lower mutual information may appear as frequently or even a bit more frequently than genes with higher mutual information. Regarding the single gene signature results, we agree that single gene signatures are more susceptible to extrinsic sources of variation unrelated to radiation exposure. However, simpler signatures may be necessary under laboratory conditions that limit the amount or complexity of testing, e.g. space radiation assays performed by astronauts."
}
]
}
] | 1
|
https://f1000research.com/articles/7-233
|
https://f1000research.com/articles/7-741/v1
|
14 Jun 18
|
{
"type": "Software Tool Article",
"title": "iSEE: Interactive SummarizedExperiment Explorer",
"authors": [
"Kevin Rue-Albrecht",
"Federico Marini",
"Charlotte Soneson",
"Aaron T.L. Lun",
"Kevin Rue-Albrecht",
"Federico Marini",
"Charlotte Soneson"
],
"abstract": "Data exploration is critical to the comprehension of large biological data sets generated by high-throughput assays such as sequencing. However, most existing tools for interactive visualisation are limited to specific assays or analyses. Here, we present the iSEE (Interactive SummarizedExperiment Explorer) software package, which provides a general visual interface for exploring data in a SummarizedExperiment object. iSEE is directly compatible with many existing R/Bioconductor packages for analysing high-throughput biological data, and provides useful features such as simultaneous examination of (meta)data and analysis results, dynamic linking between plots and code tracking for reproducibility. We demonstrate the utility and flexibility of iSEE by applying it to explore a range of real transcriptomics and proteomics data sets.",
"keywords": [
"visualization",
"interactive",
"R",
"Bioconductor",
"genomics",
"transcriptomics",
"proteomics",
"shiny"
],
"content": "Introduction\n\nInteractive data exploration is critical to the analysis and comprehension of data generated by high-throughput biological assays, such as those commonly used in genomics. Exploration drives the formation of novel data-driven hypotheses prior to a more rigorous statistical analysis, and enables diagnosis of potential problems such as batch effects and low-quality samples. To this end, visualisation of the data using an intuitive and interactive interface is crucial for enabling researchers to examine the data from different perspectives across samples (e.g., experimental replicates, patients, single cells) and features (e.g., genes, transcripts, proteins, genomic regions).\n\nMost existing tools for interactive visualisation of biological data are designed for specific assays and analyses, e.g., pRoloc for proteomics (Gatto et al., 2014), shinyMethyl for methylation (Fortin et al., 2014), HTSvis for high-throughput screens (Scheeder et al., 2017). Opportunities for customisation are generally limited, making it difficult to re-use the same visualisation software for new technologies or experimental designs where different aspects of the data are of interest. Moreover, standalone tools such as the Loupe Cell Browser from 10x Genomics (Zheng et al., 2017) do not easily integrate into established analysis pipelines such as those based on the R statistical programming language (R Development Core Team, 2008). This complicates any coordinated use of these tools with a reproducible, transparent, and statistically rigorous analysis.\n\nHere, we present the iSEE software package for interactive data exploration. iSEE is implemented in R using the Shiny framework (Chang et al., 2017) and exploits data structures from the open-source Bioconductor project (Gentleman et al., 2004), specifically the SummarizedExperiment class. iSEE allows users to simultaneously visualise multiple aspects of a given data set, including experimental data, metadata and analysis results. Dynamic linking and point selection facilitate the flexible exploration of interactions between different data aspects. Additional functionalities include code tracking, intelligent downsampling of large data sets, custom colour scale specification and tour construction. We demonstrate the capabilities of iSEE by applying it to a diverse range of real data sets.\n\n\nOperation\n\nThe iSEE software package requires R version 3.5.0 or higher, along with packages from Bioconductor version 3.7 or higher. The interface is initialised with a single call to the iSEE() function, accepting a SummarizedExperiment object (Huber et al., 2015) as input. Any analysis workflow that generates a SummarizedExperiment object is supported.\n\n\nMotivation for using the SummarizedExperiment class\n\nEach instance of the SummarizedExperiment class stores one or more matrices of experimental observations as “assays”, where rows and columns represent genomic features and biological samples, respectively. For instance, individual assays may represent gene expression matrices, either in the form of raw counts or normalised values. In addition, per-feature or per-sample variables are stored in the “rowData” and “colData” slots, respectively; these may include experimental metadata as well as analysis results.\n\nThe flexibility of the SummarizedExperiment class is the driving factor behind its broad deployment throughout the Bioconductor ecosystem. SummarizedExperiment objects are currently used in analysis pipelines for RNA sequencing (Love et al., 2014), methylation (Aryee et al., 2014) and Hi-C data (Lun et al., 2016), amongst others. Package developers can also easily use the base SummarizedExperiment class to derive new bespoke classes for particular applications, such as the SingleCellExperiment class for single-cell ‘omics data. By accepting SummarizedExperiment objects as input, iSEE immediately offers interactive visualisation for a variety of data modalities. This complements the state-of-the-art analysis workflows and methodologies already available in R/Bioconductor packages.\n\n\nInterface implementation\n\nAll data aspects stored in a SummarizedExperiment can be simultaneously examined in the multi-panel layout of the iSEE interface (Figure 1A). The interface layout is built using the shinydashboard package (Chang & Borges Ribeiro, 2018), with colour-coded panels to visualise each data aspect. Individual panel types include:\n\nColumn data plots, for visualising sample metadata stored in the colData slot of the SummarizedExperiment object.\n\nFeature assay plots, for visualising experimental observations for a particular feature (e.g. gene) across samples from any assay in the SummarizedExperiment object.\n\nRow statistics tables, to present the contents of the rowData slot of the SummarizedExperiment object.\n\nRow data plots, for visualising feature metadata stored in the rowData slot of the SummarizedExperiment object.\n\nHeatmaps, to visualise assay data for multiple features where samples are ordered by one or more colData fields.\n\nReduced dimension plots, which display any two dimensions from pre-computed dimensionality reduction results (e.g., from PCA or t-SNE). These results are taken from the reducedDim slot if this is available in the object supplied to iSEE.\n\niSEE uses a customisable multi-panel layout (A) that simultaneously displays one or more panels of various types, where each panel type visualises a different aspect of the data. New panels of any type can be added (i), and all panels can be removed, reordered or resized (ii). Panel types are available to visualise sample-based reduced dimensionality embeddings (iii), sample-level metadata (iv), and experimental observations across samples for each feature (v). Other panel types include row statistics tables (vi), to facilitate searching across features and their metadata; heatmaps (vii), to visualise experimental observations for multiple features; and feature-level metadata plots. Panels of each type are colour-coded for ease of interpretation. (B) Information can be transmitted between panels according to a user-specified scheme. Here, the selection of feature X in the row statistics table determines the y-axis of the feature assay plot, and colours the samples in the reduced dimension plot by the expression of X. Selection of points in the reduced dimension plot (dotted blue line) also determines the samples that are shown in the column data (i.e., sample metadata) plot; further selection of points in the column data plot determines the samples that are shown in the heatmap.\n\nEach sample is represented as a point in column data, feature assay and reduced dimension plots. Similarly, each feature is represented by a point in row data plots. For these panel types, a scatter plot is automatically produced if the selected variables on the x- and y-axes are both continuous. If exactly one variable is categorical, points are grouped by the categorical levels and a (vertical or horizontal) violin plot is produced with points scattered within each violin. If both variables are categorical, a “rectangle plot” is produced where each combination of categorical levels is represented by a rectangle with area proportional to the frequency of that combination. Points are scattered randomly within each rectangle. For ease of interpretation, the rectangle plot collapses to a mirrored bar plot when one of the categorical variables only has one level.\n\nSample-based points can be coloured according to the values of any sample-level metadata field in the colData slot or by the assay values of a selected feature. Similarly, feature-based points can be coloured according to any feature-level metadata field in the rowData slot. Heatmaps are coloured according to the expression values of the selected features in the chosen assay, with additional colour annotation for each of the colData fields used to order the samples. In all cases, the variable to use for colouring can be dynamically selected for each plot. This enables users to easily examine relationships between different variables in a single plot.\n\nBy default, colour maps for categorical and continuous variables are taken from the ggplot2 (Wickham, 2009) and viridis packages (Garnier, 2018), respectively. However, iSEE also implements the ExperimentColorMap class, which allows users to specify arbitrary colour maps for particular variables. Each colour map is a function that returns a vector of distinct colours of a specified length, and will be called whenever the associated variable is used for point colouring in a particular panel. The returned colours will be mapped to factor levels for categorical variables, or used in colour interpolation for continuous variables. For categorical variables, the function may also return a constant vector of named colours corresponding to the levels of a known factor. Colour maps can be specified for individual variables; for all assays, all column data variables, or all row data variables (with different functions for continuous or categorical variables); or for all categorical or continuous variables. This provides a convenient yet flexible mechanism for customisation of colouring schemes within the interface.\n\nA key feature of iSEE is the ability to dynamically transmit information between panels (Figure 1B). Users can define and reorganise arbitrary links between “transmitting” and “receiving” panels, whereby selections in transmitting panels control the inclusion and appearance of the corresponding data points in receiving panels. This feature facilitates exploration of the relationships between different aspects of the data. For example, users can easily determine co-expression patterns of genes in a particular region of a reduced dimensionality embedding – this is achieved by selecting points in a reduced dimension plot (using the standard rectangular brush or a lasso selection) and transmitting that selection to any number of feature assay plots.\n\nThis linking paradigm extends to multiple panels, whereby a panel can transmit to multiple receivers, and a receiving panel can transmit its own selection to another plot. Chains of linked plots allow users to mimic the arbitrarily complex gating strategies often found in analyses of flow cytometry data (Finak et al., 2014). With iSEE, this concept is extended to any assay data, feature-level or sample-level metadata present in a SummarizedExperiment object, providing a powerful framework for interrogating multiple interactions between data aspects. Row statistics tables can also transmit to various plot types, by selecting a table row to control the colouring of sample-based points; or by defining a subset of features to visualise in a heatmap. Furthermore, row data plots can transmit to row statistics tables, whereby selection of points in the former will subset the latter.\n\niSEE automatically memorises the exact R code that was used to generate every plot, extending previous work by Marini & Binder (2016). This code is fully accessible to users at any time during the run-time of the interface. By integrating the code reported by iSEE into their own scripts, users can easily reproduce the results of any exploratory analysis. Similarly, the code required to reproduce the current state of the interface can also be reported. This can be used in startup scripts to launch an iSEE instance in any preferred layout, including the panel organisation, variable selection, colouring schemes, links between panels and even individual brushes and lasso selections.\n\nRow statistics tables can be augmented with dynamic annotation based on the selected row, linking to online resources such as Ensembl (Zerbino et al., 2018) or Entrez (Coordinators, 2017). For large data sets, points can be downsampled in a density-dependent manner to accelerate rendering of the plots, improving the responsiveness of the interface without compromising the fidelity of the visualisation. Users can also include a bespoke step-by-step “tour” of their data set via the rintrojs package (Ganz, 2016), guiding the audience through an examination of the salient features in the data.\n\n\nUse cases\n\nTo demonstrate iSEE’s functionality, we used it to explore a plate-based single-cell RNA sequencing (scRNA-seq) data set involving 379 cells from the mouse visual cortex (Tasic et al., 2016). This demonstration guides the user through the main features of the iSEE interface including the multi-panel layout, colouring and dynamic linking.\n\nAn interactive tour of this use case can be viewed here.\n\nWe applied iSEE to a larger scRNA-seq data set involving 4,000 peripheral blood mononuclear cells (PBMCs), generated by 10x Genomics (Zheng et al., 2017). This demonstration explores the differences between different methods for distinguishing cells from empty droplets in droplet-based scRNA-seq protocols (Lun et al., 2018).\n\nAn interactive tour of this use case can be viewed here.\n\nWe applied iSEE to bulk RNA sequencing data from The Cancer Genome Atlas (TCGA) project, using a subset of expression profiles involving 7,706 tumor samples (Rahman et al., 2015). This demonstration examines the elevation of HER2 expression in a subset of breast cancer samples.\n\nAn interactive tour of this use case can be viewed here.\n\nFinally, we explored a mass cytometry study involving more than 170,000 PBMCs from multiple donors before and after stimulation with BCR/FcR-XL (Bodenmiller et al., 2012). We used iSEE to visualise and refine a gating analysis to obtain B cells, and to investigate differences in expression of the functional marker pS6 after stimulation.\n\nAn interactive tour of this use case can be viewed here.\n\n\nConclusion\n\niSEE provides a general interactive interface for visual exploration of high-throughput biological data sets. Any study that can be represented in a SummarizedExperiment object can be used as input, allowing iSEE to accommodate a diverse range of ‘omics data sets. The interface is flexible and can be dynamically customised by the user; supports exploration of interactions between data aspects through colouring and linking between panels; and provides transparency and reproducibility during the interactive analysis, through code tracking and state reporting. The most obvious use of iSEE is that of data exploration for hypothesis generation during the course of a research project. However, we also anticipate that public instances of iSEE will accompany publications to enable authors to showcase important aspects of their data through guided tours.\n\n\nSoftware availability\n\nThe iSEE package is available at https://doi.org/doi:10.18129/B9.bioc.iSEE (Soneson et al., 2018) under an MIT license.\n\nSource code of the development version of the package is available at https://github.com/csoneson/iSEE.\n\nCode for the demonstrations and tours is available at https://github.com/LTLA/iSEE2018.\n\nArchived source code of the version reported in this article and interactive tours is available from http://doi.org/10.5281/zenodo.1247374 (Rue-Albrecht et al., 2018)\n\n\nData availability\n\nData used in the described use cases is available from the following articles:\n\nhttp://doi.org/10.1038/nn.4216 (Tasic et al., 2016)\n\nhttp://doi.org/10.1038/ncomms14049 (Zheng et al., 2017)\n\nhttps://doi.org/10.1093/bioinformatics/btv377 (Rahman et al., 2015)\n\nhttps://doi.org/10.1038/nbt.2317 (Bodenmiller et al., 2012)",
"appendix": "Competing interests\n\n\n\nNo competing interests were declared.\n\n\nGrant information\n\nATLL was supported by core funding from Cancer Research UK [award no. 17197 to JM]. The work of FM is supported by the German Federal Ministry of Education and Research (BMBF 01EO1003).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank the organisers and participants of the European Bioconductor Meeting 2017, where the idea for this package was first conceived. We also thank members of the Bioconductor community for their helpful suggestions. Finally, we thank John Marioni and Mark Robinson for their helpful comments on the manuscript.\n\n\nReferences\n\nAryee MJ, Jaffe AE, Corrada-Bravo H, et al.: Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics. 2014; 30(10): 1363–1369. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBodenmiller B, Zunder ER, Finck R, et al.: Multiplexed mass cytometry profiling of cellular states perturbed by small-molecule regulators. Nat Biotechnol. 2012; 30(9): 858–867. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChang W, Borges Ribeiro B: shinydashboard: Create Dashboards with ’Shiny’. R package version 0.7.0. 2018. Reference Source\n\nChang W, Cheng J, Allaire JJ, et al.: shiny: Web Application Framework for R. R package version 1.0.5. 2017. Reference Source\n\nFinak G, Frelinger J, Jiang W, et al.: OpenCyto: an open source infrastructure for scalable, robust, reproducible, and automated, end-to-end flow cytometry data analysis. PLoS Comput Biol. 2014; 10(8): e1003806. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFortin JP, Fertig E, Hansen K: shinyMethyl: interactive quality control of Illumina 450k DNA methylation arrays in R [version 2; referees: 2 approved]. F1000Res. 2014; 3: 175. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGanz C: rintrojs: A wrapper for the intro.js library. J Open Source Softw. 2016. Publisher Full Text\n\nGarnier S: viridis: Default Color Maps from ’matplotlib’. R package version 0.5.1. 2018. Reference Source\n\nGatto L, Breckels LM, Wieczorek S, et al.: Mass-spectrometry-based spatial proteomics data analysis using pRoloc and pRolocdata. Bioinformatics. 2014; 30(9): 1322–1324. 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\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Perry M, Ing-Simmons E: Infrastructure for genomic interactions: Bioconductor classes for Hi-C, ChIA-PET and related experiments [version 2; referees: 2 approved]. F1000Res. 2016; 5: 950. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLun AT, Riesenfeld S, Andrews T, et al.: Distinguishing cells from empty droplets in droplet-based single-cell rna sequencing data. bioRxiv. 2018. Publisher Full Text\n\nMarini F, Binder H: Development of applications for interactive and reproducible research: a case study. Genomics Comput Biol. 2016; 3(1): e39. Publisher Full Text\n\nNCBI Resource Coordinators: Database Resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2017; 45(D1): D12–D17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Development Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2008. Reference Source\n\nRahman M, Jackson LK, Johnson WE, et al.: Alternative preprocessing of RNA-Sequencing data in The Cancer Genome Atlas leads to improved analysis results. Bioinformatics. 2015; 31(22): 3666–3672. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRue-Albrecht K, Marini F, Soneson C, et al.: Interactive SummarizedExperiment Explorer. Zenodo. 2018. Data Source\n\nScheeder C, Heigwer F, Boutros M: HTSvis: a web app for exploratory data analysis and visualization of arrayed high-throughput screens. Bioinformatics. 2017; 33(18): 2960–2962. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoneson C, Lun A, Marini F, et al.: iSEE: Interactive SummarizedExperiment Explorer. R package version 1.0.1. 2018. Data Source\n\nTasic B, Menon V, Nguyen TN, et al.: Adult mouse cortical cell taxonomy revealed by single cell transcriptomics. Nat Neurosci. 2016; 19(2): 335–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2009. Publisher Full Text\n\nZerbino DR, Achuthan P, Akanni W, et al.: Ensembl 2018. Nucleic Acids Res. 2018; 46(D1): D754–D761. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng GX, Terry JM, Belgrader P, et al.: Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017; 8: 14049. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "35043",
"date": "19 Jun 2018",
"name": "James W. MacDonald",
"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 iSEE package was developed to allow people to easily perform exploratory data analysis with data that are stored in a Bioconductor SummarizedExperiment object. A SummarizedExperiment container allows researchers to store one or more matrices of data, where the columns represent samples, and the rows represent either genomic positions or genomic features (genes, exons, transcription start sites, etc). In addition to the matrices of data, the SummarizedExperiment also contains two additional objects that describe the samples (the colData) and the rows (the rowData or rowRanges).\niSEE allows users to interactively plot the underlying data from a SummarizedExperiment, and also choose subsets of the data based on either interactive selection of data in a plot, or by selecting samples or genomic regions based on the colData or rowData. The chosen subsets can then be linked to other plots in the Shiny Dashboard. This simplifies what could be a complex process, allowing both experienced R users a quick way to check over their data, and allowing less experienced R users the ability to do things that they otherwise might not have been able to do.\nAll the underlying code generated while making interactive changes is saved and can be printed out later, in order to make the exploratory data analysis reproducible. This is an excellent feature, particularly for those who want to share observations with colleagues that may not be local.\n\nThe only negative for this package is that, being based on the Shiny framework, to allow a colleague to explore the data requires that the colleague either have R, iSEE, and all its dependencies installed, or that you have a server running all necessary packages that you can point the colleague to. This limits sharing with people who are not R savvy, but is a function of how Shiny works, rather than the iSEE package.\nThis is a high quality package, and given the generalizability of the SummarizedExperiment package, is applicable to a whole range of different data types. Given the ease of use, self documenting features, and applicability to multiple data types, this package will likely become very popular for exploratory data analysis.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "35042",
"date": "20 Jun 2018",
"name": "Lorena Pantano",
"expertise": [
"Reviewer Expertise small rna",
"mirna",
"visualization",
"report",
"integration"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors show an interactive visualization tool for a very common data type used for many of the packages in Bioconductors (SummarizedExperiment). It has enough flexibility to explore all kind of information the object can contain, an interactive tool based on Rshiny, is customizable so it can be adapted to each user.\nI only have minor some comments:\nTutorial 2: step 10 gets the text box in the upper left of the windows, but I think it should be at other position since it says to change the y-axis of the plot. I think this happens when the user doesn't follow the instruction to click on to some button that should expand the menu with more options.\n\nIt would be nice the tour re-start from the position it was left, with an option to start over. It happened many times that I click accidentally outside the box and I had to start over.\n\nIn the cases the object doesn't have reducedDim for more than the 2 dimensions shown in the plot. I tried to use 3, and it gave an error. Maybe a more informative error would help the user to understand that there is no that information.\n\nI am not totally sure how to use the rintrojs package to generate a tool. It would be nice a reference to some documentation on how to do it or clarification if I am not understanding this correctly.\n\nFor the features mentioned like code tracking and additional functionality, it would be nice to have a link to the vignette in the paper so the user can jump into how to get it done.\n\nI think it would be nice to make available a docker image with all the requirements to run iSEE installed. It would promote the use of the tool a lot among bioinformaticians working with non-computational researchers.\n\nIt is nice to change the color for all the variables. I would add an example on how to change the palette for all categorical since the code would be slightly different than the one for continuous variables. It would make the user quickly using that option and avoid silly errors.\n\nI don't know if this is possible as it is right now, but it could be an option to load a RDA/RDS file containing the SE object instead of creating an app only for that data? That would open the door to deploy the tool independent of the data. For instance, I can see a scenario where iSEE is installed in a docker container, where the user just starts the image and when opening the browser at localhost:8787, there is an option to load a file with the object.\n\nCongrats on the tools!\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "35044",
"date": "27 Jun 2018",
"name": "Alejandro Reyes",
"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 implement an interactive tool, called iSEE, to perform exploratory analyses for high-throughput experiments. The tool inputs a Bioconductor core structure, the SummarizedExperiment object (coerced into a SingleCellExperiment object) and builds an interactive interphase for data exploration. iSEE provides several tools for data exploration by plotting features of an assay along with sample metadata, feature metadata, and reduced representations of the assays. Furthermore, iSEE enables users to interact with the plots and to dynamically link panels with different representations of the data. The analyses performed using iSEE are reproducible, since the code that was run through the graphic interphase can be downloaded.\n\nOverall, the manuscript presents a very good idea and the code implementation is of great quality. iSEE will be very useful for people without programming background to perform basic analyses. I believe that the success of this tool will depend on whether the authors continue to develop it based on feature requests from users.\n\nI don’t have major concerns. However, I do have some recommendations to increase the interest of potential users.\nEnable users to select more than one group of samples from the dimensionality reduction plots. Furthermore, it would be very useful to enable users to fill new columns of colData based on the interactive grouping of samples.\n\nEnable users to retrieve an R data object if the initial input was modified during the analysis.\n\nIn the context of single-cell or large-scale analyses, it would be helpful to implement tools for differential abundance analyses and gene set enrichment analyses. For instance, one could think of an implementation where users manually define groups of cells from tSNE/PCA plots, retrieve the genes that are differentially expressed between these groups, and extract the pathways that are enriched among the differentially expressed genes.\n\nWhen grouping samples manually on the tSNE/PCA plots, the violin plots of individual features (for example, genes) could be stratified based on these selections (e.g. plot one violin per group of selected points in the “Feature assay plot” panel). In the current implementation, it is only possible to colors the points within the violin plot, which makes difficult to compare distributions between groups of samples.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-741
|
https://f1000research.com/articles/7-737/v1
|
13 Jun 18
|
{
"type": "Case Report",
"title": "Case Report: Drug induced hyperkalemia presenting as acute flaccid quadriparesis",
"authors": [
"Santhosh Narayanan",
"Divya Prakash",
"Divya Prakash"
],
"abstract": "Hyperkalemia is a life-threatening dyselectrolytemia that commonly affects cardiac conductivity and contractility. Ascending paralysis affecting the extremities associated with hyperkalemia is not commonly seen. Here we report a case of flaccid quadriparesis in a patient who was taking potassium sparing diuretic for cardiac disease. An electrocardiogram showed typical signs of hyperkalemia. The patient was administered antihyperkalemic measures, which led to a dramatic improvement in symptoms. Hyperkalemic paralysis is a completely reversible emergency condition and should always be considered when dealing with acute onset flaccid paralysis.",
"keywords": [
"Hyperkalemia",
"neuromuscular paralysis",
"Electrocardiography"
],
"content": "Introduction\n\nHyperkalemia is a frequently encountered clinical problem. Hyperkalemic paralysis occurs primarily in genetic defects due to sodium channelopathy. Though numerous other etiologies can cause hyperkalemia, neuromuscular paralysis is unusual. Prompt diagnosis and management ensues complete and rapid reversal of symptoms1. This atypical presentation is a challenge to the clinican. In the case presented herein, early initiation of treatment saved the patient without the need of invasive ventilation or hemodialysis.\n\n\nCase report\n\nA 66 year old man presented with a history of weakness of both lower extremities followed by weakness in the upper extremities for a duration of 12 hours. There was no history of numbness, paresthesia or sensory loss. The patient did not have any symptom suggestive of cranial nerve involvement, raised intracranial tension and autonomic dysfunction. He had a past history of coronary artery disease (anterior wall myocardial infarction with moderate left ventricular dysfunction) and had undergone percutaneous transluminal coronary angioplasty. The patient was taking Aspirin (75 mg once daily), Ramipril (2.5 mg once daily) and Spironolactone (25 mg once daily) for the past year.\n\nOn examination the patient was attentive, oriented and hemodynamically stable. Nervous system examination revealed hypotonia of all four limbs with a power of grade 2. Deep tendon reflexes were absent. Remaining neurological examination and other system examination were within normal limits. Clinically a possibility of acute inflammatory demyelinating polyradiculoneuropathy was considered. Laboratory investigations are depicted in Table 1.\n\nAn electrocardiogram (ECG) showed tall peaked T waves (Figure 1), prolonged PR interval and loss of P waves conspicuous of hyperkalemia (Figure 2). The patient was immediately initiated on antihyperkalemia measures, where the dose was guided by clinical response. Calcium gluconate, insulin-dextrose solution infusion, nebulization with beta 2 agonist and oral potassium binding agent (calcium polystyrene sulfonate) were administered. ACE inhibitors (Ramipril 2.5 mg once daily) and Spironolactone (25 mg once daily) were stopped. Arterial blood gas analysis revealed mild metabolic acidosis.\n\nSerum potassium levels subsequently normalized after 12 hours of treatment. The patient showed rapid clinical improvement, without the requirement of haemodialysis. He was able to walk on the second day and on examination had grade 5 power in all limbs. An ECG taken after correction of hyperkalemia (after 12 hours of treatment) showed normal sinus rhythm. The patient was detected to have deranged renal function with evidence of chronic kidney disease using ultrasonography, which might have contributed to his hyperkalemia. A portable chest radiograph did not show evidence of fluid overload. Nerve conduction was normal.\n\nOn follow up after two months the patient was doing well without any neurological symptoms. His serum creatinine was 2.4 mg/dl and serum potassium was 4 meq/l. He is being monitored regularly by the nephrologist.\n\n\nDiscussion\n\nClinical manifestations of hyperkalemia are often nonspecific. Patients may present with vague aches and pains, muscle cramps, fatigue or sometimes palpitations. Cardiac arrhythmias are life threatening. Neuromuscular paralysis is not a common presentation of hyperkalemia and is frequently seen with hypokalemia. There are only isolated reports of hyperkalemia presenting as flaccid paralysis2,3. Weakness in hyperkalemia occurs in an ascending manner, starting in the lower limbs and sometimes leads to respiratory distress requiring invasive ventilation. In the present case, weakness progressed rapidly, but did not involve the respiratory muscles and the patient improved without the need of respiratory support.\n\nThe exact molecular mechanism by which hyperkalemia produces neurological dysfunction is not yet elucidated. It is hypothesized to be due to abnormal membrane depolarization. Resting membrane potential of cell membranes is maintained by the difference in concentration of extracellular and intracellular potassium ions. When the extracellular potassium increases in hyperkalemia, it blunts the transmission of nerve impulse to muscle fiber4.\n\nIn a report by Evers et al., renal dysfunction was the most common cause of secondary hyperkaemic paralysis, as in our case5 Other etiologies include excessive dietary intake of potassium, metabolic acidosis, medications that inhibit the renin angiotensin aldosterone system, medications that cause transcellular shift of potassium (Digoxin) and increased tissue catabolism as in tumor lysis and trauma. Succinylcholine used for rapid sequence intubation can cause fatal hyperkalemia, and sometimes result in a similar presentation6. Serial electrocardiographic changes in hyperkalemia are described by Aslam et al.,7 Electrocardiographic changes may not always correlate with degree of hyperkalemia.\n\n\nConclusion\n\nThe present case highlights the importance of considering hyperkalemia as a differential diagnosis in patients presenting with acute flaccid paralysis. Regular monitoring of serum potassium should be done in patients on potassium sparing diuretics and other potassium altering drugs. Patients with chronic kidney disease should be carefully watched for development of any complication of hyperkalemia.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "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\nWahab A, Panwar RB, Ola V, et al.: Acute onset quadriparesis with sine wave: a rare presentation. Am J Emerg Med. 2011; 29(5): 575.e1–2. PubMed Abstract | Publisher Full Text\n\nPanichpisal K, Gandhi S, Nugent K, et al.: Acute quadriplegia from hyperkalemia: A case report and literature review. Neurologist. 2010; 16(6): 390–3. PubMed Abstract | Publisher Full Text\n\nKumar KS, Ramakrishna C, Padmanabhan S, et al.: Hyperkalemic quadriparesis in a patient of ESRD. Indian J Nephrol. 2005; 15: 108–9. Reference Source\n\nDutta D, Fischler M, McClung A: Angiotensin converting enzyme inhibitor induced hyperkalaemic paralysis. Postgrad Med J. 2001; 77(904): 114–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEvers S, Engelien A, Karsch V, et al.: Secondary hyperkalaemic paralysis. J Neurol Neurosurg Psychiatry. 1998; 64(2): 249–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPang YL, Tseng FL, Tsai YC, et al.: Suxamethonium-induced hyperkalaemia in a patient with a normal potassium level before rapid-sequence intubation. Crit Care Resusc. 2006; 8(3): 213–214. PubMed Abstract\n\nAslam S, Friedman EA, Ifudu O: Electrocardiography is unreliable in detecting potentially lethal hyperkalaemia in haemodialysis patients. Nephrol Dial Transplant. 2002; 17(9): 1639–1642. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "127249",
"date": "28 Mar 2022",
"name": "Pasquale Esposito",
"expertise": [
"Reviewer Expertise Clinical Nephrology"
],
"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 read this paper that I found interesting, even if I think that some aspects of this case should be more deeply discussed.\nFor example, it could be useful for the readers to clearly explain what were the possible causes of hyperkalemia in your patient.\nMoreover, I suggest adding a paragraph about the potential differential diagnosis of hyperkalemia-related flaccid paralysis, including, for example, the case of hyperkalemic periodic paralysis (see D'Ercole et al., 20211).\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
},
{
"id": "143549",
"date": "27 Jul 2022",
"name": "Emma Matthews",
"expertise": [
"Reviewer Expertise neurology",
"neuromuscular",
"channelopathies",
"genetics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 case of a 66 year old man with acute ascending weakness affecting all four limbs, with a background of cardiac disease. Examination reveals a purely motor lower motor neuron deficit. Subsequent investigations demonstrate hyperkalaemia with ECG changes presumed secondary to potassium sparing diuretics and impaired renal function.\nThere are several \"learning points\" in this case report which the authors highlight nicely:\n1. Flaccid weakness due to deranged potassium is most commonly reported with hypokalaemia and much less commonly with hyperkalaemia. This means in a patient such as this hyperkalaemia may not immediately enter the differential and there could be delay to diagnosis. This is potentially dangerous given the ECG changes and associated risk of cardiac arrhythmia.\nI think one point the authors may expand on is that dyskalaemia (whether high or low) can be a cause of flaccid paralysis or limb weakness so in a patient presenting with this picture an ECG should be a test done sooner rather than later. The answer of course usually comes with the blood results but these can take several hours to be returned and if ECG is only done after the identification of potassium level there may have been time wasted in initiating treatment to prevent arrhythmia.\nI wonder if the paralysis occurred in this case because the potassium was so high >9, and this may also be highlighted or discussed.\n2. The patient's potassium was thought to have become elevated due to impaired renal function. This is the second learning point highlighted, that patients on potassium sparing diuretics require regular renal and electrolyte monitoring.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-737
|
https://f1000research.com/articles/7-733/v1
|
12 Jun 18
|
{
"type": "Data Note",
"title": "A chronic protocol of bilateral transcranial direct current stimulation over auditory cortex for tinnitus treatment: Dataset from a double-blinded randomized controlled trial",
"authors": [
"Ali Yadollahpour",
"Miguel Mayo",
"Nader Saki",
"Samaneh Rashidi",
"Arash Bayat",
"Miguel Mayo",
"Nader Saki",
"Samaneh Rashidi",
"Arash Bayat"
],
"abstract": "Preliminary studies have demonstrated the therapeutic potential of transcranial direct current stimulation (tDCS) for chronic tinnitus. However, the findings are controversial and most of the studies investigated effects of a single session of tDCS and short after-effects, ranging from hours to days. To our knowledge, there is no published study investigating the effects of a chronic protocol of bilateral tDCS over auditory cortex (AC) with one month follow-up in a double blinded randomized clinical trial. This dataset presents the results of a double-blinded placebo controlled trial investigating the effects of chronic protocol (10 sessions) of tDCS over AC with 1 month follow-up. The data of the two groups, real tDCS (n=25) and sham tDCS (n=15), are reported. The dataset includes three main data groups: patient- and tinnitus-specific data, data of the primary and secondary outcomes, and data on the adverse effects of and tolerability to tDCS. The first group includes demographic information, audiometric assessments, and tinnitus-specific characteristics. The second group includes tinnitus handicap inventory (THI) scores, tinnitus loudness, and tinnitus related distress based on 0-10 numerical visual analogue scale (VAS) scores. The values of the primary and secondary outcomes for pre-intervention and at different time points following interventions are presented. THI scores pre-intervention and immediately post-intervention and at 1 month follow-up; the scores of tinnitus loudness and distress scores for pre-intervention, and immediately, 1 hour, 1 week, and at 1 month after the last stimulation session are presented. Moreover, the adverse effects of and tolerability to the tDCS were assessed using a customized questionnaire after the last tDCS session. This dataset can be used alone or in combination with other datasets using advanced statistical analyses and modeling to investigate the treatment efficacy of tDCS in chronic intractable tinnitus.",
"keywords": [
"Transcranial direct current stimulation",
"repeated sessions",
"chronic protocol",
"Tinnitus",
"Auditory cortex",
"Tolerability",
"Adverse effects"
],
"content": "Introduction\n\nTinnitus is a phantom auditory perception in the absence of external sound that affects 10–15% of the world adult population, present in different forms, including buzzing, hissing, pulsatile, ringing and pulsatile tone1,2. Neuroimaging studies have shown that tinnitus is a complex disorder involving a large network consisting of multiple overlapping brain networks including primary and secondary auditory cortices, as well as specific non-auditory areas and limbic processes3,4. Several pharmaceutics agents are used for tinnitus treatment; however, a large portion of the patients are resistant to the treatment, which usually induce severe comorbidities, such as anxiety, sleep disturbances and depression2. There is currently no definitive medication-based treatment for tinnitus1–3. Different research groups have proposed and developed several non-pharmacological interventions, including cognitive behavioral therapies, hearing aids, neurofeedback, and noise-masking techniques5–7. Despite the development of different non-pharmaceutical techniques including cognitive behavioral therapies8, noise-masking modalities9, and neurofeedback10, the efficacies of these treatments for tinnitus are limited.\n\nTranscranial direct current stimulation (tDCS) has potential therapeutic efficacy for different neuropsychiatric disorders5–7,11,12 and also that capability of enhancement of cognitive functions in healthy individuals13. Single and repeated-session protocols of tDCS over the auditory cortex (AC) have shown promising outcomes14,15; however, the findings are controversial and most of the studies that have been conducted investigated the effects of single-session protocols of tDCS and short after-effects, ranging from hours to some days15,16. To achieve an effective tDCS protocol, different clinical studies with large sample sizes and robust designs should be conducted to identify the effective electrode montage and stimulation parameters. In addition, chronic protocols with long enough follow-up assessments should also be considered. For long-term tDCS protocols, assessing the possible adverse effects is necessary. In this regard, we have designed a comprehensive project to assess the efficacy of different protocols of tDCS with different electrode montages over different sites of the brain for intractable tinnitus17. Here, we report the dataset of a clinical trial designed as a double-blinded randomized placebo-controlled trial to investigate the effects of repeated sessions of tDCS on tinnitus symptoms. To the best of our knowledge, this is the first randomized clinical trial to investigate the effects of repeated sessions of tDCS on intractable chronic tinnitus symptoms and comorbid depression and anxiety with a 1-month follow-up. The main feature of these data is that the patients, the researcher who evaluated the outcomes, and the researcher who performed the statistical analyses were blinded to the study.\n\n\nMethods\n\nThis dataset presents the results of a double blinded randomized placebo controlled clinical trial investigated the effects of chronic protocol of bilateral tDCS over AC in intractable chronic tinnitus (n=40)18. It should be noted that the main study protocol was designed with three arms (anode, cathode, and placebo, each with 30 patients)18. However, due to several reasons we decided to conduct and report the results of the two arms of the protocol as a separate study. The two arms were anode (real group) and placebo groups. The remaining arms of the original study are under recruitment phase and will be conducted with two arms of cathode (n=30) and placebo (n=30). The main reason was difficulty in recruiting enough patients who meet the inclusion/exclusion criteria so that we could not conduct the whole study as a single trial within a season, and the duration of the study took a long time. In addition, we observed a significant different therapeutic outcomes between the real (anode arm) and placebo groups and decided to report the two arms as a separate study. The primary outcomes of this study are expected to be reported within approximately 4 months.\n\nThe data consist of three main groups: the first group is demographic information, tinnitus characteristics and audiometric assessments of the patients; the second group are data of the primary and secondary outcomes at pre- and post-intervention; and the third group are data concerning the adverse effects and tolerability of tDCS. Table 1 presents the demographic information, tinnitus characteristics (including tinnitus quality, laterality and duration), audiometric assessments, and the primary and secondary outcomes for pre- and post-intervention for the participants in the real tDCS. The corresponding data for the sham tDCS group are presented in Table 2. Table 3 and Table 4, respectively, present the adverse effects and tolerability of tDCS for real and sham tDCS groups using a customized questionnaire (Supplementary File 1).\n\nTHI, Tinnitus Handicap Inventory; pre, pre-intervention; post, post-intervention; post-i, immediately after intervention; post-1h, at 1 hour post-intervention; post-1w, at 1 week post-intervention; post-1m, at 1 month post-intervention. Tinnitus loudness and distress ranged 0–10, where 0 indicates the lowest level and 10 indicates the highest tolerable level. aTinnitus quality codes: R, ringing; B, buzzing; H, hissing; HU, humming; T, ticking; HPW, high-pitched whistling; TH, thumping; C, cicadas; P, pulsating. bTinnitus side: L, left; R, right, R = L, bilateral with no lateralization; R>L, bilateral lateralizing more to the right side; L>R, bilateral lateralizing more to the left side. cClass of hearing loss: N, normal hearing threshold (<20 dB); L, mild hearing loss (20–40 dB); M, moderate hearing loss (41–70 dB); S, severe hearing loss (70–90 db); P, profound hearing loss (<90 db).\n\nTHI, Tinnitus Handicap Inventory; pre, pre-intervention; post, post-intervention; post-i, immediately after intervention; post-1h, at 1 hour post-intervention; post-1w, at 1 week post-intervention; post-1m, at 1 month post-intervention. Tinnitus loudness and distress ranged 0–10, where 0 indicates the lowest level and 10 indicates the highest tolerable level. aTinnitus quality codes: R, ringing; B, buzzing; H, hissing; HU, humming; T, ticking; HPW, high-pitched whistling; TH, thumping; C, cicadas; P, pulsating. bTinnitus side: L, left; R, right, R = L, bilateral with no lateralization; R>L, bilateral lateralizing more to the right side; L>R, bilateral lateralizing more to the left side. cClass of hearing loss: N, normal hearing threshold (<20 dB); L, mild hearing loss (20–40 dB); M, moderate hearing loss (41–70 dB); S, severe hearing loss (70–90 db); P, profound hearing loss (>90 db).\n\nNA, not applicable. aC, under cathode; A, under anode. Very high tolerability indicates the subject could easily tolerate the tDCS sessions.\n\naC:under cathode, A:under anode, C-A: between cathode and anode; very high tolerability indicates the subject could easily tolerate the tDCS sessions.\n\nAll of the experimental procedures of the study were approved by local ethical committee of Ahvaz Jundishapur University of Medical Sciences (AJUMS), Ahvaz, Iran (registration code: IR.AJUMS.REC.1394.639), and were completely in accordance with the ethical standards and regulations of human studies of the Helsinki declaration (2014). After the enrolment, objectives, possible benefits, and side effects of the study were clearly explained to the patients and all patients filled and signed a written consent form for participation in the study.\n\nPatients with intractable chronic tinnitus (n=40) were randomly assigned into real tDCS (n=25; female, 14; male, 11; age, 47.52 ± 7.51 years; disease duration, 7.48 ± 3.99 years) and placebo tDCS (n=15; female, 8; male, 7; age, 47.67 ± 7.96 years; disease duration, 7.60 ± 3.60 years) in a parallel allocation and a double-blinded randomized controlled clinical trial (Figure 1). Due to the difficulty in the patient recruitment and high number of the patients who did not complete the treatment we reduced the size in the anode and placebo group (n=25 vs n=30 for the anode group and n=15 vs n=30 for the placebo group). The change in the groups’ size was approved by the local ethics committee of Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran (registration code: IR.AJUMS.REC.1394.639). The treatment consisted of daily sessions of 20 minutes of 2 mA current for 5 consecutive days per week and 2 consecutive weeks with 35 cm2 electrodes. Both groups were matched in age, gender, ethnicity, and audiometric main characteristics. The patients, the researcher who evaluated the outcomes during the experiments and at the follow-up period, and the researchers who performed data analyses were blinded to the type of protocol. The experimental procedures of the present study, including tDCS sessions and outcomes evaluations, were performed at the Bioelectromagnetic Clinic in Ahvaz Imam Hospital, an affiliated Hospital to AJUMS, Iran. The study was registered as a clinical trial in the Iranian Registry of Clinical Trial (IRCT2016110124635N6), which is the registration for the original clinical trial design18.\n\nThe direct current was applied through a saline-soaked pair of carbon electrodes (35 cm2) and delivered by a tDCS device (OASIS ProTM; Mind Alive, Inc., Edmonton, Alberta, Canada). The tDCS protocol consisted of 2 mA current, daily for 20 min, over 5 consecutive days per week for 2 consecutive weeks (total, 10 sessions). The anode was placed over left AC (halfway T3 - F7) and cathode over right AC (halfway T4 - F8) with 35 cm2 electrodes. In the placebo tDCS, the electrode montage was the same with real tDCS except that the device was turned off after 30 s after the start of session without the patient knowing it19. During treatment, the patients were asked to remove all metal-based jewelry from the head and neck.\n\nBefore the start of the first session of tDCS, intervention, all patients underwent complete audiometric and neurological assessments by expert specialties and the data were recorded along with the demographic data (Table 1 and Table 2). The recorded variables consisted of the tinnitus quality, lateralization, duration, and class of hearing loss in both ears (Table 1 and Table 2). The lateralization or tinnitus side is the dominant side of the head where the patient experiences the tinnitus and was classified as left (L), right (R), bilateral with no lateralization (R = L), bilateral lateralizing more to the right side (R>L), bilateral lateralizing more to the left side (L>R). In addition, hearing assessments were conducted in an acoustically isolated chamber with pure-tone audiometry using an AC 40 dual-channel Audiometer (Intracoustics Co., Denmark). The hearing thresholds were recorded over the frequency ranges of 250 to 8000 Hz for air conduction and 500 to 4000 Hz for bone conduction pathways, according to the modified Hughson–Westlake Method proposed by ANSI 199720. Pure-tone audiometry was considered normal whenever the hearing thresholds at all frequencies were below 20 decibels hearing level (dBHL). The class of hearing loss in both ears was classified as normal hearing threshold (<20 dB), mild hearing loss (20–40 dB), moderate hearing loss (41–70 dB), severe hearing loss (70–90 db), and profound hearing loss (>90 db). The hearing class was determined as the average of threshold in 250, 1000, 2000, and 4000 Hz20. The THI scores were assessed at pre-intervention, immediately after intervention, and at 1 month after the last stimulation. The tinnitus loudness and distress were recorded using a numeral 0–10 VAS rating scale before intervention, and immediately, 1 hour, 1 week, and 1 month after last stimulation.\n\nPrevious studies have demonstrated that tDCS is relatively safe with no serious side effects11,15,19. It should be mentioned that most of the conducted studies on the safety profile of tDCS assessed during single-session tDCS with different current intensities. For this study, considering the chronic protocol of tDCS, the adverse effects of and tolerability to the chronic protocol of tDCS were assessed using a customized questionnaire (Supplementary File 1). We used a five-point Likert-type scale for each adverse effect and the tolerability. In addition, the site (under cathode, under anode, other (mention) and time of sensation (beginning, middle, and end of session), and duration of the sensation (very short, some minutes, throughout, and after termination of session) end for each session were recorded. The questionnaire was filled before the start of the tDCS intervention and after the last session.\n\nThe measured data were recorded in customized designed forms which were available in print. All the collected data were entered into the specific forms in the Excel (Microsoft Office, 2010). The data entry was double-checked by two independent researchers. After validation of the data by a third researcher, the data were checked for wrong and out-of-range values and any dispute was resolved referring to the print version of the dataset. For the missing data, the missing data point was imputed using linear interpolation to reach the averaged values of the sequential closest values. After the validation, the data were sorted and then entered into statistical package of SPSS (Version 20, Windows) for analyses.\n\n\nUtility and discussion\n\nThis dataset has two main advantages over similar studies so far conducted. First, tinnitus-specific features as well as audiometric measures, including the laterality and quality of tinnitus, pure tone auditory threshold and class of hearing impairments for both ears, are assessed and recorded for all patients. This feature allows researchers to investigate the possible correlations between the response and/or non-response to tDCS and each of these variables, along with gender and history of tinnitus. However, to find reliable correlations, it is necessary to increase the volume of such datasets, and the similar data from other clinical trials can help to build a comprehensive dataset of tinnitus-specific features. The second feature is that this dataset presents the effects of relatively long-term tDCS exposure (10 sessions, one session daily) on tinnitus symptoms; the primary and secondary outcomes were measured at several time points following intervention for 1 month. The outcomes were assessed at different time points covering the short and long term after effects. A 1-month follow-up is relatively long compared to the similar studies14–16 conducted so far; however; it is necessary to increase the follow-up assessments further to several months. Such datasets can be used alone or in combination with other datasets, using advanced statistical analyses and modeling to identify the influencing parameters of tDCS as well as patient-specific features correlating with the therapeutic outcomes of the tDCS.\n\n\nConclusion\n\nThis dataset presents the effects of tDCS in tinnitus symptoms in a double-blinded randomized placebo-controlled trial with a 1-month follow-up. Considering the long-term tDCS exposure, the AEs and tolerability to tDCS were also presented. One of the limitations of this study was relatively low sample size, and the imbalance in sample size between the real and sham tDCS groups. For the latter limitation, which could result in overestimation of the effect size in the real tDCS groups, the authors could conduct bootstrap analyses to determine whether overestimation or underestimation occurred in each group. In line with this trial, our group is conducting a series of clinical trials to reach an effective tDCS treatment for tinnitus. Adding the dataset of other trials to this data will allow researchers to quantitatively evaluate the factors influencing the treatment efficacy and to build different treatment models for tinnitus based on the tinnitus-specific features.\n\n\nData availability\n\nThe datasets of this study are freely available in the Mendeley repository under a CC BY 4.0 license, DOI: https://doi.org/10.17632/8d8wrk62vy.121.\n\n\nAbbreviations\n\nAC: auditory cortex; tDCS: transcranial direct current stimulation; THI: tinnitus handicap inventory; BDI-II: Beck Depression inventory; BAI: Beck Anxiety Inventory; dBHL: decibels hearing level.",
"appendix": "Competing interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis study is financially supported by Ahvaz Jundishapur University of Medical Sciences (AJUMS), Ahvaz, Iran (Grant No: U-94187).\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\nAuthors would like to thank the personnel of Khuzestan Cochlear Implant Center, Ahvaz, Iran for their assistance in patients recruiting and regular monitor.\n\n\nSupplementary material\n\nSupplementary File 1. Adverse effects and tolerability questionnaire.\n\nClick here to access the data.\n\n\nReferences\n\nSanchez L: The epidemiology of tinnitus. Audiol Med. 2004; 2(1): 8–17. Publisher Full Text\n\nEggermont JJ: Pathophysiology of tinnitus. Prog Brain Res. 2007; 166: 19–35. PubMed Abstract | Publisher Full Text\n\nElgoyhen AB, Langguth B, De Ridder D, et al.: Tinnitus: perspectives from human neuroimaging. Nat Rev Neurosci. 2015; 16(10): 632–42. PubMed Abstract | Publisher Full Text\n\nWeisz N, Müller S, Schlee W, et al.: The neural code of auditory phantom perception. J Neurosci. 2007; 27(6): 1479–84. PubMed Abstract | Publisher Full Text\n\nShiozawa P, da Silva ME, Cordeiro Q, et al.: Transcranial direct current stimulation (tDCS) for the treatment of persistent visual and auditory hallucinations in schizophrenia: a case study. Brain Stimul. 2013; 6(5): 831–3. PubMed Abstract | Publisher Full Text\n\nGeorge MS, Padberg F, Schlaepfer TE, et al.: Controversy: Repetitive transcranial magnetic stimulation or transcranial direct current stimulation shows efficacy in treating psychiatric diseases (depression, mania, schizophrenia, obsessive-complusive disorder, panic, posttraumatic stress disorder). Brain Stimul. 2009; 2(1): 14–21. PubMed Abstract | Publisher Full Text\n\nSchlaug G, Renga V, Nair D: Transcranial direct current stimulation in stroke recovery. Arch Neurol. 2008; 65(12): 1571–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRauschecker JP, Leaver AM, Mühlau M: Tuning out the noise: limbic-auditory interactions in tinnitus. Neuron. 2010; 66(6): 819–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVanneste S, Plazier M, der Loo Ev, et al.: The neural correlates of tinnitus-related distress. NeuroImage. 2010; 52(2): 470–80. PubMed Abstract | Publisher Full Text\n\nkhoramzadeh S, Saki N, Davoodi I, et al.: Investigating the Therapeutic Efficacy of Neurofeedback Treatment on the Severity of Symptoms and Quality of Life in Patients with Tinnitus. Int J Ment Health Addict. 2016; 14(6): 982–92. Publisher Full Text\n\nYadollahpour A, Jalilifar M, Rashidi S: Transcranial Direct Current Stimulation for the Treatment of Depression: a Comprehensive Review of the Recent Advances. Int J Ment Health Addict. 2017; 15(2): 434–43. Publisher Full Text\n\nMondino M, Haesebaert F, Poulet E, et al.: Efficacy of cathodal transcranial direct current stimulation over the left orbitofrontal cortex in a patient with treatment-resistant obsessive-compulsive disorder. J ECT. 2015; 31(4): 271–2. PubMed Abstract | Publisher Full Text\n\nYadollahpour A, Asl HM, Rashidi S: Transcranial direct current stimulation as a non-medication modality for attention enhancement: A review of the literature. Research Journal of Pharmacy and Technology. 2017; 10(1): 311–6. Publisher Full Text\n\nFrank E, Schecklmann M, Landgrebe M, et al.: Treatment of chronic tinnitus with repeated sessions of prefrontal transcranial direct current stimulation: outcomes from an open-label pilot study. J Neurol. 2012; 259(2): 327–33. PubMed Abstract | Publisher Full Text\n\nHyvärinen P, Mäkitie A, Aarnisalo AA: Self-Administered Domiciliary tDCS Treatment for Tinnitus: A Double-Blind Sham-Controlled Study. PLoS One. 2016; 11(4): e0154286. PubMed Abstract | Publisher Full Text | Free Full Text\n\nForogh B, Mirshaki Z, Raissi GR, et al.: Repeated sessions of transcranial direct current stimulation for treatment of chronic subjective tinnitus: a pilot randomized controlled trial. Neurol Sci. 2016; 37(2): 253–9. PubMed Abstract | Publisher Full Text\n\nYadollahpour A, Bayat A, Rashidi S, et al.: Dataset of acute repeated sessions of bifrontal transcranial direct current stimulation for treatment of intractable tinnitus: A randomized controlled trial. Data Brief. 2017; 15: 40–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBayat A, Mayo M, Rashidi S, et al.: Repeated sessions of bilateral transcranial direct current stimulation on intractable tinnitus: a study protocol for a double-blind randomized controlled trial. F1000Res. 2018; 7: 317. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoreisz C, Boros K, Antal A, et al.: Safety aspects of transcranial direct current stimulation concerning healthy subjects and patients. Brain Res Bull. 2007; 72(4–6): 208–14. PubMed Abstract | Publisher Full Text\n\nLloyd LL, Kaplan H: Audiometric interpretation: a manual of basic audiometry. University Park Press; 1978. Reference Source\n\nYadollahpour A, Mayo M, Rashidi S, et al.: Effects of long term transcranial direct current stimulation over auditory cortex on tinnitus symptoms and adverse effects. Mendeley Data, v1. 2018. Data Source"
}
|
[
{
"id": "34973",
"date": "03 Jul 2018",
"name": "Myles Jones",
"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 is well designed, double-blinded and randomized. Despite some non-completions the study is still reasonably ‘well-powered’, in that the n is reasonable compared to that typically reported in the literature. There is a clear flow diagram explaining how subjects were assigned to conditions. There is a wide range of outcome measures including those of auditory function, perception of tinnitus and well-being. There are several outcome time points which is also a strength of the design. The introduction may have liked to have provided a rationale for only investigating cathodal rather than anodal TDCS. Presumably the putative inhibitory actions of cathodal TDCS (rather than the excitatory actions of anodal TDCS) would be expected to reduce tinnitus symptoms.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "38489",
"date": "08 Oct 2018",
"name": "Saeid Mahmoudian",
"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 well-protocol using bilateral transcranial direct current stimulation over auditory cortex for tinnitus treatment. Tinnitus is quite common, yet not well understood. Understanding the etiology and neural basis of tinnitus would help in developing treatment: both alleviation and even remediation. In addition brain research on tinnitus can also reveal important information on the auditory system and its principles in general. As in many neurological conditions, the challenge in studying tinnitus lies in its heterogeneity and the comorbidity with other conditions affecting auditory processing. When regarding tinnitus the hearing loss is perhaps the most obvious mixing factor. In my opinion, this protocol provides a fairly good starting point towards investigating the effects of a chronic protocol of bilateral tDCS over auditory cortex (AC) with one-month follow-up in a double-blinded randomized clinical trial. The study is reasonably well planned and conducted. However, there are some concerns about the efficacy and safety of tDCS and finding a specific electrode location on the patient's scalp in order to achieve optimal electrode positioning.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-733
|
https://f1000research.com/articles/7-732/v1
|
12 Jun 18
|
{
"type": "Research Article",
"title": "Measuring the effectiveness of maternal delivery services: A cross-sectional and qualitative study of perinatal mortality in six primary referral hospitals, Kenya",
"authors": [
"Richard Ayah",
"Dismas Ongore",
"Alfred T.O. Agwanda",
"Dismas Ongore",
"Alfred T.O. Agwanda"
],
"abstract": "Background: The effective performance of hospitals is critical to overall health system goal achievement. Global health system performance frameworks are often used as part of global benchmarking, but not within low and middle-income countries as part of service delivery performance measurement. This study explored the utility of perinatal mortality as a measure of hospital effectiveness. Methods: A cross sectional, mixed methods study of six primary referral hospitals, differentiated by ownership, was conducted from 10th June to 9th October 2015. Monthly summary hospital data of maternal delivery services (MDS) were abstracted to determine the perinatal mortality. Tests of associations were used to correlate bed turnover, skilled staffing, method of delivery and perinatal mortality. Additionally, 40 questionnaire interviews were held with hospital board members and the management team to assess the availability of standard operating procedures (SOP) in MDS. Qualitative data was analysed thematically. Results: All six hospitals reported having SOP in managing MDS. The average perinatal mortality rate for all the hospitals was 24.63 per 1,000 live births. However, a perinatal death was 2.6 times more likely in public hospitals compared to private hospitals (29.8 vs 11.4 per 1,000 births respectively). The average caesarean section rate for all hospitals was 25.9%, but the odds of a caesarean section were 1.67 higher in a private hospital compared to a public hospital (P<0.001 95% CI: 1.58-1.77). Perinatal mortality was associated with bed turnover ratio (R squared 0.260, P=0.001), and skilled staff availability (R squared 0.064,P<0.001). Discussion: The high perinatal mortality reported in public hospitals may be due to high bed turnover and relatively low caesarean section rate. Input measures of performance such as reporting standards of care and staffing levels are not useful performance indicators. Perinatal mortality as a performance indicator may be an ideal measure of the effectiveness of hospitals.",
"keywords": [
"Health system effectiveness",
"Hospital performance",
"perinatal death"
],
"content": "Introduction\n\nThe hospital sector represents approximately 45–69% of government health expenditure in sub-Saharan Africa and the effective performance of the sub-sector is therefore critical to overall health system goal achievement1. Globally considerable efforts have been made to develop health system performance (HSP) assessment frameworks that take into consideration the peculiarities of health systems and the multiplicity of stakeholders in health with different perspectives2. But HSP benchmarking is often done between countries as part of a global health comparison, rather than being used at a subnational level, where policymakers in low income countries with high disease burden seek to understand how well the delivery of healthcare meets the needs of citizens3.\n\nIn Kenya, overall health status is measured by indicators including life expectancy, and under-five and maternal mortality4. However, health system performance is measured mainly through process input indicators such as health per capita spend and human resource availability5. This disconnect leads to poor performance accountability defined as “demonstrating and accounting for performance in the light of agreed-upon performance targets focusing on services, outputs and results”6. An ideal health system performance indicator would link hospital process and outcomes to overall health system effectiveness, allow for hospital comparisons, be sensitive to outcomes under the control of the health system and ensure provider accountability6,7.\n\nMaternal delivery service (MDS) indicators and outcomes, such as skilled delivery levels, coverage of caesarean sections and neonatal mortality, are sensitive indicators of the effectiveness of the whole health system8. The core impact indicators are also well defined; however the Every Newborn Action Plan (ENAP), launched in 2014, recognised that efforts are needed to improve data quantity and quality, with only 17 countries that have a policy for reporting and reviewing stillbirths and neonatal deaths9,10.\n\nRoughly one third of 363 interventions in the Kenya Essential Package for Health focus on reproductive health11. Despite the focus, Kenya did not meet the Millennium Development Goal (MDG) target for maternal deaths of 147 per 100,000 live births by 2015, and little advancement has been made in reducing mortality among newborns, which now accounts for 45% of all child deaths4,12. Facility-based delivery has gained traction as a key strategy for reducing perinatal mortality in developing countries13. In Kenya, healthcare provision is devolved to the 47 counties, which provide care to geographical defined populations14. In the delivery of MDS, primary referral hospitals are expected to provide comprehensive emergency obstetric care, which includes all basic emergency obstetric care interventions and caesarean sections15,16.\n\nEfforts to reduce maternal mortality and morbidity in low-resource settings often depend on global standards and indicators to assess obstetric care. However these standards often do not take into account the local context especially in terms of skill and resource availability17. Moreover, using a national average does not provide timely and accurate measurements of levels and trends at local levels, which are crucial to assess progress, allow benchmarking and provide policymakers with the data to prioritize the areas of greatest need18.\n\nThis study explored the utility of perinatal mortality as a measure of hospital effectiveness in six primary referral hospitals in Kenya.\n\n\nMethods\n\nA cross-sectional study of six primary referral hospitals in Kiambu and Nairobi Counties differentiated by ownership was conducted. In 2013, Kiambu County was estimated to have a population of 1,838,397 including 59,191 pregnant women19. In Kiambu, there were six faith-based, one private and four government hospitals. Nairobi County’s population in 2013 was estimated at 3,554,261 including 172,143 pregnant women19. Nairobi had four faith-based, seven private and two government hospitals. Kiambu and Nairobi Counties were chosen for this study because compared to the national averages (32%), health facilities in Kiambu and Nairobi counties (40% and 48%, respectively) had above average maternal health service readiness19. Census data analysis of the county Maternal Mortality Ratio (MMR) estimated Kiambu and Nairobi at 230 and 212 per 100,000 live births, respectively, roughly half the national average (495 per 100,000)20.\n\nAll the level four health facilities, that is primary referral hospitals, were picked from the list of hospitals in the two counties. The hospitals were grouped according to ownership, public (government), not for profit, faith-based and for profit hospitals. In the two counties there were six public, eight private and ten faith-based hospitals. Hospitals that did not offer maternal delivery services were excluded. A list of all public hospitals was developed and computer generated random numbers were used to select three government hospitals, which were selected and then matched by bed capacity with two faith-based and one for profit hospitals across both counties.\n\nData was collected from 10th June to 9th October 2015. Monthly summary hospital data of patients who had been admitted to the maternity unit of each selected hospital in the period 1st January 2014 – 31st December 2014 were abstracted between 10th June and 9th October 2015 to determine: number of patients admitted, type of delivery, skilled staff per 1,000 deliveries, length of stay, bed capacity, bed turnover ratio, caesarean section rate, number of perinatal deaths, perinatal mortality per 1,000 live births.\n\nAdditionally, 40 questionnaire interviews were held with board members and members of the hospital management team to assess the availability of standard operating procedures in MDS. In each hospital, a minimum of three board members (including the chair, chief executive and one other), and ensuring that at least one third of members were interviewed. For each hospital management team, the medical superintendent, hospital nursing officer in charge, administrator and nurse in charge of maternity unity were interviewed. Consequently, the combined participants from the six facilities provided at least 40 interviewees - an adequate medium size sample pool of interviews (Baker and Edwards, 2012). Informed written consent was sought with interviews audio recorded except where participants were uncomfortable, only field notes were taken (Supplementary File 1). The length of stay was determined by abstracting dates of admission and discharge from 200 randomly selected patient files from each hospital.\n\nEffectiveness of MDS was defined as the extent to which the hospital manages all major causes of maternal and newborn mortality as measured by the perinatal mortality rate. The World Health Organization defines perinatal mortality as the “number of stillbirths and deaths in the first week of life per 1,000 total births”. The perinatal mortality rate was calculated as: (No. of perinatal deaths / total No. of births (still births + live births)) x 1000.\n\nCorrelations and tests of associations of chi-square (X2) were used to show the relationships between MDS patients, bed turnover, average length of stay, skilled delivery staff, bed capacity and patient outcomes of normal, caesarean section; and perinatal mortality. Data was analysed using the Statistical Products and Service Solutions (SPSS) and MS-Excel.\n\nQualitative data was analysed thematically, by manually reviewing the transcripts. Using priori codes emanating from the questionnaire, a code book (Supplementary File 2) was developed that provided a working analytical framework that was then used to code the transcripts. Two independent coders reviewed the transcripts and consequently agreed on emergent codes and resultant thematic findings.\n\nEthical clearance was obtained from the Ethics and Research Committee of Kenyatta National Hospital and University of Nairobi (P128/03/2015). To facilitate carrying out the study, administrative consent was obtained from both Kiambu County and Nairobi County to facilitate access to the hospitals. Before starting data collection at each hospital, written consent was obtained from each facility in-charge. Respondents in the study were asked to provide informed written consent before being interviewed.\n\n\nResults\n\nThe six hospitals ranged in maternity bed capacity from 13 – 70 with a median of 55 beds. Total deliveries in the calendar year ranged from 381 at the 13 maternity bed private hospital to 8,279 at a 70 bed public hospital. The bed turnover ratio ranged from 29 – 163 with a median of 80. The lowest number of perinatal deaths was 1, while the highest was 208. The average length of stay varied from 0.7 to 5.1 days and was associated with perinatal mortality P<0.001, 95%CI: 0.6472–0.7542) (Table 1).\n\nPFP=Private for Profit Hospital; P=Public Hospital; FBO=Faith-based Organisation Hospital\n\nThe average caesarean section rate for the all the hospitals was 25.9%. When the public hospitals (P) were grouped together and compared to the private [for profit (PFP) and faith based organisation hospital (FBO)], public hospitals had caesarean section rates of 18.4%, 23% and 27.1%, (P2-Kiambu, P3-Nairobi and P1-Kaimbu, respectively), while the private hospital caesarean sections rates were 31.6%, 42.5% and 43.4% (FBO1-Kiambu, PFP-Nairobi, FBO2-Kiambu, respectively). The odds of a caesarean section were 1.67 higher in a private hospital compared to a public hospital (P<0.001 95% CI: 1.5833-1.7763). The number of perinatal deaths per 1,000 live births in private hospitals were 2.62, 9.91, 13.15 (PFP-Nairobi, FBO2-Kiambu, FBO1-Kiambu, respectively), while in the public hospitals they were 25.12, 29.74, 39.17 (P2-Kiambu, P3-Nairobi, P1-Kiambu, respectively) (Table 2).\n\nThe perinatal death rate was 2.6 times higher in public hospitals (29.76 per 1,000 births) compared to private hospitals (11.39 per 1,000 births). The number of skilled delivery staff available per 1,000 patients were as follows: P1-Kiambu, 4; P2-Kiambu, 7; FBO1-Kiambu, 15; P3-Nairobi, 22; FBO2-Kiambu, 34; PFP-Nairobi,235. Despite the wide range of skilled delivery staff availability, there was an association between the skilled staff availability and the perinatal mortality (R squared 0.260, P<0.001). The bed turnover ratio and perinatal mortality were associated (R squared 0.064, P<0.001).\n\nFrom the 40 interviews conducted the following information was found. All six hospitals reported having standard operating procedures in managing MDS. Three of the facilities, P2-Kiambu, P1-Kiambu and FBO2-Kiambu, reported having annual work plans. All six hospitals had a scheme of service and code of conduct for their employees. None reported having been inspected by the national Ministry of Health, but county health teams had visited all the hospitals in the past year except for PFP hospital. With respect to inspection by a non-MOH regulator such as the National Environmental Agency, only FBO2-Kiambu and PFP-Nairobi had interactions (Table 3).\n\n‘X’ = availability of method of standards of care, MoH – Ministry of Health; SOP – standards operating procedures\n\nAll the hospitals reported having SOPs in managing maternal delivery services. All had a scheme of service and code of conduct for their employees. None reported having been inspected by the national ministry of health, but county health teams visited.\n\n“…I want to say that the county comes to the ground very often and especially to check and monitor maternal outcomes and hold discussion…which we do with them...” (Hospital Management Team Member, FBO1-Kiambu).\n\nHalf the facilities reported having annual work plans. With respect to following external regulations only the non-government hospitals were subjected to some inspection by non ministry of health regulator such as NEMA. Of interest is that the hospitals with the worst perinatal mortalities reported to have the same number of methods to maintain standards of service as the hospital with the best mortality figures (Table 2).\n\nRespondents at senior management level also reported a lack of engagement with ministry of health at county and national level in strategy development of maternal delivery services.\n\n“…The county ministry of health have not given us an opportunity to contribute towards the formation of the county health strategy...we would want to be equal partners in provision of strategy and implementation strategy...we will go for discussions… they will set a circular and we are told from now on xyz will be happening and you are not involved...” (Hospital Management Team Member, FBO2-Kiambu.\n\n\nDiscussion\n\nCounty referral hospitals play an increasingly significant role in maternal delivery services. Kenya recorded an increase in the proportion of facility based deliveries from 44% in 2008 to 61% in 201420. A total of 186,688 deliveries in 2014 occurred in such facilities, roughly 15% of all deliveries in Kenya21. In this study the total number of deliveries were 24,534 in 2014. The average length of stay for patients admitted for maternal delivery services was just under 2.7 days (median 2 days). This is in line with global practice; in a 92 country review of hospital the mean length of stay after child birth ranged from 1.3 to 6.6 days with the majority of women staying too short a time to receive adequate postnatal care22. However, the average masked considerable differences between the different hospitals, with the public hospitals discharging patients in less than 24 hours compared to two days for private hospitals. Developing countries have reported neonatal infection rates 3–20 times that of developed countries due to poor intra-partum and postnatal infection-control practices23. Neonatal infection in the first week of life account for 26% of neonatal deaths in sub-Saharan Africa24. The short time available to monitor the mother and newborn could explain the association between the high bed turnover ratio in the public hospitals and high perinatal mortality.\n\nThis study reported the average perinatal mortality rate for all the hospitals at 24.63 per 1,000 live births. This is a little higher than the national average of 22.5 per 1,000 live births in 201525. Yet it is reported that in sub-Saharan Africa the risk of perinatal mortality is 21% higher for home compared to facility-based deliveries13. In agreement with our study results, a Bangladesh study looking at whether facility delivery modified the risk of intra-partum related perinatal deaths found that the risks were higher for facility deliveries compared to home deliveries26. The reported perinatal mortality in the present study compared poorly with an assessment of facility quality and association with neonatal mortality in Malawi; it found an average of 17 per 1,000 live births with the newborn mortality rate of 28 per 1,000 births at low-quality facilities and of 5 per 1,000 births at the top 25% of facilities27. However, perinatal mortality in South Africa was reported at 33.4 deaths per 1,000 births in 201328; while a cross sectional descriptive study of eight major hospitals in Dar es Salaam in January 2009 established a perinatal mortality rate was 44/1000 births (range: 17 – 147)29.\n\nPublic hospitals had a perinatal mortality rate 2.6 times higher than private hospital (29.8 and 11.4 per 1,000 live births respectively) in the present study. In Bangladesh, the risk of perinatal mortality in a public health facility was twice that of a private facility, with the difference being attributed to quality of care26. Emergency obstetric care including caesarean section has been recommended as the first priority intervention in reducing stillbirths30. One of the key differences observed in our study was that mothers in private hospitals were almost twice as likely to undergo a caesarean section compared to those in a public hospital. The reported difference here is consistent with studies that show that regardless of a woman’s risk and contextual factors, for profit hospitals are more likely to perform caesarean sections compared to not for profit hospitals because of financial incentives31. As expected, the hospitals with higher caesarean section rates also reported lower perinatal mortality32.\n\nPrevious studies have shown strong associations between patient mortality and low staffing levels33,34. However, there was weak association between the ratio of skilled birth attendant and patients, indicating that perhaps primary referral hospitals had met the threshold of minimum number of staff required. This finding contrasts with a study that found that the presence of a doctor at birth reduced maternal and infant mortality35. Since staffing numbers weakly predicted mortality rates, it can be hypothesised that the quality of clinical decision making in not identifying mothers requiring caesarean sections was poorer in those hospitals with relatively low caesarean section rates.\n\nAll the hospitals in this study reported having at least two, with a median of three, methods of monitoring standards of care, including SOPs in managing maternal delivery services. Public hospitals in particular had more methods compared to private, though none reported having Ministry of Health oversight. This finding may appear to contradict an assessment of the quality of maternity care in an Indian metropolitan city that concluded that public hospitals practices fell short of evidence-based guidelines, while there was relative overuse of interventions in private hospitals36. However, it is known that reporting having certain quality improvement methodologies is not enough to lead to an outcome of high standards of care without an imbedded culture of quality improvement37. It is possible that a weakness in the health system building blocks of leadership and health information are the weak links in not enabling health workers to champion the process of improvement and use of best practice guidelines to monitor performance10.\n\nIn this study, none of the hospitals reported having been inspected by the national Ministry of Health, but county health teams had visited the three public hospitals. Despite these visits and the reporting of available protocols to ensure quality of care, the public hospitals reported worse perinatal outcomes. Quality-of-care audits have been promoted as useful in identifying and changing suboptimal care, and therefore reducing perinatal mortality; however a study in South Africa, did not demonstrate that quality-of-care audits improved perinatal mortality28. However, in the monitoring of external non medical regulations only the non-government hospitals were subjected to inspection by a non ministry of health regulator such as the National Environmental Agency, symptomatic of many policies in developing countries where there are often conflicting existing laws and legislations and overlapping bureaucratic mandates leading to policy implementation failure38. This uneven treatment of hospitals does not augur well for policymakers and hospital boards efforts to relate hospital performance to overall health system effectiveness.\n\nFocusing on a specific service such as MDS allows for greater comparison between hospitals because patient heterogeneity, which can be a major factor in measuring effectiveness, is reduced1,39,40. This study chose mortality as the health outcome because it is relatively easy to measure and therefore likely to achieve valid results. However, the study relied on hospital records, and therefore there may have been elements of underreporting. Perinatal mortality includes live births and still births, which is a comprehensive indicator for assessing outcomes of both intrapartum and immediate post-partum care services; however without perinatal audit systems in place, there is often underreporting of stillbirths13,41.\n\n\nConclusions and recommendations\n\nThe study demonstrates that the average perinatal mortality in primary referral hospitals was high with considerable variation between public and private hospitals. This is despite all the hospitals reporting having various methods of maintaining standards of care. While there was considerable variance in size and patient numbers among the hospitals, staffing levels were not associated with perinatal mortality, suggesting that the quality of clinical decision making as measured by the caesarean section rate was a factor in improving outcomes. Given the heterogeneity of primary referral hospitals, the use of perinatal mortality as a performance indicator to measure the effectiveness of maternal delivery services and hold hospitals to account in relation to the entire health system is recommended.\n\n\nData availability\n\nDataset 1: Effectiveness of maternal delivery services and perinatal mortality in six primary referral hospitals. DOI, 10.5256/f1000research.14862.d20606442\n\nThe datasets generated and/or analysed during this study, other than those provided herein, are not publicly available as ethical restrictions apply to publicly sharing the qualitative interviews transcripts due to potentially identifiable information detailed in the transcripts. Excerpts of the data are however available from the corresponding author on reasonable request and approval by the Kenyatta National Hospital/University of Nairobi- Ethics Review Committee (KNH/UoN-ERC) by contacting ayah@uonbi.ac.ke or uonknh_erc@uonbi.ac.ke.",
"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\nThe authors would like to acknowledge the Boards of the hospitals, study participants, Kenneth Mutai and Kellen Karimi.\n\n\nSupplementary material\n\nSupplementary File 1: Hospital governance questionnaire.\n\nClick here to access the data.\n\nSupplementary File 2: Code Book Perinatal Mortality.\n\nClick here to access the data.\n\n\nReferences\n\nZere E, Mbeeli T, Shangula K, et al.: Technical efficiency of district hospitals: evidence from Namibia using data envelopment analysis. Cost Eff Resour Alloc. 2006; 4: 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTashobya CK, da Silveira VC, Ssengooba F, et al.: Health systems performance assessment in low-income countries: learning from international experiences. Global Health. 2014; 10: 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNoor AM: Subnational benchmarking of health systems performance in Africa using health outcome and coverage indicators. BMC Med. 2015; 13: 299. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinistry of Health: Kenya Health Policy 2014–2030. Nairobi; 2014. Reference Source\n\nMinistry of Medical Services Kenya: Ministry of Medical Services Strategic Plan 2008–2012. Nairobi; 2008. Reference Source\n\nMartin Hilber A, Blake C, Bohle LF, et al.: Strengthening accountability for improved maternal and newborn health: A mapping of studies in Sub-Saharan Africa. Int J Gynecol Obstet. 2016; 135(3): 345–357. PubMed Abstract | Publisher Full Text\n\nAllin S, Grignon M: Examining the role of amenable mortality as an indicator of health system effectiveness. Healthc Policy. 2014; 9(3): 12–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMills S, Chowdhury S, Miranda E, et al.: REDUCING MATERNAL MORTALITY: Strengthening the World Bank Response. Washington (DC); 2009. Reference Source\n\nMoxon SG, Ruysen H, Kerber KJ, et al.: Count every newborn; a measurement improvement roadmap for coverage data. BMC Pregnancy Childbirth. 2015; 15 Suppl 2: S8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKerber KJ, Mathai M, Lewis G, et al.: Counting every stillbirth and neonatal death through mortality audit to improve quality of care for every pregnant woman and her baby. BMC Pregnancy Childbirth. 2015; 15 Suppl 2: S9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinistry of Public Health and Sanitation: Ministry of Public Health and Sanitation Strategic Plan 2008–2012. Nairobi; 2008. Reference Source\n\nMurphy GA, Gathara D, Aluvaala J, et al.: Nairobi Newborn Study: A protocol for an observational study to estimate the gaps in provision and quality of inpatient newborn care in Nairobi City County, Kenya. BMJ Open. 2016; 6(12): e012448. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChinkhumba J, De Allegri M, Muula AS, et al.: Maternal and perinatal mortality by place of delivery in sub-Saharan Africa: A meta-analysis of population-based cohort studies. BMC Public Health. 2014; 14: 1014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRepublic of Kenya: The Constitution of Kenya, 2010. Kenya: National Council for Law Reporting with the Authority of the Attorney General; 2010. Reference Source\n\nMinistry of Health: Kenya Health Sector Referral Strategy 2014–2018. Gov Kenya. 2014; 1–44. Reference Source\n\nNCAPD, Ministry of Medical Services, Sanitation M of PH: Kenya Service Provision Assessment (SPA) 2010. Nairobi; 2010. Reference Source\n\nSpangler SA: Assessing skilled birth attendants and emergency obstetric care in rural Tanzania: the inadequacy of using global standards and indicators to measure local realities. Reprod Health Matters. 2012; 20(39): 133–141. PubMed Abstract | Publisher Full Text\n\nColson KE, Dwyer-Lindgren L, Achoki T, et al.: Benchmarking health system performance across districts in Zambia: A systematic analysis of levels and trends in key maternal and child health interventions from 1990 to 2010. BMC Med. 2015; 13: 69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinistry of Health: Government of Kenya, 2014: Kenya Service Availability and Readiness Assessment Mapping (SARAM). Nairobi; 2014.\n\nKenya National Bureau of Statistics & ICF Macro: Kenya Demographic and Health Survey 2014. Nairobi; 2015. Reference Source\n\nNjuguna J, Kamau N, Muruka C: Impact of free delivery policy on utilization of maternal health services in county referral hospitals in Kenya. BMC Health Serv Res. 2017; 17(1): 429. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCampbell OM, Cegolon L, Macleod D, et al.: Length of Stay After Childbirth in 92 Countries and Associated Factors in 30 Low- and Middle-Income Countries: Compilation of Reported Data and a Cross-sectional Analysis from Nationally Representative Surveys. PLOS Med. 2016; 13(3): e1001972. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZaidi AK, Huskins WC, Thaver D, et al.: Hospital-acquired neonatal infections in developing countries. Lancet. 2005; 365(9465): 1175–88. PubMed Abstract | Publisher Full Text\n\nSeale AC, Mwaniki M, Newton CR, et al.: Maternal and early onset neonatal bacterial sepsis: burden and strategies for prevention in sub-Saharan Africa. Lancet Infect Dis. 2009; 9(7): 428–438. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeats EC, Ngugi A, Macharia W, et al.: Progress and priorities for reproductive, maternal, newborn, and child health in Kenya: a Countdown to 2015 country case study. Lancet Glob Heal. 2017; 5(8): e782–e795. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhanam R, Baqui AH, Syed MIM, et al.: Can facility delivery reduce the risk of intrapartum complications-related perinatal mortality? Findings from a cohort study. J Glob Health. 2018; 8(1): 010408. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeslie HH, Fink G, Nsona H, et al.: Obstetric Facility Quality and Newborn Mortality in Malawi: A Cross-Sectional Study. PLoS Med. 2016; 13(10): e1002151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAllanson ER, Pattinson RC: Quality-of-care audits and perinatal mortality in South Africa. Bull World Health Organ. 2015; 93(6): 424–428. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNyamtema AS, Urassa DP, Pembe AB, et al.: Factors for change in maternal and perinatal audit systems in Dar es Salaam hospitals, Tanzania. BMC Pregnancy Childbirth. 2010; 10: 29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhutta ZA, Yakoob MY, Lawn JE, et al.: Stillbirths: what difference can we make and at what cost? Lancet. 2011; 377(9776): 1523–38. PubMed Abstract | Publisher Full Text\n\nHoxha I, Syrogiannouli L, Luta X, et al.: Caesarean sections and for-profit status of hospitals: systematic review and meta-analysis. BMJ Open. 2017; 7(2): e013670. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcclure EM, Goldenberg RL, Bann CM: Maternal mortality stillbirth and measures of obstetric care in developing and developed countries. Int J Gynaecol Obstet. 2007; 96(2): 139–146. PubMed Abstract | Publisher Full Text\n\nKane RL, Shamliyan T, Mueller C, et al.: Nurse Staffing and Quality of Patient Care. Evid Rep Technol Assess (Full Rep). Rockville, MD; 2007; (151): 1–115. PubMed Abstract | Free Full Text\n\nAiken LH, Clarke SP, Sloane DM, et al.: Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002; 288(16): 1987–1993. PubMed Abstract | Publisher Full Text\n\nYego F, D’Este C, Byles J, et al.: A case-control study of risk factors for fetal and early neonatal deaths in a tertiary hospital in Kenya. BMC Pregnancy Childbirth. 2014; 14: 389. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNagpal J, Sachdeva A, Sengupta Dhar R, et al.: Widespread non-adherence to evidence-based maternity care guidelines: a population-based cluster randomised household survey. BJOG. 2015; 122(2): 238–247. PubMed Abstract | Publisher Full Text\n\nRaven J, Hofman J, Adegoke A, et al.: Methodology and tools for quality improvement in maternal and newborn health care. Int J Gynaecol Obstet. 2011; 114(1): 4–9. PubMed Abstract | Publisher Full Text\n\nBanchani E, Tenkorang EY: Implementation challenges of maternal health care in Ghana: the case of health care providers in the Tamale Metropolis. BMC Health Serv Res. 2014; 14: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOsei D, d'Almeida S, George MO, et al.: Technical efficiency of public district hospitals and health centres in Ghana: a pilot study. Cost Eff Resour Alloc. 2005; 3: 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlank J, Valdmans V: Evaluating Hospital Policy and Performance: Contributions from Hospital Policy and Productivity Research. 1st edition. Oxford: JAI Press, Elsevier Ltd; 2008. Reference Source\n\nBhutta ZA, Darmstadt GL, Haws RA, et al.: Delivering interventions to reduce the global burden of stillbirths: improving service supply and community demand. BMC Pregnancy Childbirth. 2009; 9 Suppl 1: S7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAyah R, Ongore D, Agwanda ATO: Dataset 1 in: Measuring the effectiveness of maternal delivery services: A cross-sectional and qualitative study of perinatal mortality in six primary referral hospitals, Kenya. F1000Research. 2018. Data Source"
}
|
[
{
"id": "34970",
"date": "16 Jul 2018",
"name": "Faith Yego",
"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 provides perinatal mortality as a performance indicator for measuring the effectiveness of hospitals. This study is important since most studies usually rely on maternal mortalities as indicators. However, since most perinatal deaths are underreported, especially in developing countries, the findings of this study provide a basis for documenting these deaths to report them to use the data to come up with interventions to improve service delivery.\nThe literature review is current and accurately presented. However, some references need to be checked and corrected for 11,14,15, 16, 20, 42.\nThe methodology is appropriate and scientifically sound.\nStatistical analysis is adequate though a review by a statistician would be recommended. Study findings are adequately presented in tables however some issues indicated below need to be clear:\nIt would be good to know what tier the public hospitals (p1, P2 and P3) were to explain certain factors like caesarean section rates based on patient turnover and hospital level. If they are referral hospitals then high CS rates could be due to the kind of emergency cases they receive.\nWhat is the difference between PFP and FBO? Aren’t they both private just that one is for profit and the other not for profit? This would help clarify the odds ratio for caesarean section rates.\nSince the study relied on hospital records, it would be important to understand how they handled missing data and how what impact it had on their findings.\nAuthors should provide a table showing correlations and test of associations\nDiscussion\nThe link between high CS rates and “because of financial incentives” is unclear whether there is a proven causal relationship between the two to justify the statement.\nThe authors mention in the final paragraph on Pg. 6 about inspection of health facilities by the MOH. Since these facilities are under the County government, and with devolution there are structures in place to ensure inspection of health facilities and roles are clear for national and County government. The statement provided in this paragraph may need to be revised to consider the current structures that are in place to monitor and regulate the County hospitals and clearly bring out the role of national government in these facilities.\nRecommendations for future/further research needs to be provided.\n*Revise some grammatical errors in text\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "47065",
"date": "29 Apr 2019",
"name": "Abiodun S. Adeniran",
"expertise": [
"Reviewer Expertise Feto-maternal medicine",
"Prenatal screening",
"Management of HIV in pregnancy."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comment: It is important to recognise that the focus of the study is of interest in contemporary obstetrics. However, the methodology employed by the authors is inappropriate which limited the information which the authors were able to extract from the hospital records of the participants. Therefore, the result obtained was ‘overstretched’ and ‘over-interpreted’ while attempting to use it to answer the research questions leading to inferences that are not derivable from the results. While the authors imply causal relationships between the measured parameters and perinatal mortality, there is insufficient evidence from the results to establish these claims. It is therefore unacceptable to draw such ‘strong’ causal conclusions without evidence as was done in this article. A retrospective study is not the appropriate study design to make such conclusions attributed to it in this article.\nTitle:\nInsert ‘in’ before Kenya\nAbstract:\n\nBackground: This is poorly worded and may be confusing; it should be revised to make it clearer. Change ‘discussion’ to conclusion. However, the opinion expressed in this sub-section is not derivable from available evidence in the study; it is more of assumptions and repetition of reports of previous studies. I do not agree that CS rates of 18.4%, 23% and 27.1% were too low. Authors to relate this to the recommended WHO CS rate.\nIntroduction:\nThis is too lengthy. It should end with a statement of the aim/objectives. The objective should be incorporated into the introduction instead of making it a separate sub-section.\nMethodology:\nThe information about the study facilities are grossly inadequate: I am interested in more information on the setting, cadres of staff, who takes decisions, were all patients reviewed by a doctor or the doctor is invited after complications has set in, were there on-call doctors who slept in the hospital, who monitored the labour and who performed the CS? The selection method for individual facilities especially the non-public is not clearly defined. How was sample size determined for each component of the study? How did you assess the qualitative component? Was it individual interview or focused-group discussion? Your definition of effectiveness of MDS as ‘the extent to which the hospital manages all major causes…’ is not clear. How was this assessed? There is no evidence that the bed capacity of the facilities were matched as claimed by the authors- what is the evidence of matching in 25,55 and 70 versus 13, 62, 54? The matching should be in terms of individual facilities, not aggregated. Do you mean that the 13 bedded facility was performing CS regularly? Are the public facilities free? Generally private facilities are more expensive compared to public hospitals; thus, low socioeconomic people are more likely to patronise them. The effect of low social class on perinatal outcome therefore becomes a confounding factor in this comparison with those from private facilities. In addition, the status of the mother and fetus at presentation is a well-known factor that affects perinatal mortality; this was not evaluated for in this study. In addition, antenatal and early labour events no doubt affect perinatal outcome, these were not considered. How many cases were emergencies, what were the identified complications? Do any of these facilities conduct maternal and perinatal death reviews? How many facilities had neonatologist? Which mode of anaesthesia are available and used. Availability of resuscitation facilities and mode of anaesthesia no doubt can affect perinatal mortality. For the qualitative component, I was expecting responses on mode of presentation, identified delays, how soon emergencies are attended to, limitations of care, issues about whether complaints were made earlier and administration's response, etc.\nResults:\n\nThe information provided is not enough.\nDiscussion:\nLine 3: change ‘were’ to ‘was’ Line 3: delete ‘just’. Lines 13 – 23: There is no evidence from this study to confirm that ‘the short time available to monitor the mother and newborn’ was responsible for the perinatal mortality. This can only be scientifically deduced if you compare women with good to those with poor perinatal outcome. This was not done. Page 4, first paragraph: You seem to be mixing up a number of issues here and the references are inappropriate. The ‘21% higher mortality perinatal mortality of facility over home deliveries’ does not imply a mortality rate of 21% as you suggest. Rather it is a comparison. Do you imply that it is better to deliver at home rather than hospital? Since the discussion and comparison here is not for home versus facility delivery, the comparison is inappropriate and will confuse the readers. It is ‘dangerous’ to encourage home delivery without skilled birth attendant in this century. Please remember that other reasons for the apparent ‘increased mortality’ for facility deliveries will include the fact that it is the ‘bad cases’ that typically comes to the hospital in low-resource countries. Thus, it will be unfair to expect too much in cases of severely compromised mother and fetus. We need to know the state of both mother and fetus at arrival- how many came with intrauterine fetal death, fetal distress or critically ill mothers? I wish to state that the ‘higher fatality’ from hospital deliveries may be difficult to establish because facility data is captured unlike home deliveries which are not recorded. On what data is the comparison made then- verbal reports? 2nd paragraph: there is no evidence to suggest that the quality of care in the public hospitals was lower than the private ones. Therefore, you are over-interpreting your limited data. 3rd and 4th paragraphs- there are no evidence in the study indicating a causal relationship. 5th paragraph: Although there was no monitoring by external agencies, the study did not explore in full the institutional audit and evaluation processes. Therefore, we cannot assume as was done by the authors. This study is seriously limited by inadequate parameters to validate the strong causal relationships claimed by the authors.\nReferences:\n\nMany references do not have last pages. Authors should insert the last pages while online only articles should have the DOI number inserted.\nRecommendation: This article cannot be indexed in its current form. I suggest that the authors should revise it along either of the suggestions below:\nIf they choose to maintain the available results, the article should be revised into a description while all causal claims not derivable should be deleted. Authors may choose to retrieve additional information as outlined above that will enable them to scientifically validate the causal claims.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-732
|
https://f1000research.com/articles/7-731/v1
|
12 Jun 18
|
{
"type": "Case Report",
"title": "Case Report: Corpus callosal apraxia following acute ischemic stroke",
"authors": [
"Santhosh Narayanan",
"Gomathy Subramaniam",
"Gomathy Subramaniam"
],
"abstract": "The corpus callosum is a compact structure that connects the right and left cerebral hemispheres. Here we report the case of a 50 year old woman who presented with features of corpus callosum apraxia, initially mistaken as psychiatric symptom by her relatives. Computed tomography and magnetic resonance of brain confirmed the diagnosis of acute ischemic infarct in the body of the corpus callosum. Isolated stroke involving the corpus callosum is rarely reported in literature and is a diagnostic challenge due to atypical clinical features.",
"keywords": [
"Corpus callosum",
"apraxia",
"disconnection syndrome",
"stroke"
],
"content": "Introduction\n\nCorpus callosum contains densely packed white matter tracts essential for coordination and integration of the two sides of brain. Pathology affecting the corpus callosum is very rare. Clinically it manifests with apraxia, cognitive dysfunction, disturbances of memory and may be mistaken for psychosis. Tumours (glioma, lymphoma, meningioma, metastasis), demyelinating diseases, trauma and congenital malformations can involve the corpus callosum. However, due to its rich collateral blood supply, isolated vascular lesions are extremely uncommon1. Here we report the case of a 50 year old woman who presented with features of corpus callosum apraxia, initially mistaken as a psychiatric symptom.\n\n\nCase report\n\nA 50 year old woman presented with a history of abrupt onset of numbness on the left side of her body. She had no history suggestive of raised intracranial tension, cranial nerve dysfunction or motor involvement. The patient was admitted with a provisional diagnosis of sensory stroke considering the rapidity of onset. On the second day of hospitalisation, she found it hard to transfer objects from her left hand to the right hand, and found it hard to execute finer activities with left hand. Her family mistook these manifestations as psychiatric symptoms. The patient had a past medical history of systemic hypertension for 12 years and was on tab. amlodipine 5 mg once daily.\n\nOn examination on the second day of admission, the patient was conscious, oriented to time, place and person. She had a body mass index of 22.1 kg/m2 and vital signs were stable. Tests for attention, registration, recall were normal. Her immediate and remote memory was intact. Cranial nerves were normal. She had grade 5 power in all limbs. Sensory system examination revealed apraxia of the left hand, which was diagnosed as ideomotor apraxia: the patient was unable to perform pantomime commands with her left hand and also had astereognosis of the left hand. The patient did not have any involuntary movements in the same limb and could perform all activities with her right hand, but was unable to do the same with the left. There were no cerebellar signs nor signs of meningeal irritation. These features were suggestive of callosal disconnection syndrome.\n\nComputed tomography of brain revealed an acute infarct in the body of the corpus callosum (Figure 1). Magnetic resonance imaging (MRI) of the brain showed hyperintense signals in the body of corpus callosum in T2 weighted FLAIR images (Figure 2). Diffusion weighted images showed diffuse restriction confirming the presence of an acute infarct. Electroencephalogram was normal. Further workup was done to look for the cause of thrombosis. Echocardiography was normal. Arterial Doppler of the cerebral vessels did not reveal any abnormality. The patient’s complete blood count, erythrocyte sedimentation rate and peripheral smear were normal. Antinuclear antibody test and retroviral screening were negative.\n\nThe patient was treated with antiplatelets (tab. Ecosprin 150 mg once daily and tab. Clopidogrel 75 mg once daily), statins (tab. Atorvastatin 40 mg once daily), and antihypertensives (tab. Amlodipine 5mg once daily). She showed mild clinical improvement at 4 months of follow up (MRI showed resolution of edema and chronic infarct), although apraxia of the left hand persisted.\n\n\nDiscussion\n\nThe corpus callosum contains the white matter bundle that connects the interhemispheric neurons and serves to connect the cortical and subcortical structures of the brain. The corpus callosum is present only in placental mammals and reflects the transition towards increased degree of myelination seen in higher animals. Anatomically it is divided into the rostrum, genu, body and splenium from anterior to posterior2. The corpus callosum receives a blood supply from three arterial networks. The anterior communicating artery (medical callosal and subcallosal branches), anterior cerebral artery (through the pericallosal artery), posterior cerebral artery (through the posterior pericallosal branch); each supplies different parts of the corpus callosum3.\n\nClinical manifestations of callosal infarcts include callosal disconnection syndrome (as in our case), neuropsychiatric symptoms, gait disorders and alien hand syndrome4. Callosal disconnection manifests in different ways based on the part of the callosum affected. Lesions in the anterior part of callosum presents as unilateral unresponsiveness to touch (tactile anomia), difficulty in calculation, difficulty in copying drawings, and unilateral agraphia (inability to write with one hand). Posterior lesions cause visual unresponsiveness on one side (visual anomia) and a phenomenon called “alexia without agraphia” (able to write, but unable to read). Lesions in the body of the callosum results in unilateral ideomotor apraxia, difficulty to use objects with one hand, and inability to transfer objects from one hand to the other. This occurs in the left hand in majority of the cases, as the left is the dominant hemisphere in most individuals5. MRI is the diagnostic test of choice for the localisation of a callosal infarct. Differential diagnoses for callosal infarcts include Marchiafava Bignami syndrome (callosal demyelination in chronic alcoholism) and multiple sclerosis6,7.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "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\nMurthy SB, Chmayssani M, Shah S, et al.: Clinical and radiologic spectrum of corpus callosum infarctions: Clues to the etiology. J Clin Neurosci. 2013; 20(1): 175–177. PubMed Abstract | Publisher Full Text\n\nTüre U, Yaşargil MG, Krisht AF: The arteries of the corpus callosum: a microsurgical anatomic study. Neurosurgery. 1996; 39(6): 1075–85; discussion 1084-5. PubMed Abstract | Publisher Full Text\n\nSaito Y, Matsumura K, Shimizu T: Anterograde amnesia associated with infarction of the anterior fornix and genu of the Corpus Callosum. J Stroke Cerebrovasc Dis. 2006; 15(4): 176–7. PubMed Abstract | Publisher Full Text\n\nKasow DL, Destian S, Braun C, et al.: Corpus callosum infarcts with atypical clinical and radiologic presentations. AJNR Am J Neuroradiol. 2000; 21(10): 1876–1880. PubMed Abstract\n\nYuan JL, Wang SK, Guo XJ, et al.: Acute infarct of the corpus callosum presenting as alien hand syndrome: evidence of diffusion weighted imaging and magnetic resonance angiography. BMC Neurol. 2011; 9: 142. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuwanwela NC, Leelacheavasit N: Isolated corpus callosal infarction secondary to pericallosal artery disease presenting as alien hand syndrome. J Neurol Neurosurg Psychiatry. 2002; 72(4): 533–536. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBourekas EC, Varakis K, Bruns D, et al.: Lesions of the corpus callosum: MR imaging and differential considerations in adults and children. AJR Am J Roentgenol. 2002; 179(1): 251–257. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "42656",
"date": "14 Jan 2019",
"name": "Radek Ptak",
"expertise": [
"Reviewer Expertise Clinical neuropsychology"
],
"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 focuses on a specific clinical sign of disconnection following corpus callosum damage, unilateral apraxia. The authors point out that focal damage to the corpus callosum is rare and may present a diagnostic challenge. I only partly agree, as damage to the corpus callosum is at least occasionally observed following ischemic stroke to one of the pericallosal arteries, though in this case the damage is often not confined to the corpus callosum, but also involves medial frontal cortex and cingulate cortex. I also think that such injury presents only a diagnostic challenge when the examiner does not specifically focus on symptoms of callosal disconnection. The body and genu of the corpus callosum carry fibers connecting pre-motor and parietal cortices of the two cerebral hemispheres which are necessary for the coordination of bilateral actions. When reading this case report I was surprised to notice that the authors observed apraxia of the left hand, but did not examine callosal transfer by asking the patient to execute bilateral actions. Besides alien hand syndrome, corpus callosum damage is known to result in diagonistic dyspraxia. This is a movement disorder of the left arm/hand characterized by movements that are antagonistic or symmetrical to the right hand, as well as the occasional inability to move the left arm (though force is typically preserved). Similarly to alien hand syndrome, this disorder suggests an impairment of voluntary control, and some patients report that their hand feels disobedient or has 'its own personality'. Such impairments appear sufficiently mysterious to be mistaken as psychiatric symptoms. The current case is interesting because of the isolated corpus callosum damage, but one would have wished a thorough examination of disconnection signs other than unilateral apraxia.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-731
|
https://f1000research.com/articles/7-274/v1
|
05 Mar 18
|
{
"type": "Review",
"title": "Role of extracellular matrix in breast cancer development: a brief update",
"authors": [
"Manoj Kumar Jena",
"Jagadeesh Janjanam"
],
"abstract": "Evidence is increasing on the crucial role of the extracellular matrix (ECM) in breast cancer progression, invasion and metastasis with almost all mortality cases owing to metastasis. The epithelial-mesenchymal transition is the first signal of metastasis involving different transcription factors such as Snail, TWIST, and ZEB1. ECM remodeling is a major event promoting cancer invasion and metastasis; where matrix metalloproteinases (MMPs) such as MMP-2, -9, -11, and -14 play vital roles degrading the matrix proteins for cancer spread. The β-D mannuronic acid (MMP inhibitor) has anti-metastatic properties through inhibition of MMP-2, and -9 and could be a potential therapeutic agent. Besides the MMPs, the enzymes such as LOXL2, LOXL4, procollagen lysyl hydroxylase-2, and heparanase also regulate breast cancer progression. The important ECM proteins like integrins (b1-, b5-, and b6- integrins), ECM1 protein, and Hic-5 protein are also actively involved in breast cancer development. The stromal cells such as tumor-associated macrophages (TAMs), cancer-associated fibroblasts (CAFs), and adipocytes also contribute in tumor development through different processes. The TAMs become proangiogenic through secretion of VEGF-A and building vessel network for nourishment and invasion of the tumor mass. The latest developments of ECM involvement in breast cancer progression has been discussed in this review and this study will help researchers in designing future work on breast cancer pathogenesis and developing therapy targeted to the ECM components.",
"keywords": [
"Extracellular matrix",
"breast cancer",
"metastasis",
"matrix metalloproteinases"
],
"content": "Introduction\n\nBreast cancer (BC) accounts for 25% of all cancer cases in women, and 12% of overall cancer cases worldwide1. The extracellular matrix (ECM) plays a crucial role in breast cancer (BC) progression, invasion, and metastasis; thus, elucidating the role of ECM will help in designing therapies targeting different ECM components. At present, mortality in any form of cancer accounts for 98% due to metastasis. Comprehensive studies are currently going on related to the involvement of ECM in BC progression, and this review focuses on the latest developments in this regard with possible molecular targets for therapies.\n\nThe EMT (process of losing epithelial characteristics and gaining mesenchymal properties) is the beginning step of metastasis. About 90% deaths from BC are due to invasion and metastasis, and EMT plays a significant role involving different transcription factors (TFs) and signals2–5. It induces metastasis through ECM disruption and metabolism reprogramming. Aberrant cancer metabolism promotes EMT which further aggravates metabolism (especially glucose metabolism) through EMT-specific TFs such as Snail and TWIST6. Platelets and platelet-derived TGF-β promote epithelial-mesenchymal-like transition and promote metastasis in vivo7. Snail is a transcriptional repressor of E-cadherin (cell-cell adhesion molecule), and E-cadherin loss is a hallmark of EMT2. Snail and TWIST cooperate inducing another TF, ZEB18 (significant inducer of EMT, invasion, and metastasis), which is triggered by extracellular hyaluronic acid (HA). Furthermore, ZEB1 induces HAS2 synthesis, promoting HA production in a positive feedback loop and its expression is correlated with ZEB1 expression in poor prognosis tumors. HAS2 also has a role in TGF-β-induced EMT9.\n\nVarious ECM-remodeling enzymes are induced in BC promoting stem/progenitor signaling pathways and metastasis. Major ECM proteins induced are fibrillar collagens, fibronectin, specific laminins, proteoglycans, and matricellular proteins and these could be potential drug targets for therapy10. Matrix metalloproteinases (MMPs) degrade ECM proteins promoting invasion and metastasis. The MMP-11 (stromelysin-3) seems facilitating tumor development through apoptosis inhibition. However, it suppresses metastasis in animal models, exhibiting different roles in tumor progression11. β-D mannuronic acid (BDM) is a MMP inhibitor, inhibiting MMP-2 and MMP-9 involved in invasion, metastasis, and angiogenesis12. BDM possesses anti-metastatic activity and inhibits tumor growth by suppressing inflammatory chemokine and tumor–promoting cytokines13. MMP-14 located on the cell surface, is a potential target to stop metastasis and a novel antibody-mediated MMP-14 blockade seems to limit hypoxia and metastasis in triple negative breast cancer (TNBC) models14. Loss of ECM integrity by plasmin facilitates cancer cell spread15,16 and plasmin-induced ECM degradation may be controlled by lipoprotein-A (competitive inhibitor of plasminogen)17. Vitamin C seems to be very important curbing tumor growth, and metastasis as ECM integrity requires vitamin C17. The Lox (Lysil oxidase) family of genes enhances ECM fibrosis through collagen cross-linking and it seems down-regulation of LOXL4 promotes BC growth and lung metastasis in mice18. The LOXL2 protein catalyzes cross-linking of ECM components collagen and elastin and is involved in cancer progression and metastasis. The intracellular LOXL2 shows EMT induction and Snail-1 stabilization, and LOXL-2/Snail-1-mediated E-cadherin down-regulation promotes lung metastasis of BC without affecting ECM stiffness19.\n\nThe enzyme procollagen lysyl hydroxylase-2 required for collagen synthesis, increases breast tumor stiffness, promotes metastatic tumors in lymph nodes and lungs. Matrix stiffness promotes tumor progression and invasion of ER+ type BC20. The hardened ECM drives invasion and metastasis through ERK1/2 signal up-regulation and JAK2/STAT5 signal down-regulation. The enzyme heparanase cleaves heparan sulfate, promoting tumor invasion and metastasis. ER stress during chemotherapy enhances the heparanase activity21. The MMTV-heparanase mice promoted growth and metastasis of breast tumor cells to lungs suggesting a role for heparanase in BC progression22. Elemene (extract of Curcuma erhizoma plant), is an anticarcinogenic phytochemical showing effects by down-regulating heparanase expression (potential target for heparanase)23. The heparin and nanoheparin derivatives show their anti-cancer activities by reducing BC cell proliferation and metastasis24.\n\nTumor cells recruit tumor-associated macrophages (TAMs), which become proangiogenic by secreting VEGF-A which nourishes tumor cells and build a vessel network for their invasion. Hypoxia also induces macrophages to produce more VEGF and suppress immune response, promoting invasion25. Cancer-associated fibroblasts (CAFs) are involved in tumor development, progression, inflammation, metastasis, and build resistance to cancer therapy through secretion of hormones, cytokines, growth factors, etc. and cross-talk with other stromal cells, cancer cells, and ECM. CAFs can be potential therapeutic targets in BC26. Cancer cell proliferation and migration is induced by activated fibroblasts derived from endothelial-to-mesenchymal transformation27. Adipocytes have a significant role in cancer progression, ECM remodeling, phenotype changes of CAFs, and resistance to cancer therapy28.\n\nIntegrins, the primary receptors of MECs for ECM, act as sensors of epithelial microenvironment. Their altered expression seems to disorganize ECM and promotes metastasis29. Increased MEC proliferation occurs due to enhanced activity of integrin signaling (b1-, b5-, and b6- integrins) by co-activating the oncogenes for enhanced growth factor signaling. Protein ECM1 is involved in angiogenesis, promoting TNBC migration and invasion30. Protein Hic-5 (focal adhesion scaffold/adaptor protein) promotes mammary duct formation. Focal adhesions of cells are attached to ECM and transduce signals from ECM to cell. Hic-5 is up-regulated in CAFs of BC, involved in EMT and invadopodia formation facilitating invasion, migration and metastasis31. The sustained directionality of tumor cells to a vessel is promoted by a chemotactic gradient of hepatocyte growth factor (HGF) produced from vessel endothelium. This directional streaming is possible by HGF/c-Met signaling pathway between endothelial cells and tumor cells; and c-Met inhibitors could be a potential target to block tumor cell streaming and metastasis32.\n\n\nConclusion\n\nThe ECM constitutes a complex of structural proteins and its reorganization is essential during cancer progression. ECM proteins provide biochemical signals to induce EMT and initiate metastasis, progression of cancer to advanced stage. ECM remodeling enzymes like MMPs play an essential role in these processes. The tumor microenvironment, platelet-derived mitogens and chemokines, granulocytes and stromal cells help cancer cells achieve intravascular transit and metastasis to target site. In addition, various ECM proteins such as integrins, collagen and fibronectin engage in cell adhesion, invasion and metastasis. All these elements of the ECM are critical for cancer progression and hence targeting ECM is a prospective approach for targeted drug discovery and cancer therapy.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nFerlay J, Soerjomataram I, Dikshit R, et al.: Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015; 136(5): E359–86. PubMed Abstract | Publisher Full Text\n\nWang Y, Zhou BP: Epithelial-mesenchymal transition in breast cancer progression and metastasis. Chin J Cancer. 2011; 30(9): 603–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Craene B, Berx G: Regulatory networks defining EMT during cancer initiation and progression. Nat Rev Cancer. 2013; 13(2): 97–110. PubMed Abstract | Publisher Full Text\n\nPeinado H, Olmeda D, Cano A: Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer. 2007; 7(6): 415–428. PubMed Abstract | Publisher Full Text\n\nPuisieux A, Brabletz T, Caramel J: Oncogenic roles of EMT-inducing transcription factors. Nat Cell Biol. 2014; 16(6): 488–494. PubMed Abstract | Publisher Full Text\n\nHuang R, Zong X: Aberrant cancer metabolism in epithelial-mesenchymal transition and cancer metastasis: Mechanisms in cancer progression. Crit Rev Oncol Hematol. 2017; 115: 13–22. PubMed Abstract | Publisher Full Text\n\nLabelle M, Begum S, Hynes RO: Direct signaling between platelets and cancer cells induces an epithelial-mesenchymal-like transition and promotes metastasis. Cancer Cell. 2011; 20(5): 576–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDave N, Guaita-Esteruelas S, Gutarra S, et al.: Functional cooperation between Snail and twist in the regulation of ZEB1 expression during epithelial to mesenchymal transition. J Biol Chem. 2011; 286(14): 12024–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPreca BT, Bajdak K, Mock K, et al.: A novel ZEB1/HAS2 positive feedback loop promotes EMT in breast cancer. Oncotarget. 2017; 8(7): 11530–11543. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInsua-Rodríguez J, Oskarsson T: The extracellular matrix in breast cancer. Adv Drug Deliv Rev. 2016; 97: 41–55. PubMed Abstract | Publisher Full Text\n\nZhang X, Huang S, Guo J, et al.: Insights into the distinct roles of MMP-11 in tumour biology and future therapeutics (Review). Int J Oncol. 2016; 48(5): 1783–93. PubMed Abstract | Publisher Full Text\n\nMirshafiey A, Khorramizadeh, MR, Saadat F, et al.: Chemopreventive effect of M2000, a new anti-inflammatory agent. Med Sci Monit. 2004; 10(10): PI105–PI109. PubMed Abstract\n\nHosseini F, Hassannia H, Mahdian-Shakib A, et al.: Targeting of crosstalk between tumor and tumor microenvironment by β-D mannuronic acid (M2000) in murine breast cancer model. Cancer Med. 2017; 6(3): 640–650. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLing B, Watt K, Banerjee S, et al.: A novel immunotherapy targeting MMP-14 limits hypoxia, immune suppression and metastasis in triple-negative breast cancer models. Oncotarget. 2017; 8(35): 58372–58385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRath M, Pauling L: Plasmin-induced proteolysis and the role of apoprotein(a), lysine and synthetic analogs. Orthomolecular Med. 1992; 7: 17–23. Reference Source\n\nChoong PF, Nadesapillai AP: Urokinase plasminogen activator system: A multifunctional role in tumor progression and metastasis. Clin Orthop Relat Res. 2003; (415 Suppl): S46–S58. PubMed Abstract\n\nCha J, Roomi MW, Kalinovsky T, et al.: Lipoprotein(a) and vitamin C impair development of breast cancer tumors in Lp(a)+; Gulo-/- mice. Int J Oncol. 2016; 49(3): 895–902. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChoi SK, Kim HS, Jin T, et al.: LOXL4 knockdown enhances tumor growth and lung metastasis through collagen-dependent extracellular matrix changes in triple-negative breast cancer. Oncotarget. 2017; 8(7): 11977–11989. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalvador F, Martin A, López-Menéndez C, et al.: Lysyl Oxidase-like Protein LOXL2 Promotes Lung Metastasis of Breast Cancer. Cancer Res. 2017; 77(21): 5846–5859. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarcus CE, O'Leary KA, Brockman JL, et al.: Elevated collagen-I augments tumor progressive signals, intravasation and metastasis of prolactin-induced estrogen receptor alpha positive mammary tumor cells. Breast Cancer Res. 2017; 19(1): 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Y, Liu H, Huang YY, et al.: Suppression of endoplasmic reticulum stress-induced invasion and migration of breast cancer cells through the downregulation of heparanase. Int J Mol Med. 2013; 31(5): 1234–42. PubMed Abstract | Publisher Full Text\n\nBoyango I, Barash U, Fux L, et al.: Targeting heparanase to the mammary epithelium enhances mammary gland development and promotes tumor growth and metastasis. Matrix Biol. 2018; 65: 91–103. PubMed Abstract | Publisher Full Text\n\nZhang Y, Sun X, Nan N, et al.: Elemene inhibits the migration and invasion of 4T1 murine breast cancer cells via heparanase. Mol Med Rep. 2017; 16(1): 794–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAfratis NA, Karamanou K, Piperigkou Z, et al.: The role of heparins and nano-heparins as therapeutic tool in breast cancer. Glycoconj J. 2017; 34(3): 299–307. PubMed Abstract | Publisher Full Text\n\nObeid E, Nanda R, Fu YX, et al.: The role of tumor-associated macrophages in breast cancer progression (review). Int J Oncol. 2013; 43(1): 5–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJung YY, Kim HM, Koo JS: The role of cancer-associated fibroblasts in breast cancer pathobiology. Histol Histopathol. 2016; 31(4): 371–8. PubMed Abstract | Publisher Full Text\n\nMina SG, Huang P, Murray BT, et al.: The role of shear stress and altered tissue properties on endothelial to mesenchymal transformation and tumor-endothelial cell interaction. Biomicrofluidics. 2017; 11(4): 044104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChoi J, Cha YJ, Koo JS: Adipocyte biology in breast cancer: From silent bystander to active facilitator. Prog Lipid Res. 2018; 69: 11–20. PubMed Abstract | Publisher Full Text\n\nGlukhova MA, Streuli CH: How integrins control breast biology. Curr Opin Cell Biol. 2013; 25(5): 633–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGómez-Contreras P, Ramiro-Díaz JM, Sierra A, et al.: Extracellular matrix 1 (ECM1) regulates the actin cytoskeletal architecture of aggressive breast cancer cells in part via S100A4 and Rho-family GTPases. Clin Exp Metastasis. 2017; 34(1): 37–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoreczny GJ, Ouderkirk-Pecone JL, Olson EC, et al.: Hic-5 remodeling of the stromal matrix promotes breast tumor progression. Oncogene. 2017; 36(19): 2693–2703. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeung E, Xue A, Wang Y, et al.: Blood vessel endothelium-directed tumor cell streaming in breast tumors requires the HGF/C-Met signaling pathway. Oncogene. 2017; 36(19): 2680–2692. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "31504",
"date": "09 Apr 2018",
"name": "Andrew R. Craig",
"expertise": [
"Reviewer Expertise Cancer metastasis",
"tumor microenvironment",
"matrix metalloproteinases",
"immunotherapy"
],
"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 review by Jena and Janjanam provides some updates on progress in our understanding of the complex interactions between breast cancer cells and surrounding extracellular matrix (ECM) during tumor progression and metastasis. The review also mentions the growing appreciation of the interplay of stromal cells such as tumor-associated macrophages and fibroblasts on this process. While I applaud the efforts to synthesize some recent findings in this review, each subtopic within the review would warrant a deeper dive into the literature and what this means for developing new therapies. The recent discovery of the interstitium as a fluid compartment, as opposed to a rigid ECM barrier, that facilitates metastasis within lymph nodes and blood vessels1 would warrant mention in a future review on this topic. There is also evidence emerging that ECM can form a barrier to cytotoxic immune cell recruitment. What might this mean for therapies targeting matrix metalloproteinases or cancer-associated fibroblasts that alter the ECM? How can the altered ECM improve responses to chemotherapy or immunotherapy in breast cancer?\n\nOverall, the topic is very interesting and timely, but the review needs to be more focused to achieve significant depth, and to truly guide thinking about new therapeutic strategies for metastatic breast cancer.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? No\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly",
"responses": [
{
"c_id": "3715",
"date": "12 Jun 2018",
"name": "Manoj Jena",
"role": "Author Response",
"response": "All the queries have been answered in this version."
}
]
},
{
"id": "32907",
"date": "09 Apr 2018",
"name": "Yunus A. Luqmani",
"expertise": [
"Reviewer Expertise Molecular biology",
"endocrinology of breast cancer"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very brief overview of the various complex processes that are thought to influence the migration of breast cancer cells from their primary site into and through the extracellular matrix, a pre-requisite for metastatic dissemination through the vascular system. The importance of the ECM in this process and its potential as a therapeutic target has only relatively recently been recognized, but the extent of interest has grown rapidly. Although this report does not add much novelty to the existing literature it is nevertheless a useful succinct summary of some of the major influencing factors/processes that are involved in breast cancer progression that inter-relates to the ECM. It is well written and easy to follow, particularly for those less familiar with this field.\nThe title refers specifically to the role of the ECM. But EMT is a process occurring within the cancer cells prior to interaction with the ECM so it’s description is rather out of place here. Perhaps instead the authors may speculate as to how alterations or signals originating from the ECM might initiate EMT in some cells of the tumour mass (platelet derived-TGFβ is one factor mentioned but this is also secreted from cancer cells). This would add something new as little is known about the triggers (as opposed to the pleithora of mediators) of EMT. A short description of the differences in the tumour ECM from that around normal cells would be also be useful.\nMinor points:\nStatements that 98% of any cancer mortality and 90% of BC is due to metastasis should be removed or compelling evidence cited to support these figures. Whilst it may be true for BC, primary tumours in vital organs such as brain, lung, liver etc are likely to be as responsible for mortality, if not more, than their metastases.\n\nA few grammatical corrections: lysyl and builds and MEC abbreviation to be explained.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "3714",
"date": "12 Jun 2018",
"name": "Manoj Jena",
"role": "Author Response",
"response": "All the queries have been answered in this version."
}
]
},
{
"id": "32720",
"date": "09 May 2018",
"name": "Ren Xu",
"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\nECM is the major component of tumor microenvironment. The review briefly summarized function of ECM in breast cancer progression. Authors discussed function of MMPs and LOXs in ECM remodelling and cancer progression. The review also touched the roles of EMT and CAF in ECM remodelling. However, none of these aspects has been covered thoroughly and discussed in detail. The review provides some information about ECM and cancer progression, but lacks big picture of ECM function and detail information about ECM deposition and remodelling.\nMajor comments:\nECM degradation is only part of ECM remodelling during cancer development. Increased ECM deposition is also associated with cancer progression. The authors need provide more and detail information about what and how ECM deposition contributes/represses cancer progression.\n\nFunction of ECM in cancer progression is very complicated. Some ECM proteins promote cancer progression, some may suppress cancer progression. It is crucial to classify it in the review.\n\nEMT and CAF are both involved in ECM remodelling. Detail information about how EMT and CAF regulate ECM remodelling need to be provided. It is also important to discuss the function of other stromal cells in ECM remodelling in cancer tissue.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly",
"responses": [
{
"c_id": "3713",
"date": "12 Jun 2018",
"name": "Manoj Jena",
"role": "Author Response",
"response": "All the queries have been answered in this version."
}
]
}
] | 1
|
https://f1000research.com/articles/7-274
|
https://f1000research.com/articles/7-729/v1
|
12 Jun 18
|
{
"type": "Research Article",
"title": "Regenerative collagen biomembrane: Interim results of a Phase I veterinary clinical trial for skin repair",
"authors": [
"Andreas Kaasi",
"João F. Lima-Neto",
"José A. Matiello-Filho",
"Mário H.S. Calejo",
"André L. Jardini",
"Paulo Kharmandayan",
"João F. Lima-Neto",
"José A. Matiello-Filho",
"Mário H.S. Calejo",
"André L. Jardini",
"Paulo Kharmandayan"
],
"abstract": "Background: The availability of commercial tissue engineering skin repair products for veterinary use is scarce or non-existent. To assess features of novel veterinary tissue engineered medical devices, it is therefore reasonable to compare with currently available human devices. During the development and regulatory approval phases, human medical devices that may have been identified as comparable to a novel veterinary device, may serve as predicate devices and accelerate approval in the veterinary domain. The purpose of the study was to evaluate safety and efficacy of the biomembrane for use in skin repair indications. Methods: In the study as a whole (3 year total length), 15 patients (animals), dogs and cats (male/female, <8 years) with skin lesions of different etiologies considered difficult to heal (size, >2 cm), with a wound depth equivalent to 2nd/3rd degree burns are to be studied from Day 0 to Day 120-240, post-application of the biomembrane. This interim report covers the 5 patients assessed to date and deemed eligible, of which 3 enrolled, and 2 have completed the treatment. Wound beds were prepared and acellular collagen biomembranes (Eva Scientific Ltd, São Paulo, Brazil) applied directly onto the wounds, and sutured at the margins to the patient's adjacent tissue. Wound size over time, healing rate, general skin quality and suppleness were assessed as outcomes. Qualitative (appearance and palpation) and quantitative (based on Image Analysis of photographs) wound assessment techniques were used. Results: Both patients’ wounds healed fully, with no adverse effects, and the healing rate was comparable in both, maxing out at approximately 1 cm2/day. Conclusions: Early results on the biomembrane's safety and efficacy indicate suitability for skin repair usage in veterinary patients.",
"keywords": [
"biomembrane",
"biofabrication",
"tissue engineering",
"medical device",
"collagen",
"wound healing"
],
"content": "Introduction\n\nEver since the pioneering work of Rheinwald & Green (1975), on skin epidermis (keratinocyte) cultures, Yannas & Burke (1980) on hybrid (biologic/synthetic) dermis/epidermis skin substitutes, and Bell et al. (1981) on fully biologic dermis and epidermis skin substitutes, a number of companies and products engaged with tissue engineering technologies have developed and consolidated in the market. In particular, early clinical research (Phase 1 and Early Phase 1) has tended to be characterized by a tight relationship between the industry-academia-hospital stakeholders of this archetypical medical device triad.\n\nThe tissue engineered devices that have made the most headway, clinically and commercially, include Epicel®, Integra®, Dermagraft®, Apligraf® and Alloderm®. These products have all had to make decisions and compromises whether to mimic only the dermis, only the epidermis, or both. Whether to employ only synthetic biomaterials, only natural biomaterials, or both. And whether to include only living dermis-type (fibroblast) cells, only epidermis-type (keratinocyte) cells, or both, or alternatively, to let the device be acellular. Review articles written by Eisenbud et al. (2004) and Groeber et al. (2011) provide a more in-depth comparison of these products and others.\n\nTissue engineered devices for veterinary use are non-existent, or scarce, at best. The same can be said about veterinary clinical trials in general, which lack a global consensus concerning the rules and regulations governing them. Usually, veterinary clinical trials with patients having spontaneously occurring diseases are treated analogously to animal experimentation with experimental subjects with induced diseases (Fürdös et al., 2015), requiring Ethics Committee approval for the use of Animals for Experimentation (the same way that this is required for “traditional” animal experimentation). By contrast, countries may not regard animal clinical trials as animal experiments, thus not requiring an Ethics Committee approval by law. This would present a less restrictive approach. A third possibility would be to implement a set of rules and regulations specific for veterinary clinical trials, with a separate Ethics Committee for this purpose. This would present a more restrictive approach.\n\nMany researchers debating on the topic of veterinary clinical trials (Fürdös et al., 2015; Vail, 2007) acknowledge the attractiveness of letting veterinary clinical trials pave the way for human use of the same therapy. To make the cross-applicability of veterinary/human data as smooth as possible, and while more specific veterinary clinical trial rules and regulations remain absent or non-enforced, it may be advantageous to frame veterinary clinical trials to conform with rules, regulations and standards governing human clinical trials. One aspect of interest is the sample size and terminology of the “phase” of the trial.\n\nIn the human domain, the U.S. Food and Drug Administration (FDA) defines the number of participants appropriate for each phase as follows: Phase 1, 20–100 participants; Phase 2, “several hundred” participants; Phase 3, 300–3000 participants; Phase 4, “several thousand” participants (https://www.fda.gov/ForPatients/Approvals/Drugs/ucm405622.htm, retrieved on 12 Jan 2018).\n\nThe U.S. National Institutes of Health (NIH) has a similar definition for Phases 1 (n= 20–80) and 3 (n=100–3000), but offers a precise definition for Phase 2, at 100–1000 participants. The NIH, unlike the FDA, does not provide any specific participant number for Phase 4 (https://www.nih.gov/health-information/nih-clinical-research-trials-you/basics, retrieved on 12 Jan 2018).\n\nA so-called “Early Phase 1”, previously known as Phase 0, is defined as an “exploratory study involving very limited human exposure to the drug, with no therapeutic or diagnostic goals” (https://clinicaltrials.gov/ct2/help/glossary/phase, retrieved on 12 Jan 2018). Again, the descriptor “very limited human exposure” does not offer precise guidance on numbers, but a figure of <15 participants appears to be the consensus for this type of study (Gawai et al., 2017).\n\nWe believe a sample size of 15, being the borderline number between the above mentioned “Early Phase 1” and “Phase 1” (as defined by FDA and NIH, above) is a suitable sample size for veterinary clinical trials, being an economically and practically feasible number to execute in a veterinary medicine context (usually not backed up by heavy industry or institutional funding), whilst at the same time offering the desired property of translationability of veterinary clinical trial findings to human applications.\n\nIn this study, we set out to treat skin lesions in dog and cat patients referred to our clinical practice, where conventional clinical treatment had proved itself incapable of satisfactorily healing the lesions. In order to form suitable exclusion criteria with respect to wound size, we considered the following reference parameter: maximum lesion/defect size for a spontaneous and unconfined tissue regeneration to occur in vivo, without special treatment = 1 cm (Dorin et al., 2008). The same value is also given in the literature in a burn context, when faced with a full thickness, third degree burn (Papini, 2004). Thus, we chose to concentrate on wounds larger than this reference size.\n\nIn assessing the wounds and progression of healing, for qualitative data, we ordered that photographs be taken as often as possible, and, inasmuch possible, using certain reference angles of photography that were easy for the patient owners to employ when sending photographs from home. These same photographs formed the basis for a quantitative wound healing assessment, using the photographs as raw data for an image analysis powered calculation of wound area/healed area over time. The approach chosen required minimal or no investment, as opposed to commercial wound image scanner devices (see e.g., Bills et al., 2016; Davis et al., 2013) that can cost upwards of US$10,000. In addition to the experiences sought through this clinical trial, pertaining to the clinical outcome of the study, the experiences obtained with this image analysis method were seen as secondary benefits of the study that also would be useful to perfect the methodological aspects of the study for the remainder of the study period.\n\n\nMethods\n\nThis study has been reported using CONSORT guidelines (see Supplementary File 1 for a completed checklist).\n\nThe protocol to this ongoing, 36-month, phase I, quasi-experimental, interventional clinical trial is available as Supplementary File 2.1\n\nResults in this interim report include data from animals who were assessed and eligible (n=5) and enrolled (n=3) in the trial, with treatment completed (n=2), at the time of writing (Figure 1). Therefore this article presents interim results, spanning 1 year of the 3 year study length.\n\nThe study protocol abided by the principles of the WMA Declaration Of Helsinki – Ethical Principles For Medical Research Involving Human Subjects (https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/), adapted to veterinary practice.\n\nThe Ethics Committee of Sorocaba Veterinary Hospital approved the study protocol before the study began (Ref. no. 20161012-1), and all patient (animal) owners provided written informed consent for their pets’ participation before undergoing any procedure related to the present study. The study was carried out at Sorocaba Veterinary Hospital.\n\nThe objective of the present study was to obtain clinical experiences with the use of a proprietary collagen biomembrane medical device for skin repair in dogs and cats, and to assess its safety and efficacy for this indication.\n\nDogs and cats, of both sexes, aged up to 8 years, with an etiology of skin lesion of different causes, such as burns, traumas, surgical excisions, were deemed eligible for treatment with the collagen biomembrane. The lesions should be equivalent to second or third degree burns, i.e., extending into the dermis or subcutaneous layers of the skin, or loss of skin tissue altogether. The size of the lesion should measure 2 cm or more (length/width/diameter as per the lesion geometry). This figure is double the reference maximum lesion/defect size (1 cm) for spontaneous and unconfined tissue regeneration in vivo to occur, without special treatment, as noted above. The rationale of setting the size criterion at double that figure, considered the maximum size for spontaneous healing (1 cm), was that the healing of a lesion/defect size of this magnitude (2 cm) would be unquestionably attributed to the treatment regime. Key exclusion criteria included lesions only of the epidermis (equivalent to a first degree burn), in addition to skin lesions stemming from chronic disease. These criteria are summarized in Table 1.\n\nEach patient was photographed on the operating table, just prior to the first surgical procedures, from a variety of angles. Inasmuch possible, a decision on the most suitable angle and focal distance for future photographs, would take place during these initial photographs. Ideally, a single angle and focal distance would provide for a general wound healing assessment, as supposed to requiring multiple angles, especially since later follow-up time points depend on owners’ own photographs, with little or no control by study researchers. A ruler would be placed adjacent to the treatment area for later dimensional analyses, or, alternatively, direct measurements would be taken of reference distances such as eye-midline, knee-hip joint, and recorded on file.\n\nUnder general anesthesia (preanesthetic agents: acepromazine 0.02 mg/kg and meperidine 3 mg/kg, both administered intramuscularly; induction agent: propofol, 3 mg/kg, administered intravenously; maintenance agent in cats: propofol, 0.2 mg/kg/min; maintenance agent in dogs: oral isoflurane administered through tracheal intubation with a universal anesthetic vaporizer, dose being subject to anesthetist’s clinical assessment and varied as required during the surgery), the treatment area was prepared by shaving away fur adjacent to the treatment area and cleaned with saline and iodine. If the affected area had been subject to natural scarring, with the formation of hard, unhealthy, fibrotic scar tissue, this was surgically removed, exposing subcutaneous tissue. In this case, the tissue would be fixed in 10% formalin and submitted to histological analysis, using standard procedures. If the treatment area was an unhealed, deepithelialized wound bed, but with no fibrotic scar tissue formation, then shallow scalpel incisions were made scattered throughout the treatment area, causing local bleeding, with the aim of promoting regeneration and integration with the later application of the collagen biomembranes. In any case, debridement of lacerated, devitalized or contaminated tissue was performed prior to the application of the collagen biomembranes.\n\nAcellular collagen biomembranes measuring 8 × 3 cm were obtained from Eva Scientific Ltd, São Paulo, SP, Brazil (http://www.evascientific.com/products/biomembranespet/). According to the manufacturer, the biomembranes are made aseptically from collagen type I sourced from rat tail tendon (Ref. E-003RT050M1), using the manufacturer’s proprietary biofabrication apparatus. The resulting biomembranes possess a mean thickness of 200 to 300 μm, with a regular pattern of local depressions/elevations of approximately 100 μm and 2.5 mm apart center-to-center (see Figure 2).\n\nThe biomembrane measures 8×3 cm, and is scattered with small ridges throughout the structure, measuring approximately 100 μm in height, and a spacing of 0.25 cm between each elevated ridge. The standard presentation of the biomembrane is with a synthetic elastomeric membrane underlying the biomembrane itself; when applied to the treatment area, the device (double-layer synthetic membrane / biomembrane structure) is placed biomembrane-side down onto the treatment area, and the synthetic membrane carefully peeled away with a forceps, leaving behind the biomembrane (see Supplementary File 3 for footage of this application technique).\n\nDepending on the size and shape of the treatment area, one or several collagen biomembranes were used to cover the area, extending over the wound margin by a few millimeters. At the margins, the biomembranes were sutured to non-treatment tissue using the simple interrupted suture technique. If one or more biomembrane edges were facing away from the non-treatment tissue margin (i.e., into the wound, kissing or overlapping another biomembrane’s non-treatment tissue margin), the surgeon would make a decision on whether or not to make an internal suture; the biomembrane’s capacity to adhere to the wound bed and commence integrative regeneration discourages the use of the internal suture.\n\nAfter applying and suturing all biomembranes, alginate hydrogel (Curatec®, LM Farma, São José dos Campos, SP, Brazil) was generously applied onto the treated areas. At this stage, the patients’ general anesthesia would be retracted. The patients were then medicated with a long-duration injectable antibiotic drug (cefovecin sodium, 8 mg/kg), and the treatment areas covered up with gauze tape.\n\nUpon awakening from the anesthesia, the first dose of an antifungal oral drug (ketoconazole, 10 mg/kg or itraconazole 5 mg/kg) was administered, with the treatment carried through for the duration as prescribed by the drug manufacturer.\n\nThe sutures were removed upon attesting that the biomembrane had adhered well to adjacent tissue structures and when the sutures would not provide any useful additional mechanical support to the host tissue-wound-biomembrane system integrity. This could happen as early as Day (D)3 after the surgery.\n\nApproximately every two days, the gauze tape was removed, and the wound cleaned with physiological saline and extra alginate hydrogel applied, as needed, and covered up again with fresh gauze tape, to prevent infection. This care was continued until the wound had completely healed.\n\nFollow-up time points for patient evaluation were defined as D0 post-surgery, D10, D15, D30 and a long-term time point at D120–D240 (4–8 months). At these time points, patients had photographs taken of the treatment area at roughly identical angles and distance to those taken immediately prior to the surgery. These photographs were used as basis for quantitative size assessments, as detailed below. When possible, a qualitative assessment of skin tissue quality would be done, either by a researcher associated with the study or by the owner. If another surgical procedure was scheduled for a study participant, for any reason, a small skin biopsy was attempted to be taken from the treatment area, but the study’s premise was that end-point histology would not be necessary to provide assertive conclusions to the treatment effectiveness, based on clinical assessment and healing rates.\n\nThe wounds were monitored for adverse effects such as infection, stagnation of healing for more than 15 days, and necrosis. If any adverse effects were confirmed, in the first 1–2 days, then the biomembranes were excised or scraped off host tissue. If adverse effects were confirmed in later stages (>D3), the biomembranes were not excised due to the fact that by this time they had absorbed significantly into the host tissue, rendering removal difficult or impossible. In this case, the patients were treated symptomatically and added to the “Discontinued intervention” group, as indicated in the Study Flow Diagram (Figure 1).\n\nIn order to use a quantitative wound size measurement method based on two-dimensional image analysis, the photographs would ideally depict the wounds at an angle as normal (90°) as possible to the idealized, hypothetically two-dimensional wound area.\n\nUsing ImageJ 1.49v (Schneider et al., 2012) and the “Straight line” tool, a straight-line was traced for a known distance. Usually, the tool was applied to trace the long dimension of a biomembrane, using sutures applied at either end as cue points for the extremities, and, assuming the sutures were placed some 2 millimeters from the edges of the biomembranes, using the “Set scale” tool that distance in pixels (as traced in the previous step) was taken to be equivalent to a distance in centimeters, 7.5 cm, slightly less than the nominal 8.0 cm biomembrane length (Figure 3). Subsequent uses of ImageJ’s “Measure” tool then output a distance in cm (distance) or cm2 (area), which were used for the analyses.\n\nThe first part of the method is to use the Set scale function on the image analyzed for a known distance, visible on the photo. In this case, the distance chosen was the length of the biomembrane (with the sutures used for cue points of approximate termini of the biomembrane, useful for later time points when the original biomembrane had been completely absorbed into the patient tissue), with a correction applied to account for the inclination of the target plane, using a trigonometric approach.\n\nSuch analysis assumes the wound plane to be in the same plane as the observer (camera). Whenever photographs diverged from this standard photographing procedure, a semi-quantitative assessment was made to estimate the approximate inclination angle relative to the observer (camera). This angle was then used to estimate the true distance of wound plane by projection of the observer (camera) plane distance onto the inclined wound plane, using a simplified trigonometric approach, as outlined in Figure 4.\n\n(A) The known distance (b) traced was, for photographs captured at an angle exactly normal to the wound plane, used for the “Set scale” function as-is, to trace the wound margin for use with the “Measure” function to obtain the wound area. (B) For photographs captured at an inclined angle, a semi-quantitative estimate of the inclination angle (theta) was made and a simplified trigonometric system used to approximately determine the known distance's (c) adjacent cathetus b’ (b’≈b; the unknown distance in the camera plane) measurement in the observer (camera) plane. Then, using “Set scale”, the distance traced in b was taken as equal to b’=c*cos(θ). (C and D) Thereafter, the wound margins were traced using the “Polygon selection” tool for subsequent measurement. This patient’s wound at this time point had formed a “pinch point” whereby the wound area above the eye had become discontinuous with the wound area adjacent to the eye (originally, these two were continuous) due to the advancing epithelialization, thus forming two distinct wound areas, 1 and 2, that were measured (using “Analyze-Measure”) as shown in C and D, respectively, and the sum of this observer (camera) plane wound area projected onto the true inclined wound plane using the trigonometric system explained above. The total wound area (whether a single area or the sum of distinct areas) was calculated and recorded.\n\nThis method was applied to the photographs collected for every time point, for every individual, and every wound region, and a total wound area recorded in a table. The whole process was repeated, albeit using the same photograph, but redoing the set scale, the approximate polygon wound area/margin delineations, and measure function, thus producing a duplicate measurement for statistics. For each individual and time point, the mean wound area and the standard deviation were calculated, along with the coefficient of variation.\n\nAnother inferred parameter, the “Wound healing rate” was calculated using the mean wound areas, defined as:\n\nA˙x=ΔAxΔtx\n\nwhere\n\nΔAx=Atime point x−Atime point x−1\n\nis the current wound area (Atime point x) measurement minus the previous wound area (Atime point x-1) measurement, and\n\nΔtx=tx−tx−1\n\nis the absolute time difference in days between the current wound measurement’s nominal time point and the previous’ nominal time point.\n\nThe calculated values for total wound area (Atime point x) and wound healing rate (Ȧtime point x) were tabulated and plotted graphically against nominal time points. A third parameter, wound healing area at a given nominal time point, %Atime point x, was defined as:\n\n%Atime point x=(max(A)−Atime point x)max(A)\n\nwhere\n\nmax(A)\n\nis the maximum wound area for the series.\n\nThe choice of using the maximum value of the series, rather than the initial value, counteracted the effects of the imprecision of the method; at times, the second (D10) and third (D15) time point measurements, ended up having higher values than the initial value. Albeit not theoretically impossible (wound growing instead of shrinking/healing), the clinical, qualitative assessment did not suggest this was the case in any individual studied, and the choice was therefore made to use the maximum value as the common reference point for the series, as explained.\n\nThe quantitative wound area measurements, using Image Analysis as the method, were obtained in duplicate. Basic statistical analysis was performed on these n=2 samples to provide a mean value (“AVERAGE function”), the standard deviation (“STDEV function”) and coefficient of variation (defined as the standard deviation divided by the mean value, for each sample). The software used for analysis was Excel 2011 (Microsoft Corporation, Redmond, WA, USA).\n\nIn the cases where pre-surgery and/or post-surgery biopsies of the affected area were collected, these were fixed in 10% formalin, embedded in paraffin, sectioned at 4 μm and subjected to the hematoxylin and eosin stain (H&E). The same protocol was applied to an unused biomembrane of identical specification to those used in the treatment group, to serve as a comparator for patient tissue histological samples. The stained tissue sections were visualized in an upright microscope with automatic scale bar incorporation.\n\n\nResults\n\nPatient N01 was approximately 6 years old (rescued animal, exact age unknown) at the time of surgery, female dog, with a skin lesion due to trauma (dog fight), disfigured fibrotic overgrowth of area adjacent to left eye with an area of approximately 10 cm2.\n\nPatient N04 was approximately 1 year old (rescued animal, exact age unknown) at the time of surgery, male cat, loss of skin equivalent to 2nd degree burn after car accident trauma with an area of approximately 120 cm2 (note: approximately 40 cm2 of the total lesion area was treated with the biomembranes, the remainder was treated using conventional wound treatment methods and healed at a slower rate than the biomembrane-treated areas).\n\nThe photography series presented a favorable outcome in terms of wound closure with the formation of healthy skin in the areas treated, in both patients studied in this interim report. In both patients, vascularization of the biomembrane-treated areas was observed clinically by the marked redness of the areas treated. The neo-skin that was formed presented itself as supple and robust, as opposed to a tissue that could form itself as hard and/or fragile, respectively. Moreover, in both patients, the areas treated ended up, at the end of the follow-up period, having a partial or complete coverage of fur, indicating the possible formation of new hair follicles in the treated areas, or, alternatively, a migratory mechanism whereby hair follicles in adjacent healthy areas were tugged toward the center of the wound/treated area, as part of the gradual wound closure process. The photographs collected are shown in Figure 5.\n\nIn patient N01, old fibrotic scar tissue was removed surgically and two biomembranes applied and sutured to adjacent tissue. By time point D30, the wound had fully healed. In patient N04, only part of the lesion (approximately 1/3) was treated on D0 with the biomembrane (as seen in the D0 post-surgery photograph). The non biomembrane-treated area (without biomembrane in the D0 post-surgery photograph) had in fact been treated with another wound dressing two weeks before, but this treatment proved to be ineffective. The subsequent regeneration of the biomembrane-treated areas, opportunely led to a snowball effect that eventually would lead to the complete healing also of the non biomembrane-treated areas. The photograph in the last time point shown here, having a very wide bracket of Day 120 to 240, was in fact taken on day 129; follow-up visits were also done at the end of the D120-240 time point interval, but photographs are not available. Nonetheless, the clinical data obtained at this final assessment confirm complete wound closure with clinically excellent parameters, such as suppleness, robustness, lack of contracture and with unrestricted mobility in the region in question (as seen on examination of the patient). The photographs provided basis for these qualitative assessments, in addition to the quantitative assessment aided by image analysis. In the case of N04, only the areas treated with the biomembrane de facto were used in the statistical analyses (even though the snowball effect led to healing of non biomembrane-treated areas, but at a slower rate).\n\nWorthy of an additional explanation, N04’s fur coat color/tonality changes significantly from D30 to the last time point, D120–240. Such changes are not uncommon in cats, especially in individuals that have not yet reached adult age, as was the case here. Other variables, such as temperature/season change (here, fall-winter), sunlight exposure, diet, hormones and stress also play a role, as do genetics and illness (here, the significant change may be due to the transition from a “sick” or “recovering” phenotype to a “healthy” phenotype).\n\nThe photographs collected were analyzed as described in the Methods, giving rise to values for wound area, healed area and healing rate, as tabulated in Table 2. Graphs for Wound area vs Time, Healing rate vs Time and Healing rate vs Healed area, based on the values shown in Table 2, are shown in Figure 6. The wounds of the two patients studied differed somewhat in size, which in turn made it more difficult to generalize features of healing in the wounds and individuals treated. However, it was noted that, in general, the healing rate accelerates during the first phase of the wound healing, reaching a maximum at D15–30. When looking at the Healed area percentage at maximum Healing rate, the results indicate that maximum Healing rate coincides with a Healed area percentage of around 50% or more. Upon investigating the general shape of the curves, in particular in Figure 6B curve depicting Healing rate vs Time, the first part of the curves (from the beginning of the data set to maximum value) have a comparable, accelerating shape, but the latter part (from maximum value to the last data point) have differing shapes, possibly owing to lack of suitable complementary time/data points between D30 and D120–240.\n\nObtained from analysis of data collected at the time points as per the study protocol, from Day 0 to Day 120–240. The values for parameter A, wound area, were calculated using an image analysis approach of photographs taken at the different time points, from which the two other parameters: %A, healed area; and Ȧ, healing rate, were calculated. x, mean; SD, standard deviation; CV, coefficient of variation.\n\n(A) Total wound area vs Time. (B) Wound healing rate vs Time. (C) Wound healing rate vs Wound healing percentage. Error bars indicate +1 SD from mean data point.\n\nAs described in the Methods, pre-surgery histology would only be done with the patient’s pre-existing scar tissue, removed during the first stage prior to the application of biomembranes. Moreover, post-surgery histology of a biopsy taken from the biomembrane-treated healed wounds would only be done in conjunction with another surgical procedure executed at a later date, i.e., no biopsy would be taken solely for the purpose of obtaining post-surgery histology data.\n\nIn this interim report, patient N01’s pre-existing scar tissue was removed, and the corresponding histomicrograph is shown in Figure 7 (A and C). In the case of patient N04, no pre-existing scar tissue was available for histology and the biomembranes were applied directly to the wound bed as described above. At the time of writing, no later surgical procedures were scheduled for either of the two individuals covered by this interim report, and, as such, no post-surgery histology data is available. Figure 7 (B and D) shows a histomicrograph of a biomembrane representative of the biomembranes applied in the study.\n\n(A and C; 20 × and 4 × magnification of same specimen) Hematoxylin and eosin (H&E) histomicrograph of N01’s pre-surgery tissue in the treatment area, exhibiting numerous hallmarks of fibrotic scar tissue. Epidermis is flattened, with no undulation on the surface, no epidermal rete ridges extending into the dermis, and no dermal papillae. The uppermost papillary dermis presents a dense collagenous matrix, and the lowermost, reticular dermis exhibits highly acidophilic regions with denser bundles of collagen. (B and D; 20 × and 4 × magnification of same specimen) H&E histomicrograph of a representative collagen biomembrane (i.e., equivalent to the implanted biomembranes prior to their utilization). Horizontally aligned thin collagen fibers with some larger pores interspersed throughout the structure, along with smaller pores with a more uniform distribution. The lowermost part of the structure presents increased acidophilicity and collagen density. This particular part of the histomicrograph shows the local elevation of approximately 100 μm in height. These elevations are part of the characteristics of the biomembrane, and are present throughout the structure, spaced approximately 2.5 mm from one another.\n\n\nDiscussion\n\nThe present clinical trial did not have a randomized sample assigned to a treatment and a placebo arm; rather, the sole sample group was the treatment group, and a quasi-experimental control group (pseudo-placebo arm; Figure 1) was taken as the clinicians’ assessment of how the individual in question would fare without the treatment with the biomembrane. In both patients reported in this interim report, the biomembranes were capable of stimulating formation of quality skin tissue and complete wound closure in the time frame studied, indicating efficacy. The clinicians’ assessment (J.F.L.N. and M.H.S.C) was that the wounds would take approximately 3 × longer to heal compared to the time it took using the biomembrane treatment, or, possibly, the wounds would not have healed at all, or formed fibrotic scar tissue, indicating superiority compared to conventional wound therapy. No adverse effects were noted, indicating safety. In addition to the objectives of the study pertaining to skin repair, an additional apparent benefit of the biomembrane treatment was seen with the apparent stimulation of fur growth, a characteristic that will be investigated in more detail for subsequent patients enrolled in the study. The image analysis method for quantitative size assessments is robust and accurate for photographs taken perpendicular to the wound (commercial systems also exclusively mandate perpendicular image capturing), but when taken at an angle, even with the trigonometric correction as detailed herein, the final size will suffer and not present good accuracy. Ideally, all photographs should be taken completely perpendicularly to the wound to maximize accuracy (and precision).\n\nThe enrolment of patients has been slow, due to cost considerations, but agreements have been made with new partner clinics to ensure the total number of 15 patients is reached by the end of the study period. We believe the biomembrane is well suited for wound healing in veterinary patients, and foresee other uses as well, e.g. as a periosteum substitute. Moreover, due to the similarities between the veterinary medical device studied in this report and human medical devices based on tissue engineering technologies, we expect that the study may have repercussions beyond the veterinary domain and into the human domain, and vice versa.\n\n\nData availability\n\nDataset 1: Biomembrane data files. This dataset contains the unedited photographs from the study. DOI, 10.5256/f1000research.15138.d206373 (Kaasi et al., 2018).",
"appendix": "Competing interests\n\n\n\nThe first author (A.K.) is the founder and shareholder of Eva Scientific Ltd, where the biomembrane used in this clinical trial was developed and is sold.\n\nThe co-authors are veterinary practitioners of Sorocaba Veterinary Hospital (J.F.L.N., J.A.M.F. and M.H.S.C.) and full-time employees of University of Campinas (A.L.J. and P.K.). All co-authors read and agreed to the manuscript and had full access to the data, and attest to the study’s impartiality as a whole, as well as declare no competing interests.\n\n\nGrant information\n\nThis work was supported by the Council for Scientific and Technological Development (CNPq 573661/2008-1) and São Paulo Research Foundation (FAPESP 2008/57860-3) grants awarded to Biofabris (The National Institute of Biofabrication) and by a FAPESP grant awarded to Eva Scientific Ltd (FAPESP 2014/22799-3).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: CONSORT checklist.\n\nClick here to access the data.\n\nSupplementary File 2: Study protocol.\n\nClick here to access the data.\n\nSupplementary File 3: Footage of application of double-layer regenerative collagen biomembrane.\n\nClick here to access the data.\n\n\nFootnotes\n\n1 There are as yet no widely known veterinary clinical trial registries, apart from the American Veterinary Medicine Association’s (AVMA) Animal Health Studies Database. This database is the first of its kind, and was only 4 months old at the time the present trial was conceptualized and commenced (October 2016). The authors only became aware of its existence in subsequent months, and the study details were submitted to the database thereafter; however, the AVMA informed the authors that they were unable to accept studies from outside the U.S.A. and Canada, with the reason being they do not have the resources to consider study listings in the context of animal use laws and regulations that vary from country to country. As such, it was decided to include the study protocol as a Supplementary File to this article. It is also available on the Eva Scientific website (http://www.evascientific.com/products/a161012001/).\n\n\nReferences\n\nBell E, Ehrlich HP, Buttle DJ, et al.: Living tissue formed in vitro and accepted as skin-equivalent tissue of full thickness. Science. 1981; 211(4486): 1052–1054. PubMed Abstract | Publisher Full Text\n\nBills JD, Berriman SJ, Noble DL, et al.: Pilot study to evaluate a novel three-dimensional wound measurement device. Int Wound J. 2016; 13(6): 1372–1377. PubMed Abstract | Publisher Full Text\n\nDavis KE, Constantine FC, Macaslan EC, et al.: Validation of a laser-assisted wound measurement device for measuring wound volume. J Diabetes Sci Technol. 2013; 7(5): 1161–1166. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDorin RP, Pohl HG, Filippo RE, et al.: Tubularized urethral replacement with unseeded matrices: what is the maximum distance for normal tissue regeneration? World J Urol. 2008; 26(4): 323–326. PubMed Abstract | Publisher Full Text\n\nEisenbud D, Huang NF, Luke S, et al.: Skin Substitutes and Wound Healing: Current Status and Challenges. Wounds. 2004; 16(1): 2–17. Reference Source\n\nFürdös I, Fazekas J, Singer J, et al.: Translating clinical trials from human to veterinary oncology and back. J Transl Med. 2015; 13(1): 265. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGawai AA, Shaikh F, Gadekar M, et al.: A review on: Phase ‘0’ clinical trials or exploratory investigational new drug. Turkish Journal of Pharmaceutical Sciences. 2017; 14(1): 84–89. Publisher Full Text\n\nGroeber F, Holeiter M, Hampel M, et al.: Skin tissue engineering--in vivo and in vitro applications. Adv Drug Deliv Rev. 2011; 63(4–5): 352–366. PubMed Abstract | Publisher Full Text\n\nKaasi A, Lima-Neto JF, Matiello-Filho JA, et al.: Dataset 1 in: Regenerative collagen biomembrane: Interim results of a Phase I veterinary clinical trial for skin repair. F1000Research. 2018. Data Source\n\nPapini R: Management of burn injuries of various depths. BMJ. 2004; 329(7458): 158–160. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRheinwald JG, Green H: Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell. 1975; 6(3): 331–343. PubMed Abstract | Publisher Full Text\n\nSchneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012; 9(7): 671–675. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVail DM: Cancer clinical trials: development and implementation. Vet Clin North Am Small Anim Pract. 2007; 37(6): 1033–1057; v. PubMed Abstract | Publisher Full Text\n\nYannas IV, Burke JF: Design of an artificial skin. I. Basic design principles. J Biomed Mater Res. 1980; 14(1): 65–81. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "35668",
"date": "19 Jul 2018",
"name": "Kaveh Memarzadeh",
"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 entitled, “Regenerative collagen biomembrane: Interim results of Phase I veterinary clinical trial for skin repair” by Kaasi et al explores the application of acellular collagen membranes for veterinary wound healing purposes in dogs and cats. The study provides the preliminary results to what may be considered a phase I clinical study on animal subjects that had been victims of injuries with the immediate need of care. The manuscript is well written with a good structure and the data presented is interesting and worth publishing. There are, however, some minor clarifications and modifications that are suggested to improve the manuscript before publication.\n\n1: The study presents an interesting idea of letting veterinary clinical trials pave the way for human use and equivalent therapy. Currently, there is a lack of global consensus concerning the rules and regulations governing veterinary clinical trials. The establishment of such rules and regulations could possibly make the translational process from veterinary subjects to man more flexible for some medical devices by providing substantial data and thereby allow a faster route to regulatory approval. It is suggested that the authors add a couple of sentences in the discussion section on the potential prospect of the proposed wound healing product for human applications and the long-term objectives or plans of the authors for the clinical translation of the product.\n2. There is not sufficient information on the characteristics of the collagen biomembrane to fully differentiate and appreciate the current approach. Please provide more information concerning the main attributes of the collagen biomembrane and a comparison with similar marketed products or products in development. How does the presented collagen membrane differ from other biomembranes including cell-incorporating and synthetic alternatives?\n3. It is somewhat unclear with the current status of the animal studies. Please provide additional information regarding the current status with the number of completed animal subjects and information and timeline for the remaining studies and subjects\n4. If there were any adverse effects noticed such as infection and necrosis after application of the biomembrane, what measures are there to investigate this further? Did this occur in this current study? If yes, please provide some insight as to why you think this was the case.\n5. Regarding photographing the patients, how do the authors explain the method of ‘Semi-quantitative assessment’? More detail is required about this method. Going forward, the method of photographing animals should be standardised in order to lessen the chance of errors in the study.\n\n6. There are no clear controls for each model presented. We appreciate that this is a very challenging task to achieve at this stage, however going forward a ‘quasi-experimental control group’ will not suffice. We would advise the researcher to have solid controls in place when recruiting subjects for the next studies.\n\n7. The histological evidence provided here is very limited. This needs to be improved upon significantly. The histological data presented in this paper is only for one patient and only in a pre-surgical stage ( A and C). The researchers need to elaborate on the reasons why they were unable to provide small histological samples from patients after treatment. The local elevations present on the biomembrane and pores are highlighted in Figure 7, however, the biological significance of each characteristic is missing. It would also be great if the researchers made the scales more visible on the H&E images.\n8- The authors refer to an apparent fur growth on one of the patients, however, this is not backed by supporting evidence as the numbers are very low. Please refrain from using terms such as ‘benefit’ and ‘stimulation’ in these instances and just state what has been observed and report fully when enough evidence is available.\n\n9. Please provide an indication of the time generally required to get a veterinary medical device on the market.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "34932",
"date": "27 Sep 2018",
"name": "Iris Ribitsch",
"expertise": [
"Reviewer Expertise Veterinary medicine",
"regenerative medicine",
"tissue engineering"
],
"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\nIntroduction: The introduction focuses on an important topic which is drawing more and more attention in medical sciences: The idea of “one health” - the collaborative effort of multiple disciplines – to attain optimal health for people, animals and the environment. The authors point out the necessity to conduct veterinary clinical trials in a standardized manner in order to obtain valid and meaningful results to facilitate the translation into human medicine. Currently there is a lack of global consensus concerning the rules and regulations governing veterinary clinical trials.\n\nHowever, this discussion (although very important) should not make the bigger part of the introduction. Instead please include a discussion of solid scientific evidence on the similarities and discrepancies of human, feline and canine wound healing and why it may be valid to extrapolate results from one species to the other. Furthermore, please offer and overview of the research question, the state of the art in the field, how the current state may be improved, what is new about the collagen membrane used in this study compared to other similar products and how and why the collagen membrane employed in this study may therefore contribute to an improvement of the current state of the art. Currently the collagen membrane, which is the core of this paper, is not described whatsoever.\n\nMethods: The materials and methods section offers a description of the study design of the envisaged phase 1 clinical trial. It gives in detail guidelines how to proceed at certain stages in the course of the study and based on findings which may be encountered within the scope of the study (e.g. “If the affected area had been subject to natural scarring, with the formation of hard, unhealthy, fibrotic scar tissue, this was surgically removed, exposing subcutaneous tissue. In this case, the tissue would be fixed in 10% formalin and submitted to histological analysis, using standard procedures. If the treatment area was an unhealed, deepithelialized wound bed, but with no fibrotic scar tissue formation, then shallow scalpel incisions were made scattered throughout the treatment area, causing local bleeding, with the aim of promoting regeneration and integration with the later application of the collagen biomembranes.”\n\n“If any adverse effects were confirmed, in the first 1–2 days, then the biomembranes were excised or scraped off host tissue. If adverse effects were confirmed in later stages (>D3), the biomembranes were not excised due to the fact that by this time they had absorbed significantly into the host tissue, rendering removal difficult or impossible. In this case, the patients were treated symptomatically and added to the “Discontinued intervention” group…”).\n\nAlso detail on how data should be obtained and by whom are included (“When possible, a qualitative assessment of skin tissue quality would be done, either by a researcher associated with the study or by the owner. If another surgical procedure was scheduled for a study participant, for any reason, a small skin biopsy was attempted to be taken from the treatment area….”).\n\nHowever, the materials and methods should provide detailed information on methods and analysis carried out particularly for the current 2 patients included into the study.\n\nThe anaesthesia regimen which is described in great detail is not so relevant for the interpretation of the study results as e.g. the preparation of the skin prior to surgery which may have considerable influence on the healing process. Please provide more information.\n\nResults: Following up clinical patients especially long term is very challenging and depends on the cooperation of the animal owners and their willingness and time to take their animals to the follow up examinations etc. Using series of pictures which can be taken either by the owners or the treating veterinarian at defined time points to document the healing process will certainly facilitate cooperation. Detailed instructions when and how (distance camera to animal, angle camera to wound etc.) are offered in the methods section. However, the photos in the paper (Figure 5) do unfortunately not look standardised even when comparing only the pictures taken from one patient, which hampers comparability.\nIs the work clearly and accurately presented and does it cite the current literature? No. Is the study design appropriate and is the work technically sound? No. Are sufficient details of methods and analysis provided to allow replication by others? Yes If applicable, is the statistical analysis and its interpretation appropriate?\n\nMeaningful statistical analysis on n=2 is impossible.\n\nA detailed description of the results is more useful in this case.\nAre all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Partly\n\nThe results of the study are interesting and encouraging and (so I am convinced) worth further research efforts. However, in order to improve the quality of the clinical study standardisation should be increased, the photo documentation should be improved, mandatory re-checks with the treating veterinarian should be implemented during which a healing score is obtained based on healing rate and a qualitative assessment of newly formed skin tissue. Histologic evaluation pre and post-surgery should be implemented. Also inclusion of a control group would considerably increase the quality of the study. The clinicians assessment (“The clinicians’ assessment was that the wounds would take approximately 3 × longer to heal compared to the time it took using the biomembrane treatment, or, possibly, the wounds would not have healed at all, or formed fibrotic scar tissue, indicating superiority compared to conventional wound therapy.) and a ‘quasi-experimental control group’ are not sufficient. I certainly agree with the author´s that it wouldn´t be possible and ethically debateable to include a control group receiving no treatment, but a control group receiving the current gold standard of wound treatment would be feasible and still increase the quality of the study.\n\nI strongly suggest publishing the results from the two patients as case reports because they are valuable and interesting for the scientific community and veterinary clinicians, but at the current stage unfortunately not sufficient for an original research article. However, I am looking forward to read more and see more results to come from further patients enrolled into this interesting and promising clinical trial and whish the authors all the best for the continuation of their study.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "41549",
"date": "11 Jan 2019",
"name": "Kalan Bastos Violin",
"expertise": [
"Reviewer Expertise Biomaterials",
"Experimental and Comparative Pathology",
"Pre-clinical Trials",
"Histopathology",
"Veterinary Medicine",
"Immunohistochemistry",
"Lectinhistochemistry",
"Image Analysis",
"Microscopy"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 seems to provide enough evidence for the proposed goal, which basically was to induce skin repair in a critical size wound equivalent to 2nd to 3rd degree burns in depth, using an engineered biomembrane based on collagen type I.\nThe biggest challenge was to gather enough subjects eligible to the study, presenting only 2 cases isn't ideal, but expected since real cases for clinical trials are difficult to respond to all criteria to be included. For further studies these criteria should be flexed to maximize the number of subjects included, specially if the exclusion of those were related to unsuccessful results, cut to not compromise the product. A larger number of subjects is desirable and for clinical trials mandatory. Nevertheless for this level of initial trial and concept it is acceptable and the results were remarkable.\nAnother point to be kept for the following research is to give a molecular approach to the biopsies. Basic histology just present the morphological aspects of the tissue and it would be interesting to gather from these same samples an immunohistochemistry analysis to characterize the molecular content of the formed tissue looking for proteins related to repair and healing, but also for inflammatory and regulatory markers that play an important role in those process. This would give a enormous depth to the consolidation of the material and present more reliable and straightforward data for comparison.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-729
|
https://f1000research.com/articles/7-727/v1
|
12 Jun 18
|
{
"type": "Software Tool Article",
"title": "Automating the PathLinker app for Cytoscape",
"authors": [
"Li Jun Huang",
"Jeffrey N. Law",
"T. M. Murali",
"Li Jun Huang",
"Jeffrey N. Law"
],
"abstract": "PathLinker is a graph-theoretic algorithm originally developed to reconstruct the interactions in a signaling pathway of interest. It efficiently computes multiple short paths within a background protein interaction network from the receptors to transcription factors (TFs) in a pathway. Since December 2015, PathLinker has been available as an app for Cytoscape. This paper describes how we automated the app to use the CyRest infrastructure and how users can incorporate PathLinker into their software pipelines.",
"keywords": [
"Network Biology",
"Shortest Paths",
"Pathway Reconstruction",
"CyREST API"
],
"content": "Introduction\n\nPathLinker is an algorithm that automates the reconstruction of any human signaling pathway by connecting the receptors and transcription factors (TFs) in that pathway through a physical and regulatory interaction network1. In a comprehensive quantitative evaluation on NetPath pathways, PathLinker achieved higher reconstruction accuracy than several other state-of-the-art algorithms. In addition, PathLinker’s novel prediction that the cystic fibrosis transmembrane conductance regulator (CFTR), an ion-channel receptor, is involved in the Wnt pathway was experimentally validated1. In general, PathLinker can be used to connect any set of sources to any set of targets in a given network.\n\nThe PathLinker app for Cytoscape is an implementation of this algorithm. The PathLinker app2 was first released in the middle of December 2015. While it is possible to use the PathLinker app in conjunction with other Cytoscape apps, it can be cumbersome to create such workflows using the Cytoscape user interface. It is also challenging to reproduce the results of these workflows.\n\nHere we present an CyREST-based API that allows users to incorporate PathLinker algorithms into their own software pipelines. The PathLinker application programming interface (API) will facilitate automated analysis of complex networks in reproducible workflows, including in conjunction with other CyREST enabled Cytoscape apps3.\n\n\nMethods\n\nThe PathLinker API allows external software (written in languages such as Python and R) and tools (e.g., Jupyter Notebook) to access PathLinker functions via the REST protocol. The API follows the OSGi design pattern. The PathLinker API exposes two functions via JAX-RS annotations that allows them to be discovered by CyREST3 and be made available to external callers via REST. We have documented these functions using Swagger annotations to meet the Cytoscape Automation documentation standards. The Swagger annotation allows the PathLinker API functions to be accessed via the Swagger user interface along with other API functions exposed by CyREST.\n\nWe substantially updated and refactored the PathLinker codebase to follow the principles of the OSGi modular design and to remove redundant code. Specifically, we refactored the code for generating the k shortest paths, network visualization, and functions related to the user interface (e.g., generating information in the “Result Panel”) into distinct Task classes that can be managed by the Task Manager in the Cytoscape API. The current design enables us to use the same codebase for running PathLinker through the Cytoscape user-interface as well as through the REST API. This modular design will facilitate easy expansion of the app in the future, e.g., by implementing additional subnetwork-finding algorithms.\n\nThe PathLinker API is accessible directly through the Swagger user interface within Cytoscape or by using any REST-enabled client. Note that users must have installed v1.4 of the PathLinker app and Cytoscape v3.6.0 or higher. Moreover, an instance of Cytoscape must be running on the user’s computer. In the “Use Cases” section, we describe a sample workflow using py2cytoscape and provide an example Jupyter Notebook.\n\nAs shown in Figure 2, the PathLinker API provides two POST functions, “/pathlinker/v1/currentView/run” and “/pathlinker/v1/networkSUID/run”, which run PathLinker on the currently selected network and on the given networkSUID respectively. Both functions send user-selected source nodes and target nodes and a set of parameters to the PathLinker Cytoscape app.\n\nThe app computes the k shortest simple (loopless) paths that connect any source to any target in the network specified in the POST function, generates a subnetwork that contains these paths and a view of this subnetwork, creates a table in the Result Panel that contains these paths (See Figure 3b), and adds a “path rank” column to the Edge Table that contains the rank of the first path in which each edge appears. The POST functions return the computed paths, the SUIDs of the subnetwork and subnetwork view, and the name of the “path rank” column created by the app.\n\nWe summarize the parameters of the API functions and their outputs below.\n\nAPI parameters\n\nThe API functions have the same set of parameters as the user can set in the user interface of the PathLinker app (see Figure 1a and Figure 1b). The user should provide these parameters in JSON format, whether they use the Swagger user interface or invoke the PathLinker API in code or via external tools. Table 1 contains an overview of the parameters, their types and a brief description. For a detailed description of the parameters, please see the Swagger documentation or documentation of the PathLinker app.\n\nThe last column records whether an equivalent option is present in the user interface of the PathLinker app. T/F - True/False.\n\nAPI Output\n\nThe PathLinker API functions returns a response in JSON format with the following fields:\n\nsubnetworkSUID: The SUID of the subnetwork created in Cytoscape.\n\nsubnetworkViewSUID: The SUID of the subnetwork view created in Cytoscape.\n\npathRankColumnName: The name of the edge column created in the subnetwork’s edge table containing the rank of the first path in which each edge appears. The user can programmatically access this table using the CyREST API.\n\npaths: The list of paths generated by the algorithm sorted by the path rank. This list contains the same information as the result table in the “Result Panel” created by the PathLinker app. Each entry in the list contains the following fields that describe a path:\n\nrank: The rank of the path. For example, “100” for the 100th shortest path.\n\nscore: The total weight of the edges in the path.\n\nnodeList: The list of nodes in the path. Each element in this list appears in the “name” column in the node table of the network provided as input to PathLinker. If multiple paths have the same score, then we order them lexicographically by their node lists.\n\nIf the user sets skipSubnetworkGeneration to true in the input to the API functions, then the PathLinker app does not generate the subnetwork and its view. Therefore, the API output will not contain the first three attributes, i.e., subnetworkSUID, subnetworkViewSUID, and pathRankColumnName.\n\nThe functions in the PathLinker API implement careful checks of the input. The function return the following error codes to provide meaningful feedback to the user:\n\n400: Invalid user input. The JSON response lists all faulty input along with a reason why the input is correct, allowing the user to correct all the errors at the same time.\n\n404: Current network or network with the given SUID does not exist.\n\n422: No paths found. This error message indicates that PathLinker could not find any path connecting the source(s) to the target(s) in the given network. This error can occur if every source is disconnected from every target.\n\n\nUse cases\n\nWe provide two Jupyter Notebooks that illustrate the PathLinker API. The first notebook implements a use case on a simple network with 11 nodes and 13 edges, shown in Figure 3a. The notebook shows how to use Python and the py2cytoscape library to 1) load this network into Cytoscape, 2) call the PathLinker API with a set of parameters (Figure 1), 3) view the computed paths and subnetwork, and 4) save the paths and/or subnetwork image to a file. The API function call in the notebook asks PathLinker to find the two shortest paths connecting “a” to “e” or to “h” while treating the network as undirected. It also asks PathLinker to return more than two paths if their scores are tied with the second path.\n\nEven though we set k = 2, PathLinker returned three paths; Figure 3b shows the subnetwork containing these paths. PathLinker did so because we used the “Include tied paths” option, and the second and third path both had a score of three (they contain three edges each). In this example, we used the “Treat network as undirected” option because even though the edges in the network were intended to be undirected, py2cytoscape treats networks imported from the Python NetworkX package as directed.\n\nThe second notebook implements a more complex example that we presented in the paper describing the PathLinker app2. Here, we used PathLinker to compute and analyze a network of interactions connecting proteins that are perturbed by the drug Lovastatin. The PathLinker API functions now allow us to automate this use case enabling fast reproduction of the results, as shown in the Jupyter Notebook.\n\n\nSummary\n\nThe subnetwork returned by the PathLinker API is another network in the current Cytoscape session. Hence, this network is amenable for analysis by other CyREST APIs and/or Cytoscape apps. Some examples include calling the Diffusion app’s API on the subnetwork, or modifying the subnetwork using standard CyREST API calls such as adding visual styles to the subnetwork or calling different layout algorithms.\n\nIn the near future, we plan to provide users more freedom in determining which column in the Node Table should act as the input for source and target fields. We will also implement additional sub-network generation algorithms. We will implement these new features in parallel in the PathLinker app and in the REST API.\n\nWe hope that these additions will increase making PathLinker a method of choice in the network biology community.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSoftware available from:http://apps.cytoscape.org/apps/pathlinker\n\nThe Cytoscape app source code and sample Jupyter Notebooks are available at https://github.com/Murali-group/PathLinker-Cytoscape\n\nArchived source code as at time of publication: http://dx.doi.org/10.5281/zenodo.12523084\n\nLicense: GNU General Public License version 3",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nNational Science Foundation grant [CCF-1617678] supported this research. The research is also partially based upon work supported by the Office of the Director of National Intelligence (ODNI), Intelligence Advanced Research Projects Activity (IARPA), via the Army Research Office (ARO) under cooperative Agreement Number [W911NF-17-2-0105]. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the ODNI, IARPA, ARO, or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon.\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\nRitz A, Poirel CL, Tegge AN, et al.: Pathways on demand: automated reconstruction of human signaling networks. NPJ Syst Biol Appl. 2016; 2: 16002. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGil DP, Law JN, Murali TM: The PathLinker app: Connect the dots in protein interaction networks [version 1; referees: 1 approved, 2 approved with reservations]. F1000Res. 2017; 6: 58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOno K, Muetze T, Kolishovski G, et al.: CyREST: Turbocharging Cytoscape Access for External Tools via a RESTful API [version 1; referees: 2 approved]. F1000Res. 2015; 4: 478. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang LJ, Law J, Gil D, et al.: Murali-group/PathLinker-Cytoscape: v1.4.1 (Version 1.4.1). Zenodo. 2018. Data Source"
}
|
[
{
"id": "34942",
"date": "25 Jun 2018",
"name": "Alexander Pico",
"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\nIdentifying paths through complex networks is a common use case. The PathLinker app provides a nice approach via GUI or scripting. The methods and use cases are detailed and clear. The provided Python Notebook is especially handy.\nMy only suggestions for future versions of this work would be to support table column-based specification of sources and targets (already mentioned in your summary) as well as an R Markdown version of your script. The RCy3 package should make it easy to translate as it is analogous to py2cytoscape.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "34937",
"date": "17 Jul 2018",
"name": "Stefan Wuchty",
"expertise": [
"Reviewer Expertise Systems biology",
"bioinformatics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to 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 ' Automating the PathLinker app for Cytoscape' by Huang, law and Murali introduces an API solution of their PathLinker software that was previously published. Although the PathLinker algorithm was made available through a Cytoscape app it turned out that its usage through the Cytoscape framework was initially cumbersome. With the new solution the authors used the CyREST API that allows to tap the results of the PathLinger app when used in conjunction with Cytoscape to be incorporated in user specific pipelines. As such, the authors allow through their API to use a Cytoscape app and results thus obtained to link to methods outside the Cytoscape framework. In particular, API output can easily be tapped using python and R code as well as using the Junyper notebook.\n\nThe proposed API appears very useful for researchers who want to consider the advantages of the PathFinder app w/o being constrained by the Cytoscape framework. The current manuscript describes the basic steps to get a user going. As a small change I would describe the examples in a bit more detail to show a fully worked example. In particular, Fig. 3 goes in that direction but is a bit generic. Working a real example would help with the transparency of the underlying approach.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-727
|
https://f1000research.com/articles/6-409/v1
|
03 Apr 17
|
{
"type": "Review",
"title": "Autophagy and airway fibrosis: Is there a link?",
"authors": [
"Anudeep Kota",
"Deepak A. Deshpande",
"Mehra Haghi",
"Brian Oliver",
"Pawan Sharma",
"Anudeep Kota",
"Deepak A. Deshpande",
"Mehra Haghi",
"Brian Oliver"
],
"abstract": "In the past decade, an emerging process named “autophagy” has generated intense interest in many chronic lung diseases. Tissue remodeling and fibrosis is a common feature of many airway diseases, and current therapies do not prevent or reverse these structural changes. Autophagy has evolved as a conserved process for bulk degradation and recycling of cytoplasmic components to maintain basal cellular homeostasis and healthy organelle populations in the cell. Furthermore, autophagy serves as a cell survival mechanism and can also be induced by chemical and physical stress to the cell. Accumulating evidence demonstrates that autophagy plays an essential role in vital cellular processes, including tissue remodeling. This review will discuss some of the recent advancements made in understanding the role of this fundamental process in airway fibrosis with emphasis on airway remodeling, and how autophagy can be exploited as a target for airway remodeling in asthma and chronic obstructive pulmonary disease.",
"keywords": [
"asthma",
"COPD",
"airway remodeling",
"airway mesenchymal cells"
],
"content": "Introduction\n\nAutophagy is an evolutionarily conserved pathway for the turnover of organelles and proteins by lysosomal-dependent processing1. During autophagy, newly formed double-membrane structures, called autophagosomes, encapsulate cytoplasmic material, such as dysfunctional or damaged organelles or proteins. The autophagosomes then fuse with lysosomes, thus delivering the sequestered cargo for lysosomal-dependent degradation2, as described in Figure 1. In the last decade, autophagy has emerged as a fundamental process involved in tissue and cellular homeostasis, and thus has been implicated in maintaining basal physiologic (healthy) and adaptive pathophysiologic responses (disease)2–4. There is increasing evidence to suggest that autophagy can impact the pathogenesis and/or progression of many human diseases2,4,5, including neurodegenerative diseases6, cancer7, heart diseases8,9 and immune disorders (reviewed in 3,10).\n\nThe autophagy pathway proceeds through several phases, including initiation (formation of a preautophagosomal structure leading to an isolation membrane, or phagophore), vesicle elongation, autophagosome maturation and cargo sequestration, and autophagosome–lysosome fusion. In the final stage, autophagosomal contents are degraded by lysosomal acid hydrolases and the contents of the autolysosome are released for metabolic recycling.\n\nFibrotic airway remodeling remains a key pathological feature correlating with a decline in lung function and disease progression in both asthma and chronic obstructive pulmonary disease (COPD) patients11,12. While there is no treatment for preventing development of airway remodeling, cell fate phenomena, such as autophagy, have been shown to play a role in asthma and COPD pathogenesis. Here, we review the progress in understanding how autophagy can contribute to airway fibrosis and the emerging strategies to target this process for therapeutic benefit.\n\n\nEvidence of autophagy in asthma\n\nAsthma is a chronic inflammatory disease of the lungs, characterized by airway inflammation, airway hyperresponsiveness and tissue remodeling. It affects more than 300 million people worldwide and this number is estimated to escalate to 400 million by 202513.\n\nThe role of autophagy in pulmonary diseases has gained attention in the past decade, but only very few studies have described the role of autophagy in asthma. One of the very first studies that described a direct role between asthma and autophagy revealed the association of the ATG5 gene in the pathogenesis of asthma14. This study found a genetic association in 1338 adult patients with asthma and the expression of the autophagy gene ATG5. ATG5 encodes the ATG5 protein of 276 amino acids. During autophagy, the ATG5 protein interacts with ATG12 and ATG16 to form a ATG12-ATG5-ATG16 complex. This complex is associated with autophagosomal membrane elongation by interaction with ATG3, leading to ATG8-phosphatidyl ethanolamine formation15. This association was further validated in another study with 312 asthmatic and 246 control children, which showed that genetic variants in ATG5 are associated with pathogenesis of childhood asthma16,17. Furthermore, a study by Poon et al. revealed the role of ATG5 in adult asthma, and also found an increased number of autophagosomes in fibroblast and epithelial cells from severe asthmatics when compared to healthy volunteers15. Recent studies show that there is emerging evidence for the role of autophagy in both eosinophilic18 and neutrophilic asthma19, and convey its link to severe asthma and fibrotic tissue remodeling.\n\nA recent study by Ban et al. investigated the role of autophagy in sputum granulocytes, peripheral blood cells and peripheral blood eosinophils of severe and non-severe asthmatics20. They found increased autophagy in the immune cells from the severe asthmatics when compared to non-severe and healthy controls. This clearly indicates that induction of autophagy in immune cells is associated with severe asthma. By contrast, a study conducted by Akbari’s group reveals the induction of neutrophilic airway inflammation and hyperreactivity on deletion of CD11 cell specific ATG5 mice. In addition, in this study augmented neutrophilic inflammation in Atg5(-/-) mice is IL-17A driven and glucocorticoid resistant21.\n\nIn our own hands, we have found increased signatures of key autophagy genes in the lungs of asthmatic patients when compared with non-asthmatics, suggesting that basal autophagy is higher in asthma (unpublished data). Furthermore, we also found increased expression of autophagy proteins in the lung tissue obtained from chronic mouse model of HDM-induced asthma and this expression was found to correlate with pro-fibrotic signaling (Smad) and extracellular matrix protein (collagen) in the lung (unpublished data).\n\nThese data suggest that autophagy and airway fibrosis occur together with allergic insult, and act as a key driver for airway remodeling in allergic asthma. The current literature clearly indicates that the autophagy-phenomenon may be a crucial driver in the pathogenesis of asthma, particularly in severe forms of the disease, with an unknown underlying mechanism. The therapeutic modulation of autophagy with novel inhibitors may lead to the development of a new class of drugs for severe asthma.\n\n\nEvidence of autophagy in COPD\n\nCOPD is a progressive lung disease characterized by accelerated decline in lung function over time. Its most common pathological feature includes emphysema and chronic bronchitis. Airway obstruction in COPD in associated with formation of peribronchial fibrosis, increased wall thickness and excess mucus secretion, especially in the smaller airways22. Exposure to cigarette smoke is one major cause of COPD; however only 25% of smokers develop COPD, which suggests the existence of numerous other factors contributing to COPD (such as genetic predisposition and oxidative stress)23,24. The role of autophagy in COPD seems to be more complex than anticipated, as some studies showed its impairment25–27, while others suggest it facilitates disease pathogenesis28–31. More recently, the role of selective autophagy (such as mitophagy, ciliophagy and xenophagy) in COPD pathology has been proposed31.\n\nThe very first demonstration of autophagy in COPD was shown by Chen et al., where authors found increased autophagy markers in the lungs of COPD patients28. The authors also found a similar expression of elevated autophagy markers at various stages of the disease. This suggests that altered autophagy could be a key regulator in the pathogenesis and progression of COPD. The same study also showed that the autophagy marker LC3II/LC3I was significantly increased in lungs from patients with α-1 anti-trypsin deficiency when compared with non-COPD donors28. In addition, many other studies have shown the increased expression of autophagy markers both in vitro and in vivo when exposed to cigarette smoke extract16,28,29,32, which explains increased loss of alveolar epithelial cells as seen in emphysema.\n\nMoreover, to investigate the role of autophagy in chronic bronchitis, Lam and colleagues demonstrated that induction of autophagy leads to shortening of cilia in mouse tracheal epithelial cells exposed to cigarette smoke30. They further found that autophagy gene deficient mice (Becn1+/- or Map1lc3B-/-) were resistant to the shortening of cilia in tracheal epithelial cells when exposed to cigarette smoke, demonstrating a direct role of autophagy in this process30. Recent studies have demonstrated that selective autophagy (namely mitophagy) plays an important role in regulating mitochondrial function, which in turn has a crucial role in COPD pathogenesis33. However, the specific role of mitophagy in tissue injury mediated by cigarette smoke remains obscure and requires further study31.\n\nOverall, autophagy plays a key role in COPD pathogenesis, especially in the development of emphysema, but the underlying mechanisms by which it promotes emphysema and bronchitis in COPD is not clear.\n\n\nAutophagy and fibrotic airway remodeling\n\nThe pathogenesis of COPD and asthma is typified by structural changes in the lung, collectively known as airway remodeling, which is characterized by basement membrane fibrosis, epithelial goblet cell hyperplasia, deposition of extracellular matrix proteins and smooth muscle hypertrophy34,35. Current therapies provide very limited benefit on airway remodeling12,36–40, thus identifying new drug targets that can prevent or reduce airway remodeling in asthma and COPD is vital to reverse structural changes that determines the underlying cause of the disease34,41. TGFβ1 is a well-known regulator of inflammation and fibrotic remodeling in COPD and asthma, and it is upregulated in both diseases11. TGFβ1 is the most abundant isoform of the TGF-β family and is secreted by most immune cells, including airway epithelial, smooth muscle, fibroblast and endothelial cells. TGFβ1 acts through the Smad dependent pathway and independent pathways, like mitogen activated protein kinases (MAPKs) namely p38 MAPK and c-Jun-N-terminal kinase, leading to the accumulation of extracellular matrix (ECM).\n\nAutophagy-mediated TGFβ1-induced fibrosis plays a key role in the pathogenesis of heart and kidney diseases42,43, and recent studies in human airway smooth muscle cells have demonstrated that TGFβ1 induced autophagy is required for collagen and fibronectin production, while silencing of key autophagy-inducing proteins Atg5 and Atg7 leads to reduction in pro-fibrotic signaling and ECM protein release44,45. It is now believed that increased production of matrix proteins, as seen in airway remodeling, requires huge resources of energy, which is compensated by enhanced autophagy flux within the cell45. Our own data (unpublished) demonstrates that there is a simultaneous induction of autophagy and pro-fibrotic signaling in human airway mouse models of allergic asthma. We propose that blockade of autophagy in the lungs can lead to reduction in airway fibrosis, as seen in airway remodeling in asthma and COPD. Recent observations suggest that autophagy exhibits a circadian rhythm and this rhythmic activation of autophagy is regulated by the transcription factor C/EBPβ46. These findings need to be further evaluated in the context of asthma and COPD, as circadian rhythm genes can also be a potential target to modulate autophagy46.\n\n\nTherapeutic strategies: Novel autophagy modulators\n\nThe mechanistic insight of disease pathogenesis in chronic airway diseases, such as asthma and COPD, is complex. Altered structural changes in the lung correlates with poor lung function, severity of disease and response to therapy in asthma and COPD. Current therapy does not target the underlying cause/s of the disease and does not restore the structural integrity of the airways12,36–40; therefore, novel mechanisms, such as autophagy, pose attractive therapeutic options in airway remodeling. With the emerging role of autophagy in asthma and COPD and our increased understanding in the disease pathogenesis, we believe that the autophagy pathway can be exploited for therapeutic benefit. One of the exciting features of this pathway is that it can be inhibited at several steps, including initiation and vesicle elongation, autophagosome formation and maturation, and autophagosome lysosome fusion. Emerging evidence suggests that autophagy inhibition at the very first step is beneficial in an ovalbumin-induced model of asthma in mice, where 3-methyl adenine, a class III PI3K inhibitor, reduced the expression of autophagy marker LC3B with simultaneous reduction in airway eosinophilia18. However, there is a need for more research in exploring the role of various PI3K inhibitors in the context of autophagy dependent-airway remodeling, using the most suitable models of asthma and COPD, as modulation of autophagy by the PI3K pathway can be an attractive target for both severe asthma and COPD47–50.\n\nBafilomycin A1 is a late phase potent inhibitor of autophagy, which acts through the inhibition of vacuolar type H+-ATPase that is present on lysosomes. Inhibition of vacuolar type H+-ATPase by bafilomycin A1 leads to acidification of lysosomes and inhibits their fusion with autophagosomes51. Previous studies also indicate that bafilomycin inhibits the infection and airway inflammation induced by rhinovirus52. Chloroquine is a well know anti-malarial and anti-rheumatoid agent, and it is also used for blocking autophagy at a late stage by lysosomal dysfunction (lysosomal lumen alkalizers)51. The most important shortcoming of chloroquine is that this molecule blocks autophagy only at higher concentrations. Lys01 and Lys05 were recently developed modified versions of chloroquine, which are 10-fold more potent in autophagy inhibition51,53,54; however these require further validation in asthma and COPD models.\n\nKinases, such as adenosine monophosphate activated protein kinase (AMPK) and ULK1 (mammalian orthologue of yeast protein kinase Atg1) are required for autophagy. AMPK senses nutrient deficiency and positively regulates ULK1, leading to activation of autophagy. Moreover, it has been reported that loss of either of these kinases results in defective autophagy (mitophagy)55. Modulating the AMPK-ULK1 pathway may be an option for treatment of chronic airway diseases. Thus, there is a need to identify the right autophagy mechanism in the progression and development of airway remodeling in asthma and COPD, which can become a realistic drug target to prevent structural changes in the lung.\n\n\nConclusions and future perspectives\n\nBased on the current scientific data, the autophagy pathway is directly linked to asthma and COPD pathophysiology and may act as a potential contributor to fibrotic airway remodeling, as described in Figure 2. Future research should be directed towards understanding the specific pathway and the mechanism/s leading to the pathophysiological activation of autophagy in disease. Efforts should be made in developing a modulator of autophagy as a new therapeutic strategy for treatment of airway remodeling in asthma and COPD. In the future, we propose to study the role of circadian rhythm genes in the activation of the autophagy pathway and understand whether modulating circadian rhythm genes, using novel small chemical compounds, has any therapeutic benefit in animal models of asthma and COPD.\n\nAn altered autophagy pathway is seen in response to cellular stress in asthma and chronic obstructive pulmonary disease (allergens or smoke), leading to activation and crosstalk between structural airway and immune cells. This further leads to impairment of autophagy causing degradation of intracellular constituents, providing an energy resource for ECM protein biosynthesis, and releasing mediators of inflammation and profibrotic signaling, which collectively lead to airway remodeling in the lung. ECM: extracellular matrix; TGFβ: transforming growth factor-beta.",
"appendix": "Author contributions\n\n\n\nPS, BO, DD, MH conceived the plan of the review. PS and AK 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 was supported by the Centre for Health Technologies, Chancellors Fellowship Research Program, the National Institute of Health, and the National Health and Medical Research Council.\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\nYorimitsu T, Klionsky DJ: Autophagy: molecular machinery for self-eating. Cell Death Differ. 2005; 12(Suppl 2): 1542–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEskelinen EL, Saftig P: Autophagy: a lysosomal degradation pathway with a central role in health and disease. Biochim Biophys Acta. 2009; 1793(4): 664–73. PubMed Abstract | Publisher Full Text\n\nKlionsky DJ: Autophagy: from phenomenology to molecular understanding in less than a decade. Nat Rev Mol Cell Biol. 2007; 8(11): 931–7. PubMed Abstract | Publisher Full Text\n\nRavikumar B, Sarkar S, Davies JE, et al.: Regulation of mammalian autophagy in physiology and pathophysiology. Physiol Rev. 2010; 90(4): 1383–435. PubMed Abstract | Publisher Full Text\n\nMizushima N, Levine B, Cuervo AM, et al.: Autophagy fights disease through cellular self-digestion. Nature. 2008; 451(7182): 1069–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee JA: Autophagy in neurodegeneration: two sides of the same coin. BMB Rep. 2009; 42(6): 324–30. PubMed Abstract | Publisher Full Text\n\nRosenfeldt MT, Ryan KM: The multiple roles of autophagy in cancer. Carcinogenesis. 2011; 32(7): 955–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartinet W, De Meyer GR: Autophagy in atherosclerosis: a cell survival and death phenomenon with therapeutic potential. Circ Res. 2009; 104(3): 304–17. PubMed Abstract | Publisher Full Text\n\nGottlieb RA, Mentzer RM: Autophagy during cardiac stress: joys and frustrations of autophagy. Annu Rev Physiol. 2010; 72: 45–59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevine B, Mizushima N, Virgin HW: Autophagy in immunity and inflammation. Nature. 2011; 469(7330): 323–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang YC, Zhang N, Van Crombruggen K, et al.: Transforming growth factor-beta1 in inflammatory airway disease: a key for understanding inflammation and remodeling. Allergy. 2012; 67(10): 1193–202. PubMed Abstract | Publisher Full Text\n\nPrakash YS, Halayko AJ, Gosens R, et al.: An Official American Thoracic Society Research Statement: Current Challenges Facing Research and Therapeutic Advances in Airway Remodeling. Am J Respir Crit Care Med. 2017; 195(2): e4–e19. PubMed Abstract | Publisher Full Text\n\nBoulieri A, Hansell A, Blangiardo M: Investigating trends in asthma and COPD through multiple data sources: A small area study. Spat Spatiotemporal Epidemiol. 2016; 19: 28–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoon AH, Chouiali F, Tse SM, et al.: Genetic and histologic evidence for autophagy in asthma pathogenesis. J Allergy Clin Immunol. 2012; 129(2): 569–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoon A, Eidelman D, Laprise C, et al.: ATG5, autophagy and lung function in asthma. Autophagy. 2012; 8(4): 694–5. PubMed Abstract | Publisher Full Text\n\nCaveolin-1 Protects Against Cigarette Smoking Induced Autophagic Cell Death and Emphysema in Chronic Obstructive Pulmonary Disease. In A13. INSIGHTS IN COPD PATHOGENESIS. A1001. Publisher Full Text\n\nMartin LJ, Gupta J, Jyothula SS, et al.: Functional variant in the autophagy-related 5 gene promotor is associated with childhood asthma. PLoS One. 2012; 7(4): e33454. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu JN, Suh DH, Trinh HK, et al.: The role of autophagy in allergic inflammation: a new target for severe asthma. Exp Mol Med. 2016; 48(7): e243. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPham DL, Ban GY, Kim SH, et al.: Neutrophil autophagy and extracellular DNA traps contribute to airway inflammation in severe asthma. Clin Exp Allergy. 2017; 47(1): 57–70. PubMed Abstract | Publisher Full Text\n\nBan GY, Pham DL, Trinh TH, et al.: Autophagy mechanisms in sputum and peripheral blood cells of patients with severe asthma: a new therapeutic target. Clin Exp Allergy. 2016; 46(1): 48–59. PubMed Abstract | Publisher Full Text\n\nSuzuki Y, Maazi H, Sankaranarayanan I, et al.: Lack of autophagy induces steroid-resistant airway inflammation. J Allergy Clin Immunol. 2016; 137(5): 1382–1389.e9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurgel PR, Bourdin A, Pilette C, et al.: [Structural abnormalities and inflammation in COPD: a focus on small airways]. Rev Mal Respir. 2011; 28(6): 749–60. PubMed Abstract | Publisher Full Text\n\nKirkham PA, Barnes PJ: Oxidative stress in COPD. Chest. 2013; 144(1): 266–73. PubMed Abstract | Publisher Full Text\n\nKoyama H, Geddes DM: Genes, oxidative stress, and the risk of chronic obstructive pulmonary disease. Thorax. 1998; 53(Suppl 2): S10–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVij N, Chandramani P, Westphal CV, et al.: Cigarette smoke induced autophagy-impairment accelerates lung aging, COPD-emphysema exacerbations and pathogenesis. Am J Physiol Cell Physiol. 2016; ajpcell.00110.2016. PubMed Abstract | Publisher Full Text\n\nTran I, Ji C, Ni I, et al.: Role of Cigarette Smoke-Induced Aggresome Formation in Chronic Obstructive Pulmonary Disease-Emphysema Pathogenesis. Am J Respir Cell Mol Biol. 2015; 53(2): 159–73. PubMed Abstract | Publisher Full Text\n\nFujii S, Hara H, Araya J, et al.: Insufficient autophagy promotes bronchial epithelial cell senescence in chronic obstructive pulmonary disease. Oncoimmunology. 2012; 1(5): 630–641. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen ZH, Kim HP, Sciurba FC, et al.: Egr-1 regulates autophagy in cigarette smoke-induced chronic obstructive pulmonary disease. PLoS One. 2008; 3(10): e3316. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim HP, Wang X, Chen ZH, et al.: Autophagic proteins regulate cigarette smoke-induced apoptosis: protective role of heme oxygenase-1. Autophagy. 2008; 4(7): 887–95. PubMed Abstract | Publisher Full Text\n\nLam HC, Cloonan SM, Bhashyam AR, et al.: Histone deacetylase 6-mediated selective autophagy regulates COPD-associated cilia dysfunction. J Clin Invest. 2013; 123(12): 5212–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRyter SW, Choi AM: Autophagy in lung disease pathogenesis and therapeutics. Redox Biol. 2015; 4: 215–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhu L, Barrett EC, Xu Y, et al.: Regulation of Cigarette Smoke (CS)-Induced Autophagy by Nrf2. PLoS One. 2013; 8(4): e55695. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMizumura K, Cloonan SM, Nakahira K, et al.: Mitophagy-dependent necroptosis contributes to the pathogenesis of COPD. J Clin Invest. 2014; 124(9): 3987–4003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChung KF: The role of airway smooth muscle in the pathogenesis of airway wall remodeling in chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2005; 2(4): 347–54; discussion 371–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Alawi M, Hassan T, Chotirmall SH: Transforming growth factor β and severe asthma: a perfect storm. Respir Med. 2014; 108(10): 1409–23. PubMed Abstract | Publisher Full Text\n\nRoyce SG, Moodley Y, Samuel CS: Novel therapeutic strategies for lung disorders associated with airway remodelling and fibrosis. Pharmacol Ther. 2014; 141(3): 250–60. PubMed Abstract | Publisher Full Text\n\nRoth M: Airway and lung remodelling in chronic pulmonary obstructive disease: a role for muscarinic receptor antagonists? Drugs. 2015; 75(1): 1–8. PubMed Abstract | Publisher Full Text\n\nPauwels B, Jonstam K, Bachert C: Emerging biologics for the treatment of chronic rhinosinusitis. Expert Rev Clin Immunol. 2015; 11(3): 349–61. PubMed Abstract | Publisher Full Text\n\nKume H, Fukunaga K, Oguma T: Research and development of bronchodilators for asthma and COPD with a focus on G protein/KCa channel linkage and β2-adrenergic intrinsic efficacy. Pharmacol Ther. 2015; 156: 75–89. PubMed Abstract | Publisher Full Text\n\nPelaia G, Vatrella A, Maselli R: The potential of biologics for the treatment of asthma. Nat Rev Drug Discov. 2012; 11(12): 958–72. PubMed Abstract | Publisher Full Text\n\nSoltani A, Walters EH, Reid DW, et al.: Inhaled corticosteroid normalizes some but not all airway vascular remodeling in COPD. Int J Chron Obstruct Pulmon Dis. 2016; 11: 2359–2367. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhavami S, Cunnington RH, Gupta S, et al.: Autophagy is a regulator of TGF-β1-induced fibrogenesis in primary human atrial myofibroblasts. Cell Death Dis. 2015; 6: e1696. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing Y, Choi ME: Regulation of autophagy by TGF-β: emerging role in kidney fibrosis. Semin Nephrol. 2014; 34(1): 62–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAutophagy Regulates Tgf-Beta1 Induced Fibrosis In Human Airway Smooth Muscle Cells. In A67. SURVIVOR-RCMB: AUTOPHAGY, SENESCENCE, AND SURVIVAL.A2110. Publisher Full Text\n\nZeki AA, Yeganeh B, Kenyon NJ, et al.: Autophagy in airway diseases: a new frontier in human asthma? Allergy. 2016; 71(1): 5–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMa D, Lin JD: Circadian regulation of autophagy rhythm through transcription factor C/EBPβ. Autophagy. 2012; 8(1): 124–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu YT, Tan HL, Shui G, et al.: Dual role of 3-methyladenine in modulation of autophagy via different temporal patterns of inhibition on class I and III phosphoinositide 3-kinase. J Biol Chem. 2010; 285(14): 10850–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDowdle WE, Nyfeler B, Nagel J, et al.: Selective VPS34 inhibitor blocks autophagy and uncovers a role for NCOA4 in ferritin degradation and iron homeostasis in vivo. Nat Cell Biol. 2014; 16(11): 1069–79. PubMed Abstract | Publisher Full Text\n\nRonan B, Flamand O, Vescovi L, et al.: A highly potent and selective Vps34 inhibitor alters vesicle trafficking and autophagy. Nat Chem Biol. 2014; 10(12): 1013–1019. PubMed Abstract | Publisher Full Text\n\nStark AK, Sriskantharajah S, Hessel EM, et al.: PI3K inhibitors in inflammation, autoimmunity and cancer. Curr Opin Pharmacol. 2015; 23: 82–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang YP, Hu LF, Zheng HF, et al.: Application and interpretation of current autophagy inhibitors and activators. Acta Pharmacol Sin. 2013; 34(5): 625–635. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYamaya M, Sasaki H: [The pathogenesis and therapy of virus infection-induced senile bronchial asthma]. Nihon Ronen Igakkai Zasshi. 2000; 37(6): 464–8. PubMed Abstract | Publisher Full Text\n\nAmaravadi RK, Winkler JD: Lys05: a new lysosomal autophagy inhibitor. Autophagy. 2012; 8(9): 1383–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcAfee Q, Zhang Z, Samanta A, et al.: Autophagy inhibitor Lys05 has single-agent antitumor activity and reproduces the phenotype of a genetic autophagy deficiency. Proc Natl Acad Sci U S A. 2012; 109(21): 8253–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEgan DF, Shackelford DB, Mihaylova MM, et al.: Phosphorylation of ULK1 (hATG1) by AMP-activated protein kinase connects energy sensing to mitophagy. Science. 2011; 331(6016): 456–61. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "21469",
"date": "10 Apr 2017",
"name": "Eugene Roscioli",
"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\nSummary statement Kota et al. present an interesting and challenging subject for review – the interrelationship between autophagy in the context of airway diseases that exhibit airway remodelling as a major pathological feature. Challenging because we still have a lot to learn about both airway remodelling, and indeed the diseases such as asthma and COPD that potentiate fibrotic airway changes, and further still the contribution of increased or blocked autophagy as a driver, or parallel phenomena in this scenario. Overall they provide a needed and good start to the discussion, that will no doubt pioneer further debate in these areas off airway pathobiology. Hence, one leaves this review thinking about the link between fibrosis and autophagy, but also, wondering what insight they gained to direct new lines of inquiry.\n\nMajor comments As stated, this is by no means an easy literature review to compose. At first one would think this is due to limited publication of such studies. However, a PubMed search using “Fibrosis[Title] and Autophagy[Title]”, show there are 83 hits for this query. While the vast majority are not respiratory in nature, it would be beneficial if Kota et al. could use mechanistic insights from other diseases to draw hypothetical parallels with the situation that may be occurring in the airways, which they elude to in paragraph 2 under the heading “Autophagy and fibrotic airway remodelling”, but little supposition is made in regards to this thereafter. Indeed, this paragraph is possibly the only one which directly address the subject matter stated by the title. Refreshingly however, they generously state new unpublished finding (on three occasions) from their laboratory which provides insight for the reader and which have the potential to direct further lines if understanding.\n\nA current question for autophagy is: is it a block in autophagic flux leading to (or contributing to) the cellular dyshomeostasis, or is autophagy increased because it is unable to manage the protracted/chronic disease stressors? While Kota et al. touch upon this question for COPD, and correctly state that there is significant conjecture with regards to this, subsequent discussion mainly lent to the increased autophagy scenario, with an extended discussion with regards to cilia (ciliophagy). Indeed, we now know the relationship between cilia and autophagy is a complex dual interaction whereby they foster each other’s activity/generation rather than a one way situation where increased autophagy leads to the degradation of cilia (Pampliega et al., 2016). Hence, some new insights/thoughts which could weigh into the autophagic flux vs. increased autophagy debate would be welcomed.\n\nFurther, airway fibrosis occurs as a result of the chronic inflammation (in addition to e.g. altered smooth muscle-epithelia interactions, between themselves and the inflammatory cells), while autophagy is known to be an anti-inflammatory mechanism. Hence, beyond TGFβ, some insights into how the dysregulation of autophagy leads to fibrosis driven by complex inflammatory changes at the airways would have been informative.\n\nHence, while this review most certainly warrants publication, and marks yet another achievement for the Oliver laboratory, the informed reader may ask for more than the start of the dialogue presented in this particular review.\n\nMinor comments The manner of composition on occasions lacks maturity. For example, “ATG5 encodes a protein of 276 amino acids”, is not needed in a review article. The chronological presentation of concepts/studies, should be forfeited to a style of writing that groups and discusses processes to illustrate a pattern of phenomena that helps orient/inform the reader. Avoid using definite statements such as “clearly indicates”, particularly in a review, and especially in areas where we have so much more to learn such as is the case for chronic lung disease and autophagy. Finally, approximately one third of the 55 references are review articles.",
"responses": [
{
"c_id": "3525",
"date": "11 Jun 2018",
"name": "Pawan Sharma",
"role": "Author Response",
"response": "Reviewer 1: We thank reviewer for the constructive feedback. Reviewers’ comments are addressed one by one below (comments in italics followed by our response).C1: As stated, this is by no means an easy literature review to compose. At first one would think this is due to limited publication of such studies. However, a PubMed search using “Fibrosis[Title] and Autophagy[Title]”, show there are 83 hits for this query. While the vast majority are not respiratory in nature, it would be beneficial if Kota et al. could use mechanistic insights from other diseases to draw hypothetical parallels with the situation that may be occurring in the airways, which they elude to in paragraph 2 under the heading “Autophagy and fibrotic airway remodelling”, but little supposition is made in regard to this thereafter. Indeed, this paragraph is possibly the only one which directly address the subject matter stated by the title. Refreshingly however, they generously state new unpublished finding (on three occasions) from their laboratory which provides insight for the reader and which have the potential to direct further lines if understanding.R1: We agree with the reviewer about the complexity of the topic being reviewed. However, we would like to state that this mini-review is focussed on “lung-related” fibrotic changes in asthma and COPD. We haven’t discussed other lung diseases such because the autophagy pathway is quite distinct in those diseases which will make it difficult to bring them under the same category. This review has tried to discuss the dysregulated autophagy pathway in asthma and COPD pathology as literature suggests there are some parallels in autophagy signalling mechanisms. Further, TGFbeta driven fibrotic changes that are dependent and mediated through autophagy are discussed in this review which also occur in other organs such as liver and kidney fibrosis.C2. A current question for autophagy is: is it a block in autophagic flux leading to (or contributing to) the cellular dyshomeostasis, or is autophagy increased because it is unable to manage the protracted/chronic disease stressors? While Kota et al. touch upon this question for COPD, and correctly state that there is significant conjecture with regards to this, subsequent discussion mainly lent to the increased autophagy scenario, with an extended discussion with regards to cilia (ciliophagy). Indeed, we now know the relationship between cilia and autophagy is a complex dual interaction whereby they foster each other’s activity/generation rather than a one-way situation where increased autophagy leads to the degradation of cilia (Pampliega et al., 2016). Hence, some new insights/thoughts which could weigh into the autophagic flux vs. increased autophagy debate would be welcomed.R2. Reviewer has asked a very interesting question with relevance to autophagy: whether it is being blocked or increased during disease pathology. We believe it is too early to say that autophagy flux is blocked or there is a net increase in autophagy that could lead to disease pathology. Published literature in COPD suggests that that there is an increase in autophagy in disease, and this has been also demonstrated in the cigarette smoke-induced experimental COPD models with additional increase in ciliophagy contributing to reduced mucus clearance (Choi’s group). In asthma, it is not yet clear whether there is a block in autophagy flux or there is a net increase in autophagy, though recent data demonstrate increase number of autophagosomes in asthma that associates with lung function and ECM protein in humans (Poon et al). In experimental models of allergic asthma there is an evidence to believe that autophagy is increased in severe asthma models and inhibition of autophagy in vivo may be a viable option.C3. Further, airway fibrosis occurs as a result of the chronic inflammation (in addition to e.g. altered smooth muscle-epithelia interactions, between themselves and the inflammatory cells), while autophagy is known to be an anti-inflammatory mechanism. Hence, beyond TGFβ, some insights into how the dysregulation of autophagy leads to fibrosis driven by complex inflammatory changes at the airways would have been informative.R3. We agree with the reviewer that there are complex cell-cell interactions and multitude of signalling pathways which eventually contributes to structural changes in the airways. Keeping in mind the scope of this review we have focussed on TGFβ-dependent mechanisms that are related to the activation of autophagy in the lung. We have now added text to the MS to reflect these changes. C4. The manner of composition on occasions lacks maturity. For example, “ATG5 encodes a protein of 276 amino acids”, is not needed in a review article. The chronological presentation of concepts/studies, should be forfeited to a style of writing that groups and discusses processes to illustrate a pattern of phenomena that helps orient/inform the reader. Avoid using definite statements such as “clearly indicates”, particularly in a review, and especially in areas where we have so much more to learn such as is the case for chronic lung disease and autophagy. Finally, approximately one third of the 55 references are review articles.R4: We have now modified the text in the MS to reflect these changes."
}
]
},
{
"id": "22182",
"date": "02 May 2017",
"name": "Kimberley C.W. 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\nSummary:\nThis review clearly demonstrates the need for more studies investigating the role of autophagy in respiratory disease. It is good that the authors added some unpublished data into the review article as support of evidence.\n\nMajor comments:\nFigure 2 can be improved. Eg: at the arrow between Autophagy and Cell Growth/ECM release, the authors can add the words increase energy resource so the readers understand the association between Autophagy and Cell Growth/ECM release. Authors should consider adding ‘Genetics’ into the figure. The authors could add ‘circadian rhythm’ into the figure with a ‘?’ to show that it is a factor for further investigation.\n\nFor the collectively structural changes of airway remodeling, please include smooth muscle hyperplasia (James et al. 2012; Am J Respir Crit Care Med).\n\nIn the autophagy and fibrotic airway remodelling section, the paragraphs could be rearranged. Eg do not split TGFb1 between the paragraphs. The section about circadian rhythm seems to be a better fit in the Therapeutic strategies: Novel autophagy modulators section, as it is something to further evaluate as a potential target.\n\nAre there any genetic component as evidence of autophagy in COPD? For example: Chen et al. 2015; J Formos Med Assoc.\n\nShould have a reference for the sentence regarding TGFb1 leads to accumulation of ECM. ECM could be abbreviate earlier in the article.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "3526",
"date": "11 Jun 2018",
"name": "Pawan Sharma",
"role": "Author Response",
"response": "Reviewer 2 We thank reviewer for constructive feedback. Reviewers’ comments are addressed one by one below (comments in italics followed by our response).C1: Figure 2 can be improved. Eg: at the arrow between Autophagy and Cell Growth/ECM release, the authors can add the words increase energy resource, so the readers understand the association between Autophagy and Cell Growth/ECM release. Authors should consider adding ‘Genetics’ into the figure. The authors could add ‘circadian rhythm’ into the figure with a ‘?’ to show that it is a factor for further investigation.R1: We have modified the figure accordingly.C2: For the collectively structural changes of airway remodeling, please include smooth muscle hyperplasia (James et al. 2012; Am J Respir Crit Care Med).R2: We have now added this with the reference. C3: In the autophagy and fibrotic airway remodelling section, the paragraphs could be rearranged. Eg do not split TGFb1 between the paragraphs. The section about circadian rhythm seems to be a better fit in the Therapeutic strategies: Novel autophagy modulators section, as it is something to further evaluate as a potential target.R3: It has been moved to the section on novel autophagy modulators.C4: Are there any genetic component as evidence of autophagy in COPD? For example: Chen et al. 2015; J Formos Med Assoc.R4: Not as far as we know.C5: Should have a reference for the sentence regarding TGFb1 leads to accumulation of ECM. ECM could be abbreviated earlier in the articleR5: Reference added."
}
]
},
{
"id": "21520",
"date": "04 May 2017",
"name": "Sukhwinder Singh Sohal",
"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\nSummary –\nThe current manuscript has reviewed the literature on relationship between autophagy and airway fibrosis in chronic lung disease, especially asthma and COPD. This topic is of great interest and very little work has been done to understand the underlying mechanisms. I think this article is a timely reminder and will stimulate research into this area, which is much needed. Overall, this a well presented article but I have some suggestions, which will enhance readability and understanding on the diseases discussed.\n\nMajor comments –\nBefore authors go into the details for role of autophagy in asthma/remodeling, I think they need to add a separate section on airway remodeling and autophagy first.\n\nPlease list the airway remodeling changes described in asthma so far and then discuss which might be related to autophagy.\n\nI think authors need to add a reference if there is any work done on fibroblast populations in asthma to make the case for fibrosis or suggest as potential future work.\n\nSimilar changes I think need to be made for the COPD section, especially listing airway remodeling changes described so far, for example Rbm fragmentation, vascularity, epithelial mesenchymal transition (EMT) etc.\n\nAuthors stated in the evidence for autophagy in COPD section that major features are emphysema and chronic bronchitis. I think it is very important to describe small airway destructive pathology in COPD. The classic fixed airflow obstruction described in COPD comes from the small airways. These airways are destroyed by up to nearly 40% well before the diagnosis of COPD. So damage starts quite early on and it is this site where fibrosis really occurs and EMT is central to this. Airway disease is the primary phenomenon in COPD.\n\nEGFR has also been regarded as one of the important drivers of airway remodeling in asthma and COPD, that needs to be discussed.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "3527",
"date": "11 Jun 2018",
"name": "Pawan Sharma",
"role": "Author Response",
"response": "Reviewer 3: We thank reviewer for constructive feedback. Reviewers’ comments are addressed one by one below (comments in italics followed by our response).C1: Before authors go into the details for role of autophagy in asthma/remodelling, I think they need to add a separate section on airway remodelling and autophagy first.R1: We have now added this to the text.C2: Please list the airway remodelling changes described in asthma so far and then discuss which might be related to autophagy.R2: We have now added new text to the MS.C3: I think authors need to add a reference if there is any work done on fibroblast populations in asthma to make the case for fibrosis or suggest as potential future work.R3: We have added references related to fibroblasts and potential future direction in this area.C4: Similar changes I think need to be made for the COPD section, especially listing airway remodelling changes described so far, for example Rbm fragmentation, vascularity, epithelial mesenchymal transition (EMT) etc.R4: We have now modified the text and added these suggestions as well.C5: Authors stated in the evidence for autophagy in COPD section that major features are emphysema and chronic bronchitis. I think it is very important to describe small airway destructive pathology in COPD. The classic fixed airflow obstruction described in COPD comes from the small airways. These airways are destroyed by up to nearly 40% well before the diagnosis of COPD. So, damage starts quite early on and it is this site where fibrosis really occurs and EMT is central to this. Airway disease is the primary phenomenon in COPD.R5: These are valid points made by the reviewers, we have modified the text in the MS.C6: EGFR has also been regarded as one of the important drivers of airway remodeling in asthma and COPD, that needs to be discussed.R6: We agree that EGFR is an important player in driving airway remodeling but the current review focusses on autophagy pathway and how it is linked to airway fibrosis."
}
]
}
] | 1
|
https://f1000research.com/articles/6-409
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.