On 2016 May 27, C. Alan Titchenal commented:

There appears to be a discrepancy in this abstract. It states that ". . . those reporting aspartame intake had a steeper positive association between BMI and glucose tolerance" (i.e. as BMI increases, so does glucose tolerance) and then it concludes that aspartame consumption is associated with "obesity-related impairments in glucose tolerance." It also seems that their conclusion could suggest that overweight/obese people with impaired glucose tolerance are more likely to opt for artificial sweeteners in attempts to decrease their carbohydrate intake.

On 2016 Jun 01, Kenneth Witwer commented:

These data cannot be used to conclude differential expression of miRNAs between soybean strains (only one sample per strain), but the authors are commended for publishing this study, which is interesting and potentially instructive for several reasons. Four RNA samples were prepared: one each from two soybean strains that were genetically modified using modern biotechnology and one each from the two parent strains that were genetically modified by traditional breeding approaches. A sequencing library was prepared from each RNA sample.

The first interesting aspect of the study is that the small RNA profile of soybean was overwhelmingly dominated by fragments of longer RNAs, especially the rRNAs, such that only single digit percentages of miRNAs are mapped. It would be good to compare these results with other soybean studies to determine if sample quality or processing issues contributed to degradation. To this end, it would be helpful to know more about the quality control that was mentioned. rRNA degradation fragments are variable in length and well conserved (in contrast with erroneous suggestions of species-specificity, e.g. Zhou Z, 2015), and they should not be confused with miRNAs. A study of one family, previously called "MIR2911," showed that the family does not associate with RNA regulatory machinery (Yang J, 2015) and is thus unlikely to be involved in RNA regulation, within or across organisms.

A second and more substantive contribution of the study comes from the 14 newly predicted miRNAs. Although at low abundance and not reportedly associated with "GM" strains, these may be worth some follow-up through bioinformatics and perhaps even in the lab.

As a third matter of note, the study reminds us to check data and not rely only on abstracts and conclusions in the text. Unfortunately, no conclusions can be drawn about differential expression of miRNAs between parental and "GM" strains, since there are no biological replicates. More reported differences were found between the two traditionally genetically modified "parent" strains than between parent and biotech, yet curiously this is not emphasized. The very minor reported differences, which are inconsistent between the two parent-daughter strain pairs, could be due to any number of factors that have nothing to do with biotechnology-introduced modifications, from growing, harvesting, storage, shipping, and other handling conditions (all out of the control of the authors...presumably these beans were not grown side-by-side in the same field); to a variety of technical factors introduced during isolation of RNA, library prep, and sequencing; to chance.

Finally, the study prompts reflection on the term "genetically modified," which applies to all strains examined here. The popular media definition is used in this study to refer only to molecular biotechnology-produced strains, not the parents produced by legacy methods of genetic modification.

On 2016 Jun 02, Michael Tatham commented:

Is SUMO5 a pseudogene?

There are known to be many SUMO pseudogenes in humans (http://www.ncbi.nlm.nih.gov/pubmed/12383504). A fair position when confronted with a claim that a new SUMO paralog has been discovered is to assume it is a non-expressed pseudogene until otherwise convincing evidence is provided. This paper lacks one piece of critical evidence supporting the idea that SUMO5 really exists as a protein, and that is the presence of endogenous protein.

When BLAST searched, the nucleotide sequence of SUMO5 (originally termed SUMO13 according to the authors’ GenBank entry: FJ042790.1), returns a top hit of “Homo sapiens SUMO1 pseudogene 1 (SUMO1P1), non-coding RNA Sequence ID: ref|NR_002189.3|”. The only difference is a single nucleotide T23 (in SUMO13/SUMO5), which is C in SUMO1P1. This may be a primer synthesis error or a DNA sequencing error.

To put beyond reasonable doubt that SUMO5 is not a pseudogene at least two pieces of new experimental evidence showing SUMO5 is expressed in cells is required. I can think of three good ways to do this:

(1) Mass-spectrometric evidence of a peptide unique to SUMO5.

(2) Cross-reaction of a SUMO5-specific antibody with an endogenous protein.

(3) Editing of the genome to insert an epitope tag into the endogenous SUMO5 gene, with the intention of detecting the protein using an antibody specific to the tag.

All three of these pieces of evidence will be strengthened by parallel studies comparing cells with and without SUMO5-specific knock-down.

RTPCR experiments intending to detect mRNA are particularly uninformative given that DNA contamination often leads to false-positives. This is especially true for SUMO5 given the fact the gene is intronless, a notable characteristic of pseudogenes.

On 2016 Aug 17, Karyn Schmidt commented:

Please note SLNCR1 and LINC00673 refer to the same non-coding transcript. While this manuscript was being prepared, multiple other manuscripts were published referring to the lncRNA as LINC00673: Childs EJ, 2015, Shi X, 2016, Zheng J, 2016, and Invalid PMID: 27260162.

On 2016 May 29, Sonia Kharay commented:

Effect size, confidence intervals and P value misinterpretations are explained with ample examples. It is an exemplary article on errors in presentation of statistical data.

On 2016 May 28, Arnaud Chiolero MD PhD commented:

This unexpected review is welcomed: it is well written, readable by non-statisticians, and comprehensive. Highly recommended for students in public health and epidemiology and for health researchers.

On 2017 Mar 24, Atanas G. Atanasov commented:

I started regularly drinking kefir recently and really experience positive effects, thanks a lot for putting together this interesting review. I have featured the manuscript at: http://healthandscienceportal.blogspot.com/2017/03/the-health-promoting-and-microbiota.html

On 2017 Jun 01, Lawrence Moon commented:

There is a typo regarding SPSS Syntax in Supplementary Information. prepop should read preop

Thanks to Merrick Strotton for finding it!

On 2016 Nov 05, Lawrence Moon commented:

You may also find this short PDF useful if you have empty cells in your dataset that represent missing data.

https://www.yorksj.ac.uk/media/content-assets/schools/psychological-social-sciences/documents/How-to-enter-missing-data-in-SPSS.pdf

On 2016 Nov 05, Lawrence Moon commented:

I will happily help people learn to analyse their own data using this protocol if you get in touch by email. Best wishes, Lawrence

On 2016 May 20, Donald Forsdyke commented:

FACTS BEFORE COUNTERFACTS

Counterfactual explorations can provide intriguing insights (1). But we have to be sure that the facts themselves are correct in the first place. In the context of Weldon, reference to "bad-tempered conflict with Mendel’s followers," really means conflict with William Bateson. While it is correct that the doctrinaire "Mendelian ‘genes for’ approach is increasingly seen as out of step with twenty-first-century biology" (1), for Bateson (1861-1926) the approach was also seen as out of step with twentieth-century biology.

Well aware of developmental and environmental factors, Bateson recognized that the biochemical characterization of genes should be high on the twentieth century research agenda. Thus, near the end of his life he declared that “Our knowledge of the nature of unorganized matter must first be increased. For a long time we may have to halt” in getting to grips with the underlying biological principles (2). However, he argued forcefully for looking beyond the visible characters of an organism (its conventional phenotype) to what we now regard as its genome phenotype (3). It was here that the answer to Darwin's fundamental question - the origin of species - was likely to lie.

Weldon had allied himself with Pearson whose brilliant work (later built on by Fisher), was to create modern biostatistics. But those were early days and they made elementary mistakes that Bateson was quick to point out. For example, Bateson would have bridled at the idea that “first year biologists” could serve as a reliable “control” against which to compare “second year humanities undergraduates” (1). Yes, we should “study Mendel, but let him be part of his time”(1). And as related by Meijer (4), Mendel followed the statistics of his time. Indeed, his results have withstood the test of time.

(1) Radick G (2016) Teach students the biology of their time. Nature 533:293 Radick G, 2016

(2) Cock AG, Forsdyke DR (2008) "Treasure Your Exceptions." The Science and Life of William Bateson. Springer, New York.

(3) Forsdyke, D. R. (2010) George Romanes, William Bateson, and Darwin's "Weak Point." Notes Rec R Soc Lond 64:139-154.

(4) Meijer OG (1982) The essence of Mendel’s discovery. In: Gregor Mendel and the Foundations of Genetics. Orel V (ed). The Mendelianum of the Moravian Museaum, Brno, pp. 173-200.

On 2016 Oct 04, Atanas G. Atanasov commented:

None

On 2016 Jun 06, Thomy Tonia commented:

We would like to thank Melissa Vaught for her interesting and stimulating thoughts on our paper. She states that the title and description of our trial’s objective is misleading. However, it is clearly stated in the abstract and in the introduction that we randomised IJPH papers into two arms one of which involved IJPH social media exposure at 3 time points; moreover, our Methods section describes this intervention in detail.

Melissa Vaught makes a very valid point stating that we did not include any measures of reach/exposure such as re-tweets, shares or other users’ dissemination of our papers in social media etc. We have touched upon this issue when we discussed the limitations of our study.

She also refers to potential confounding factors. Given the randomised design of our study, however, systematic confounding cannot be an issue; our analysis is what one would call “intention to treat” in clinical trials. There are some interfering factors that we could not control, which we tried to outline in our paper. This is why we standardised our intervention as much as possible, randomising only original articles (and not systematic reviews that generally tend to receive more attention) and stratifying for open access status. We have not issued any press releases for any of the randomised papers but we could not control for any press or other exposure coming from other sources such as authors’ institutions.

Although we agree that SM strategies could influence outcomes, the studies by Fox CS, 2016 and Adams CE, 2016 cited by Melissa Vaught did not directly compare different intensities of interventions such as multiple postings per day versus less frequent postings. Additionally, as we mention in our limitations, we could not have the data on paper views which could be different from the download data.

Melissa Vaught mentions that only the study by Sorenson 2014 looked at SM promotion by the journal. However, the studies by Fox CS, 2015 and Fox CS, 2016 which were randomised and bigger in size also assessed the promotion of papers from Circulation journal on their own social media and found no effect on article views.

We fully agree with the last remark of Melissa Vaught. Indeed, having social media presence as a journal can have benefits that lie beyond downloads and citations and this is one of the reasons our Journal is active on social media.

Indeed, social media exposure and its effects on scientific is a complex and multi-layered subject. Our trial looked at only one specific aspect of this, namely the effects of exposure of our own papers on our own social media channels. Nevertheless, we believe that it adds useful information to the existing body of literature. We hope that future RCTs will assess the effect of other measures of reach and exposure in order to give a fuller picture of the effect.

On 2016 May 24, Melissa Vaught commented:

This interesting article presents the impact of a journal's social media activity on article-level metrics - one of very few randomized trials on the topic. Notably it uses a traditional (citation counts) and an alternative metric (article downloads) as outcomes.

However, the title and descriptions of the trial's objective and intervention are a bit misleading. The authors state, "We sought to investigate whether exposing scientific papers to social media (SM) has an effect on article downloads and citations." But the study does not evaluate social media exposure. Rather, it measures the effect of a journal promoting its content via its own social media channels. The study did not include any measures of reach or exposure, such as retweets, shares, or likes of their posts. It also did not examine whether other users disseminated articles in the trial via social media.

Still, this study addresses a pertinent question: Does publisher promotion of publications via social media influence reading or citation? For IJPH, there is no apparent association. Of course, there are potential confounding factors, some beyond the journal's control. Did IJPH or authors' institutions issue press releases for any publications in the trial? Did any articles garner attention in mainstream media or on social media? Would easier access to publications affect outcomes? The authors report no difference in outcomes for social media promotion vs. control, when stratified for open access; however, there were few open access articles in the trial—10% in control, and 5% in intervention group.

Using citations as an outcome is a valuable contribution of this trial. Other randomized trials have evaluated effects of social media promotion by publishers on article views, which may moderately correlate with citations. In two separate trials (Fox CS, 2015, cited by the authors, and Fox CS, 2016), Circulation found that Facebook and Twitter posts did not significantly affect views at 7 or 30 days. By contrast, the Cochrane Schizophrenia Group (Adams CE, 2016) found a significant increase in 7-day page views for systematic reviews shared on Twitter and Weibo (a popular social media site in China).

Social media strategies could influence outcomes. By using Weibo, Adams CE, 2016 extended reach to a distinct audience. Post frequency and schedules likely affect audience reach as well. For instance, tweets posted at the same time weeks apart (as in the IJPH trial) might not be as effective as multiple posting over one or two days (as in Fox CS, 2016 and Adams CE, 2016). Overall social media engagement and post volume could also come into play.

The authors note that many observational studies show positive correlation between social media exposure and article views, downloads, or citations. They suggest the difference may be due to design (randomized vs. observational). Contexts and sizes of the studies could also contribute to differences in results. Only Sorenson M, 2014 looked at social media promotion by the journal; this was only for a small number of publications (3 promoted and 3 non-promoted). Other cited studies looked at third-party or global social media activity, for a handful to thousands of publications.

Ultimately this trial evaluates whether a journal's social media activity affects reading or citation of its own publications. So who follows journals on social media, and why? Social media uptake in scholarly communities is growing. But it's far from replacing other methods for discovering articles (Ware M, 2015, pp. 52-5, 134-5). Researchers are still likely to read articles recommended by colleagues (Tenopir C, 2015), and we might expect this trend to be reflected in social media. In other words, individuals on social media may have a degree of influence that publishers do not. The benefits of journal social media accounts may lie with other outcomes.

On 2016 Jun 29, David D Leedahl commented:

In the accompanying editorial1, Calkins offers some insightful comments regarding the relatively high rate of stroke or TIA observed in the recent investigation by Gillinov et al.2 We would like to offer two important considerations related to these findings.

First, cardiac surgeons and physicians caring for cardiac surgical patients in the immediate postoperative period have a reluctance to initiate anticoagulant strategies for fear of bleeding complications. Accordingly, bridging strategies with parenteral anticoagulants are infrequently used when initiating warfarin therapy for POAF, potentially resulting in a greater length of time until therapeutic anticoagulation is achieved. Delaying anticoagulant initiation for one to two days to avoid bleeding and another two to three days to become therapeutic may leave patients unprotected for a longer period of time than recognized.

Secondly, using direct-acting oral anticoagulants for the treatment of POAF may improve the time to therapeutic anticoagulation and lower the bar for surgeons to decide to anticoagulate their patients. Although our hospital has some published experience with this approach,3 the effect on stroke or TIA incidence in patients with POAF remains uncertain.

Cornelius Dyke, M.D. Sanford Health, Fargo, ND 58122

David Leedahl, Pharm.D. Sanford Health, Fargo, ND 58122

1. Calkins H. Is Less More for the Treatment of Atrial Fibrillation after Cardiac Surgery? N Engl J Med 2016;374:1977-8.
2. Gillinov AM, Bagiella E, Moskowitz AJ, et al. Rate Control versus Rhythm Control for Atrial Fibrillation after Cardiac Surgery. N Engl J Med 2016;374:1911-21.
3. Anderson E, Johnke K, Leedahl D, Glogoza M, Newman R, Dyke C. Novel oral anticoagulants vs warfarin for the management of postoperative atrial fibrillation: clinical outcomes and cost analysis. Am J Surg 2015;210:1095-102; discussion 102-3.

On 2016 May 20, Mayer Brezis commented:

Did the physical activity increase in the intervention group? Did insulin sensitivity improve? Is it possible that this was the reason for the lack of effect?

On 2016 May 20, Clive Bates commented:

The paper raises well-known fire and explosion risks from lithium-ion batteries, which arise in any products that use them. However, the paper goes on to draw an inappropriate and alarmist conclusion that e-cigarettes constitute a "significant public health risk" on the basis two case studies and four inconclusive media reports.

The problem is that they apply this alarming label without defining what constitutes a significant public health risk, a concept that must depend on the magnitude of the risk involved and how this compares to other risks.

Comparison to smoking-related accidents and injuries

The most important feature of e-cigarettes is that they are alternatives to smoking cigarettes. The authors have not made a comparison with smoking-related accidents or injuries (let alone disease) or made an estimate of the net impact that arises from consumers who switch from smoking to e-cigarette use. The accidents and injuries associated with smoking do actually appear like a "significant public health risk". The U.S. National Fire Protection Service provides useful data:

In 2011, U.S. fire departments responded to an estimated 90,000 smoking-material fires in the U.S., largely unchanged from 90,800 in 2010. These fires resulted in an estimated 540 civilian deaths, 1,640 civilian injuries and $621 million in direct property damage; deaths were down substantially from the year before.

Home structure fires dominated all these measures of loss except for fire incidents. In 2011, an estimated 17,600 smoking-material home structure fires caused 490 civilian deaths (19% of all home structure fire deaths), 1,370 civilian injuries and $516 million in direct property damage. The other 72,400 smoking-material fires in 2011 were mostly outdoor fires (60,200 fires in trash, vegetation and other outdoor combustibles).

Comparison to other risks

Some broader context may be useful - each year there is one significant injury episode for every eight Americans. U.S. data from CDC:

Morbidity

Number of medically attended injury and poisoning episodes in the population: 39.5 million

Episodes per 1,000 population: 126.3

Mortality

Number of injury deaths: 192,945

Deaths per 100,000 population: 60.2

Just about every practice in everyday life involves some risk: cooking, eating, walking, showering, using a laptop, walking up stairs, anything that involves electricity, heat or glass etc.

For example, according to the U.S. National Fire Protection Service home grills account for damage and injuries as follows:

In 2009-2013, U.S. fire departments responded to an average of 8,900 home fires involving grills, hibachis, or barbeques per year. These 8,900 fires caused annual averages of 10 civilian deaths, 160 reported civilian injuries, and $118 million in direct property damage. Almost all the losses resulted from structure fires

How does the e-cigarette compare to the ubiquitous and much-loved BBQ?

There are also persistent reports of exploding iPhones. Should iPhones be designated a "significant public health risk"?

Conclusion

Claims that products or behaviours cause significant public health risks need to be proportionate and reflect both absolute risks (how much harm is caused) and relative risks - i.e. set in appropriate context by reference to risks arising from other common products and activities. Risk is a quantitative concept or it is meaningless.

The authors did not reflect on whether there is a net public health gain in terms of accident risk through the substitution of smoking by e-cigarette use, which is by far the most common pattern of use. It is irresponsible to draw attention to one risk associated with a behaviour, without considering whether it reduces more significant risks. Given the rarity of e-cigarette battery incidents and the preponderance of smoking-related fires, it seems more likely that the net impact on accidents through e-cigarette use will be a significant reduction and therefore a public health win.

This is not a call for complacency but for proportionality. Regulators could have addressed fire risks from e-cigarette batteries some years ago, but have been spending their time on grandiose and excessive regulatory schemes that have yet to achieve anything and may not even solve this problem when they finally apply.

On 2017 Oct 19, Julien Bobe commented:

Thank you for pointing this out. We will add a corrigendum to the original publication.

On 2017 Oct 17, Anne Niknejad commented:

Error in the list of RNA-seq raw sequence data accessions from NCBI SRA:... ', SRP058863 (cave fish), SRP058865 (Atlantic cod), and SRP058863 (surface fish).'

should be read

', SRP058863 (cave fish), SRP058865 (Atlantic cod), and SRP058866 (surface fish).

On 2016 Jul 21, Donald Forsdyke commented:

THEORY-DRIVEN RESEARCH

The results of Enard et al. Enard D, 2016 "draw a broader picture where adaptation against viruses involves not only the specialized antiviral response, but also the entire population of host proteins." Indeed, Petrov has remarked: "Organisms have been living with viruses for billions of years" so on theoretical grounds alone "those interactions have affected every part of the cell."

Given recent disparagement of theoretical work Lander ES, 2016, it is nice to see results that are consistent with theory (e.g. Trends Immunol (2002) 23:575-579; Paper with Endnotes). Enard et al. now "conservatively estimate" that "viruses have driven close to 30% of all adaptive amino acid changes in the part of the human proteome conserved within mammals." Such "virus interacting proteins" vastly exceed the known proteins that regularly engage in immune responses to viruses (e.g. protein kinase R).

This is consistent with the 2002 postulate of the existence of intracellular protein "immune receptors" Forsdyke DR, 2002. Thus, over evolutionary time a protein that primarily evolved for a distinct function, but also happened to cross-react with some virus component, would in addition be selected by virtue of the latter function.

On 2017 Oct 06, Lily Chu commented:

Thank you for the response, Dr. Loy. Many medical resources used by clinicians continue to conflate CFS with SEID and suggest that cognitive behavioral and/or graded exercise therapy are appropriate for patients who fit SEID criteria despite there being no trials of CBT and GET using SEID criteria. I was concerned your group's article might contribute to this confusion.

(In fact, the US Centers for Disease Control and Prevention are working on removing CBT and GET as treatments from their ME/CFS website. See: http://www.npr.org/sections/health-shots/2017/10/02/554369327/for-people-with-chronic-fatigue-syndrome-more-exercise-isnt-better.)

On 2017 Sep 01, Bryan Loy commented:

Thank you for commenting on our paper, and for your contributions to this field. We agree that the ME, CFS, and SEID case definitions are not the same, which is why we provided a citation of each major case-defining paper in the first sentence of our introduction. We then wrote in the “Data searches” section that we were looking for papers using any of the 3 major case definitions, and cited them again. We further acknowledge the differences in the “Diagnostic characteristics” section of the paper, where we coded studies based on the criteria used to diagnose the participants within each study that was included in the meta-analysis. As you point out in table 1, we only found studies that had used the CDC or Fukuda criteria. We wrote ME/CFS/SEID in the title, abstract, and paper to be consistent with the naming used in the IOM report “Beyond Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: Redefining an Illness” and the first sentence in the report brief, “Between 836,000 and 2.5 million Americans suffer from myalgic encephalomyelitis/chronic fatigue syndrome-commonly referred to as ME/CFS” (https://www.nap.edu/resource/19012/MECFS_ReportBrief.pdf). While our conclusions may pertain only to people diagnosed using the CDC or Fukuda criteria, we wanted to still refer to the disease as ME/CFS/SEID since ME and SEID have been recommended as new names for CFS. We hope the results of this initial summary of the existing literature will encourage future research that includes other case definitions, and improves the lives of people living with the disease.

On 2017 Aug 24, Lily Chu commented:

I am one of the co-authors for the SEID (Systemic Exertion Intolerance Disease) case definition. The SEID and ME case definitions are not the same as the CFS case definition. Otherwise we would not have different names for them. The authors of this study seem to think they are equivalent to each other. The title and abstract is misleading: they should be clear that ALL of the studies examined by the authors (in Table 1) used only the 1994 Fukuda CFS case definition. Consequently, any results and conclusions drawn are only about CFS.

On 2017 Aug 30, Hilda Bastian commented:

Although the authors draw conclusions here about cost and effectiveness of simply offering badges if certain criteria are met, the study does not support these claims. There are, for example, no data on costs for the journal, peer reviewers, or authors. Any conclusions about effectiveness are hampered by the study's design, and the lack of consideration and assessment of any potentially negative repercussions.

It was not possible for the authors to study the effects of offering badges alone, as this intervention was part of a complex intervention: a package of 5 co-interventions, announced by the journal in November 2013 to begin taking effect from January 2014 (Eich E, 2014). All were designed to improve research transparency and/or reproducibility, and signaled a major change in editorial policy and practice. Any manuscript accepted for publication after 1 January, while being eligible for these badges, was also subject to additional editorial requirements of authors and reviewers. All authors submitting articles from 2014 faced additional reproducibility-related questions before submission, that included data disclosure assurances. Other authors have shown that although these did not all lead to the changes sought, there was considerable impact on some measures (Giofrè D, 2017).

Data on the impact on submissions, editorial rejections, and length of time until publication of accepted articles is not provided in this paper by Kidwell and colleagues. These would be necessary to gain perspective on the burdens and impact of the intervention package. I had a look at the impact on publications, though. It is clear from the data as collected in this study, and from a more extended timeframe based on analysis of date of e-publication, that the package of interventions appears to have led to a considerable drop in publication of articles (see my blog post, Absolutely Maybe, 2017). The number of articles receiving badges is small. During the year in this study from the awarding of the first badge, it was about 4 articles a month. That first dropped, then rose since, while at the same time the number of publications by Psychological Science has dropped to less than half the rate it was in the year before this package of interventions was introduced, leading to a substantial increase in percentage, while the absolute numbers of compliant articles remains small.

Taken together, it appears that it was likely there was a process of "natural selection", on the side of the journal and authors, leading to more rigorous reporting and sharing among the reduced number of articles reaching publication. The part that badges alone played in this is unknowable. Higher rates of compliance with such standards have been achieved without badges at other journals (see the blog post for examples). There is some data to suggest that disinclination to data disclosure is part of a range of practices adopted together more by some psychology researchers than others, in one of the studies that spurred Psychological Science to introduce these initiatives (<PMID:26173121). The data in Giofrè D, 2017 tend to support the hypothesis that there is a correlation between some of the data disclosure requirements in the co-interventions, and data-sharing (see my follow-up blog post).

In addition to not considering a range of possible effects of the practices, or being able to isolate the impact of one of the set of co-interventions, the study used only one data extractor and coder for each article. This is a particularly critical potential source of bias, as assessors could not be blinded to the journals, and the badging intervention was developed and promoted from within the author group.

It would be useful if the authors could report in more detail what was required for the early screening question of "availability statement, yes or no". Was an explicit data availability statement required here, whether or not there was indeed additional data than was included in the paper and its supplementary materials?

It would be helpful if the authors could confirm the percentage of articles eligible for badges, where the offer of a badge was rejected.

At the heart of this badge approach for closed access journals, is a definition of "open-ness" that enables potentially serious limitation of the methodological information and key explanatory data available outside paywalls. In de-coupling the part of the study included in the paper from the study's data, and allowing it be inaccessible to many who could potentially use it or offer useful critique, the intervention is promoting a limited form of open-ness. The trade-off assumed is that this results in more open-ness than there otherwise would be. However, it may have the reverse effect, for example, by encouraging authors to think fully open access doesn't matter and can be foregone with pride and without concern, and if journals believe this "magic bullet" is an easy way out of more effective intensive intervention.

Disclosures: I have a long relationship with PLOS (which has taken a different approach to increasing openness), including blogging at its Blog Network, and am a user of the Open Science Framework (which is produced by the group promoting the badges). My day job is at NCBI, which maintains literature and data repositories.

This comment was updated with the two references and data on the question of correlation between data disclosure and data sharing on 1 September, after John Sakaluk tweeted the Giofrè paper to me.

On 2016 May 27, Hans Morreau commented:

Massive Chromosomal Loss with Subsequent Whole Genome Doubling is seen in a Wide Variety of Rare Tumor Types:

The authors Zheng et al. performed an impressive molecular characterization of 91 cases of adrenal cortical carcinoma (ACC). The integrated analysis is the way to understand the behaviour of this rare disease far better and hopefully will lead to better treatment options. They conclude that there is a subset of ACCs showing massive chromosomal loss with subsequent whole genome doubling (WGD). The authors state that this chromosomal loss leads to a hypodiploid karyotype. Such a phenomenon “was only matched by chromophobe renal cell carcinoma”. The latter is partly correct. In 2012 we published the occurrence near-haploidisation with or without subsequent endoreduplication (whole-genome doubling) in oncocytic follicular thyroid carcinoma (FTC-OV) and anaplastic thyroid carcinoma (ATC) derived from FTC-OV (Corver et al., 2012). In combined SNP array analysis and DNA content analysis the genomes of these lesions were seen as near-homozygous genomes (NHG) with DNA indices of 0.6-1.4 depending on the absence or presence of endoreduplication. Wagle et al. independently confirmed our observations in the New England Journal of Medicine with the description of one ATC patient who showed a spectacular treatment response upon giving the mTOR inhibitor everolimus. The patient’s ATC derived from FTC-OV and high density SNP analysis clearly identified NHG in the tumor. The phenomenon of NHG with or without endoreduplication is similar to the pattern that is seen by Zheng et al, although the terminology to describe this is slightly different. In 2014 we also showed that in a subset of ACC and parathyroid carcinoma NHG and endoreduplication can be seen, especially in tumors with oncocytic metaplasia (Corver et al., 2014). As seen by Zheng et al the “allelic states” (Corver et al., 2008) in ACC indicated the presence of more chromosomal breakpoints than seen in FTC. In fact similar observations of NHG or widespread chromosomal loss with endoreduplication of the complete genome has been described in peripheral chondrosarcomas (Bovee et al., 2000) and a subset of childhood acute lymphoblastic leukemia (Holmfeldt et al., 2013) and other uncommon cancers (Mandahl et al., 2012). It is intriguing what is eventually responsible for the widespread chromosomal loss with or without endoreduplication. In our model we proposed a stepwise process that might be related to metabolic processes, something that still needs to be proven. It might now be a step forward in understanding the underlying biology of endoreduplication/WGD with the observation of Zheng et al. that TERT expression is higher in the WGD group of ACCs. In conclusion the combined analysis of different tumor types with massive chromosomal loss with subsequent WGD might lead to further insights underlying this remarkable process.

References:

Zheng S, Cherniack AD, Dewal N, Moffitt RA, Danilova L, Murray BA, Lerario AM, Else T, Knijnenburg TA, Ciriello G, et al. (2016). Comprehensive Pan-Genomic Characterization of Adrenocortical Carcinoma. Cancer Cell. 29(5):723-36.

Corver, W. E., Ruano, D., Weijers, K., den Hartog, W. C., van Nieuwenhuizen, M. P., de Miranda, N., van Eijk, R., Middeldorp, A., Jordanova, E. S., Oosting, J., et al. (2012). Genome haploidisation with chromosome 7 retention in oncocytic follicular thyroid carcinoma. PLoS ONE 7, e38287.

Wagle N, Grabiner BC, Van Allen EM, Amin-Mansour A, Taylor-Weiner A, Rosenberg M, Gray N, Barletta JA, Guo Y, Swanson SJ, et al.(2014) Response and acquired resistance to everolimus in anaplastic thyroid cancer. N Engl J Med. 371(15):1426-33.

Corver, W. E., van, W. T., Molenaar, K., Schrumpf, M., van den Akker, B., van, E. R., Ruano, N. D., Oosting, J., and Morreau, H. (2014). Near-haploidization significantly associates with oncocytic adrenocortical, thyroid, and parathyroid tumors but not with mitochondrial DNA mutations. Genes Chromosomes Cancer 53, 833-844.

Corver, W. E., Middeldorp, A., Ter Haar, N. T., Jordanova, E. S., van Puijenbroek, M., van Eijk, R., Cornelisse, C. J., Fleuren, G. J., Morreau, H., Oosting, J., and van Wezel, T. (2008). Genome-wide allelic state analysis on flow-sorted tumor fractions provides an accurate measure of chromosomal aberrations. Cancer Res 68, 10333-10340.

Bovee, J. V., van Royen, M., Bardoel, A. F., Rosenberg, C., Cornelisse, C. J., Cleton-Jansen, A. M., and Hogendoorn, P. C. (2000). Near-haploidy and subsequent polyploidization characterize the progression of peripheral chondrosarcoma. Am J Pathol 157, 1587-1595.

Holmfeldt, L., Wei, L., Diaz-Flores, E., Walsh, M., Zhang, J., Ding, L., Payne-Turner, D., Churchman, M., Andersson, A., Chen, S. C., et al. (2013). The genomic landscape of hypodiploid acute lymphoblastic leukemia. Nat Genet 45, 242-252.

Mandahl, N., Johansson, B., Mertens, F., and Mitelman, F. (2012). Disease-associated patterns of disomic chromosomes in hyperhaploid neoplasms. Genes Chromosomes Cancer 51, 536-544.

Hans Morreau, also on behalf of Willem Corver and Tom van Wezel, Dept of Pathology, Leiden University Medical Center The Netherlands Email: j.morreau@lumc.nl.

Note: This comment was also posted on the website of Cancer Cell attached to the manuscript of Zheng et al 2016. This will however not be visible in the PubMed domain. I do not have conflicts of interest to declare.

On 2016 Jun 21, Dinesh Barupal commented:

Some links to DBs for getting pesticides lists in the supplement file are broken/not working - below are the updated links -

• KEGG Pesticide - https://pubchem.ncbi.nlm.nih.gov/classification/#hid=8
• ChEBI Pesticide - https://www.ncbi.nlm.nih.gov/pccompound/?term=ChEBI[SourceName]+AND+Pesticide
• MESH Pesticide - http://www.ncbi.nlm.nih.gov/pccompound?LinkName=mesh_pccompound&from_uid=68010575
• EPA Pesticide - https://iaspub.epa.gov/apex/pesticides/f?p=CHEMICALSEARCH:50:0
• EU Pesticide - http://ec.europa.eu/food/plant/pesticides/eu-pesticides-database/public/?event=activesubstance.selection&language=EN
• ToxRefDB - https://www.epa.gov/aboutepa/about-national-center-computational-toxicology-ncct

On 2016 Jun 16, thomas samaras commented:

The increase in BMI is certainly puzzling in view of a recent study by Aune et al.(2016) that found a BMI of 20-22 was best for non-smokers over the long term. The study involved a meta-analysis of 230 studies involving 30 million subjects and almost 4 million deaths. When BMI increases, most biological factors get worse, starting for a BMI of less than 21(Lamon-Fava). This includes increases in blood pressure, total cholesterol, LDL, TG and lower HDL. Other studies have found that insulin and IGF-1 increase with BMI. Higher levels of insulin and IGF-1 have been found to be inversely related to longevity. Also, APO-B increases with BMI and is related to a higher risk of heart disease. In addition, APO-A decreases with increasing BMI and this trend is associated with increasing heart disease.

On 2017 Mar 20, Andrew Ibrahim commented:

Here is the visual abstract: https://twitter.com/AnnalsofSurgery/status/768877921406681088

On 2017 Jun 30, Pertti Hakkinen commented:

The authors and readers might be interested in the following publications (these focused on research conducted in the United States):

1) Neil, N., Slovic, P., and Hakkinen, P.J. 1993. Mapping consumer perceptions of risks. Chemical Manufacturers Association, Washington, D.C.

2) Hakkinen, P.J. and Leep, C.J. 1996. Industry’s use of risk, values, perceptions, and ethics in decision making. In: Handbook for environmental risk decision making. Values, perceptions, & ethics. C.R Cothern, ed. CRC Lewis Publishers, Boca Raton, FL.

This is an excerpt from the 1996 publication: “How can risk perceptions of chemical and consumer products be assessed? In 1993, the Chemical Manufacturers Association published a document based on research sponsored by P&G. This document demonstrates how consumer perceptions of risk can be quantified and examined, presents examples of consumer risk perception data for a wide range of consumer products, and discusses the importance of assessing and considering perceptions of risk as part of the development of a product. The 47 household chemical products covered in this document are shown in Table 2. Each of these products was evaluated for the 15 topic areas shown in Table 3.”

On 2016 Jul 05, Tom Kindlon commented:

A (science) blogger has written about why he feels these results might be relevant to people with ME/Chronic Fatigue Syndrome* in this: "The Exercise Intolerance in POTS, ME/CFS and Fibromyalgia Explained?" http://www.healthrising.org/blog/2016/07/04/exercise-intolerance-fibromyalgia-chronic-fatigue-pots-explained/

*The National Academy of Medicine (previously called Institute of Medicine) after reviewing the evidence suggested that Chronic Fatigue Syndrome be renamed Systemic exertion intolerance disease (SEID) last year.

On 2016 Aug 01, Gary Goldman commented:

The study by Kawail et al demonstrates that HZ incidence has increased over a 60 year period from “0.76 per 1000 person-years (PY) (95% confidence interval [CI], .63–.89) in 1945–1949 to 3.15 per 1000 PY (95% CI, 3.04–3.26) in 2000–2007” with a 2.5% per year rate over the time period after adjusting for age and sex (adjusted incidence rate ratio, 1.025 [95% CI, 1.023–1.026]; P < .001). [1] The authors do not provide an explanation for this increase, yet state, “This increase is unlikely to be due to the introduction of varicella vaccination….”

Unfortunately, Kawail et al do not report details of how widespread the uptake of varicella vaccine was in Olmstead County during 2000-2007. In fact, there were still substantial outbreaks of varicella during 2000 and 2001 (with a period of 9 months that the vaccine was unavailable). Minnesota did not require varicella vaccination for students entering kindergarten and 7th grade who lacked proof of having had chickenpox until 2004. In 1999, based on a CDC National Immunization Survey, varicella vaccine coverage was 61.6% among children aged 19-35 in Minnesota. [2] Approximately 50% of children aged less than 10 years must be vaccinated to effectively reduce exogenous boosting to initiate a noticeable increase in HZ incidence. Likely, only during 2003 through 2007 was varicella vaccination sufficiently widespread in Olmstead County to begin influencing (increasing) HZ incidence rates.

It was first hypothesized by Hope-Simpson [3] that “The peculiar age distribution of zoster may in part reflect the frequency with which the different age groups encounter cases of varicella and because of the ensuing boost to their antibody protection have their attacks of zoster postponed” [3]. However, prior to 1999, only limited studies existed that supported this hypothesis. For example, in 1983, Arvin et al. noted a boost in cell-mediated immunity (CMI) in 71% of adults who were exposed to varicella patients in the family [4]. In 1995, Terada et al. reported that Japanese pediatricians aged 50–69 who received multiple VZV exposures, demonstrated HZ incidence rates one-half to one-eighth that of the general population [5]. Gershon et al. in 1996 showed an immunologic boost that reduced the risk of HZ among leukemic children by reexposure to VZV, either by vaccination or by close exposure to varicella [6]. A 1998 study by Solomon found that pediatricians who had a greater incidence of exposure to VZV had lower rates of HZ than psychiatrists who had the lowest VZV exposure rates [7]. More recent studies by Thomas et al. [8] and Salleras et al. [9] have also demonstrated that re-exposure to VZV via contacts with children was associated with reduction in the risk of HZ in adults. In more recent times, during the varicella vaccination era, additional studies, including those derived from the Antelope Valley Varicella Active Surveillance Project (VASP) in a community with widespread varicella vaccination, have emerged that have provided data that validates Hope-Simpson’s hypothesis. [10-12]. In support of the VASP trends [10-11] between 2000 and 2006, a more recent 2013 study by Guzzetta et al suggests that each episode of exposure to VZV increases protection against HZ and that ‘‘this mechanism may be critical in shaping HZ patterns’’. [12]

Thus, while Kawail et al is unable to provide an explanation for the mean 2.5% annual increase in HZ incidence over six decades, one logical hypothesis for the increase is related to societal changes. Living conditions have gradually changed during the past six decades from multiple generations of families residing in a single household, to more independent living as children aged 18 years and over have increasingly moved out of their parent's home and elderly and sedentary individuals are placed in care facilities. Interesting, a Census Bureau reports, “In 2010, there were 40.3 million people aged 65 and above, comprising 13% of the overall population. (This total is 12 times the number it was in 1900, when this group constituted only 4.1% of the population.)” [13] Thus, due to changing trends in society structure and increases in the elderly population, parents and grandparents generally have fewer opportunities for exogenous boosting of their cell-mediated immunity due to reduced contact to children infected with varicella—giving rise to the steady increase in HZ incidence—especially in decades prior to the varicella vaccination era.

In the Antelope Valley VASP, varicella vaccination was widespread in the community with approximately 50% of children aged <10 years vaccinated by 2000. By 2000, exogenous exposures to natural varicella (producing immunologic boosts) were dramatically reduced, especially after a marked decline in varicella seasonality in 1999. After 2 years of active HZ surveillance (2000 and 2001), the number of HZ case reports had maintained or increased in every adult age category except elderly adults aged 70 years and older. Using the paired t-test, there was a statistically significant (p < 0.042; t = 2.95, dF = 4) 28.5% increase in HZ case reports from 158 to 203 from 2000 to 2001 for the population aged 20–69 years. [10] Moreover, the 2007 VASP annual summary to the CDC presents data (not ascertainment corrected) demonstrating a statistically significant increase in HZ incidence rates, from 3.90 to 4.70 per 1,000 p-y in from 2006 to 2007 among adults aged 50 years and older. [11]

Finally, there is a further worrying reason to suspect future increases in HZ incidence. When a child is administered the varicella (or Oka-) strain and then later is exposed to either a child with natural chickenpox or an adult with herpes zoster (almost always due to the reactivation of the natural or wild-type strain), the child harbors two similar but possibly sufficiently heterologous strains--that now are both subject to reactivation. This could potentially double the chances for the reactivation of herpes zoster, or at a minimum, at least increase the chances for reactivation of shingles relative to the pre-varicella vaccination era. [14]

[1] Kawai K, 2016 [2] http://www.cdc.gov/vaccines/imz-managers/coverage/nis/child/index.html [3] HOPE-SIMPSON RE, 1965 [4] Arvin AM, 1983 [5] Terada K, 1995 [6] Gershon AA, 1996 [7] Solomon BA, 1998 [8] Thomas SL, 2002 [9] Salleras M, 2011 [10] Goldman GS, 2013 [11] Goldman GS, 2014 [12] Guzzetta G, 2013 [13] Raphel A. Trends and statistics relating to U.S. seniors, elderly: Census Bureau 2014 report. August 5, 2014. http://journalistsresource.org/studies/society/public-health/trends-statistics-relating-us-seniors-elderly-census-bureau-2014-report [last accessed 07/26/2016] [14] Weinmann S, 2013

On 2017 May 25, Cicely Saunders Institute Journal Club commented:

This paper was discussed on 2.5.17, by students on the KCL Cicely Saunders Institute MSc in Palliative Care

The paper explores the experiences of cleaning staff in a hospital setting on various wards. We thought it was a very interesting and thought-provoking topic which had not been examined in depth before. The study uses a mixed methods approach with qualitative interviewing and focus groups, followed by a questionnaire study for quantitative data.

The study gives an interesting perspective of staff who have regular and prolonged contact with patients, but are little studied in this regard. It prompted discussion regarding the input that non clinical staff can give.

Recruitment for the qualitative part was difficult leading to low numbers and saturation not being reached. The recruited participants were all female and many were non-native speaking.

According to the COREQ criteria, we would have liked more information about who the researchers were, their relationship with the participants, the settings of the interviews and whether this could influence their responses.

It would also have been beneficial to understand how the questionnaire was developed from the themes identified in the qualitative part of the study and of any attempt at validation etc. We were unclear if they meant to use this questionnaire for future research or service implementation, like for instance psychological support to "non-clinical" staff.

We noticed the long time interval between the study taking place (questionnaires were distributed in 2008) and publication, and it would be useful to know the reasons behind this and if there was any follow up research in the intervening years.

Future research would be interesting in other settings: for example in a hospice where patients die more frequently than various hospital areas or among other components of "non clinical" staff, as for instance secretaries or volunteers.

Anna Oriani, Sarah Hanrott and Konstantina Chatziargyriou

On 2016 Nov 22, Alex Vasquez commented:

Correspondence on this article has been published but as of 22 Nov 2016 (more than a month after publication on 19 Oct 2016) has not been appropriately linked; the citation is as follows: Vasquez A. Correspondence regarding Cutshall, Bergstrom, Kalish's "Evaluation of a functional medicine approach to treating fatigue, stress, and digestive issues in women.” Complement Ther Clin Pract. 2016 Oct 19 doi: 10.1016/j.ctcp.2016.10.001 http://www.ctcpjournal.com/article/S1744-3881(16)30107-4/abstract

Critiques of this article are described in the following paragraphs.

1. Description of treatments: The authors state that treatments were “personalized” but provide little data on how or why the treatment was allocated other than to divide patients into two groups for the “adrenal protocols” which are faintly outlined in the footnotes; no dosages are provided as all of the product formulations appear to be proprietary. Table 1 which outlines the “adrenal protocols” shows the treatments to be remarkably similar between the high and low salivary cortisol groups. The authors administered “DHEA drops” and “pregnenolone drops” but the varying dosages of the hormones, administered three times each per day, are not provided. Authors need to provide the actual dosages of all treatments, including hormones.

2. Use of hormonal therapy: The authors need to have described the laboratory methodology. Salivary DHEA levels were found to increase, decrease, and change unreliably, thus rendering these results worthless. Table 5 of the results shows that Cortisol/DHEA ratio started at 5.2 and resulted at 12 at the end of the study. Consistently throughout the medical literature, respectively higher levels of cortisol and lower levels of DHEA are causatively associated with insulin resistance, intra-abdominal obesity, hippocampal atrophy, and bone loss[1,2]; to the contrary of the bulk of the peer-reviewed literature, these authors present these results as beneficial changes. The “DHEA drops” and “pregnenolone drops” are not described either in dosage or formulation. Pregnenolone and DHEA are mood-elevating neurosteroids[3-5]. DHEA administration raises androgens and estrogens[6] which promote cancers[7,8]. The authors should have described appropriate serologic and clinical follow-up, risk considerations, and limits to duration of treatment. By failing to describe the dose of the steroid hormones used, the authors present that these hormones can be used without regard to dosage or duration; this presents a potential hazard by modeling unsafe use.

3. Accurate description of functional medicine: The authors define functional medicine as “The functional medicine model is focused on restoring optimal functioning of 3 body systems: hormonal, digestive, and detoxification.” The authors’ sweeping statement “Restoring these 3 body systems has positive effects on stress, energy, fatigue, digestive issues, and quality of life” has no citations. The definition of functional medicine provided by these authors is discordant with more authoritative descriptions published by the Institute for Functional Medicine[9] and International College of Human Nutrition and Functional Medicine.[10] The authors use the terminology “adrenal and digestive cleanse protocols” without definition, justification, or citation. In scientific publications, unique statements require substantiation and citation.

4. Controlling for external influences: The authors describe the dates of the study as “September 2014 through April 2015” but did not control for vitamin D levels and sunlight exposure which are known to affect mood and cognition[11,12]. The severity of depression, pain, and “functional disorders” has been reported to vary seasonally with exacerbations in fall and winter and alleviation in spring and summer.[13,14] The “lifestyle and nutritional counseling” included a 1-hour in-person coaching session at the start of the study, followed by various telephone contacts and “online group sessions” including “nutrition coaching and follow-up with diet compliance.” Patients may have felt better simply as a result of season change or social contact.

5. Disclosure of commercial interests: The authors deny conflict of interest and that “The study sponsors had no involvement in the study design; in the collection, analysis, and interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.” However, the authors do not disclose the identity of the sponsors. One of the authors (Kalish) is a consultant/speaker for a large distributor of nutritional supplements[15] and also a consultant/speaker/endorser of the laboratory used in this study[16]; these relationships are not disclosed in the article. According to the authors' interview[17] and press release[18] regarding their publication, they provide "training in functional medicine" and seminars related to the treatments used in this study. The authors fail to provide the identity of the study sponsors and the nature of the conflict(s) of interest.

6. Fecal microbiologic testing; antimicrobial treatments: The authors fail to detail the laboratory methodology for the microbiologic testing. Per the footnote describing the mastic product, the authors do not clearly describe the identity nor the ingredients of this treatment.

7. This clinical trial was not registered: The authors make no mention of having registered this clinical trial, and no record of the trial is available at https://clinicaltrials.gov/.

While the functional medicine approach to healthcare is science-based, eclectic, and effective[19], the field is not benefited by poorly conducted research, especially that which does not accurately reflect the practice, which employs poor methodology, and/or which uses steroid hormones without appropriate risk/benefit considerations. This publication is not an accurate representation of the clinical practice of functional medicine. The study design and description present several errors, the methods are not reproducible, the treatments were not sufficiency described, and the study cannot be either validated or refuted scientifically due to design flaws and/or inadequate description of treatments.

[1] Brown et al. Biol Psychiatry. 2004 Jan 1;55:1-9 [2] Ferrari et al. Eur J Endocrinol. 2001 Apr;144:319-29 [3] Brown et al. Neuropsychopharmacology. 2014 Nov;39:2867-73 [4] Fung et al. J Autism Dev Disord. 2014 Nov;44:2971-7 [5] Gaby AR. Alt Med Rev 1996;1:60-69 [6] Arlt et al. J Clin Endocrinol Metab. 1999 Jun;84:2170-6 [7] Folkerd E, Dowsett M. Breast. 2013 Aug;22 Suppl 2:S38-43 [8] Stoll BA. Eur J Clin Nutr. 1999 Oct;53:771-5 [9] Vasquez A. Reprint from Textbook of Functional Medicine: Web-like interconnections of physiological factors. Integrative Medicine 2006 April;5:32-37 [10] Vasquez A. Textbook of Clinical Nutrition and Functional Medicine. ICHNFM.ORG; 2016:134-146 [11] Kent et al. Environ Health. 2009 Jul 28;8:34 [12] Vieth et al. Nutr J. 2004 Jul 19;3:8 [13] Schlager et al. Compr Psychiatry. 1995 Jan-Feb;36:18-24 [14] Schlager et al. Br J Psychiatry. 1993 Sep;163:322-6 [15] emersonecologics.com/drkalish and emersonecologics.com/Webinars-DrKalish. June 10, 2016 [16] "BioHealth Adrenal Testing – Dan Kalish", "Premenopause Hormone Profiles: In Brief – Dan Kalish", and "Functional Lab Testing – Dan Kalish". biohealthlab.com/multimedia-archive/. June 17, 2016. [17] Functional Medicine Plan Treats Fatigue, Stress and Digestive Issues in Women. medicalresearch.com/author-interviews/functional-medicine-plan-treats-fatigue-stress-and-digestive-issues-in-women/25012. June 23, 2016. [18] Research Conducted with Mayo Clinic Practitioners and Kalish Institute Confirms Efficacy of Functional Medicine. businesswire.com/news/home/20160524005553/en/ June 23, 2016. [19] Vasquez A. Inflammation Mastery, 4th Edition. http://www.ICHNFM.ORG 2016 Dr Vasquez, Director ICHNFM and Consultant to Biotics Research

On 2016 Sep 27, James Tsung commented:

Link to Case Videos: https://youtu.be/BnCqMV_ETtM?list=PLHg2xyua_Kjtk47mTO0qK-t0q-8oRLZzs

On 2016 Jun 17, Lydia Maniatis commented:

(This is a response to the abstract only). The authors conclude that perception of distance ratios is relatively accurate, but this result is purely contingent on the stimuli. Perception of relative distances can be more or less accurate. The outcome depends strictly on the structure of the retinal stimulus and the rules of organisation built into the visual system. By choosing a particular physical stimulus to project to the retina, the authors produced relatively accurate outcomes; by choosing a different one, they could have produced very different outcomes. Pictorial images, for example, can obviously produce very inaccurate perception of relative distances. Unless the authors explicitly discuss the relationship between the structure of the environment and the structure of the retinal projection, they haven't added anything to the conversation.

On 2016 Jul 27, Duke RNA Biology Journal Club commented:

This is a summary of a journal club discussion:

This is one of four articles using similar imaging techniques to study translation in living cells published at the same time. These publications add to the growing number of techniques used to image translation such as mature fluorescent proteins Yu J, 2006, TRICK Halstead JM, 2015, and RNA-binding protein/mRNA co-fluorescence Wu B, 2015. The technique presented in this article is similar to the last except that it uses Suntag to image the nascent chain and PP7 aptamers on the mRNA. The colocalization of the two represents active translation in polysomes.

The technique is novel because co-localization is detected as the protein is translated. This brings fluorescent V4-peptide antibodies into concentrated foci at a single point, and can thus be used to follow multiple rounds of translation. Because of this, only detection of the translated protein is needed and indeed, past the first figure, the mRNA fluorescence is not shown. Fast changes in translation can be detected as shown using the ATF4 ORF construct translational response to stress shown in Figure 4 with the possibility of extending the time of tracking to hours by anchoring the mRNA Yan X, 2016 or using fast 3D imaging techniques. One unusual observation the authors made was the vast heterogeneity of transcript translation within a single cell; at any given time only a subset of the transcripts undergo translation and translation rates may vary depending on as yet unknown factors. A related observation is the diffusion of polysomes within the cell: polysomes translating cytosolic transcripts have slower diffusion rates in the perinuclear region of the cell compared to the cytoplasm. This could be due to the restrictive architecture of a membranous area but the exact mechanism remains unknown. A second surprising observation indicates mRNAs that have begun translation and are associated with polysomes can be transported in dendrites, contrary to earlier reports Besse F, 2008. However, the authors cannot detect if translation is temporarily stalled during transport.

While this technique makes substantial findings in the area of single transcript translation behavior, there are limitations. All in all, these images are dots that respond to translation inhibitors, meaning the resolution is not good enough to detect codon resolution and should be coupled with other techniques to verify observations and determine their mechanism. Additionally, since detection of the nascent chain wouldn’t be detected until the majority of the V4 peptides were translated, initiation would be overlooked; however, TRICK is an existing technique for studying the first round of translation.Our main criticism with this technique is the extensive construct engineering that must be performed which raises concerns over disturbing the mRNA and protein functions from both the PP7 aptamers, the Suntag peptides and an ornithine decarboxylase tag to facilitate rapid degradation of the protein. These engineering steps add over 2 kb to the original gene. Additionally, an antibody against the Suntag and a fluorescent PP7 coat protein must be expressed in the cytosol. While the constructs studied did not cause harm to the cell, each construct of interest must be tested individually. Along this line, while there is the possibility to multiplex by changing the aptamer loop or peptide-antibody combination, it would be difficult to multiplex above two individual transcripts. Thus large-scale studies involving individual translation dynamics of mRNA subsets would remain time consuming and technically challenging.

A quick comparison with the three other papers show agreements among all of them Iwasaki S, 2016 however, there is a great opportunity to learn by reading the papers to compare experimental approaches of three groups. We look forward to see what novel findings this technique uncovers as it becomes adopted in different laboratories.

On 2016 Jun 22, Lydia Maniatis commented:

Reading the title of this article, one might get the idea that we learn the 3D structure of objects from 2D views. Looking at it a little more closely, we might experience some confusion as to the distinction being made between “structure” and “shape.” Isn't it a little tautological to say that grasping structure depends on grasping shape? And what do they mean by “format”?

In fact, the idea that we learn 3D structure from 2D views, which the authors want us to take for granted, is false. But isn't it the case that “a large body of research shows that viewpoint-invariant recognition is learned by viewing 2-D examples of an object (Bulthoff, Edelman, & Tarr, 1995; Hayward & Tarr, 1997; Tarr, Williams, Hayward & Gauthier, 1998)”? In fact, the references cited in no way support this bold claim. One of them is irrelevant, and the other two make claims that are highly qualified, tentative, and in need of corroboration. Given that nothing more definite seems to have developed in the two decades or so since these hopeful publications, the relevant body of research seems neither large nor solid. Another claim is also unsupported by its accompanying citation: “It is thought that during learning people use structural information in the 2-D image to infer the 3-D structure of the object (Nakayama, Shimojo, & Silverman, 1989).” In fact, the cited article addresses issues of figure ground segmentation, and not 3-D shape. Its text doesn't even include the terms 3-D, structure, or shape. So the authors are giving readers a false impression that their project is on empirically solid ground.

In fact, the claim that 3-D structure is learned from 2-D views has long been known to be false on logical and empirical grounds. In a more general sense, the idea that we either can or do learn to perceive in 3-D from 2-D experience is not defensible. No one has to learn to see 3-D objects; it happens automatically. “Newborns can recognize...the same rectangle placed at different slants.” (http://nwkpsych.rutgers.edu/~alan/Gilchrist_NN_2003.pdf). Individual 2D views produce 3D percepts; these percepts may or may not be veridical, but knowledge does not affect the basic process. Each view entails its own necessary 3D perceptual response based on shape rules that assume characteristics such as convexity. View “x” will not produce a different shape percept even if we know the one we see is false, just as a 3D-looking flat image does not look flat even though we know it is. We can't learn to see a Necker cube as flat, from no matter how many angles we inspect it.

The idea that we learn 3-D structure also fails for logical reasons. How does a collection of 2D images, a pack of flat cards, in effect, become a - qualitatively different - 3D percept? On what basis is this conversion made? Is each new sample - of the infinite number of possible samples - a new piece of information? Multiple 2D views don't resolve the fundamental ambiguity of the projection/3D percept relationship. The Ames window is not correctly seen despite a complete set of views around an axis. Shape ambiguity is resolved on the basis of rules of organization, not on the basis of collecting ever new views which are themselves ambiguous. (Even when we know a shape is unchanging, the necessary link between individual views and 3D percepts can produce a changing percept, e.g. https://www.youtube.com/watch?v=jRqYkQz0G-g). Furthermore, how or when does the visual system draw the conclusion that a series of projections, such as are shown to subjects in the task chosen by these authors, are projections of the same single and unchanging 3D shape?

Don't the study's subjects exhibit learning of 3D structure form 2D views? If they did it would be a kind of miracle, but the results don't require us to believe in miracles. Subjects' (rather poor) performance in no way requires that their exposure to a series of views of a very strangely-shaped object produce a consistent, learned, 3D percept. Memory for details can do the trick (a single sharper angle, for example, in one of two objects to be discriminated can allow them to eliminate it without having grasped the whole or even part of the 3D shape). And a lot of the time they are just guessing.

A final thing that should be clear to researchers with some knowledge of perceptual phenomena is that the results of this task are totally contingent on the shapes used. The authors here crudely divide objects into novel and familiar. But whether projections of a single, novel object consistently produce the same 3D percept (which, as we saw, applies to some objects viewed by infants) depends on the shape of the object. There is no absolute expectation of constancy. Different, more rational (as far as the expectations of the visual system) shapes would have produced better outcomes than the random ones used here; others may have produced worse. The numbers here are meaningless because the authors haven't analyzed the role of shape.

As for the finding with regard to “format,” it is well-known that both line drawings and chiaroscuro drawings can both produce good 3D shape percepts, and that silhouettes tend to look flat, lacking any indication of relief. If the results of this study had not borne this out, these facts would have remained intact.

In sum, the results of this study can be interpreted as consistent with a false (on the basis of logic and experiment) premise because the task was not designed to distinguish between this (im)possibility and more plausible (and theoretically uninteresting) alternatives.

It would be nice if people doing research in perception spent a little time actually learning the basics.

p.s. In their discussion, Tian et al: "suggest that stereo provides a behavioral benefit only if it resolves ambiguity in the interpretation of the 3-D structure of objects that cannot be resolved from other sources of information."

This is very similar to the titular claim of an earlier article by Pizlo, Li, Steinman (2008): "Binocular disparity only comes into play when everything else fails; a finding with broader implications than one might suppose."

On 2016 Sep 15, Kelly Drew commented:

Thank you for this comprehensive contribution to our understanding of pre-hibernation remodeling in the Syrian hamster. Our studies in the arctic ground squirrel suggest that pre-hibernation remodeling and torpor onset involves increased sensitivity of A1 adenosine receptors with a subsequent inhibition of thermogenesis. See Olson et al., Circannual rhythm in body temperature, torpor, and sensitivity to A₁ adenosine receptor agonist in arctic ground squirrels. J Biol Rhythms. 2013 Jun;28(3):201-7. doi: 10.1177/0748730413490667. PubMed PMID: 23735499; PubMed Central PMCID: PMC4423736.

and Jinka et al., 2011 Season primes the brain in an arctic hibernator to facilitate entrance into torpor mediated by adenosine A(1) receptors. J Neurosci. 2011 Jul 27;31(30):10752-8. doi: 10.1523/JNEUROSCI.1240-11.2011. PubMed PMID: 21795527; PubMed Central PMCID: PMC3325781.

On 2017 Aug 25, Robert Badgett commented:

The website referred to in this letter for calculating the time in therapeutic range for the population of patents at a clinical site is http://qitools.github.io/

On 2016 May 13, thomas samaras commented:

This study certainly refutes the earlier studies reporting that some overweight was healthy. The World Cancer Research Fund Report (2007)recommended a BMI as low as possible within the normal range of 18.4-24.9. Fontana and Hu (2014) recommended a BMI of 20-21. Also, Hosegood found the optimum BMI for women in S. Asia was about 18.5-19.5. In addition, there are many biological parameters that become worse with increasing BMI from the low end of the BMI range. A few are listed below. Most involve undesirable increases in levels of the following parameters: Homocysteine, C-Reactive Protein, Sex Hormone Binding Globulin (higher is better), left ventricular mass, blood pressure, adiponectin (higher is better), mTOR, total cholesterol, low-density lipoprotein, high-density lipoprotein (higher is better), IGF-1, insulin, glucose, ApoB, and ApoA-1 (higher is better).

All of these have been found to increase mortality due to one or more of these causes: CVD, cancer, type 2 diabetes and all-cause mortality. Therefore, it is hard to see how higher levels of BMI have health benefits with the exception of special cases, such as frail, elderly people who lack adequate nutrition, suffer from digestive problems or have a chronic illness.

On 2016 Jun 26, David Smith commented:

All statistical maps derived from the re-analysis with GingerALE 2.3.6 have been uploaded to the NeuroVault repository: http://neurovault.org/collections/1406/

We emphasize that the re-analysis confirmed that all of our key claims are valid.

On 2016 Jul 22, Christopher Tench commented:

Excellent. This is science!

On 2016 Jun 24, David Smith commented:

Thanks for pointing this out. Our original analyses used a conservative threshold, and we have confirmed that our key results hold with a cluster-level FWE of p = 0.05, as implemented in the latest release of GingerALE (v2.3.6).

Within the main text, two clusters failed to replicate in our new analyses. First, the inferior lateral occipital cortex (iLOC) cluster in Figure 3C (amygdala seed with emotion studies) did not pass cluster-level FWE of p = 0.05 (with a cluster-forming thresholds of p = 0.001 or p = 0.005). Second, the dorsolateral prefrontal cortex (DLPFC) cluster in Figure 3B (DLPFC seed with cognitive control studies) did not pass cluster-level FWE of p = 0.05 (with a cluster-forming thresholds of p = 0.001 or p = 0.005). Additionally, our re-analyses failed to confirm the findings in Supplemental Figure 2.

We will reach out to the journal (HBM) to address these issues and indicate that our key claims were not affected by the bug in GingerALE (which was corrected after our article was accepted).

Thanks again for bringing this to our attention.

On 2016 Jun 14, Christopher Tench commented:

This article uses a version of GingerALE (2.3.3) that has a known bug, which produces false positive results. The bug was fixed at 2.3.6. There is no evidence of statistical significance.

On 2016 Jun 21, Andrew W Swartz commented:

Based upon post hoc questionnaire analysis, Shoag et al. claim that the PLCO was confounded by more prostate cancer (PCa) screening in the control group than in the intervention group. This seems highly unlikely because it is incompatible with other more objective outcomes of this trial.

First, the cumulative incidence plots for PCa [1,2] are consistent with comparison of a screened group to an unscreened (or at least less screened) group. During the first four years of screening the intervention group was diagnosed with PCa at a rate 48% greater than the control group [2]. Then, after the intervention period, the rate of prostate cancer diagnosis in the intervention group drops to that of the control group and the lines go parallel, implying similar screening exposure from that point forward. These are the expected plot shapes for screening with overdiagnosis. The Shoag et al. claim directly implies that the group with the overdiagnosis (the intervention group) was the LEAST screened. This seems implausible, as it would be contrary to all screening theory and experience.

Second, the Shoag et al. claim is also incompatible with the previously reported tumor stage-shifting. The intervention group had more favorable staging than the control group [1,3]. Though the magnitude of this effect is smaller in the PLCO than other PCa screening trials, the direction still favors the intervention group. It is highly unlikely that the intervention group would have better tumor staging if the control group had more screening.

The overdiagnosis and tumor stage-shifting are objectively measured outcomes which indicate that there was more screening in the intervention group than the control group. Therefore Shoag et al. have more likely discovered a bias in the questionnaire answers [4,5] than a flaw in the PLCO. We should also consider the possibility that this is simply the product of the “researcher degrees of freedom” which inherently accompany reanalyses[6].

Andrew W. Swartz MD, Emergency and Family Medicine, Yukon-Kuskokwim Health Corporation, Bethel, Alaska

1] Andriole G, Grubb R, Buys S, Chia D, Church T, Fouad M, et al. Mortality results from a randomized prostate-cancer screening trial. New England Journal of Medicine. 2009;360:1310–9. Andriole GL, 2009 Full Text

2] Andriole GL, Crawford DE, et al. Prostate Cancer Screening in the Randomized Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial: Mortality Results after 13 Years of Follow-up. J Natl Cancer Inst. 2012;104:125–132. Andriole GL, 2012 Full Text

3] Ilic D, Neuberger MM, Djulbegovic M, Dahm P. Screening for prostate cancer. Cochrane Database of Systematic Reviews 2013, Issue 1. Art. No.: CD004720. DOI: 10.1002/14651858.CD004720.pub3. Ilic D, 2013

4] Hebert JR, Clemow L, Pbert L, Ockene IS, Ockene JK. Social desirability bias in dietary self-report may compromise the validity of dietary intake measures. Int J Epidemiol. 1995;24(2):389-98. Hebert JR, 1995

5] Adams SA, Matthews CE, Ebbeling CB, et al. The Effect of Social Desirability and Social Approval on Self-Reports of Physical Activity. Am J Epidemiol. 2005;161(4):389-98. doi:10.1093/aje/kwi054. Adams SA, 2005 Full Text

6] Christakis DA, Zimmerman FJ. Rethinking reanalysis. JAMA. 2013;310(23):2499-500. doi: 10.1001/jama.2013.281337. Christakis DA, 2013

On 2016 Jul 17, Lise Bankir commented:

Thank you Dr Pontzer for these interesting comments.

I apologize for saying that you forgot "cooking" since you had indeed mentioned it, briefly. I may have put too much emphasis on cooking.

Actually, I was very impressed by this brillant TED conference. But I am not a specialist of metabolism and energy needs, and I probably did not appreciate well the relative importance of different factors.

On 2016 Jul 05, Herman Pontzer commented:

Thank you for the comment. We haven't forgotten about the importance of cooking: we include it among the critical human adaptations that make more energy available (reference 25, Carmody et al. 2011). In the penultimate paragraph of the main text, we write:

"...the adoption of cooking²⁵ ... effectively increase[s] the net energy gained from foraging, and may have had an essential role in the evolutionary expansion of the hominin energy budget."

However, as we discuss in the paper, increased food energy intake is necessary but not sufficient in accounting for the suite of metabolically costly human traits. To take advantage of the extra calories from cooking (or from other dietary changes that increase the mean calories/gram of food) requires evolved physiological changes to increase the metabolic rate. Previously, the prevailing view of metabolic evolution in humans (and other mammals) assumed that metabolic rates were similar across the hominoids, with no difference in calories/day (accounting for size) across humans and other apes. Our paper shows that humans have undergone an evolutionary increase in metabolic rate that accounts for the (estimated) extra caloric expenditure on brains, reproduction, etc. It will be interesting, in future analyses, to begin parsing the contributions from cooking, dietary changes, and other adaptations to the evolution of human metabolic rate.

On 2016 Jun 24, Lise Bankir commented:

In this interesting paper, the authors explore the mechanisms by which the human lineage has experienced an acceleration in metabolic rate, providing energy for larger brains. I think they forgot an important aspect of human evolution, that is the benefit of COOKING our food. This made food nutrients much more efficient and allowed a better energy supply to our body at a lesser cost. This is very well explained in a TED conference entitled "What is special in the human brain?" by Suzana Herculano-Houzel. Here is the link.

www.ted.com/talks/suzana_herculano_houzel_what_is_so_special_about_the_human_brain.html?utm_medium=on.ted.com-static&utm_campaign=&utm_content=awesm-publisher&utm_source=t.co&awesm=on.ted.com_BrainSoup

On 2016 Oct 24, Sean Ekins commented:

the attached slides bring this article and this effort upto date http://www.slideshare.net/ekinssean/open-zika-presentation

On 2016 Aug 10, Joaquim Radua commented:

Re: the previous comments, please note that under the null hypothesis of no differences between groups, only 1 out of 20 studies should show differences between groups, which is absolutely not the case when randomizing coordinates or blocks of voxels. Random coordinates and similar approaches, which randomize the location of the findings rather than the individuals between groups, are not a valid way to exactly test this hypothesis. Rather, they are only used to yield approximated p-values that, appropriately thresholded, return a map similar but slightly more conservative than that of FWE-corrected p-values in mega-analyses. Voxel-based meta-analytic methods are young and there is room for improvement, but they are based on evidence.

On 2016 Aug 09, Christopher Tench commented:

This is not a result of confusion, but of the definition of statistical inference. Uncorrected p-values do not control the type 1 error rate. A meta-analysis is performed to improve estimates, and is a statistical problem demanding statistical methods. To threshold at an arbitrary p-value controls neither the FDR nor the FWE, so no quantitative evidence that the results are critical of the null hypothesis is available. You cant know if the results are true positives without doing the full experiment, but meta-analysis is used for the case where the full experiment (mega analysis) has not been done. The one, and only, thing that can be done is to make sure that the null hypothesis is appropriately rejected; arguably the whole point of statistical inference. That requires either FWE or FDR control. Using just random coordinates and an uncorrected p-value will produce results that are apparently publishable, but obviously incorrect. Without any estimate of error rate, there is no quantifiable evidence that the results are meaningful.

On 2016 Aug 01, Joaquim Radua commented:

Re: the previous comment, I think there may be some unfortunate confusion. Raw p-values of current voxelwise meta-analyses have not the same meaning as usual p-values because they are not derived from the usual null hypothesis (“there are no differences between groups”), but from another null hypothesis (“all voxels show the same difference between groups”). Thus, up to the moment one of the only ways to "approximately" know if the results of a voxelwise meta-analysis are neither too liberal nor too conservative is to compare them with the results of a mega-analysis of the same data, and that's what it was done.

On 2016 Jun 13, Christopher Tench commented:

The methods employed provide no control over the type 1 error rate. There is no evidence that the results are statistically significant.

On 2017 Jun 12, Xue-hai Liang commented:

The sequence data related to this paper has been deposited to GEO, with the accession number: GSE99658.

On 2016 May 24, Atanas G. Atanasov commented:

Consumer Health Digest Scientific Abstract based on this article is available at: https://www.consumerhealthdigest.com/general-health/obesity-metabolic-inflammation-and-insulin-resistance.html

On 2017 Feb 01, Martine Crasnier-Mednansky commented:

It is astounding that the authors totally ignore the specific effects of cAMP on the lag phase of the glucose-lactose diauxie. Not only does addition of cAMP eliminate the diauxic lag, it also clearly impairs growth on glucose (see figure 1 in Ullmann A, 1968). An increased level of cAMP triggers a 'leaky' expression of CRP-cAMP-dependent genes and operons (including the lactose operon) thereby affecting growth on glucose. Leaky expression of genes reduces fitness in glucose, with a trade-off for a shorter diauxic lag (or, as in figure 1 mentioned above, a complete elimination of the lag phase resulting in biphasic growth). In the Escherichi coli glucose-lactose diauxie, there is a correlation between the cAMP level and the cost-benefit trade-off.

On 2017 Jun 27, Seán Turner commented:

According to the JCM web site, the type strain of Lutibacter profundi is JCM 30586. Strain JCM 30585 is the type strain of "Mariprofundus micogutta" Makita et al. 2017 [PubMed Id 27766355].

On 2016 May 01, Daniel Tsin commented:

We performed appendectomies using a suprapubic port when the vaginal port access was not feasible in : Tsin DA, Colombero LT, Mahmood D, Padouvas J, Manolas P. Operative culdolaparoscopy: a new approach combining operative culdoscopy and minilaparoscopy.J Am Assoc Gynecol Laparosc. 2001 Aug;8(3):438-41. PMID 11509789

We later did a small series of 4 appendectomies and 7 cholecystectomies. http://www.slideshare.net/tsin/minilaparoscopy-as-an-alternative-to-natural-orifice-surgery-7832507

On 2016 Apr 28, Donald Forsdyke commented:

RNAS ENCODING HBZ AND EBNA1 PROTEINS ARE BOTH PURINE-LOADED

The sensitive quantification of HBZ protein levels in various clinical conditions (1) is a major advance. Unlike other HTLV1 proteins, the latency-controlling HBZ protein is encoded by the antisense strand. Thus, whereas the main-gene-encoding ‘top’ strand of the latent virus is pyrimidine-rich, the complementary, ‘bottom,’ strand that encodes HBZ is purine-rich. Likewise, most genes in Epstein-Barr virus are encoded by pyrimidine-rich strands, but the latency-controlling EBNA1 protein is encoded by a purine-rich strand.

The speculation that this purine-loading militates against the formation of double-stranded RNA and hence dampens the host immune response (2) is in keeping with:

(a) the “extremely low expression and immunogenicity of HBZ in natural HTLV-1 infection,”

(b) the possibility that “the most important actions of HBZ, which are critical to HTLV-1 persistence, are exerted at the RNA level, and not the protein level,” and

(c) the view that “minimized HBZ protein translation is a sophisticated viral strategy for evasion from the host T cell response.”

(1) Shiohama et al. (2016) Retrovirology 13:29 Shiohama Y, 2016

(2) Forsdyke DR (2014) Microbes and Infection 16, 96-103 Forsdyke DR, 2014

On 2017 Jul 03, P Jesper Sjöström commented:

Thanks for your interest in our work. I would like to make the following points:

1. We did not actually say 'no connections in mature cortex' -- that quote is certainly not lifted from Mizusaki et al Nat Neurosci 2016. We said "In fact, it was recently reported that, surprisingly, pyramidal cells in visual cortex of mature animals do not seem to interconnect at all, neither bidirectionally nor unidirectionally12," where 12 refers to Jiang et al. We thus say that Jiang et al report that PCs do not seem to interconnect, we do not say that there are no PC-PC connections in mature cortex. What Jiang et al state and what our opinion about that statement is, those are different things.

2. The intention of that passage in Mizusaki et al is to point out that Brunel is using my data from Song et al as a gold standard, but this may or may not be appropriate, since my connectivity data was acquired from a developmental snapshot in time (just after eye opening, typically postnatal day 14-16), whereas Brunel is in fact focussing on the functioning of the mature brain, when circuits are wired up. Our intention was thus to acknowledge that my own data need not be the ground truth, and this has important implications for the validity of the Brunel study. The Tolias study provides an alternative view: "the most compelling and consistent difference across experiments is the age of the animals tested, suggesting that mature cortical circuits are not identical to developing circuits." Such a developmental difference would important in the context of the Brunel study. Again, this is not necessarily my opinion, but as scientists, we have to acknowledge this possibility.

3. In the Tolias study, they report in Fig S14 that they found precisely zero L5 PC-PC connections even after 150 attempts, which is in stark contrast to my connectivity data presented in Song et al. Indeed, if you do a Chi-squared test for 931/8050 versus 0/150, you will find that this is a highly significant difference. We can debate the accuracy of the Tolias measurement (like they do in Barth et al Science 2016 353:1108, as you point out), but if we do so, we should also debate the accuracy of my measurements in juveniles, as presented in Song et al. While it is true that my data in Song et al is more in line with e.g. Thomson et al Cereb Cortex 2002 than with Jiang et al, the key point in the context of Brunel's theoretical study is that the ground truth is not necessarily well established.

In summary, I certainly believe in my own connectivity data set, and I think Brunel's study provides a very compelling theoretical framework for explaining such connectivity patterns, but I feel obliged to point out a few possible caveats associated with my connectivity data set. Jiang et al provide one such key caveat. I hope this clarifies somewhat.

On 2017 Jun 27, Gabriele Scheler commented:

There is a technical comment from Barthetal2016 in Science which sums this up, and gives more detail as to why the connectivity in Xiangetal2015 is underestimated.

On 2017 Jun 27, Gabriele Scheler commented:

Very surprisingly, Mizusaki etal report: " In fact, it was recently reported that, surprisingly, pyramidal cells in visual cortex of mature animals do not seem to interconnect at all, neither bidirectionally nor unidirectionally (12)." The reference (12) is to Xiangetal2015, where they state: "Finally, the connectivity among mature pyramidal neurons, particularly among L5 pyramidal neurons, was much lower than the connectivity among pyramidal neurons within the same range of intersoma distance in juvenile slices [figs. S13B (average, 91 ± 4μm) and S14 and supplementary text]". The percentages noted in the table S14 range from 4,8% to 0% for ~150 connections tested. Quite clearly, the percentages are low, but to summarize this as 'no connections in mature cortex' does not seem to be adequate. Note that the number of connections tested is not very high and that "evoking unitary excitatory or inhibitory postsynaptic potentials [uE(I)PSPs] on postsynaptic neurons with brief depolarizing current pulses applied in presynaptic neurons" may not be sufficient to map all synaptic connections.

On 2017 Sep 08, Youhe Gao commented:

A strategy named 4F-acts was proposed a few years ago trying to minimize false positives and false negatives. Fast Fixation is necessary to study real-time protein-protein interactions under physiological conditions. Fast formaldehyde crosslinking can fix transient and weak protein interactions. With brief exposure to a high concentration of formaldehyde during the crosslinking, the complex is crosslinked only partially, so that the complex is small enough to be resolved by SDS-PAGE, and the uncrosslinked parts of the proteins can be used for identification by shotgun proteomics. Immunoaffinity purification can Fish out complexes that include the proteins of interest. Because the complex is covalently bound, it can be washed as harshly as the antibody-antigen reaction can stand; the weak interactions will remain. Even if the nonspecific binding can persist on the beads or antibody, it will be eliminated at the next step. To Filter out these complexes, SDS-PAGE is used to disrupt non-covalent bonds, thereby eliminating uncrosslinked complexes and simultaneously providing molecular weight information for identification of the complex. The SDS-polyacrylamide gel can then be sliced on the basis of the molecular weight without staining. All the protein complexes can be identified with the sensitivity of mass spectrometry rather than sensitivity of the staining method. The advantages are the following: (i) The method does not involve tagging. (ii) It does not include overexpression. (iii) A weak interaction can be detected because the complexes can be washed as hard as the antigen-antibody reaction can stand as the complexes are crosslinked covalently. No new covalent bond can form as a false positive result. (iv) The formaldehyde crosslinking can be performed at the cellular, tissue, or organ level fast enough so that the protein complexes are fixed in situ in real time. Proteome Science 2014, 12:6 doi:10.1186/1477-5956-12-6

On 2016 Jun 19, Lukas Marti commented:

We thank Dr. Stanley Goldberg for pointing out, that we do not include in our PSP operation a levatorplasty. We therefore do not as he calls it "tackle the perineal hernia". It might be that adding a levatorplasty to perineal procedures as Rehn-Delorms, Altemeier's sigmoidectomy and the described PSP, lowers recurrence rate. But there is no randomized controlled trial to prove that. On the other hand adding a levatorplasty would make the procedure more complex and therefore it would become probably more prone to complications, something we want to avoid in the very frail patients to which we offer a PSP operation.

On 2016 Jun 01, Dr, Stanley Goldberg commented:

This operation does nothing for the perineal hernia which should be attacked in a perineal operation.

On 2016 Aug 10, Joaquim Radua commented:

Re: the previous comments, please note that under the null hypothesis of no differences between groups, only 1 out of 20 studies should show differences between groups, which is absolutely not the case when randomizing coordinates or blocks of voxels. Random coordinates and similar approaches, which randomize the location of the findings rather than the individuals between groups, are not a valid way to exactly test this hypothesis. Rather, they are only used to yield approximated p-values that, appropriately thresholded, return a map similar but slightly more conservative than that of FWE-corrected p-values in mega-analyses. Voxel-based meta-analytic methods are young and there is room for improvement, but they are based on evidence.

On 2016 Aug 09, Christopher Tench commented:

This is not a result of confusion, but of the definition of statistical inference. Uncorrected p-values do not control the type 1 error rate. A meta-analysis is performed to improve estimates, and is a statistical problem demanding statistical methods. To threshold at an arbitrary p-value controls neither the FDR nor the FWE, so no quantitative evidence that the results are critical of the null hypothesis is available. You cant know if the results are true positives without doing the full experiment, but meta-analysis is used for the case where the full experiment (mega analysis) has not been done. The one, and only, thing that can be done is to make sure that the null hypothesis is appropriately rejected; arguably the whole point of statistical inference. That requires either FWE or FDR control. Using just random coordinates and an uncorrected p-value will produce results that are apparently publishable, but obviously incorrect. Without any estimate of error rate, there is no quantifiable evidence that the results are meaningful.

On 2016 Aug 01, Joaquim Radua commented:

Re: the previous comment, I think there may be some unfortunate confusion. Raw p-values of current voxelwise meta-analyses have not the same meaning as usual p-values because they are not derived from the usual null hypothesis (“there are no differences between groups”), but from another null hypothesis (“all voxels show the same difference between groups”). Thus, up to the moment one of the only ways to "approximately" know if the results of a voxelwise meta-analysis are neither too liberal nor too conservative is to compare them with the results of a mega-analysis of the same data, and that's what it was done.

On 2016 Jun 14, Christopher Tench commented:

The method employed here offers no protection against type 1 error rate. It cant be considered a meta-analysis. Uncorrected p-values provide no evidence of statistical significance when large numbers of voxel-wise tests are performed.

On 2017 Jul 19, David Mage commented:

The authors have done a thorough survey of the sudden unexpected infant death (SUID) investigations in several U.S. states. They report on the autopsy and death scene investigations (DSI) of the case infants performed to determine which if any of the three conditions considered to form SUID may be involved: [ICD-10 sudden infant death syndrome (SIDS R95); unknown causes (UNK R99); accidental suffocation and strangulation in bed (ASSB W75)]. However, it is well known that the DSI have an inherent problem if the infant under review has been moved upon discovery by the parents or other caregiver: “cases of sudden death are associated with so much shock to the family that an accurate history is impossible to obtain” (Farber S. NEJM 211 (4):157;1934); “The suddenness of death so stuns the family that attempts to obtain a reliable history on the first visit to the scene are frequently unsuccessful” (Werne J, Garrow I. AJPH 37(June):678;1947); When a possible SIDS infant is immediately picked up on discovery the final resting position is reconstructed from subjective recall, which complicates the DSI (Willinger M, James LS, Catz C., Pediatric Pathology 11: 677-684, 1991). The writer notes that to calculate accurately the reduction of oxygen and the increase in carbon dioxide from possible rebreathing exhalations at the infant’s nares, the distance between the nares and the possible obstruction must be known to the millimeter, that is lost when the infant is picked up on discovery (Mage DT. Forensic Sci Med Pathol 9:2013; 283.) The reader should note that when a SUID infant is picked up on discovery and the DSI is compromised, the ICD-10 code chosen (R95 or R99 or W75) may be incorrect but the totals for SUID remain unchanged.

On 2016 Sep 22, Natalie Clairoux commented:

Free version of this article available after July 1, 2017 at http://hdl.handle.net/1866/13955

On 2016 May 04, Martine Crasnier-Mednansky commented:

There are several discrepancies in this paper which cannot be reconciled. For example, figure 1a and table S1 indicate a growth rate of 0.76 hr^-1 for maltotriose with ammonia as nitrogen source. The glucose-maltotriose diauxie with ammonia (Figure 2a, upper and middle panel) indicates a growth rate for maltotriose of 0.37 hr^-1. In diauxie, the growth rate on each sugar is characteristic of that sugar (Monod, 1942). Therefore, the glucose-maltotriose diauxie should exhibit a growth rate of 0.76 hr^-1 for maltotriose. It is likely that the presence of the reporter gene in the diauxie experiment (figure 2) is affecting the growth rate on maltotriose.

On 2016 Jun 02, Clive Bates commented:

The measured markers are not a reliable proxy for disease risk

The authors are careful not to attribute serious disease risk to these measured changes in vascular function. That caution is well-founded. It is worth pointing out that similar effects are observed in coffee drinkers. See for example:

Papamichael CM, 2005 Effect of coffee on endothelial function in healthy subjects: the role of caffeine

The study examined similar effects as they arise from coffee consumption and concluded:

In conclusion, coffee exerts an acute unfavourable effect on the endothelial function in healthy adults, lasting for at least 1 h after intake. This effect might be attributed to caffeine, given that decaffeinated coffee was not associated with any change in the endothelial performance.

A comparison with coffee would have made a worthwhile additional arm to this study and may have helped with interpreting whether there is a material risk of harm. If the effects are similar to those experienced by coffee drinkers, that might have provided valuable reassurance for smokers making an informed choice about quitting smoking by switching to e-cigarettes.

The role of nicotine

The authors suggest several possible mechanisms for the observed changes in vascular function, including changes induced by nicotine. This is a promising explanatory hypothesis because achieving a satisfactory nicotine dose is common to both smoking and e-cigarette use. One might expect, therefore, similar physiological changes to arise from both ways of taking nicotine if nicotine is the primary cause of these observations.

However, that would also be reassuring to e-cigarette users. Nicotine has well-documented effects on the body but these effects have not been found to be a significant cause of the diseases attributed to smoking, including cardiovascular disease. The effect of nicotine separated from tobacco smoke exposure has been studied through assessments of medical nicotine replacement therapy (discussed on TreatTobacco.net) and snus, a form of smokeless tobacco that delivers high doses of nicotine - see Lee PN, 2013.

The danger of over-interpreting these findings

No one should underplay any potential risks from e-cigarettes. But neither should activists jump to exaggerations about disease risk based on these findings. Overstating the risks of e-cigarettes is logically and ethically equivalent to understating the risks of tobacco smoking. The latter was the practice of tobacco companies 30 years ago and it is essential to challenge the former in today's public health discourse.

As an example, one campaigner has already declared that this study means "e-cigarettes could be half as dangerous as conventional cigarettes" here.

I hope the authors will distance themselves from such extreme over-interpretation of their work. The authors could have pre-empted the need for that by explaining in the discussion that disease is not an inevitable consequence of these changed markers of vascular function.

To be fair, the authors do point to the need for further studies, "to clarify the chronic vascular effects of E-cigarette smoking". To which they might have added "if any" for completeness.

On 2016 Jul 07, GianCarlo Panzica commented:

Further comments on this topic are on Bourguignon JP, 2016 and Kortenkamp A, 2016

On 2016 Aug 22, Anthony Jorm commented:

The Royal Australian and New Zealand College of Psychiatrists (RANZCP) has recently published clinical practice guidelines on schizophrenia Galletly C, 2016, eating disorders Hay P, 2014 and mood disorders Malhi GS, 2015. These guidelines contain a mixture of evidence-based recommendations where there were relevant intervention studies, and consensus-based recommendations where relevant studies did not exist. The consensus-based recommendations comprised a substantial proportion of the guidelines for mood disorders (59%) and schizophrenia (46%), but less so for eating disorders (10%), indicating that expert consensus is an important source of guidance on best practice in psychiatry.

Given the substantial contribution of expert consensus to these guidelines, it is important that the methods for establishing this consensus are adequate. The Australian National Health and Medical Research Council (NHMRC) has published requirements for development of clinical practice guidelines, but these do not give much guidance on how this should be done, simply mandating that “The method used to arrive at consensus-based recommendations or points (e.g. voting or formal methods, such as Delphi) is documented”. (National Health and Medical Research Council. Procedures and requirements for meeting the 2011 NHMRC standard for clinical practice guidelines. Melbourne: National Health and Medical Research Council; 2011.)

Another potential source of criteria for evaluating the quality of methods for developing consensus-based recommendations comes from research on ‘wisdom of crowds’ Lorenz J, 2011 Kattan MW, 2016 Baumeister RF, 2016. Based on such research, Surowiecki has proposed four conditions necessary for a group to make good decisions (Surowiecki J. The wisdom of crowds: why the many are smarter than the few. London: Abacus; 2004.): 1. Diversity of expertise. A heterogeneous group of experts will produce better quality decisions than a homogeneous one. For guidelines developers, this may mean that the experts should come from a range of relevant disciplines, including consumer experts where appropriate. 2. Independence. The experts must be able to make their decisions independently, so that they are not influenced by others. For guidelines developers, this means that voting on consensus-based recommendations is carried out privately so that strong individuals cannot dominate the group. 3. Decentralization. Expertise is held by autonomous individuals working in a decentralized way. For guidelines developers, it is important to specify what sources of information the experts had available to them. 4. Aggregation. There is a mechanism for coordinating and aggregating the group’s expertise. For guideline developers, this could involve an independent person who runs the voting and gives feedback to the group.

If we take these four conditions as appropriate for judging the quality of methods for developing consensus-based recommendations, how well do the RANZCP guidelines meet them?

An indication of diversity of expertise is the disciplinary composition of the guideline working groups. There was limited diversity for all working groups, with non-psychiatrists comprising 3 of the 8 members for eating disorders, 4 out of the 12 members for mood disorders and 2 out of the 10 members for schizophrenia working group. There were no consumer or carer members of any of the working groups. While the mood disorder and schizophrenia guidelines included consensus-based recommendations s for indigenous peoples, it is not stated whether any of the working groups included indigenous members.

The mood disorders and schizophrenia working groups did not appear to involve independent decision making. Both groups had discussions until consensus was reached. The eating disorders guidelines did not give relevant information about whether there was independence.

All three guidelines state that consensus-based recommendations were based on collective clinical and research knowledge and experience. The eating disorder guidelines additionally state that level IV articles were considered where higher-level evidence was lacking and this informed the consensus-based recommendations.

After drafting, all guidelines had input from a broader group of expert advisers with a wide diversity of expertise. However, it is not clear whether these advisers had the potential to persuade working group members to change consensus-based recommendations.

Where the guidelines included consensus-based recommendations relevant to indigenous peoples, it is not clear what sources of cultural expertise these were based on.

None of the guidelines state how judgements were aggregated to determine consensus. The mood disorders guidelines state that agreement on consensus-based recommendations was “in most cases unanimous but allowed one committee member to abstain”. The other guidelines did not define what constituted consensus.

In conclusion, there are major weaknesses in the procedures used to determine consensus-based recommendations for all three guidelines. These are lack of independence in decision making by experts, a lack of a formal mechanism for aggregating judgments, and a lack of diversity of expertise, particular in areas where consumers and carers could contribute and where cultural expertise is relevant.

While NHMRC gives quite detailed guidance on how to develop evidence-based recommendations, there is little guidance on best practice for developing consensus-based recommendations. While many of these weaknesses would be overcome by using formal consensus methods such as the Delphi process, there is a need for NHMRC and similar agencies to produce more rigorous quality standards for development of consensus-based recommendations.

On 2016 Sep 25, Lydia Maniatis commented:

I would respectfully ask the authors to justify their fundamental assumption, as this is straightforwardly asserted in their opening sentence:

“Visual sensitivity is limited by both the strength of the neural signals, and the noise in the visual nervous system1 (Levi DM, Klein SA, Chen I. 2005).”

I am interested in the reference to “the noise in the visual nervous system.” What is the basis of the claim that “noise in the visual nervous system" affects the percepts generated under the conditions of this study? What is the nature of this noise, where does it occur, and what is the evidence for it?

The single reference provided for the claim doesn’t provide answers as it, too, takes the it for granted. It does provide two other supporting references; but on inspection, these are equally inadequate.

According to Levi et al (2005) :

“It has been recognized for well over a century that visual perception is limited by both the strength of the neural signals, and by the noise in the visual nervous system…The notion that internal noise in the visual system acts like light, even in the absence of a stimulus, i.e., dark light, led Barlow (1957) to formulate a very influential model of visual detection which posits that visual sensitivity is limited “by the difficulty of distinguishing a weak signal from the background of spurious signals, or ‘noise’, which occurs without any light signal at all”. Barlow (1957) quantified the dark noise by determining the amount of actual light that produced the same amount of noise in the eye that is present in the dark, using the now widely used prescription for quantifying the noise, known as the equivalent input noise technique (Pelli, 1990).”

To clarify, Barlow (1957) is suggesting that there are “low levels of intrinsic retinal noise even in complete darkness.” He is interested in absolute thresholds. He states that his results “fit the theoretical predictions for the case of stimuli of short duration and small area (against a background of large area) for a range of background intensities up to about 10 scotopic trohinds [i.e. very, very low intensities].”

Levi, Klein and Chen (2005) however, are not referring to absolute thresholds or to retinal noise. They tell us that Pelli (1990); Pelli (1981) provided a new description of the equivalent input noise as a contrast function.

The theoretical rationale provided by Pelli (1990) is as follows:

Thus [in the case of electronic amplifiers] we can measure [“specify” would be more accurate] the intrinsic noise by finding an equivalent input noise. This is sometimes called 'referring the amplifier's noise to its input'. Essentially the same approach can be applied to vision. Indeed, this is analogous to Barlow's (1957) dark light measurements. By a similar analogy we can apply this idea to the contrast domain.”

More casual assumptions follow as Pelli (1990) elaborates the application of this analogy to the human visual system.

I submit that vague analogies should not be admitted the status of theoretical arguments.

It should be noted that Zomet et al (2016) in no way test the casually adopted and elaborated assumption that I am challenging. They simply interpret data as though it were true; their procedure can’t tell if it’s false. (The claim is, in fact, so vague that it’s not clear how one would go about testing it).

Postscript: The criteria for publication in Scientific Reports are that: "In this journal a paper is not assessed based on its perceived importance, significance or impact, but it is enough to be just scientifically valid." Testable hypotheses, and testing of assumptions prior to applying them as though they were true, is a fundamental criterion of scientific validity, which I believe is absent in the present case.

On 2016 Sep 08, Andrea Messori commented:

Active rheumatoid arthritis with inadequate response to methotrexate monotherapy: incremental benefit for infliximab plus methotrexate versus methotrexate alone

Andrea Messori, Sabrina Trippoli HTA Unit, ESTAR, Regional Health Service, Firenze, 50100 Italy

One strength of the meta-analysis by Hazlewood and co-workers [1] is that all pharmacological interventions aimed at rheumatoid arthritis have been evaluated in terms of efficacy. Since infliximab biosimilar has recently become available and its cost is lower than that of infliximab originator, evaluating the magnitude of the incremental benefit for infliximab plus methotrexate versus methotrexate alone is of interest.

In patients with active rheumatoid arthritis with inadequate response to methotrexate monotherapy, Hazlewood and co-workers [1] report a total of 5 randomised studies.

According to the end-point of ACR50 response at 30 weeks, the results found in these trials are shown in Table 1. The heterogeneity assessment (carried out according to Higgins and Thompson [2]) showed an I squared of 0% (with p=0.71).

If one analyses these data based on traditional pairwise meta-analysis (OMA software, Open Meta-Analyst, version 4.16.12, Tufts University, url http://tuftscaes.org/open_meta/), the pooled risk difference in favour of infliximab plus methotrexate vs methotrexate alone is +19.7% (95% confidence interval: +15.4% to +24.1%; fixed-effect model).

References

1. Hazlewood GS, Barnabe C, Tomlinson G, Marshall D, Devoe D, Bombardier C. Methotrexate monotherapy and methotrexate combination therapy with traditional and biologic disease modifying antirheumatic drugs for rheumatoid arthritis: abridged Cochrane systematic review and network meta-analysis. BMJ. 2016 Apr 21;353:i1777.

2. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002 Jun 15;21(11):1539-58.

Table 1. Data of ACR50 response reported in 5 randomized trials: comparison between Remicade plus methotrexate (R+MTX) vs methotrexate alone (MTX). Detailed references for the 5 trials are given in the Appendix.

ACR50 response

Abe 2006: R+MTX=15/49 (30.6%) vs MTX=4/47 (8.5%)

Lipsky et al 2000 (ATTRACT): R+MTX=18/86 (20.9%) vs MTX=7/88 (8.0%)

Mac Isaac 2014: R+MTX=6/30 (20.0%) vs MTX=0/31 (0.0%)

Westovens et al 2006 (START): R+MTX=110/360(30.6%) vs MTX=33/361 (9.1%)

Zhang 2006: R+MTX=38/87 (43.7%) vs MTX=22/86 (25.6%)

Overall crude rate R+MTX=187/612 (30.5%) vs MTX=66/613 (10.8%)

Appendix

-Abe T, Takeuchi T, Miyasaka N, et al. A multicenter, double-blind, randomized, placebo controlled trial of infliximab combined with low dose methotrexate in Japanese patients with rheumatoid arthritis. J Rheumatol 2006;33:37-44

-Lipsky PE, Van Der Heijde DMFM, St. Clair EW, et al. Infliximab and methotrexate in the treatment of rheumatoid arthritis. New Engl J Med 2000;343:1594-602

-Mac Isaac KD, Baumgartner R, Kang J, et al. Pre-treatment whole blood gene expression is associated with 14-week response assessed by dynamic contrast enhanced magnetic resonance imaging in infliximab-treated rheumatoid arthritis patients. PLoS ONE 2014;9 (12) (no pagination)

-Westhovens R, Yocum D, Han J, et al. The safety of infliximab, combined with background treatments, among patients with rheumatoid arthritis and various comorbidities: a large, randomized, placebo-controlled trial.[Erratum appears in Arthritis Rheum. 2007 May;56(5):1675 Note: Dosage error in article text]. Arthritis Rheum 2006;54:1075-86

-Zhang FC, Hou Y, Huang F, et al. Infliximab versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: A preliminary study from China. APLAR Journal of Rheumatology 2006;9:127-30

On 2016 Apr 27, Ana Herrmann commented:

The authors state that 1) "animals were used for only one experiment and in a single behavioral test, to reduce interference from apparatus exposure", and later they state that 2) "Animals were allowed to freely explore the novel tank for 6 min, after which they were removed from it and exposed to the scototaxis tank". Which one was it?

On 2016 May 24, Stefan Hofmann commented:

See: http://www.ncbi.nlm.nih.gov/pubmed/27009055

On 2016 May 06, Margaret Sampson commented:

The initial posting for this article contained an error in the order of the authors, as did the article posted on the JCE website on 18 April 2016. The correct author order for "The Single-Case Reporting guideline In BEhavioural Interventions (SCRIBE) 2016 Statement" is as follows: Tate, Perdices, Rosenkoetter, Shadish, Vohra, Barlow, Horner, Kazdin, Kratochwill, McDonald, Sampson, Shamseer, Togher, Albin, Backman, Douglas, Evans, Gast, Manolov, Mitchell, Nickels, Nikles, Ownsworth, Rose, Schmid, and Wilson. As a check, please ensure that Vohra is the 5th author (Posted May 6, 2016, revised May 19, 2016)

On 2017 Mar 02, Melissa Rethlefsen commented:

I thank the authors for including the full version of their search strategies in the appendix. This is a valuable tool for readers to more comprehensively analyze the systematic review's methodology and outcomes.

In the appendix containing the full search strategies, the MEDLINE search and the Embase search appear identical. It is unclear if these databases were searched simultaneously using Ovid's multifile searching capability, or whether the search listed for MEDLINE is in fact not the search used for the MEDLINE search. It is of course also possible that the same search was used for each database, though searched individually.

The primary complication with the MEDLINE search as reported is that many of the search terms used are Embase-specific Emtree terms that are not searchable in MEDLINE. This includes: "crossover procedure"; "double blind procedure"; "single blind procedure"; "triple blind procedure"; "animal"; "nonhuman"; "human"; "human cell"; and "ulcerative colitis." Because these are not the correct MeSH terms for these concepts, searches for these terms as indicated would produce no results. This unfortunately impacts two major components of the search, from the recommended Cochrane Embase sensitive search for randomized control trials (lines 1-18) and the disease state section (lines 21-24). Multifile Ovid searching does try to map thesaurus terms to corresponding thesaurus terms across databases, which may resolve some of these issues. However, a best practice would be to include both Emtree and MeSH terms in a search instead of relying on multifile term mapping. If the authors did rely on multifile searching, it would be appropriate to note this in the methods section to alert readers.

In addition, lines 26 and 27 both search for the MeSH heading "Mesalamine," and lines 28 and 29 both search for the MeSH heading "Sulfasalazine." This was perhaps adapted from the previous version of this review (Feagan BG, 2012), which also utilizes this duplication. It is not clear why these terms were duplicated in either case, though this version of the review did update the rest of the search strategy to be more comprehensive and also more specific.

On 2016 Aug 18, Andrew Kewley commented:

Expression of concern over un-interpretable statistics of multiple candidate gene association studies.

The lack of replication of candidate gene association studies has led to much discussion and debate. It is generally considered that each candidate cannot be considered to be an independent hypothesis, hence the alpha level needs to be corrected for multiple comparisons. There is no discussion of this in the paper.

This research group have published multiple candidate gene manuscripts in addition to this study [1,2,3,4] (Most of which are not listed on pub-med), and may have other unpublished data. Hence it is difficult to know what alpha should be used for significance, after correcting for multiple comparisons. If correcting for 678 comparisons, then none of the results reach significance and the overall conclusion should be quite different.

In general, future genetic studies should utilise genome-wide association methodology to avoid such biases.

[1] Examination of Single Nucleotide Polymorphisms (SNPs) in Transient Receptor Potential (TRP) Ion Channels in Chronic Fatigue Syndrome Patients. Sonya M. Marshall-Gradisnik, Peter Smith, Ekua W. Brenu, Bernd Nilius, Sandra B. Ramos and Donald R. Staines Immunology and Immunogenetics Insights 2015:7 1-6

[2] Genotype Frequencies of Transient Receptor Potential Melastatin M3 Ion Channels and Acetylcholine Muscarinic M3 Receptor Gene Polymorphisms in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis Patients. Marshall-Gradisnik, SM; Chacko, A; Johnston, S; Smith, P; Nilius, B; et al. Immunology and Immunogenetics Insights 8 (2016): 1-2.

[3] Examination of Single Nucleotide Polymorphisms in Acetylcholine Receptors in Chronic Fatigue Syndrome Patients. Marshall-Gradisnik, Sonya; Smith, Peter; Nilius, Bernd; Staines, Donald R. Immunology and Immunogenetics Insights 7 (2015): 7-20.

[4] Killer Cell Immunoglobulin-like Receptor Genotype and Haplotype Investigation of Natural Killer Cells from an Australian Population of Chronic Fatigue Syndrome/Myalgic Encephalomyelitis Patients. T. K. Huth, E. W. Brenu, D. R. Staines, and S. M. Marshall-Gradisnik. Gene Regulation and Systems Biology 2016:10 43-49

On 2016 Aug 24, Ben Goldacre commented:

This trial has the wrong trial registry ID associated with it on PubMed: both in the XML on PubMed, and in the originating journal article. The ID given is NCT0255810. We believe the correct ID, which we have found by hand searching, is NCT02558101.

This comment is being posted as part of the OpenTrials.net project^[1] , an open database threading together all publicly accessible documents and data on each trial, globally. In the course of creating the database, and matching documents and data sources about trials from different locations, we have identified various anomalies in datasets such as PubMed, and in published papers. Alongside documenting the prevalence of problems, we are also attempting to correct these errors and anomalies wherever possible, by feeding back to the originators. We have corrected this data in the OpenTrials.net database; we hope that this trial’s text and metadata can also be corrected at source, in PubMed and in the accompanying paper.

Many thanks,

Jessica Fleminger, Ben Goldacre*

[1] Goldacre, B., Gray, J., 2016. OpenTrials: towards a collaborative open database of all available information on all clinical trials. Trials 17. doi:10.1186/s13063-016-1290-8 PMID: 27056367

* Dr Ben Goldacre BA MA MSc MBBS MRCPsych<br> Senior Clinical Research Fellow<br> ben.goldacre@phc.ox.ac.uk<br> www.ebmDataLab.net<br> Centre for Evidence Based Medicine<br> Department of Primary Care Health Sciences<br> University of Oxford<br> Radcliffe Observatory Quarter<br> Woodstock Road<br> Oxford OX2 6GG

On 2017 Mar 29, Zvi Herzig commented:

Despite the fact that this study relates to e-cigarette (EC) vapor dissolved in a liquid medium, the authors surprisingly attribute outcomes to fine particulate matter:

… our findings suggest it is not nicotine or toxic combustion product that activates the hemostatic system, but instead the hemostatic system is most responsive to fine particulate matter

This conclusion is puzzling because particulate matter of EC aerosol are in fact liquid droplets. When these are merged with the liquid medium, their former particle sizes become irrelevant. (Solid particles emitted from EC hardware are at quantities below safety limits Farsalinos KE, 2015 and thus unlikely to be of significance).

It is also unclear what the causes or the relevance of these in vitro platelet effects are. For example, propylene glycol and glycerol are not rapidly metabolized here as they are in vivo. Indeed, Hom et al write:

There are some important limitations to the work presented here, including the use of an ex vivo static system, without negative hemostatic regulators (such as endothelial cells), to evaluate the effects of e-vapor on platelet functions.

In contrast, a recent clinical study of smokers switching to Tobacco Heating System 2.2 (whose emissions also include the limited cigarette smoke toxicants present in EC vapor Schaller JP, 2016) shows a level of reduction in platelet activation among switchers approaching those of quitters Lüdicke F, 2018:

Reductions in 8-epi-prostaglandin F2α (biomarker of oxidative stress), 11-dehydro-thromboxane B2 (biomarker of platelet activation)… occurred in the menthol tobacco heating (mTHS) system group compared with the menthol conventional cigarette group. The changes in the mTHS group approached those in the smoking abstinence group.

This would indicate that the outcomes of Hom et al are not manifested in real use. While it can be countered that perhaps EC flavorants not emitted by heated tobacco have induced the platelet effects, this is unlikely considering that outcomes were similar for two disparate EC products tested.

It would also be interesting to consider the effects of air bubbles present from bubbling the aerosol through the extraction buffer. It's well-known that air bubbles induce platelet activation Sandgren P, 2011 and this might help explain the strange outcome.

On 2016 May 23, Jonathan Heald commented:

Posted on behalf of the American Academy of Sleep Medicine:

These recommendations for the use of adaptive servo-ventilation for the treatment of congestive heart failure related central sleep apnea are updates to the 2012 recommendations found in the guideline "The treatment of central sleep apnea syndromes in adults: practice parameters with an evidence-based literature review and meta-analyses." These recommendations serve as an essential supplement to the 2012 practice parameter document:

Aurora RN, 2012

On 2016 Jun 06, Christopher Southan commented:

Since the molecular structure of HM71224 is not disclosed the journal has allowed the publication of work that cannot be reproduced. This consequently obfuscates the clinical trial entry from Eli Lilly

https://clinicaltrials.gov/ct2/show/NCT02628028

On 2017 Sep 13, Seán Turner commented:

According to the CCM catalog of strains, the type of Paenibacillus cucumis Kampfer et al. 2016 is CCM 8655, not CCM 8653. The authors have later assigned CCM 8653 as the type of Isoptericola cucumis Kampfer et al. 2016 (PMID 27045419) which is consistent with the assignment in the CCM catalog of strains.

On 2017 Sep 13, Seán Turner commented:

None

On 2016 Jul 14, Marcus Munafò commented:

This study provides deeper theoretical insights than might be imagined from a cursory reading. Evolutionary life history theory addresses how organisms, including humans, vary in the allocation of resources to growth, survival and reproduction across the life course. Life history strategies are defined by key decisions that trade off finite resources against competing demands to achieve reproductive goals [1, 2]. Both across and within species, the trade-off between the allocation of resources to growth and to reproductive efforts results in organisms employing varying strategies that can be characterised, loosely, as ‘slow’ or ‘fast’. A ‘slow’ life history strategy is characterized by later maturity and proportionally greater investment of resources in a smaller number of offspring, while a ‘fast’ life history strategy involves more effort directed towards reproduction, such as earlier puberty and sexual activity [2, 3]. Typically, animals pursue a fast life history strategy (‘live fast, die young’) under adverse environmental conditions of high extrinsic mortality. In such conditions, as Day and colleagues suggest, earlier sexual maturation may increase reproductive fitness. However, Day and colleagues also suggest that associations between earlier puberty and a greater propensity for risk-taking behaviours, lower educational attainment, and adverse health outcomes, contradicts this evolutionary perspective. On the contrary, directing effort into risky and aggressive behaviour is best considered as an important part of a fast life history strategy, in which competition for mates is crucial [1]. Viewed in this light, adolescent behaviours of unprotected sex and earlier pregnancy, as well as violence, law breaking, and substance abuse, can be seen as central components or consequences of a fast strategy in which the future is discounted relative to the present [1, 4]. Variation in these physiological and behavioural traits in response to changing environments may reflect a suite of adaptations – psychological and somatic – that increase fitness on average despite poor social and health consequences [5], as resources are diverted from longevity towards short-term reproductive goals under conditions of adversity.

The human life history literature suggests strategies are contingent on variation in the environment rather than variations in genotype. Day and colleagues illustrate the central role genetics can play in testing key hypotheses arising out of life history theory, where reliance on observational data has previously limited causal inference. The key point is that variants identified in GWAS studies may reflect environmental (and hence modifiable) risk factors, as well as direct genetic effects [6]. An ever-increasing range of social and psychological traits relevant to life history theory are proving tractable to GWAS, including, for example, social deprivation [7] and endocrine biomarkers including cortisol [8] and sex steroids [9]. Genetic instruments associated with social outcomes could be used as a proxy for environmental harshness, while instruments for endocrine status may provide causal evidence for status seeking and reproductive behaviours known to occur in other species, but that are not amenable to experimental analysis in humans. The results of emerging GWAS therefore provide us with unprecedented opportunities for testing the predictions of life history theory, with implications for our understanding of the determinants of health and social behaviour. The elegant demonstration by Day and colleagues of causal links between earlier puberty and both key reproductive traits and adverse outcomes provides compelling evidence that evolutionary life history theory can shed light on human reproductive behaviour, health and disease, even in a modern Western environment. Combining causal analysis from genetic epidemiology and theoretical insights from evolutionary models may help to launch a science of evolutionary social epidemiology, combining knowledge of ‘how’ exposures lead to certain outcomes with an understanding of ‘why’ these relationships may exist. Reframing our understanding of the development of significant social and health outcomes in this way may lead to novel insights to inform future interventions.

Rebecca Lawn, Abigail Fraser, Marcus Munafò and Ian Penton-Voak

1. Ellis, B.J. and Bjorklund, D.F. Beyond mental health: an evolutionary analysis of development under risky and supportive environmental conditions: an introduction to the special section. Developmental Psychology, 2012. 48(3): p. 591.
2. Chisholm, J.S., Death, hope and sex: Steps to an evolutionary ecology of mind and morality. 1999, Cambridge: University of Cambridge Press.
3. Ellis, B.J., et al., Fundamental dimensions of environmental risk. Human Nature, 2009. 20: p. 204-268.
4. Simpson, J.A., et al., Evolution, stress, and sensitive periods: the influence of unpredictability in early versus late childhood on sex and risky behavior. Developmental psychology, 2012. 48: p. 674-686.
5. Gluckman, P.D. and Beedle, A.S. Match fitness: development, evolution, and behavior: comment on Frankenhuis and Del Giudice (2012). Developmental Psychology, 2012. 48: p. 643– 646.
6. Gage, S.H., et al., G= E: What GWAS can tell us about the environment. PLoS Genet, 2016. 12: e1005765.
7. Hill, W.D., et al., Molecular genetic contributions to social deprivation and household income in UK Biobank (n= 112,151). bioRxiv, 2016.
8. Bolton, J.L., et al., Genome wide association identifies common variants at the SERPINA6/SERPINA1 locus influencing plasma cortisol and corticosteroid binding globulin. PLoS Genet, 2014. 10: e1004474
9. Vandenput, L. and C. Ohlsson, Genome-wide association studies on serum sex steroid levels. Molecular and cellular endocrinology, 2014. 382: p. 758-766.

On 2016 May 15, Arnaud Chiolero MD PhD commented:

In a life course epidemiology perspective, this is a great review to understand strengths and limitations of intergenerational approach of prevention of obesity and related conditions. In this domain, the words are elegant and the concepts are appealing; the evidence is however weak, letting us short of knowing how to act. We need trials to go further.

On 2017 Mar 02, Simina Boca commented:

Code for analyzing the processed data available at https://github.com/SiminaB/DMD-metabolomics

On 2016 Jul 26, Fernando Castro-Chavez commented:

Dear Reader, This is my most recent article and in it I present the match between the Vigesimal Numerical System of the Maya Culture from the Yucatan Peninsula, in Mexico, and the twenty amino acids with suggested systems of anatomical mnemonics in the Education of Molecular Biology, i.e., using the twenty appendages of our extremities (fingers and toes), and another to remember the purines and pyrimidines with hands and feet. Furthermore, by the suggestion of my chemistry tutoring student Ana Baleva, I have also explored the Icosahedron as an alternate way to represent the twenty amino acids for the Graphics Design in Bioinformatics, containing in one of its triangular cells, an on/off switch. To see the rest of a suggested set of correspondences, please go to my Facebook page and (if you wish) befriend me: https://www.facebook.com/fernando.castrochavez.90. I deeply appreciate feedback, and/or constructive comments, for this and for The Rest of my: Publications in PubMed, by Fernando Castro-Chavez:. And my LinkedIn: https://www.linkedin.com/in/fernando-castro-chavez-6744a322. Sincerely, Fernando Castro-Chavez. P.D. An updated image inspired by this research:

The First Ten plus Zero Numerals of the Vigesimal Mayan System

On 2016 Jun 29, Fernando Castro-Chavez commented:

None

On 2017 Nov 30, Joe G. Gitchell commented:

Readers may find the follow-up letter to this article (http://ajph.aphapublications.org/doi/full/10.2105/AJPH.2016.303456) and the authors' reply to the letter (http://ajph.aphapublications.org/doi/full/10.2105/AJPH.2016.303476) of interest.

On 2016 Apr 25, Dita Gratzinger commented:

Dr. Itkin and colleagues provide a tour de force overview of hematopoeitic stem and progenitor cell maintenance and trafficking within the context of distinctive capillary/arteriolar and sinusoidal vascular beds in the mouse. In parallel to their findings in mouse, we have previously described the analogous vascular beds in intact human marrow, particularly the nestin+ capillary/arteriolar bed, which is tightly wrapped in SMA/CD146+ pericytes/vascular smooth muscle which in turn is wrapped in CD271+ mesenchymal stromal cells (MSCs), as distinct from the nestin-negative sinusoidal endothelium in direct contact with MSCs; the MSCs are then in direct contact with CD34+ hematopoietic progenitor or stem cells Flores-Figueroa E, 2012. CD105 is a specific sinusoidal vascular marker and nestin as a specific capillary/arteriolar marker in intact human bone marrow, allowing for quantification of the distinct vascular beds in the setting of myelodysplastic syndromes Ewalt MD, 2016.

On 2016 Apr 14, Fernando Suarez-Obando commented:

Great paper

On 2016 Dec 15, Victoria MacBean commented:

Plain English Summary:

Parasternal intercostal electromyography (EMGpara) is a new way to measure breathing difficulty. Research needs to be carried out because body parts used in breathing, like the lungs, need to be properly checked over for breathing problems to be managed, but the testing methods aren’t always suitable for children who are very young or ill. The parasternal intercostal muscles are muscles that move at the same time as the diaphragm (a thin sheet of muscle under the lungs) when you breathe in and out. EMGpara measures signals from the brain which are sent to these muscles without putting any instruments into the body, so it is ideal for children. EMGpara was measured using stickers on the front of the chest while the participants (92 healthy, 20 wheezy and 25 with a machine (ventilator) to help them breathe) were breathing in and out in a resting state. For the wheezy children, measurements were taken before and after a substance to widen air passages (reliever inhaler, or bronchodilator) was used; for the critically ill children, these were taken during ventilator-assisted breathing, then with just mild air pressure to keep the airways open (continuous positive airways pressure). It was found that as age, weight and height increased, EMGpara decreased. This is because when children are growing up, big changes take place in the respiratory system, decreasing the effort needed for breathing. EMGpara in the healthy children was the lowest; in the wheezy children it was higher before the bronchodilator was used, dropping to similar levels to the healthy children afterwards. In the critically ill children, EMGpara was higher than in the wheezy children when the ventilator was used, and even higher with continuous positive airways pressure when they were having to breathe without support. This study has shown that measuring EMGpara is possible in children of a range of ages and levels of health. The results from the healthy children have shown important age-related changes in EMGpara, and those from the wheezy and critically ill children have shown that EMGpara is affected by changes in how hard the breathing muscles have to work because of different diseases and treatments. EMGpara could be a really helpful method in testing the breathing ability of patients who are usually difficult to assess.

This summary was produced by Lottricia Millett, Year 12 student from Burntwood School, London, as part of the investigators' educational outreach programme.

On 2016 Aug 31, Joanna Hudson commented:

We thank Ben Goldacre and the opentrials.net project for bringing this to our attention. We have informed the journal of this anomaly and will ensure all future publications are cited with the correct trial registry ID.

On 2016 Aug 24, Ben Goldacre commented:

This trial has the wrong trial registry ID associated with it on PubMed: both in the XML on PubMed, and in the originating journal article. The ID given is NCT023528702. We believe the correct ID, which we have found by hand searching, is NCT02352870.

This comment is being posted as part of the OpenTrials.net project^[1] , an open database threading together all publicly accessible documents and data on each trial, globally. In the course of creating the database, and matching documents and data sources about trials from different locations, we have identified various anomalies in datasets such as PubMed, and in published papers. Alongside documenting the prevalence of problems, we are also attempting to correct these errors and anomalies wherever possible, by feeding back to the originators. We have corrected this data in the OpenTrials.net database; we hope that this trial’s text and metadata can also be corrected at source, in PubMed and in the accompanying paper.

Many thanks,

Jessica Fleminger, Ben Goldacre*

[1] Goldacre, B., Gray, J., 2016. OpenTrials: towards a collaborative open database of all available information on all clinical trials. Trials 17. doi:10.1186/s13063-016-1290-8 PMID: 27056367

* Dr Ben Goldacre BA MA MSc MBBS MRCPsych<br> Senior Clinical Research Fellow<br> ben.goldacre@phc.ox.ac.uk<br> www.ebmDataLab.net<br> Centre for Evidence Based Medicine<br> Department of Primary Care Health Sciences<br> University of Oxford<br> Radcliffe Observatory Quarter<br> Woodstock Road<br> Oxford OX2 6GG

On 2016 May 12, Heidi Schulz commented:

Note that according to HGVS guidelines, the novel variant described as c.717delG in the paper should be referred to as c.718del p.Val240*.

On 2016 Apr 20, Anderson Santos commented:

Dear users of the MED pipeline, The MED web server address has been migrated to this new one: http://med.compbio.sdu.dk/ in the Denmark. The MED web server administrator is negotiating a web address forward but no responses from Saarland till now. An important clue about MED: keep the proteins names small and simple, avoid punctuation characters or your processing can fail. I hope to correct such weakness in a further version. Best regards, Anderson Santos - Author

On 2016 Apr 30, Clive Bates commented:

Missing the Point

The priorities elaborated in Matthew Myers' editorial are misplaced will have only trivial impact on the cigarette market. I have proposed five more fundamental tobacco-related regulatory issues for FDA or Congress to address. Please see this e-letter response to the journal Tobacco Control: e-letter Missing the point

An expanded version of this is available with links in this posting The tobacco control high command has lost its way - what we learn from its views on FDA priorities

On 2016 May 27, Carmine Pariante commented:

We are grateful to Dr. Carroll for his comment, which has allowed us to clarify this point. Dr. Carroll is right that we use 'specific' as indicating the outcome of statistical analyses. However, we refer to pathways and genes 'specifically' regulated in depressed patients (or, indeed, in the group that does not develop depression) because these genes/pathways are only regulated in the depressed (or in the non-depressed) group. In this sense, we are using 'specific' to differentiate these genes/pathways from the many genes/pathways that are regulated equally in both groups. We also agree that there are limitations in using IFN-alpha as a model of ordinary major depressive disorder, although we are confident that our findings are relevant to at least the subgroup of depressed patients with high levels of peripheral inflammation. We hope that this reply is helpful.

On 2016 May 08, Bernard Carroll commented:

Several issues arise with this report. 1. Authors claimed that "specific signatures" of gene expression changes distinguished depressed versus non-depressed patients. After reading the full paper plus the Supplemental material it is clear that specificity was not established. No data were presented on conditional probabilities or on diagnostic confidence or prognostic confidence. 2. Multiple, misleading, references to specificity in the article actually just denote statistically significant differences in gene expression between the 2 groups of patients. 3.No group-wise differences in cytokine concentrations were found. 4. Validity of IFN-alpha induced depression as a model of ordinary major depressive disorder has not been established.

On 2016 Jul 20, Rohan Maddamsetti commented:

While gene flow is certainly important in the evolution of gut microbiota, given current evidence I do not believe that gene flow is a good explanation for synonymous variation in E. coli. I've written a short post on my blog explaining my thinking: http://rohanmaddamsetti.weebly.com/blog/a-retraction-of-sorts

On 2017 Apr 27, Randi Pechacek commented:

Katherine Dahlhausen wrote an engaging blog post on microBEnet about this review. She explains how this review emphasizes our "knowledge gaps in the understanding of the fate of antibiotics in our environments."

On 2016 Apr 13, Anders von Heijne commented:

Recommended reading - Note the important aspect of including the scale/lexicon as a footnote in all radiology reports!!

On 2016 Apr 16, Aiguo Ren commented:

The lowest and highest point estimates of NTD prevalence were obtained from two studies in China. However, these two studies are not comparable in terms of geographical area, time period, and, most importantly, method of calculation. The estimate from Beijing was obtained from a maternity hospital and the numerator was the number of NTD cases in pregnancies of 28 weeks of gestation or greater. In Beijing, every pregnant woman receives ultrasound scanning for fetal structural defects around 20 weeks of gestation, and virtually all fetuses with NTDs will be detected by 28 gestational weeks, and terminated based on informed choice of the mother. This explains, in a larger part, the low NTD prevalence in Beijing. Other national or hospital-bases data from China suffer from the same problem. Please see our most recent report PMID: 26879384.

On 2016 Jun 22, Vatsalya Vatsalya commented:

Research highlight at NIAAA June 2016 Council Meeting: http://www.niaaa.nih.gov/about-niaaa/our-work/advisory-council/directors-reports-council/niaaa-directors-report-institute-12

On 2016 Aug 01, Joaquim Radua commented:

Re: the previous comment, I think there may be some unfortunate confusion. Raw p-values of current voxelwise meta-analyses have not the same meaning as usual p-values because they are not derived from the usual null hypothesis (“there are no differences between groups”), but from another null hypothesis (“all voxels show the same difference between groups”). Thus, up to the moment one of the only ways to "approximately" know if the results of a voxelwise meta-analysis are neither too liberal nor too conservative is to compare them with the results of a mega-analysis of the same data, and that's what it was done.

On 2016 Jul 22, Christopher Tench commented:

Control in statistics is not achieved by fixed p-value threshold when there are many tests. There are very fundamental reasons for using FWE and FDR, and as a minimum these must be estimated to quantify the risk of false positives. Applying a fixed p-value and fixed cluster extent is a bit like having a fixed steering wheel in a car: it will work really well when the road is straight, but fail terribly when there is a bend! Controlled methods necessarily adapt themselves to the data. Without them, there is no quantifiable evidence that an effect is truly critical of the null.

I appreciate the validation study. Unfortunately it only tested known positive data. In statistics this cant be considered validation as it promotes confirmation bias. I am aware of the claim that the threshold is equivalent to corrected 0.05, but unfortunately this is not correct. Correction implies adaptation to the data and its null, but fixed thresholds are not able to do this.

On 2016 Jun 21, Fabio Richlan commented:

1) Whether or not a study can be considered a meta-analysis does not depend on the use of a correction for multiple comparisons. 2) As usual for coordinate-based meta-analyses, statistical significance was assessed by a permutation test (thresholded at a voxel-level (height) of p < 0.005 and a cluster-level (extent) of 10 voxels). Note that the use of a cluster extent threshold is a way of controlling false positives. In addition, only voxels with z > 1.0 were considered statistically significant. Based on an empirical validation, it was shown that, at least for seed-based d Mapping (SDM), this combination of voxel-level and cluster-level thresholds optimally balances sensitivity and specificity and corresponds to a corrected p-value of 0.05 (http://www.ncbi.nlm.nih.gov/pubmed/21658917). More details on the SDM method can be found at http://www.sdmproject.com/

On 2016 Jun 15, Christopher Tench commented:

The method employed in this study offers no control over the 10⁵ statistical tests performed. The results are can not be considered statistically significant or a meta-analysis.

On 2017 Apr 09, Harri Hemila commented:

A manuscript version is available at HDL.

On 2016 Apr 30, Morten Oksvold commented:

Please note that this publication is actually a retraction of the Leukemia article. There is no indication in this PubMed notice that this is a retraction.

I will strongly suggest that all retractions in the future are written clearly in the beginning of the article title.