On 2016 Apr 09, Lydia Maniatis commented:

Turning a blind eye to the well-established role of stimulus structure in all aspects (including thresholds, salience, etc) of perception is a good method for proliferating either reliably falsifiable generalisations or generalisations that can't be challenged because they're so vague.

In other words a. results are contingent on the stimulus and a different stimulus could well lead to different results, and b. words like "weak target" "highly visible stimuli" describe the percept not the stimulus, and so avoid specifying conditions in physical terms.

Anyway, what is interesting about "perisaccadic compression"?

On 2016 Dec 18, Laura Williams commented:

Our online journal club discussed this paper on May 18, 2016. https://www.youtube.com/watch?v=N5x4t0eRts8

On 2016 Aug 18, Peter Hajek commented:

Cigarette smoke killed the cells within 24 hours, e-cig vapour damaged them after extended exposure over up to 8 weeks. The abstract ignores this. The authors of this study are Yu et al., the authors listed above are not responsible for this paper, they in fact published a critique explaining the problems in it and in the accompanying media release - see below.

On 2016 May 19, Clive Bates commented:

For clarity, it is important to point out that the abstract above is drawn from a different paper Yu V, 2016. It is not the abstract for this PubMed entry.

This PubMed entry actually refers to a highly critical commentary on the Yu V, 2016 paper and the abstract shown above. Please use the full-text link at the top right of this page or access the commentary here:

Holliday R, Kist R, Bauld L. Commentary on Yu et al, Evidence-Based Dentistry (2016) 17, 2–3. doi:10.1038/sj.ebd.6401143

On 2017 Sep 23, Karim Abbas commented:

I see the data in Figure 3 confusing and might involve a bias. Figure 1 clearly depicts a stable PS-LTP recorded for 15 minutes every day for successive 14 days. However, in Figure 3b the magnitude of PS-LTP with anisomycin (but without "reactivation" (Sic1)) looks decreasing from the values between 0-3 days, albeit non significantly (I guess the lack of significant differences was due to high SD; the values of SD were not reported in the article but can be seen in the figures). There is no explanation though for that tendency of decreasing PS magnitude compared to stable values depicted in Figure 2. More intriguing is why the recording interval following "reactivation" was restricted to 2 days (total 5 days), which is much shorter than the one depicted in Figure 1 (14 days)?

On 2016 Sep 19, Saurabh Gupta commented:

This article discusses which of the vascular channels may represent fifth arch artery based on current knowledge of embryogenesis.

On 2016 Sep 19, Saurabh Gupta commented:

Fifth arch artery, one of the most controversial topics in the embryology of cardiovascular system.

On 2016 May 09, M Mangan commented:

An interesting discussion of this study can be found here: http://www.marklynas.org/2016/04/deformed-gmo-franken-butterflies-not-fast/

Edit to add: a response from the Rothamsted researchers can be found here as well: <http://rothamsted.ac.uk/sites/default/files/Comment on Hixson et al. 2016_Deformed butterfly wings when feeding on artificial diets containing EPA and DHA.pdf>

On 2016 Apr 19, S. Hong Lee commented:

Many thanks for your comments. The two models (response variables) do not simply represent association between genome-wide markers, previously associated with schizophrenia, and woman’s age at first birth. The two models (response variables) are the functions derived from the relationship between the schizophrenia risk in children and their mother’s age, studied in a large national-wide study (please see reference #11 in the paper). So what we tested was that the relationship between the schizophrenia risk in children and their mother’s age found from the previous (totally independent) study was associated with the relationship between schizophrenia risk profile score (inferred from SNP effects estimated in another independent SCZ GWAS data set) and age at first birth. Although we used SNP effects to infer risk profile score, we did not think the test should be corrected for the number of SNPs. We used a response variable and one set of risk profile score (N x 1) in a single test (not multiple sets of risk profile score).

In addition, I would like to explain a bit more about the variance explained by the predictor in the target data set (R2 in Table 2). The variance explained by the predictor in the target data set is not the actual variance of the underlying effects. It means it can increase more when there is less sampling error (i.e. with a larger samples size). For example, if you estimate SNP effects in a discovery (or training) data set, and the estimated SNP effects are projected into an independent target data set, the R2 explained by the predictor (from the estimated SNPs effects) depends on the estimation error of the SNP effects. So, if the sample size in the discovery data set (SCZ GWAS in our case) is increased, the R2 can be increased (hence lower p value).

On 2016 Mar 31, Anastasia Levchenko commented:

Reading the paper, I did not see a clear mention that the calculated p-values (Table 2) were corrected for multiple testing (and if they were, it is not clear which statistical method was used). For instance, “statistically significant” association between a genome-wide marker and a phenotype is represented by a p-value ≤ 5E-08, if the Bonferroni correction is used. Don’t the two models (response variables), used in the present study, represent association between genome-wide markers, previously associated with schizophrenia, and woman’s age at first birth? If they do, then it is not clear based on what grounds the authors claim that the p-value = 4.1E-04 is “statistically significant” in the present study setting. Near the end of Discussion the authors state that “A caveat of our findings is that the genetic association between risk of SCZ and delayed AFB was only marginally significant given the number of analyses performed, perhaps reflecting smaller sample size and correspondingly larger standard errors for women with delayed AFB, and so require replication in a larger sample”, probably referring to the p-value = 1.4E-02. Again, the reader sees no support for the statement that p = 1.4E-02 is “significant”, although “marginally”.

On 2016 Mar 29, Ludovic Barault commented:

Hi, People who are interesting in this research area should also read some of the recent studies (from our lab and others) currently unreferenced in the above manuscript:

• Tumor MGMT promoter hypermethylation changes over time limit temozolomide efficacy in a phase II trial for metastatic colorectal cancer. PMID: 26916096

• Dose-Dense Temozolomide in Patients with MGMT-Silenced Chemorefractory Colorectal Cancer. PMID: 26538496

• Digital PCR quantification of MGMT methylation refines prediction of clinical benefit from alkylating agents in glioblastoma and metastatic colorectal cancer. PMID: 26113646

• Activity of temozolomide in patients with advanced chemorefractory colorectal cancer and MGMT promoter methylation. PMID: 24379162

• A phase II study of temozolomide in patients with advanced aerodigestive tract and colorectal cancers and methylation of the O6-methylguanine-DNA methyltransferase promoter. PMID: 23443801

• Promoter CpG island hypermethylation of the DNA repair enzyme MGMT predicts clinical response to dacarbazine in a phase II study for metastatic colorectal cancer. PMID: 23422094

• Works from B. Kaina's lab

Sincerely,

On 2016 Apr 04, Donald Forsdyke commented:

COLLECTIVE GENE FUNCTIONS SHOULD BE TAKEN INTO ACCOUNT

Gene products have both individual and collective functions (e.g. the Donnan equilibrium; 1). Wang et al. suggest that female susceptibility to autoimmune diseases may reflect incomplete dosage-compensation, which results in overexpression of certain X chromosome-located, immunity-related, genes (2). However, they refer to studies by Scofield and colleagues [ref. 11] on diseases with changes in numbers of entire X chromosomes (e.g. XXY), indicating biallelic expression of many genes. Scofield has noted that collective, as well as specific, gene functions must be considered (3). This is in keeping the hypothesis that the cytosolic aggregation pressure exerted by the protein collective is immunologically important (4, 5). <br>

1.Loeb J (1921) Donnan equilibrium and the physical properties of proteins. 1. Membrane potentials. J Gen Physiol. 3:667-90. Loeb J, 1921<br>

2.Wang J et al. (2016) Unusual maintenance of X chromosome inactivation predisposes female lymphocytes for increased expression from the inactive X. Proc Natl Acad Sci USA doi/10.1073/pnas.1520113113 Wang J, 2016<br>

3.Dillon SP et al. (2012) Sex chromosome aneuploidies among men with systemic lupus erythematosus. J Autoimmun 38:J129-J134.Dillon SP, 2012<br>

4.Forsdyke DR (2009) X chromosome reactivation perturbs intracellular self/not-self discrimination. Imm Cell Biol (2009) 87:525-528. Forsdyke DR, 2009<br>

5.Forsdyke DR (2012) Ohno's hypothesis and Muller's paradox: sex chromosome dosage compensation may serve collective gene functions. BioEssays 34:930-933. Forsdyke DR, 2012

On 2017 Feb 17, Arturo Martí-Carvajal commented:

Trial authors did not report randomization process. Randomization is oly mentioned into abstract. Where are inclusion and exclusion criteria? This "RCT" is not clear.

On 2016 May 23, Maurizio Battino commented:

The post released by AG Atanasov, revealed the wide interest that CoQ has arisen in health debates. He explained, in a a few words and for a public which is not professionally involved in health care, where, how and why CoQ has a critical/strategic role in our physiology and well-being. When I began to investigate CoQ antioxidant properties in health and disease in addition to the well-known role in the mitochondrial respiratory chain of this molecule, about 25 years ago, I was considered one of the pioneer in the field and several criticisms arose from reviewers (e.g., it took more than 2 years to publish DOI: 10.1016/0003-2670(91)80070-A and DOI: 10.1016/0098-2997(94)90016-7 and also DOI: 10.1024/0300-9831.69.4.243 suffered important delays). 25 years later, CoQ is widely recognized as an active compound with important features which go beyond its primitive redox role and also those antioxidant properties my group proposed at the beginning. Nowadays, we have clear proofs of its involvement in several pathologies and we directly found evidences it has, among others, key roles in aging (doi: 10.1093/gerona/glv063, doi: 10.1016/j.freeradbiomed.2011.02.004, doi: 10.1016/j.mad.2009.11.004, DOI: 10.1023/A:1023754305218), in periodontal diseases (doi: 10.1016/j.phrs.2014.10.007, doi: 10.1016/j.fct.2009.06.026, doi: 10.1016/j.numecd.2009.03.003), in fibromialgia (doi: 10.1089/ars.2013.5198, doi: 10.1089/ars.2013.5260), in Papillon-Lefevre Syndrome (DOI:10.1080/1071576031000083116) and in modulating the effects of nutrients as outlined by this very last review (doi: 10.3390/molecules21030373).

On 2016 May 12, Atanas G. Atanasov commented:

Consumer Health Digest Scientific Abstract of this article is available at: https://www.consumerhealthdigest.com/anti-aging/coenzyme-q-role-in-dietary-therapy-against-aging.html

On 2016 Apr 22, Toby Gibson commented:

Dr. Gack, Thank you for taking the time to comment. One interesting way in which NS3 could potentially regulate 14-3-3 epsilon would be to bind across the top of the groove. Such a binding mode could act as a gatekeeper for phosphopeptide entry/exit into the deep binding pocket. This presupposes that the interaction is direct and strong enough and for this, bacterially expressed proteins need to be assessed in vitro for their dissociation constant.

Incidentally, regarding the work that you are building upon (Liu et al., 2012, PMID:22607805), there was no attempt to identify a candidate motif. But again, RIG-I is also unlikely to have a 14-3-3 phosphopeptide binding mode. RIG-I is almost entirely structured with a few short interdomain linkers. I don’t see good candidates for canonical 14-3-3 binding motifs in RIG-I. So, in this whole system, in my opinion there are no strong clues toward elucidating the molecular details of how 14-3-3 epsilon might be interacting. Assuming that this interaction also validates in vitro, there is some useful structural biology to be done.

On 2016 Apr 19, Michaela Gack commented:

Dr. Gibson’s comment about our manuscript solely stated that the viral RxEP motif identified in the Dengue NS3 protein does not fulfill all of the criteria of conventional 14-3-3-binding motifs found in cellular 14-3-3-interacting proteins. The main conclusions of our study, that the NS3 protein of Dengue virus binds to 14-3-3epsilon and thereby antagonizes RIG-I translocation to mitochondria and type I interferon induction, were never questioned by Dr. Gibson.

Using mapping experiments and site-directed mutagenesis, we identified that the 64-RxEP-67 motif in the Dengue NS3 protein is essential for 14-3-3epsilon binding. Mutation of the central Glu66 (E66) residue to positively-charged Lys66 (K66) markedly reduced the binding of NS3 to 14-3-3epsilon. Additional mutation of Arg64 to Lys64 (termed ‘NS3(KIKP)’ mutant) led to a nearly complete loss of binding to 14-3-3epsilon. These data together with the similarity of the viral RxEP motif with the cellular 14-3-3-binding motif Rxx(pS/pT)xP, where pS/pT indicates a phosphorylated Ser/Thr residue, suggested that the negatively charged E66 serves as a phosphomimetic residue. Although our study indicated that the RxEP motif is required for 14-3-3epsilon binding, our data also suggested that additional as-yet-unidentified residues in the Dengue NS3 protein are important for 14-3-3epsilon interaction, as introduction of the RxEP motif into the NS3 protein of Yellow Fever virus, which is unable to bind 14-3-3epsilon, did not result in gain of 14-3-3epsilon binding. Thus, the full characteristics of the binding mode of Dengue NS3 and 14-3-3epsilon and their binding affinity remain to be determined in future studies, as well as additional residues in NS3 and 14-3-3epsilon that are engaged in the interaction.

On 2016 Apr 08, Toby Gibson commented:

Is there an RxEP motif in Dengue NS3 protein?

This comment addresses the plausibility that the reported RxEP motif in Dengue NS3 is a genuine 14-3-3 binding motif and as such a target for rational design of therapeutics. It does not address other aspects of this published work, in particular that the NS3 protease plays a role in antagonising the cellular relocation and antiviral activity of RIG-I.

Viral mimicry of cellular protein motifs is very common and frequently provides a useful insight into the cellular mechanisms that are being disrupted (Davey et al., 2011). Given our interest in these interactions, I was keen to read this new viral motif paper. Given the current importance of flaviviral infections in human health, it then seemed desirable to post here the explanation as to why RxEP cannot be a phosphomimetic for 14-3-3 proteins. It is my hope that this comment will be of value to the Dengue research field e.g. in how follow up experiments might be planned.

There are several molecular structural issues relevant to 14-3-3 / phosphopeptide interactions:

Proline is disallowed at the +1 position. There is no ambiguity about this. Phosphopeptide arrays probed by (yeast) 14-3-3 clearly show that Pro is disallowed at +1 (Panni et al., 2011). The structural explanation is that there is no physical space for proline because the backbone peptide bond NH group hydrogen bonds to the 14-3-3 protein (residue Asn176 in the epsilon isoform) as part of local geometry that orients the adjacent phosphorylated side chain (See e.g. Fig. 1D, Molzan et al., 2012). Biologically, rejection of proline at +1 is crucial for separate readout of sites phosphorylated by basophilic kinases (PKA, CAMK etc.) vis-a-vis proline-directed kinases (MAPKs, CDKs etc.) as discussed by Panni et al. (2011).

Arginine is typically present at either positions -3 or -4 in many well-studied phosphopeptides binding to 14-3-3 proteins. Presence of Arginine at position -2 is also sometimes observed and structures show that the residue orients to H-bond with the phosphate group. Therefore Arg at the -2 position is not an issue in the present context.

14-3-3 phosphomimetics. We already know that aspartic acid cannot function as a mimetic (de Chiara et al., 2009). This means that the proposed RLDP motif in WNV NS3 can’t work. I am unaware whether a glutamic acid mimetic has previously been evaluated in vitro for its binding affinity. The reason why it may not be possible for any 14-3-3 phosphomimetic is that the PO3- group is complemented by three H-bond donor ligands: in epsilon they are Arg57, Arg130, Tyr131 (Fig. 1D, Molzan et al., 2012). The single negative charge and planar geometry of the carboxyl group rule out an equivalent interaction for glutamate.

The well-studied 14-3-3-binding phosphomotifs are overwhelmingly, perhaps exclusively, found in regions of intrinsically disordered protein (IDP). As the authors note, the motif postulated in NS3 is on the protein surface but is very well folded. Structurally-embedded motif candidates require a high standard of proof. It is essential that the affinity of the interaction between purified NS3 and 14-3-3 be measured in vitro. In our recent motif discovery guidelines paper, we have explained why a short linear motif is never reliably assigned unless in vitro binding assays using purified components have been undertaken in addition to the in-cell experiments (Gibson et al., 2015).

In my view the experiments reported here also do not rule out indirect interaction modes. The most direct experiment, in Fig. 1G, used HEK293T cell extracts for the GST-NS3 bead attachment and this might allow bridging proteins to be complexed with NS3 that can secondarily bind the added 14-3-3.

To recap, the data presented in the paper here indicate that NS3 and 14-3-3 epsilon associate within the same macromolecular complex. In my view, whether the interaction may be direct or indirect has not been unambiguously determined.

I wish to conclude with two straightforwardly testable predictions:

1. A synthetic peptide encompassing the viral RxEP sequence will bind 14-3-3 epsilon with an affinity that is millimolar or weaker when measured by ITC or an equivalent solution-based assay.

2. If NS3 directly and convincingly binds 14-3-3 epsilon at micromolar or nanomolar affinity, when measured by ITC or equivalent, it will do so by an interface that does not involve phosphomimicry.

References

Davey NE, Travé G, Gibson TJ. How viruses hijack cell regulation. Trends Biochem Sci. 2011 Mar;36(3):159-69.

de Chiara C, Menon RP, Strom M, Gibson TJ, Pastore A. Phosphorylation of S776 and 14-3-3 binding modulate ataxin-1 interaction with splicing factors. PLoS One. 2009 Dec 23;4(12):e8372.

Gibson TJ, Dinkel H, Van Roey K, Diella F. Experimental detection of short regulatory motifs in eukaryotic proteins: tips for good practice as well as for bad. Cell Commun Signal. 2015 Nov 18;13:42

Molzan M, Weyand M, Rose R, Ottmann C. Structural insights of the MLF1/14-3-3 interaction. FEBS J. 2012 Feb;279(4):563-71.

Panni S, Montecchi-Palazzi L, Kiemer L, Cabibbo A, Paoluzi S, Santonico E, Landgraf C, Volkmer-Engert R, Bachi A, Castagnoli L, Cesareni G. Combining peptide recognition specificity and context information for the prediction of the 14-3-3-mediated interactome in S. cerevisiae and H. sapiens. Proteomics. 2011 Jan;11(1):128-43.

On 2016 Mar 24, Jeromy Anglim commented:

Thanks for posting a comment. I just wanted to add a few thoughts on your points.

Relationship between type D and outcomes were not moderated by illness group. For us, this was the important conclusion that we want researchers to think about. There is a lot of research done examining Type D in specific illness groups. To some extent implicit in such research is the idea that the effect of Type D might vary based on illness type. While this may be true, this study presents some evidence that at least for the illness groups and variables studied, this is not the case.

The effect of negative affect on social support and the effect of social inhibition on health behaviours failed to reach statistical significance. I would not say that it failed. This is actually an important finding that supports the idea that the subscales of Type D provide different correlates (i.e., negative affectivity is related more to affective processes, and social inhibition is related more to social processes). Presumably, this is as we would expect. Although if you wanted to be critical, there is the idea that NA is merely neuroticism, and SI is a mix of neuroticism and introversion (see Horwood, Anglim, Tooley, 2015). Whether that is a problem probably depends on how primary you view Big 5 personality.

Limitations: I think we note the limitations you mention in the limitations section. That said, those limitations only relate to certain points of the paper. For me personally, I think that the paper has two fairly important implications for researchers working with Type D personality. The first relates to the general lack of variation in effects by group as discussed above. Second, we did a comparative regression analysis comparing a range of different scoring systems for Type D personality. The results suggested that binary Type D is a poor predictor, and there was limited evidence for NA by SI interactions effects. Rather entering NA and SI as two separate predictors generally resulted in the best prediction of outcomes. This goes agains the implicit claims of Type D that there is an interactive effect and that cut-offs are appropriate for Type D. Importantly, all these analyses also speak to the novelty of the Type D construct and the rationale for choosing the two particular subscales for inclusion.

Thus, for me, the paper provides a nuanced and critical assessment of the predictive validity of Type D personality. In particular, I'd encourage other researchers working with Type D personality (and some are doing this already) to run the comparative regression analyses in their samples.

On 2016 Mar 23, Andrew Kewley commented:

Given that the main hypothesis was unsupported by the evidence, it seems strange that there is no mention of this in the abstract.

The hypothesis was stated: "that functional somatic syndromes, conditions that are characterized primarily by general somatic complaints of unclear etiology, such as chronic fatigue syndrome or fibromyalgia, may be more susceptible to the effects of Type D personality than illnesses of known etiology such as type 2 diabetes or arthritis."

The authors did not find a difference in the prevalence of "Type D personality" between the two chronic illness groups. Likewise, the effect of negative affect on social support and the effect of social inhibition on health behaviours failed to reach statistical significance between the two chronic illness groups after correcting for multiple comparisons.

The study had key limitations, in particular it was based on a convenience sample and self-report questionnaires rather than objective measures of symptom impact and social support. The sample size was adequate.

While there was much speculation in the discussion, the most likely directional association between the questionnaire results and illness is simply that chronic illness contributes to affective suffering.

On 2016 Apr 29, Geriatric Medicine Journal Club commented:

This article was critically appraised at the April 2016 Geriatric Medicine Journal Club (follow #GeriMedJC on Twitter). While there were concerns about the level of commercial influence in the included studies, the overall young age of the patients, the lack of inclusion of other treatment modalities (exercise, narcotics), and the absence of analysis of harm; this paper brought forth important doubts about effectiveness of acetaminophen. How to reconcile the known harms of NSAIDs and the clinical effectiveness on pain and function is the essence of ongoing discussion.

Did you miss the #GeriMedJC tweetchat? Check out the transcript in our Archives section: http://gerimedjc.utorontoeit.com/index.php/2015/12/01/missed-gerimedjc-all-archived-here/

On 2016 Jun 21, Evelina Tutucci commented:

We have also recently discussed Nelles et al. Nelles DA, 2016. Since we are interested in developing new techniques for studying gene expression and mRNA localization at the single molecule level, a potential tag-less system to detect mRNAs in fixed and live cells would be a further advance. As pointed out by the Duke RNA Biology journal club we think that Nelles et al. represents an attempt to apply the Cas9 System to detect endogenous mRNA molecules. Unfortunately, no evidence is presented to demonstrate that this system is ready to be used to study gene expression at the single molecule level, as the MS2-MCP system allows. The RNA letter by Garcia and Parker Garcia JF, 2015 showed that in S. cerevisiae the binding of the MS2 coat protein to the MS2-loops diminished tagged mRNA degradation by the cytoplasmic exonuclease Xrn1. However, these observations were not extended to higher eukaryotes. Previous work from our lab described the generation of the beta-actin-MS2 mouse, whereby all the endogenous beta-actin mRNAs were tagged with 24 MS2 loops in the 3’UTR (Lionnet T, 2011, Park HY, 2014). This mouse is viable and no phenotypic defects are observed. In addition, control experiments were performed to show that the co-expression of the MS2 coat protein in the beta-actin-MS2 mouse allowed correct mRNA degradation and expression (Supplementary figure 1b, Lionnet T. et al 2011). Furthermore, multi-color FISH (Supplementary figure 6, Lionnet T. et al 2011) showed substantial co-localization between the ORF FISH probes and MS2 FISH probes, demonstrating the validity of this model. We think that the observations by Garcia and Parker are restricted to yeast because of the short half-life of their mRNAs, wherein the degradation of the MS2 becomes rate-limiting. Based on our extensive use of the MS2-MCP system, we think that higher eukaryotes may have more time to degrade the high affinity complexes formed between MS2-MCP, providing validation for this system to study multiple aspects of gene expression. In conclusion, we think that the MS2-MCP system remains to date the best method to follow mRNAs at the single molecule level in living cells. For the use of the MS2-MCP system in S. cerevisiae we have taken the necessary steps to improve it for the study of rapidly degrading mRNAs and are preparing this work for publication.<br> Evelina Tutucci and Maria Vera, Singerlab

On 2016 May 17, Duke RNA Biology Journal Club commented:

These comments were generated from a journal club discussion:

We were excited to read and discuss this paper as many of us have questions pertaining to mRNA localization. This technique theoretically allows for imaging of mRNAs without genetic manipulation meaning mRNAs at native expression levels can be tracked in live cells. However, as with many cutting-edge papers, more work is needed before this will become commonplace in the lab.

Most current methods to track mRNAs in live cells involve aptamer based methods which require genetic manipulation of mRNA PMCID: PMC2902723. Additionally, the most commonly used aptamer system, the MS2-MCP system, has become controversial in light of recent findings that the MS2 coat protein stabilizes the aptamer bearing constructs Garcia JF, 2015. In this paper, Fig 1F and 1G replicated these findings and also reassured us that the RCas9 technique would not have the same downfall. While this is certainly a good thing, we were unconvinced this technique was better than FISH (Fig 2), other than having the potential for live cell imaging.

Unfortunately, we found the live cell imaging, which was limited to Fig 3B, to be disappointing. First, we observed that unless an mRNA is strictly localized, as in stress granules, live imaging shows a diffuse mass within the cytosol. Second, imaging was performed with ACTB mRNA which is highly abundant. We don’t think live cell imaging would work as well for low abundance mRNAs due to high background signal. Finally, while specialized imaging software can detect the pile-ups of mRNA in localized foci, we are concerned that tracking individual mRNA may prove a hurdle. Cell models for mRNA localization are large cells such as fibroblasts and neurons, we would be interested to see the ability of this system within these cell types.

One major flaw with this system is the lack of ability to monitor nuclear localized RNA such as lncRNA or splicing machinery. Since since the RCas9 and sgRNA have to first be produced in the nucleus, the majority of the signal in all the figures came from the nucleus. There is a split-GFP-PUM-HD system that has been used to successfully track mRNA in mitochondria Ozawa T, 2007. Perhaps a similar concept could be used with the Cas9 system. This would prove an advantage to the Pumilio system since only the sgRNA needs to be modified instead of the entire protein.

Overall, this is a great start towards a new, tagless, method of mRNA tracking. We look forward to future developments and improvements of this exciting technique.

On 2017 Jul 01, David Keller commented:

A randomized, controlled trial of ethanol for reduction of cardiovascular mortality is needed

The reduction of overall mortality observed with moderate regular ethanol ingestion is driven by the much larger and more significant reduction in cardiovascular mortality which is observed [1]. The beneficial effects of moderate ethanol consumption on the lipid profile are significant, along with hypothetical benefits from antithrombotic and relaxation-inducing effects of ethanol. Many cardiologists have lamented the lack of quality data that would allow them to prescribe moderate alcohol intake to patients with elevated cardiovascular risk, and no elevated risk of addictive or hepatic adverse effects of alcohol. In addition, it is clear that the topical carcinogenic effects of ethanol on the gastrointestinal epithelium vary directly with the cytotoxic effects caused by ethanol concentration, with the greatest risk of esophageal cancers seen with high-proof beverages such as straight whiskey, and no significant elevation of risk due to comparatively dilute beer, with its 5% ethanol concentration causing little if any topical cytotoxicity.

I am calling for a randomized, placebo-controlled interventional trial of ethanol, taken at bedtime to reduce automobile accidents as a source of traumatic mortality, in persons at elevated risk of cardiovascular mortality. Expectation effects could be minimized by administration of the ethanol in capsules, several of which might be required. Oil of peppermint could be included in both the ethanol capsules and in the placebo, for further masking when belching or reflux occur. We have been speculating and correcting observational data for long enough. If a randomized, controlled, interventional trial of ethanol proves to have powerful benefits in persons at high risk of cardiovascular mortality, then populations with average cardiovascular risk can be tested next. Such clinical trials are long overdue. The potential benefits of ethanol are substantial, if properly ascertained.

Reference

1: Vogel RA. Alcohol, heart disease, and mortality: a review. Rev Cardiovasc Med. 2002 Winter;3(1):7-13. Review. PubMed PMID: 12439349.

On 2017 Jul 01, David Keller commented:

A randomized, controlled trial of ethanol capsules for reduction of cardiovascular mortality is called for

The reduction of overall mortality observed with moderate regular ethanol ingestion is driven by the much larger and more significant reduction in cardiovascular mortality which is observed. The beneficial effects of moderate ethanol consumption on the lipid profile are significant, along with hypothetical benefits from antithrombotic and relaxation-inducing effects of ethanol. Many cardiologists have lamented the lack of quality data that would allow them to prescribe moderate alcohol intake to patients with elevated cardiovascular risk, and no elevated risk of addictive or hepatic adverse effects of alcohol. In addition, it is clear that the topical carcinogenic effects of ethanol on the gastrointestinal epithelium vary directly with the cytotoxic effects caused by ethanol concentration, with the greatest risk of esophageal cancers seen with high-proof beverages such as straight whiskey, and no significant elevation of risk due to comparatively dilute beer, with its 5% ethanol concentration causing little if any topical cytotoxicity.

I am calling for a randomized, placebo-controlled interventional trial of ethanol, taken at bedtime to reduce automobile accidents as a source of traumatic mortality, in persons at elevated risk of cardiovascular mortality. Expectation effects could be minimized by administration of the ethanol in capsules, several of which might be required. Oil of peppermint could be included in both the ethanol capsules and in the placebo, for further masking when belching or reflux occur.

We have been speculating and correcting observational data for long enough. If a randomized, controlled, interventional trial of ethanol proves to have powerful benefits in persons at high risk of cardiovascular mortality, then populations with average cardiovascular risk can be tested next. Such clinical trials are long overdue. The potential benefits of ethanol are substantial, if properly ascertained.

Reference

1: Vogel RA. Alcohol, heart disease, and mortality: a review. Rev Cardiovasc Med. 2002 Winter;3(1):7-13. Review. PubMed PMID: 12439349.

On 2016 Apr 12, Martin Hofmeister commented:

Do not forget published reviews

I thank Dr Mat Eil Ismail et al, for their interesting article "Preoperative physiotherapy and short-term functional outcomes of primary total knee arthroplasty", published in the March 2016 issue of the Singapore Medical Journal (SMJ). There is one aspect worth mentioning. In my opinion, systematic reviews of the past five years should be included in an original article consistent with good scientific practice. The reviews of Kwok et al, Jordan et al, Simmons et al and the Australian Agency for Clinical Innovation about the evidence of preoperative physiotherapy on outcomes following total knee arthroplasty are not mentioned in the discussion (1-4). The results of the SMJ study reconfirm the above-mentioned reviews: No statistically significant effect in patient key outcomes (1-4). I refer readers to the latest meta-analysis "Does preoperative rehabilitation for patients planning to undergo joint replacement surgery improve outcomes?" that can be found in the February 2016 issue of the BMJ Open (5).

REFERENCES

1) Kwok IH, Paton B, Haddad FS. Does Pre-Operative Physiotherapy Improve Outcomes in Primary Total Knee Arthroplasty? - A Systematic Review. J Arthroplasty. 2015;30:1657-63. Kwok IH, 2015

2) Jordan RW, Smith NA, Chahal GS, Casson C, Reed MR, Sprowson AP. Enhanced education and physiotherapy before knee replacement; is it worth it? A systematic review. Physiotherapy. 2014;100:305-12. Jordan RW, 2014

3) Simmons L, Smith T. Effectiveness of pre-operative physiotherapy-based programmes on outcomes following total knee arthroplasty: a systematic review and meta-analysis. Phys Ther Rev. 2013;18:1-10. http://www.tandfonline.com/doi/abs/10.1179/1743288X12Y.0000000035?journalCode=yptr20

4) NSW Agency for Clinical Innovation (NSW ACI). Musculoskeletal Network: NSW Evidence Review: preoperative, perioperative and postoperative care of elective primary total hip and knee replacement. Chatswood, Australia: Agency for Clinical Innovation, 2012. Available at: http://www.aci.health.nsw.gov.au/__data/assets/pdf_file/0020/172091/EJR-Evidence-Review.PDF. Accessed 22 March 2016.

5) Wang L, Lee M, Zhang Z, Moodie J, Cheng D, Martin J. Does preoperative rehabilitation for patients planning to undergo joint replacement surgery improve outcomes? A systematic review and meta-analysis of randomised controlled trials. BMJ Open 2016;6:e009857. Wang L, 2016

On 2016 Apr 30, Ivan Oransky commented:

Ben Ashby, who reviewed a later paper by Hanna Kokko, one of the authors, in which she corrected an error in this paper, says it should not have been retracted: http://retractionwatch.com/2016/04/29/why-that-evolution-paper-should-never-have-been-retracted-a-reviewer-speaks-out/

On 2016 Mar 29, Wan map Ni commented:

It sounds effective. But in fact the gates could be more effective. Furthermore, it is not suitable for many patients with hematological malignancy , especially MDS.

On 2016 Jul 10, Jihoon Yoon commented:

I found that the author cited my case report in this article. (Citation No. 121.) However, the citation is inappropriate. The patient in our case represented duplication in 14q11.2 to 14q21.1 region, but the genes mentioned in this article, AMN and HSP90AA1, are located on 14q32 which does not duplicated in our case. There is no overlapping area with the genes and the duplicated region in our case. Therefore, the author should revise the citation or the content. (Contacted with the author.)

On 2016 May 04, Simon Young commented:

This review raises a number of issues related to the content and interpretation of articles that are cited.

1. Paragraph 6 on page 347, about side effects of TRP, cites two placebo controlled trials of TRP (Thomson et al and Steinberg et al) and mentions side effects of TRP. However, the article fails to mention that the study by Thomson et al reported no significant differences in side effects between TRP and placebo, and that in the study of Steinberg et al only 1 of 20 side effects measured (dizziness) was significantly greater after TRP group than after placebo. The same paragraph also contains the sentence “In a study in which 3 g TRP daily was administered to participants for 12 weeks, one patient reported diarrhea as a side-effect of TRP intake (Van Praag et al., 1972)”. However, the van Praag article states “The precursor we used was not tryptophan but dl-5-HTP”. 5-Hydroxytryptohan has a different side effect profile from that of TRP.

2. Paragraph 6 on page 347 mentions that the dose of TRP in the study of Steinberg et al was 6g daily for 3 months and suggests that “Such high doses might not be recommendable not only because of such side-effects, but also because the TPH enzyme is likely to already be saturated by a dose of TRP up to 3g”. While it is true that the daily dose was 6g it was given in three doses of 2g each (after breakfast, after lunch, and before bed), which would not be likely to saturate TRP hydroxylase, and explains why only 1 of 20 side effects measured was significantly greater in the TRP group than in the placebo group. Also, the article did not mention that TRP altered social behavior, specifically a decrease in irritability, in the study of Steinberg et al.

3. The article lists the dose of TRP in the study of Volvaka et al as “6g for 3 weeks”. It fails to mention that the daily dose of 6g was given in divided doses, either 3 or 4 times per day. The article lists the dose of TRP in the study of Nemzer et al as 100mg/kg daily for 1 week. TRP was given in divided doses in the morning and afternoon. The article gives the dose of TRP in the study of Cerit et al as “2.8g per day for 6 days”. The TRP was given in 3 doses per day, 0.8g in the morning and afternoon and 1.2g in the evening. These differences are important as a divided dose is likely to increase serotonin synthesis for longer during each day than the same amount given as a single dose, as well as possibly reducing side effects.

4. Five of the articles listed in Table 1 and discussed in the text involve a technique called acute TRP depletion (ATD). The discussion of all these articles lacks important information. In ATD participants are given a mixture of amino acids that is devoid of TRP (T-), or the same amount of amino acids plus the appropriate amount of TRP as in a balanced protein (B mixture), which for the studies described is 2.3g Young SN, 2013. When the T- mixture is given the amino acid mixture induces protein synthesis and TRP in the blood and tissues is incorporated into protein. Over 5 hours TRP levels fall dramatically and the rate of serotonin synthesis in human brain can decline by more than 90% Nishizawa S, 1997. The B mixture, which contains TRP, will not increase brain TRP or serotonin synthesis, for reasons explained in the paragraph 2 of section 1.1 of the article. Sometimes in ATD studies an additional 8g of TRP is added to the B mixture to raise brain TRP (T+). The problems in the discussion of these articles are: (i) The study of Marsh et al compared the T- mixture, the B mixture, and fasting participants. None of these treatments would increase brain TRP so this was an ATD study, not a TRP supplementation study. The study demonstrated that ATD increases aggressive responding relative to the B mixture. This does not necessarily mean that TRP supplementation would have decreased aggression relative to the B mixture. The fact that the B mixture decreased aggressive responding relative to the fasting condition is irrelevant as the B treatment does not increase brain TRP. While the B and T- mixtures were given under blind conditions, the fasting day occurred after both days in which amino acid mixtures were given, and the participant knew they were not receiving an amino acid mixture. Any effect could have been due to lack of blinding or an order effect, issues not mentioned in the article. Also (a minor point) the article mentions “a control condition (i.e. a low monoamine diet)”. The participants were on a low monoamine diet throughout the study, and the control condition was fasting. (ii) The study of Bjork et al (2000) compared ATD treatment with a T+ treatment containing 10.3g TRP, and a fasting condition. The conclusion given in Fig. 1 of the article by Steenbergen et is that TRP supplementation “Decreased aggressive responding”. However, the article by Bjork et al (2000) states that “TRP depletion increased aggression relative to TRP loading in aggressive men”. Any difference could have been due to increased aggression after ATD, or decreased aggression after TRP loading, or both, so this study did not demonstrate an effect of TRP loading. (iii) The study of Cleare and Bond looked at the effect of ATD and a T+ mixture containing 10.3g TRP. As reported by Steenbergen et al TRP supplementation “was found to reduce self-report ratings of angriness, quarrelsomeness, hostility, and annoyance, but only for males with high trait levels of aggression”. These were changes over time after administration of the amino acids and were relatively small (at most 20%). Given the absence of any control group with unaltered TRP levels these changes cannot necessarily be attributed to the effect of TRP supplementation. (iv) The study of Finn et al compared the effect of a B mixture and a T- mixture. As there was no treatment that increased brain TRP this paper should not have been included in the article.

5. Section 3 of the article includes the statement “if the TPH1 enzyme in the gut is very active, more TRP is converted there and less will be available to pass through the BBB and be converted into 5-HT in the brain. Thus TRP might have less impact on social behavior in individuals with highly active TPH1 enzyme.” Under normal circumstances only about 1% of ingested TRP is converted into serotonin by TPH (see section 4, paragraph 5 of the article and SJOERDSMA A, 1956). Only in a rare condition, the carcinoid syndrome, which is due to neuroendocrine tumors metastasizing in the liver and producing large amounts of serotonin Molina-Cerrillo J, 2016, is metabolism via TPH quantitatively significant, with up to 60% of TRP converted to serotonin SJOERDSMA A, 1956. The main catabolic route of TRP is along the kynurenine pathway, and flux along this pathway is increased in patients with major depressive disorder Teraishi T, 2015. However, TRP is an effective antidepressant according to a Cochrane review Shaw K, 2002. Therefore, it is unlikely that increased peripheral catabolism of TRP will mitigate the effect of TRP on social behavior to any great extent.

6. Section 4 of the article states “some of the effects of TRP administration on social behavior might in fact be a result of enhanced sleep and mood”. An effect on sleep cannot apply to the majority of the studies discussed, which were carried out during a single day. When TRP was given for many days TRP was usually given in a divided dose with only a small dose (usually around 1g) given in the evening, sometimes with dinner. When TRP is given as a hypnotic it is usually given shortly before bedtime, as plasma TRP rises quickly after ingestion. Whether the low doses of TRP given in the evening in the studies lasting longer than a day would have had an effect on sleep is not clear. In relation to improved mood as a mediator of more positive social behavior, the study of Moskovitz et al (2001) demonstrated that TRP resulted in more positive social behavior without any change in mood.

On 2016 Jun 28, Lydia Maniatis commented:

No real objection to this study, which is refreshingly sensible. I want to note, though, that the term "geometric figure-ground cues" is no more specific than the term "good shape cues," both of which are essentially place-holders for further specification (e.g. convex shape). I want to note this because "good shape" is a Gestalt concept which was trivialised and dismissed, but which evidently is necessary. It's not reducible to points, orientations, angles, "features," "signals," or probabilities. (The probability of a particular seen shape can be, essentially, zero.)

On 2016 May 11, Wan map Ni commented:

When will flow cytometry immunophenotyping be included in the classification?

On 2017 Dec 07, M Mangan commented:

There was a major correction to this paper, but it's not linked from this page. Their concentrations were 1000x off. de Aguiar LM, 2017 94-101].")

On 2016 Dec 06, Atanas G. Atanasov commented:

Excellent, thanks a lot for sharing your experience, I have enjoyed a lot reading :-)

On 2017 Oct 29, David Keller commented:

Pharmacists should be enlisted to help physicians monitor PDMP data

Recommendation #9 is: "Clinicians should review the patient’s history of controlled substance prescriptions using state prescription drug monitoring program (PDMP) data to determine whether the patient is receiving opioid dosages or dangerous combinations that put him or her at high risk for overdose. Clinicians should review PDMP data when starting opioid therapy for chronic pain and periodically during opioid therapy for chronic pain, ranging from every prescription to every 3 months."

Prescribing physicians should obtain and review the patient's PDMP data, if possible, prior to initiating any opioid prescription. In addition, pharmacists must be enlisted to check the PDMP databases whenever they fill a prescription for a Schedule 2, 3 or 4 medication. If the pharmacist discovers evidence of concurrent opioid prescriptions or other red flags, they should inform the prescriber and the patient.

The 2016 CDC Guideline for Prescribing Opioids for Chronic Pain establishes a de facto standard of care for prescribing opioids, which will enhance patient safety, and thereby increase access to opioids for patients who really need them. Prescribers will gain a degree of protection from lawsuits for adverse outcomes related to opioid therapy by documenting careful adherence to the 12 recommendations in the CDC Guidelines. This adherence (and documentation) will increase the cost of caring for pain patients, as measured in physician time and effort.

Pharmacists should be required to cross-check the PDMP databases with every opioid fill and refill, and transmit their findings to the prescribing physician, especially red-flag findings. This will be an easy task for pharmacists, and will free up physician time to perform the extensive discussions and documentation required by these Guidelines.

On 2016 Sep 19, Preben Berthelsen commented:

The consequence of the statistical problems in this study must be considered before accepting the authors’ contention that dexmedetomidine is better than saline in ICU patients with agitated delirium.

In the power/sample size calculation and the pre-trial ClinicalTrials registration, a difference of at least 20 hours in ventilator-free time was defined as the minimum difference of clinical interest. The authors found difference of only 17 hours. This finding was statistically significant but not clinically important as the authors - a priori - had defined the 20 hour difference as the minimum difference of interest.

The trial was stopped early when the sponsor – the manufacturer of dexmedetomidine – lost patience due to slow recruitment of patients. As a consequence, only 71 of the 96 stipulated patients could be included in the final statistical analysis – severely limiting the power of the trial. The authors state in the paper that “no data analysis by the study investigators had occurred prior to this decision.” The decision alluded to is the decision taken by the pharmaceutical company sponsoring the trial. But the essential point is when or why the investigators decided not to finish the trial according to the original plan. Was the decision taken before or after the finding of a statistically significant result? The answer cannot be found in the paper.

Additionally, analyses of the “hard” secondary end points (length of ICU stay, hospital stay or mortality) do not support the authors’ view that dexmedetomidine is superior to saline in ICU patients with agitated delirium.

In conclusion, I find that the authors’ hypothesis that dexmedetomidine is effective in ICU patients with agitated delirium not proven. Moreover, I suspect that the chance of reproducing the positive result of this trial to be no better than fifty-fifty.

P.G.Berthelsen. MD, MIA, DCHA. Charlottenlund, Denmark

On 2017 Jun 26, David Marks commented:

This trial was neither randomised, nor controlled, and needs to be retracted. The trial did not compare an abrupt method to a gradual method, as stated, it compared an abrupt method with a mixed bag of complicated gradual methods about which it is impossible to draw solid conclusions. The lead authors have either declared significant conflicts of interest in products used in the research (Aveyard and West) or the ICMJE Form for Disclosure of Potential Conflicts of Interest is incorrect, being the form for the wrong study(Michie). The conclusion that abrupt cessation produces superior cessation rates to gradual cessation cannot be maintained on the basis of this flawed trial for the reasons given below. 1) The gradual-cessation group received short-acting nicotine replacement therapy (NRT) and nicotine patches before the quit day. The abrupt-cessation group received only nicotine patches before the quit day. The treatments are therefore confounded with different pre-exposure levels of NRT. 2) Eligible smokers were booked for an appointment with a research nurse during which the study was explained, eligibility was confirmed, and written informed consent was obtained. What research training did the so-called 'research nurses' receive or were the key study personnel basic grade practice nurses given the task of running the trial? 3) The gradual-cessation group were were supposed to reduce smoking to half of the baseline amount by the end of the first week (known as visit −1) and to a quarter of the baseline amount at the end of the second week (visit 0) in daily increments using a complex variety of procedures that were likely to have been confusing and difficult to follow. 4) The nurses provided the gradual-cessation group with nicotine patches, 21 mg/d, and a choice of short-acting NRT products (gum, lozenges, nasal spray, sublingual tablets, inhalator, or mouth spray) during the reduction period. For such products as gum and lozenges, the instruction was to use 1 dose per cigarette missed. Again, apart from the confounding, and different pre-exposure for the gradual-reduction group, the procedure is unnecessarily complex. 5) Before quitting, participants in the abrupt-cessation group were asked to use nicotine patches, 21 mg/d, but no short-acting NRT. Nicotine patches were used in this group before the quit day, a protocol that aimed to balance the effect between groups. Instead of aiming to balance the effect between groups, there should have been precise balancing, otherwise the trial cannot be described as 'controlled'. 6) Allocation of participants was the responsibility of the so-called 'research nurse' who put patients into blocks of 2, 4, and 6. This allocation was manifestly non-random: “After the participant granted consent, the research nurse opened sealed, numbered envelopes in turn. However, for pairs (for example, husband and wife), one person was allocated randomly and the other was allocated to the same group”. This was a clustered method of allocation, not randomisation. 7) The loss of 300 potential participants also raises questions about how the remaining 697 differed from the original applicants. 8) Unsurprisingly, given their complicated and non-matched treatment, significantly fewer participants in the gradual-cessation group attended visit 0, (67.0% [229 of 342] vs. 83.4% [296 of 355] in the abrupt-cessation group; P < 0.001). Fewer participants in the gradual-cessation group (61.4% [210 of 342]) than the abrupt-cessation group (71% [252 of 355]) (P = 0.007) made a quit attempt. In sum, the trial was a mish-mash of umatched 'treatments', one of which was actually three different treatments counted as one, both confounded by differing pre-cessation exposure to a variety of NRT products chosen by the participants themselves. The trial was carried by 'research nurses' working in GP practices using a batch method for allocating participants. A deliberately uncontrolled, improperly randomised trial, confounded by different NRT products between groups.

On 2017 Aug 09, Tom Kindlon commented:

Questioning the exclusion from a diagnosis of CFS of individuals whose symptoms improve with rest

When I read this paper I was concerned that genuine patients could be excluded unnecessarily.

A recent paper[1] I believe highlights well the point I am concerned with:

"Additionally, there is likely to be a good amount of variability in how this case definition is used. In particular, the potential for variation in the methods used to assess substantial reduction has not yet been adequately explored. Operationalizing key concepts outlined in the Fukuda criteria is important. For example, it would be useful to find a reproducible way to specify fatigue as outlined in Fukuda [1]: “chronic fatigue that is of new or definite onset (i.e., not lifelong). The fatigue is not the result of ongoing exertion. The fatigue is not substantially alleviated by rest.” To this end, others have outlined a way to define “lifelong,”3 which is indeed a challenging task [23].

Let’s examine how Unger and colleagues [3] operationalized “not substantially alleviated by rest.” First the person would need to answer “no” to fatigue was made a lot better by rest to fulfill this requirement. But if they responded “yes” to fatigue was made a lot better by rest, they could be included if their fatigue was relieved by rest “some of the time,” “a little of the time,” “or hardly ever.” They would be not included if they said that their fatigue was relieved by rest “all of the time” or “most of the time.” The problem with this approach stems from the fact that much of the time, rest does relieve fatigue symptoms for many patients with CFS. However, for these patients, rest is not fully curative and does not increase the stamina and endurance necessary to carry on life tasks. Therefore, while it is important to operationalize this part of the Fukuda case definition, it is critical to do so in a way that distinguishes between those who’s rest fully eliminates the symptom complex and those form whom this does not occur (e.g., patients with CFS). It is equally important to determine if CFS induced fatigue is result of ongoing exertion. The failure of the Unger et al. article [3] and the empiric criteria to address this key issue of ongoing exertion causing the fatigue is problematic. In other words, unless questions have been carefully crafted and validated, a person could meet the CFS diagnosis whose fatigue is mainly due to excessive exertion, and with lifestyle issues such as being over-committed."

References:

1 Jasaon LA, Gleason K, Fox P (2017) The implications of using a broad versus narrow set of criteria in research. J Med Therap 1: DOI: 10.15761/JMT.1000116

On 2016 Mar 15, Tom Kindlon commented:

Possible errors in Table 4?

Minor point: I think there is a good chance there are errors in Table 4 in terms of listing which groups are different statistically on some criteria. For example, for both Role Physical and Social Functioning, it says the only differences are between M1 only and M2 only but it looks very likely that M1 only would also be different from M1/M2 given that compared to M2, the scores for M1/M2 are worse again, the SEMs are smaller and the sample size is bigger.

On 2016 Mar 15, Tom Kindlon commented:

Depression scores in this follow-up study are very different to scores in original study (looking solely at the Reeves et al. (2005) operationalization)

Leonard Jason and colleagues previously raised concerns about the Reeves et al. (2005) chronic fatigue syndrome (CFS) criteria [which have also been described as an operationalisation of the Fukuda et al (1994) criteria] (1-4). In particular, Jason and colleagues were concerned that some people who did not have CFS might get diagnosed with CFS using this new set of criteria. They found some evidence to support this concern in a study of those with major depressive disorder who did not have CFS: 38% were found to satisfy these new criteria for CFS(4).

Looking solely at the current study, it would look like there might have been little basis for these concerns. Of 71 people classified with CFS in the current study, only one (1.4%) had a Zung self-rating depression scale (SDS) (5) score of >=60. The mean SDS score for the 71 CFS participants was 44.78 (calculated from the data in Table 4) (6).

However, it should be noted that the SDS (depression) scores in the follow-up study are very different from the scores in the original Georgia cohort(7). Of the 113 people diagnosed with CFS in the original Georgia cohort, data for 112 (99.1%) was published(7). The average SDS score was considerably higher at 56.2. Possibly more revealingly, 40.2% had a SDS score of >=60. As described in the paper, the SDS scale provides an index score and categories reflecting no (<50), mild (50-59), moderate (60-69), and severe (>=70) depression.

I am not sure why there should be such a large difference in a cohort between the initial and follow-up studies in the rate of those with moderate or severe depression (40.2% vs 1.4%). But it does mean that caution should be used in terms of interpreting the findings reported in the current paper and their significance regarding the Reeves et al. (2005) criteria (1,6).

References:

[1]. Reeves WC, Wagner D, Nisenbaum R, Jones JF, Gurbaxani B, Solomon L, Papanicolaou DA, Unger ER, Vernon SD, Heim C. Chronic fatigue syndrome--a clinically empirical approach to its definition and study. BMC Med. 2005 Dec 15;3:19.

[2]. Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. The chronic fatigue syndrome; a comprehensive approach to its definition and study. Ann Int Med 1994, 121:953-959.

[3]. Jason LA, & Richman JA. How science can stigmatize: The case of chronic fatigue syndrome. Journal of CFS 2007;14:85-103.

[4]. Jason LA, Najar N, Porter N, Reh C. Evaluating the Centers for Disease Control's empirical chronic fatigue syndrome case definition. Journal of Disability Policy Studies 2009;20;93.

[5]. Zung WW, Richards CB, Short MJ. Self-rating depression scale in an outpatient clinic: further validation of the SDS. Arch Gen Psychiatry.1965;13(6):508-515.

[6]. Unger ER, Lin JM, Tian H, Gurbaxani BM, Boneva RS, Jones JF. Methods of applying the 1994 case definition of chronic fatigue syndrome - impact on classification and observed illness characteristics. Popul Health Metr. 2016 Mar 12;14:5.

[7]. Heim C1, Nater UM, Maloney E, Boneva R, Jones JF, Reeves WC. Childhood trauma and risk for chronic fatigue syndrome: association with neuroendocrine dysfunction. Arch Gen Psychiatry. 2009 Jan;66(1):72-80.

On 2016 Mar 13, Jim Woodgett commented:

I commend you Marek for the most honest title ever applied to a review of the proteins associated with Wnt signalling (although could also be applied to most of what is published in all fields of biology).

On 2016 Apr 03, Michael Brennan commented:

This is consistent with current guiding principles in pain management. The identification of regional variability is new and begs several new questions.

What will likely surprise many is the relatively low rate of opioid "abuse or dependence" identified. Some will claim the relatively low number as validation of opioids and as a new indication of relative safety for the prescribing of opioids. Meanwhile others will criticize process, sample or look for bias to explain number as being too low.

Regardless of ones perspective, this tool offers the next generation in patient assessment.

On 2016 Apr 28, Jorg Rosgen commented:

The processes in question are associated with such small energetic changes, that everything happens within the range of normal thermal fluctuations. That's why protein folding can be both spontaneous and reversible.

On 2016 Apr 14, Ariel Fernandez commented:

Walter:

In so far as we agree that protein folding in vitro is spontaneous under suitable renaturation conditions, the process is thermodynamically irreversible. Hence, the intermediate states of the system, comprised of protein chain and solvent statistical bath, are irretrievable [1]. In thermodynamic terms there is no such thing as a spontaneous reversible process. Ariel Fernandez

[1] Ariel Fernandez Stigliano. Biomolecular Interfaces, Chapter 3 (Springer, Heidelberg, 2015) http://link.springer.com/book/10.1007/978-3-319-16850-0

On 2016 Apr 13, S Walter Englander commented:

I disagree with several issues in this comment.

1. The protein folding question is not a sterile one. It is centrally important to understand how proteins fold and why they fold in that way. The protein folding reaction is essential to all living things, arguably the most important reaction in all of biology. It determines many currently forefront biological behaviors and diseases.

2. Folding is not irreversible. At the molecular level, free energy downhill processes are spontaneous but certainly not irreversible. Proteins spontaneously fold energetically downhill to their lowest free energy native state, as per Anfinsen, but even under fully native conditions they continue to unfold and refold repeatedly over time, cycling through all possible higher energy states. One result is that all higher energy states, including those that carry the major U to N folding flux, are populated at equilibrium, each one according to its Boltzmann factor [K = e^-G/RT].

3. The effort is far from futile. Hydrogen exchange (HX) methods can observe that cycling and characterize the major intermediates (structure, G, ASA). For cytochrome c we found four major partially folded intermediates. Each differs from the next by one native-like foldon unit. In the present paper we used advanced HX MS methods to define the major partially folded cyt c states DURING kinetic folding. They are the same as the high free energy states we found before at equilibrium native conditions. One conclusion is that cyt c and, we suspect, proteins in general fold through defined N-like intermediates, adding one foldon unit at each step.

4. Summary: Fifty plus years after Anfinsen there is still not general consensus about how protein folding works and why it works that way. The reason is that it has been so hard to define folding intermediates and pathways, although not impossible as Fernandez asserts. Our HX experiments indicate that proteins are made of cooperative foldon units that provide built-instructions for the folding process. Stepwise folding puts a sequence of cooperative foldon units into place, one at a time in an ordered foldon-dependent pathway (the HOW question), just as they are evolutionarily tailored to fit together in the native protein (the WHY question).

All of this is true for cyt c (the present paper) and for some other proteins as well (see paper for refs).

On 2016 Apr 05, Ariel Fernandez commented:

In a recent paper, Hu et al. [1] reported a careful and detailed characterization of the folding pathway for a soluble protein. I am always confused by this kind of results. Under the appropriate enabling conditions, protein folding in vitro is known to be spontaneous [2] and therefore thermodynamically irreversible [3]. The endpoints of the protein folding process, i. e. the denatured random coil ensemble and the native state, are of course recoverable by restoring respectively denaturing or renaturing conditions [2]. Yet, at a variance with thermodynamics, protein scientists incorrectly regard this restoration of folding endpoints as meant to imply that protein folding is a reversible process [2]. Be as it may, according to the tenets of thermodynamics, the folding and unfolding pathways are untraceable and irreproducible as it would be the case for any spontaneous process [3]. In fact, the very notion of “pathway” for a spontaneous process is thermodynamically meaningless because dissipative forces intervene in such processes causing a net increase in the entropy of the universe, the hallmark of irreversibility. Thus, as with any spontaneous process, the actual intermediate states associated with the protein folding process are irretrievable. Therefore I think that the sterile controversy on whether protein folding in vitro is actually a two-state process or proceeds through intermediates is not even an issue: The two–state model is simply a realization that intermediates are irretrievable in a thermodynamically spontaneous process [4].

Notwithstanding such thermodynamic considerations, an active quest for folding intermediates continues to this day [1, 5]. In my view, this search remains futile from a thermodynamic perspective, unless some sort of paradox holds (thermodynamics is full of paradoxes) that, at the very least, needs to be properly dispelled before the saga of the quest for folding intermediates continues. To the best of my knowledge this has not been done. Real folding intermediates not only remain elusive: I am afraid they do not exist, and claims to the contrary violate the second law of thermodynamics. Ariel Fernandez

References

1. Hu W., Kan Z.-Y., Mayne L. & Englander, S. W. Proc. Natl. Acad. Sci. USA 113, 3809-3814 (2016).

2. Anfinsen, C.B. Science 181, 223-230 (1973).

3. Planck, M. Treatise on Thermodynamics, 3rd edition, Dover, New York (2010).

4. Fernandez, A. Biomolecular Interfaces (ISBN 978-3-319-16849-4), Springer, Berlin (2015).

5. Vendruscolo M. & Dobson, C.M. Nature Chem. Biol. 9, 216-217 (2013).

On 2016 Aug 30, Duke RNA Biology Journal Club commented:

This paper brought a new perspective to our discussion by focusing on the structural and biophysical characterization of RNA, in this case the mammalian ribozyme CPEB3, to gain insight into its biological function. Of particular interest to our group is understanding structure in the presence of a biologically relevant amount of magnesium, which remains a challenge in the field of RNA structural biology but is critical for understanding how RNA may function in the cell. In fact, magnesium acts as a crucial cofactor for many catalytic RNAs and is often required for folding of structured RNAs into their functionally competent state.

NMR is a powerful technique for studying biomolecular structure and dynamics at the atomic level, but much of the current NMR data describing RNA are reported in the absence of magnesium due to experimental limitations on signal sensitivity which worsens with high-conductivity samples.

This paper used previously established NMR methods such as Diffusion Ordered Spectroscopy (DOSY) and NOESY experiments to probe how CPEB3 ribozyme global conformation and local secondary structure are affected by magnesium. They were able to mitigate sensitivity issues by substituting magnesium with hexamminecobalt (III) Gonzalez RL Jr, 1999 as a probe of outer-sphere metal coordination to identify potential magnesium binding sites. While this technique is a useful starting point, concerns were raised that the authors used NOESY cross peaks of aromatic or sugar protons as the main evidence for direct magnesium binding since these resonances likely shift due to magnesium-induced conformational changes rather than site specific interaction. A better analysis would be to monitor nitrogen cross peaks such as N7 on guanines or adenines Fu DJ, 1992 which are insensitive to secondary structural rearrangements but sensitive to presence of nearby metal ions.

Nonetheless, this paper successfully characterized the impact of magnesium on secondary structure of a complicated RNA system comprised of a full-length nested double pseudoknot ribozyme which happens to be one of the few self-cleaving ribozymes identified in humans. We are excited for the development of new techniques directed toward studying specific RNA-metal interactions to better understand RNA structure as it exists in the cell.

On 2017 Feb 04, Stuart RAY commented:

Ribavirin monotherapy has been studied (summarized in Brok J, 2006), with no beneficial effect detected (an ACP Journal Club commentary was entitled, "ribavirin is not better than placebo" Chen W, 2006). This suggests that if the subset with high IFN-gamma was included in those trials, they did not benefit from ribavirin monotherapy. The comment's author may have conflated interferon gamma (a type II interferon) with interferon alfa (a type I interferon), the latter being a treatment that was enhanced by ribavirin.

On 2017 Feb 03, Muhammad Aslam commented:

In my view the patients with high IFNg plasma levels could benefit with Ribavirin only without suplementing with peginterferon.

On 2017 Sep 20, Stuart RAY commented:

This remains controversial.

On 2017 Jan 05, Melissa Rethlefsen commented:

Upon reading this article, I came across a small detail that I am curious about. In the flow diagram (Figure 1), the authors note in the final step that they found 127 reports to 15 trials. Generally, I would interpret this as the authors having located 127 references that discuss 15 different trials. In other Cochrane reviews, this is often noted as X number of references to X number of studies, and then all of the references for each study are usually listed under the main study name in the included studies list. In this case, there appear to be only 15 references to 15 studies in the included studies list. Is this a typo in the flow diagram, or are the remaining 112 references that could have been ascribed to the 15 studies erroneously omitted?

On 2016 Aug 30, Ben Goldacre commented:

This trial has the wrong trial registry ID associated with it on PubMed. The ID given is NCT102577224 - the correct ID is NCT02577224. This has been corrected in a subsequent correction published in the originating journal, but not in the PubMed metadata.

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, and this ID has already been corrected within the journal itself; we hope that this trial’s text and metadata can also be corrected in PubMed.

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 Mar 27, Bruno Ramalho Carvalho commented:

When we talk about humanization of care, whether in medical area or not, we are talking about the comprehension that the assisted person is the owner of her instincts, needs and desires. Also, that the person is the owner of her past, present and future. Humanization refers to the understanding that any intervention somehow violates the line between the intimate and the notorious. And that exceeding this limit is not always a well received act, even if it was allowed. That is, the intervention can be both visit, as can be invasion. And in the context of humanization only the first option is accepted.

On 2016 Jun 11, Daniel Schwartz commented:

The HOSPITAL score can be used online and in mobile apps here: http://qxmd.com/calculate/calculator_353/hospital-score

Conflict of interest: Medical Director, QxMD

On 2016 Apr 26, Annalisa Forlani commented:

We are thankful to Dr. Cooper for his comment and deep knowledge of the literature. However, we have not included the suggested citation for different reasons, as discussed below.

Images from previous reports of pulmonary hyalinosis may resemble our case. However, there are several histological, histochemical and ultrastructural findings that we believe are not consistent with such diagnosis.

Based on previous reports (Billups et al., 1972; Dagle et al., 1976), hyaline bodies are intracellular globules usually distributed within the cytoplasm of macrophages and multinucleated cells, occasionally calcified and strongly positive for Periodic Acid-Schiff (PAS), crystal violet and Oil Red O. The histological lesions are described as multifocal granulomas effacing smaller bronchi and the subpleural tissue at the margins of lobes.

In our case, the eosinophilic granular material was predominantly extracellular and filled the alveolar spaces without any other pulmonary architectural changes. Moreover, histochemistry revealed a negative Von Kossa and Congo red reaction and only variably positivity for PAS.

Ultrastructurally, the substance in the alveolar lumina was composed by short lamellar thick fascicles rather than whorled intracytoplasmic lamellar membranes as reported for pulmonary hyalinosis. Therefore, our findings were strongly suggestive of accumulation of abnormal surfactant rather than degenerate cells.

Pulmonary alveolar proteinosis and pulmonary hyalinosis are both rare and poorly characterized conditions whose pathogenesis is still not entirely understood. In 1976, Dagle and collaborators hypothesized a similar pathogenetic mechanism, even an overlap between the two . However, no additional efforts have been made toward a better understanding of the two entities.

In conclusion, given the differences between the abovementioned studies on pulmonary hyalinosis and our findings, a comparison with pulmonary hyalinosis seems unnecessary, and no additional comments should be included in our manuscript.

References 1. Billups LH, Liu SK, Kelly DF, Garner FM. Pulmonary granulomas associated with PAS-positive bodies in brachycephalic dogs. Vet Pathol 1972; 9: 294-300. 2. Dagle GE, Filipy RE, Adee RR, Stuart BO. Pulmonary hyalinosis in dogs. Vet Pathol 1976; 13: 138-142.

On 2016 Apr 12, Timothy K Cooper commented:

The authors should consider pulmonary hyalinosis of brachycephalic dogs: Pulmonary Granulomas Associated with PAS-Positive Bodies in Brachycephalic Dogs. Billups LH, et al. Veterinary Pathology 1972: 9(5):294-300.

On 2016 May 10, Kenneth Katz commented:

Obviously an important and well crafted study. However. while I may have missed it, I do not see what I would have considered a key control: that a different flavivirus, such as Dengue, or West Nile, was unable to infect the progenitors in this system. After all, it is concluded that the infection observed is what distinguishes Zika from other flaviviruses, and implied that this accounts for the equally distinguishing, and tragic, neonatal consequences of infection.

On 2016 Mar 13, Lise Bankir commented:

It may be useful to remind that loop diuretics act on the thick ascending limb upstream of the macula densa, whereas the thiazide-type diuretics work donwstream the macula densa. Moreover, loop diuretic block the sodium-chloride cotransporter in the cells of the macula densa. The tubulo-glomerular feedback that permanently regulates the GFR is thus abolished, allowing the GFR to go up. This produces a permanent "hyperfiltration" that is known to induce renal damage when sustained for long periods. Thiazide diuretics, acting beyond, the macula densa should not influence the GFR

On 2016 Mar 28, Atanas G. Atanasov commented:

Consumer Health Digest Scientific Abstract based on this paper is available at: https://www.consumerhealthdigest.com/general-health/controlling-obesity-balance-omega-6-omega-3-fats-ratio.html

On 2017 Feb 28, Tom Kindlon commented:

A major limitation was not mentioned: no objective outcome measures were used

I was amazed to read the long (906-word) limitations section and find no mention of the limitation that the results rely solely on subjective outcome measures[1].

The most obvious outcome measure to use would have been actometers which measure activity levels objectively. The equipment was available to the researchers as it was used at baseline ["Physical activity was assessed with an actometer, a motion-sensing device worn at the ankle for 14 days"].

The importance of the use of such a measure can be seen in the results of an earlier study using the the same or very similar intervention on people with chronic fatigue syndrome[2]. That study involved two of the current research team with one of them being its corresponding author. That paper reported improvements in the intervention group on the CIS fatigue severity, SIP8 total score and SF–36 physical functional questionnaires (which were also used in the current study). Subsequently to that, data from the actometers were reported in a paper co-authored by three of the current team[3]. Both the intervention group and the control group had the same change in activity, 4.3 units, during the trial. The intervention group finished at a mean of 67.8 units, significantly less than the actometer scores for healthy controls of 91.

Numerous response biases could be at play in this nonblinded study with such interventions causing participants to report improvements without their objectively-measured levels of functioning having improved.

If actometers were used during or after the current study, it is important that the researchers should now release such data, rather than delay for years as they have done with some trials before[3].

References:

1 Janse A, Wiborg JF, Bleijenberg G, Tummers M, Knoop H. The efficacy of guided self-instruction for patients with idiopathic chronic fatigue: A randomized controlled trial. J Consult Clin Psychol. 2016 May;84(5):377-88.

2 Knoop H, van der Meer JW, Bleijenberg G. Guided self-instructions for people with chronic fatigue syndrome: randomised controlled trial. Br J Psychiatry. 2008 Oct;193(4):340-1.

3 Wiborg JF, Knoop H, Stulemeijer M, Prins JB, Bleijenberg G. How does cognitive behaviour therapy reduce fatigue in patients with chronic fatigue syndrome? The role of physical activity. Psychol Med. 2010 Aug;40(8):1281-7.

On 2016 May 18, MATTHEW MESELSON commented:

Meiotic Sex in Bdelloid Rotifers

Debortoli et al. conclude that the patchwork pattern of sequences shared within the group of three isolates of the bdelloid Adineta vaga they sequenced is “…unlikely to arise in cases of PTH (Oenothera-like) meiosis since haplotypes are transferred as entire blocks…” and therefore that “Genetic exchange among bdelloid rotifers is more likely due to horizontal gene transfer than to meiotic sex”. But this assumes without justification that homologous HGT cannot occur in species with Oenothera-like meiosis, for which we have reported evidence in the bdelloid Macrotrachela quadricornifera (Signorovitch et al. 2015 Genetics 200: 581-590). And it does not take account of the possibility that gene conversion followed by outcrossing would also contribute to such a patchwork pattern, even in Oenothera-like systems.

Moreover, the set of three individuals studied by Debortoli et al., in which the shared sequences are considerably diverged, is not well suited to the detection of sex in an outcrossing population that may include numerous distinct Oenothera-like complexes. For that purpose, one should select individuals whose shared sequences are identical or nearly so in order to enrich for direct descendants of the F1 from a particular cross. Otherwise, further crossing is likely to replace shared complexes with others, removing the evidence for the transfer of entire haplotypes. It is therefore important to note that the shared sequences in the three individuals we studied were either identical or very nearly so, allowing us to observe the specific and unusual pattern of sharing expected for Oenothera-like meiosis.

Debortoli et al. suggest that HGT of very long fragments, rather than sexual transfer of entire haplotypes, may explain the pattern of sharing we observed. Considering the large size of the M. quadricornifera genome, some 1500 mb, and the fact that there are 10 chromosomes, it is exceedingly unlikely that the sequences from all four regions we studied reside on one of the horizontally transferred segments of DNA required by their suggestion and that the four allelic sequences reside on the other. Moreover, the results of FISH in other bdelloid species suggests that at least three and quite possibly all four regions we studied reside on separate chromosomes.

While awaiting full genome sequencing of the allele-sharing isolates of M. quadricornifera, present evidence argues strongly for the occurrence of sexual reproduction with Oenothera-like meiosis.

Ana Signorovitch, Jae Hur, Eugene Gladyshev, Matthew Meselson

On 2017 Jun 16, Jacob H. Hanna commented:

The last three papers from Smith group describing human transgene free "Reset cells", including this one, have failed to describe ability to generate teratomas. Mouse naive pluripotent cells have the intrinsic "self organizing capacity" to enter a formative/primed state after in vivo SC injection and make teratomas within 4-8 weeks. I find this stunning and wonder whether the human "reset" cells being induced do not qualify to be annotated as pluripotent cells at all. I hope the authors can clearly point out and directly address this critical caveat.

On 2016 Mar 12, Sudan Prasad Neupane commented:

Indeed a very important study. However, I found the conclusions were hardly based on the findings of the present study. In my opinion, the conclusions should be focussed on a deeper understanding of IPV in the studied setting, rather than drawing a sketch of possible interventions. Congratulations on a good study, we need more of these.

On 2016 Sep 06, Morten Oksvold commented:

Please note that part of this study is not reliable due to report of falsified/fabricated data in figure 2A:

"...Respondent falsified and/or fabricated the results in Figure 2A by erasure of a band in the blot image for LYST/CHD-4 that was present in the original data".

A full report has been published by ORI:

http://ori.hhs.gov/content/case-summary-cullinane-andrew-r

On 2017 Mar 12, Andrea Messori commented:

Promoting the use of Markov simulation models to study outcomes of total knee arthroplasty

Andrea Messori

HTA Section, ESTAR Toscana, Regional Health Service, Firenze, Italy

Correspondence to: Dr. Andrea Messori, PharmD, HTA Unit, ESTAR Toscana, Regional Health Service, Via San Salvi 12, 50100 Firenze, Italy. andrea.messori.it@gmail.com Fax: +39-05-74701319

In patients receiving total knee arthroplasty (TKA), simulation studies employing Markov models are increasingly being used [1-4]. The aim of these studies is to determine the “typical” clinical outcomes expected on the long term and to generate estimates of cost/effectiveness. If we consider these modelling tools, most of the simulation software published thus far shares the following characteristics:

a) Health states. The model implements, with minimal variations across different models, the health-states shown in Figure 1 along with the corresponding transitions from one health state to another. The probabilities of individual transitions are shown in Figure 1; these probabilities can be adjusted depending on the specific intervention under examination;

b) Clinical outcomes after TKA. The following outcomes can occur after the first surgery: i) successful outcome; ii) complications; iii) death; the same outcomes can occur also after a repeat surgery for arthroplasty revision.

c) Life expectancy. The life-expectancy attributed to the simulated patients is determined by considering: a) the age-related and gender-related mortality of a healthy population [5]; b) the mortality attributable to arthroplasty surgery. These two factors are separately managed in different sections of the Markov model (see Figure 1).

d) Utilities and estimation of QALYs. Utility of patients is assumed to be around 0.72 [6] after surgery. Over the pre-specified time horizon (e.g. 20 years), QALYs are computed on the basis of the health states of the model, their utilities, and the corresponding transition probabilities.

e) Discounting. The annual discount rate (e.g. 3.5%) is incorporated in the calculation of QALYs according to standard discounting techniques [6].

As regards the practical use of these computer programs, the simulation models published in the past years are essentially based on two software tools: on the one hand, some researchers have used a general-purpose spreadsheet (namely: Excel by Microsoft) to develop these Markovian programs; on the other, other researchers [3] have used a specific, commercial program (in most cases: TreeagePro by Treeage Software Inc., Williamstown, Massachusetts, USA). The Markovian subroutines written under Excel, as well as the tools developed under Treeage, share a negative characteristic because they are not freely available. Even NICE does not provide these tools when a Technology Appraisal is released. The unavailability of these programming tools is a serious hurdle that limits the scientific advancement of cost-effectiveness research on TKA. Hence, in the present report we have tried to facilitate the application of Markov models in the setting of TKA by developing a simulation software which is an improved version of the tools previously employed for specific research projects [3]. Our simulation model, that can be downloaded from the following link http://www.osservatorioinnovazione.net/papers/total_knee_arthroplasty.trex, is designed to be run under TreeagePro version 2011 (or subsequent versions). The input variables for the model are shown in the legend to Figure 1. The output of the program is represented by the estimate of total QALYs per patient accrued over the pre-specified time horizon. The software manages only the clinical part of these simulations; however, cost data can be added quite easily by introducing new sections of programming.

References

[1] Losina E, Walensky RP, Kessler CL, Emrani PS, Reichmann WM, Wright[ EA, Holt HL, Solomon DH, Yelin E, Paltiel AD, Katz JN. Cost-effectiveness of total knee arthroplasty in the United States: patient risk and hospital volume. Arch Intern Med. 2009 Jun 22;169(12):1113-21.

[2] Bedair H, Cha TD, Hansen VJ. Economic benefit to society at large of total knee arthroplasty in younger patients: a Markov analysis. J Bone Joint Surg Am. 2014 Jan 15;96(2):119-26.

[3] Pennington M, Grieve R, Black N, van der Meulen JH. Cost-Effectiveness of Five Commonly Used Prosthesis Brands for Total Knee Replacement in the UK: A Study Using the NJR Dataset. PLoS One. 2016 Mar 4;11(3):e0150074.

[4] Mari K, Dégieux P, Mistretta F, Guillemin F, Richette P. Cost utility modeling of early vs late total knee replacement in osteoarthritis patients. Osteoarthritis Cartilage. 2016 Dec;24(12):2069-2076.

[5] ISTAT. Tavole di mortalità della popolazione italiana—Ripartizione: Italia—Maschi—Anno: 2005, Report of 2010. http://demo.istat.it/unitav2012/index.html?lingua=ita (last accessed 7 May 2014).

[6] Mason J, Drummond M, Torrance G. Some guidelines on the use of cost effectiveness league tables. BMJ. 1993 Feb 27;306(6877):570-2.

[7] Jørgensen CC, Kehlet H; Lundbeck Foundation Centre for Fast-track Hip and Knee Replacement Collaborative group.. Time course and reasons for 90-day mortality in fast-track hip and knee arthroplasty. Acta Anaesthesiol Scand. 2017 Apr;61(4):436-444.

Figure 1. States of the Markov model and transition probabilities.

The starting point of the simulation model is a Markov node (circled M) from which six branches originate. The explanation for these six branches is the following: 1) surgery for TKA; 2) follow-up after first surgery (and also the occurrence of revision surgery): 3) follow-up after revision surgery; 4) follow-up after first surgery with complications; 5) follow-up after revision surgery with complications: 6) death. Second-level branches regard events defined according to the accompanying labels. The symbols adopted in this scheme reflect the syntax required by the Treeage software: Ο, probabilistic node;◄, terminal node.

Abbreviations: RWD, reward (which in this model represents the incremental increase in quality- adjusted survival).

The graph of Figure 1 can be downloaded from the following link: http://www.osservatorioinnovazione.net/tenders/tka.gif

On 2016 Mar 16, Daniel T Gilbert commented:

Our Technical Comment has elicited lengthy responses from several colleagues and counter-responses from us. For those who have not been following this conversation, here is a brief synopsis:

OPEN SCIENCE COLLABORATION: “We have provided a credible estimate of the reproducibility of psychological science.”

GILBERT ET AL: “No, you haven’t, because (1) you violated the basic rules of sampling when you selected studies to replicate, (2) you did unfaithful replications of many of the studies you selected, and (3) you made statistical errors.”

OPEN SCIENCE COLLABORATION & OTHERS: “We don't think we made statistical errors.”

Several colleagues wish to challenge our Point 3 while conveniently ignoring Points 1 or 2. But it requires no sophisticated mathematics to see that Points 1 and 2 are simple facts to which the OSC fully admits, and that these simple facts are by themselves sufficient to repudiate the OSC’s claim. We continue to believe that our Point 3 is correct, but even if it were entirely wrong, the conclusion that OSC2015 does not provide a credible estimate of the reproducibility of psychological science is inescapable, and it remains the one and only conclusion of our Technical Comment. Interested readers will find our full discussion HERE

On 2016 Mar 09, Daniël Lakens commented:

Invalid statistical conclusions in Gilbert, King, Pettigrew, and Wilson (2016)

Gilbert, King, Pettigrew, and Wilson (GKPW; 2016) argue that the Reproducibility Project (Open Science Collaboration, 2015) provides no evidence for a ‘replication crisis’ in psychology. Their statistical conclusions are meaningless due to a crucial flaw in their understanding of confidence intervals. The authors incorrectly assume that ‘based on statistical theory we know that 95% of replication estimates should fall within the 95% CI of the original results’. This is incorrect. When original and replication studies have identical sample sizes, 83.4% of confidence intervals from a single study will capture the sample statistic of a replication study. This is known as the capture percentage (Cumming & Maillardet, 2006).

GKPW use data from Many Labs (another large-scale replication project, Klein et al., 2014) to estimate the expected capture percentage in the Reproducibility Project when allowing for random error due to infidelities in the replication study, and arrive at an estimate of 65.5%. They fail to realize that the capture percentage for studies with different sample sizes (in the Many Labs project ranging from 79 to 1329) can be any number between 0 and 1, and can’t be used to estimate ‘infidelities’ in replications in general. Most importantly, the capture percentage observed for replications in the Many Labs dataset does not generalize in any way to the expected capture percentages between original and replication studies in the Reproducibility Project.

Nevertheless, GKPW conclude that the capture percentage in a subset of Reproducibility Project studies overlaps with the “the 65.5% replication rate that one would expect if every one of the original studies had reported a true effect.” Due to this basic statistical misunderstanding, the main claim by GKPW that ‘the reproducibility of psychological science is quite high’, based on the 65.5% estimate, lacks a statistical foundation, and is not valid.  

References

Cumming, G., & Maillardet, R. (2006). Confidence intervals and replication: Where will the next mean fall? Psychological Methods, 11(3), 217–227. http://doi.org/10.1037/1082-989X.11.3.217

Gilbert, D., King, G., Pettigrew, S., & Wilson, T. Comment on 'Estimating the reproducibility of psychological science', Science. (4 March 2016), Vol 351, Issue 6277, Pp. 1037a-1037b.

Klein, R. A., Ratliff, K. A., Vianello, M., Adams, R. B., Bahník, Š., Bernstein, M. J., … Nosek, B. A. (2014). Investigating Variation in Replicability: A “Many Labs” Replication Project. Social Psychology, 45(3), 142–152. http://doi.org/10.1027/1864-9335/a000178

Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716–aac4716. http://doi.org/10.1126/science.aac4716

On 2016 Mar 06, Sanjay Srivastava commented:

I have written about the analyses in Gilbert et al. Technical Comment elsewhere. Some key points:

(1) The comment proposes to define a "successful" replication as one where the replication effect is contained within the original study's confidence interval. However, it interprets this based on an incorrect definition of a confidence interval. Even more seriously in my view, the comment does not adequately address how using confidence intervals to gauge replication success will be affected by the power of original studies.

(2) The comment claims that high-powered replications have a high success rate, and bases this claim on Many Labs 1 (Klein et al., 2014), stating that ML1 had a "heartening" 85% success rate. However that is incorrect. Using the same replication metric Gilbert et al. define at the start of their comment and use everywhere else in their Technical Comment, Many Labs 1 had only a 40% success rate, which is similar to the Reproducibility Project.

(3) The analysis of replication "fidelity" is based on original authors' judgments of how well replication protocols matched original protocols. However, the analysis by Gilbert et al. combines 18 nonresponses by original authors with 11 objections, labeling the combined group "unendorsed." We do not know whether all 18 nonresponders would have lodged objections; it seems implausible to assume that they would have.

In my view these and other issues seriously undermine the conclusions presented in the Gilbert et al. technical comment. Interested readers can see more here: Evaluating a New Critique of the Reproducibility Project

On 2016 Mar 05, Dorothy V M Bishop commented:

My reading of this comment is that it maintains we should not expect high reproducibility for psychological studies because many are looking at effects that are small and/or fragile - in the sense that the result is found only in specific contexts. If that is so, then there is an urgent need to address these issues by doing adequately powered studies that can reliably detect small effects, and, once this is done, establishing the necessary and sufficient conditions for the effect to be observed. Unless we do that, it is very hard to distinguish false positives from effects that are genuine, but small in size and/or fragile - especially when we know that there are two important influences on the false positive rate, namely publication bias and p-hacking. I discuss these issues further on my blog here: http://deevybee.blogspot.co.uk/2016/03/there-is-reproducibility-crisis-in.html

On 2016 Aug 11, Eran Elhaik commented:

None

On 2016 Aug 09, Eran Elhaik commented:

A response to the criticism: Responding to an enquiry concerning the geographic population structure (GPS) approach and the origin of Ashkenazic Jews-a reply to Flegontov et al. by Das et al. (2016) is here: https://arxiv.org/abs/1608.02038

On 2016 Aug 04, Debbie Kennett commented:

Two critiques of this paper, from both a linguistics and a genetics perspective, have now been published:

1) Aptroot M, 2016 “Yiddish language and Ashkenazic Jews: a perspective from culture, language, and literature”. Genome Biol Evol. 2016 Jul 2;8(6):1948-9.

2) Flegontov P, 2016 “Pitfalls of the geographic population structure (GPS) approach applied to human genetic history: A case study of Ashkenazi Jews”. Genome Biol Evol. 2016 Jul 7. pii: evw162. [Epub ahead of print].

On 2016 Jul 01, David Reardon commented:

Once again, these researcher have failed to provide a breakdown of how a history of prior pregnancy loss (miscarriage or termination of pregnancy) effects mortality rates. This is a serious oversight since other studies have shown that a history of pregnancy loss is a significant risk factor for elevated mortality rates.

On 2016 Apr 14, Eduardo Eyraa commented:

This article did not cite a previous work Sebestyén E, 2015 where isoform swiches in cancer were already described between tumor and normal samples using TCGA data for 9 different cancer types; as well as between subtypes for breast tumors, lung squamous carcinoma and colon tumors from TCGA. In this previous article a specific switch in CTNND1 was already described for the basal breast tumors. If you are considering citing this publication, please consider whether you should also cite Sebestyén E, 2015.

On 2016 Jun 12, Christopher Southan commented:

Pointers to the F1000 recommendation and molecular mapping details at https://cdsouthan.blogspot.se/2016/06/bace1-inhibitor-post-for-f1000.html

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 NCT0239396. We believe the correct ID, which we have found by hand searching, is NCT02393976.

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 Mar 20, Amanda Capes-Davis commented:

STR loci used in today's STR profiling kits come from multiple chromosomes, and the technique can generate a full STR profile even in the presence of microsatellite instability. So I find the absence of STR loci in these profiles from CABA I puzzling.

Some additional testing is needed to further explore these findings.

1) It is important to perform separate species testing to confirm that CABA I is of human origin. STR profiling is typically considered species-specific, however, it has been clearly documented that related species can be detected. This has been documented previously by Almeida et al (http://www.ncbi.nlm.nih.gov/pubmed/22059503), Ren et al (http://www.ncbi.nlm.nih.gov/pubmed/22206866), and others. STR profiles generated from non-human species can produce patterns similar to those seen here.

2) Human cell line STR profiles have clearly defined quality criteria, including some requirements that are unique to cell lines (see ANSI/ATCC ASN-0002-2011 Authentication of Human Cell Lines: Standardization of STR Profiling). To my eye, the electropherograms seen here do not meet all quality criteria. It would be helpful to see other cell lines used as positive controls alongside CABA I data, to demonstrate adherence to quality criteria, in addition to the "typical male" and "typical female" results shown here.

On 2016 Apr 16, David Keller commented:

Your helpful explanations clarify the study protocol very well. I have deleted my erroneous comments which were based, as you correctly noted, on a misunderstanding. Thank you, Dr. Buse.

On 2016 Apr 05, John B Buse commented:

I believe that you have misunderstood the protocol. At randomization, those assigned to the combination of insulin degludec and liraglutide (IDegLira) stopped their glargine and started 16 dose-steps of IDegLira (16 units of degludec and 0.6 mg of liraglutide. They then titrated IDegLira twice a week based on their average fasting plasma glucose by -2, 0 or +2 dose steps of IDegLira aiming for a fasting plasma glucose of 72-90 mg/dl. The maximum dose of IDegLira was 50 dose steps (50 units of insulin degludec and 1.8 mg of liraglutide). The IDegLira patients did not continue the glargine. So, our conclusion is that for patients inadequately controlled on glargine 20-50 units, switching glargine to IDegLira is superior to continued titration of glargine. There is a remaining question as to what to do with a patient inadequately controlled on the maximum dose of IDegLira. That has not been studied as of yet.

On 2016 Mar 10, David Keller commented:

None

On 2016 Mar 09, John B Buse commented:

As explained in the paper, the comparison was made to glargine to examine the common clinical scenario of inadequately controlled diabetes treated with basal insulin. Glargine is the most commonly prescribed insulin formulation in the world. There are prior comparisons of IDegLira versus degludec in DUAL-1 (Gough, et al. Lancet Diabetes Endocrinol. 2014 PMID: 25190523) and in DUAL-2 (Buse, et al. Diabetes Care 2014. PMID: 25114296). There are also studies that have compared glargine to degludec head to head, e.g., Rodbard Diabet Med. 2013 PMID: 23952326. Thank you for your comment.

On 2016 Mar 09, David Keller commented:

None

On 2016 Jun 21, Jean-Michel Claverie commented:

An alternative interpretation to these results has been proposed in: Claverie JM, Abergel C. CRISPR-Cas-like system in giant viruses: why MIMIVIRE is not likely to be an adaptive immune system. Virol Sin. 2016 Jun 13. [Epub ahead of print] PubMed PMID: 27315813. see also: https://pubpeer.com/publications/3480B9DE6C9330B0747034C330BA6A

On 2016 Apr 03, Lydia Maniatis commented:

I don't see why this discussion is still going on. Anyone who has read PubPeer's blog posts will understand that a. they have solid arguments based on the public interest and thus that b. they have no reason to back down on their publishing model which c. is very popular for users and d. no one can make them do it against their will. End of story. MB's claims, on the other hand, turn a blind eye to important facts.

Given MB's general hostility to anonymity in the context of scientific discourse, I'm having trouble imagining how he rationalises the anonymity of reviewers of submissions for publication. Why isn't it a problem that the potential critic of the submission is cravenly hiding (as he might put it) behind anonymity? What's the danger of being up front?

On 2016 Apr 02, Boris Barbour commented:

None

On 2016 Apr 02, Boris Barbour commented:

The important issue here is the reluctance/refusal of Michael Blatt to engage in a substantive analysis of the pros as well as the cons of anonymous commenting, a recurring theme in this thread.

The questions about the negotiations to publish a reply to Michael's original editorial in Plant Physiology represent a distraction from the more fundamental issues. However, because he is creating the impression that we have been untruthful and have something to hide, I reluctantly respond again on this point.

Michael, as I said, we felt your initial suggestions were unfair. They did improve when we pushed back, as I have been happy to confirm. However, as I did not spread misinformation, I'm not apologising for it. Specifically, that you attempted to impose constraints that (at least we felt) were unfair is true, so I'm not apologising for having said that either.

You requested permission to publish our email exchange. We do so below.

Some context will be helpful in understanding why we did not reach agreement. Michael had just published a 3-page editorial in which he deployed a combination of insinuation and plausible deniability to associate us with notions such as voyeurism (peeping at published articles...), going through dirty laundry and money grabbing. A completely neutral editor-in-chief covering a controversial issue might have considered allowing us a reply of the same length, published at the same time (we must have missed the invitation to do so...) or as soon as possible afterwards. But Michael was in the conflicted position of also being chief prosecutor, having turned Plant Physiology into his personal propaganda vehicle (three editorials attacking PubPeer so far). Although there was a degree of mistrust and we were skeptical that he would be able to dissociate his conflicted roles, we decided to explore the possibility of replying to the same audience. As Michael was well aware, speed was of the essence, with any delay affording him the comfort of monopolising the "news cycle". From our point of view, truth was struggling to get her boots on, and any delay would reduce the effectiveness of our response.

I have edited email adresses, boilerplate (signatures and embedded emails) and some whitespace in the chain below (posted in one or more comments below because of PMC size limits). Emphasis and text in square brackets have been added by me.

On 2016 Apr 02, Michael R Blatt commented:

Boris

I shall take this as your apology for propagating misinformation and suggesting that I “tried to impose various unfair constraints” on any response from your and your PubPeer colleagues.

Thank you for your (eventual) candour.

Mike

On 2016 Mar 31, Boris Barbour commented:

Dear Michael,

There is no real contradiction on the format negotiations. After some to-and-fro, your final offers were indeed relatively generous given "journal constraints". But by that time we had come to realise that we didn't need to satisfy ourselves with the "halfway" we were working towards. As anybody who has tried to correspond with a journal knows, the process can feel extremely restrictive compared to the freedom and immediacy of a blog post.

Anyway, the point of the above comment was to correct rapidly three possible implications ambiguously left open (and predictably seized upon by a twitter denizen): i) that you'd offered to give us equal airtime, ii) spontaneously, and iii) that we hadn't felt able to counter your arguments. That's why I gave a bit more background about the process.

On 2016 Mar 31, Michael R Blatt commented:

Boris

Again, I think you do me a disservice. Given the constraints of publishing in a scientific journal, I did my utmost to meet you halfway and not limit your effectiveness (for example, engineering a way around the time lag between submission, acceptance, and final publication so that your response might be published instantly). The email thread I refer to above bares this out. Once more, I am happy to share it here with your approval (yes, vetos can work both ways).

As for any mis-reading of the latest editorial (or any of the others, for that matter), I can only say that it is always possible to take a statement out of context and twist it into someting altogether different, no matter how precise the text. The context in this case was of an offer to respond in Plant Physiology, nothing more or less. If this was misconstrued to imply that you declined to make any response whatsoever (which, clearly, is not the case as you've noted), I can only say that this was not my intention.

On 2016 Mar 31, Boris Barbour commented:

Dear Michael,

You made several suggestions ("tried to impose constraints") that would, coincidentally of course, have limited the effectiveness of our reply to your editorial: shorter, later, hobbled, elsewhere. I didn't invent the list of issues I gave (and the problem was of course your veto not ours). Sure, the restrictions weren't untypical of journal correspondence and the power that editors are accustomed to wielding. And, yes, we might have been able to work something out. But we decided it was just not worth the struggle when we could post instantly in our desired format. In one way we acknowledge that was a mistake, because it has proven exceptionally difficult to engage you in any discussion of specifics.

We didn't say that the statement about us declining to publish in Plant Physiology was wrong, we just felt that it might mislead people (just as it misled Leonid Schneider) into believing that we had avoided the debate. Those tempted by that interpretation are invited to read this thread, and, of course, our replies to your editorials.

On 2016 Mar 31, Michael R Blatt commented:

None

On 2016 Mar 31, Boris Barbour commented:

Michael Blatt has published yet another editorial attacking PubPeer, containing an incomplete and potentially misleading statement:

"An offer to respond had been made to Brandon Stell of PubPeer, who ultimately declined."

We at PubPeer requested the opportunity to put our case to the readers of Michael's original editorial. He agreed in principle but tried to impose various unfair constraints ("no more than 3 points", "limit on text", interleaved rebuttals, publication veto, etc). In addition, as the timing of the new piece shows, we might have had to wait 5 months and Michael's decision for our reply to appear. The process reminded us why journal correspondence sucks so much and indeed why PubPeer was created in the first place. So we decided to publish our response immediately as a blog post.

Readers of this thread can follow our largely unsuccessful attempt to draw Michael into a joint, open and even-handed evaluation of the pros and cons of anonymous post-publication peer review. Given his preference for preaching (several times) to a captive and passive audience, we shall just have to wait and see how scientists in general and the plant community in particular, which is by no means united on this matter, votes with its feet.

On 2016 Mar 23, Jaime A. Teixeira da Silva commented:

From my personal experience a PubPeer from what I have observed is that there are all kinds of anonymous: those with a desire to hold an academic discussion, as if in a journal club; those with valid, succinct claims; those with wild, but plausible claims; those with wild, and sometimes unsubstantiated claims; those with simple observations or concerns; those who have come to troll; those who have come to abuse, make libelous comments, or harass.

Comments by the last group tend to be flagged and removed by the moderator(s), who are likely Boris Barbour and the other two PubPeer management figures. But all others remain, which is what makes PubPeer so conflictual, because it has attractive and highly unattractive aspects.

One will never know the identity of all these types of anonymous commentators, and except for the use of extremely bad language, slang, or downright libelous name-calling (e.g. calling someone a fraud), we need this type of platform to allow a free level of discussion that is never possible with any journal's comment platform. Most scientists will know how to differentiate the wheat from the chaff, and can discern valid criticisms or concerns from noise, evasion or deflection. The most important thing is if what is written, either as a bounce from PubMed Commons, or directly here at PubPeer, has any value, and to whom?

In my opinion, PubPeer serves for me as a platform to begin to show how sad the state of affairs is in plant science. Comments might not always be perfect, or tone-perfect, and you will find that will ultimately always create enemies or irritate those who oppose you, or your ideas. But this is a risk that comes with using an anonymous tool. Those who use PubPeer should know that these risks exist.

I think the anonymous vs named argument is a dead horse. It is quite obvious that there are three groups: those who understand, and appreciate, anonymity; those who will always be skeptical and critical of it, and ultimately shun it; and those who see some benefit, and also some risk, but who would likely never venture to use it, either because they are of a traditional class of scientists/editors, or because they fear.

I think that ultimately that what is lacking is the respect and recognition of one of these groups of the other two. And because there is a lack of recognition and/or respect, there will always be frustration and passionate defense of the home turf opinions. That is so evident in the responses by select members of the public or scientific community to Prof. Blatt's two editorials.

I can personally see where Prof. Blatt's fears and concerns are coming from, and I respect his opinion and point of view, because that's all the editorials represent. I might not necessarily agree with his views in their entirety, but I understand that we need to respect his position, or at worst, respect his position in a civil way. Ultimately, one has to ask: has Blatt been a valuable asset to the plant science community, even if within his own restricted niche at Plant Physiology, and has something positive come from these two editorials?

The answers to these two questions are more than evident.

I thus suggest a new trajectory, at least for plant science. PubPeer has shown, in already hundreds of cases, that there are problems with the plant science literature. Problems that neither leaders like Blatt, Kamoun, or Zipfel knew or detected. But problems that ultimately drew them into the conflict that is, broadly speaking, a literature that is problematic, even in the top level plant science journals.

We only need two things to make this recipe of correcting the literature work: a) the recognition that there are problems and that they need to be corrected; b) action, i.e., getting editors and publishers to recognize these errors formally, and correcting the ills of the traditional peer review process.

Unless a) and b) take place, this whole discussion surrounding the anonymous voice is meaningless.

In closing, I should add that not all anonymous commentators are the same, and that not all necessarily agree with the position, or choice of words, employed by Boris Barbour or PubPeer.

On 2016 Mar 28, Jens Sommer commented:

Dear Michael,

I know about the complications of double blind review and disclosure of reviewers. As the reviewers see the reference list and self reference is part of scientific writing it is almost imposible to hide author's identity, so most authors don't care.

About disclosure of the reviewers: It is not essential to insist on disclosure. Let the reviewers decide. In addition allow the authors to rate the quality of the reviews (anonymously?).

As long as we want to improve our knowledge (and scientific progress) we will need skilled reviewers, not just many reviewers. But again, this has been discussed before.

On 2016 Mar 22, Michael R Blatt commented:

Dear Jens,

Forgive me for not responding to your last paragraph. These issues have been addressed time and again (e.g. in my editorial and in the discussion below with Boris Barbour).

As for your two, numbered points, you may be aware that several journals have tried and/or do offer a double-blind review process, including several of the Nature journals. Only a very tiny percentage of authors ever take up this option, however, and realistically it is often difficult to hide the authors' identities (see the editorial from Chris Surridge in Nature Plants last September for more information).

Complete disclosure, as you propose in your second point (and if I understand you correctly), is also problematic. I think most editors would argue that, were they to insist on such disclosure, then it would be very difficult indeed to secure reviewers. Of course, editors are generally acknowledged; all journals publish the list of their editors on the journal masthead and some journals include the names of the handling editor with each published article (e.g. PNAS). As for the social contract involved in considering a manuscript for publication, I have commented on this in my editorial of last October.

On 2016 Mar 21, Jens Sommer commented:

Thank you for the editorials and thanks to all comments. This gives hope for the future of the scientific community and scientific progress.

Maybe it is just the point in time, when we need or accept anonymous comments.

Is it important to have

1) a double blind review process (authors, reviewers) and 2) a complete disclosure (authors, reviewers and editors) after rejection or acceptance?

While the first is essential to get an unbiased review, I expect the second to improve the quality of reviews and thus the quality of articles. At least my idea of reviewing an article is to improve it, and the communication with the authors is more like an anonymous discussion.

As the review process includes more than one reviewer it is interesting to see that the process sometimes fails completely (false acceptance). So if it takes time to review an article properly, why shouldn't we see the names of the reviewers, who supported the authors in getting their work published?

Finally, when the article is published and a discussion starts any reasonable comment will be welcomed - anonymous or named. Do we really need regulation if there is more than one free platform with high-quality comments and good usability? Why don't we let people figure out what suits their needs best?

On 2016 Mar 23, Daniel Corcos commented:

I would be grateful if you could show me potentially toxic comments. As for the toxicity of anonymous reviewers and bad editor choice, I know too many examples, but I only have to mention the case of CRISPR role discovery by Francisco Mojica, which has been rejected by many high impact journals for 2 years (http://www.cell.com/cell/pdf/S0092-8674(15)01705-5.pdf). You may say that two years delay for a basic paper does not harm much, but when it comes to medicine, it can be terribly harmful.

On 2016 Mar 18, Michael R Blatt commented:

Daniel, I guess we will have to disagree on the toxicity potential of anonymous comments.

On 2016 Mar 18, Daniel Corcos commented:

Mike, I agree that PubPeer comments can be wrong and misleading, but their advantage is that they can be seen by everybody. I prefer anonymous comments that I can read to hidden criticism. Rejecting a paper for spurious motives makes certainly more harm than comments in PubPeer. I must say that I am not in favor of anonymity and I hope that this debate will lead to openness of review. With time, this would allow a full evaluation of the harm done by some renowned scientists to the progress of knowledge.

On 2016 Mar 18, Michael R Blatt commented:

Daniel, I agree that overtly offensive comments are usually obvious as such to the reader. What is much more worrying about anonymous commenting is its potential to spread untruths and to do so without accountability. So I cannot agree with you that anonymous commenting is always bland and harmless. Subtle rumours can have just as serious consequences for an individual as an undue rejection.

On 2016 Mar 17, Daniel Corcos commented:

Mike, we are often aware that major breakthrough papers had been initially rejected by many journals, especially when the authors were not renowned. It would be interesting to know who were the experts and if the editor has ceased to ask them to review papers after considering that rejection was undue but for ordinary people like us, reviewers remain anonymous. On the other side, offensive anonymous comments have no great consequences because readers can judge by themselves, whereas undue rejection has much more consequences for the authors and for science.

On 2016 Mar 17, Michael R Blatt commented:

Daniel, it is a common mistake to equate blind (confidential) peer review with anonymity. There is a world of difference here. In assessing the potential of a manuscript for publication, an editor will often turn to one or more known experts in the field for their opinions. The editor will know the identity and expertise of these individuals, so their advice is most certainly not anonymous. Please have a read of my editorial from October 2015.

Of course, we might discuss the pros and cons of open (non-confidential) review; however, this is not the same discussion as that of anonymity. Mike

On 2016 Mar 14, Daniel Corcos commented:

If anonymity "has no place in scientific critique" as Blatt argues, then it should have no place in peer review. For this reason, Blatt's position is untenable.

On 2016 Mar 22, Michael R Blatt commented:

As you will have noted, I too am in favour of open (non-anonymous) debate.

On 2016 Mar 19, Lydia Maniatis commented:

I'm against artificial and unnecessary obstruction of open scientific debate, either via selection by a self-proclaimed "meritocracy" or any other means. Open debate is messy but the alternatives, arguably, are messier.

On 2016 Mar 17, Michael R Blatt commented:

Hello Lydia. Do I understand you correctly, then, that you are in favour of meritocracy in science?

Mike

On 2016 Mar 17, Michael R Blatt commented:

None

On 2016 Mar 08, Lydia Maniatis commented:

I would like for the moment to single out the following argument/counterargument from the article, because it argues that science is not/should not be democratic.

View attributed to those in favor of anonymity: “Anonymity is essential to protect fundamental rights and free speech in a global democratic society.”

Blatt's rebuttal: “Yes, science is a “massively cooperative undertaking,” to quote one of my PubPeer commenters,1 but that does not mean it is democratic. Science requires substantial training; its foundations are logic and reasoning; it builds on the merits of knowledge and expertise; it is not a ‘one man, one vote’ endeavor with universal enfranchisement. To argue otherwise is manifestly absurd. “

First, arguments appealing to “manifest absurdity” are not worthy of scientific debate, whether signed or anonymous. Second, logic and reasoning are the property of every human being, and their use is very often and very demonstrably absent from the scientific literature (which is the reason editors so fiercely protect the published literature from dangerous “Letters to the Editor.”) Third, there is no degree that can confer infallibility to any individual; likewise, no individual should be denied the right to have their arguments evaluated ON THEIR MERITS (something very different from the misleading "one man one vote" argument.)

People who don't feel comfortable with the responsibility to defend their positions (scientific and otherwise) by argument and not on the basis of membership in a closed society of initiates do not understand how science progresses, and how it stalls.

On 2016 Mar 08, Lydia Maniatis commented:

I think that the “warped worldview” arguments made by Dr. Blatt and others in opposition to anonymous critiques could more appropriately be levelled at them. Their main concern seems to be that the material welfare and (relatedly) personal reputation of individuals within the scientific community will be threatened by anonymous trolls whose only aim is to sully reputations by suggestive but ill-founded attacks. (The critics of anonymity seem less concerned about the benefits to the public interest that PubPeer's editors have demonstrated and documented to have followed from the enabling of anonymous posting).

Let's assume that such trolls exist and are even in the majority (though I don't believe this to be the case). How, in the context of a healthy, intellectually sharp, critically-minded scientific community will their efforts have an influence? Why would the targets' astute colleagues, grant reviewers, academic employers, etc. allow specious, unmerited criticism to influence their views or choices? If, on the other hand, decision-makers in the community are not equipped to separate the wheat from the chaff (whether we are referring to criticism of scientists or the scientists academic productions), then this is indeed a warped world, and in such a world the documented public interest value of anonymous criticism surely outweighs any nuisance value to individuals. Relatedly, Dr. Blatt states early on in his editorial that he is “Putting aside the issues of policing for fraud and whistleblowing for the moment....” I would be interested if he would come back to this issue, and particularly in his plan for creating a system where anonymity could be automatically enabled for comments falling in the category of “policing for fraud and whistleblowing” while denying it for silly comments. Who would decide, a priori, which commenters/comments get the privilege?

On 2016 Mar 08, Lydia Maniatis commented:

Blatt says: "Ultimately, it is a warped worldview, indeed, in which scientists are so fearful of engaging that they never challenge others’ research and ideas openly, whether online or in publication."

Barbour notes that: "Plant Physiology has no functional feedback mechanism..."

Perhaps Dr. Blatt should consider helping to unwarp the world by enabling signed, open publication of criticism in his journal...

On 2016 Mar 28, Michael R Blatt commented:

Dear Boris:-

Thank you, but no apologies are necessary.

However, I think we are now going around in circles. You continue to use the ‘volume’ argument which I am not prepared to accept and, if you think about it, I suspect neither are you. As devil’s advocate, I could point out that the volume of good research still overwhelming outweighs the bad (see the references in my first editorial and further discussion in the editorial to be published next week), just as you argue in favour of the “overwhelming majority” of PubPeer comments that you claim are useful compared to the “tiny minority” that are antisocial, ethically unsound and/or defamatory.

You will see my point, I hope. So let’s not beat this one to death. We are not going to resolve the problem by defending corners or looking for the lowest common denominator. I am convinced that to find a solution it will be necessary to look outside the box, so to speak.

I’m happy to continue our discussion, but I don’t think anyone is particularly interested in following this thread much further. So I suggest we now do so by email. If we do come to a solution, then of course we will want to share this with the community, either through PubMed Commons or some other way.

Mike

On 2016 Mar 27, Boris Barbour commented:

Dear Michael,

I apologise if I have misconstrued your position, which I thought was a good deal more negative.

Anyway, let's work on the common ground a little.

In favour of anonymous comments, some disseminate useful information that: 1) reduces wasted research based on unreliable research, 2) diminishes errors in clinical trials and medical guidelines arising from unreliable publications, 3) therefore saves careers, taxpayers' money and lives.

Against anonymous comments ("abuse, misrepresentation, sock-puppetry, and other antisocial or ethically unsound behaviours"): 4) unjustified denigration of reputations, 5) no declaration of conflicts of interest, 6) no information about commenter's status, 7) no discussion of equals

Let's weigh the "costs" and "benefits" of the anonymous comments on PubPeer as they are; we'll worry about how to influence their nature later. So do 4-7 outweigh 1-3, taking into account their relative frequencies?

I would say that the most extreme negative outcome is damage to somebody's reputation. But how much damage can be done without convincing ammunition? Remember also that researchers can always defend themselves by explaining, showing data etc, in the case of a truly unfortunate misunderstanding. So, even if reputations probably can be damaged slightly by the ill-intentioned, I would contend that it is difficult to cause severe unjustified damage to somebody's reputation on PubPeer.

In contrast, it is highly likely that rapid dissemination of information can save a PhD or post-doc from wasting 6 months to a year trying to build on some exciting but unreproducible result. In today's competitive environment that unproductive time may spell the end of a young career, and taxpayer's money will have been poorly spent. There are no doubt clinical trials in progress based upon flawed research - deeply unethical - and there are - hopefully rare - cases of flawed research causing erroneous medical decisions. In these cases, rapid dissemination of information could save lives. Examples where one wishes information had been made public (and acted upon) earlier include the Poldermans case mentioned in our blog and the Wakefield MMR vaccination scandal.

So in terms of extreme outcomes, do you agree that saving lives, taxpayers' money and research careers outweighs slight damage to researchers' reputations?

What about frequencies of the different types of comments? From having read nearly all of the ~50000 comments that have appeared on PubPeer over 3.5 years, I am happy to report that the overwhelming majority report valid signs of low-quality research or misconduct - the sort of comment that could lead to benefits 1-3. Only a tiny minority might be suspected of trying to run down the reputations of other researchers unfairly.

Based upon the importance of disseminating information to readers and the observed low frequency of comments appearing to abuse the system, we have concluded that the anonymous comments appearing on PubPeer are very clearly beneficial on average. Therefore they should be encouraged. Do you agree?

A question that I would like to keep separate and analyse next is what can be done to tilt the balance further towards beneficial comments (including your desire to convert anonymous to nonanonymous comments).

On 2016 Mar 25, Michael R Blatt commented:

Dear Boris,

There’s nothing grand in these statements nor are they exempt from a cost-benefit analysis. It just happens that my measures of cost and benefit are (obviously) different from yours. Of course, it may be that we can still find common ground, and I would hope this is the case.

As to your question “Is it a good thing to alert readers … to possible problems?”, clearly the answer is yes. I have said so repeatedly in my editorials and here on PubMed Commons. However, in my opinion, this needs to be done in a way that does not open the door to abuse, misrepresentation, sock-puppetry, and other antisocial or ethically unsound behaviours. I don’t think this is a particularly difficult concept, even if its solution is more complex in practice.

On 2016 Mar 22, Boris Barbour commented:

Dear Michael,

You appeal to "foundational arguments" and "principles", but sounding grand doesn't exempt them from a cost-benefit analysis.

I'll ask you just one question, the one you have avoided answering over 2 editorials and all the discussion here: is it a good thing to alert readers of publications to possible problems?

We could call it the foundational principle of PubPeer...

COI statement: see my original post.

On 2016 Mar 22, Michael R Blatt commented:

Dear Boris,

I really do not think that we are so far apart in our views. We both are dismayed by some of what we see in scientific publishing and communication today, and we both want the same for the scientific community as a whole. Where we differ is only in some details of the means to this end.

The point you raise is of measures, ‘averages’, and quantity, rather than of principle. I do not doubt that there are many comments on PubPeer that are thoughtful and constructive. I certainly never suggested that all comments on PubPeer “abuse the system” (nor did I ever suggest coersion, so let us not confuse the issue here). The point on which we differ is whether the quantitative argument for anonymity that you pose outweighs the foundational arguments I have set out against it. I think not.

You raise the analogy to the utility of cars and whether these should be banned. Of course analogies are poor vehicles (pun intended) for ideas, but let’s follow it for a moment. It would be virtually impossible to ban anonymous commenting from social media, just as it is impossible to ban reckless driving (I recall you had this discussion with Philip Moriarty previously). However, this is not to say that either should be actively encouraged. There are norms for interpersonal interaction that we generally follow and that protect civil society (e.g. accountability), just as there are rules of the road and legal requirements (e.g. the need for a driving license) that are there to protect us when we are on the road.

I think it is always important to look for other ways to a solution. Answers sometimes come from taking an entirely different perspective rather than looking for the common denominator. So, to follow your analogy one step further, rather than banning cars (and anonymous commenting), would it not be better to make them less attractive as a whole while making the use of public transport (and of open, accountable commenting) more attractive? Are we not both in a position to influence the process of PPPR?

I alluded, at the end of my March 2016 editorial, to what I hope will be an approach to such a ‘third solution.’ It comes straight out of discussions with Leonid Schneider who, I think you will recall, was originally one of my fiercest critics last October. I am convinced this alternative is worth a try and, at this point, have a number of my opposite numbers from other publishers on board. You may be convinced as well in due course. Again, I hope that I will have much more to say on this matter later this year.

On 2016 Mar 19, Boris Barbour commented:

The key issue is whether or not anonymous comments are beneficial ON AVERAGE. If they are or can be made so (PubPeer implements guidelines to favour useful comments), then such comments should be encouraged. Your arguments focus purely on the negatives and you systematically avoid consideration of the benefits of comments that happen to be anonymous. We can agree that it is possible to abuse anonymous commenting. And anonymity does enable commenters to forego in-depth discussion with meritorious professors. However, a balance needs to be struck and you have still made little attempt to do that. Thus, even if abuse is possible, that doesn't mean that all comments do abuse the system. In fact, the great majority of anonymous comments on PubPeer are perfectly factual and some highlight matters of genuine importance to readers, disseminating that information without delay. At PubPeer we have weighed both the advantages and the disadvantages of the anonymous comments we publish. We are convinced that their overall effect is overwhelmingly beneficial, despite a small number of awkward cases. So we shall continue to enable anonymous commenting.

Having been around the houses of this argument a few times without making much progress, maybe an analogy will be helpful. Would you ban cars because sometimes people get run over? Or would you take into consideration the fact that they are a useful means of transportation? I'd like to see you take into consideration the potential benefits of the content of anonymous comments.

We agree that we should all strive to create a system in which researchers feel able to comment freely and transparently. But we don't have a magic wand to create that environment. You at least are in a position of power to implement some changes, but that will require supportive and constructive action, not coercion. The coercive approach has failed in the past: our direct experience on PubPeer has shown that many useful comments will only be made if anonymity is available. In other words, there is no way to make all useful commenting non-anonymous, you can only suppress the majority of comments, including many useful ones, by (hypothetically) forbidding anonymity.

You continue to confuse research that contains known flaws (including overinterpretation) when produced with that which doesn't. Although all research is indeed potentially, eventually falsifiable, the use of small sample sizes, inappropriate statistics, unverified cancer cell lines etc, etc (the list is long) is known today to generate unreliable research. You can't expect researchers to predict the future, but it's not unreasonable to ask them to avoid known mistakes (respecting the "state of the art"). Moreover, isn't it precisely your job as a journal editor to draw this line? Do you really not recognise this distinction? In any case, PubPeer simply allows comments and questions; the site makes no judgement.

Regarding the arsenic life paper, I'll leave you, as a practising editor-in-chief, to interpret the (admittedly inconsistent) COPE guidelines on the matter. Here are a couple of key quotes:

"Journal editors should consider retracting a publication if ... they have clear evidence that the findings are unreliable, either as a result of misconduct (e.g. data fabrication) or honest error (e.g. miscalculation or experimental error)."

"Retraction should usually be reserved for publications that are so seriously flawed (for whatever reason) that their findings or conclusions should not be relied upon."

COPE guidelines

Finally, we obviously agree that "science does not end with publication". Nobody at PubPeer has ever said otherwise. Indeed, the whole raison d'être of the site is to enable science to continue after publication, something that traditional journals have not always embraced.

COI statement: see my original post.

On 2016 Mar 18, Michael R Blatt commented:

I’ll reply to both your comments here, Boris. I did address all of the standard, conceptual arguments around anonymity in my second editorial, and I will be discussing some of these and other aspects of anonymity again next month with Jaime Teixeira da Silva.

I believe you wish to point out, as your central argument, that the traffic on PubPeer is far greater than on PubMed Commons, for example, and you ascribe this to encouraging anonymous comments. Your numbers may be correct – I am not in a position to comment one way or the other – but I do dispute your underlying assumption that traffic volume equates with scientific value. I raised this point in my October 2015 editorial, as did Philip Moriarty both in the PubPeer threads that followed and in his discussion with you in the Times Higher Education in December 2015.

I maintain, furthermore, that anonymous commenting encourages grubby comments and nefarious behaviours that undermine the very scientific community you want to build. So, in my opinion, encouraging anonymous commenting is counterproductive and, ultimately, self-defeating. Again, you will find all of the arguments in my editorials, so I’ll not revisit them here.

As for my misinterpreting your definition of ‘ultimately unreliable’ “in the sense of ‘unreproducible’, ‘low-quality’, ‘known to be wrong’, [and] ‘overinterpreted’”, I agree there is such a thing as ‘bad science.’ Ben Goldacre has much to say about this. However, I think you need to be more cautious in calling for sweeping retractions on the basis of your definitions. Can you issue a blanket statement of unreproducibility without first seeking to reproduce each set of data and explaining why it is unreproducible, for example? Where do you draw the line between interpretation and overinterpretation? And when does overinterpretation become grounds for vilification?

Again as a specific example, I agree that the Science paper on arsenic-based life was overinterpreted and included experimental methods that were insufficient to meet the exacting standards expected for such a claim. On this basis alone there is a strong argument to say that it should never have found its way into the journal. However, my understanding is that the results were not low-quality per se; they were demonstrably reproducible; and they did lead (ultimately!) to detailed knowledge of a transporter with a remarkable selectivity for phosphate over the structurally similar arsenate anion. As I noted before, “science does not end with publication. Publication is only the beginning of scientific debate. Progress often arises from what, in hindsight, is ‘ultimately unreliable’, and its cornerstone is open debate.”

On 2016 Mar 07, Boris Barbour commented:

The words 'ultimately unreliable' in the title of this editorial are a quote from the text of our blog Vigilant scientists. By accident or by design, Blatt deforms their meaning completely. We meant "unreliable" in the sense of "unreproducible", "low-quality", "known to be wrong", "overinterpreted", while "ultimately" simply added emphasis (meaning something like "most importantly"). In contrast, Blatt interprets the word pair to include the meaning "eventually improved upon". In other words, we were discussing research that is of low-quality or wrong according to the state of the art at the time of publication, while he lumps such poor work with outstanding research containing no known defects at publication but upon which even greater discoveries are subsequently built. Thus, he gives the example of Hodgkin and Huxley's explanation of the action potential building upon Cole and Curtis' measurements of axonal impedance, characterising the latter authors' work as "ultimately unreliable". Nothing could be further from our intended meaning, which should have been abundantly clear from the context of the blog. In particular, we used the terms "unreliable" and "unreproducible" interchangeably, gave numerous examples and references relating to unreproducible and low-quality research, and gave no examples of the sort Blatt mentions.

So there is a clear criterion of reliability that Blatt did not consider: does a paper contain known problems according to the state of the art at the time of publication? Papers that fail this test are "unreliable" and the relevant information should be disseminated to the readers. Applied to examples in the editorial, this test would classify the arsenic life paper as unreliable and Cole and Curtis as reliable, while Blatt considers both to be "ultimately unreliable". Coming from the editor-in-chief of a high-quality journal, this seems to be questionable relativism.

COI: see previous post.

On 2016 Mar 06, Boris Barbour commented:

This editorial by Michael Blatt, editor-in-chief of Plant Physiology, follows up a previous one, Vigilante Science; both attack the anonymous commenting enabled by PubPeer (see "COI" below). PubPeer has responded to both editorials at Vigilant scientists.

In his follow-up, Blatt completely avoids addressing our central argument in favour of anonymity, which is that our priority as a community should be to disseminate information about publications to readers and users as rapidly and as widely as possible, a process encouraged by anonymity. As Plant Physiology has no functional feedback mechanism and because Blatt has refused to join any discussions on PubPeer, maybe he would like to respond here, at least to address our principal argument in favour of anonymous commenting?

Potential conflicts of interest: I am a co-organiser of PubPeer and wrote most of their two blogs on this subject. These views are expressed in a personal capacity, not as an official PubPeer position.

On 2016 May 27, Eduardo ANGLES-CANO commented:

Comment on "We hypothesized that low factor XII reduces kallikrein formation and consequently the release of bradykinin..." I suggest an alternative explanation, to the oedema hypothesis; a low bradykinin may results in insufficient stimulation to release tPA by the endothelium leading to inefficient thrombus lysis. This is particularly pertinent if we consider that thrombus persistence is finally due to an insufficient fibrinolytic response.

On 2016 Mar 15, Wichor Bramer commented:

Well, one can imagine, as there are 120 reviews, some have limited their dataset to English language articles only, others have translated foreign language articles. Likewise, some reviews I performed the searches for have included non published articles from registries, where they do make an important difference compared to reviews that only included published articles (such as Jaspers L, 2016, but that is not used for this research).

Some reviews excluded conference papers (especially if the number of hits was high in the reviewers eyes, we resort to using that to reduce the number of hits), others included them. I must say that I don't see why these would not be found in embase/medline, as this is particularly a problem when searching embase, while medline hardly includes detailed conference proceedings.

In this research we only looked at the included references that had been published in a journal, and we considered conference proceedings, published as supplements to journals to fall into that category.

Regarding searching cochrane central, this results will be shown in upcoming articles from partially overlapping data, i must say that sofar for the 2500 included reviews of 60+ Published reviews Cochrane Central has not identified one single included reference that was not also retrieved by another database.

In my opinion, when doing a systematic review, the authors should aim to find all relevant articles that can answer the research question. If that is not the goal, then it should not be called a systematic review, they can combine three MeSH terms in PubMed, extract some conclusions and automatically generate a rapid review.

On 2016 Mar 15, Hilda Bastian commented:

Many thanks, Wichor and Dean - that's really helpful. Still not clear on whether there was a language restriction or not. I looked at a couple of the reviews you link to (thanks!), but couldn't see an answer in those either.

On the question of implications for reviews: being included is a critical measure of value of the search results, but with such major resource implications, it's not enough. One of the reasons more detail about the spread of topics, and the nature of what was not found is important, is to explain the difference in these results compared to other studies (for example, Waffenschmidt S, 2015, Halladay CW, 2015, Golder S, 2014, Lorenzetti DL, 2014).

Even if studies like this don't go as far as exploring what it might mean to the conclusions of reviews, there are several aspects - like language - that matter. For example, the Cochrane trials register was searched and other places as well. If studies were included from these sources based only on abstracts from conference proceedings for example, then it's clear why they may not be found in EMBASE/MEDLINE. Methodological issues such as language restriction, or whether or not to include non-journal sources, are important questions for a range of reasons.

One way that the potential impact of studies can be considered is the quality/risk of bias assessment of the studies that would not have been found. As Halladay CW, 2015 found, the impact of studies on systematic reviews can be modest (if they have an impact at all).

Disclosure: I am the lead editor of PubMed Health, a clinical effectiveness resource and project that adds non-MEDLINE systematic reviews to PubMed.

On 2016 Mar 15, Wichor Bramer commented:

Dear Hilda,

Thank you for your insightful comments, much appreciated. I have left comments via PubMed Commons before, but have never received any from other researchers. I will respond to your comments point by point:

1) As we described in the last line of the second to last paragraph of the methods section of our paper, we searched all three databases post-hoc for included references.

2) We searched the largest Ovid Medline files comprising Ovid MEDLINE® In-Process & Other Non-Indexed Citations. For clarity for endusers at Erasmus MC this is the only Medline database shown, and it is referred to as Medline, though it includes non-Medline PMC records. Articles retrieved from PubMed, the subset as supplied by publishers, were not classified as resulting from Medline Ovid searches, but rather as unique results from PubMed publisher subset (a classification not used in this article, but that will be used in other articles from partially overlapping datasets).

3) As you pointed out, Bramer WM, 2015 is not a systematic review. After article acceptance, I realized it would have been wise to limit our study to medical research questions only (this being the only non-medical topic). Not all 120 searches have resulted in published systematic reviews. In some cases, the process is is ongoing and in others results were used to create other end products, such as clinical practice guidelines, grant proposals and chapters for theses. In 47 of the searches used in this research the resulting articles have been published in PubMed. That selection can be viewed via http://bit.ly/bramer-srs-gs.

4) Criteria for searches to be included this research were that

a) researchers had requested librarian-mediated searches because they intended to write a systematic review (in that view, the title should be read as 120 systematic review requests)

b) titles and abstracts for the results for all databases had been reviewed

c) the full text of the relevant abstract had been critically read and

d) the resulting relevant references had been reported to us or were extractable from the resulting publication.

Whether the searches result in finished published systematic reviews is independent of the search process. Retrospectively, it would have been wise to include a paragraph on this in the article.

5) One of the peer reviewers also mentioned the expected difference between certain topics, and advised us to investigate that relation. However, it would be very complicated to group 120 unique and diverse topics systematically and even within broad subjects such as surgery or pediatrics one can expect variation between research questions. For very distinct topics such as nursing or psychology one can expect differences, because of the need to search Cinahl, respectively PsycINFO, but these research topics were scarce among our set. We do not believe huge differences were to occur regarding the performance of GS between different topics, as the overall performance remains too low. We did observe that for uncomplicated questions GS performed better than for search strategies with many synonyms.

6) We chose not to investigate in detail what the missed studies would have meant for the conclusion of the reviews. Partially because of the vast number of topics, but also because we feel this does not add value to our conclusion about coverage, precision and recall. If searches in GS were likely to find fewer than 40% of all relevant references, or in Embase a high likelihood that fewer than 80% were retrieved, expected recall is too low for the systematic review, no matter what the quality was of retrieved results. In follow-up research where best database combinations are compared (in that case for published medical systematic reviews, so only partially overlapping with this set) we plan to investigate in detail why certain references were found by GS but not by traditional databases. One of the reasons could be that articles are retrieved from lower quality journals, as GS lacks quality requirements for inclusion, however there can be other reasons.

Kind regards,

Wichor Bramer

On 2016 Mar 12, Hilda Bastian commented:

An interesting and very useful study of Google Scholar (GS). I am unclear, though, about the methods used to compare it with other databases. The abstract includes this step after the systematic review authors had a final list of included studies: "All three databases were then searched post hoc for included references not found in the original search results". That step is clearly described in the article for GS.

However, for the other 2 databases (EMBASE and MEDLINE Ovid), the article describes the step this way: "We searched for all included references one-by-one in the original files in Endnote". "Overall coverage" is reported only for GS. Could you clarify whether the databases were searched post hoc for all 3 databases?

I am also unclear about the MEDLINE Ovid search. It is stated that there was also a search of "a subset of PubMed to find recent articles". Were articles retrieved in this way classified as from the MEDLINE Ovid search? And if recent articles from PubMed were searched, does that mean that the MEDLINE Ovid search was restricted to MEDLINE content only, and not additional PubMed records (such as those via PMC)?

There is little description of the 120 systematic reviews and citations are only provided for 5. One of those (Bramer WM, 2015) is arguably not a systematic review. What kind of primary literature was being sought is not reported, nor whether studies in languages other than English were included. And with only 5 topics given, it is not clear what role the subject matter played here. As Hoffmann T, 2012 showed, research scatter can vary greatly according to the subject. It would be helpful to provide the list of 120 systematic reviews.

No data or description is provided about the studies missed with each strategy. Firstly, that makes it difficult to ascertain to what extent this reflects the quality of the retrieval rather than the contents of the databases. And secondly, with numbers alone and no information about the quality of the studies missed, the critical issue of the value of the missing studies is a blank space.

Disclosure: I am the lead editor of PubMed Health, a clinical effectiveness resource and project that adds non-MEDLINE systematic reviews to PubMed.

On 2016 Jun 28, GRAHAM COLDITZ commented:

Important data here directing us to further study of in utero exposures and prostate cancer risk. Stronger results from measured weight than self-reported birthweight points to the underlying mechanisms likely being more important than previously thought.

On 2016 Mar 07, Martine Crasnier-Mednansky commented:

Escherichia coli cells, when 'pre-induced' in the presence of the artificial inducer TMG, synthesize β-galactosidase in the presence of glucose. COHN M, 1959 stated: "The effect of pre-induction is to restore in the presence of 10^-3 M glucose about 50 per cent of the maximal differential rate obtainable on succinate". The observation the maximal rate was not reached in the presence of glucose led the authors to argue, indeed incorrectly, that glucose was a preferential metabolic source for yielding high internal levels of repressor. Such observation however will have an explanation later on with the discovery of the 'cAMP effect' on β-galactosidase synthesis, in agreement with the finding by COHN M, 1959 that carbon sources presently known to elicit higher cAMP levels (particularly succinate, lactate and glycerol, see Epstein W, 1975) were found to be non-inhibitory (i.e. allowing maximal differential rate). Anke Becker’s final statement, that inhibition of lactose permease by unphosphorylated Enzyme IIA^Glc (leading to inducer exclusion) is primarily responsible for CCR of the lac operon, is therefore inappropriate as cAMP via its receptor protein (simultaneously designated as CRP Emmer M, 1970 and CAP Zubay G, 1970) also plays a role in CCR of the lac operon. Furthermore, Jacques Monod (1942) reported diauxie was attenuated - but not eliminated - when the cells were pre-induced (adapted to the less preferred 'B' sugar). Diauxie was however eliminated by addition of exogenous cAMP Ullmann A, 1968. Therefore, inducer exclusion and the level of cAMP both contribute to CCR of the lac operon.

Lastly, unphosphorylated EIIA^Glc does not inhibit adenylate cyclase. The current model of regulation postulates dephosphorylation of Enzyme IIA^Glc during glucose transport interferes with the activation of adenylate cyclase by phosphorylated Enzyme IIA^Glc.

On 2017 Mar 06, Robert J Maier commented:

Drs. McNichol and Sievert make some good points about the interpretation of our results. While we observed H2-augmented growth and CO2 uptake into cell-associated material, we did not show that CO2 contributes the main source of carbon. Therefore, the terms mixotrophy or chemolithoheterotrophy would seem to be accurate to describe our data, and not the term we used, chemolithoautotrophy. From our results, we cannot conclude Helicobacter is an autotroph.

On 2017 Feb 16, Jesse McNichol commented:

Kuhns et al (2016) provide evidence that the gastric pathogen Helicobacter pylori can use molecular hydrogen as an energy source. Increased growth yields and inorganic carbon incorporation both support the ability of H. pylori to gain metabolically useful energy from hydrogen. However, the use of the term chemolithoautotrophic to describe these findings is not correct.

The term chemolithoautotrophy is accurately defined as a metabolic mode that derives energy from chemical compounds (chemo-; as opposed to light or photo-), electrons from inorganic sources (-litho-) and carries out net fixation of inorganic carbon (-autotrophy; (2)). While the -litho- portion of this term has been used to describe heterotrophic organisms that oxidize inorganic compounds to supplement their metabolism (3), the -autotroph portion of this term can only be applied where carbon dioxide can serve as the predominant source of carbon for biosynthesis.

Since abundant organic carbon was present in the growth medium in this study, it is unclear if CO2 accounted for the main source of carbon for H. pylori. In addition, although the authors do observe CO2 uptake into biomass this does not prove that autotrophic carbon fixation occurred. Anaplerotic carbon fixation occurs as a series of carboxylation reactions that replenish intermediates in the citric acid cycle (4) or during fatty acid synthesis (5). As a normal process during heterotrophic growth, it explains the observed incorporation of CO2 in the absence of hydrogen. While such carboxylating enzymes do indeed incorporate inorganic carbon into biomass, the growth mode of an organism can only be considered autotrophic if they have complete pathways for using inorganic carbon as the main source for cellular biosynthesis (6).

This point is illustrated considering the importance of the higher activity and abundance of the acetyl-CoA carboxylase enzyme in the presence of hydrogen observed by Kuhns et al (2016). While this enzyme is indeed responsible for the carboxylation of acetyl-CoA to malonyl-CoA, the CO2 thus incorporated is lost during the condensation of malonyl-CoA subunits during lipid synthesis (5). Its higher activity may therefore simply be the result of higher levels of lipid synthesis associated with increased growth in the presence of hydrogen.

It should be simple to clarify whether autotrophic carbon fixation likely occurred during these experiments. The authors could estimate how much carbon was needed to support the observed increase in cell density, and compare this estimate with the amount of inorganic carbon incorporated into biomass. Unless the amount of inorganic carbon fixed represents a dominant fraction of H. pylori's cell carbon, chemolithoheterotrophic would be a more accurate term for the results observed by Kuhns et al (2016). Indeed, such chemolithoheterotrophic growth with hydrogen has been previously observed in other organisms (7).

A final point is worth mentioning. True autotrophs are well-known among the Epsilonproteobacteria (6,8), which employ the reverse tricarboxylic acid (rTCA) cycle for carbon fixation (9). Therefore, the absence of RuBisCO reported by Kuhns et al (2016) is not surprising given that autotrophic Epsilonproteobacteria do not use this enzyme for carbon fixation. The key enzyme that allows the rTCA cycle to run in a reductive direction is ATP-citrate lyase (6); however, the genes encoding this enzyme are absent in H. pylori strain 26695 (10). Since it lacks this enzyme and is thought to have a complete (albeit non-canonical) oxidative citric acid cycle (11), the current genomic evidence also argues against the possibility of autotrophic carbon fixation in H. pylori.

Jesse McNichol, Postdoctoral Scholar, Chinese University of Hong Kong; Simon F. S. Li Marine Science Laboratory, Shatin, Hong Kong; mcnichol at alum dot mit dot edu

Stefan Sievert, Biology Department, Woods Hole Oceanographic Institution; Woods Hole, MA, 02543, USA; ssievert at whoi dot edu

References:

1) Kuhns LG, Benoit SL, Bayyareddy K, Johnson D, Orlando R, Evans AL, Waldrop GL, Maier RJ. 2016. Carbon Fixation Driven by Molecular Hydrogen Results in Chemolithoautotrophically Enhanced Growth of Helicobacter pylori. Journal of Bacteriology 198:1423–1428.

2) Canfield DE, Erik Kristensen, Bo Thamdrup. 2005. Thermodynamics and Microbial Metabolism, p. 65–94. In Donald E. Canfield, EK and BT (ed.), Advances in Marine Biology. Academic Press.

3) Muyzer DG, Kuenen PJG, Robertson DLA. 2013. Colorless Sulfur Bacteria, p. 555–588. In Rosenberg, E, DeLong, EF, Lory, S, Stackebrandt, E, Thompson, F (eds.), The Prokaryotes. Springer Berlin Heidelberg.

4) Kornberg HL. 1965. Anaplerotic Sequences in Microbial Metabolism. Angew Chem Int Ed Engl 4:558–565.

5) Voet D, Voet JG. 2010. Biochemistry 4th edition. Wiley, Hoboken, NJ.

6) Hügler M, Sievert SM. 2011. Beyond the Calvin Cycle: Autotrophic Carbon Fixation in the Ocean. Annu Rev Marine Sci 3:261–289.

7) Kiessling M, Meyer O. 1982. Profitable oxidation of carbon monoxide or hydrogen during heterotrophic growth of Pseudomonas carboxydoflava. FEMS Microbiology Letters 13:333–338.

8) Campbell BJ, Engel AS, Porter ML, Takai K. 2006. The versatile ε-proteobacteria: key players in sulphidic habitats. Nature Reviews Microbiology 4:458–468.

9) Hügler M, Wirsen CO, Fuchs G, Taylor CD, Sievert SM. 2005. Evidence for Autotrophic CO2 Fixation via the Reductive Tricarboxylic Acid Cycle by Members of the ε Subdivision of Proteobacteria. J Bacteriol 187:3020–3027.

10) Tomb J-F, et al. 1997. The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388:539–547.

11) Kather B, Stingl K, van der Rest ME, Altendorf K, Molenaar D. 2000. Another Unusual Type of Citric Acid Cycle Enzyme in Helicobacter pylori: the Malate:Quinone Oxidoreductase. J Bacteriol 182:3204–3209.

On 2016 Mar 31, Damien Chaussabel commented:

I read your paper with great interest; this is excellent work with clear translational potential, so first of all congratulations on getting it published!

We are currently establishing a science education program that builds on the availability of large amounts of data in public repositories.

In this context we will encourage students/trainees to examine new noteworthy publications and to identify and share with the authors observations that may extend or build upon their original findings.

This is one of our first attempts! We hope that the exercise may also prove helpful to you:

A first observation is that in blood stimulated in vitro with a wide range of immune agonists, including pathogen-associated molecular patterns, heat-killed bacteria and cytokines, the patterns of induction of PKM2, IL6 and IL1B at the transcriptional level are rather distinct:

• IL1B shows high induction by Heat-killed E. coli (HK E. coli), Staph (HKSA) and Legionella (HKLP) as well as Flagellin, PAM3, R837, Zymosan, CpG; and some induction by LPS, RSV and TNF. http://www.interactivefigures.com:80/dm3/miniURL/view/L3 (just in case here is a 20 sec video showing how to view the data: https://youtu.be/MAtDFnLeeG0)

• IL6 shows high induction by Heat-killed E. coli (HK E. coli – which dominates the responses) and some induction by Heat-killed Staph (HKSA) and Legionella (HKLP), Flagellin, PAM3, Zymosan and CpG http://www.interactivefigures.com:80/dm3/miniURL/view/L4

• As for PKM2, first, induction levels are not as dramatic; second strikingly no response is seen following culture with HK E. coli. Compared to the “ex-vivo control” (=culture without stimulation); the most induction is seen with HKSA and HKLP, with possibly some induction seen with LPS, Flagellin, PAM3, Zymosan or TNF http://www.interactivefigures.com:80/dm3/miniURL/view/L2

The explanation might be trivial (at least to you!), but I found such disconnect puzzling, especially with regards to differences in levels of induction by HK E. coli.

Also, would you assume that the induction of PKM2 transcription correlates with dimerisation and nuclear translocation?

If that were the case would this phenomenon be driven by ROS released as a result of an innate response to bacteria from for instance neutrophils, independent of a change in the cellular glucose metabolism?

Thanks!

On 2016 May 15, Arnaud Chiolero MD PhD commented:

A brillant paper to understand the distance between the hype of clinical genomic and the reality.

On 2016 Apr 29, Suresh Panneerselvam commented:

Sir/Madam, This is an interesting article. Thank you for the article. I like the title very much.

The mention of TLR2 as intracellular instead of extracellular looks odd in the third paragraph "The intracellular TLRs consist of TLRs 2, 3, 7, 8, 9 and 10". Although, it is mentioned extracellular in the Figure.

In addition, it seems to me that TLR10 is also extracellular for eg; in this paper http://www.pnas.org/content/111/42/E4478.full.pdf (Figure 5E) there is a mention.

On 2016 Sep 22, Natalie Clairoux commented:

Free version of this article available after April 1, 2017 at http://hdl.handle.net/1866/13706

On 2016 Nov 13, James Tsung commented:

Link to Lung Ultrasound Videos: https://youtu.be/1Fuv-dEK-bc?list=PLHg2xyua_KjtS_bBnMWlmJH1ujUPnA92n

On 2016 Jul 03, Christopher Southan commented:

The sterol-linked endosomally targeted β-secretase inhibitor structure is neither disclosed here nor in the primary reference, rendering that part of the study irreproducible