On 2019-11-26 17:00:02, user Babak Naghizadeh Safa wrote:

Now published in J Roy Soc Interface https://royalsocietypublish...

On 2019-11-26 16:58:25, user Vahe Demirjian wrote:

The chasmosaurine specimens YPM 2016 and AMNH 5402 are interpreted as more similar to Vagaceratops irvinensis by Campbell et al. (2019), who remove these specimens from Chasmosaurus belli and place Vagaceratops back in Chasmosaurus.

Campbell, J. A., Ryan, M. J., Schroder-Adams, C. J., Holmes, R. B., & Evans, D. C. (2019). Temporal range extension and evolution of the chasmosaurine ceratopsid ‘Vagaceratops’ irvinensis (Dinosauria: Ornithischia) in the Upper Cretaceous (Campanian) Dinosaur Park Formation of Alberta. Vertebrate Anatomy Morphology Palaeontology, 7, 83-100. https://doi.org/10.18435/va...

On 2019-11-26 15:06:25, user Klaus H. Kaestner wrote:

Great study! What is the expression of Foxl1 and Foxf1 in the various telocyte- populations? I think you should cite Aoki et al as well, as this was the first description of Foxl1 positive telocytes as critical signaling centers. I hope you submit your study to CMGH! All the best, Klaus

On 2019-11-26 14:48:46, user Rob Moran wrote:

Interesting study! It's particularly interesting that a ColV plasmid is contributing to virulence and antibiotic resistance in ST101. It sounds like (lines 255-256) the resistance gene region in pEC121.B might be derived from RR1, which we've described previously (<br /> doi: 10.1089/mdr.2017.0177). If so, this represents further dissemination of this plasmid lineage.

On 2019-11-26 14:39:53, user Tomáš Hluska wrote:

Hi, just a short comment after quick look.

Please, use the standard abbreviation for trans-zeatin, i.e. tZ. <br /> In description of Figures I'd switch it to "Quantification of Proline, total flavonoid content (TFC), phenylpropanoids (PPs)", etc.<br /> You need to improve the English significantly. Right now, some of the sentences really do not make much sense.<br /> Good luck with your work.

On 2019-11-26 10:34:45, user Aliaksei Chareshneu wrote:

Is it the entire receptor molecule or a subunit/domain that you have studied?

On 2019-11-26 00:18:45, user José H. Oliveira wrote:

Figure 4 is missing.

On 2019-11-25 19:00:24, user Rich Abdill wrote:

This paper makes references to supplementary tables and figures that I can't find in the PDF—are they available?

On 2019-11-25 13:15:05, user Erik wrote:

Which (bulk) RNA-seq platforms is this algorithm suitable for?<br /> Should I specify it while running the package?

Thanks,<br /> Erik

On 2019-11-25 09:30:11, user Daniel Žucha wrote:

Dear readers, <br /> I would like you to be notified that this preprint has been already published in the Clinical Chemistry journal with minor changes. It is accessible under DOI: 10.1373/clinchem.2019.307835. The direct link to this site is forthcoming shortly.

On the behalf of authors,<br /> Daniel Zucha

On 2019-11-25 08:05:01, user Alessandro Cellerino wrote:

I am not conviced by your age transform. It is well known that aging, defined as reduction in the performance of organs and systems, starts in humans in the fourth decade (see for example https://www.ncbi.nlm.nih.go... and Labrador retrievers at age two are not even mature, as every Lab owner knows.<br /> Also Labrador retrievers are carriers of POMC mutations and prone to obesity and metabolic syndrome, so they may actually suffer from accelerated aging.

On 2019-11-25 00:24:22, user MarkGB wrote:

Dear authors: the formula for age comparison is bunk. Your research for comparison of DNA age markers is great - but using it to create an aging comparison formula is a parlor trick. No dog at two is the mental, or physical, equivalent of a 40 year old person. And the reduced slope of aging at advanced years, indicated by the formula, does not replicate a reasonable expectation of life experience or expectancy.

I'm going to stick with the seat-of-the-pants formula of 10 years for each of the first two, and 6 thereafter. It gives me more useful results, and communicates better to inexperienced dog owners.

Creating a formula based on a single data point does not a full analysis make.

On 2019-11-22 20:07:23, user Luis Morais wrote:

I am also a proud dog owner, and you are all missing the point of the research.... <br /> The formula was designed to aproximate the dogs genetic aging with ours. Have you have seen an 100 years dog? Thought so. Therefore the formula won't apply.

And of course your dog isn't a 40 or 50 something year old human. It may be in the "spring of his youth" and yet it's aging at a way faster rate than you!

Their claim is not that our dogs have old person/dog personality, but that their body is degrading that much faster than ours is.

On 2019-11-21 18:35:39, user Guido Governatori wrote:

Absolute non-sense. There are problems with the formula: 100 human-years corresponds to 75 dog-years. 90 human-years to 39 dog-years, 80 human-years to 22 dog-years, 75 human-years to 16 dog-years. <br /> 75 human-years is less than 10% increase of the human life expectancy rate they consider in the paper (70 years), while 16 is over 30% increase of Labrador life expectancy (12 years).<br /> Their figure 2D is already off chart of the stage of life classification.

On 2019-11-19 19:57:31, user Oscar Salguero wrote:

I'm no researcher or anything, but I noticed if you add 16 instead of 31, the results feel much more satisfying I would say. My dog (5 yrs) according to this formula is about 56. Which seems really old. It also doesn't make sense when a dog is 1 year old they are 31. When I used the formula but added 16 instead of 31, this means when a dog lives for one year, they are 16. This makes more sense, to me at least. And now that he is five, it would make him close to 42, which seems more accurate since he is about halfway through how long he is expected to live for. Again not a scientist, but instead a curious dog owner.

On 2019-11-19 02:40:29, user Jingo Balls wrote:

Um, no. Having been graced with the companionship of many dogs on my life, I can say conclusively that this formula is far from complete or accurate. My 2.5 year old Lab/Viszla mix is 45 years old in human years?? I am 45 years old. My dog is definitely a young 20-something. She is in top physical shape and absolutely bonkers at times. I, on the other hand, am a serenely calm, completely rational and self-regulating adult hooman who enjoys measured exercise (but not too much) and all other enjoyable activities in moderation.

My dog is not a 45 year old dog-human.

On 2019-11-18 14:38:33, user Dmitriy Kozubov wrote:

Absolutly nonsens. Maybe there is some correlation between methylomes and aging. But the sugestion that "These results establish methylation not only as a diagnostic age readout but as a cross-species translator of physiological aging milestones" contradicts obvious fisiological facts. 9 month dog is not sexually mature, that corresponds according to this article 26 year old human. 3 years old dog corresponds 48.5 y.o. human. The resuls probably show that aging of all species follows common logariphmic relation. But not more. And has nothing in common with "cross-species translator of physiological aging".

On 2019-11-25 02:51:01, user Edward Y. Sheffield wrote:

Brief explanation:<br /> According to the bioRxiv team’s email, this paper’s “false information” means “false name”. It is the pen name (“Edward Y. Sheffield”) I used before.At the beginning, I was really not confident on this paper due to its “strange” results. Since this work was further verified recently, I am glad to reveal my true name. Thanks for your understanding!<br /> Yaohua Xie (Edward Y. Sheffield)

https://www.linkedin.com/in...

On 2019-11-24 11:43:42, user Edward Y. Sheffield wrote:

By the way, the previous version seems to be accessible also.<br /> Please click the tab "Info/History" at the top to view it.

On 2019-11-05 02:29:59, user Edward Y. Sheffield wrote:

Hello! We have written a "Questions&Answers.pdf" according to our recent discussion with peers. It can be found on the following page: https://zenodo.org/record/3...<br /> If anything inappropriate is found in the material (or this study), please do not hesitate to tell us.

On 2019-11-23 19:54:27, user Ting wrote:

I find this interesting. I have many questions.

Are there available antibodies or small molecules to show that breaking up the filaments would revert the phenotype to that of the smaller particles?

I can see slower bursts in Supporting video 2 vs Supporting video 1 which means slower fusion of pellet vs sup but it is not clear in the manuscript the average fusion time for the sup vs pellet extracted from these experiments. Maybe I missed it in the text or figure.

How do short particles vs long filaments interact with cells before cell death? Fusion is difficult to image in cells, but are filaments fusing with endosome slower (as expected from the in vitro fusion assay)? Fusion on cells is more complicated as it involved protein receptors, so it would be good to compare the early steps of viral infection in cells between short particles and long filaments.

When a human is infected with either influenza filaments or small particles, would the immune system clear away the filaments faster than the small particles because they are bigger and much easier to detect and fuse more slowly with cells?

How are long filaments made? Are they made on plasma membrane? Is it an error in budding off the membrane? <br /> Are certain membrane domains (like lipid rafts?) allowing filament formation? Or do small particles fuse after budding off the membrane?

What is the HA/NA ratio/distribution in long filaments? Do they fuse slower because of less HA or the shape is less conducive for membrane fusion? Are there specific domains on the virus that are dense with HA?

Editing suggestions: <br /> Could the words sup and pellet be put in lower case rather than capitalized? It is a bit jarring to see these capitalized words pop up as one is reading.<br /> Could you put all the Fab concentration data on one graph (Fig 4) and use different color? Also can the lines in the graphs be removed?

On 2019-11-23 08:53:01, user Ricman wrote:

No supplementary information?

On 2019-11-22 20:57:41, user Davidski wrote:

Hello authors,

What are the Y-chromosome haplogroups of the 25 males? I can't see them listed anywhere in the manuscript.

On 2019-11-23 05:08:40, user Awais wrote:

You have written Malus domestica throughout your manuscript which is wrong. You need to correct it to Malus x domestica.

On 2019-11-22 12:13:53, user Olivier Gandrillon wrote:

Dear authors

You write in your discussion : « In particular, it will help us to better understand whether increased plasticity, as manifested in increased cell-to-cell variability, is a general feature that precedes cell commitment or whether this is restricted to specific systems such as hepatoblast differentiation. »

What can already be established is that is is NOT restricted to hepatoblast differentiation.

Indeed this has been demonstrated in <br /> 1. Murine ES cells (Stumpf et al., 2017; Semrau et al., 2017; Moris et al., 2018)<br /> 2. EML cells (Mojtahedi et al., 2016)<br /> 3. Chicken erythroid progenitors ((Richard et al., 2016; Guillemin et al., 2019)<br /> 4. Murine hematopoiesis (Wiesner et al., 2018)

This might be argued for evidence for a rather general feature IMHO.

Oliiver Gandrillon

Guillemin, A., Duchesne, R., Crauste, F., Gonin-Giraud, S., and Gandrillon, O. (2019). Drugs modulating stochastic gene expression affect the erythroid differentiation process. PLOS ONE 14, e0225166.

Mojtahedi, M., Skupin, A., Zhou, J., Castano, I.G., Leong-Quong, R.Y., Chang, H., Trachana, K., Giuliani, A., and Huang, S. (2016). Cell Fate Decision as High-Dimensional Critical State Transition. PLoS Biol 14, e2000640.

Moris, N., Edri, S., Seyres, D., Kulkarni, R., Domingues, A.F., Balayo, T., Frontini, M., and Pina, C. (2018). Histone Acetyltransferase KAT2A Stabilizes Pluripotency with Control of Transcriptional Heterogeneity. Stem Cells.

Richard, A., Boullu, L., Herbach, U., Bonnafoux, A., Morin, V., Vallin, E., Guillemin, A., Papili Gao, N., Gunawan, R., Cosette, J., et al. (2016). Single-Cell-Based Analysis Highlights a Surge in Cell-to-Cell Molecular Variability Preceding Irreversible Commitment in a Differentiation Process. PLoS Biol 14, e1002585.

Semrau, S., Goldmann, J.E., Soumillon, M., Mikkelsen, T.S., Jaenisch, R., and van Oudenaarden, A. (2017). Dynamics of lineage commitment revealed by single-cell transcriptomics of differentiating embryonic stem cells. Nat Commun 8, 1096.

Stumpf, P.S., Smith, R.C.G., Lenz, M., Schuppert, A., Müller, F.-J., Babtie, A., Chan, T.E., Stumpf, M.P.H., Please, C.P., Howison, S.D., et al. (2017). Stem Cell Differentiation as a Non-Markov Stochastic Process. Cell Systems 5, 268–282.

Wiesner, K., Teles, J., Hartnor, M., and Peterson, C. (2018). Haematopoietic stem cells: entropic landscapes of differentiation. Interface Focus 8, 20180040.

On 2019-11-22 10:56:20, user Henry wrote:

There's a video of the fish climbing the wall discussed in the paper at:

https://gallery.jkwillis.in...<br /> They love it!

On 2019-11-22 05:34:19, user Davidski wrote:

Hello authors,

On page 10 of your manuscript you refer to Eastern European populations such as hunter-gatherers from Ukraine and the Yamnaya people as "eastern Eurasian".

Please keep in mind that in English speaking countries the term eastern Eurasia is associated with East Asia rather than Eastern Europe. In other word, places like Japan and Korea.

Also, I'm not sure about your claim that the admixture between early farmers and steppe pastoralists was gradual, considering that early Corded Ware samples from Poland and the Baltics are almost indistinguishable from Yamnaya samples in terms of genome-wide ancestry. This suggests to me that the proto-Corded Ware migrations involved the rapid expansion of steppe ancestry into Central Europe with very little admixture along the way.

On 2019-11-21 21:53:28, user Tommy Vo wrote:

Very nice and interesting work to actually see the gradient of chromatin based on compactness. The observation of a "return to compaction" near the TES is reminiscent of recent findings that transcription termination machinery can promote heterochromatin formation.

Vo, T., et al. (2019). CPF Recruitment to Non-canonical Transcription Termination Sites Triggers Heterochromatin Assembly and Gene Silencing.

Chalamcharla VR, et al. (2015). Conserved factor Dhp1/Rat1/Xrn2 triggers premature transcription termination and nucleates heterochromatin to promote gene silencing.

On 2019-11-21 21:40:03, user Keith Hughitt wrote:

Great article! Few quick thoughts/notes:

1. In addition to distinguishing between competitive and self-contained methods, there has also been a lot of focus recently on sample-permuting and gene-permuting / single-sample / "pre-ranked" methods.
2. Similarly, choice of ranking metric also has a significant impact on performance and may be worth considering in future benchmarks.
3. It may be useful to use consistent colors to refer to each of the methods across the figures.
4. I know "10" is a nice round number and there are also already about 200 methods out there, so it is not feasible to compare all of them, but fgsea is one other method that I've had a good experience with over the past couple years that has been gaining a lot of traction and may be interesting to include in future benchmarks.

On 2019-11-21 21:11:47, user Mark M wrote:

One needs to cautious interpreting this study, as it utilizes gremlien whole-body AKT2 ko mice. Given the role of insulin signaling via AKT in liver and fat, the cell-autonomous effects of AKT signaling in muscle cannot be ascertained using an whole-body insulin-resistant AKT2 KO model.

On 2019-11-21 19:48:23, user Charles Warden wrote:

As a minor comment, please just call me "Charles Warden" instead of "Dr. Charles Warden," since I don't have a PhD :)

https://sites.google.com/si...

Very exciting work - thank you very much for acknowledging me!

On 2019-11-21 18:42:37, user jamesian wrote:

Can you help me communicate this more widely by .. indulging in a metaphor? If iAGE is a clock what are the hypothetical numerals 1-12 that show "iAGE?"

On 2019-11-17 15:24:56, user Sebastian Aguiar Brunemeier wrote:

Very interesting. One wonders why CXCL9 would be elevated in the first place.

On 2019-11-21 14:42:49, user ganesha rai wrote:

Authors used a questionable LDHA inhibitor FX11 for their study. I do not see any rationale using FX11 when there are other more advanced inhibitors available. FX11 is a pan inhibitor that hits many other targets and the data generated using this small molecule is always questionable

On 2019-11-21 14:40:18, user Manuel Martinez Garcia wrote:

Hi all,<br /> nice results but bad that this papers does not say any word on the first published paper showing nanopore sequencing in viral metagenomics, really! https://peerj.com/articles/...<br /> and also in introduction talks about different methods to capture the viral diversity: fosmids, metagenomics AND forget one of the last one, single virus genomics, which has been very informative in marine systems. Definitely, it has different aspects to be improveb

On 2019-11-21 09:57:21, user Pavel Tomancak wrote:

This version has a mistake in Figure 4A which is showing LifeAct data instead of Myosin (they look very similar). We apologise for the error.

On 2019-11-21 09:22:02, user franck raynaud wrote:

The modeling part of this paper is now published on iScience:<br /> https://www.cell.com/iscien...

On 2019-11-21 07:33:52, user Saravanan vijayakumar wrote:

Though the authors created "Consensus_NR" from the previous data set for this study, the authors should clearly state what is "Positive" and what is "Negative" data set. This is because, for instance (on random check) IEDB ID 107354, which is classified as "POSITIVE" in the "Consensus_NR" data set is reported to have 10 B-cell assays in IEDB, out of which 8 are negative 1 is positive and 1 is low-positive. Does this data should be in Negative? Similarly, IEDB ID 93172 which is also classified as "POSITIVE" in this study, but the IEDB reports 4 B-cell assay out of which 2 are negative and 2 are positive-low! Hence, authors should clearly mention what is "POSITIVE" and what is "NEGATIVE" according to this study. Also, it will be great if authors provide the sequences of both positive and negative data set they used in this study as supplementary, rather than only IEDB ids, because IEDB doesn't allow bulk epitope download via IDs.

On 2019-11-16 11:36:22, user Katerina Nastou wrote:

We have changed the description of LBEEP to mention that 'The data set <br /> used to train this method was constructed by using only exact epitopes, <br /> instead of epitope containing regions, which have been used as training <br /> material in the past, making LBEEP a pioneer method in that respect. <br /> The exact epitopes used for training were isolated from IEDB and are 5 <br /> to 15 amino acid residues long and thus LBEEP is better suited for <br /> predictions in that range. '<br /> Unfortunately we tested the method using all three models, and the<br /> confirmed model performs slightly worse than the balanced model in<br /> our data set. This will be clearly now stated in the article and<br /> results from this test will be provided. Moreover, it should be<br /> mentioned that the fact that LBEEP is constructed to predict exact<br /> epitopes is clearly stated, and is the reason why we chose a<br /> window size of length 14 to test the method, where we showed that<br /> LBEEP performs much better. Unfortunately, we could not use this<br /> window size for the consensus predictor, since all other methods<br /> do not perform as well (a fact also mentioned in the text). We<br /> plan to change the manuscript text according to your comments<br /> before submission. Thank you for taking the time to comment on our<br /> article.

On 2019-11-11 12:59:33, user Saravanan vijayakumar wrote:

LBEEP uses "exact epitopes" for training the model and not the "epitope containing regions". LBEEP is build on "Exact epitopes" (as defined by IEDB) and is constructed for predicting exact epitopes of length between 5 and 15 (clearly documented in LBEEP article). No other methods adopted "Exact Epitope" ! The reason for poor performance of LBEEP on this tested data is because a) Tested peptides may be Epitope containing regions rather than exact epitopes and b) as mentioned in this study that LBEEP is not trained on epitopes greater than 15 length (however, confirmed mode of LBEEP is trained with peptide >15 and this study should have been tested LBEEP using confirm mode). The authors of this article should mention that the LBEEP is constructed using "Exact Epitope" and designed to predict the same, which I feel is not mentioned in this article. This may mislead the readers about the LBEEP tool.

On 2019-11-21 00:43:46, user Charles Warden wrote:

While I haven't directly worked with Nanopore data myself, there was a clear effect on species assignments in this paper (when using full barcodes versus mini-barcodes), as you can see in the comment:

https://www.biorxiv.org/con...

For Illumina, I am not sure how you are defining "full" versus "mini" barcodes, but I have definitely encountered enough issues with "index hopping" that I am at least a little concerned about discussions with increased multiplexing (even though cross-contamination can occur at various steps).

In terms of what I can currently cite, I have some notes on Biostars about possible QC metrics for de-multplexing:

https://www.biostars.org/p/...

While the PhiX is kind of a special case (which I believe you typically only see with single-barcode samples, even though I would guess there are probably other events that are more difficult to define). However, since PhiX spike-ins don't have a barcode, that is one example of a de-multiplexing issue includeed in that discussion:

https://www.biostars.org/p/...

Otherwise, I noticed drops in I2 (i5) index quality scores (with 100x100 reads), which might be a red flag for some runs (and a possible disadvantage to having a second barcode to assign samples).

While I think the most important point for Illumina de-multiplexing is that these problems are often batch effects (hard to see unless you consider a large number of batches), there was a more limited example with a small fragment size (to show that you can get at least some wrong sample assignments, even with Illumina dual-barcodes):

https://github.com/cwarden4...

On 2019-11-20 19:34:44, user Yuwei wrote:

Very interesting paper. I have a somewhat naive question: How do selectivity and diffusivity influence each other? It seems a little bit contrasting to see that the gel is more permeable for NTF2 while its diffusion is actually slower than mCherry in the gel. Does that mean the selective permeation is mainly dictated by how easy/fast the molecule can enter the gel?

On 2019-11-20 17:58:22, user John Greally wrote:

Now published https://www.nature.com/arti...

On 2019-11-20 17:31:12, user Connor Rosen wrote:

I thank the authors for this investigation and I am excited by the potential of this analysis. However, without the full gene expression information it is difficult to fully understand the role of Langerhans cells and the impact of their absence. The authors should provide the full data (e.g. tables of differentially expressed genes in each cell type), and the accession number for the underlying RNA-seq data for re-analysis.

On 2019-11-20 17:02:26, user Julia wrote:

Hi, nice work! Would love to try the protocol. What is the PH for the lysis buffer (100 mM Tris, 1% sodium deoxycholate, 10 mM TCEP, 15 mM 2-chloroacetamide)? Is pre-isolation of the membrane fraction not necessary?

On 2019-11-19 20:05:32, user Leeya Engel wrote:

Featured in our webinar on maskless photopatterning of EM grids! https://www.youtube.com/wat...

On 2019-11-19 15:05:37, user stefano cagnin wrote:

Several manuscripts, not considered in this, were published demonstrating the importance of single myofiber analyses (e.g. Chemello et al 2019 Cell Rep; Alessio et al 2019 NAR; Chemello et al 2015 Genom Data; Mammucari et al 2015 Cell Rep; Chemello et al 2011 PlosOne; Schiaffino et al. Histol. Histopathol. 2019; Murgia et al Cell Rep 2017; Murgia et al EMOBO Rep 2015).

On 2019-11-19 11:28:15, user Thomas Blankers wrote:

Dear authors,

we enjoyed reading this preprint in our journal club. This is very interesting research and the finding that the genetic architecture of intra-individual variability in behaviors may be largely uncoupled from that of inter-individual variability in individual means, is fascinating. As even the phenotypic side of intra-individual variability is poorly studied, we felt like this section would benefit from some more elaboration of expectations and observations. For example, even though you state “cross-test correlations demonstrate that individuals show consistency in their level of behavioral predictability both within and across test-type”, the magnitude of the two correlations across testing paradigms are very low and only borderline significant (in a sea of highly significant within-testing paradigm correlations). My naïve expectations would be more and stronger coupling across tests, but I guess that depends on many factors.

It is also unclear how the phenotypic correlations and their sources of error would shape the coupling of genetic architectures of intra-individual trait variability in different test paradigms (QTL for IVV traits same to only occasionally overlap between testing paradigms). We felt like a somewhat broader introduction to the different forms of intra-individual variability and which ones specifically you are addressing here, the null expectations for how consistent these behaviors ought to be across testing paradigms, and which scenarios would result in more or less overlap among QTL (both overlap between means and IVV and between IVV testing paradigms) would help guide the reader a lot.

To illustrate what I mean: Looking at figure 2, the two chromosomes where QTL for inter and intra-individual variability overlapped (10 and 24), were also chromosomes where IVV measured under different paradigms overlapped, whereas most of the other QTL did not overlap. So, what does a unique genetic architecture for inter and intra-individual variation mean when there is also limited overlap between different experiments for the same behavior? And the overlapping QTL for IVV and QTL for means seem to have small confidence regions, so maybe these are also the QTL of high effect? This prompts the question: How much of the variance is explained by shared versus unique QTL? And, what would be a reasonable expectation under different true distributions of causal loci? All these questions are likely to be answered when the context of intra-individual variability, the expectations as to how much of the genetic bases explored here should be shared, and what the observed amount of sharing/independence means biologically are more clearly outlined in the introduction and discussion. Best of luck and please do not hesitate to contact me if you have any questions.

On 2019-11-19 07:18:08, user Andrew Teschendorff wrote:

In response to the recent comment by Halperin, we have implemented TCA using the tcareg function. We find that our results are ***unchanged*** whether we use tcareg, or whether we infer the tensor first and then perform regression for each cell-type separately. Indeed, it should make absolutely no difference in the simulation scenarios we have been considering. Halperin is unable to explain why it should make a difference on the simulated datasets. In our hands, TCA suffers from a very low precision or high FDR, a key performance metric which, we repeat , was completely ignored in the TCA paper. In our bioRxiv preprint we provide a long list of issues (FDR was ignored in TCA paper, no bidirectional changes, highly ambiguous specification of simulation models, wrong use of metric to compare methods that aim to identify different things (DMCs vs DMCTs),...etc etc), which Halperin et al have not yet addressed. We note that these issues have absolutely nothing to do with how TCA is implemented, so the claim that our criticism is irrelevant is completely unfounded.

On 2019-11-17 23:04:04, user Eran Halperin wrote:

We have to strongly disagree with the comment by Teschendorff. As in several cases in Jing et al., Teschendorff makes another false claim about the TCA paper in his comment below: We do provide in the TCA package an option to learn the tensor, which is of interest (and works well, as demonstrated in the TCA paper), however, TCA should be applied differently for the task of association testing (i.e., EWAS). Specifically, we used Equation (13) in the Methods of the TCA paper for association testing; we clarified this in the paragraph that follows Equation (13) in our paper: "In this paper, whenever association testing was conducted, we used this direct modeling of the phenotype given the observed methylation levels."

Importantly, in his commentary, Teschendorff does not acknowledge the fact that there are two innovative components in the TCA paper: (1) inferring a three-dimensional tensor of cell-type-specific levels from two-dimensional bulk data, and (2) direct modeling of phenotypes as having cell-type-specific effects, given the observed methylation levels, which allows to integrate over the hidden tensor information; as pointed out in the TCA paper (and instructed in the vignette and manual of the TCA package), this is the preferred way to perform EWAS using TCA. While the estimates of the tensor may also be used for EWAS (as performed by Jing et al.), this option is substantially less powerful, as it does not take into account the differences in variance between samples. For more details see Equation (13) in the TCA paper.

Also, we concur that there is value in the CellDMC paper as a benchmarking paper for previous methods. However, our argument is that CellDMC is not a new approach (although in their own words, in the CellDMC paper Tschendorff and his colleagues present it as a “novel statistical algorithm”), as the same method has been previously applied to gene expression (Westra et al., Plos Genetics 2015, Shen-Orr et al., Nature Methods 2011), while to the best of our knowledge, TCA is a new approach, with its advantages and disadvantages.

Finally, we would like to emphasize that we disagree with most of the claims made by Jing et al. in their paper, however, these claims are irrelevant as long as they present irrelevant results based on an irrelevant application of TCA. If any of the reviewers or editors of Jing et al. would be interested in a more detailed criticism of their claims, we will be happy to provide it, although we do not think that it is needed at this point.

On 2019-11-14 15:46:37, user Andrew Teschendorff wrote:

We have to strongly disagree with the comment by Halperin. First of all, the paper itself does not state that the method TCA needs to be run with tcareg, and indeed the authors of the TCA package provide all the required functions to learn the tensor-object, which is after all the task that gives name to their algorithm: TCA=tensor composition analysis. If Halperin now claims that the tensor they infer with their tensor function is of low quality, then this further undermines their method. And indeed, if the tensor function in the TCA package is wrong why do the authors make it available? In fact, the inference of the tensor is the key innovative aspect of their paper, and so if the authors now claim that it ain't working, then the package and functions should be removed.<br /> Beyond this, Halperin offers no counterarguments for the multiple other issues (e.g. no assessment of the PPV or precision was done in the original paper, no comparison of computational efficiency was provided, no clarification of which simulation scenarios were considered). <br /> In summary, we acknowledge that TCA is innovative, but that does not automatically qualify the method as being reliable or useful in an EWAS context. This requires a much more comprehensive assessment in terms of multiple performance metrics, and importantly in multiple real EWAS datasets, not just one example as done in the TCA paper. <br /> While the use of statistical interaction terms have been considered in previous studies on gene expression data and in the 2 cell-type scenario, our CellDMC paper does demonstrate that the concept works for multiple cell-types, in multiple real EWAS datasets and over a wider range of simulation scenarios than those considered by previous studies, including the TCA paper.

On 2019-11-13 23:25:31, user Eran Halperin wrote:

Jing et al. did not use the right function in the R package of TCA (Rahmani et al., Nat Comm 2019) and therefore their evaluation of TCA is irrelevant and their conclusions throughout the paper are invalid. Specifically, Jing et al. did not use the 'tcareg' function for performing EWAS; the 'tcareg' function is demonstrated and explained in detail in the manual and vignette of the R package (and the full technical details behind 'tcareg' are in the original TCA paper; all the analyses in the TCA paper were performed using that same model).

Regardless of the irrelevance of this commentary by Jing et al., importantly, as demonstrated in the TCA paper, TCA is able to find new replicating cell-type-specific associations that were not found prior to the availability of the TCA method; particularly, they were not found using CellDMC (the method described by the authors of this commentary in another paper). Importantly, unlike TCA, which is an original and innovative method, we would like to emphasize that CellDMC is, in fact, the same method that was previously proposed by others (e.g., Westra et al., Plos Genetics 2015, Shen-Orr et al., Nature Methods 2011); these works were unfortunately not credited by the authors of CellDMC.

On 2019-11-19 03:17:54, user Fong tian Wong wrote:

The plasmids are now available on addgene! <br /> https://www.addgene.org/Fon...

On 2019-11-18 22:26:25, user Michael Hoffman wrote:

This manuscript is incomplete as the authors have withheld a description of methods that would be required to interpret the data, critically examine the procedures used, or reproduce the study as in usual scientific practice. The manuscript says the authors will provide it at a future date. Other readers may wish to postpone consideration of this until such time as the authors decide to follow the most basic scientific standards, such as describing methods in detail in posted manuscripts.

On 2019-11-18 17:34:23, user Maria G. Castro wrote:

Congratulation to Flor Mendez, graduate student in my lab, and all the lab colleagues who contributed to make this research on developing an immunotherapy for DIPG a reality! Phase 1 clinical trials in the planning stages.

On 2019-11-18 14:12:24, user Mark Ashby wrote:

To read a response to this pre-print, see here: https://ecoevorxiv.org/68h3w<br /> To read our original peer-reviewed critique of the EMBER project, see here:<br /> https://besjournals.onlinel...

On 2019-11-18 08:30:14, user Wolfgang Graier wrote:

Congratulation to the team, this a very nice and excellent work. May I ask to kindly consider our recent findings when discussing your excellent findings:

1. Klec, C; Madreiter-Sokolowski, CT; Stryeck, S; Sachdev, V; Duta-Mare, M; Gottschalk, B; Depaoli, MR; Rost, R; Hay, J; Waldeck-Weiermair, M; Kratky, D; Madl, T; Malli, R; Graier, WF. Glycogen Synthase Kinase 3 Beta Controls Presenilin-1-Mediated Endoplasmic Reticulum Ca²⁺ Leak Directed to Mitochondria in Pancreatic Islets and β-Cells.<br /> Cell Physiol Biochem. 2019; 52(1): 57-75.

2. Klec, C; Madreiter-Sokolowski, CT; Ziomek, G; Stryeck, S; Sachdev, V; Duta-Mare, M; Gottschalk, B; Depaoli, MR; Rost, R; Hay, J; Waldeck-Weiermair, M; Kratky, D; Madl, T; Malli, R; Graier, WF. Presenilin-1 Established ER-Ca2+ Leak: a Follow Up on Its Importance for the Initial Insulin Secretion in Pancreatic Islets and β-Cells upon Elevated Glucose.<br /> Cell Physiol Biochem. 2019; 53(3): 573-586.

I think your data truly are very important and, in my humbling opinion, together with our findings mentioned above add to the current correction on the principles of the mechanisms of insulin secretion.

Thank you very much and good luck in publishing.<br /> Best,<br /> Wolfgang Graier

On 2019-11-18 08:14:45, user Tanai Cardona Londoño wrote:

"Our mutation rate estimates for baboons raise a further puzzle in suggesting a divergence time between apes and Old World Monkeys of 67 My, too old to be consistent with the fossil record; reconciling them now requires not only a slowdown of the mutation rate per generation in humans but also in baboons."

If the divergence time of OWM and apes based on the fossil record is actually 35 My, to reconcile rates and the fossil record wouldn't you need an acceleration of the mutation rates rather than a slowdown?

Faster rates leads to younger divergence time estimations because at a faster rate you need less time to accumulate the same amount of mutations... no? :)

On 2019-11-18 04:30:40, user Trudy Oliver wrote:

Interesting connection re: B-catenin. Wnt/Bcatenin pathway is well known for inducing MYC. We find that only the MYC-high subset of small cell lung cancer (SCLC) has low ASS1 and sensitivity to ADI-PEG20. Please see Chalishazar et al, Clin Can Res, 2019 for a much more thorough assessment in SCLC.

On 2019-11-17 06:21:05, user Wacław Jan Kroczek wrote:

Greetings,<br /> With all due respect, the allegations in the article, namely, that supercentenarians are concentrated into regions with no birth certificates and short lifespans is false. By the study of the Gerontology Research Group, International Database on Longevity and Max Planck Institute for Demographic Research among others, we know that supercentenarians are extremely rare, yet an existent and validated population. The Gerontology Research Group takes the global view on the subject, using stable and defined validation criteria outlined by Dr. Poulain of Belgium in the Supercentenarians monograph (2010). I invite all of you who would like to learn more about the GRG data and its characteristics.<br /> www.grg.org<br /> Sincerely,<br /> Waclaw Jan Kroczek<br /> GRG Administrator for Case Validation Reports

On 2019-11-16 00:56:33, user James Mallet wrote:

I loved this paper -- well, I would, wouldn't I?!! But I think there's some editing issues in the figures and legends that I quickly noticed. It helps to put the legends on the same page as the figures if at all possible! In Fig. 3, A, C, E are labeled H. melpomene, but in the legend it says they are erato, and also B,D,F vice-versa. And in spite of a legend, I couldn't see any Fig. 5 at all! But I am so happy you're finally publishing this cool work, so these critiques are meant as a help, not a criticism.

On 2019-11-15 21:41:18, user Williams Lab wrote:

In this study we show that the empirical resolution of mapping based on cis eQTL analysis will be roughly 0.4–0.8 Mb for loci associated with LODs of >5 (see Figure 4, and extrapolating from current eQTL data sets with n = 60 to 80 BXD types and replication of only 2 per strain).

Resolution could be quite easily pushed down to ~100 kb by using a highly selective (small) diallel cross of BXDs with other inbred lines—the fully sequenced Collaborative Cross strains, the Phenome Project strains (also sequenced), and 50 fully inbred Nachman strains (PMID: 30248095, sequencing in progress for some).

In my opinion, resolution is NOT the current problem in forward genetics. GWAS has us beat in terms of precision and always will. The core problem is the sparsity of systematic phenome data that enables highly scalable genome-to-phenome prediction. This is an area where animal geneticist need to get their act together—to move from maps to mechanism to prediction that actually are translationally relevant to human populations.

On 2019-11-15 21:12:39, user Tyler Square wrote:

Cyclostomes (lampreys and hagfishes) are not established as being 2R, and they are unaddressed here. It seems like you are actually addressing gnathostomes and the gnathostome common ancestor, not "all vertebrates" and their ancestor (per the first sentence of your abstract).

On 2019-11-15 10:10:14, user Alexis Verger wrote:

This work is a nice addition to the previous work by the Hahn and Klevit's labs deciphering the molecular mechanisms of Mediator complex subunits recruitment by transactivation domains (TADs). Here they show that despite no clear sequence homology, the Gal4 and Gcn4 TADs bind the same surface of Med15 ABDs domains. In addition the same region of Gal4 TAD was known to interact with Gal80 via a tight structured complex, suggesting that the structured binding partner of an intrinsically disordered protein (IDP) dictates the type of interaction.

Please find below some general comments that I hope will be of some interest.

• I am not completely convinced that the interfaces of the Med15 ABDs with Gcn4 and Gal4 is sequence-independent. Analysis of the backbone chemical shifts for the Gal4/ABD1 complex indicate that residues 861-869 adopt helical structure upon binding (Figure 3D). This is reminiscent of the situation observed between Gcn4 cTAD/ABD1 that adopts helical character upon binding Med15 (Brzovic et al. 2011). Intriguingly, as stated by the authors, region 2 of Gal4 TAD contains a sequence (YNYLF) included in the helical structure and that resembles the short motif in GCN4 cTAD (WXXLF) critical to bind ABD1 hydrophobic cleft. Interestingly mutation of Gal4 TAD YLF to AAA has a strong effect in transcription activation (Figure 2) and Y865W (mimicking WXXLF of GCN4) possesses higher transcription activity compared to wt. It could be interesting to test in FP and/or ITC the effects of YLF ->AAA and Y865W mutations and see if there is a correlation between transcriptional activation potential and Med15 binding.

• Given that the affinity and binding mode of GCN4 and Gal4 TADs for MED15 are similar (Table 1), did the authors try competition experiments to investigate whether GCN4 TAD can compete with Gal4 TAD for binding to Med15 (and maybe implying the WXXLF motif )?

• Concerning Table 1 - it could be nice to add 1 or 2 ITC curves for better illustration. My understanding is that protein concentrations used for ITC are quite high (mM) 500X above the Kd.

• I understand that Gal4 TAD 828-881 is soluble but not 840-881. Did you try a shorter version 855-870 (corresponding roughly to the helical structure) in NMR ?

• The Gal4 region that binds to Gal80 overlaps with the one that binds to Med15. Is the Gal80 interface very different from the ABDs interface ?

On 2019-11-15 10:02:41, user Benjamin Bauz wrote:

Wonderful<br /> So many things to learn from the Turkana Project<br /> Can't wait for the next paper

On 2019-11-14 22:54:28, user Jubin Rodriguez wrote:

Also, this title would read better: A Simple Bioinformatics Pipeline for Reference-based Transcriptome Analyses. The current title gives a misleading impression that de novo transcriptome analysis (for e.g., in non-model organisms for which reference genomes or decent quality reference genomes are not available) is also covered.

On 2019-11-14 22:42:18, user Jubin Rodriguez wrote:

In my opinion, an incomplete tutorial; for example, where is the R code for generating the volcano plots. Also, it is important to show on the plot or as a separate heatmap the names of some of the top upregulated or downregulated genes in the dataset under study.

On 2019-11-11 05:58:46, user Saurabh Gayali wrote:

Doesn't sound like a complete paper but a tutorial. Would be great if you dockerize all tools in a single container and provide step by step guide. That should build a comprehensive tool and increase face value to this article.

On 2019-11-14 17:34:47, user Lynsey Hall wrote:

Between pre-print and final publication, the gene set analysis was re-run using a mixed linear model framework to account for LD induced correlations between gene expression and to covary for gene length and number of SNPs in the gene. This changed the results from the 5 gene sets listed in the pre-print to 2 gene-sets: abnormal<br /> CNS synaptic transmission and antigen processing and presentation of<br /> peptide antigen via MHC class I (GO:0002474). We also compared the data more thoroughly to the existing literature (which evolved substantially between our initial submission to a journal in August 2018, and our submission to HMG in June 2019). Lastly, we offered a more thorough and empirical analysis of why the TWAS based gene set results differ from the GWAS based gene set results of the same data.

On 2019-11-14 04:14:22, user Probe wrote:

Nice work. There are previous studies that use activity based probes to ascertain spatially resolved enzyme activity. Might not want to overlook these:<br /> https://www.nature.com/arti...<br /> https://www.sciencedirect.c...<br /> and other applications of the same approach.

On 2019-11-13 18:57:38, user Artem Babaian wrote:

It was a lot of fun to do this research, if you're interested in working on macp{Psi} shoot me an email and lets see if we can collaborate.

On 2019-11-13 16:39:53, user Jackie Thompson wrote:

Great to see this poster format proliferating! I'm a bit confused and concerned about calling these "prereg" posters -- from my understanding of this preprint, the posters don't serve as a preregistration; rather, the process serves as a form of pre-study peer review. It doesn't make sense to me to preregister and only THEN get feedback!

On 2019-11-13 16:20:45, user Nick R Waterfield wrote:

I would love to know if G-quadruplexes are used in bacteria as well....any hints?

On 2019-11-12 17:06:42, user Lobding wrote:

Our first BioRxiv preprint reveals the in vivo existence of RNA G-quadruplex structures! Plants are so powerful!

On 2019-11-13 16:09:39, user Desseyn wrote:

The published paper in Biol. Open (https://bio.biologists.org/... is not exactly the same as the preprint posted on bioRxiv. The preprint has been modified slightly for publication in Biol. Open.

One author name is misspelled in the preprint (BioRXIV). Readers should read Marc Le Bert instead of Marc Lebert.

On 2019-11-13 14:34:27, user Pietro Pichierri wrote:

Interesting work. I would suggest reading also:

https://doi.org/10.1371/jou...

On 2019-11-13 11:38:26, user Aaqib Sohail wrote:

Dear Travaglini,<br /> Thank you for sharing the article, very comprehensive study. Really liked it.<br /> I was trying to access the script from github, but the link is not work. Can you check if it is valid?<br /> Aaqib

On 2019-11-13 11:25:08, user Anton Kulaga wrote:

Guys, gene selection is broken, at least in Firefox

On 2019-11-13 10:55:19, user Simon Rasmussen wrote:

We are not allowed to post our reply on biorxiv so we posted it to figshare. Please have a read: https://doi.org/10.6084/m9....

On 2019-11-12 17:08:59, user Lobding wrote:

Changing the splicing fate by tuning the RNA structure features!

On 2019-11-12 08:10:48, user Biorobothuman wrote:

This is awesome! Have you tried any other proteins in the place of Smt3, and do you think this method could be used to probe structural stability or confirmation changes of the "pore blocking" protein?

On 2019-11-12 05:32:08, user John Chris wrote:

As a BCBA who uses positive and negative reinforcement regularly, I think more clarity is needed by the authors as to the process used by the schools. Negative reinforcement is meant to increase a specific behavior through the removal of a stimulus. Typically an aversive stimulus is removed. Putting sunglasses on is negatively reinforced by removing the bright light from my eyes. The student that can't do the math is taught to ask for a break or help to avoid the difficult classwork.

When dog training schools that are using positive punishment, that means they are introducing an aversive to stop a behavior. A pull on the leash is meant to decrease (or stop) the dog's continued walking. The pull is added to that situation to reduce the behavior you don't want. Shock collars are similar - an electric shock is added to stop the dog from running away.

If negative reinforcement is viewed as an aversive, it would appear to be preceded by the positive punishment. For example, I shock the dog until it stops running away. When it does this, I stop shocking the dog. This would appear to be negative reinforcement. However, it is just positive punishment to stop the running away behavior. The fact that the do stops and the shock stops could be interpreted as negative reinforcement, but that would be over-generalizing that negative reinforcement is an aversive approach to training. In this situation I described it would appear to be an after-effect.

Analyzing it using the 3-term contingency yields this analysis:<br /> Antecedent Behavior Concequence<br /> 1) Dog sees bird Dog runs toward bird Owner shocks dog<br /> 2) Owner shocks dog Dog stops running Owner stops shocking dog

On 2019-11-12 00:15:30, user Bob Dolin wrote:

Great article. Is there a public repository of CTML-encoded trials?

On 2019-11-11 19:15:52, user Maciej Śmiałek wrote:

Any comments and suggestions, also this which Journal will be interested - are very welcome.

On 2019-11-11 16:31:57, user Li Ding wrote:

that's interesting work, but I have a question to the authors about the result of reducing voltage at central electrode (5 to 1 kV). Yes, this caused ions flying slower so they can survive 4s acquisition time. Did you try to reduce the acquisition time say to 1.6 S and keep CE voltage -5kV unchanged? This in principle should give you same data quality as ion oscillated same number of cycles as if it is in -1kV and for 4s. Ion should not dye and yet you save the time! Do you have such result? If it is not as good, why?

On 2019-11-11 14:24:28, user Mathilde Boumasmoud wrote:

Very interesting paper! I am curious whether you looked from closer to the non-circularized sequences of the assemblies (e. g. plasmid 3 of E6988)? Do you assume these to be incompletely sequenced plasmids or assembly artifacts? Or could they possibly reflect the presence of 'linear plasmids'?

On 2019-11-11 09:16:28, user Colin Logie wrote:

Hi there,

The first part of this manuscript is about 4C results. It will make much more sense when you read the paper for a second time (recommended option for CTCF and TAD researchers). <br /> ALTernatively, start reading the results section of this manuscript at the first 'genome-wide' section which is "CTCF site spatial distribution analysis reveals orientation biases" and only after you have read all the genome-wide results start to read the first parts of the results section.

Sorry, I understand it is unusual these days to ask editors, reviewers and readers to re-read a paper, what with all the papers we all want to read. But, those who wish to build further on our results may find they wish to read the paper more than once. This pre-print was mainly written for them. Thanks for citing this pre-print in your publications!

Colin

On 2019-11-10 15:50:52, user David Curtis wrote:

It might be worth also citing these other papers which used similar approaches to analyse most of these exomes:

https://link.springer.com/a...<br /> https://www.nature.com/arti...

On 2019-11-10 15:23:11, user Daniel Yero wrote:

Happy to share our results on the structure and function of the soluble perisplamic component of the ABC transport system Ttg2 (otholog to Mla) in Pseudomonas aeruginosa. This system has evolved in a curious way to shape the multi-resistance of this species.

On 2019-11-09 13:44:29, user Anil wrote:

Congratulations on a nice work on delineating the transcriptional architecture of enhancers.

But, terms like "promoter", "core promoter" and "TSS" in our collective understanding have been etched in our minds to reflect gene-proximal elements necessary for gene transcription. Our knowledge of widespread intergenic transcription is only a decade old, and we are getting to learn deeper about how these transcriptional units function, and this paper is a great contribution towards that. Now, calling the similar elements within gene-distal enhancers by the very same name creates confusion. I believe there is a strong need to distinguish enhancer-embedded promoter elements from the classical meaning of such terms that normally refer to such elements at the 5' end of genes.

I suggest the authors, instead, use "ePromoter", "core ePromoter" and "eTSS" to reflect such elements that are embedded within enhancers and that primarily initiate enhancer transcription - a prefix "e" signifying the elements' location at/within enhancers.

On 2019-11-09 03:30:57, user Qiongshi Lu wrote:

There is a typo in the link to our summary statistics files. The correct link is ftp://ftp.biostat.wisc.edu/... <br /> This will be addressed in an update next week

On 2019-11-08 22:48:59, user Tyson V. Sharp wrote:

A very interesting manuscript. The notion that LIMD1 forms phase separated condensates is not a novel finding as we have reported twice (James V et al PNAS 2010 and Bridge KS et al Cell Reports) that LIMD1 is required for the formation of P-bodies which it also localises to. This should be addressed in the work? You kindly referenced our other two papers where we define LIMD1 as a TSG. Why not the two that describe P-body biology? This would seem critical for your study?

Why not look at Ajuba and WTIP as controls for functional overlap? Or indeed controls for the specifics of this LIMD1-FA function? <br /> What happens if you KO Ajuba And LIMD1 together?

That would be very interesting indeed? I have many more questions but this will do for starters!

Very nice work indeed!

Tyson

On 2019-11-08 16:26:18, user V Blaine wrote:

THCA might be the diet pill that could revolutionize the obesity industry. I suppose it is related to cannabinoid hyperemesis syndrome. And it is based on the theory that reducing a drug causes the opposite of what the drug causes. THC increases appetite and THCA causes limited appetite. I am only giving my two cents based on my experiences so that researchers can perhaps test some of these hypotheses as cannabis becomes more accepted.

On 2019-11-08 14:34:43, user Joan Slonczewski wrote:

Our revised manuscript, "Acid Experimental Evolution of the Extremely Halophilic Archaeon Halobacterium sp. NRC-1 Selects Mutations Affecting Arginine Transport and Catabolism," is now posted here:<br /> https://www.researchsquare....

On 2019-10-22 01:26:36, user Joan Slonczewski wrote:

Thanks to Friedhelm Pfeiffer for pointing out that the gene annotated nhaC3 actually is "arginine/ornithine antiporter ArcD" as shown by physiology (PMID:18930051 DOI:10.1016/j.febslet.2008.10.004). The arginine antiporter makes sense for acid regulation, a fascinating parallel with previous findings for E. coli. We have updated our annotations throughout, and will post the revision shortly.

On 2019-11-08 09:19:06, user Gopal Gowane wrote:

Good start! Indeed that is required, however, cost for genotyping is really a big hurdle. For cattle its fine, but we cant thi9nk for sheep and goat

On 2019-11-07 09:02:29, user atul Kumar Pandey wrote:

The article explains the complexity associated with Juvenile Hormone functions in Insects with particular attention on Bumble bee, Bombus terrestris.

On 2019-11-06 17:07:46, user Steve wrote:

The methods section of this manuscript is incomplete and poorly described, which makes interpreting the data impossible.

On 2019-11-05 21:32:31, user CJ Battey wrote:

Very interesting work! I have no background in METE but have been studying some spatial population genetic models that come to similar conclusions and so might be useful to reference just in the interests of drawing some parallels across subdisciplines. Apologies if this is all familiar to you.

Briefly, Wright and Malecot both tried to model genetic differentiation in continuous space in the mid 20th century (see e.g. Malecot's "The Mathematics of Heredity" and Wright's several papers from 1942-1946 on isolation by distance). These models imagine individuals dispersing across a landscape and finding mates in a local region, with fitness (i.e. the number of offspring) independent of local population density. These models yield nice analytic solutions for e.g. heterozygosity of a population, but Felsenstein (1974, "A Pain in the Torus...") showed that the model leads to individuals clustering in space over time. In the worst case of an infinite one-dimensional habitat all individuals eventually end up in the same cluster. That was, frankly, a bummer for population genetics because it meant the mathematically tractable Wright and Malecot models did not actually produce realistic spatial patterns, so the field mostly moved to discretized models and dropped continuous space.

One solution to the clustering problem, however, is to add density dependance to the model -- either by adjusting the number of offspring or the probability of survival. This doesn't lead to nice analytic solutions but does allow you to simulate individuals reproducing in continuous space without clustering over time. A few studies used density-dependance in this way, including Doebeli and Dieckmann 2003 ("Speciation along environmental gradients"), though to my knowledge no one studied that aspect in detail from a population genetics point of view. I have a preprint up where we do study genetic variation of individuals simulated in continuous space with density-dependent fitness (https://www.biorxiv.org/con..., for which we should have a new revision up in a couple weeks that will include a relevant figure (also here, in case you're interested). Anyways this is all a very long way of saying that in our simulations adding density-dependance also leads to less spatial agglomeration, which is consistent with your findings using a very different model. Which: neat. Good luck with the review process. Best, https://uploads.disquscdn.c...

CJ

On 2019-11-05 20:54:18, user Diego Crespo wrote:

Link to the full paper: https://dev.biologists.org/...

On 2019-11-05 15:38:16, user David Curtis wrote:

It might have been appropriate to cite some of my publications implementing an approach to incorporate functional weights into burden-testing:

https://www.nature.com/arti...<br /> https://link.springer.com/a...<br /> https://www.ingentaconnect....

This approach is implemented in the GENEVARASSOC and SCOREASSOC programs. Although it isn't identical to yours, it's similar enough in spirit to be worth referring to.

On 2019-11-04 23:47:42, user Misha Koksharov wrote:

I have a couple of questions on the 1SY food which is typically used by your labs:

1) How important is the use of autolysed Brewer's yeast (MP Bio cat # 103304) vs the non-autolysed Brewer's yeast (MP Bio cat # 903312)? In most subsequent publications after Bass et al, 2007 it is stated that the autolysed yeast are used (e.g. Wong et al, 2009). Though in Toivonen et al 2007 and in this publication it is #903312 (non-autolysed).

2) How good is the 1SY food as a "standard" food for continuous routine fly keeping?

On 2019-11-04 13:55:44, user AspiringPolymath wrote:

Please consider the logic for eukaryotic complexity that you use in the second sentence in an alternative biological context: blue whales (larger) and water fleas (more genetic material), would be, by your definition, more complex than humans. Also, can you please describe why lipid compartmentalization is more complex than the protein compartmentalization seen in prokaryotes? Assuming the answer is because it allows multicellularity, consider that plants cannot produce their own nitrogen and so could not exist without prokaryotes being present. Is eukaryotic complexity derived from being the majority structural component of a plant?

On 2019-11-04 07:49:30, user Nikhil Ratna wrote:

Extremely important research. I have a question.<br /> In the methods it is given as, "We removed 133 individuals with a<br /> comorbid diagnosis of bipolar disorder, schizophrenia, schizotypy or schizoaffective disorder (since these are likely to share risk genes for psychiatric disorders independently of their HD status)"<br /> What is the basis of classifying people as psychiatric patients of HD independent of HD status? Since psychiatric symptoms in HD are of broadest range and can present in any stage of the disease, how is it possible to say that the psychiatric symptoms in an individual are related to or not specifically to CAG status.

On 2019-11-03 17:07:23, user Björn Rissiek wrote:

If you want to improve cell recovery and vitality for flow cytometry analyzes you could use the ARTC2.2 blocking nanobody s+16a. We and others have shown that ARTC2.2.-mediated ADP-ribosylation of the P2X7, which is expressed by various memory T cell populations, kills alot of those T cells when they are ex vivo incubated at 37 °C.

See also:<br /> https://www.fasebj.org/doi/...<br /> https://jlb.onlinelibrary.w...<br /> https://www.frontiersin.org...

On 2019-11-03 11:19:28, user Jubin Rodriguez wrote:

Also, I want to use Shengwei's growthpred tool but there isn't any sort of documentation available with it and docker pull command fails (I wonder if this is due to growthpred bein a private repository?)

On 2019-11-03 10:40:59, user Jubin Rodriguez wrote:

Very interesting study! I was curious to know from the personal experience of the authors if GRiD exhibits better accuracy than iRep with regards to estimation of replication rates (of course I understand from reading your paper that both methods apparently do not fare that great with slowly growing bacteria!)? The reason I ask is cos' I'm getting a lot inconsistent replication rate values between iRep and GRiD for the same MAGs. For example I have a MAG (75.1% completeness and zero contamination & strain heterogeneity; 12X coverage) for which iRep estimates the replication index value to be 1.95 while GRiD outputs a value of 1.24. My gut feeling is that the GRiD estimated value is more closer to reality here but there's no way for me to be absolutely sure here since I've not made any real-lab measurements like you've done for your study. I was hoping to know if you would have any sort of inputs here to share from your own experiences?

On 2019-11-03 09:39:28, user Tobias Aurelius Knoch wrote:

we would like to point the readers of this manuscript to the following list of some publications for consideration of its background and importance. for the authors of the below works…

PATENT: Method for analysing the interaction of nucleotide sequences in a three-dimensional DNA structure. Knoch TA, Grosveld FG, International Patent Application WO2014IB02485 2014.

Targeted Chromatin Capture (T2C): a novel high resolution high throughput method to detect genomic interactions and regulatory elements.<br /> Kolovos P, van de Werken HJ, Kepper N, Zuin J, Brouwer RW, Kockx CE, Wendt KS, van IJcken WF, Grosveld F, Knoch TA.<br /> Epigenetics Chromatin. 2014 Jun 16;7:10. doi: 10.1186/1756-8935-7-10. eCollection 2014.

The detailed 3D multi-loop aggregate/rosette chromatin architecture and functional dynamic organization of the human and mouse genomes.<br /> Knoch TA, Wachsmuth M, Kepper N, Lesnussa M, Abuseiris A, Ali Imam AM, Kolovos P, Zuin J, Kockx CEM, Brouwer RWW, van de Werken HJG, van IJcken WFJ, Wendt KS, Grosveld FG.<br /> Epigenetics Chromatin. 2016 Dec 24;9:58. doi: 10.1186/s13072-016-0089-x. eCollection 2016.

Dynamic properties of independent chromatin domains measured by correlation spectroscopy in living cells.<br /> Wachsmuth M, Knoch TA, Rippe K.<br /> Epigenetics Chromatin. 2016 Dec 24;9:57. doi: 10.1186/s13072-016-0093-1. eCollection 2016.

The 3D structure of the immunoglobulin heavy-chain locus: implications for long-range genomic interactions.<br /> Jhunjhunwala S, van Zelm MC, Peak MM, Cutchin S, Riblet R, van Dongen JJ, Grosveld FG, Knoch TA, Murre C.<br /> Cell. 2008 Apr 18;133(2):265-79. doi: 10.1016/j.cell.2008.03.024.

Simulation of different three-dimensional polymer models of interphase chromosomes compared to experiments-an evaluation and review framework of the 3D genome organization.<br /> Knoch TA.<br /> Semin Cell Dev Biol. 2019 Jun;90:19-42. doi: 10.1016/j.semcdb.2018.07.012. Epub 2018 Aug 24. Review.

A Guided Protocol for Array Based T2C: A High-Quality Selective High-Resolution High-Throughput Chromosome Interaction Capture.<br /> Knoch TA.<br /> Curr Protoc Hum Genet. 2018 Oct;99(1):e55. doi: 10.1002/cphg.55. Epub 2018 Sep 10.

Investigation of the spatial structure and interactions of the genome at sub-kilobase-pair resolution using T2C.<br /> Kolovos P, Brouwer RWW, Kockx CEM, Lesnussa M, Kepper N, Zuin J, Imam AMA, van de Werken HJG, Wendt KS, Knoch TA, van IJcken WFJ, Grosveld F.<br /> Nat Protoc. 2018 Mar;13(3):459-477. doi: 10.1038/nprot.2017.132. Epub 2018 Feb 8.

Light optical precision measurements of the active and inactive Prader-Willi syndrome imprinted regions in human cell nuclei.<br /> Rauch J, Knoch TA, Solovei I, Teller K, Stein S, Buiting K, Horsthemke B, Langowski J, Cremer T, Hausmann M, Cremer C.<br /> Differentiation. 2008 Jan;76(1):66-82. Epub 2007 Nov 26.

Binding of nuclear factor κB to noncanonical consensus sites reveals its multimodal role during the early inflammatory response.<br /> Kolovos P, Georgomanolis T, Koeferle A, Larkin JD, Brant L, Nikolicć M, Gusmao EG, Zirkel A, Knoch TA, van Ijcken WF, Cook PR, Costa IG, Grosveld FG, Papantonis A.<br /> Genome Res. 2016 Nov;26(11):1478-1489. Epub 2016 Sep 15

On 2019-11-03 02:31:15, user Özen Karaocak wrote:

When will the dataset be available? It's been months.

On 2019-11-02 16:02:40, user hiplady wrote:

My hearing has been bad for quite a while. Would be interested in more information on this product.<br /> Thank you

On 2019-11-01 14:24:00, user Anna James-Bott ???????????????? wrote:

Hey! Where can I find the supplementary tables?

On 2019-11-01 13:24:46, user Anna wrote:

Hey, cool method! Where can I find the supplementary tables? Thanks!

On 2019-11-01 09:02:28, user Lars Forsberg wrote:

This paper is now published online in European Journal of Human Genetics https://www.nature.com/arti...

Danielsson, M., Halvardson, J., Davies, H. et al. Longitudinal changes in the frequency of mosaic chromosome Y loss in peripheral blood cells of aging men varies profoundly between individuals. Eur J Hum Genet (2019) doi:10.1038/s41431-019-0533-z

A link to the published article on this page is forthcoming.

Lars Forsberg

On 2019-11-01 07:55:21, user Pierluigi Scerbo wrote:

Wow! A new paper about in vivo reprogramming! It is amazing and looks like my PhD paper about in vivo reprogramming of Xenopus muscles to Xenopus-IPS (XiPS) (PMID: 22232554).

On 2019-11-01 05:00:37, user Davidski wrote:

P.S.

Actually, it's worse than I thought. Your paper suggests that there's elevated North African (NAF) ancestry among present-day Poles!

On 2019-11-01 04:21:26, user Davidski wrote:

Hello authors,

There's a problem with the samples in your dataset.

The modern samples representing Poland show a dip in Yamnaya ancestry and a high in Anatolian ancestry relative to their geography. So it's extremely unlikely that they're ethnic Poles.

You might want to download the full Human Origins dataset which includes ethnic Poles genetically representative of western and eastern Poland.

This will help you produce more accurate spatio-temporal models for East Central Europe.

On 2019-10-31 22:47:24, user Ben Hindson wrote:

Thanks Caleb, Sai, Fabiana and Jason for bringing this to our attention. We just posted our response on our blog. <br /> https://community.10xgenomi...<br /> Reach out to our support team if you have questions.

On 2019-10-31 20:29:46, user Charles Warden wrote:

I had previously heard of CellProfiler (and CellProfilerAnalyst), but I think this title and abstract is a little confusing to someone new to the area.

In other words, you are providing a pipeline of scripts to use CellProfiler for a specific application, but you didn't develop CellProfiler itself.

If you re-word this, then I think that is fine. Being able to create new modules and specific protocols to expand applications for open-source software is important!

On 2019-10-31 17:29:26, user Charles Warden wrote:

I thought this was very interesting - thank you for posting this!

Have you seen any evidence that the barcoding can have any effect on mixed sample calls? If you barcode 100 (or 1000) samples, are you getting a similar number of reads for each sample?

On 2019-10-31 12:09:03, user yashvant patel wrote:

Real assessment of Vitamin B12 from Oyster mushroom, thereafter, It could be a good source of this vitamin and consumption of this mushroom would enhances the health.

On 2019-10-31 11:27:30, user Matt Iadanza wrote:

The initial model generation is really clever! One question: shouldn't this be 'ls' instead of 'cat'? https://uploads.disquscdn.c...

On 2019-10-31 10:54:21, user Saul Burdman wrote:

A slightly modified version of this article was published few days ago (as online early view) in Molecular Plant Pathology, with the following title: Show me your secret(ed) weapons: a multifaceted approach reveals a wide arsenal of type III‐secreted effectors in the cucurbit pathogenic bacterium Acidovorax citrulli and novel effectors in the Acidovorax genus (doi.org/10.1111/mpp.12877):RK-PAcqhko4VDH0teDBPdAcX4g8 "doi.org/10.1111/mpp.12877)").

On 2019-10-23 17:19:03, user Saul Burdman wrote:

A slightly modified version of this article was published today (as online early view) in Molecular Plant Pathology, with the following title: Show me your secret(ed) weapons: a multifaceted approach reveals a wide arsenal of type III‐secreted effectors in the cucurbit pathogenic bacterium Acidovorax citrulli and novel effectors in the Acidovorax genus (doi.org/10.1111/mpp.12877):RK-PAcqhko4VDH0teDBPdAcX4g8 "doi.org/10.1111/mpp.12877)").

On 2019-10-31 06:38:40, user Prem Prakash wrote:

Heartiest congratulations to Stachwell and co-workers for this seminal and groundbreaking study that was, as stated above, subsequently published as: "Genetic manipulation of cell line derived reticulocytes enables dissection of host malaria invasion requirements" (Nature Communications, 10: 3806; 2019; https://www.nature.com/arti....

This response pertains to the Stachwell et al observation on the human Cyclophilin B (CypB) receptor in relation to parasite invasion and their report that the knockout of cyclophilin B in their reticulocyte cell-line derived through in vitro differentiation of an enucleation-competent immortalized erythroblast cell line BEL-A (that caused 20% less expression of GPC) resulted in "no significant difference" in the invasive susceptibility of reticulocytes.

As the primary author of the 2017 Nature Communications paper that reported the role of Human Cyclophilin in the parasite invasion process (https://www.nature.com/arti..., it is believed the following observations would add to the discussion and analysis surrounding the seminal Stachwell study:

It is to be noted that Stachwell et al state (https://static-content.spri... that they have not tested either the anti-malarial drug and Cyclophilin inhibitor Cyclosporin A (CsA) or CDP3, the peptide inhibitor of CypB, on their WT and CypB KO BEL-A lines, because, as they emphasise, neither CsA nor CDP3 are specific to cyclophilin B over other cyclophilins. CsA binds to multiple cyclophilins, and CDP3 binds to cyclophilin A and perhaps other cyclophilins. Therefore, as Stachwell et al observe: "Our CypB KO line contains multiple known binding partners for these drugs, which would convolute the results of this experiment."

Furthermore, Stachwell et al state (https://static-content.spri... that they found "CsA to be poorly soluble, making such experiments practically challenging. Finally, BEL-A derived reticulocyte production and subsequent parasite invasion assays represent a significant time and cost investment, which, given the specificity and solubility concerns, we decided to devote to other experiments to improve the manuscript."

We believe that, in addition to being an important advance in Malaria research, the findings of Stachwell et al are entirely in keeping with the nature of Cyclophilin inhibition that was reported by us in 2017 (https://www.nature.com/arti..., for the following reasons:

A. Cyclophilins, the Cyclosporin A (CsA) binding proteins and their role in malaria pathogenesis.<br /> Cyclophilins were first identified as binder of Cyclosporin A (CsA) and hence named cyclophilins [1-4]. In humans, there are 16 Cyclophilins reported till today. Of these, Cyclophilin A (CypA) was the first cyclophilin discovered in mammalian host and is the most prevalent followed by the Cyclophilin B (CypB) [2, 5].

CypA and CypB share ~80% sequence similarity overall, and 70% identity in their core region [6]. CypB is found on the surface while CypA has been found to be secreted outside of the cell in response to inflammatory responses and oxidative stress [7-10].

When CsA binds to Cyclophilins, they form an intracellular ternary complex involving Calcineurin [11]. CsA potently blocks the enzymatic activity of cyclophilins [11]. Furthermore, extracellular CypB has been reported to mediate the incorporation of CsA in T-lymphocytes [12]. Also, because the structure of CypB and CypA is almost identical, the CsA-binding pocket in CypB has the same structure as in CypA, with CsA exhibiting similar conformation and network of interactions when it is bound to these proteins [11, 13-15]. Yet another study shows that cyclophilins were identified as erythrocyte cyclosporine-binding proteins [16]. Finally, in addition to our observation that CypB is found on the surface of the RBC, a recent study confirms the presence of CypA and CypB in the RBC ghost fraction through LC-MS/MS [17].

The antimalarial activity of CsA was identified as early as in 1981 when it was shown that CsA inhibits the parasite invasion in human cells in vitro and in vivo in a mice model [18, 19]. In another report, low dose treatment with CsA showed inhibitory effect on the mice cerebral malaria pathogenesis [20]. These studies indicate that the mode of action of CsA as antimalarial is through binding to the Cyclophilins.

Basigin has been identified as the extracellular receptor for CypA and CypB [23]. The interaction between Cyclophilins and Basigin has been implicated in many medical conditions including inflammatory responses, cardiovascular disease, and rheumatoid arthritis [1]. The interaction of cyclophilins with Basigin on the cell surface has been proven by utilising monoclonal antibodies against Basigin. Treatment of CHO cells with anti-Basigin antibodies inhibits the Cyclophilin-induced chemotaxis response [21, 23].

Summary: CsA binds to Cyclophilin B and this binding is responsible for the potent anti-malarial activity of the drug. Cyclophilin binds to Basigin.

B. Our Study I:<br /> 1. Our 2017 finding (https://www.nature.com/arti..., that cyclophilin B is a host receptor for PfRhopH3 starts with the discovery of human cyclophilin B as an interacting partner of PfRhopH3-C through screening a human lung c-DNA library using the bacterial two hybrid system (B2H). Further, in our study, this interaction has been confirmed through a range of in vitro techniques like ELISA, SPR, Far-Western Blot, Co-IP and an in vivo B2H β-galactosidase assay. (Figure 1b, c, Figure 3a, b, c, d)

1. Subsequently, to confirm that PfRhopH3 is a ligand for CypB, we performed binding and co-localisation studies with recombinant CypB on the surface of merozoite. Reciprocally, binding and co-localization of PfRhopH3-C was also studied on the RBC surface. (Figure 1d, e; Figure 2a, b; Figure 3e; Supplementary fig 2d, e)

2. Basigin has earlier been reported as a receptor/interacting partner, both for PfRH5 as well as CypB. Therefore, to confirm this we performed IFA based expression and co-localization of CypB and Basigin on the RBC surface and found that these two proteins are present together on the RBC surface. Reciprocally, Basigin and CypB bind on the merozoite and are co-localized. (Figure 4a, b, c)

3. To establish the significance of CypB and PfRhopH3 during invasion, we then looked for a peptide inhibitor from a de novo peptide library and fished out CDP3 as a binder of CypB. CPD3 was tested for the disruption of interaction of PfRhopH3-C and CypB using Bacterial Three Hybrid System (B3H) and ELISA-based competitive inhibition assay. In both experiments CDP3 was found to potently disrupt the interaction. (Figure 5; Supplementary fig. 7)

4. Notably, as CypA and CypB are found to be sharing 70% sequence identity in their core region, we obtained the recombinant CypA protein. Utilising the ELISA based interaction experiments, the interaction of CypA with RhopH3-C as well as CDP3 was tested and it was found that CypA can also interact with these two molecules very strongly. (Supplementary fig. 12)

5. Following on, we tested the effect of CDP3 on invasion through two different assays. In the first, CDP3 was added to 2% haematocrit with 1% tightly synchronised schizont stage of 3d7 and Dd2 culture. It was observed that parasitaemia decreased in response to an increasing dose of CDP3. In the second, the RBCs were first treated with different doses of CDP3 and these RBCs were then used for the invasion assay. In this experiment, we again observed a dose dependent reduction in parasitaemia. (Figure 6a, b)

Summary: CypB is expressed on the RBC surface. Both CypB and CypA – because of their sequence identity – interact not only with RhopH3C but also with the CDP3 peptide. CDP3 reduces parasitaemia in a dose-dependent manner.

C. Our study II: Cyclosporin A and malaria invasion<br /> 1. As stated above, it is well-established that Cyclophilin is the target protein and binder for Cyclosporin A (CsA). To confirm the role of host cyclophilins in invasion, we used CsA, first, to treat the RBC for 1 h (https://www.nature.com/arti.... Unbound CsA was washed out extensively and the treated RBCs were then used for invasion assay. In the second experiment, CsA was introduced directly to culture. In both assays, a significant reduction in the parasitaemia was observed, suggesting that host Cyclophilins (CypB and CypA) are directly involved in the parasite invasion. Notably, when CsA was directly added in the culture, reduction in parasitaemia was more for the same concentration as compared to the washed RBC. This observation suggest that CsA potentially blocks the host cyclophilins (CypA, CypB) as well as parasite Cyclophilins. (Figure 6c, d, e). This is a crucial point, in that the drug CsA, that binds to CypB, blocks invasion when it is introduced either i) to the RBC first, and ii) to a culture.<br /> 2. In our in vivo experiments, when CsA is administered to the Balb/C mice followed by challenging of the latter with the parasite, we found growth to be completely absent in treated mice as compared to the untreated mice (Supplementary Fig. 11a). All the untreated mice died by day 20 whereas the treated mice survived beyond day 30 (Supplementary Fig. 11b). This observation corroborates the original finding that CsA acts an antimalarial drug targeting the Cyclophilins [18-22].

Summary: Based on our experimental findings and previous reports, host cyclophilins play a role during parasite invasion. Small molecules like CsA and CDP3 can potentially block these receptors on the cell surface and prevent the parasite entry in isolation or together with their complex.

D. Our explanation in view of the Satchwell et al report: <br /> 1. There are as many as 16 host cyclophilins out of which, just to give one example, CypB shares an overwhelming similarity (80% overall) and idenditity (70% in the core region) with CypA [24]. It is entirely conceivable, therefore, that in the absence of CypB, other cyclophilins – most notably CypA, whose copy number, it should be emphasised, is greater than CypB in RBC ghost [17] – have taken its place for carrying out its receptor-mediated role. And while CRISPR/Cas9 has removed completely from the system only CypB, all other cyclophilins are still in place.

1. Additionally, while the CRISPR KO has removed only CypB, inhibitors of Cyclophilins like CsA and our peptide CDP3 (that we have shown binds also to CypA) bind and incapacitate all Cyclophilins, thereby preventing the swapping of blocked (KOed) Cyclophilin with the unblocked one.

2. This is brought out by the observation that CsA, when introduced to the RBC, and the latter then washed, blocks the invasion potently; meaning that Cyclophilins (one, maybe more, maybe all) are important for the invasion process because the drug that binds to them for its potent activity goes on to kill the parasite because of this very binding.

3. As Satchwell et al have themselves stated (https://static-content.spri... neither CsA nor CDP3 are specific to cyclophilin B over other cyclophilins. CsA binds to multiple cyclophilins, and CDP3 binds to cyclophilin A and other cyclophilins. Therefore, their CypB KO line contains multiple known binding partners for these drugs. It stands to reason that if both CsA and CDP3 have multiple cyclophilin targets, a knock-out of just one of sixteen available cyclophilins would not have the desired effect on invasive susceptibility.

E. References:<br /> 1. Yurchenko, V., et al., Cyclophilin-CD147 interactions: a new target for anti-inflammatory therapeutics. Clin Exp Immunol, 2010. 160(3): p. 305-17.<br /> 2. Wang, P. and J. Heitman, The cyclophilins. Genome Biol, 2005. 6(7): p. 226.<br /> 3. Harding, M.W. and R.E. Handschumacher, Cyclophilin, a primary molecular target for cyclosporine. Structural and functional implications. Transplantation, 1988. 46(2 Suppl): p. 29S-35S.<br /> 4. Handschumacher, R.E., et al., Cyclophilin: a specific cytosolic binding protein for cyclosporin A. Science, 1984. 226(4674): p. 544-7.<br /> 5. Galat, A., Peptidylprolyl cis/trans isomerases (immunophilins): biological diversity--targets--functions. Curr Top Med Chem, 2003. 3(12): p. 1315-47.<br /> 6. DeBoer, J., C.J. Madson, and M. Belshan, Cyclophilin B enhances HIV-1 infection. Virology, 2016. 489: p. 282-91.<br /> 7. Price, E.R., et al., Human cyclophilin B: a second cyclophilin gene encodes a peptidyl-prolyl isomerase with a signal sequence. Proc Natl Acad Sci U S A, 1991. 88(5): p. 1903-7.<br /> 8. Price, E.R., et al., Cyclophilin B trafficking through the secretory pathway is altered by binding of cyclosporin A. Proc Natl Acad Sci U S A, 1994. 91(9): p. 3931-5.<br /> 9. Nigro, P., et al., Cyclophilin A is an inflammatory mediator that promotes atherosclerosis in apolipoprotein E-deficient mice. J Exp Med, 2011. 208(1): p. 53-66.<br /> 10. Nigro, P., G. Pompilio, and M.C. Capogrossi, Cyclophilin A: a key player for human disease. Cell Death Dis, 2013. 4: p. e888.<br /> 11. Mikol, V., J. Kallen, and M.D. Walkinshaw, X-ray structure of a cyclophilin B/cyclosporin complex: comparison with cyclophilin A and delineation of its calcineurin-binding domain. Proc Natl Acad Sci U S A, 1994. 91(11): p. 5183-6.<br /> 12. Allain, F., A. Denys, and G. Spik, Cyclophilin B mediates cyclosporin A incorporation in human blood T-lymphocytes through the specific binding of complexed drug to the cell surface. Biochem J, 1996. 317 ( Pt 2): p. 565-70.<br /> 13. Kallen, J., et al., Structure of human cyclophilin and its binding site for cyclosporin A determined by X-ray crystallography and NMR spectroscopy. Nature, 1991. 353(6341): p. 276-9.<br /> 14. Ke, H.M., et al., Crystal structure of recombinant human T-cell cyclophilin A at 2.5 A resolution. Proc Natl Acad Sci U S A, 1991. 88(21): p. 9483-7.<br /> 15. Dornan, J., P. Taylor, and M.D. Walkinshaw, Structures of immunophilins and their ligand complexes. Curr Top Med Chem, 2003. 3(12): p. 1392-409.<br /> 16. Foxwell, B.M., et al., Identification of cyclophilin as the erythrocyte ciclosporin-binding protein. Biochim Biophys Acta, 1988. 938(3): p. 447-55.<br /> 17. Gautier, E.F., et al., Absolute proteome quantification of highly purified populations of circulating reticulocytes and mature erythrocytes. Blood Adv, 2018. 2(20): p. 2646-2657.<br /> 18. Thommen-Scott, K., Antimalarial activity of cyclosporin A. Agents Actions, 1981. 11(6-7): p. 770-3.<br /> 19. Nickell, S.P., L.W. Scheibel, and G.A. Cole, Inhibition by cyclosporin A of rodent malaria in vivo and human malaria in vitro. Infect Immun, 1982. 37(3): p. 1093-100.<br /> 20. Grau, G.E., D. Gretener, and P.H. Lambert, Prevention of murine cerebral malaria by low-dose cyclosporin A. Immunology, 1987. 61(4): p. 521-5.<br /> 21. Marin-Menendez, A. and A. Bell, Overexpression, purification and assessment of cyclosporin binding of a family of cyclophilins and cyclophilin-like proteins of the human malarial parasite Plasmodium falciparum. Protein Expr Purif, 2011. 78(2): p. 225-34.<br /> 22. Berriman, M. and A.H. Fairlamb, Detailed characterization of a cyclophilin from the human malaria parasite Plasmodium falciparum. Biochem J, 1998. 334 ( Pt 2): p. 437-45.<br /> 23. Yurchenko, V., et al., CD147 is a signaling receptor for cyclophilin B. Biochem Biophys Res Commun, 2001. 288(4): p. 786-8.<br /> 24. Davis, T.L., et al., Structural and biochemical characterization of the human cyclophilin family of peptidyl-prolyl isomerases. PLoS Biol, 2010. 8(7): p. e1000439.

On 2019-10-31 06:13:12, user Anonymous spectator wrote:

The usage of the term "epimutants" appears suspect here.

WGS of unstable isolates revealed no genetic changes in coding sequences involved in either caffeine resistance or H3K9me2-mediated silencing....

S. pombe does have a lot of non-coding sequences that have functional roles too. Take a random example like....ncRNA.394. Probably shouldn't rule out genetic mutations in those regions.

"In addition to a heterochromatin domain over ncRNA.394, analysis of ChIP-seq input DNA indicated that many independent unstable caffeine-resistant isolates also contained overlapping regions of chromosome III present at increased copy number"

Widespread gene duplications sound like genetic mutations.

But aside from the quasi-sarcasm, most if not all of those "new" domains have already been noted in the literature. And stress/environmental response genes have already been shown to be enriched at heterochromatin domains that form outside of the typical lab condition (ie. rich media, 30-32oC, plenty of air, etc.). On a more positive note, this work does a pretty fine job of showing that heterochromatin rewiring doesn't just happen during adaption to cold stress (previously demonstrated several times already) but are also at work during caffeine stress.

On 2019-10-24 18:48:16, user Shiv Grewal wrote:

The finding that heterochromatic gene silencing confers caffeine resistance in S. pombe provides an interesting example of the role of heterochromatin machinery in adaptive genome regulation, a phenomenon that has been well documented in previous studies [1, 2]. Indeed, previous work has shown that wild-type cells exposed to unfavorable growth conditions such as low temperature form new facultative heterochromatin domains (referred to as “heterochromatin islands”) [1]. In that study, heterochromatin machinery was found to be part of an adaptive gene control mechanism that prevents unrestrained transcriptional upregulation during reprogramming of the genome in response to environmental changes [1].

Several heterochromatin domains described by Torres-Garcia and colleagues as “new ectopic domains” have been identified previously in multiple studies (e.g. mcp7, ppr4, ssm4, ncRNA.394 and mbx2 are islands 1, 4, 6, 14, 16) [1, 3-5]. Indeed, the “epimutant” domain containing the ncRNA.394 and SPBC17G9.13 loci (island 14) is detectable in wild type cells, in particular in cells cultured at low temperature. This island contains chromosomal-internal telomeric repeats that provide a binding site for Shelterin, which nucleates heterochromatin through recruitment of Clr4 histone methyltransferase [4]. Relevant to the findings by Torres-Garcia et al., Shelterin-dependent islands were found to repress expression of nearby stress-induced genes [4]. Given the requirement for Shelterin binding to the telomeric repeat DNA sequence, the description of the observed effects as “epimutants” (defined as sequence independent heritable changes) might need to be reconsidered.

As an additional note, the original article that demonstrated the binding of Clr4 chromodomain to methylated histone H3 tail and the “read-write” mechanism is Zhang et al., 2008 [6].

Shiv Grewal<br /> National Cancer Institute<br /> National Institutes of Health<br /> Bethesda, Maryland

References cited:

1. Gallagher, P.S., M. Larkin, G. Thillainadesan, J. Dhakshnamoorthy, V. Balachandran, H. Xiao, C. Wellman, R. Chatterjee, D. Wheeler, and S.I.S. Grewal, Iron homeostasis regulates facultative heterochromatin assembly in adaptive genome control. Nat. Struct. Mol. Biol., 2018. 25: 372-383.<br /> https://drive.google.com/op...

2. Yamanaka, S., S. Mehta, F.E. Reyes-Turcu, F. Zhuang, R.T. Fuchs, Y. Rong, G.B. Robb, and S.I. Grewal, RNAi triggered by specialized machinery silences developmental genes and retrotransposons. Nature, 2013. 493: 557-560.<br /> https://drive.google.com/op...

3. Zofall, M., S. Yamanaka, F.E. Reyes-Turcu, K. Zhang, C. Rubin, and S.I. Grewal, RNA elimination machinery targeting meiotic mRNAs promotes facultative heterochromatin formation. Science, 2012. 335: 96-100.<br /> https://drive.google.com/op...

4. Zofall, M., D.R. Smith, T. Mizuguchi, J. Dhakshnamoorthy, and S.I.S. Grewal, Taz1-Shelterin promotes facultative heterochromatin assembly at chromosome-internal sites containing late replication origins. Mol. Cell, 2016. 62: 862-874.<br /> https://drive.google.com/op...

5. Vo, T.V., J. Dhakshnamoorthy, M. Larkin, M. Zofall, G. Thillainadesan, V. Balachandran, S. Holla, D. Wheeler, and S.I.S. Grewal, CPF recruitment to non-canonical transcription termination sites triggers heterochromatin assembly and gene silencing. Cell Rep., 2019. 28: 267-281 e265.<br /> https://drive.google.com/op...

6. Zhang, K., K. Mosch, W. Fischle, and S.I. Grewal, Roles of the Clr4 methyltransferase complex in nucleation, spreading and maintenance of heterochromatin. Nat. Struct. Mol. Biol., 2008. 15: 381-388.<br /> https://drive.google.com/op...

On 2019-10-29 08:38:39, user Fridtjof lund-johansen wrote:

The next thing is to use this for antibody-based interactomics!

On 2019-10-29 08:37:21, user Fridtjof lund-johansen wrote:

Very relevant for interactomics reserach!

On 2019-10-29 08:35:14, user Tim Gruene wrote:

It would be nice to see a discussion whether phasing is really based on radiation damage. Direct methods for small molecules is greatly facilitated in the presence of heavy atoms (Zinc in this case). Substracting one data set from another is prone to leave traces of the heavy atom contribution, even in the absence of radiation damage, and in particular with the great errors that ED data still have nowadays. In that case, the structure was solved simply from the Patterson map, which is not new. For ED data this has been discussed and used as early as in the 1980s (Fryer & Dorset (eds), Electron Crystallography of Organic Molecules, 1990).

Maybe you can apply the same procedure within each set, BEFORE and AFTER, with a number of arbitrary partitions. If you can reliably demonstrate that this fails, it is a good indicator that the claim in the title is indeed correct. W.r.t. literature, you may want to cite at least one of the publications of Raimond Ravelli and Max Nanao, who introduced RIP-phasing. A good reference for Sheldrick's rule is Morris & Bricogne (2004), DOI 10.1107/S090744490300163X - maybe scientifically more sound than 'in practice ~1A', The discussion w.r.t. macromolecular crystallography seems a bit far-fetched. This is a small-molecule structure, and phasing is entirely different compared to MX structures. Considering the conclusions towards MX structures: RIP phasing has remained an academic excercise since its introduction by Ravelli/Nanao 20 years ago, even for X-ray data, where data quality is much, much better than for ED data.

It is interesting that direct methods do not work with the data, when cut to 1.4A. Given the presence of Zinc, the structure should fall out even though this is beyond Sheldrick's rule. Something seems wrong with your data - do you have any idea what this might be? The fact these are ED data would not explain this sufficiently, comparing with many other ED data sets on small molecules.

Good luck with the manuscript!,

On 2019-10-29 05:58:09, user Xabier Vázquez-Campos wrote:

Is there any reference to how the scaffold extension is performed? i.e. software. It is mentioned several times but there are no details in the methods.

Also, all the Supplementary figures are missing. Only the suppl. tables are available.

On 2019-10-18 16:28:23, user °christoph wrote:

The cumulative GC skew is a valuable tool to verify correct assembly of bacterial genomes but there are caveats. Cyanobacteria frequently lack clear GC skews (not the Synechococcus/Prochlorococcus group which have clean GC skews). Also, while most GC skews show a "V" shaped minimum region, others have rather a "U" shape, which make pinpointing the minimum difficult. Also "W" shapes are not uncommon, and it is difficult to pinpoint the minimum among several "secondary minima" although the "minimum region" seems clear. Lastly, I would suggest that you rather talk of "origin region" because *the origin* should be reserved for oriC (origin for DnaA-dependent replication from oriC). In my experience, the GC skew minimum locates close to oriC but can be up to 20 kb away.<br /> Thanks for considering!

On 2019-10-29 03:19:33, user mismatch_repair wrote:

I had a number of questions/concerns about this manuscript and its co-submitted counterpart on which I would appreciate feedback from the authors:

Some of my concerns are the following:

1) The manuscript states that I-PpoI "recognizes ~20 sites in the genome." However, in addition to a number of unique sites in genes and noncoding regions, which comprise the 20 sites you refer to, I-PpoI cuts within every 28s rDNA repeat (which you mention as a target, but which seems to be counted only once). Mammalian genomes contain many identical rDNA repeats spanning multiple chromosomes, and copy number can vary by an order of magnitude between individuals in a species because these repetitive sequences are highly prone to recombination. These repeats are difficult to sequence and not annotated on the Mus musculus reference genome. Per the NCBI entry on the murine 28s gene: "The sequences coding for ribosomal RNAs are present as rDNA repeating units distributed on chromosomes 12, 15, 16, 18 and 19. The number of rDNA repeating units varies between individuals and from chromosome to chromosome, although usually 30 to 40 repeats are found on each chromosome. These rDNA repeats are not currently annotated on the reference genome." Several publications even report an ability of 28s rDNA units to undergo coordinated copy number expansion in response to deletion events.

2) The claim that it is possible to generate "non-mutagenic" DSBs by simultaneously creating hundreds of compatible sticky-end cuts throughout the genome (primarily in highly repetitive sequences) is quite unprecedented. I am not aware of any prior publications on DNA repair claiming the existence of a 100% non-mutagenic DSB. The burden of proof for this should be high. However, the evidence provided here is insufficient to support this claim. There are numerous types of mutations: point mutations, minor indels, insertion and deletion of larger chromosomal regions, duplications, inversions, and chromosomal translocations. All of the larger chromosomal rearrangements are anticipated outcomes of simultaneously freeing compatible sticky ends throughout the genome. Point mutations/minor indels may occur but at lower rates. However, these minor mutations are the only ones directly assessed, by sequencing the genome and checking mapped reads. Detecting these larger genomic rearrangements is a challenging task even for experts in the field, and it seems the sequencing efforts did not extend beyond this. The genome reads are based on 500-bp fragments, which would make detection of most of these events impossible, even if you were looking for them. In the rarer case of a chimeric 500-bp read resulting from fusion of compatible but non-homologous sequences, the read would not map to the genome and have been discarded by your analysis. In the more likely case of a fusion between 28s cuts on different regions of a chromosome or on different chromosomes, the read would merely show a normal sequence in the 500 bp surrounding the cut but it would be impossible to discern where or on which chromosome the sequence is located among the numerous repetitive tracts throughout the genome.

3) You use a few additional methods like the Surveyor assay to assess 28s mutation, but this again can only detect point mutations. Furthermore, it relies on PCR amplification so if 2 different sequences are fused, the 28s primers would no longer amplify this. And the small size to which the DNA analyzed by Southern blot was fragmented render it similarly unable to detect rearrangements. While you prepared metaphase spreads, you did not do any banding analysis which drastically limits the ability to detect chromosomal rearrangements that do not lead to obvious changes in the shape of the whole chromosome. I do not know why your ligation-based method did not detect 28s cuts, but my guess would be failed PCR. I took a look at your target amplicon and the region between the primers is 64% GC, and immediately adjacent to one of the primers is a stretch of ~70 bp with 90% GC content. This would likely make for an extremely difficult PCR- one publication describing special conditions for amplifying 28s DNA reports that "the cloning of the rDNA gene family is very difficult" and "Sequencing primers should be far from the sequences with stretches of G or C repeats." (DOI:10.17221/3960-cjas). Your baseline "aberrant metaphase" level also seems very high. Mladenov et al. (Chromosome Translocation 2018) reported that 15 aberrant chromosomes/100 metaphases as a result of 8 single I-SceI cut sites and transient transfection leads to 30% lethality in CHO cells. 12 clusters of 4 cut sites lead to 90% cell lethality.

4) Your findings of a lack of mutagenicity from I-PpoI cutting contradict a substantial number of publications using this system. Ray Monnat, whom you cite in these manuscripts, reported that "These endonuclease-induced breaks can be repaired in vivo, although break repair is mutagenic with the frequent generation of short deletions or insertions," and also mentioned that the human genome contains ~300 28s cut sites for I-PpoI across chromosomes (1999). Other publications report the same phenomenon in yeast. Your own paper from 2015 found large deletions in mice. I-PpoI has been incorporated into a "gene drive" in mosquitos to "shred" the X chromosome and prevent the birth of female offspring. I-PpoI is derived from a mobile genetic element and its evolved purpose is to catalyze the insertion of a new sequence into the genome. Yet, you claim that in your system, it causes no genetic changes. If 100% re-ligation of cut sticky ends in the presence of many other compatible sticky ends was as likely as you suggest here, restriction enzyme-based cloning would not work, nor would numerous DNA repair reporter constructs based on this principle.

I do not see in this manuscript data that is sufficient to support the claims being made. One manuscript claims that mutations do not accumulate substantially with age and the other cites multiple sources showing they do. It appears you have generated an artificial progeroid model with genomic instability due to DSBs and that is why you see the same phenotypes as human progeroid syndromes and mouse models based on DNA repair deficiencies. It is impossible to claim epigenetic changes are responsible for the observed phenotypes when you are in all likelihood causing extensive genetic changes to these mice.

I also wonder how the following statements from a 2015 paper on the same mouse model (the only difference being cell type specific rather than ubiquitous expression) can be reconciled with the current manuscripts:

2015: "To induce nuclear translocation of ERT2-I-PpoI, PpoSTOP/+; lck-Cre mice were subjected to 2–4 intraperitoneal injections of 1 mg TAM (Sigma, resuspended in corn oil) at 24 h intervals. Animals were analyzed 4 h after the final TAM injection."

2019: "ICE mice were generated by crossing I-PpoI STOP/+ mice to CreERT2/+ mice harboring a single ERT2 fused to Cre recombinase that is induced whole body (Ruzankina et al., 2007). 4-6 month-old Cre and ICE mice were fed a modified AIN-93G purified rodent diet with 360 mg/kg Tamoxifen citrate for 3 weeks to carry out I-PpoI induction."

2015: "33% of break-spanning DNA segments yielded a chimeric DNA sequence, in which one end of the I-PpoI-flanking DNA was joined to that of a second, polymorphic I-PpoI site located ∼1 Mb downstream. No evidence for aberrant junctions was observed in break-flanking DNA from lck-Cre controls, demonstrating I-PpoI-dependent formation of these distal fusions."

2019: "Unlike other methods of creating DSBs, such as CRISPR, chemicals and radiation, I-PpoI creates "sticky DNA ends" that are repaired without inducing a strong DNA damage response or a mutation (Yang et al., co-submitted manuscript)."

2015: "Although moderate transcriptional changes can be detected in DSB-bearing genes, persistent DSB formation and repair is associated with a surprisingly stable transcriptome in vivo." "Our findings further suggest that DSB repair is necessary and sufficient to ensure the maintenance and/or restoration of break-proximal gene expression profiles and, by extension, epigenetic integrity in vivo." "Consistent with cell-intrinsic epigenetic deregulation being a minor consequence of continued DSB exposure in vivo, a recent study shows that DNA damage-induced, age-associated functional decline can be attributed in large part to systemic consequences of DSBs, including cell death, tissue atrophy and the ensuing, non-cell-autonomous inflammatory response."

2019: "We present evidence that the response to DSBs changes the compartmentalization of chromatin and introduces transcriptional and epigenetic noise that closely mimics what happens during normal aging, including hallmark changes to the histone modifications, gene expression, and DNA methylation patterns." "After repair, the epigenome is reset but not completely, leading to progressive changes to the epigenetic marks and chromatin compartmentalization of the genome." "In the parlance of Waddington, the youthful epigenetic landscape is eroded to the point where cells head towards other valleys, losing their identity in a process we have termed "exdifferentiation.""

On 2019-10-28 23:45:14, user Charles Warden wrote:

Hi,

I have a minor point about a citation:

"The only other paper we identified which performed a systematic benchmarking of pseudocounts is Warden et al [2], however they limited the range of their pseudocount to be between 0 and 1; and as we’ve seen the optimal value may be much larger."

The fold-change calculation is from FPKM values. While you can do something similar with Count-Per-Million (CPM) values, that would still not be exactly the same as a pseudocount. In other words, an FPKM of 1 probably is a much more conservative threshold than a pseudocount of 1 (if you are talking about normalized counts, whose exact value can vary depending upon other samples).

Also, I apologize, but I think accessing that paper has become tricker more recently. However, you can see all of the original content here:

http://cdwscience.blogspot....

Thank you for putting together this paper!

Sincerely,<br /> Charles

On 2019-10-28 21:45:54, user Tim F. wrote:

Very nice resource - great job putting this together! I wanted to comment on transformations of haptophytes, as there seems to be a few references/methods/species you have missed. Table 2 in "Velmurugan N, Deka D. 2018. Transformation techniques for metabolic engineering of diatoms and haptophytes: current state and prospects. Appl Microbiol Biotechnol 102:4255–4267. doi:10.1007/s00253-018-8925-5", covers the approaches not mentioned, including Prasad et al. 2014, Prasad 2017, and Gruber 2009. Worth including?

On 2019-10-27 15:39:50, user Kelly Twede Brewer wrote:

Where is Figure 4?

On 2019-10-27 03:00:04, user PRADEEP KUMAR wrote:

Is the measurement of interleukin and p.gingivalis is enough to establish the role in cardiovascular disease?<br /> I doubt it <br /> We need stronger evidence

On 2019-10-26 04:14:18, user K E Castro wrote:

Hi. Thanks for posting this paper. Overall, I believe it represents a complete and well-thought-out set of experiments that provide valuable insights into the CRG observed phenomenon, as well as into the functioning of the CDCP1 pathway within renal cells. It also offers a potential explanation and application into the mechanism of invasion and metastasis of certain cancer cells. This knowledge, as the paper suggests, may lead to new and innovative therapeutic targets.

Having said this, there are several items that may aid in improving the presentation of this scientific story. The following is a general list with some supporting references.

1. The headers in the results section are often incomplete or inconsistent.

Certain of the headers in the results section do not match up with what is shown in the text that follows or in the conclusions reached. For example, in the first of the six results sections, the paper states that CDCP1 is required for HGF signaling in MDCK cysts but does not mention the prominent role of Src phosphorylation (or lipid rafts) shown in many of the experiments. A header more consistent with the conclusions reached in the section that references Src and its co-location with CDCP1 in lipid rafts would be more accurate and informative. While this is brought up later, the foundation of where it is discovered is in the DRM analysis in this section and thus it might be prudent to include it in the header for the section. Alternatively, these inconsistent headers may be solved by splitting the section. For example, the first section could be divided into the identification of the role of Src and then a separate section created on the identification of the role of CDCP1 .

1. Discussion information is intermixed in the results section.

There are several examples of background information being interspersed in the results section. Some of this is helpful and needed for understanding of the results, but other items are duplicative or even distract from the information provided. For example, in the first results section the paper provides information on the MDCK model in order to justify its use which is duplicative of the information in the introduction, where it is more appropriate. Similarly, in the third results section there is the discussion of the indirect activation of MMPs (as related to STAT3) placed as additional proof of the involvement of STAT3. As there is already more direct evidence of the activation of STAT3 in the other experiments presented, this seems better placed in the EV or other support sections of the paper.

Often these sorts of diversions from the results can be found in narrative that cites the EV figures at the end. My suggestion is that if the figure is important enough to provide in the results section, then it should be in the main set of figures, and vice versa. The EV figures should be for background information only, as further support. This is not generally followed and adds some difficulty in getting the full impact of the paper. See for example, EV1B which is a good visual for Src-MERs effect on MDCK cells, a key factor in the story being told.

1. The figure legends are not informative of what is going on in the experiment,

The figure legends being separate from the figures and text make it difficult to follow along . This problem is exasperated by the legends generally providing limited descriptions of dosing or methods, rather than a clear explanation of what is in the figures. As one example, in EV3A it is unclear, without reading the text, how each of the mutant EGFP cells impact the experiment. The legend simply describes that these cells were incubated in the presence of Dox for the indicated time periods.

A more difficult problem is that several items in the figures are never explained even in the text, like for example the Myc-tags which show up in several places. An explanation in the legend would cure this issue.

As a result of the figures failing to explain themselves, it appears (and may be true) that some of the figures do not match the text. For example, in EV5F the text discusses the degradation of the ECM, but the figure shows the percentage of cyst protrusion as impacted by Marimastat, a general MMP inhibitor.

1. Lipid rafts are assumed, but not shown, to be the location of interaction after the first set of experiments.

In the third results section, the occurrence of CDCP1 activating STAT3 in lipid rafts is referenced, but none of the experiments in this section show such a locational affiliation. In view of the final conclusions later reach in the paper, it would seem beneficial, if not necessary, to offer experimental results of a colocation in lipid rafts as was done in the DRC experiments connecting Src and CDCP1.

1. The connection between the described pathway interaction with Myc and CyclinD1 is suggested and even used in experiments, but not explained in the paper.

A clearer explanation of the relationship between the described pathway, STAT3 specifically, and Myc and CyclinD1 should be provided if references to these proteins in the figures, results and discussion continue to be included. Someone who knows could connect the dots between these proteins and cell proliferation, but this connection is not discussed or tested in this paper. In any case, it would be an improvement and less confusing to take these references out of the third results section and Figure 3B, and put it with the support EV information.

1. References are not footnoted.

Cited references are not footnoted. A short version of the reference is included with the text and then listed in full by alphabetical order at the end of the paper. This makes it more difficult to find and follow the reference to the source. A footnote and even a linked connection between the information provided and the referenced citation would assist the reader greatly. <br /> <br /> Discussion of some Minor items:

1. ‘Occurred’ is misspelled in the last results section describing the immunofluorescence analysis of Met and STAT3 in wild type renal tubules.

2. The EV 2A figure legend is duplicated.

3. The textual explanation in the third part of the results section misstates which figure is which – Figure 3E shows proliferation and Figure 3F describes protrusion (contrary to text).

4. There are also a couple of figures that represent the complete pathway process that I think would be helpful at the start in the abstract rather than in the middle or end -- specifically 4G or 6H/EV6B 2A (repeated). They show the ultimate mechanistic flow being identified by the paper and thus would be helpful to readers in understanding what is being presented. Also, a figure like EV8A would seem to be related to 6B and C, and valuable to show the CDCP1 side of the pathway being described (which is missing in the other two figures). Finally, figures like 6G and EV8E do not seem to be necessary. One shows one type of MMP while the other shows another type doing the same thing (nothing new). Since these are not part of the main argument, either one or both could be deleted or moved to supporting materials without impact.

I hope this has been of some value.

On 2019-10-25 22:44:48, user Bryan Ivan Ruiz wrote:

The manuscript by Xu et.al proposes a post-transcriptional mechanism required to maintain mitochondrial dynamics modulated by Clock. Using both in-vitro and in-vivo experiments, the authors clearly show a novel role of the Clock in regulating mitochondrial morphology, specifically in the post-transcriptional regulation of Drp1. This post-transcriptional model is novel and has not been previously characterized in hepatocytes. The investigators have conducted a significant amount of biologically relevant experiments to test and validate to validate their conclusions. Through a combination of assays, immunofluorescence, rt-PCR and western blots, the data shows that Clock modulates Drp1 activity via mRNA degradation thus controlling mitochondrial fission. Further the authors provide a potential new roll for mitochondrial fission repressor Mdivi-1 through the evident mitigation of ROS production, the reduction of NAFLD and the recovery of the membrane potential of ClockΔ19 mitochondria. This is a novel and significant mechanism in regulating mitochondrial homeostasis in a circadian context. However, despite the amount of work that has been done in this manuscript, there are important issues that must be addressed. <br /> Minor Issues<br /> • In line 135-137 the authors state that the mitochondrial matrix of ClockΔ19 mice hepatocytes was deeply stained implying a pH change may observed in the mitochondria (Figure 1B). This is inaccurate observation as Figure 1A displays less stained mitochondria in the ClockΔ19 hepatocytes vs Wt hepatocytes. In order to remediate this pH levels should be quantified and more representative images should be utilized. <br /> • In Figure 3 the authors state that in ClockΔ19 mice display a smaller mitochondrial surface ranging approximately 0.1-0.3 μm2 However the bar plot displays the majority of ClockΔ19 mice have a slightly larger mitochondrial surface ranging from 0.3-0.5 μm2. The overall conclusion is still sound however changes in the manuscript must be made to report finding this accurately.<br /> Major Issues<br /> • In Figure 1E The authors use ATP6 as a marker to observe mitochondrial morphology. The use of ATP6 is supported since the authors show the relative protein concentrations of ATP6 are equivalent in both Wt and ClockΔ19 mice mitochondria (Fig 3C). However due to the differences in fluorescence intensity a change in mitochondrial morphology is difficult to support. In order to remedy this the authors could report two images of similar fluorescence intensity.

• Line 214 the authors claim mitochondrial fusion genes (Opa1, Mfn1 and Mfn2) display decreased mRNA expression in ClockΔ19 vs Wt. This is untrue as Mfn1 was shown to have higher mRNA expression in the ClockΔ19 sample (Figure 3D). The authors can consider in situ hybridization to verify their claims. <br /> • The authors further claim fission protein DRP1 expression was increased in ClockΔ19 mice (Figure 3E). However in the results reported this is untrue, when observing the blot both DRP1 and phospho-DRP1’s do not appear to show increased band density. This claim must be redacted or band density must be calculated and show significant difference. However, the claim that FIS1 is elevated in expression is supported.<br /> • In Figure 5E mtDNA proteins are used to demonstrate the inhibition of mitochondrial fission by Mdivi-1. This is an indirect way to demonstrate inhibition of mitochondrial fission. In order to further support this point the investigators can consider blotting for Fis1 in tandem with the mitochondrial proteins reported in Figure 5E.

On 2019-10-25 21:29:49, user Leobardo Corona wrote:

The research done by Hohmeier et al. was aimed at finding a method to enhance β-cell proliferation while retaining glucose-stimulated insulin secretion, both of which have limited methods available. Through testing, the lead compound GNF-9228, demonstrated a strong ability to stimulate β-cell proliferation selectively and insulin secretion, as well as protecting against cytotoxic stress. The authors did a great job at explaining their experimental approach and the flow of their experiments made sense, especially to someone with a limited background on the topic. The abstract highlights the goal and major findings well, but the title would be stronger if it were more specific by naming the molecule itself. The title also fails to ignore the effects on human islet cells. The discussion clearly states the main findings, issues, and future directions.

Type 1 and Type 2 diabetes involve the loss of β-cell mass, the mechanism of which is explained in the Discussion, and could be more ideal if implemented in the Introduction. This loss contributes to elements of β-cell failure: loss of β-cell proliferative capacity, low insulin secretion, and susceptibility to cytotoxic stress. Overexpression of the transcription factor Nkx6.1 stimulates β-cell replication and enhances glucose-stimulated insulin secretion (GSIS), this overexpression leads to a robust induction of the VGF gene. Due to this, a screening strategy involving a VGF reporter gene was used to identify three compounds that displayed a high capacity to stimulate human islet cell proliferation The paper mentions that these compounds were screened out of 41 total compounds that were studied on their effects on islet cell proliferation, GSIS, and protection against apoptotic<br /> stress but that data is not provided. Two out of the three compounds, GNF-9228<br /> and GNF-1346, were shown to increase Nkx6.1 mRNA and GNF-9228 was chosen to<br /> further study its activity.

GNF-9228 was shown to significantly increase human islet cell proliferation, while being specific to β-cells with a small effect on α-cells and no effect on σ-cells. Next, they saw that pre-treatment of GNF-9228 increased insulin secretion when compared to a DMSO control. The title of this figure states that GNF-9228 stimulates insulin secretion for both ran and human islets, but only the data for human islets is provided (Figure 4). Apart from<br /> increasing β-cell proliferation, GNF-9228 demonstrated the ability to protect<br /> 832/13 cells from both ER stress and cytokine-induced cytotoxicity. When<br /> compared to Dyrk1A inhibitor GNF-4877, GNF-9228 had higher VGF induction and no<br /> clear NFAT nuclear translocation, suggesting that GNF-9228 activates cell proliferation<br /> via a different pathway distinct from Dyrk1A inhibitors. Lastly, all three compounds<br /> displayed poor pharmacodynamic properties and future work will focus on<br /> developing a chemically modified version of GNF-9228 to enhance its bioavailability<br /> for in vivo testing as well as understanding its mechanism of action. This paper has made remarkable findings, but there are a few concerns I would like to see addressed.

Figure 1 seems a little out of place, it would be good to include the<br /> other two compounds to see their representative hits or explain why these two<br /> compounds were showcased. In Figure 2, GNF-9228 and GNF-1346 showed a higher<br /> induction of Nkx6.1 mRNA when compared to the control, but significance between<br /> the two treatments in rat cells and not just DMSO-treated cells would further<br /> support the statement that GNF-9228 is considered the most potent. In human<br /> cells, there is no significant trend on Nkx6.1 mRNA or expression of VGF and<br /> c-myc, but the paper still stated that GNF-9228 is the most potent when looking<br /> at the data, both GNF-9228 and GNF-1346 seem very similar. The statement that<br /> GNF-9228 is the most potent out of the three compounds is misleading since only<br /> two compounds were tested and data was shown. Even though GNF-4088 has a similar<br /> structure to GNF-9228, it would still be valid to see how it performs in comparison<br /> to the other two compounds to increase validity to the statement made in the<br /> paper. It would have also been great to see the effects GNF-1346 had on the<br /> other experiments apart from Figure 6.

As previously stated, GNF-1346 was used in Figure 6 to compare activation<br /> of NFAT nuclear translocation between GNF-1346, GNF-9228 and two Dyrk1A<br /> inhibitors (GNF-4877 and harmine). Both GNF-9228 and GNF-1346 had a low percent<br /> of NFAT nuclear translocation, with GNF-1346 having less at the highest<br /> concentration. In 6b, both also exhibit a similar pattern of NFATc<br /> localization, these two findings suggest they have a similarity and would be<br /> interesting to see if they had a similar activity of the VGF-luc reporter in<br /> 6c. There is a small mislabeling in 6b where the legend mentions 10µM of the<br /> different treatments were used but the representative images show different<br /> concentrations, it is difficult to determine which concentration is the correct<br /> one as it is not mentioned in the Results section either.

On 2019-10-25 19:13:36, user 4gab wrote:

The manuscript by De Belly et al. sought to expand upon how changes in cellular morphology and forces decide the fate of embryonic stem cells. In their paper, the authors chose to probe this question through the use of embryonic stem cells which change from a round to a spread morphology as they further differentiate. When these stem cells are changed to media without inhibitors, such as GSK3 and MEK inhibitors, they adopt the aforementioned spread morphology. The cells also exhibit decreased Nanog, a marker of naive pluripotency, and upregulated Otx2 transcription factor. After observing increased membrane blebbing prior to embryonic stem cell exit from naive pluripotency, the authors decided to probe cell membrane tension as a possible mechanism for exit due to low membrane cortex attachment. Their trap assay experiment demonstrates that cells which have exited a naive pluripotency state possess significantly lower membrane tension.

The decreased membrane tension observed during exit of the naive pluripotent state prompted the authors to look at Ezrin- Radixin- Moesin (ERM), a linker between the plasma membrane and the actomyosin cortex, which regulates membrane tension. They find a decrease in phosphorylated active ERM (pERM) during exit of naive pluripotent state further supporting their hypothesis of lowered membrane tension mediating further cell differentiation. Furthermore, the evidence presented in this manuscript supports the activation of GSK3 and the subsequent degradation of Beta-catenin as possible mechanisms for the decrease in pERM during exit of naive pluripotency. The authors bolstered their support of endocytosis as a major player in the exit cell state through fluid uptake assays showing increased endocytosis once the cells adopted a spread morphology. An increase in Rab-5a rescued defects in early differentiation through the induction of Rab- 5a mediated endocytosis. The FGF/ERK pathway was indicated by FGF receptor internalization and pERK live imaging to be the signaling pathway regulating membrane tension during further cell differentiation. Ultimately, De Belly et al. conclude that first embryonic stem cells experience a decrease in membrane tension, then an increase their rate of endocytosis, and the activation of ERK leads to the exit of these cells from naive pluripotency. This manuscript makes a significant contribution to our field. It also possesses a strong logical flow of experiments and corresponding analysis. However, a few results presented here prompt some unanswered questions.

First, in Figure 1H the level of pERM does decrease dramatically after 2i+L media removal. The observed decrease in pERM indicates a change in differentiation state believed to be due to low membrane tension. Even so, there seems to be a significant increase in pERM from T4 to T24. It might be prudent to address such an increase in pERM and speculate as to why the increase might be occurring over the time course of the experiment. The quantification of this difference may additionally help to explain whether such an increase in pERM between these timepoints is significant or not. Also, this figure does not include data for T48. T48 is indicated as the end point of the transition out of naive pluripotency in Figure 1A. Even though data is shown in the Supplemental Figure 3A as to pERM and GAPDH at T48, this data does not aid the reader in the comparison of the levels of pERM to the levels of pERM at other time points using the same controls. The T48 time point is also used throughout the manuscript for the readout of various experiments. The inclusion of the pERM levels at T48 in Figure 1H might shed some light on the phosphorylation state of pERM at the end of the exit from the naive pluripotent state and add to the continuity of this figure.

Secondly, Figures 2A and B aimed to demonstrate how a constitutively active form of Ezrin (iEZR_CA) led to embryonic stem cells maintaining their high membrane tension. Therefore, this mutant would allow the authors to test the expression of various differentiation state markers and perform phenotype rescue assays. Figure A shows pictures at T48 in control and iEZR_CA mutants along with their corresponding quantification in Figure B. The interpretation of the data states that even T48 maintained high membrane tension similar to ES cells. However, the data presented in Figure B does not show data for T48 and only shows data for T24. Figure A also does not present any images from T24. One way the authors could clarify their claim is to show pictures in Figure A for the phenotype of wild type and iEZR_CA at T24. Also, in Figure B if the authors showed data for trap force in the wild type at T48 and iEZR_CA at T48 they would create a more logical connection between Figures A and B. The author’s argument that ES cells maintained high membrane tension even at T48 would be strengthened through the inclusion of this additional data.

Throughout the manuscript De Belly et al. utilize Nanog and Otx2 as markers of differentiation state in embryonic stem cells. Some figures show data for both Nanog and Otx2 or reference supplemental figures. Examples include Figures 1C, 2C and D, and Supplemental Figure 4D. A majority of experiments use Nanog as the sole marker of naive pluripotency while only some utilize Otx2 as a marker of exit from naive pluripotency. This occurs particularly in later experiments in the manuscript where the authors do not look at Otx2 and only assay for Nanog. Figures 2G- I connect to Supplemental Figure 5a and these figures only probe for Nanog and not Otx2. Figures 3E, I, and J along with their Supplemental Figures 6i-J also only probe for Nanog. Due to the fact that they defined the exit state as low Nanog and high Otx2 expression, inclusion of both markers in each assay where the experiment probes for Nanog would support their interpretation of the differentiation states and more closely align with their description of these markers.

On 2019-10-15 22:29:00, user Alfonso Martinez Arias wrote:

Much of interest in this work. I agree that while Transcription Factors are important for pluripotency, it is worth exploring the role that cell structure, adhesion and the cytoskeleron play in the establishment of this state. May this observation be related to the requirement for membrane bound ß-catenin for pluripotency (Faunes et al. Development 2013 https://dev.biologists.org/...<br /> and to the findings of Hrs and Mapksp1/Mp1 in a screen for controllers of the exit from pluripotency (Betschinger et al.Cell 2013 https://linkinghub.elsevier...

Also, if it is cell shape that matters, have the authors try to plate their naïve cells in E-Cadherin coated plates? As cells in 2i have very high levels of E-Cadherin, when plated in this conditions will spread and wonder what will happen to pluripotency.

Very good work and direction of enquiry