On 2021-02-28 08:26:57, user Christian Monsivais wrote:

My colleagues and I recently chose to discuss this paper in one of our journal club sessions because we were interested in your work on creating a DNA vaccine, for the SARS-CoV-2 virus that currently plagues our world. I found that this paper was able to properly establish VIU-1005’s ability to stimulate a robust immune response, help prevent infection of host cells, and provide longer coverage when delivered intramuscularly. That being said, here are some recommendations to make this paper become more strong and clear. One thing that would have been interesting to see would have been the addition of a different virus, such as a MERS virus, as a sort of control in Fig. 4. I believe that this would have allowed for the reader to affirmingly be convinced that there is a significant efficiency in producing neutralizing antibodies between the VIU-1005 immunized models and that of other commonly known viruses. Another suggestion includes adding an experiment similar to that of Fig10, where instead of just showing the data of the typical needle system, a separate mice group should be immunized using the needle-free system only as well. This last suggestion would definitely strengthen the story being told throughout the paper about ways to make this vaccine more efficient, specifically on the delivery method.

On 2021-02-27 19:34:10, user Tao Zhang wrote:

At first, thank you very much to the authors and co-authors for this study. It is good to see that more and more studies about the peopling of East Asian (and Eastern-Eurasia generally) are being published.

However I have serious concerns regarding some arguments and results in this study. This includes the hypothetical clade "Andamanese/Hoabinhians with ancient Tibetans and Jomon period Japanese". As I can see in the preprint, this clade is largely based on the findings of McColl et al. 2018, which proposed a partial shared ancestry between the Jomon period people (samplified by one sinlge sample IK002 "Ikawazu Jomon") and Hoabinhians/Onge as well as the unusual link to haplogroup D (D1/D-M174), taking into account that this haplogroup has an estimated age of 60,000 years, with the Japanese lineage D1a2/D-M55 53,000 years. McColl et al. 2018 estimated the IK002 sample having about 44% Onge/Hoabinhian like ancestry and 56% East Asian like ancestry (identical to the study resutls here). However more recent studies such as Gakuhari et al. 2019 (published in 2020) rejected the conclusions of McColl et al. 2018 and did not find any evidence for a shared ancestry between the Ikawazu Jomon sample IK002 and Onge/Hoabinhians. However, they found evidence for geneflow between Jomon period people with coastal East Asians in Northeast and Southeast Asia, with up to 41% geneflow in the Ami and Atayal ancestors (Native Taiwanese). (See:https://www.nature.com/arti...

Further studies, such as Boer et al. 2020 similarly rejected McColl et al. 2018 findings and concluded that there is no shared ancestry between the Ikawazu Jomon (IK002) and two other Jomon samples from Hokkaido and Honshu (such as Sanganji Jomon sample), with "basal Asians" such as Hoabinhians or Tianyuan. (See:https://www.cambridge.org/c...

Citing Boer et al. 2020 :

The Jōmon do not share a special relationship with Hòabìnhians as previously suggested (McColl et al., 2018). Tests of genetic similarity do not show Hòabìnhians or the Jōmon sharing exceptionally high genetic similarity with each other.

Another study by Yang et al. 2020 expanded on the findings of Boer et al. 2020, similarly rejecting a close genetic relationship between Jomon and Onge/Hoabinhians. See: (https://science.sciencemag....

The question is now, why did the authors ignore these new findings and took the "early stage" data from McColl et al. 2018 for granted? In the paper it does not seem that these data got tested for accuracy?

This problem and possible error is affecting the data results of the whole study. Did the authors take into accound aditional migrations and admixture currently not detected? Why the Andamanese get counted as unmixed proxy? There were several hints that the Andamanese recived admixture from the mainland (several times). Citing Chaubey et al. 2015, the Andamanese (samplified by the Onge) have 32% East Asian like ancestry. Citing Mondal et al. 2017, the Onge paternal lineage D1a3 (D1a2b) splitted from a Tibeto-Burmese tribe of Northeast India (Riang/Tripuri) about 7,000 years ago. How did this clade come into the Andamanese/Onge? Why do the Great Andamanese have different paternal lineages such as P, K, M and S? D is common in East Asian Highlanders and Japanese (suggested to be a Jomon lineage, but still debated) and strangely in Andamanes. However other Hoabinhian like people lack haplogroup D but have K, F, P, M or S, found in other Negrito and Melanesion populations of Insular Southeast Asia and Oceania.

The conclusions made by the authors to link haplogroup D to the Andamanese/Hoabinhian clade seems rather unlikely. The diversity of D clades (to be specific D1 clades) point to a northern route and dispersal (i.e. Altai route and following dispersal into the regions of East Asia). See this map as example:(https://commons.wikimedia.o...

Haplogroup D1 (D-M174) has an age of about 60,000 years and originated in Central Asia and diverged between Tibet and the Altai mountains (Hammer et al.). The Japanese subclade D1a2 or D-M55 splitted from other D1 clades about 53,000 years ago (Mondal et al. 2017).

Citing Mondal et al. 2017:

This strongly suggests that haplogroup D does not indicate a separate ancestry for Andamanese populations. Rather, haplogroup D was part of the standing variation carried by the OOA expansion, and later lost from most of the populations except in Andaman and partially in Japan and Tibet.

How do the authors explain their suggestion linking haplogroup D to a hypothetical clade uniting all populations which carry haplogroup D? Do they suggest that Aboriginal Australians and Mongolians also form one clade based on haplogroup C? Are Europeans with haplogroup R (P1) closer to Papuans with haplogroup M/S than pre-Indo-European Europeans with lineages such as I or J? I thought it was already generally accepted that genetic ancestry should not be based solely on haplogroups?

Boer et al. 2020 found evidence for geneflow of an divergent East Asian population into both the Jomon and less into ancient Southeast Asians. This geneflow can be better linked to the spread of haplogroup D (D1 and its subclades). Did the authors take into account geneflow into the Andamanese? Taking them as a 100% proxy seems rather unusual and unlikely. Especially after the findings of Chaubey et al. 2015 with about 32% East Asian admixture (samplified by Han) in the Andamanese (samplified by Onge).

It would be great if the authors can comment on this point. I have written this rather fast and hope that a reviewer will take a deeper look onto this and may help to improve this paper.

Nevertheless, thank you and good luck!

On 2021-02-27 16:18:30, user G Jeffrey Elian wrote:

This is an interesting study that requires a lot more research. Not being peer reviewed throws up a big red flag. From a non academic prospective, if metformin is so widely use and until Covid, lifespans have been increasing, it probably can't be that harmful. I know. I know there are many, many other factors that make this possible....just saying.

On 2021-02-27 15:32:44, user Matt wrote:

Highly original, ambitious and very interesting paper. I'd like to point to a couple of other recent preprints on Biorxiv which may be relevant and put some questions I have:

• The paper "A curated dataset of modern and ancient high-coverage shotgun human genomes" 10/2020 (https://www.biorxiv.org/con... provides some very high coverage ancient Native American and Siberian samples - AHUR_2064(10962 BP), USR1 (11425 BP), Kolyma1 (9775 BP), Anzick-1 (12712 BP), Sumidouro5 (10403 BP).

Those could provide an excellent set to extend the ancient-modern comparison on fragmentation of Neanderthal sequence? As these samples largely a clade with living Native American people and have somewhat comparable dating to the Stuttgart and Loschbour samples. There would be no confound of any other factor - differential long term population size (expected to break up fragments?) or admixture or selection against Neanderthal variants - would affect results.

• Another paper sharing an author with the above, "Inferring population histories for ancient genomes using genome-wide genealogies" 02/2021 (https://www.biorxiv.org/con.... This paper again covers a large number of high-coverage ancient whole genome sequences. Part of the paper discusses excess TCC>TTC mutations.

This paper makes a persuasive argument that the TCC>TTC signal is heavily correlated to late Upper Paleolithic Anatolian Hunter-Gatherer ancestry, and lacks a relationship with generalized West Eurasian ancestry (e.g. it is absent or very weakly detectable in an early Iranian Neolithic agriculturalist and a European mesolithic hunter gatherer).

Further they identify that "the driver of this mutation rate change, which may have been of genetic or environmental nature,was already extinct by the Mesolithic" (e.g TCC/TTC prop is stable and constant post 10kybp), and is not seen in Sunghir Upper Paleolithic individual dated 30KYBP and is present more weakly with isolation-by-distance from Anatolia in Late Upper Paleolithic and Mesolithic samples. (Ref Figure 6e)

The authors state: "We note that while the cause of this mutation rate elevation remains uncertain, our results would fit well with a genetic cause within a specific ancient population (for example a mutation in some repair protein, transiently present). If, alternatively, the cause is environmental, it appears highly localised in both time and place, and this seems potentially harder to explain."

This identification of a localised, intense signal, is very compatible with Harris and Pritchard's inference that - https://elifesciences.org/a... - "We used the enrichment of TCC→TTC mutations as a function of allele frequency to estimate when this mutation pulse was active. Assuming a simple piecewise-constant model, we infer that the rate of TCC→TTC mutations increased dramatically ~15,000 years ago and decreased again ~2000 years ago."

How much are these indications (direct and indirect) of a pulse compatible with your paper's proposal that this mutational signal (and C>T enrichment generally in West Eurasians) is associated with reduced parental age over time?

Explaining a reduction in generation length by 5 years (30->25) over a period of 40k as giving rise to the enriched C>T mutation pattern is a very plausible idea, but if this is happening within a "highly localised" 10k, then assuming linearity a more intense, shorter reduction in generation length of 4x as much (i.e. 20 years? so 30->...?).

Some general notes or suggestions: it might also be helpful to readers of the paper to quantify the contribution of this mutation pattern to the overall accumulation of derived alleles. (The data is there in Table S6). There is also a possible typo in Table S6 where C>T’ is given as T>C'. Having a more similar amount of whitespace in Figure 3 would be useful to understand similar the relationships are across mutational spectra.

On 2021-02-27 08:08:51, user Shyam Bhakta wrote:

Very useful design rules deduced!<br /> Minor point: RNase, not RNAse, since no "acid" in ribonuclease.

On 2021-02-27 03:36:52, user Manwa Li wrote:

Japan has developed a Antiviral Coating Agent Prestige Shield ( http://www.darg.com.hk/page... ) which contain copper nanoparticles in the market mainly on contruction site and hotels. Mr. Miyamoto start studing Copper against influenza virus in 1998.

Let's fight against COVID-19 together! Please contact me email: manwa@darg.com.hk

On 2021-02-27 01:33:37, user Lorenzo Calviello wrote:

Nice paper and idea!<br /> Did you try using random forest regression (using feature importance as a predictor for gene-set relevance) as an alternative to regularized regression?

On 2021-02-26 21:35:36, user Aamir Rana wrote:

What about IVM precipitation? IVM is water insoluble. When you dilute it from 1% IVM injectable solution to PBS, it may precipitate. So, what about the bioavailability of IVM?

On 2021-02-26 10:39:37, user Simon wrote:

Interesting work. Just letting you know that JSON is JavaScript Object Notation and not JavaScript Open Notation as stated in the manuscript. I also believe that the term "JSON pulse" should be rephrased, I think I get what you mean by it but it is confusing.

On 2021-02-26 09:35:40, user Guilherme Valente wrote:

Hello, fascinating research!<br /> Since you have many points along growth curve, what about model your systems by ODEs on network? Then you could assess how edges and nodes are shifting along the growth, conditions and cells.

On 2021-02-26 08:28:41, user Sharon wrote:

Really nice information. amino acids is actully a big source of protein. I also like this information about bcaa-<br /> https://www.ajinomoto.com/a...

On 2021-02-26 04:57:25, user Takahiro Anzai wrote:

There is an error in the amino acid sequence of the MFSD13A_HUMAN in Fig.S6b.<br /> For example, here it is 135-GMTYTIMDLHHH-146, but it is actually 135-DGFLTLVDLHHH-146.<br /> Please check the sequence with UniProtKB (https://www.uniprot.org/uni....

On 2021-02-26 01:38:45, user Nina Gilshteyn wrote:

I really enjoyed how your department’s visualized your iBAQ scores similarly to how RNA seq data is visualized (heatmaps, UMAP, PCA etc…); however, your department should repeat these experiments in order to generate scRNAseq data that is comparable to the scProteinseq data acquired by your novel methodology. The scProtein seq experiments need to be repeated and conducted at the same time as scRNAseq data acquisition in order to mitigate batch effects from acquiring the data at different time points. Perhaps you could acquire funding for such future experiments by citing this work. The following recommendations are for this paper. The conclusion of the celltrails analysis was difficult to trust but I think that measuring how the states change between the cells and over pseudotime is a very interesting thing. I think it would be easier to trust how data from different cells under different experimental conditions can be represented by proteome changes over time for 1 abstract cell. I think it'd be better to put the cell trails analysis in the supplements if there isn't enough space to elucidate the value behind the jargon. Furthermore, scRNAseq clustering by dimension reduction is prima facie inaccurate so scProtein seq by the same method is likely inaccurate. Why did you use the best practice, umap, for figure 4 and PCA for figure 5? I think its better to address the distortion before someone else does or before someone believes that this math is an accurate representation of the real. A better way to salvage this study would be compare scRNA Seq data analysis to scProteinseq analysis in order to bring more information about how dimension reduction distorts data. I think the average Jaccard distance might be a good metric for comparing scproteinseq data to scRNA data ( see reference). I also think it is important to be consistent with the dimensionality reduction methodology unless stated otherwise or else it looks like you fitted the data to support your hypothesis. All in all very good work with the improvements in mass spec data and I think this may be very useful if truth and understanding were made a priority over publication and title.

Cooley et. al, “ A novel metric reveals previously unrecognized distortion in dimensionality reduction of scRNA-Seq data.” bioRxiv 2019 https://doi.org/10.1101/689851

On 2021-02-24 23:21:11, user Ziyan Wu wrote:

I really appreciated this paper and I think it did a nice job in showing the accuracy and precision of the TIFF method. At the same time, I think a couple of edits could be made to help the clarity.

For example, figure 2(c) includes the proteins that had more than two valid values. These proteins, shown blue and red in figure 2(b), are shown red in figure 2(c). Therefore, changing the color scheme may be helpful to prevent misunderstanding.

Also, I really like the Venn Diagram, while can be a little confusing when considering what the “98%” in the overlapped part means.

Besides, figure 3(b) includes some data from previous research. And I noticed the standard method used in that paper is LC-MS. Since the one for the other figure is MS-MS, I think it will be nice to label LC-MS directly in figure 3(b).

In the end, personally, I hope there can be a distinct color scheme for figure 3(d), which shows the different pair-wise coefficients in different colors so that the comparison can be easier.

In general, my colleagues and I believe your research is really important, and that is also why we chose it for our journal club. Overall, the work was very exciting and intriguing to read. I hope this feedback helps in strengthening your paper.

On 2021-02-24 06:19:18, user AHMAD KASSEM wrote:

My colleagues and I have recently chosen your paper for our journal club. This paper was excellent to illustrate the use of Mass spectrometry in biomedical research. I think the new TIFF method is promising and could lead to a breakthrough in the realm of single-cell proteomics. Your group has done an excellent job in the overall presentation of data; however, I have a couple of suggestions that might strengthen this paper. Firstly, fig 3. (d) presents the pair-wise correlation of the ten cells using the TIFF method, but it has no control. I suggest you could show the pair-wise correlation for the ten cells analyzed with the standard method. Secondly, all the bar-graphs and violin plots except for fig.5 (d) were not analyzed for statistical significance. The lack of statistical significance analysis made it trickier to interpret the data, so I would suggest adding them wherever possible. Finally, your group mentioned the importance of applying both sc proteomics and sc-RNA sequencing to get the optimal results. I believe that including other methods such as RNA-seq or even traditional biochemical methods (i.e., Western blot or Flow cytometry) to provide further evidence to the results obtained by the TIFF method could dramatically increase the creditability of your results.

On 2021-02-26 00:58:02, user Jimmy Walker wrote:

So did they ever revise this article? Funny how it makes total sense and yet no one wants to acknowledge it or at least try to dispute the theory.

On 2021-02-25 21:48:59, user Tim Triche, Jr. wrote:

Some of the product numbers for the ADTs in supplement 1 are duplicated (specifically, GYPA/CD235a is Biolegend product number 349117, and CXCR4/CD184 is Biolegend product number 306531).

Incredible resource and lovely paper.

On 2021-02-25 19:44:23, user Cluster 7 wrote:

Interesting findings. The preprint is missing supplemental figures.

On 2021-02-25 07:00:20, user Daniel P Faith wrote:

The claim is that <br /> “new uses were encountered at an increased rate in more phylogenetically dispersed communities (Table 1). More phylogenetically diverse tree communities therefore tend to represent a greater number of ecosystem services than tree communities of equal richness but lower phylogenetic diversity.”<br /> But note that the measure of phylogenetic dispersion, PSV, does not correspond to phylogenetic diversity (PD), so one cannot conclude that PD captures more uses. <br /> by the way, PSV can go up when a species goes extinct.<br /> Other - <br /> “Nature’s contributions to people”(NCP) is not identical to “ecosystem services” (for background, see e.g.<br /> Faith, D. P. (2018) Avoiding paradigm drifts in IPBES: reconciling “nature’s contributions to people,” biodiversity, and ecosystem services. Ecology and Society 23(2):40<br /> Your paper lists benefits of PD but leaves out the most fundamental – biodiversity option value, defined by IPBES (Diaz et al. 2015, p. 14) as “the ‘option values of biodiversity’, that is, the value of maintaining living variation in order to provide possible future uses and benefits.”<br /> Indeed, that Forest et al paper was a test of PD for biodiversity option value<br /> And note biodiversity option value is the basis for the existing IPBES link between PD and one of the most critical NCPs “maintenance of options”; the same PD indicator is now proposed by CBD to monitor this NCP and biodiversity option value within the post-2020 framework<br /> Also – the Mazel et al arguments are already well-countered e.g. in Owen, Nisha R., Gumbs, Rikki, Gray, Claudia L., Faith, Daniel P. (2019) Global conservation of phylogenetic diversity captures more than just functional diversity. Nature Communications<br /> Dan Faith

On 2021-02-25 04:32:39, user Miles Davenport wrote:

For a study using mixed effects to model SARS-CoV-2 immunity see:

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

On 2021-02-25 04:29:39, user Paul Wolf wrote:

I thought the E484K mutation was the signature of the variants from South Africa and Brazil, and am concerned if it's arisen independently in New York. Remember NY was hit hard by the coronavirus early on. The variants with E484K seem to appear later, after a lot of people have acquired immunity to the original (wild) strain. Bloom Labs just published more research today, testing antibodies against the different variants, including Brazilian and South African,(and Californian, although that's because of a different mutation), which all escape the same antibodies.

On 2021-02-25 04:11:17, user Nikolas Haass wrote:

Now published: Biophys J. 2021 Feb 19:S0006-3495(21)00154-5. doi: 10.1016/j.bpj.2021.02.017. Online ahead of print.

On 2021-02-25 03:55:18, user Colin Dunstan wrote:

This is an interesting and insightful manuscript advancing the technologies available for in vitro developmental patterning of human stem cells with promise for 3D fabrication of human tissues accomplished via rational design of architecture, mechanical properties and composition of a 3D printed matrix scaffold.

On 2021-02-25 01:31:11, user Paul Wolf wrote:

If the variants from California, South Africa and Brazil escape the same antibodies, do they pose the same threats of reinfection and to vaccine efficiency? Are they more or less the same? Will these types of variants dominate over B117, or occupy a different niche?

On 2021-02-25 00:06:18, user Dave Daversa wrote:

An update: this preprint has been published and a link will be forthcoming

On 2021-02-24 08:52:56, user Vágási Csongor I. wrote:

This study has been published in Proceedings of the Royal Society B<br /> https://royalsocietypublish...

On 2021-02-24 03:46:12, user Tim Flegel wrote:

This is a very interesting paper with information that is worth following up. However, I wonder whether the authors recorded information regarding the growth and survival of the shrimp in the test and control groups to determine whether there were any differences. Growth and survival, together with feed conversion ratio (FCR) are the issues of most concern to farmers, and I expect that the reviewers of this manuscript may ask that at least growth and survival information be added to this manuscript for publication in an aquaculture journal. It doesn't matter whether or not there were any differences.

On 2021-02-23 23:06:07, user Joshua Mylne wrote:

This is now published at RSC Advances under a different title "Improved herbicide discovery using physico-chemical rules refined by antimalarial library screening" https://doi.org/10.1039/D1R...

On 2021-02-23 22:41:28, user Wendi wrote:

When are you planning to submit your paper for peer review and publication? Why has this not been done yet, but you are actively registering "long haulers" for treatment?

On 2021-02-23 20:47:09, user Guillaume Mas wrote:

Stimulating work with Skp and SurA that do confirm current hypothesis that chaperones are able to expand their client proteins even in the absence of source of energy such as ATP.

The experiments at different Skp concentrations (Fig. 2) are particularly interesting as you have chosen to work at near physiological concentrations. In our recent Skp publications, we have shown that in this concentrations range Skp populates an equilibrium between an inactive disordered monomeric state and the canonical active folded trimeric state (PMID: 33087350). This equilibrium being shifted toward Skp3 in presence of OmpX/tOmpA.

Would integrating the fact that Skp exist as a mix of inactive/active species explain why only ~80% of uOmpX molecules were in a Skp3–uOmpX complexed state even at the highest concentration of Skp (2.5 μM)?

Guillaume

On 2021-02-23 20:39:23, user Michael J. wrote:

Hey nice paper, thanks for sharing. Side note: I thought Kraken2 shouldn't be used for relative abundance calculations.

https://github.com/DerrickW...<br /> https://microbe.net/2017/04...<br /> https://github.com/merenlab...

On 2021-02-23 18:55:05, user Charles Warden wrote:

Hi,

Thank you very much for posting this pre-print.

Am I correctly understanding is that the main goal is to help parse information from the existing database?

For example, I am correct in understanding that functions to test/compare application of various scores to yourself is not included?

This is essentially what I have done here, and I am curious if your package can help do something similar (for other scores):

http://cdwscience.blogspot....

Thank You,<br /> Charles

On 2021-02-23 16:29:26, user disqus_xiJ0WI39mp wrote:

Very interesting study! I checked for the antibody used to detect LOXHD1 but I could not find it in the mat and met section of the ms. Can you disclose?<br /> Thanks!

On 2021-02-23 16:02:02, user Mark Saper wrote:

Thank you for a novel description of putative multi-barrel OMBBs. Regarding GfcD discussed on page 10, you mention "a middle globular peptidoglycan-binding domain," How did you determine this? Did you get Psi-blast hits with putative PG-binding proteins? I would be very interested in this info and/or perhaps you want to reference this observation in your final paper. - MAS, saper@umich.edu

On 2021-02-23 15:13:18, user Jon Fisher wrote:

This is an interesting analysis, but I see two aspects that would make it hard for it to get traction. First, as noted in the abstract, poor enforcement is a major problem now, and rapidly expanding PAs into areas suitable for agriculture would make enforcement much harder (even if more resources were available). Second, it appears that this centers around the need of species to keep almost all of their range, but doesn't seem to consider the needs of people to grow enough food and have livelihoods. Starting solely with an ecosystem need without considering capacity for enforcement nor opportunity cost to human well-being makes this pretty non-convincing to anyone who isn't already strongly supportive of conservation as a top priority.

On 2021-02-23 15:04:29, user Team Thomma wrote:

In a revision, this manuscript (BIORXIV/2018/249565: Nuclear and mitochondrial genomes of the hybrid fungal plant pathogen Verticillium longisporum display a mosaic structure) has been combined with manuscript BIORXIV/2018/341636 (Homogenization of sub-genome secretome gene expression patterns in the allodiploid fungus Verticillium longisporum), and updated with additional analyses as well as a novel genome sequence. Based on these additional data and analyses, our previous conclusion that the mitochondrial genome displays a mosaic structure was no longer retained. The revised manuscript can be found as an update to BIORXIV/2018/249565, entitled "The interspecific fungal hybrid Verticillium longisporum displays sub-genome-specific gene expression" (doi: https://doi.org/10.1101/341636.

On 2021-02-23 12:20:45, user Karam Yacoub wrote:

Unveiling the unique role of Smad4 in preventing chronic intestinal inflammation. It was a pleasure to contribute to this n…

On 2021-02-23 10:34:14, user marodon wrote:

The figures are of poor resolution even in the pdf. Please correct. Thanks

On 2021-02-23 10:05:01, user Marcel Tarbier wrote:

Very exciting work! Congrats to all authors! Particularly enjoyed figure 4, certainly got inspired for my own data analysis!

We discussed it in a journal club and were wondering how it performs in comparison to other methods. The approach appears to be very similar to RAID (Gerlach et al. 2019). Would be great to discuss the pro's and con's in comparison.

We also collaborated with a lab from Uppsala University that had developed a cool technique to profile RNA and protein in the same single cells called SPARC (Reimegård, et al. 2019 bioRxiv). Maybe also worth checking out. ;)

On 2021-02-23 03:48:25, user Rajeev wrote:

Very interesting results. A leap of hope for TNBC patients.

On 2021-02-23 03:09:33, user BJRao wrote:

What is the functional consequence of the so called SPAD association with NS?

On 2021-02-22 13:05:13, user Natal van Riel wrote:

Possibly also of relevance: Yuan et al (2016) Flux Balance Analysis of Plant Metabolism: The Effect of Biomass Composition and Model Structure on Model Predictions, https://doi.org/10.3389/fpl...

On 2021-02-22 12:34:36, user chhaminderkaur wrote:

I presented this manuscript during a lab meeting and enjoyed reading it as<br /> well as presenting it. The manuscript studies the modulation of protein<br /> expressed from the arginine transporter gene of Toxoplasma gondii based on the amount of arginine available in the media. The basic hypothesis has been demonstrated through enough data backed by Western blot analysis performed at different arginine<br /> concentrations and time intervals. Furthermore, the data shows a direct and<br /> clear evidence of uORF-mediated translational regulation of the arginine<br /> transporter gene by mutational analysis. During the discussions, we had a few<br /> questions and comments, which are given below.<br /> 1.Figure 1B: For bradyzoite induction at high pH, the methods section clearly states<br /> that the parasites were maintained in a media with high pH for 6 days. However,<br /> it is not clear for how many days the parasites have been starved of arginine<br /> before the experiment is performed. This data point will be crucial in<br /> eliminating the hypothesis that the bradyzoite induction did not happen due to<br /> less time given for them to differentiate.<br /> 2.Previous studies show that arginine starvation leads to the conversion of the<br /> fast-replicating tachyzoites to the slow-growing bradyzoites (Fox et al., 2004).<br /> This would mean that the absence of the arginine transporter, TgApiAT1 would<br /> lead to the formation of bradyzoites. However, the authors show that the<br /> presence of another cationic amino acid transporter, TgApiAT6-1 can also<br /> transport arginine with low affinity. This shows that the parasite has a<br /> fail-safe mechanism in case TgApiAT1 is not expressed due to other unknown<br /> reasons. The presence of a fool-proof mechanism suggests that arginine<br /> concentration plays a vital role in deciding the stage of the parasite.<br /> 3.Gene IDs of TgApiAT1 and TgApiAT6 are not mentioned.<br /> 4.In Plasmodium falciparum, PfAAT1 (PF3D7_0629500)<br /> is the homologous arginine transporter. The transcript of this amino acid<br /> transporter harbours 11 uORFs in its leader sequence. Our analysis shows that these<br /> uORFs encode for peptides that do not share any similarity with the conserved<br /> yeast arginine attenuator peptide (AAP). It is possible that P. falciparum might be employing another mechanism to regulate the expression of this gene.

On 2021-02-21 21:47:08, user Tim Vines wrote:

This study has an issue because the authors can't tell when preprints have already been through peer review at a journal: these are obviously going to be similar to the published version, and the comparison can't tell us about the reliability of manuscripts prior to journal peer review. Their conclusion that pre-prints are almost as reliable as published articles is therefore not all that reliable.

Fortunately, this is relatively easy to address, by e.g. excluding preprints that were posted within 100 days of the publication date, and by finding preprints that acknowledge reviewers. Excluding preprints that only post one version would be another helpful step. Lastly, the authors should plot the level of text differences against time between the preprint appearing and the publication date – one would predict that the greatest text differences would appear for the preprints that took longest to get published.

On 2021-02-21 17:15:59, user Roger Benson wrote:

Ok, thanks for the reply. This is a sample of the text (from your abstract) that gave me the impression that you were discussing crocodylomorphs rather than just crocodilians: “Crocodilians and their allies have survived several mass extinction events. However, the impact of the K-Pg mass extinction event on crocodylomorphs is considered as minor or non-existent”. ‘Crocodylomorphs’ is also used in the parts of the Introduction that I was referring to. Anyway, consider changing this during review, or whatever to reflect that you intended to refer to Crocodylia. The implication is quite different.

On 2021-02-21 11:57:08, user monika sharma wrote:

This new Innovative app will be helpful to researches. I hope more new researchers will utilize such great platform.

On 2021-02-19 13:48:13, user robin khosla wrote:

Good Work. Very Informative app to survey the target reserach groups for initial stage researchers.

On 2021-02-21 11:07:13, user simakali wrote:

Very interesting work that I have unfortunately only just discovered. Many congratulations on a great study identifying a transcriptional role for APOBEC2. We reported a transcriptional function for APOBEC3B some years ago, not yet replicated by others. I wonder if there are similarities in modes of action of these two APOBECs?

On 2021-02-21 10:44:49, user Stuart Cook wrote:

Very nice manuscript. For the link to IL11, I think the references below would be much stronger as they relate to cardiac fibrosis/fibroblasts whereas the current reference (16) is specific to lung. Happy to help with IL11-related follow on experiments, if useful.

Schafer et al. 2017. “IL-11 Is a Crucial Determinant of Cardiovascular Fibrosis.” Nature 552 (7683): 110–15.

Lim, Wei-Wen, et al 2021. “Antibody-Mediated Neutralization of IL11 Signalling Reduces ERK Activation and Cardiac Fibrosis in a Mouse Model of Severe Pressure Overload.” Clinical and Experimental Pharmacology & Physiology, January. https://doi.org/10.1111/144....

Corden, Ben, et al. 2020. “Therapeutic Targeting of Interleukin-11 Signalling Reduces Pressure Overload-Induced Cardiac Fibrosis in Mice.” Journal of Cardiovascular Translational Research, June. https://doi.org/10.1007/s12....

On 2021-02-21 07:35:49, user Zhang Liang wrote:

--model {Turner,Zuker,ZukerS,ZukerL,ZukerC,Mix,MixC}<br /> Folding model ('Turner', 'Zuker', 'ZukerS', 'ZukerL', 'ZukerC', 'Mix', 'MixC')

May I know where I can find details about this parameter? Thanks.

On 2021-02-20 20:03:48, user Paul A. I wrote:

Hi everybody,<br /> I'm just a nursing student with limited knowledge about biochemistry but reading this article makes me think about one possibility. Someone which get the first vaccine dose and get infected with the virus after 1-2 weeks, when the quantity of antibodies is not enough to fight efficiently the virus, could this lead to natural selection of a mutation resistent to the antibodies?

On 2021-02-20 19:59:03, user Ekaterina Shelest wrote:

Some more remarks.The second one is the most important!

1. It is not accurate to say that FunOrder is the firsttool based solely on genomic data: “first program giving a prediction about core genes in fungal BGCs based solely on genomic data.” CASSIS is purely genomic based, as is in fact antiSMASH, depending on what you call “genomic data”. <br /> Moreover, strictly speaking, FunOrder is NOT genomic-based. You do not use any genomic information. You use pre-selected protein sequences for blasting and then run some phylogenetic analysis.

2. I just noticed an interesting mistake, which probably has led to many misunderstandings. It seems that you call all genes that are not involved directly in the biosynthesis, like TFs and transporters, “gap genes”. This is a huge mistake. The words “gap genes” are indeed in use, but they mean a different thing. They mean those genes that are completely unneeded for the production of the SM and essentially do not belong to the cluster, albeit they “sit” between cluster genes. In my previous comments, every time I used the words “gap genes” I meant exactly this: the genes that do not functionally belong to the cluster; they are usually not co-expressed with it. This does not refer to genes like TFs, transporters, tailoring enzymes, etc., because they are essential for the cluster function. No product will be produced without them. To illustrate, all genes marked with blue in Fig 1 are NOT gap genes; they are legitimate cluster members. <br /> I think this mistake clarifies a lot. You should understand that the genes you considered as dispensable and “non-essential” are same necessary for the cluster functioning as those that are directly involved in the synthesis. Regarding the cluster evolution, they can be<br /> expected to co-evolve with the other cluster genes with the same success.

On 2021-02-20 19:31:38, user Ekaterina Shelest wrote:

Further major concerns.

The FunOrder is positioned as a tool for “automated identification of essential genes in a BGC”; (for people who deal with BGCs, this means all cluster genes, because usually clusters are compact and spare genes are rare). But the input is already a set of BGC genes, so, first of all, the clusters are not really identified. We can only speak about some refined annotation. Given that the emphasis is made on biosynthetic genes and not all BGC genes, it is only partly refined. This makes all the statements about the importance of better cluster annotation, provided in the introduction, obsolete. Secondly, where the input BGC genes come from? In case of a new genome, will this be a set of genes in some vicinity of the PKSs and NRPSs (if yes – in which?)? Or a result of preliminary BGC annotation with antiSMASH and/or CASSIS? This should be specified. For known genomes and BGCs, again, what is the source of the BGC information? MIBiG, antiSMASH, other databases, literature? Where the examples used in this study were taken? Table 2 provides MIBiG IDs but not for all clusters; where the others come from?

MATERIAL AND METHODS <br /> FunOrder - Workflow

1. Practically the only part of the tool that deals with evolutionary questions is treeKO. This is fine. But it is not clear to me, if the “speciation history” is shown by the authors of treeKO as less significant in detection of co-evolution, why do you consider it at all? What’s the point of a combined measure that includes something that is less trustable and informative (“speciation history”, in this case)? The examples are not convincing; if you want to use a measure, you should show it’s useful.

2. I did not understand what was the point of making a curated proteome database. In which sense is it curated? Did you filter something out? If yes, what, on which principles? Is it just a collection of 134 proteomes from JGI and NCBI? Could you please explain the principle on which they were selected? One can blast against all ascomycetes in JGI and get many more hits for the query genes. Why limiting yourselves to just 134? Many of which are of the same genera? If the reason is just to rename the sequences assigning a species identifier, this can be done with any genome/proteome with a simple script, no need to keep the proteomes in a special database.

Performance evaluation.

Hmm… I was puzzled by the effort of manual comparison of 102 control BGCs, each with at least 3 genes. Did I understand it correctly, was it literally manual? Why did you do that? (Was it a practical assignment to a class of students?) I had a feeling that this manual assessment was then used as a gold standard to set up a threshold for the tool. But why? Why not simply select parameters of treeKO, which would allow to re-identify the true positive BGC genes? Eventually, this is what was done, setting up the treeKO parameters;<br /> I don’t understand the sense of the manual evaluation step.

Measures of the performance.

Here we come to an interesting part. <br /> The worries start with this: “we calculated three measures (two measures for the positive control BGCs and one for the negative control BGCs)”. In general, positive and negative controls are treated identically. Otherwise, they are not controls. Or did you mean something different?

Speaking about the proposed measures themselves, they are confusing. To start with, TP, TN, FP, FN are already defined with clear definitions and there is no need to re-define them. What you measure in your experiment and put in a confusion matrix ARE already TP, FP, and so on. A phrase like “obtained values for FCGM and ERM were classified as true positives (TP) or false negatives (FN), and the values for NCV were classified as true negative (TN) or false positives (FP).” is bewildering. You cannot classify ERM or ECGM or anything based on them into TP, FN, etc., because you use the real (measured) TP, FN, FP to calculate ERM, ECGM, and NCV! It seems that you are going in circles.

Probably you haven’t noticed that your notations “a”, “b”, “c”, correspond to FN, FP, P. The “number of genes necessary for the biosynthesis of a SM, that did not cluster with the other necessary genes in the FunOrder analysis” to me translates into “genes that we expected to be there but haven’t found”, which is a typical FN. So, your “a” from equation 1 is the FN. Moreover, your FCGM is not a new measure but just the sensitivity, or true positive rate (TPR), or recall, this is evident if you use standard notations:

a=FN; c=P; c-a=P-FN=TP; => (c-a)/c=TP/P=TPR.

What’s the point of inventing new notations?<br /> ERM is nothing else than accuracy: <br /> By definition ACC=(TP+TN)/(P+N)<br /> ERM=1-(a+b)/d; A=FN; b=FP (if there were no other genes that should not belong to the cluster); d=P+N; =><br /> ERM=1-(FN+FP)/(P+N)=(P+N-FN-FP)/(P+N)=(TP+TN)/(P+N)=ACC

I must also point out that the way how the equations are written is… a bit strange. It’s some brackets obsession there. There is no need for brackets in expression like 1-a/c, the division goes before subtraction anyway. Same for a/d+b/d; moreover, you are allowed to sum up the fractions. The scary expression for NCV looks actually like this:<br /> 1-g/2d(d-1)

No need for three classes of brackets, especially between the factors of the multiplication.

Regarding the NCV, I did not fully understand what is meant by g. It is defined as a “number of … distances in all matrices” but this does not make sense. Is it the number of genes of the considered cluster on strict and combined distances at selected thresholds, in other words, genes that fulfil the condition to be considered as clustered? If yes, then this is just TP. If no, what is it, then? It’s also not clear, why 2d(d-1)? In general, could you please explain how this NCV measure was defined, derived and why?

Results and discussion: <br /> “In our experience, evaluating only the numerical values is not enough for a thorough analysis of a BGC and it is necessary to consider all provided visualisations for a thorough data interpretation“ – Usually visualisations are used for illustration or as supportive material. The idea of computational tools is to switch from human interpretations, which may be biased, to something more systematic, isn’t it? There are ways to extract the results of cluster analyses and operate with numbers.<br /> By the way, the Fig. 3 legend is mixed up.

Performance evaluation <br /> As I think that all metrics are calculated incorrectly, further discussion of the results is senseless. But if the metrics were correct, they could be hardly considered as good. <br /> This is not surprising because, as I said, we shouldn’t expect that all genes in the clusters are co-evolving.

More comments to come!

On 2021-02-19 12:41:50, user Ekaterina Shelest wrote:

The are two major concerns in regard to the aims and the main idea of this work, and they are interconnected. <br /> The first is the concept of co-evolution of the BGC genes. I agree that the genes belonging to the same biosynthetic pathway will most likely co-evolve; the question is, do we expect that ALL of them will co-evolve? We can see that the cluster regions are mostly syntenic (not always, by the way), but we can also see that genes appear and disappear from the syntenic clusters. So why should we expect that a BGC co-evolves as a whole? Some new genes can be recruited to the cluster much later than it has been formed; in this way, the organism can start producing a new substance. You just add, e. g., a tailoring enzyme and get something new. However, the authors take for granted that all genes in BGCs are cemented together forever and from the beginning of time. <br /> In addition to possible recruitment of biosynthetic genes we have a cohort of non-biosynthetic ones that have a good chance not to co-evolve but be just recruited. TFs, transporters, and other auxiliary genes are more likely (than some enzymes) to be recruited to an existing cluster not having shared evolutionary history with it. They are more numerous in genomes and “at hand” to be taken into a cluster. However, they are crucial for the production of the SM even though they don’t synthesise anything. To start with, a cluster will not be even expressed without its TF (if it happens to have one). <br /> So I would expect that some genes co-evolve, the other don’t. This, by the way, is illustrated by high number of false negatives in your analyses.

In any case, if there is some major idea of an approach or a phenomenon, on which this approach is based, and if it is not self-evident, it must be justified, explained, and referenced. There are zero references to articles about BGC genes co-evolution; not too surprising as I haven’t found any papers in PubMed as well. This means, that you have first to substantiate your idea. In the way it is written now, the article causes a major question that is not answered, discussed or anyhow addressed. I dare to suggest that the paper would gain a lot if it were not about the tool per se but about application of this<br /> method to investigation of cluster gene co-evolution.

The second serious problem for me is what the authors actually understand as a cluster and what they aim to identify with their tool. A BGC contains not only biosynthetic genes. Transcription factors, transporters, other auxiliary genes are also indispensable parts of a cluster. However, in the discussion of the “exemplary analysis” of Lov, the authors with a light heart tell us that LovE does not cluster with necessary genes in PCA for any distance measure but it’s fine because it’s just a TF and does not participate in the biosynthesis.<br /> I am left with the impression that the authors are happy not to find a TF as a part of the detected cluster. So what is a BGC, in this case? What do the authors plan to find? Only the genes immediately involved in the synthesis? But this is NOT the definition of a BGC. If the tool is indeed oriented on identification of only biosynthetic genes, excluding all other BGC genes, this must be explicitly stated and discussed. In this case, although it is an interesting idea, I doubt that it will be very useful for BGC annotation and especially for practical applications, as in many cases it’s crucial to know the cluster-related TF (e.g., to overexpress it). It will also not help refining cluster borders and excluding gap genes, which are two aims formulated in the beginning of the paper. For the borders, it’s obvious: if you don’t aim at finding all cluster genes, you risk losing flanking ones. For the gap genes, why are you so sure they will not co-evolve together with the cluster? As a<br /> part of landscape? <br /> This second concern is mainly caused by the way how the authors serve their ideas. If the Introduction did not make such a great emphasis on the necessity of better detection of the BGCs and in particular their cluster borders, and on the practical purposes of the BGC identification (support of the lab work, etc.), I wouldn’t be left with the impression that their major aim is to detect the whole clusters. As it turns out, the aim is to detect only the co-evolved genes; in fact, this is said directly in the title: “Identification of essential biosynthetic genes” but the introduction changes the expectations, and the way how the whole paper is written does not add clarity to this. As I’ve mentioned above, if cleverly applied, the tool is a great way to investigate the cluster evolution. The paper will gain a lot if you change the angle under which you describe it.

I have many more remarks and concerns. I’ll add them in the next comment.

On 2021-02-20 09:31:33, user Rudolf A Roemer wrote:

A very nice paper by Stephen using flexibility analysis. Sad to see it's not yet published. Have a look at our paper https://rdcu.be/cfvcQ which applies flexibility analysis to nearly 300 PDB structures of SARS-CoV-2 related proteins, including the protease studied here and, of course, the spike protein structures.

On 2021-02-20 05:27:42, user Chaitanya wrote:

Id love to read the preprint, but I have a fundamental question. Is it possible to use phylogenetic tree approaches on sequences for their base content, with no reference to the information content? Do you use a sliding window?

On 2021-02-20 02:01:48, user Hateful BlackMan wrote:

Very interesting study. I would love to see one performed on Bronze Age Nubians.(Kerma)

On 2021-02-18 19:33:07, user Asten Bryant wrote:

I would like to see an analysis that shows how much North African specific ancestry these people had.

On 2021-02-19 18:55:54, user Johanna N. wrote:

Great work!<br /> It would be interesting if you could test some tri- or tetra-ortho substituted PCB and see whether they are more potent, as suggested in my publication (Nyffeler et al., 2018).<br /> Moreover, do you know whether the cells you were working with are derived from neural crest cells? I believe the peripheral nervous system as well as part of the ear are derived from neural crest cells.

On 2021-02-19 17:58:21, user Fraser Lab wrote:

The main protease (Mpro) from SARS-CoV-2 is an attractive drug target for antiviral development. Although a substantial amount of work has focussed on developing inhibitors of Mpro, there are currently no Mpro inhibitors approved for clinical use. One strategy to accelerate the development of an antiviral drug focuses on repurposing existing protease inhibitors for use against SARS-CoV-2. If the potency or selectivity of an existing drug is insufficient, then the structure of Mpro with the drug of interest may help to inspire/guide modifications to achieve the required potency and selectivity. This manuscript by Kneller et al. investigates the structural basis for binding of telaprevir to SARS-CoV-2 Mpro. Telaprevir belongs to the α-ketoamide class of protease inhibitors, and was approved for treating a subset of hepatitis C infections (telaprevir was discontinued in 2015 due to the emergence of other more effective therapies with fewer side effects). This class of inhibitor achieves potency in part through the formation of a reversible covalent bond with the catalytic serine/cysteine, a reaction that involves the shuffling of protons in the Cys-His catalytic dyad of Mpro. The aim of this manuscript was to map the protonation states of the Mpro-telaprevir complex using neutron crystallography, and to understand the importance of changes in protonation state to inhibitor binding. The authors have previously published the structure of apo Mpro at room temperature using neutron crystallography (https://www.jbc.org/article... and X-ray crystallography (https://www.nature.com/arti..., and the structure of the Mpro-telaprevir complex at room temperature using X-ray crystallography (https://www.cell.com/struct....

The major strength of this paper is the use of neutron crystallography to investigate the change in protonation that occurs upon binding of the covalent inhibitor telaprevir. Growing crystals for neutron diffraction can be a formidable challenge and the structure presented in this work is an impressive achievement. The mapping of protons in the Mpro-telaprevir complex shows that the catalytic histidine (His41) remains protonated after telaprevir binding, while two other histidine residues in the active site (His163 and His164) swap protonation states.

The major weakness of this paper is a failure to adequately address the second part of the aim: namely, what is the relevance of the protonation states observed by neutron crystallography to the binding of covalent Mpro inhibitors? How will this information be used to modify existing protease inhibitors to increase potency/selectivity against Mpro, or to guide the design of new Mpro inhibitors?

Below are two major points, and several minor points, that should be addressed before this manuscript is finalized.

Major point 1<br /> The authors chose telaprevir because it allowed them to grow crystals of suitable size for neutron studies. They have also previously reported in vitro inhibition data for this compound. The manuscript would be improved by adding a more frank assessment of telaprevir’s potential and some more background information. E.g. is telaprevir active in SARS-CoV-2 viral assays (it appears not https://pubs.rsc.org/en/con... Is telaprevir being evaluated in clinical trials for treating SARS-CoV-2 infection? Importantly, the authors should inform readers that the drug was withdrawn from the market in 2015 (https://www.ncbi.nlm.nih.go....

Major point 2<br /> The authors claim that the structural information reported in this manuscript will guide precise tailoring of Mpro inhibitors (e.g. P2 L29). The manuscript would be improvised if the authors articulated the modifications that they would make to telaprevir (or other α-ketoamides) to increase potency. This discussion should include a description of how these modifications are related to the structural information obtained by neutron diffraction and how that would compare to the assumptions based on (even simple) computational calculations of the protonation states from the X-ray structures alone. In light of this, they should consider a bit more nuance to the claim that the “Detailed structures of the ligand- free and ligand-bound drug targets are essential to steer drug design efforts in the right direction” (P5 L21).

A good start would be to articulate the relative contribution of the hydrogen bonds mapped by neutron diffraction to the stabilization of the Mpro-telaprevir complex compared to the interactions made by the rest of telaprevir. E.g. the hydrogen bond between the hemithioketal hydroxyl and His41 is said to be weak - P9 L9 - what modifications should be made to strengthen this bond, and is it plausible that this would have an impact on binding affinity?

Minor points<br /> P2 L16: “Near physiological temperature” seems imprecise. Why not just say the temperature (e.g. 25°C)?

P4 L5: “previous studies” - the two studies referenced are on Mpro from SARS-CoV, not SARS-CoV-2 - please include an appropriate reference.

P4 L16 and P7 L3: Reconsider using the term “catalytic water” for the water that hydrogen bonds to the catalytic histidine. To me, a “catalytic water” would be a water nucleophile involved in hydrolysis, not a buried water that stabilizes a catalytic residue.

P4 L18-26: Please complete the description of the Mpro catalysed hydrolysis of a peptide bond e.g. what is predicted to happen after formation of the first tetrahedral intermediate? Why are α-ketoamide inhibitors such as telaprevir resistant to hydrolysis (e.g. no suitable leaving group)?

Figure 1C: This panel could be improved by showing (and annotating) the catalytic cysteine.

P5 L3: “...neutron diffraction data can be collected at near-physiological (room) temperature avoiding possible artifacts produced during cryo-cooling and cryo-protection of protein crystals necessary for the mitigation of radiation damage in X-ray cryo-crystallography” The authors have previously reported structures of Mpro from X-ray data collected at 100 K and at room temperature. Were there artifacts in these structures caused by radiation damage and/or cryo-cooling? See this recent paper for a full discussion of these issues on this protein: https://journals.iucr.org/m...

P5 L24-28: Perhaps a reference or two might be helpful here.

P6 L9: “Remarkably, the net electric charge of +1 in the active site cavity is conserved between the two structures.” Why is this remarkable or surprising? Please include an explanation.

P6 L33: “ Hence, one might suggest that the hemithioketal-His41 pair could maintain charge separation, keeping the positive charge on His41 and having the deprotonated negatively charged hydroxyl” - is this reasonable? Can you give an estimate for the pKa of the hemithioketal hydroxyl? The author’s previous paper on the Mpro-telaprevir structure shows the hydroxyl protonated (https://www.ncbi.nlm.nih.go... - Fig. 3B). Why was this protonation state assumed in the previous work, but raised as a point of contention here?

P6 L38: Have the authors calculated Fo-Fc maps with the deuterium modeled on the His41 NE2 rather than the hemithioketal oxygen? Do these maps support their proposed protonation pattern?

P6 L51: Please include a reference for “may not be strong”.

Figure 2A and 3A: Please label the catalytic cysteine.

Figure 2A: Would it be more appropriate to show a bias-free electron density map here rather than the 2Fo-Fc map? Either a map calculated prior to ligand placement, or a map calculated using a bias-removing method (e.g. phenix.polder).

P7 L3-L16: It would be easier to interpret this paragraph if the points were illustrated in a figure. How confident are the authors that a 0.2 Å difference is real, given the resolution of the structures?

P7 L39: Was the high mobility modelled by decreasing the occupancy or increasing the B-factor? Or both?

P8 L13: Consider revising the claim that shortening a hydrogen bond leads to “considerable strengthening” of the bond. This is imprecise (e.g. https://pubs.acs.org/doi/pd...

P9 L11: “The hemithioketal hydroxyl makes a short, but probably weak, hydrogen bond with the neutral His41 in Mpro-Telaprevir that would be required to enhance inhibitor binding affinity.” I’m not sure I understand this sentence. Consider rephrasing.

P9 L16-36: This section would be easier to follow if the relevant figures were referenced throughout.

P13 L42: What program was used for integration?

P14 L28: “a script was run” - it would be helpful if this script was made publicly available.

Galen Correy and James Fraser (UCSF)

On 2021-02-19 15:20:57, user Samuel Furse wrote:

This manuscript has now been peer-reviewed and accepted for publication at Communications biology. Link to follow!

On 2021-02-19 11:16:17, user sofiakarkampouna wrote:

This study is now published in Nature Communications, available below https://www.nature.com/arti...

On 2021-02-19 11:09:42, user Jibby N Frantz wrote:

The lead measles-vectored COVID-19 vaccine candidate (MV-ATU2-SARS-CoV-2) described in this paper was not introduced into clinical trial. Our data emphasize that a strong and stable expression of the spike antigen from an early promoter of measles virus genome is crucial for high immunogenicity and protection.

On 2021-02-19 09:26:47, user upfeffer wrote:

Hi,<br /> I saw you cited our work on a highly invasive but yet not more aggressively metastases forming subpopulation of MDA-Mb-231 cells. When we did the work we supposed there could be mechanisms of coordinated or collective migration where the more invasive population increases the chance of less invasive but more metastatic populations to reach their destination. Very interesting, so. I would be happy to share our MDA-MB-231 populations with you if you intend further research in this direction.

On 2021-02-19 09:18:00, user Stéphane Ory wrote:

Hi, nice work. I was wondering why you looked at Dyn2 and not Dyn1 which is expected to be the major form of Dynamin in neurons?

On 2021-02-19 00:21:50, user Peter Hickey wrote:

FYI the 'Technical Overview mapper' table is slightly truncated at the bottom.<br /> Edit: It looks like this has also affected all the figures in the paper - all seem to be truncated at the bottom

On 2021-02-18 21:31:48, user Dr. Tyeen C Taylor wrote:

This article is in review at Frontiers in Forests and Global Change

On 2021-02-18 17:46:15, user Morita Lab wrote:

In this study, Corelle took a labor-intensive approach to visually screen hundreds of fluorescence microscope images and analyzing membrane domain-associated proteins biochemically. We welcome your comments on our new manuscript!

On 2021-02-18 16:44:30, user Gurjot Singh Sidhu wrote:

Very nice article! But I am unable to find supplementary material. Can someone help? Thanks.

On 2021-02-18 15:18:44, user Aldert Zomer wrote:

Some people ran into issues with installing and/or running RFPlasmid. We have made installation of the software easier by including a Conda option and we have clarified several issues and improved the manual. We're performing a user survey to gather some information about usability and if there are more issues with RFPlasmid or its dependencies. We're hoping we can improve user experience using the feedback we receive.<br /> The link to the survey is here:<br /> https://docs.google.com/for...

On 2021-02-18 10:54:32, user Dom San wrote:

this has been published?

On 2021-02-18 07:49:06, user Florian Privé wrote:

Please note that this work is now included as a Supplementary Note in https://doi.org/10.1101/202....

On 2021-02-18 07:07:29, user Jeroen Zegers wrote:

Link to published article in Journal of Neural Engineering:<br /> https://iopscience.iop.org/...

On 2021-02-18 06:24:22, user Ruoying Zheng wrote:

I totally enjoyed reading this paper. The experiment designs are great. It would be better if some parts of the article can be improved. In Fig 5, the biological model should be always mice instead of using human erythrocytes in 5A and using infected mice in 5C. Having a consistent biological model can rule out unnecessary variables. If both mice and human biological models are used here, then a vitro experiment about mice erythrocytes malarial infection and related treatments should be added here. In 5C, infected erythrocytes should have a higher FITC-A fluorescent binding properties than uninfected erythrocytes. More detailed explanation should be added in 5C to clarify the difference between each group. In the linear graphs of Fig 6A and 6C, different groups should have different colors, which would make it easier to read. As for the cell pictures in Fig 6A and 6B, the color scheme is not unified, the colors of the parasites should be the same in each picture so it would be less confusing. Besides, more background information about DHA(active metabolite of all artemisinin compounds) and why the author set up experiments to test efficacy of DHA+Alisporia should be explained.

On 2021-02-17 22:44:45, user Rebecca Altshuler wrote:

I really appreciated this paper. I think the research is very promising and could provide a potential solution to the issue of drug-resistant malaria strains. However, I think a couple edits could be made to enhance clarity and effectiveness. I think that the reasoning behind using DHA specifically in combination with Alisporivir could have been made more explicit in the background section. In Figure 4C, it was unclear to me why CsA was not used in addition to Alisporivir when conducting the growth inhibition assay for the artemisinin resistant strain R539T. I think this data would enhance the paper by illustrating how effective Alisporivir was in comparison to CsA in decreasing parasitemia. It could also be made clearer in figure 4C whether there were no parasites observed in the schizont stage or whether this stage was not accounted for. Some more minor edits might also improve clarity of the figure, like including the name of each strain somewhere in A-C, color coding the IC50 plots, and potentially restructuring the pie chart data into another format. One idea might be to plot the relative amount of ring, trophozoite, schizont, and pyknotic body stage parasites separately in bar charts to show numerically how the relative percent of parasites in each strain changed in response to the drugs. My colleagues and I chose this paper for our journal club because we felt that the research was important, and could potentially improve treatment options for many patients. Your efforts are much appreciated!

On 2021-02-17 17:45:07, user chikheang Soeng wrote:

Hello,<br /> My colleagues and I recently chose to present your paper in a journal club. We think that Artemisinin-resistant malaria is a major health threat and that repurposing Alisporivir as an anti-malaria drug, as demonstrated in this article, is a promising solution. <br /> I would like to share some of the comments that were brought up during our discussion. In Figure 1 and Figure 2, the interaction between Alisporivir and Cyclosporin A was demonstrated using computer simulations. However, we believe that this conclusion could be better supported by conducting an in-vitro protein binding assay. Also, in Figure 2C, the colors of the graph and the figure legends do not match, making it difficult to interpret the results. Since the main focus of this article concerns the effectiveness of Alisporivir against Artemisinin-resistant malaria, it might be a good idea to move them to the supplemental figure section. In Figure 3C, we think that the third column was mislabeled; it should be DAPI + Cyp. Otherwise, the quality of the microscopy images was excellent. When reading the methods section, the sample size could not be found and we hope to see it included there.<br /> Overall, I think that this paper is very interesting to read. I like the fact that the result section was broken down into smaller sections which makes it easy to follow. I am looking forward to reading more in the future.

On 2021-02-18 03:20:23, user Chris Bystroff wrote:

Great work! We did a journal club on this paper. This is the smoking gun for the connection of COVID and MIS-C.

On 2021-02-18 02:18:24, user zed li wrote:

In the abstract, the sentence "...and highlight ares and methods that...". Is that letter "a" missing? or what is "ares" meaning?

Best,

On 2021-02-18 02:17:20, user Tom Kelly wrote:

This work is now published. Lareau, C.A., Duarte, F.M., Chew, J.G. et al. Droplet-based combinatorial indexing for massive-scale single-cell chromatin accessibility. Nat Biotechnol 37, 916–924 (2019). https://doi.org/10.1038/s41...

On 2021-02-18 01:58:33, user Simon Castillo wrote:

Might be good to see this https://onlinelibrary.wiley...

On 2021-02-18 00:17:56, user Gernot Kaber wrote:

Nice paper! The Addgene reference numbers in Table 1 do not seem to exist in the Addgene database. Is that in the works? GK

On 2021-02-17 16:01:34, user stefan pulst wrote:

great paper on loss of fly atxn2

On 2021-02-17 15:35:10, user Kas Steuten wrote:

This article is now published in ACS Infectious Diseases: https://doi.org/10.1021/acs...

On 2021-02-17 12:09:41, user Scott Freeman wrote:

Thanks for confirmation of what Ingo Fricke & I postulated in September: https://osf.io/msu98/

On 2021-02-17 05:39:41, user Lucius Wang wrote:

This manuscript has been published in Nature Communications and can be assessed through the following link: https://www.nature.com/arti...

On 2021-02-16 19:00:29, user Andor Kiss wrote:

Welp! Going to give this one a go - very happy to see CUDA functionality builtin from the start! #GPUs #Assembler #Genome

On 2021-02-16 03:55:18, user TDNA wrote:

Very nice work! Demonstrates the importance of a second confirming allele! https://www.ncbi.nlm.nih.go...

On 2021-02-15 21:15:36, user Doraiswamy Ramesh wrote:

Thanks for your efforts towards the highly selective inhibitor, ALG-097111. Are you at liberty to disclose its Str. This will greatly facilitate a huge growth towards a potential drug. Thanks

On 2021-02-15 10:54:04, user Dr. Christos Chinopoulos wrote:

The findings support the notion that when cI function is not required, mitochondrial substrate-level phosphorylation (mSLP, PMID: 20207940) supported by NAD+ for KGDHC yielding succinyl-CoA (PMID: 23475850), requiring oxidized quinones for diaphorases for which cIII and cIV functions are critical (PMID: 24391134), is all-in all important for tumor survival (PMID: 30909720).

On 2021-02-15 08:41:09, user Shyam Bhakta wrote:

Given that the novelty here centers on SpoT and RelA, some mention is necessary of the stringent response and relaxed phenotype (namesake of relA). Common lab strains are mutants of spoT and especially relA for a reason, with DH10B being a mutant of both.

On 2021-02-15 08:14:56, user Blaz Stres wrote:

The much needed link between Biobakery tools and GTDB - MAGs (metagenome assembled genomes (https://doi.org/10.1093/mol... information and NCBI genome taxonomy is finally here. We were really looking for more streamlined and reproducible updates to HUMAnN3 database.

On 2021-02-14 22:15:21, user Anthony Leung wrote:

Please note that part of the work is now published in PNAS https://www.pnas.org/conten...

On 2021-02-14 11:40:22, user Bill Jaune wrote:

I am sorry for the scoop up of your excellent paper. You have provided solid data and talent ideas. However, I think you need to figure out which cell is expressing RANKL, since the Adipoq-cre labels both pre- and mature BMA.

On 2021-02-13 22:26:04, user Xiaoqi Feng wrote:

If you like this story, you may also like another one, about a completely different DNA methylation reprogramming event in the male germline of Arabidopsis. We show nurse-cell derived sRNAs transcribed from transposons target genes with imperfectly matching sequences in the germline, regulate meiosis, silence transposons, and determine paternal inheritance via specifying the sperm DNA methylome.<br /> https://www.biorxiv.org/con...<br /> Nurse cell-derived small RNAs define paternal epigenetic inheritance in Arabidopsis

On 2021-02-13 15:50:14, user Patrick Villiger wrote:

and now it can be found here..

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

On 2021-02-13 01:44:58, user Elizabeth Kwan wrote:

Nifty work! Our lab has previously looked at rDNA copy number in ~30 of the Schacherer Lab's 1,011 yeast strains using CHEF gels, ddPCR, and resequencing comparisons. It might be fun to compare our estimates with those from your sequencing methods.

See PMID: 31757929

On 2021-02-13 01:17:32, user Nicole Daigle wrote:

Article has been accepted, see DOI: https://doi.org/10.1016/j.a...

On 2021-02-12 20:07:15, user Mario Malicki wrote:

No authors are listed for this preprint.

On 2021-02-12 19:38:43, user Lia Maglietta wrote:

Dear authors,

I am writing you about your paper entitle “Revisiting animal photo-identification using deep metric learning and network analysis” publishen on bioRxiv.

My name is Rosalia Maglietta, and I have authored the paper “Reno ́, V., Dimauro, G., Labate, G., Stella, E., Fanizza, C., Cipriano, G., Carlucci, R. & Maglietta, R. (2019) A sift-based software system for the photo-identification of the risso’s dolphin. Ecological informatics, 50, 95–101” that you cited in your paper.

First of all, I really appreciated the contribution of your paper, as well as the increasing interest of the scientific community in the development and application of machine learning methodologies in animal photo-identification.

I would like to bring to your attention some points, on which I hope we can develop a constructive debate on this topic.

1. In your paper you wrote "A classical CNN classifier can re-identify already known individuals (usually with a softmax last layer) but will fail to identify new individuals. Indeed, the number of predicted classes must match the number of known individuals. Therefore, we crucially need a CNN-based approach that can filter out individuals unknown at the time of the analysis [Retrieving the known and unknown individuals consists in relying on the Euclidean distance computed for any pair of images.]"

Interestingly, we have already faced with a similar problem in the paper “R. Maglietta et al. Convolutional Neural Networks for Risso’s dolphins identification, IEEE Access DOI:10.1109/ACCESS.2020.2990427”. The main novelty of this paper is the development of a new method based on deep learning, called Neural Network Pool (NNPool), and specifically devoted to the photo-identification of Risso's dolphins. This new method includes the unique function of recognizing unknown vs known dolphins in large datasets with no interaction by the user.

1. In your paper you wrote "we applied state-of-the art techniques for object detection with CNNs (Lin et al., 2017) to automatically crop giraffe flanks of about 4,000 raw photographs shot in the field."

One of our previous paper (“Renò, V.; Losapio, G.; Forenza, F.; Politi, T.; Stella, E.; Fanizza, C.; Hartman, K.; Carlucci, R.; Dimauro, G.; Maglietta, R. Combined Color Semantics and Deep Learning for the Automatic Detection of Dolphin Dorsal Fins Electronics 2020, 9, 758”) approaches the problem of automatically cropping cetaceans images with a hybrid technique based on domain analysis and deep learning. Domain knowledge is applied for proposing relevant regions with the aim of highlighting the dorsal fins, then a binary classification of fin vs. no-fin is performed by a convolutional neural network.

We assume that these two papers we have published could provide you with some useful suggestions to efficiently solve your tasks, and maybe it could be interesting to compare the different strategies on different data sets (giraffe and dolphin images).

Lastly, in the paper you wrote:

“In a seminal publication, Bolger et al. (2012) first presented computer-aided photo-identification, initially for giraffes but more recently applied for dolphins (Reno ́ et al., 2019). The underlying computer technique is a feature matching algorithm, the Scale Invariant Feature Transform operator (SIFT; Lowe (2004)), where each image is associated to the k-nearest best matches. The current use of SIFT for ecologists requires human intervention to validate the proposed candidate images within a graphical interface (Bolger et al., 2011).“

We want to highlight that our algorithm, named SPIR, devoted to the automated photo-ID of Risso’s dolphins is correctly based on SIFT features and it is completely automated. In fact, it does not require any human intervention. We care about this particular skill of SPIR: it is able to automatically analyze huge amount of data, in relative short time (depending on the used processor), independently from the user. You can find more details about SPIR in the paper “R. Maglietta et al., DolFin: an innovative digital platform for studying Risso’s dolphins in the Northern Ionian Sea (North-eastern Central Mediterranean), Scientific Reports, DOI:10.1038/s41598-018-35492-3”.

I hope that the information contained in this letter could be of interest for you. I am available to discuss it with you and to consider collaborating.

Thank you for your attention.

Best regards,<br /> Rosalia Maglietta, PhD<br /> Guest Editor of the Special Issue "Statistics and Machine Learning in Marine Biology"<br /> https://www.mdpi.com/journa...

On 2021-02-12 15:50:30, user AnnotSV wrote:

Dear authors,

As we are developing AnnotSV, we have discovered the preprint of your new method for SV prioritization and we take this opportunity to make a few comments. As many others, we clearly share your view that using deep learning will refine the SV characterisation and help to take medical decisions.

However, just to let you know, since mid-December 2020 the updated ACMG guidelines for SV ranking (Riggs et al 2020) have been implemented in AnnotSV. As a result, the ACMG class assigned by AnnotSV (and therefore the associated ranking score) is now more accurate (in v3.0 rather than in v2.3) and your prediction performance comparaison is out of date.

Thank you for any clarification that you can add in your manuscript at this subject,

Best regards,

Véronique Geoffroy

On 2021-02-12 10:57:57, user Ramy Karam Aziz wrote:

Nice and a well needed database. Thanks for the citation to our Staph modeling paper among pathogens (Ref 5); however, I'd point out to a rather more relevant reference(s) from Palsson's lab:

https://www.frontiersin.org...<br /> https://www.nature.com/arti...

On 2021-02-09 18:20:06, user Sebastian Gilbert wrote:

No obligation to do so but thought I'd point out the use of CASTLE as a project name already: https://doi.org/10.1101/202... and published at the following doi https://dx.doi.org/10.1088/...

On 2021-02-12 06:34:55, user Mario Malicki wrote:

Should this be fixed to contain an abstract?

On 2021-02-12 03:23:48, user Mario Malicki wrote:

Is the published paper doi wrong, it points to the abstract in Gastro, while the paper should be the one in The Journal of Allergy and Clinical Immunology?https://doi.org/10.1016/j.j...

On 2021-02-11 16:18:27, user James Prudent wrote:

Another simple experiment is to knockdown the NKA in rodent and human line and compare with uM and nM dig respectively. Making broad statements on in vivo data without these would support your in vivo results.

On 2021-02-11 13:39:09, user Oliver Blacque wrote:

Being able to identify orthologous variation across species, in a very user friendly manner, with great visual outputs, is a super contribution. We have been using it a lot to figure out if a missense mutation in a human gene occurs at a conserved position in the C, elegans orthologue. ConVarT makes it very easy to do this. Congrats Oktay, Sebiha and their team !

On 2021-02-11 08:19:02, user Solomon Derese wrote:

It is surprising to note that there is no total protein content variability among the different samples, Do you have any plausible explanation for this observation? Did you consider the effect of ambient temperature, colder and hotter season, in the content of the fatty acids?

On 2021-02-11 06:55:34, user Jacques Fantini wrote:

Interesting study.

As far as the NTD is concerned for neutralizing anti-SARS-CoV-2 antibodies it would be fair to cite previous articles that strongly recommended to focus on this region, in particular:

1. Biochem Biophys Res Commun. 2020 Oct 10:S0006-291X(20)31924-0. doi: 10.1016/j.bbrc.2020.10.015. Leveraging coronavirus binding to gangliosides for innovative vaccine and therapeutic strategies against COVID-19. Fantini J, Chahinian H, Yahi N.

2. Synergistic antiviral effect of hydroxychloroquine and azithromycin in combination against SARS-CoV-2: What molecular dynamics studies of virus-host interactions reveal. Int J Antimicrob Agents. 2020 Aug;56(2):106020. doi: 10.1016/j.ijantimicag.2020.106020. Epub 2020 May 13. Fantini J, Chahinian H, Yahi N.

On 2021-02-11 04:25:44, user Suchita Kumar wrote:

We discussed this paper in our journal club and thought your findings on microbial metabolites identified a significant potential target for pancreatic cancer treatment.

One weakness that we identified was with Fig.1C- without a color guide in the heat map or any valuable taxonomic information, this figure was confusing to read or draw conclusions from. It was similarly confusing to compare samples in 1D because the x-axis was the genus instead of the sample and the colors changed between lean and obese mice (L2 is purple but O2 is red). Given that 1E is the main figure that illustrates the microbiome change in obese mice, we thought 1B-D could be moved to the supplementary figures.

Fig.2 effectively conveyed that the high-fat diet microbiome leads to chemotherapy resistance, but we did think it could be valuable to quantify the adiposity, fibrosis, and apoptosis from the tumor histology (2J-L). It could also be valuable to group 2A+C, 2B+D, and 2H+I for a clearer comparison. One point of confusion was the notation “Lean>Obese”, which could be mistaken to mean that the lean microbiome was becoming an obese microbiome. Using something like Obese (mLean) could eliminate this confusion. It could also be helpful to change the format of the figure legends to “(A) description” instead of “description (A)” since this proved to be confusing as well.

Overall, we thought the abstract flowed logically and the results had great structural organization. Looking forward to seeing more in the future!

On 2021-02-10 20:46:11, user Kanoa Terrazas wrote:

Hello,<br /> I enjoyed reading your paper and recently discussed it with some fellow colleagues in a journal club and wanted to share some thoughts on how the paper could be improved and a few questions.<br /> 1B) Didn't really see much significance, may be more of a supplemental figure. <br /> 1C) Figure was a little difficult to interpret. Maybe indicate families, some valuable taxonomic information, maybe show the ones that change the most on top<br /> 1D) Was difficult to compare the two groups. Maybe make it one graph side by side to directly compare change in microbes rather than two. Also maybe keep colors the same between groups if you keep it at two figures. <br /> 2A-D) Keep obese and lean mice next to each other so you can directly compare change. Maybe stick to blue and red color schemes. Also indicate the P-value for significance throughout the paper.<br /> 2G) Be consistent with using two">>".2J) Lean>>Lean wasn't mentioned in the text.<br /> 2H-I) Combine into one graph. <br /> 2J-L) Maybe quantify the change in a graph somewhere.<br /> 3B&D) Make consistent and use figure legend in both. There are no significance bars? Size of graphs differ.<br /> 3E-J) Label the top as PAC resistant and PAC sensitive. Make X-axis consistent in J.<br /> 4) Didn't find much significance with this figure. May be more of a supplemental figure.<br /> 5A) Include graph examining Q as well. <br /> 5D&E) The asterisk is confusing as to what's being compared and is significant. Also seems that from results, SAM alone is more effective than chemo?<br /> 5F&G) Not explained in the methods. Maybe quantify as well. <br /> 6) Indicate P-value. Also indicate stain used in D&E<br /> 7B) What do the numbers in the upper right mean?<br /> 7) Indicate significance/p-value! <br /> Overall, the work was very exciting and intriguing to read. Great work, look forward to its publishing.

Best regards.

On 2021-02-10 20:39:04, user Senay Beraki wrote:

Dear Dr. Kesh and colleagues,

I am an undergraduate student from UCLA and would like to share some of the comments/questions myself and other students had while analyzing your paper during a journal club session.

Figure 1: The schematic diagram of the experimental timeline in (A) was helpful and important to include in the first figure of the paper. Part (C), however, was confusing because I wasn’t sure what the green and red colors represented. Adding a legend next to the heatmap or in the caption could definitely help understand the heatmap better.

Figure 3: In part (G), it was puzzling and unclear how you were able to get the curve for the red line using those data points. Also, how is the effect of Oxaiplatin treatment in colon cancer cell line connected with the figure’s or the paper’s research question? We were not sure what the purpose of oxaiplatin was in relations to obesity and pancreatic cancer.

Figures 5, 6 and 7: There were a lot of experiments performed for these three figures including flow cytometry, immunohistochemistry, western blot, PCR array, and ELISA based analysis. However, the methodology for most of these techniques was not explained or mentioned in the methods section of the paper. It would be helpful for your readers if you could briefly explain how they were done or reference past papers on those techniques.

Overall, your paper was very interesting and exciting to read. The introduction was well written as it introduced the main components of the papers and made it easier for me to follow the overall research question. Thank you for doing such as incredible research and I hope this feedback helps in strengthening your paper.

On 2021-02-10 18:35:55, user Arnab Ghosh wrote:

The references for "worldwide dominant Asp614Gly<br /> variant, introduced a new elastase proteolysis site in the spike protein (2, 21)." needs to be updated as the statement is not mentioned in the referred papers. The following paper talked about elastase cleavage site at Spike:614G it for first time:<br /> https://www.biorxiv.org/con... (May 05, 2020); now published: https://doi.org/10.1016/j.m...

On 2021-02-10 18:31:27, user Ramy Arnaout wrote:

Fyi, our manuscript has now been peer-reviewed and published in Clinical Infectious Diseases: https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1382/6127024<br /> Thanks to all the comments, which helped us make this work better!

On 2021-02-10 14:25:10, user David Carlson wrote:

This looks like really exciting work. Just a comment - some of the figures (e.g., 1c-h) are very difficult to read, even after zooming in heavily. Perhaps increase font size?

On 2021-02-10 11:15:08, user Guy wrote:

This is great .Are there plans to make the phone app available upon publication? I would observe it would be very useful to have more details physical setup, particularly of the lightning used for the phone work (from above or below, or both) and wether the brand of 96 well plates was important to the results.

On 2021-02-10 07:24:23, user Irina Velsko wrote:

Dear Authors,

Thank you for this paper that addresses a mostly overlooked problem in metagenomics, it is an exciting push forward in the field. It is particularly relevant to the field I work in, ancient metagenomics, which deals with metagenomic data generated from historic and ancient sources. Ancient metagenome samples often have a very high proportion of reads that cannot be taxonomically classified, and determining the origins of these is of great interest. Because I hope to use your paper, based on modern data, to guide future ancient metagenome studies to address this topic, I was surprised to see you included 3 ancient metagenome studies in your analyses.

Ancient samples are affected by a set of properties that require additional processing and validation steps that modern data do not. These include damage of the endogenous DNA, as well as contamination with modern organisms. Authentic ancient DNA (aDNA) damage is characterized by short fragment lengths typically less than 100bp, and conversion of cytosines to uracils, which then become thymines in the sequencing data. The presence of such damage patterns is used to authenticate ancient samples, and to distinguish endogenous DNA from exogenous, modern contamination.

Ancient samples are obtained archaeological contexts, and are generally contaminated to at least some degree by environmental organisms derived from the burial environment. Particular steps need to be taken to remove these modern contaminants from taxonomic profiles or before assembly. Leaving them risks both assembling non-source organisms (such as assembling soil bacteria in oral samples), and of generating chimeric assemblies, which include reads from both ancient and modern sources. Indeed, the feasibility of assembling ancient metagenomes is still being assessed within the field, and the effects of aDNA properties on assembly is not well understood.

For these reasons, I recommend replacing the 3 aDNA datasets (PRJEB6045, PRJEB12831, PRJEB15334) with 3 modern data sets. Unless the ancient datasets are fully authenticated, assessed for preservation (the level of environmental contamination), and the assemblies carefully checked to remove chimeras, there are many variables affecting the outcome of the assembly process that may interfere with drawing sound conclusions for your study.

Finally, as the person who generated the data in PRJEB15334, I strongly caution the use of this version of the dataset. This data was auto-processed by the EBI-metagenomics pipeline, which does not account for any aDNA properties, and which systematically removes short DNA sequences (i.e., ancient DNA) during the quality control portion of its automated pipeline. Instead, if you would like to use this historic dataset, I recommend using PRJEB30331, which contains the full raw data of these same libraries but without the problematic auto-filtering step.

For in-depth reading about the nuances of ancient DNA analysis, I recommend the following papers as a general introduction:

A Robust Framework for Microbial Archaeology<br /> https://www.annualreviews.o...

Mining Metagenomic Data Sets for Ancient DNA: Recommended Protocols for Authentication<br /> https://doi.org/10.1016/j.t...

From the field to the laboratory: Controlling DNA contamination in human ancient DNA research in the high-throughput sequencing era<br /> https://www.tandfonline.com...

Patterns of damage in genomic DNA sequences from a Neandertal<br /> https://www.pnas.org/conten...

mapDamage2.0: fast approximate Bayesian estimates of ancient DNA damage parameters<br /> https://academic.oup.com/bi...

Separating endogenous ancient DNA from modern day contamination in a Siberian Neandertal<br /> https://www.pnas.org/conten...

Thank you, <br /> Irina Velsko<br /> velsko@shh.mpg.de

On 2021-02-09 21:52:44, user Virginia Savova wrote:

Please check out the updated version for more benchmarking against other methods

On 2021-02-09 17:01:32, user Virginia Savova wrote:

Now updated with comparison to other methods

On 2021-02-09 16:07:04, user Martijn Egas wrote:

Dear Nathan Burke and Gregory Holwell,

Thanks for an interesting paper; we discussed it in our weekly journal club of our department Evolutionary and Population Biology at the University of Amsterdam, and it is our custom to collect the comments made during the discussion and post them here in the comments. Due to lack of time these comments are rather short but do convey the main issues we discussed.

Martijn Egas (egas@uva.nl)

General:<br /> The paper appears biased towards the male perspective and ignores the female perspective (partner choice vs rival presence). We would have expected in the discussion some reflection on the female perspective.

Methodology:<br /> How do you identify the rival male and the focal male? If these are not distinguished, there may be a bias in the treatments with two males, e.g. whoever mated first got to become the focal male, and compared to single-male treatments matings may occur more quickly (because the fastest male would count in the two-male treatment).<br /> It is unclear how observations were made regarding the mating behavior – how is it established that a mating occurred? Is it possible that a female mates with both males in rapid sequence? Why is it enough to do hourly observations, and what was observed at those times? Please explain in the methods section.

Setup used:<br /> We were worried about the different feeding regimes directly prior to the experiment – why would the females be equally satiated at the start of the experiment based on these two regimes? Do females eat normally 3 flies per day? And how fast do they get hungry once they are in absence of food – within the 8 hours of the experiment?<br /> Testing for distraction may be done more directly by actively handing females big prey items, rather than by providing many flies for a long time. Why was this setup chosen?<br /> Could the lack of cannibalism be due to the type of enclosure used here, with plastic leaves etc, so that males could escape more easily than in the experiment with plastic jars that you reported on in Biology Letters 2021.

Minor comments:<br /> L. 83: abundance of prey is not considered in your experiments, only presence/absence<br /> The first paragraph of results is confusing, using mating rate on one hand and speed of approach on the other hand, and making various comparisons among the 4 available treatment without explicitly mentioning which (e.g. from l. 167 the hazard ratio mentioned there for the presence of a rival, is that the estimate for the aggregated treatments with and without food?)

On 2021-02-09 15:44:10, user Jasmina Wiemann wrote:

The data provided in this Preprint have been quantitatively analyzed in the context of a larger spectral sample set, and no similarities between the disputed spectra and edge filter artefacts were recovered. Please find the detailed response to this study here: <br /> https://t.co/nFE4pqi9wx?amp=1

On 2021-02-09 15:14:22, user Marty McFly wrote:

Interesting paper.

But with figure 4c, something went wrong.1st and 3rd picture at 28 °C look very similar, although these are different yeast strains. Just sayin :-)

Maybe it should be reviewed.

On 2021-02-09 14:22:41, user Alex Graudenzi wrote:

Dear Authors,<br /> Thank you very much for your interesting work.

We believe that there might be a misunderstanding regarding the methodology described in our quoted article "Mutational signatures and heterogeneous host response revealed via large-scale characterization of SARS-CoV-2 genomic diversity" (which is currently published on iScience, https://doi.org/10.1016/j.isci.2021.102116).

Indeed, our method processes raw sequencing data of SARS-CoV-2 samples, and this allows us to call both clonal variants, i.e., mutations with high variant frequency and typically included in consensus sequences, and "minor" variants, i.e., mutations displaying low variant frequency, which are typically excluded from consensus sequences.

As extensively discussed in our manuscript, the "mutational signatures" are then discovered by processing the profiles of those variants that are always detected as minor. <br /> Accordingly, the variants detected as clonal in at least one sample are excluded from the mutational signature analysis. <br /> This choice was specifically aimed at avoiding any possible bias related to the accumulation of clonal variants during the evolution of the virus.

We hope this clarification could be useful.

Sincerely,<br /> Alex Graudenzi, also on behalf of the other authors

On 2021-02-09 09:28:30, user Hansi Weißensteiner wrote:

Paper published on Genome Research - 2021-01-15: https://genome.cshlp.org/co...

On 2021-02-09 08:33:27, user Florian Privé wrote:

Please note that this work is now included as a Supplementary Note in https://doi.org/10.1101/202....

On 2021-02-09 06:13:39, user Aaron Hogan wrote:

cool paper! Chl fluorescence is a great tool for inferring physiological functionin of plants and this study demonstrates how it can be used to shed light on ecological processses (e.g. competetion & neighborhood interactions).

On 2021-02-08 20:11:04, user Nicholas Markham wrote:

Outstanding manuscript by Pruss and Sonnenburg! Thanks to the authors for posting on bioRxiv. These elegant experiments show how C. difficile toxin-mediated host inflammation alters the metabolic phenotype of C. difficile itself. In particular, the authors identified an inflammation-associated upregulation of the sorbitol utilization locus responsible for permitting increased C. difficile growth in the presence of sorbitol. By altering the host's ability to produce sorbitol, including the use of an aldose reductase knockout, they show differential severity of C. difficile infection. This novel work is exciting because it reveals multiple potential therapeutic targets. I imagine the reviews will be kind. I wonder if looking more at the infected ARKO mice would be helpful: fecal CFUs, fecal toxin amounts, and histopathology.

On 2021-02-08 13:46:29, user Qingyang Li wrote:

This is a fundamentally flawed model based on the assumption that vaccine can stop transmission. The currentt intramuscular vaccine cannot induce sufficient IgA which is critical for mucous immunity, so it cannot stop virus growth and shredding in the upper respiratory tract. Israel and UAE are seeing case rising despite highest vaccination rate.

On 2021-02-08 12:51:08, user D Hapangama wrote:

It is great to see confirmation of our work, published in journal of Pathology 2020. Histological 3D reconstruction and in vivo lineage tracing of the human endometrium.<br /> Tempest N, Jansen M, Baker AM, Hill CJ, Hale M, Magee D, Treanor D, Wright NA, Hapangama DK. J Pathol. 2020 Aug;251(4):440-451. doi: 10.1002/path.5478. Epub 2020 Jun 30.<br /> It is high time the true 3D microarchitecture of the human endometrium is appriciated fully.

On 2021-02-08 12:19:27, user Brandon Seah wrote:

Thank you for a fascinating study on the genomic evolution of Paramecium!

I have a question to Figure S12. The caption states that P. tetraurelia has a MIC k-mer peak at 31x followed by a peak representing the MAC contamination. However, the k-mer spectra panels for all three species look very similar (left figure column), and I do not see the peaks described for P. tetraurelia. Could it be that the wrong graphs were included in the figure?

On 2021-02-08 11:53:52, user Łukasz Sobala wrote:

Dear Authors,

This is an important and timely study, I have a few comments about it.

The cells used in the study (Vero E6 and HEK293) might be expressing an enzyme conferring an alternative pathway to glycan maturation - Golgi endomannosidase. HEK293 do express this but I don't know if data exists for Vero cells. In this case, blocking glucosidase II by swainsonine of knocking down GANAB might be less effective if the GH99 pathway is left to function. It would be interesting to try a combination therapy in this case.

Also, why does MGAT1 knockdown have a different effect in Vero and HEK293?

Thanks

On 2021-02-07 07:23:15, user Helena Kuivaniemi wrote:

The statement in the abstract that TAA is different from AAA is not new (e.g., see a review article Tromp G, Kuivaniemi H, Hinterseher I, Carey DJ. Novel genetic mechanisms for aortic aneurysms. Current Atherosclerosis Reports 12:259-266, 2010; https://pubmed.ncbi.nlm.nih.... Also, your sample size for TAA is so small that you would probably not be able to detect all the loci, thus it is unlikely that your study was powered to test if the genetic loci are the same for TAA and AAA.

On 2021-02-06 21:25:41, user Susan Fawcett wrote:

This article is now published in APPS: https://bsapubs.onlinelibra...

On 2021-02-06 18:28:13, user Uli Wellner wrote:

very interesting concept. how do you rule out the opposite theory ie duodenal / ampullary carcinoma invading the pancreas and switching to pancreatobiliary type ?