Diversity, composition, and networking of saliva microbiota distinguish the severity of COVID-19 episodes as revealed by an analysis of 16S rRNA variable V1-V3 region sequences

ABSTRACT Studies on the role of the oral microbiome in SARS-CoV-2 infection and severity of the disease are limited. We aimed to characterize the bacterial communities present in the saliva of patients with varied COVID-19 severity to learn if there are differences in the characteristics of the microbiome among the clinical groups. We included 31 asymptomatic subjects with no previous COVID-19 infection or vaccination; 176 patients with mild respiratory symptoms, positive or negative for SARS-CoV-2 infection; 57 patients that required hospitalization because of severe COVID-19 with oxygen saturation below 92%, and 18 fatal cases of COVID-19. Saliva samples collected before any treatment were tested for SARS-CoV-2 by PCR. Oral microbiota in saliva was studied by amplification and sequencing of the V1-V3 variable regions of 16S gene using an Illumina MiSeq platform. We found significant changes in diversity, composition, and networking in saliva microbiota of patients with COVID-19, as well as patterns associated with severity of disease. The presence or abundance of several commensal species and opportunistic pathogens were associated with each clinical stage. Patterns of networking were also found associated with severity of disease: a highly regulated bacterial community (normonetting) was found in healthy people whereas poorly regulated populations (disnetting) were characteristic of severe cases. Characterization of microbiota in saliva may offer important clues in the pathogenesis of COVID-19 and may also identify potential markers for prognosis in the severity of the disease. IMPORTANCE SARS-CoV-2 infection is the most severe pandemic of humankind in the last hundred years. The outcome of the infection ranges from asymptomatic or mild to severe and even fatal cases, but reasons for this remain unknown. Microbes normally colonizing the respiratory tract form communities that may mitigate the transmission, symptoms, and severity of viral infections, but very little is known on the role of these microbial communities in the severity of COVID-19. We aimed to characterize the bacterial communities in saliva of patients with different severity of COVID-19 disease, from mild to fatal cases. Our results revealed clear differences in the composition and in the nature of interactions (networking) of the bacterial species present in the different clinical groups and show community-patterns associated with disease severity. Characterization of the microbial communities in saliva may offer important clues to learn ways COVID-19 patients may suffer from different disease severities.

Lines 271-279: The author list numerous species that appear to alter among groups.Yet, there is no mention of whether abundances are significant different among groups.Furthermore, it is not clear what test is being used to identify targeted microbial taxa.Are species lowly abundant?Are rank correlations being implemented or ANCOM.Also, it would be beneficial to provide a general distribution of microbial phyla.It is not clear whether the species reside in Firmicutes, Actinobacteria, or Bacteroidetes etc.
Line 294: It would be helpful to report effect size for microbial taxa significantly different between groups.
Line 364: It is standard to report the area under curve (AUC) for random forest in order to better assess the model's accuracy.In addition, it would be beneficial to report the confusion matrix in order to better assess how many samples were correctly predicted for each group.
Line 388: What is meant by endurance?Lines 389-392: It is not exactly clear what the implications of the microbial taxa regarding the core microbiome.Does the core microbiome increase/decrease with severity.Is this related to the previously mentioned alpha diversity measures in any way?This is not entirely clear.
Line 415: The author denotes significance, but it is not evident in Table 2.It would be helpful to show p-values for Pearson correlations.Also, I think the x and y-axis for Table 2 are inverted.As currently constructed it comes off as comparing positive and negative links, but I believe you are truly comparing links between groups which would then provide the opportunity to add an additional column to report p-values.
Line 425: It would be helpful to report centrality to support this statement.
Line 430: How were subclusters identified?This wasn't stated in methods.Did you use k-means clustering when generating networks or are you basing this on a network property such as closeness.This should be explained and reported either way.
Lines 439-442: It would be helpful to define how you are grouping the microbial taxa into clusters.
Line 501: What test was applied to denote significance?Line 514: For Figure 6A, what does the x-axis represent?
Lines 584-589: I agree with the author that comparison of microbial taxa across severity is rare and makes it difficult to compare to previous studies, but it would be helpful to offer an explanation as to why certain bacteria are favored for a given state.The author does this for the asymptomatic state, it would be helpful to add some clarity as the results section list a number of different species.
Lines 643-654: I'm not exactly how this is accurate.The author did not assess connectivity i.e., if a portion of the network is disrupted how many links are loss within the network.

Additional comments:
It would be helpful to discuss the limitations of the study.Specifically, the number of statistically significant bacteria across severity.While there are some common bacteria listed across statistical analyses, there are some bacteria that are significant in some test but not others.This should be addressed, which is likely due to variation among humans.
Given the author repeated mentioned pathogenesis and dysbiosis in the manuscript, I assumed there would be some type of dysbiosis analysis.This would complement the analyses and help discern whether the oral microbiota is associated with severity.DysbiosisR seems like an appropriate statistical analysis given the methodology of the paper.
Reviewer #2 (Comments for the Author): -line 50: please add number of patients (n=xxx) for each group.
-line 138: provide PDF copy of Ethical Committee approval.
-line 200: please report how relative abundances matrix was normalized before stepping to alpha-and beta-diversity calculations.
-line 219: this paragraph is not clear.Definite prevalence and abundance cutoffs were chosen prior to analyze alfa-betadiversity and the others analysis?Please explain.-line 226: replace "co-occurrence" with "co-abundance" throughout the manuscript.
-line 235: as for microbiota data, also Picrust2 data need to be transformed before making any statistical operation, please write how data were transformed (see advise by Douglas and Langille).
-line 266: please add 95% ellipse CI for each group.
-line 266: please report for all the beta-diversity metrics the P values for inter-group comparison.
-line 268: this graph is not meaningful to the readers, because no insights are given of DA among discriminant species (severitydriven P values ?). Please do a RF or VIP plot with variable importance.Or, at least draw P values for each species among all groups (Kruskal-Wallis) with post-hoc test (Mann-Whitney), but on transformed data!-line 270: I advise to have an insight on 16S NCBI database, encompassing >50k genomes, updated every two-three days.In contrast, in eHOMD v3 (is that this version was used?) total number of genomes are 2087 including non-oral/non-nasal taxa.It means that some non-oral bacterial species should have been unrecognized or under-detected.
-line 279: please report objective values of Delta among relative abundances or P values when dealing with comparisons, avoid subjective jargon such as "particularly higher" and so on.
-line 296: why enhanced?In any case, Volcano plots do not visualize species with zero relative abundance in at least one group.I advise to draw a Figure1D panel reporting FoldC differences.Or, at least, a NMDS bi-plot, superimposing on Figure1B the eigenvalues of bacterial drivers, we should see on bottom-right corner species such as A.baumanni, Prevotella spp., Veillonella spp., etc.. -line 363: please cluster the bacterial species by the cohorts, reporting in descending order of "mean accuracy" and "LDA": in this manner it's easier for the Reader to follow results in a nutshell.I would also like to see a ROC curve with selected bacterial species able to separate 1vs1 the different cohorts.
-line 413-414: this is not the usual workflow for network co-abundance analysis.The best option is to do networks with all the species, then filtering edges with Pearson coefficients (e.g.0.7) or P values (e.g.<0.01), then ruling out singletons, then assessing communities/clusters.In this way the network is unbiased for community and functional definition.I advise to calculate betweenness centrality and figure out keystone species in each cohort.I stress also the previous comment on drawing ROC curves using the selected bacterial species (here is right to use results from volcano, lefse, RF, and networks).
-line 419: transform this Table2 in an inset of Figure5 as stacked barplot.Erase then Table2.
-line 501: the Picrust2 dataset was normalized for total gene count before doing analyses?it seems to be a bias at level2, which affects also level3.
-line 645: I would see also these network metrics computed for each severity-cohort and reported in Figure5: i) Average Path length; ii) density; iii) diameter; iv) number of communities.
The author details the connection between the oral microbiota across severity in COVID-19 infections.This manuscript is timely and needed given the need to better understand impact of hospitalizations and co-infections during the pandemic, as well as health outcomes.I offer the following comments below: Line 136: It is not clear whether the authors received IRB approval to conduct the study.
Line 206: It is not clear why the author chose to use NMDS versus principal coordinate analysis (PCoA) when assessing and visualizing microbial community structure.This should be further explained in the methods section.Furthermore, the stress value is often reported when using NMDS as this signifies whether the ordination is arbitrary i.e., a stress value above .02is generally questionable typically.
Line 213: For the random forest model, the author should clarify whether an out-of-bag or a cross-folded validation was applied.These parameters can drastically change the model's accuracy and reliability.Additionally, was the random forest model generated in R or python?Line 266: The author mentions some separation in of bacterial communities, but this is not clear.Is the author referring to microbial variation?If so, what is the R-value associated with this variation and is this significant.This should be added to provide clarity.Additionally, the author previously mentioned in the discussion section applying permutational dispersion; however, this is not reported in the results.Is there homogeneity among the samples?
Lines 271-279: The author list numerous species that appear to alter among groups.Yet, there is no mention of whether abundances are significant different among groups.Furthermore, it is not clear what test is being used to identify targeted microbial taxa.Are species lowly abundant?Are rank correlations being implemented or ANCOM.Also, it would be beneficial to provide a general distribution of microbial phyla.It is not clear whether the species reside in Firmicutes, Actinobacteria, or Bacteroidetes etc. Line 294: It would be helpful to report effect size for microbial taxa significantly different between groups.
Line 364: It is standard to report the area under curve (AUC) for random forest in order to better assess the model's accuracy.In addition, it would be beneficial to report the confusion matrix in order to better assess how many samples were correctly predicted for each group.
Line 388: What is meant by endurance?Lines 389-392: It is not exactly clear what the implications of the microbial taxa regarding the core microbiome.Does the core microbiome increase/decrease with severity.Is this related to the previously mentioned alpha diversity measures in any way?This is not entirely clear.
Line 415: The author denotes significance, but it is not evident in Table 2.It would be helpful to show p-values for Pearson correlations.Also, I think the x and y-axis for Table 2 are inverted.As currently constructed it comes off as comparing positive and negative links, but I believe you are truly comparing links between groups which would then provide the opportunity to add an additional column to report p-values.
Line 425: It would be helpful to report centrality to support this statement.
Line 430: How were subclusters identified?This wasn't stated in methods.Did you use k-means clustering when generating networks or are you basing this on a network property such as closeness.This should be explained and reported either way.
Lines 439-442: It would be helpful to define how you are grouping the microbial taxa into clusters.
Line 501: What test was applied to denote significance?Line 514: For Figure 6A, what does the x-axis represent?Lines 584-589: I agree with the author that comparison of microbial taxa across severity is rare and makes it difficult to compare to previous studies, but it would be helpful to offer an explanation as to why certain bacteria are favored for a given state.The author does this for the asymptomatic state, it would be helpful to add some clarity as the results section list a number of different species.
Lines 643-654: I'm not exactly how this is accurate.The author did not assess connectivity i.e., if a portion of the network is disrupted how many links are loss within the network.

Additional comments:
It would be helpful to discuss the limitations of the study.Specifically, the number of statistically significant bacteria across severity.While there are some common bacteria listed across statistical analyses, there are some bacteria that are significant in some test but not others.This should be addressed, which is likely due to variation among humans.
Given the author repeated mentioned pathogenesis and dysbiosis in the manuscript, I assumed there would be some type of dysbiosis analysis.This would complement the analyses and help discern whether the oral microbiota is associated with severity.DysbiosisR seems like an appropriate statistical analysis given the methodology of the paper.

Reviewer #1
The author details the connection between the oral microbiota across severity in COVID-19 infections.This manuscript is timely and needed given the need to better understand impact of hospitalizations and co-infections during the pandemic, as well as health outcomes.I offer the following comments below: Line 136: It is not clear whether the authors received IRB approval to conduct the study.R. The study was approved by the Institutional IRB, this is now described in lines 145-147 Line 206: It is not clear why the author chose to use NMDS versus principal coordinate analysis (PCoA) when assessing and visualizing microbial community structure.This should be further explained in the methods section.Furthermore, the stress value is often reported when using NMDS as this signifies whether the ordination is arbitrary i.e., a stress value above .02is generally questionable typically.R. NMDS with Bray-Curtis dissimilarity index is suggested to visualize beta diversity, although the use of PCoA is usually more accepted.Thus, as suggested, we now show the analyses of beta diversity using principal coordinate analysis (PCoA) with unweighted and weighted UniFrac to visualize covid-associated bacterial communities.See modified figure 1B.
Line 213: For the random forest model, the author should clarify whether an out-of-bag or a cross-folded validation was applied.These parameters can drastically change the model's accuracy and reliability.Additionally, was the random forest model generated in R or python?R. For the Random Forest, we used the R Caret's package, MLeval, running 1000 trees.For cross-validation values were multiplied by 10, and the training set was 90% while the test set was 10%.In addition, for further validation an AUC frame was determined (see lined 231-236 and suppl figure 2).
Line 266: The author mentions some separation in of bacterial communities, but this is not clear.Is the author referring to microbial variation?If so, what is the R-value associated with this variation and is this significant.This should be added to provide clarity.Additionally, the author previously mentioned in the discussion section applying permutational dispersion; however, this is not reported in the results.Is there homogeneity among the samples?R. The significance in the separation of bacterial communities was analyzed by PERMANOVA, ANOSIM and PERMDISP tests as described in Methods, lines 219-22 and Results, in figure 1B.
Lines 271-279: The author list numerous species that appear to alter among groups.Yet, there is no mention of whether abundances are significant different among groups.Furthermore, it is not clear what test is being used to identify targeted microbial taxa.Are species lowly abundant?Are rank correlations being implemented or ANCOM.Also, it would be beneficial to provide a general distribution of microbial phyla.It is not clear whether the species reside in Firmicutes, Actinobacteria, or Bacteroidetes etc. R. Considering that variation of species among groups is already analyzed more in detail by enhanced volcano, Random Forest and Lefseq, we decided to remove the initial analyses previously shown in figure 1C.
Line 294: It would be helpful to report effect size for microbial taxa significantly different between groups.R. To analyze the differences in taxa abundance among groups, the Fold Change (FC) and the FDR-adjusted values were calculated using DESeq2, and for normalization geometric mean was calculated for each ASVs across all samples.The data were previously filtered based on the significant results of RandomForest and Lefseq.This is now described in Methods, lines 237-242.
Line 364: It is standard to report the area under curve (AUC) for random forest in order to better assess the model's accuracy.In addition, it would be beneficial to report the confusion matrix in order to better assess how many samples were correctly predicted for each group.

R. The AUC and the confusion matrix for the random forest is now shown in Suppl. figure 2.
Line 388: What is meant by endurance?R. By endurance we mean taxa that persist the most across all clinical groups, regardless of the changes in the microenvironment in each clinical group.See note in line… Lines 389-392: It is not exactly clear what the implications of the microbial taxa regarding the core microbiome.Does the core microbiome increase/decrease with severity.Is this related to the previously mentioned alpha diversity measures in any way?This is not entirely clear.R. The core genome estimation is meant to find out which taxa prevail the most in samples from all groups analyzed after filtering data.As mentioned in Methods lines 246-249 we filtered using taxa prevalent in over 50% of the samples with a relative abundance above 0.2.These are the Taxa that vary the least across clinical groups and are not meant to give data on increase/decrease with severity nor with diversity.
Line 415: The author denotes significance, but it is not evident in Table 2.It would be helpful to show p-values for Pearson correlations.Also, I think the x and y-axis for Table 2 are inverted.As currently constructed it comes off as comparing positive and negative links, but I believe you are truly comparing links between groups which would then provide the opportunity to add an additional column to report p-values.R. Line 501: What test was applied to denote significance?R. As described in the above comment, in the co-occurrence network analysis NetCoMi defines subclusters based on Pearson correlation to show species with significant correlation as described by color of the nodes.This is described in Methods, lines 251-256.
Line 514: For Figure 6A, what does the x-axis represent?R. The x-axis in Figure 7A-D (previously 6A) represents pathway abundance as the legend describes now.Please note that the figure was edited according to reviewer´s suggestions Lines 584-589: I agree with the author that comparison of microbial taxa across severity is rare and makes it difficult to compare to previous studies, but it would be helpful to offer an explanation as to why certain bacteria are favored for a given state.The author does this for the asymptomatic state, it would be helpful to add some clarity as the results section list a number of different species.R.This is a good point that we tried to address in changes made in lines 650-663 of the Discussion section.
Lines 643-654: I'm not exactly how this is accurate.The author did not assess connectivity i.e., if a portion of the network is disrupted how many links are loss within the network.6 (that relaced Table 2) to more clearly show the contrast between groups including the total number of links.

Additional comments:
It would be helpful to discuss the limitations of the study.Specifically, the number of statistically significant bacteria across severity.While there are some common bacteria listed across statistical analyses, there are some bacteria that are significant in some test but not others.This should be addressed, which is likely due to variation among humans.R. A paragraph with limitations of the study was added in the Discussion section, lines 680-692.
Given the author repeated mentioned pathogenesis and dysbiosis in the manuscript, I assumed there would be some type of dysbiosis analysis.This would complement the analyses and help discern whether the oral microbiota is associated with severity.Dysbiosis.R. A dysbiosis score was calculated using dysbiosisR V1.0.4 and results are presented in Suppl fig 4. The analysis is also described in Methods lines 224-227 and in Results lines 311-314.

Reviewer #2 (Comments for the Author):
-line 50: please add number of patients (n=xxx) for each group.R. numbers were added as suggested.
-line 138: provide PDF copy of Ethical Committee approval.R. PDF will be uploaded as requested.
-line 200: please report how relative abundances matrix was normalized before stepping to alpha-and beta-diversity calculations R. For normalization geometric mean was calculated for each ASVs across all samples using Total Sum Scaling (TSS) as described in Methods lines 209-211, where a reference was also added, ref 21.TSS is recommended because it most accurately captures the composition of the original communities, whereas UQ and CSS distort communities.
-line 211: without a proper normalization (CLR, Gaussain-like/standardization) differences in taxa are flawed by inherent compositionality.R. see comment above.
-line 219: this paragraph is not clear.Definite prevalence and abundance cutoffs were chosen prior to analyze alfa-beta-diversity and the others analysis?Please explain.R. The paragraph on the analyses of diversity, lines 209-222 was edited to address this and other observations.
-line 226: replace "co-occurrence" with "co-abundance" throughout the manuscript.R. done, thanks -line 235: as for microbiota data, also Picrust2 data need to be transformed before making any statistical operation, please write how data were transformed (see advise by Douglas and Langille).R. We also edited the paragraph on analyses of metabolic routes, lines 267-277 to better describe the methods.
-line 266: please add 95% ellipse CI for each group.R. By suggestion of reviewer 1 this analysis was modified as illustrated in Figure 1B where some statistical values are shown.This is also described in Methods, lines 216-222.
-line 266: please report for all the beta-diversity metrics the P values for inter-group comparison.1B -line 268: this graph is not meaningful to the readers, because no insights are given of DA among discriminant species (severity-driven P values ?). Please do a RF or VIP plot with variable importance.Or, at least draw P values for each species among all groups (Kruskal-Wallis) with post-hoc test (Mann-Whitney), but on transformed data!R. Considering that variation of species among groups is analyzed in more detail by enhanced volcano, Random Forest and Lefseq, we decided to remove the initial analyses previously shown in figure 1C.

R. p-values are now shown in Figure
-line 270: I advise to have an insight on the 16S NCBI database, encompassing >50k genomes, updated every two-three days.In contrast, in eHOMD v3 (is that this version was used?) total number of genomes are 2087 including non-oral/non-nasal taxa.It means that some non-oral bacterial species should have been unrecognized or under-detected.R. True, by including the NCBI database we could identify additional species, however, eHOMD has proved to be a comprehensive microbiome database for the study of microbiome in the aerodigestive tract with a better resolution than other common databases.Still, we acknowledge the limitation of our results and accept that it can be improved with the NCBI database, this is described in Discussion, lines 687-692.
-line 279: please report objective values of Delta among relative abundances or P values when dealing with comparisons, avoid subjective jargon such as "particularly higher" and so on.R. We now indicate p-values instead of subjective phrases.
-line 296: why enhanced?In any case, Volcano plots do not visualize species with zero relative abundance in at least one group.I advise to draw a Figure1D panel reporting FoldC differences.Or, at least, a NMDS bi-plot, superimposing on Figure1B the eigenvalues of bacterial drivers, we should see on bottom-right corner species such as A.baumanni, Prevotella spp., Veillonella spp., etc.. R. Enhanced refers to enhanced-visual, where the X axis represents the FoldC.
-line 363: please cluster the bacterial species by the cohorts, reporting in descending order of "mean accuracy" and "LDA": in this manner it's easier for the Reader to follow results in a nutshell.I would also like to see a ROC curve with selected bacterial species able to separate 1vs1 the different cohorts.R. Cohorts in Figure 3B were clustered and ordered as suggested and a ROC analyses with the LDA scores is now shown in the figure -line 413-414: this is not the usual workflow for network co-abundance analysis.The best option is to do networks with all the species, then filtering edges with Pearson coefficients (e.g.0.7) or P values (e.g.<0.01), then ruling out singletons, then assessing communities/clusters.In this way the network is unbiased for community and functional definition.I advise to calculate betweenness centrality and figure out keystone species in each cohort.I stress also the previous comment on drawing ROC curves using the selected bacterial species (here is right to use results from volcano, lefse, RF, and networks).R. Analysis of network was done again using the suggested workflow and the paragraph of co-abundance modified accordingly (see lines [255][256][257][258][259][260][261][262][263][264][265] and results shown now in figure 5.
-line 419: transform this Table2 in an inset of Figure5 as stacked barplot.Erase then Table2.R. Table 2 data is now presented as a figure 6.
-line 501: the Picrust2 dataset was normalized for total gene count before doing analyses?it seems to be a bias at level2, which affects also level3.R. Details of the analysis of metabolic routes are described in lines 267-277 -line 561: see previous comment about eHOMD.R= See our reply to the previous comment -line 645: I would see also these network metrics computed for each severity-cohort and reported in Figure5: i) Average Path length; ii) density; iii) diameter; iv) number of communities.R. Additional network metrics were measured and reported in figure 6. Re: mSystems01062-22R1 (Diversity, composition, and networking of saliva microbiota distinguish the severity of COVID-19 episodes as revealed by an analysis of 16S rRNA variable V1-V3 regions sequences) Dear Prof. Javier Torres: Thank you to the authors for thoughtfully and thoroughly addressing the reviewer's comments.It is greatly appreciated Your manuscript has been accepted, and I am forwarding it to the ASM Journals Department for publication.For your reference, ASM Journals' address is given below.Before it can be scheduled for publication, your manuscript will be checked by the mSystems production staff to make sure that all elements meet the technical requirements for publication.They will contact you if anything needs to be revised before copyediting and production can begin.Otherwise, you will be notified when your proofs are ready to be viewed.
ASM policy requires that data be available to the public upon online posting of the article, so please verify all links to sequence records, if present, and make sure that each number retrieves the full record of the data.If a new accession number is not linked or a link is broken, provide production staff with the correct URL for the record.If the accession numbers for new data are not publicly accessible before the expected online posting of the article, publication of your article may be delayed; please contact the ASM production staff immediately with the expected release date.
As an open-access publication, mSystems receives no financial support from paid subscriptions and depends on authors' prompt payment of publication fees as soon as their articles are accepted.

Publication Fees:
We have partnered with Copyright Clearance Center to collect author charges.You will soon receive a message from no-reply@copyright.com with further instructions.For questions related to paying charges through RightsLink, please contact Copyright Clearance Center by email at ASM_Support@copyright.com or toll free at +1.877.622.5543.Hours of operation: 24 hours per day, 7 days per week.Copyright Clearance Center makes every attempt to respond to all emails within 24 hours.For a complete list of Publication Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees.Need to upgrade your membership level?Please contact Customer Service at Service@asmusa.org.
If you would like to submit a potential Featured Image, please email a file and a short legend to msystems@asmusa.org.Please note that we can only consider images that (i) the authors created or own and (ii) have not been previously published.By submitting, you agree that the image can be used under the same terms as the published article.File requirements: square dimensions (4" x 4"), 300 dpi resolution, RGB colorspace, TIF file format.
For mSystems research articles, you are welcome to submit a short author video for your recently accepted paper.Videos are normally 1 minute long and are a great opportunity for junior authors to get greater exposure.Importantly, this video will not hold up the publication of your paper, and you can submit it at any time.
Details of the video are: • Minimum resolution of 1280 x 720 • .movor .mp4.video format • Provide video in the highest quality possible, but do not exceed 1080p • Provide a still/profile picture that is 640 (w) x 720 (h) max • Provide the script that was used We recognize that the video files can become quite large, and so to avoid quality loss ASM suggests sending the video file via https://www.wetransfer.com/.When you have a final version of the video and the still ready to share, please send it to mSystems staff at msystems@asmusa.org.
Table 2 was questioned by the two reviewers and one suggested presenting the data as a figure, and so, figure 6 now replaces Table 2.The significance data are described in lines 476-482 and in figure 6.It would be helpful to report centrality to support this statement.3 R.Additional values to describe the nets are now included in figure 6.How were subclusters identified?This wasn't stated in methods.Did you use kmeans clustering when generating networks or are you basing this on a network property such as closeness.This should be explained and reported either way.It would be helpful to define how you are grouping the microbial taxa into clusters.R. see answer to line 430 for details.
Thank you for submitting your paper to mSystems.