Dear Dr. Bar-Joseph,

Thank you very much for submitting your manuscript "Predicting Lung Aging using scRNA-Seq Data" for consideration at PLOS Computational Biology.

As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Sushmita Roy, Ph.D.

Section Editor

PLOS Computational Biology

Sushmita Roy

Section Editor

PLOS Computational Biology

***********************

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: Authors developed a method named polyEN for prediction of ‘molecular age’ from single cell gene expression data. This method was benchmarked against existing methods in lung cell type data for smokers and non-smokers. The benchmarking is reasonable, and the results of the new methods are better than others. I have one comments about if expression is a better age predictor than methylation.

1. How to predict age for each cell type of lung from the methylation data is not clear. Did authors use single cell DNA methylation? A description of the methylation data used here is needed. Does the methylation data is profiled from the same cohort of individuals with the single cell RNA-seq data? I guess the three methods (Horvath, Hannum, and GrimAge) were trained on the bulk methylation data. If authors used single cell methylation data as input, how does it affect the results? The polyEN results are trained and tested on the similar datasets, however this is not the case for the methylation-based predictions.

Reviewer #2: Summary:

The paper presents an age prediction model for single-cell lung aging data sets. The method selects the list of gene sets from the literature related to senescence (or PCA-reduced dimensions of all genes) and calculates the mean and variance of the gene expression for the cells in a cell type. Next, it inputs first- and second-degree feature polynomials into an elastic net to predict the age of the donor's cell type as a regression task. The model is called PolyEN. After predictions, the paper performs SHAP analysis to learn relevant features for the model’s predictions. Some of the key insights from the work include: (1) PolyEN produces better results than existing bulk and single-cell age prediction models that perform regression (2) Basal cells accurately predict age for non-smoking individuals in the data (3) Smooth muscle cells and venous endothelial cells accurately predict donor age for smokers and (4) SHAP analysis reveals genes related to interesting gene ontology descriptions to understand aging better.

Major comments:

The idea of modeling the features as first- and second-degree polynomials is interesting, and the intuition behind it makes sense since the quadratic form could capture feature trends beyond the linear formulation. Investigating age regression performance that is consistent for certain cell types and gene sets helps understand aging in lungs among smokers and non-smokers. However, it would be helpful to include more detailed descriptions of the methodological setup and modeling choices to robustly support the results of the paper (see points below).

First, what was the intuition behind calculating means and variances instead of using single-cell gene expression read-out directly (as done in previous single-cell age prediction studies)?

For any normalization and dimensionality reduction done in the paper, was it ensured that it was done after the training/test split to avoid data leakage? Also, was a similar hyperparameter tuning scheme used to pick the best hyperparameters for the baseline models as well? What was the hyperparameter grid that was explored for them?

It is unclear why simple linear regression was chosen as a baseline when one could pick an elastic net with various ablations on the feature types as a baseline to better support the final formulation of the PolyEN model.

The scales in Figure 2C make it very hard to compare the boxplot performance. The cross-dataset results do not seem very good, raising questions about the model's generalizability.

It would be interesting to investigate important features obtained via the coefficients of the elastic net model. Do important genes overlap with the SHAP genes? I am also curious to learn how the paper is interpreting a SHAP importance score for a first-order versus second-order polynomial feature? Does it connect to the intuition behind using this formulation?

I could not read the text in Figure 3, which made it harder to understand the comparison of transcriptomics and methylation results.

Minor points:

A proof-reading pass of the paper (and the abstract) is highly recommended.

Higher-resolution figures with bigger font sizes are highly recommended.

CellBiAge was not included in the baseline comparison as it performed classification, but would the feature binarization step help the predictive performance on this dataset?

The main finding of the temporal analysis should be included in the main manuscript if it is mentioned. Also, its description in the Supplementary was difficult to follow.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Zhana Duren

Reviewer #2: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Dear Dr. Bar-Joseph,

Thank you very much for submitting your manuscript "Predicting Lung Aging using scRNA-Seq Data" for consideration at PLOS Computational Biology.

As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

Please note that there are still several concerns that need to be adequately addressed. 

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Sushmita Roy, Ph.D.

Section Editor

PLOS Computational Biology

Sushmita Roy

Section Editor

PLOS Computational Biology

***********************

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: Authors addressed all my comments and I recommend to accept this manuscript.

Reviewer #2: Thank you to the authors for their detailed responses. Most of my points have been addressed. (1) I appreciate clarity on the decisions behind using mean and variance across donors instead of direct readout (2) The details of the preprocessing and hyperparameter tuning are also clear.

However, some of the points need more clarification.

(1) To follow up on reviewer 1’s question: The methylation clocks work with bulk methylation data. Given, that the polyEN method’s performance gives no improvement when cell type information is combined, it would suggest that methylation clocks (bulk data) and polyEN (single-cell data) may not be directly comparable. A more direct comparison would be if all the single-cell gene expression was averaged across all cell types to create bulk information and see if transcriptome is still a better signal than methylation. Even if the donors were the same, the bulk and single-cell data would have very different distributions.

(2) I had a really hard time understanding the response for comparing PolyEN with the elastic net model. It is unclear why the decision to present the ranking in Figure S7 was made instead of just presenting the R2 and RMSE value comparisons between the models. Could something like Figure 2(c) be plotted instead for this comparison?

(3) The additional analysis of the SHAP values with coefficients is interesting. However, the observation that the top-ranked genes are different does not address the question completely. Was there a GO term analysis done for these genes like the one done for SHAP genes? Do these genes report similar GO terms or processes as SHAP genes? If they are different, what would that suggest?

(4) Also, is it possible that SHAP analysis is picking on highly expressed genes (looking at Figure 5 and 6)? If one were to rank the genes based on mean expression levels (or variance) and compare them to those ranked based on SHAP values, would one find significant overlap?

(5) Finally, despite my recommendation to do a proofreading pass and authors’ response that this was done, I found several issues with the writing. See examples below:

“led to best results for smokers” in the last line reads abrupt and vague in the abstract.

The author summary” has grammatical errors - “data profiling the gene expression”, “levels of genes in a individual”, “individual lung …. their ages”, “and that … as the best…”

Line 52: “accumulated errors at various leves” reads a bit vague (what type of errors?).

“P” capitalized in line 88 for polyEN but small in abstract.

Line 157 “utilizes the following.” ends abruptly.

Line 223 “Fleischer et al” needs a citation?

“Overview of the analysis and datasets” and “Cell type mapping and dataset integration” should probably be in the Methods section and not Results.

Line 300: “P” caplitalized for p-value but small everywhere else.

Line 297: “three previously published methods”.. is followed by listing only 2 - Fleischer et al and RAPTOR

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Zhana Duren

Reviewer #2: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Dear Dr. Bar-Joseph,

We are pleased to inform you that your manuscript 'Predicting Lung Aging using scRNA-Seq Data' has been provisionally accepted for publication in PLOS Computational Biology.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Computational Biology. 

Best regards,

Sushmita Roy, Ph.D.

Section Editor

PLOS Computational Biology

Sushmita Roy

Section Editor

PLOS Computational Biology

Feilim Mac Gabhann

Editor-in-Chief

PLOS Computational Biology

Jason Papin

Editor-in-Chief

PLOS Computational Biology

***********************************************************

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #2: Thank you for addressing all my concerns. I appreciate the new results and do not have any additional comments.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: None

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No