PONE-D-20-07163

Source Identification of Infectious Diseases in Networks via Label Ranking

PLOS ONE

Dear Dr. Huang,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Sep 13 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

M. Sohel Rahman, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data 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 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—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper “Source Identification of Infectious Diseases in Networks via Label Ranking” deals with estimating single source of infection in a network. The authors provide approaches for both the scenarios when we have complete and partial information of the nodes. For the former, they propose a label ranking based approach and for the latter, besides a direct approach, they propose a two-stage approach to estimate a source. The first stage deals with network status restoration using semi-supervised learning and the second stage in this two-stage framework is similar to the method which deals with complete information. The proposed methods seem to work well in estimating single source of infection on various synthetic and real-world datasets.

The paper is well written and organized. The main contribution of this paper is the idea of network status restoration (first stage in the snapshot observation case) and the provided techniques, i.e., estimating the actual state of each node in a snapshot graph where we have a partial observation, and restoring the initial graph status. However, I have a few comments which, upon getting properly addressed, I believe, the paper could be considered for publication.

Comment 1. About Lines 36 to 40, Lines 54 to 57, Lines 84-94 and Lines 213 to 215: There is a recent paper (Ali, S. S., et.al., EPA: Exoneration and Prominence based Age for Infection Source Identification. CIKM, Nov. 2019) which exploits and shows the importance of uninfected nodes in source identification. As, discussed in this manuscript, the paper (Ali, S.S., et al., CIKM, 2019) also talks about the prominence of a source, where a source is said to have more infected neighbor concentration than a fringe or the farthest away (from the center of an infection) node. I believe that should be added as a reference in the manuscript, besides the already included manuscript reference numbers [12] (Prakash, Aditya B., et al., ICDM, 2012) and [25] (Wang, Zheng, et al., AAAI, 2017).

Comment 2. About Lines 35 to 36: While this generally holds good, there is no mathematical proof for this. In fact, a few works show that graph centrality measures do a good job in source estimation in certain scenarios. Therefore, to inform the readers of such scenarios, besides what has already been discussed in this manuscript, I would suggest the authors add these two papers as references for the sake of unbiased dissemination of information:

1. Comin, C. H., & da Fontoura Costa, L. (2011). Identifying the starting point of a spreading process in complex networks. Physical Review E, 84(5), 056105.

2. Ali, S. S., Anwar, T., & Rizvi, S. A. M. (2020). A Revisit to the Infection Source Identification Problem under Classical Graph Centrality Measures. Online Social Networks and Media, 100061.

Comment 3. While, I don’t doubt the intentions of the authors, however, if I am not wrong, Algorithm 1 in the manuscript, i.e., BLRSI, seems starkly similar to Algorithm 1 (LPSI) of manuscript reference number [25] (Wang, Zheng, et al., AAAI, 2017). The only difference which I see is that LPSI estimates multiple sources by checking whether the label propagation score of a node with respect to its neighbors is the highest and considers it as one of the sources if that is the case. In the algorithm (BLRSI), presented in this manuscript, the highest label score amongst all the observed infected nodes is considered to be the source. I believe this has already been achieved in (Ali, S.S., et. al., A Revisit to the Infection Source Identification Problem under Classical Graph Centrality Measures. OSNEM, 2020), where, for the purpose of comparison, the authors have tuned the same algorithm (LPSI) to capture a single source in, I believe, exactly the same way as the BLRSI algorithm presented in the manuscript. Therefore, my question is what makes BLRSI algorithm different to those algorithms given BLRSI has been proposed/designed (observe the language in the Abstract Section and Line 102) in this manuscript? (Also, the analysis of BLRSI provided in Section “Algorithm analysis of Basic Label Ranking” is quite similar to the analysis of LPSI in Wang, Zheng, et al., AAAI, 2017.

Comment 4. For complete observation, while the comparison of BLRSI against NetSleuth/SSNS (manuscript ref. no. [12] (Prakash, Aditya B., et al., ICDM, 2012)) is fine, I suggest the comparison should be made against EPA as well (Ali, S.S., et al., CIKM, 2019), given EPA also exploits uninfected nodes to estimate the source of infection. Besides, EPA is a recent work which has been shown to beat NetSleuth/SSNS (Prakash, Aditya B., et al., ICDM, 2012) in single source infection identification scenario.

Comment 5. For snapshot observation, all the three proposed methods, VLRSI, TSSI-GFHF and TSSI-LGC should and can (if I am not wrong) at least be compared with one of the recent existing works. While the authors, from Lines 368-373, do argue that the existing works have different approaches in comparison to their own methods and thus the comparison is not suitable, however, in the reasons specified by them, the authors do not consider the methods (for comparison) which plainly consider the snapshot network to estimate the source from. The direct and simple way is to compare the proposed methods (especially VLRSI) with those methods which work on SIR model, for example with Reverse Infection (RI) method (Zhu et. al., Information source identification in SIR model: A sample-path-based approach, IEEE/ACM Trans. on Networking, 2016). In case, this still is not possible, the authors should explain the same.

Comment 6. I observe the average degree of the datasets used for experimentation is quite high and, hence, the diameter of these graphs would be on the smaller side and density on the higher side. Therefore, the AED might tend to appear on the good side of things. Lower average degree would mean, higher AED. This is also confirmed by the results (Table 2) achieved on Roget dataset with comparatively lower average degree tending to make AED higher. Typically, if the density of the graph is high, AED becomes small (which makes sense, since the diameter is small in the first place) and DR is not so good. This also explains why DR is so low even on ER graph (26%), when on ER graph source identification performance is generally better. I would suggest authors add density values and diameter values of each dataset in the dataset table (Table 1) and explain this relation (between graph diameter/density and AED/DR).

Comment 7. Generally, in how many iterations does BLRSI achieve convergence in the experimentations that authors have conducted? If there is some iteration number, the authors should add that in the manuscript to help future researchers work with an iterative process instead of a convergent one, wherever suitable.

Comment 8. It appears that the methods proposed in this work can estimate an infection source without having any information on the underlying model of infection propagation. Given the importance of this, I suggest the authors mention this clearly in the manuscript.

**********

6. 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: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment 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. Registration is free. 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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-20-07163R1

Source Identification of Infectious Diseases in Networks via Label Ranking

PLOS ONE

Dear Dr. Huang,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Dec 14 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

M. Sohel Rahman, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

Both reports are largely positive, albeit with minor identified issues. Please attend to the comments and resubmit as soon as possible. 

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data 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 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—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I believe all the comments have been addressed and this manuscript could be considered for acceptance. However, I need the authors to fully address the following points before the paper is published. The respected editor can personally ensure the authors incorporate these few, small changes in their accepted manuscript:

1. As the authors have admitted that BLRSI is indeed similar to LPSI (Wang, Zheng, et al., AAAI, 2017), they must explicitly state this in the manuscript especially when they start discussing BLRSI.

2. The authors have done a commendable job comparing their proposed methods with EPA (Ali, S. S., et.al., EPA: Exoneration and Prominence based Age for Infection Source Identification. CIKM, Nov. 2019). However, the language in lines 421 and 422 in the revised manuscript is not appropriate. While it is noted that EPA’s performance is not as good as the proposed methods on Roget network, the authors should understand that EPA has a good/better performance on French network (on DR), besides being better in Football and BA. Therefore, the authors should remove the generalization in these lines and talk about specific details clearly while noting that EPA outperforms the proposed methods in 3 out of 6 networks (either in DR or AED or both) and provides a stiff competition to the proposed methods in ER graph (on DR), as shown in Table 2.

3. The authors are advised to recheck the time complexity of UB, RC and SNSS (Table 3). As betweenness centrality of all the nodes in a graph involves calculating all the shortest paths between all the pairs of nodes, the complexity should be O(n^3) (see Floyd-Warshall algorithm). Similarly SNSS requires calculating eigenvalues from a matrix which generally has the time complexity of O(n^3). I believe RC on a general graph would have the complexity of O(n^3) as well. Authors need to check this again.

4. In ref. [26] in the revised manuscript, the name of one of the authors is missing. Authors are advised to correct it. Here is the full reference: Ali SS, Anwar T, Rastogi A, Rizvi SA. EPA: Exoneration and Prominence based Age for Infection Source Identification. InProceedings of the 28th ACM International Conference on Information and Knowledge Management 2019 Nov 3 (pp. 891-900).

5. In Supporting Information, "S1 Fig" is missing.

Reviewer #2: The paper “Source Identification of Infectious Diseases in Networks via Label Ranking” examines the coherence and efficiency of label ranking for exploring the possibility of a vertex or node to be the source of infection in a network (represented by graphs). The paper gives detailed comparisons with existing works to prove the superiority of their approach and at the same time, provides a fairly decent result in terms of the time complexity. The manuscript is technically sound, and the data supports the conclusions. Statistical analysis has been performed rigorously and the data has been made fully available. The manuscript is written in standard English and is presented in an intelligible fashion.

It is highly admirable that the paper’s revised version has introduced a number of new experiments as per the suggestions in the primary review. Moreover, what’s commendable is that the study acknowledges that the algorithm design and calculations are simplified due to the network’s abstractions which do not completely visualize all practicalities. The findings in Table 2 are rather impressive given the fact that BLRSI shows better results and satisfactory timings. Although SSNS shows more superiority in terms of timing, the authors have addressed this and shown that BLRSI outperforms SSNS for accuracy. The paper is enriched with lots of relevant comparisons which have made it easier to accredit. Furthermore, the authors have added extensive literature review that allows readers to comprehend the findings in a more convincing manner.

However, some minor issues can be addressed in the first revision. Firstly, the authors have provided Table 5 that analyzes the impact of the initial label vector y. It is seen that omitting the impact of either infected nodes or uninfected ones yields poor results. Although this particular observation is quite intuitive, the other observation where giving higher weights to uninfected nodes or retaining the original assignment like Eq.(1) produces superior results than giving higher weights to infected nodes could use some discussion. That is, why does punishing uninfected nodes give better results than crediting infected ones? This is an unexplored outcome of the experiment and can lead to interesting conversations or further analysis in future works. Basically, the authors could either provide some clarity on this end and explain the reasoning behind it, or perhaps provide a few more entries to the table through a small number of experiments showing that this is not an absolute outcome and will differ for datasets.

Secondly, there is a small issue with Algorithm 2. This algorithm is almost similar to Algorithm 1 with minimal differences. In both these algorithms, k is set to be the parameter that represents the number of iterations. However, although this parameter is incremented in Algorithm 1 (Line 11), Algorithm 2 does not show it. Although it is intuitive that k will be incremented for reaching convergence, having this operation in one algorithm and not having it in another shows inconsistency and can be avoided. The same applies to Algorithm 4.

Thirdly, since Algorithm 2 is almost identical to Algorithm 1 with minor differences (one of them being the definition of the initial label vector) it seems unnecessary to repeatedly show the same operations all over again. Instead, the common algorithm (particularly the iterations) could be set as a subroutine and Algorithms BLRSI and VLRSI could separately call the common subroutine with different parameters. This would also give more insight regarding where exactly the difference between Algorithm 1 and 2 lies. Although this does not seem absolutely necessary, it is something the authors can consider.

In conclusion, the paper provides some interesting observations and acknowledges the limitations. The impressive results in terms of accuracy and time complexity allow readers to comprehend the superiority of this approach.

**********

7. 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: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment 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. Registration is free. 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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Source Identification of Infectious Diseases in Networks via Label Ranking

PONE-D-20-07163R2

Dear Dr. Huang,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

M. Sohel Rahman, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data 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 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—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

**********

7. 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